Casting Hold Person [Archive] - Giant in the Playground Forums

BurgerBeast

2016-08-15, 01:08 PM

My ruling is that a spell can be cast to target something/someone that is not normally a valid target for that spell. Casting a spell on an invalid target has no effect on the target, and the spell is wasted; unless circumstances would indicate otherwise, the caster won't immediately know that the spell failed. Depending on the particular spell used, the particular target, and the particular reason it failed, I might decide to give the caster a chance (via check or save) to realize that the spell failed (or if it's obvious that the spell failed, a check or save to figure out why).

Cool. I'm going to assume that you assume that the act of waiting a spell on an invalid target also uses up the player's action for that turn. Feel free to correct me. How, then, would you handle this?

Scenario 2

Joe has cast Mirror Image.

Hero declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?

If playing on Grid, point to the target it wants to cast Hold Person on. Good luck.

If playing on ToM, describe which target it wants to cast Hold Person on. Good luck.

Alternatively, attempt to walk through the target before casting. That would be physical interaction.

Cool. So the caster must choose a target. How, then, would you handle this?

Scenario 2

Joe has cast Mirror Image.

Hero declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?

Yeah, well. Plague's stance interested me, and it's a question I didn't really look into in the other thread.

Yes, I owe Plaguescarred a big thank you for explaining himself so clearly. This process is exactly what I was trying to bring up in the other thread. It's the general process for casting a spell.

I guess I can give you a fresh start, though.

I assume Hero was not present when Joe cast the spell, and does not know which Joe is real. If that's the case, DM asks Hero which "Joe" he is targeting. If Hero comes up with an outside-the-box solution for this problem, I'd probably go with it - as Hero, I'd probably start my turn by yelling "Hey Joe!" and seeing which one reacted.

If he targets incorrectly, the spell does nothing and his action is wasted, though I'm undecided whether the spell slot is wasted and whether the player is directly given the information that he failed to pick a valid target.

Cool. Does this new perspective change anything about scenario 2?

Scenario 2

Joe has cast Mirror Image.

Hero declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?

Quoting RulesJD because he stated a rule that I base my answer on, not because my answer is directed at him.

I'd allow the spell to be cast on the illusion. I generally allow spells to be cast regardless of targeting requirements, unless it makes absolutely no sense. If an invalid target is selected for whatever reason, the action and spell are lost.

In this particular example, casting a spell at Hold Person on a Silent Image (or other illusionary target) I'd rule that it also counted as "using your action" to examine the illusion. Make an investigation check to discern the illusion for what it is.

Why?
1) Caster doesn't know if your spell fails because she selected an invalid target or a successful saving throw. She just knows it failed. And she only knows that because she doesn't have to concentrate on the spell and the target is clearly unaffected.
2) It's not a physical interaction, so it's not automatic.
3) Using actions to use Investigation to determine if things are illusions is relatively rare. Since I'm already being strict on the loss of spell slot & action side of things, I might as well be lenient by allowing the action to count as trying to Investigate the illusion at the same time.

Thanks for the explanation. How do you rule on scenario 2?

Scenario 2

Joe has cast Mirror Image.

Hero declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?

At this point, nobody has raised any of the objections I received in the other thread. So I think it's reasonable to present scenario 2 to everyone.

Scenario 2

Joe has cast Mirror Image.

Hero declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?

smcmike

2016-08-15, 01:24 PM

Joe has cast Mirror Image.

Hero declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?

You know my answer here: Joe rolls for his saving throw.

Why?

Because the Mirror Image text regarding attacks represents the only mechanical effect of the spell. This assertion is supported by the following:

1. Argument from authority: Crawford said so.

2. Argument from absence: if there were any other mechanical effects, I would expect them to be discussed in further detail. Specifically, targeted save spells would have been included in the attack mechanic if they were meant to be affected.

3. Argument from fluff: the spell creates untrackable shifting duplicates. Hero wouldn't be able to pick one out to target even if he tried. If Hero was using Firebolt, he would just say "I cast Firebolt at Joe."

4. Argument from fluff (2): Hold Person, and other targeted save spells (other than DEX saves) don't care about physical position very much. You can't dodge them, and partial cover doesn't help avoid them. The basic effect of Mirror Image is to make your physical location slightly unclear, which is not enough to foil targeted spells, just as Displacer Beast's displacement does not foil targeted spells.

5. Extension of argument from fluff (my own fluff): I think Mirror Images work like mirror images, and that a targeted non-dex spell could be cast based upon viewing the target within a mirror, so long as there is line of effect.

Tanarii

2016-08-15, 01:31 PM

At this point, nobody has raised any of the objections I received in the other thread. So I think it's reasonable to present scenario 2 to everyone.

Scenario 2

Joe has cast Mirror Image.

Hero declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?
Spell targets Joe, unless the Hold Person spell caster decides otherwise. Because Hold Person is not an attack, Mirror Image has no affect on target selection.

Why is it different from the Silent Image scenario? Because that's the mechanical rule built into Mirror Image. If they wanted it to work for all target selection, attacks and non-attacks, they should have said so.

AvatarVecna

2016-08-15, 01:49 PM

Cool. I'm going to assume that you assume that the act of waiting a spell on an invalid target also uses up the player's action for that turn. Feel free to correct me. How, then, would you handle this?

The target's validity is irrelevant in regards to the action; Hold Person takes an action to cast. If you're targeting a valid target, the spell takes effect; if you're not targeting a valid target, it doesn't.

Scenario 2

Joe has cast Mirror Image.

Hero declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?

Joe makes a saving throw against Hold Person, assuming that Joe himself is a valid target for Hold Person, because Mirror Image only affects attack rolls. Personally, I tend to think of single-target spells that don't require attacks rolls as being more similar to AoEs than attacks, just with a tiny area.

Long story short, the reason why Silent Image can cause Hold Person to be wasted, but Mirror Image can't, is because the Silent Image of Joe isn't where Joe is, so it can't target Joe, but the Mirror Image is where Joe is, so Joe gets targeted for being the targeted Joe in the targeted square. Note: this means that if, in the earlier example, the Silent Image of Joe was in the same place as the real Joe, you would only see one Joe (since the illusion would be fully covering him), and the real Joe could be targeted by targeting Hold Person at the image of Joe (although that leads into a debate about line of sight and illusory cover).

smcmike

2016-08-15, 02:01 PM

Tanarii

2016-08-15, 02:05 PM

Note: this means that if, in the earlier example, the Silent Image of Joe was in the same place as the real Joe, you would only see one Joe (since the illusion would be fully covering him), and the real Joe could be targeted by targeting Hold Person at the image of Joe (although that leads into a debate about line of sight and illusory cover).
What if the Silent Image that is completely covering Joe looks like Nancy?Can't happen. Physical interaction would cause the Silent Image to be automatically revealed.

(Edit: Nested quotes to show what the scenario is that can't happen.)

What if you know exactly where Joe is, but he's invisible?You can't target Joe, because Hold Person requires a target you can see. Spell fizzles automatically (by my ruling), or just can't be cast at all (by the "valid target required" ruling)

AvatarVecna

2016-08-15, 02:08 PM

What if the Silent Image that is completely covering Joe looks like Nancy?

What if you know exactly where Joe is, but he's invisible?

I think the sight precondition had to be taken seriously. It's not based on position.

1) I'd probably have it affect Joe still, for a lot of the same reason I wouldn't allow Disguise Self to make somebody immune to Hold Person just by disguising themselves as somebody else.

2) Being invisible does not make you untargetable, even by things that actually have attack rolls. This does not have an attack roll, so since you know what space he's in, it'll affect him. Of course, if he's invisible and successfully hiding, I'd say you have to guess which square to target, same as if you were trying to hit him with an attack roll...and if you aim it at the wrong square, it's wasted.

3) You are welcome to take it as seriously as you want. This is my ruling, because this is how I'd rule it in my game. If you feel differently, good for you on having an opinion, but it's not going to change what I think is reasonable for my game.

smcmike

2016-08-15, 02:12 PM

2) Being invisible does not make you untargetable, even by things that actually have attack rolls. This does not have an attack roll, so since you know what space he's in, it'll affect him. Of course, if he's invisible and successfully hiding, I'd say you have to guess which square to target, same as if you were trying to hit him with an attack roll...and if you aim it at the wrong square, it's wasted.

3) You are welcome to take it as seriously as you want. This is my ruling, because this is how I'd rule it in my game. If you feel differently, good for you on having an opinion, but it's not going to change what I think is reasonable for my game.

I'm not criticizing your game at all. I don't think it hurts anything to play it the way you suggest, and would be happy to play with that rule. I'm just pointing out that, per RAW, you cannot target an invisible person with Hold Person, which requires you to be able to see the target.

AvatarVecna

2016-08-15, 02:18 PM

I'm not criticizing your game at all. I don't think it hurts anything to play it the way you suggest, and would be happy to play with that rule. I'm just pointing out that, per RAW, you cannot target an invisible person with Hold Person, which requires you to be able to see the target.

It's not RAW-legal, I'll grant you. My home group just figured out pretty early on that rules regarding messing with perceptions (whether it's stealth or illusions) seem to be rather vague too often for our tastes, so we make rulings based on situations rather than figuring out what the absolute rule should be. It's why I stay out of Stealth Rules threads, and usually stay out of Illusion Rules threads: attempting to figure out a way to make the RAW make sense for every edge case in existence is a lot of time and effort being spent on something that's ultimately handled better by on-the-fly rulings made to deal with the specific weird situation that might come up in your game, rather than figuring out every possible weird combo beforehand and figuring out your ruling/RAW justification for it.

Theodoxus

2016-08-15, 02:20 PM

My ruling comes from interactions in other games (let's take World of Warcraft, in this instance).

When you attempt to cast a spell in WoW against an invalid target (healing an enemy combatant, casting a damaging spell at a friendly guard), the game notifies you that it's an invalid target (without allowing you to cast/waste resources), and you've essentially wasted the action (it's not quite the same, as everything happens in real time, so the second or two it takes to register you're casting at an invalid target isn't quite as detrimental as wasting an entire action).

However, since this is specifically interacting with an illusion, and it requires an action to interact with it - I would be perfectly happy to have that include an attempt to cast a spell at the target illusion.

So, in effect, the player designates a target, then attempts to cast the spell. I would notify the player that they have detected an illusion, and that the specific spell (Hold Person, in this instance) has no effect and hasn't spent resources (no slot loss), but that their action isn't wasted, as they now see through the illusion.

MaxWilson

2016-08-15, 02:26 PM

Alternatively I'd welcome the reverse if anyone is willing.

Scenario:

A wizard (hereafter referred to as Joe) has cast Silent Image and created an image of himself. He and the Silent Image are 10 feet apart.

A second wizard (hereafter referred to as Hero) declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?

Since you're polling for individual rulings:

Ruling: Hero attempts to cast Hold Person, but the magic fails to take effect. He feels a curious emptiness around the creature, and there is nothing for his magic to anchor itself to. He ends his action without expending any spell points but it will take a fresh action to re-cast the spell on a new target.

Reasoning: This ruling maintains a balance between the usefulness of illusions and player fun (not wasting important resources like spell slots on nothing). It's also easy to adjudicate and it fits easily into the logic of how magic works in my world.

Segev

2016-08-15, 02:31 PM

Overall, I'd say this is really in the "GM call" territory. It's one of the areas I think 5e could have been a little more explicit on, at least giving advice/guidelines in the DMG, but the question boils down to what you're "really doing" when you cast a spell.

Are you performing a closed loop that can't have its beginning without its ending, such as trying to close a circuit which, if a load is missing from its appointed place, will not complete and thus nothing happens? Or are you firing a metaphorical gun, and whether your target is there or not doesn't matter to the fact that you did, in fact, pull the trigger?

I personally go with the latter. You perform the gestures, you pour out the magical energies, but you happened to do it to something that isn't there. Woops. Nothing happens. Just like you don't get a fueled car if you pour gasoline into an illusory gas tank while the real car is ten feet away. You get gasoline all over the ground (and quite the fire hazard).

But a DM can rule the other way, too. 5e is not explicit on this.

1) I'd probably have it affect Joe still, for a lot of the same reason I wouldn't allow Disguise Self to make somebody immune to Hold Person just by disguising themselves as somebody else.Eh, different circumstances. Spells aren't name-locked, unless they say so. "I'm targeting that guy, who I think is Fred," is how I would read a player saying he's targeting a particular creature that I know is actually Frieda disguised as her brother, Fred. The creature is there. She's a valid target for the spell. The spell affects her, because she's the creature at which you aimed it.

The question here is whether aiming the spell at a "creature" that isn't really a creature does anything. If you cast hold monster at a statue because you think it's a gargoyle (you paranoid adventurer, you), does that instantly reveal it's not a creature?

I think it weakens illusions and strengthens spells into free detection abilities to treat targeting as a necessary condition to even attempt the action. "Sorry, you can't do that" tells the savvy player "ah, that's an illusion; I'll hit the other one, instead."

Foxhound438

2016-08-15, 02:34 PM

Scenario 2

Joe has cast Mirror Image.

Hero declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?

Mirror image specifically says in its own description that it only works on attacks. phb 194, "if there's ever any question weather something counts as an attack, the rule is simple: if you're making an attack roll, you're making an attack. As such, mirror image does nothing against hold person, or any other save-based spell for that matter.

If you're having trouble picturing in your head how that plays out, note in this case the school of hold person: Enchantment. There's no physical interaction with the creature, the spell tricks the target's mind into being unable to move.

RickAllison

2016-08-15, 02:38 PM

Overall, I'd say this is really in the "GM call" territory. It's one of the areas I think 5e could have been a little more explicit on, at least giving advice/guidelines in the DMG, but the question boils down to what you're "really doing" when you cast a spell.

Are you performing a closed loop that can't have its beginning without its ending, such as trying to close a circuit which, if a load is missing from its appointed place, will not complete and thus nothing happens? Or are you firing a metaphorical gun, and whether your target is there or not doesn't matter to the fact that you did, in fact, pull the trigger?

I personally go with the latter. You perform the gestures, you pour out the magical energies, but you happened to do it to something that isn't there. Woops. Nothing happens. Just like you don't get a fueled car if you pour gasoline into an illusory gas tank while the real car is ten feet away. You get gasoline all over the ground (and quite the fire hazard).

But a DM can rule the other way, too. 5e is not explicit on this.
Eh, different circumstances. Spells aren't name-locked, unless they say so. "I'm targeting that guy, who I think is Fred," is how I would read a player saying he's targeting a particular creature that I know is actually Frieda disguised as her brother, Fred. The creature is there. She's a valid target for the spell. The spell affects her, because she's the creature at which you aimed it.

The question here is whether aiming the spell at a "creature" that isn't really a creature does anything. If you cast hold monster at a statue because you think it's a gargoyle (you paranoid adventurer, you), does that instantly reveal it's not a creature?

I think it weakens illusions and strengthens spells into free detection abilities to treat targeting as a necessary condition to even attempt the action. "Sorry, you can't do that" tells the savvy player "ah, that's an illusion; I'll hit the other one, instead."

Well unless the DM tells them it's an illusion, all they know is it's an illegal target. Maybe Joe isn't actually a humanoid and instead is just a vampire in makeup? Maybe he is in an anti-magic sphere? Maybe he is immune to spells of that level? All that a player definitely knows is that the target is not a valid one.

AvatarVecna

2016-08-15, 02:39 PM

Eh, different circumstances. Spells aren't name-locked, unless they say so. "I'm targeting that guy, who I think is Fred," is how I would read a player saying he's targeting a particular creature that I know is actually Frieda disguised as her brother, Fred. The creature is there. She's a valid target for the spell. The spell affects her, because she's the creature at which you aimed it.

That's more or less what I was saying. I responded to "Would having a Silent Image of Nancy around Joe make Joe have to save?" with "Yes, because Joe's there. This is why I would have Hold Person affect somebody under Disguise Self, since they're still there even if they're not the person you think they are". It's possible my phrasing was weird, but we appear to feel similarly about the effect of casting Hold Person on somebody employing Disguise Self.

Plaguescarred

2016-08-15, 02:40 PM

Counterspell interrupts the spell while being cast. By that reasoning, the slot wouldn't be used up because the spell casting wasn't actually completed.Counterspell interrupt a spell that is being cast. If the spell countered did not followed the basic spellcasting rules to cast a spell, it wouldn't have been cast in the first place and thus never countered.

Plaguescarred

2016-08-15, 02:42 PM

Wether the spell countered by Counterspell waste a slot or not is another question though. The spell doesn't say so and i don't know if it's been clarified by a D&D staff.

Tanarii

2016-08-15, 02:47 PM

It's possible my phrasing was weird, but we appear to feel similarly about the effect of casting Hold Person on somebody employing Disguise Self.
Except Silent Image can't overlay a person. It'll automatically be revealed as an illusion due to physical interaction. But yes, in the case of disguise self, you're targeting what you believe to be a creature, and you can see it. And sure enough, it is a creature you can see.

But unlike Silent Image, the Disguise Self and Mirror Image illusions are not independent spell constructs from a creature. So they aren't actually comparable to silent image , except to show how the situation is different for Silent Image, where you actually can actually try and target the illusion separately and independently in place of a creature, thinking that is is a creature, and be wrong about it.

AvatarVecna

2016-08-15, 03:03 PM

Except Silent Image can't overlay a person.

*shrug* That's probably RAW-compliant, but my reading (and one my group agrees with) is that the physical interaction only reveals the illusion to the creature performing the physical interaction; witnessing somebody else physically interact with the illusion either warrants a free action, out-of-turn Perception/Investigation check, or is considered what the "take an action to make an Investigation check to see the illusion for what it is" part of the spell. I can certainly see somebody saying that anybody interacting with the Silent Image reveals it to everybody in line of sight, but we allow an overlay.

Regardless, I'm saying a Silent Image overlaid over the caster doesn't protect them from Hold Person because they're still in the same square, you seem to be saying that a Silent Image overlaid over the caster doesn't protect them from Hold Person because a Silent Image overlaid over the caster auto-fails because it's being physically interacted with by the caster sharing the square, so we appear to agree on the result, just not on the why.

Incidentally, my group's rulings should probably be taken with at least a grain of salt; our group has an Illusionist Wizard, so it's possible (even probable) that we're ruling them to be more powerful than they should be.

smcmike

2016-08-15, 03:17 PM

Regardless, I'm saying a Silent Image overlaid over the caster doesn't protect them from Hold Person because they're still in the same square, you seem to be saying that a Silent Image overlaid over the caster doesn't protect them from Hold Person because a Silent Image overlaid over the caster auto-fails because it's being physically interacted with by the caster sharing the square, so we appear to agree on the result, just not on the why.

Well, it's not quite the same result, and created real differences with Invisibility.

Following your physical interaction rule, what if "Joe" is actually "Joe and Steve," the gnomish illusionist brothers, and they are hiding in their creation, "Nancy," the Silent Image of a rather large lady?

Plaguescarred

2016-08-15, 03:21 PM

How ii understand it, decerning a Silent Image to be an illusion, either by physically interacting with it or by examining it by making a successful Intelligence (Investigation) check as an action, let you see through the image.

Seeing your friend's sword pass trhough the image of an Orc should be good enought to dicern that the orc is an illusion, which should in turn let you also see through it as your brain slowly realise that the silent image seen is in fact not real.

AvatarVecna

2016-08-15, 03:23 PM

Well, it's not quite the same result, and created real differences with Invisibility.

Following your physical interaction rule, what if "Joe" is actually "Joe and Steve," the gnomish illusionist brothers, and they are hiding in their creation, "Nancy," the Silent Image of a rather large lady?

Fair enough.

Depends on what ruling would make my players bitch the least. It's a toss-up between "roll randomly to see which one gets targeted" and "target which one actually cast the spell they're currently hiding in".

IShouldntBehere

2016-08-15, 03:26 PM

Before I answer that question, I guess I kind of need the answer to another question:

There is a popular street act feature Bogo, The Drum Playing Monkey. In Bogo's act he furiously plays a Drumset (very heavy on cymbals) while screaming at the top his lungs. He also shimmies and shakes as the myriad bells on vest and hat kick up a racket. He's quite popular with the kids and is as famous as one can be given the communication methods of the time.

Bill who has heard of Bogo's act and kind of walked by it once but never watched it intently walks up to a silent image of Bogo doing his whole shebang. The illusion bags the drums, has it's mouth open side in a scream and the little bells shake too and fro.

http://viralgifs.com/wp-content/uploads/2014/03/chimp_drummer.gif?fb43fe

Does bill have to interact with the illusion physically or use his action for an investigation check to determine that this ain't real?

Shining Wrath

2016-08-15, 03:33 PM

If Hero casts a Sleep spell covering both the SI and Joe, Joe falls asleep and the SI vanishes when Joe loses concentration (assuming a high enough roll for HP affected by Sleep etc). SI has zero HP, after all.

If Hero casts Fireball and blasts both Joe and SI, Joe gets hurt and the SI does not. Joe might be able to change the SI to show damage, but not as a reaction to the fireball, so the interval between BOOM and the start of Joe's turn would reveal the truth to observers.

If Hero casts Firebolt and targets SI, it passes right through SI and reveals it as an illusion.

It seems that most spells affect Joe differently than SI.

Therefore, I'd DM this as a single-target spell targeting SI reveals it as an illusion, including Hold Person. If Joe puts the SI in the same space as he so that the two cannot be distinguished then Hold Person would have a 50-50 chance of hitting Joe, but Firebolt would hit both at the same time.

Segev

2016-08-15, 03:34 PM

Before I answer that question, I guess I kind of need the answer to another question:

There is a poopular street act feature Bogo, The Drum Playing Monkey. In Bogo's act he furiously plays a Drumset (very heavy on cymbals) while screaming at the top his lungs. He also shimmies and shakes as the myriad bells on vest and hat kick up a racket. He's quite popular with the kids and is as famous as one can be given the communication methods of the time.

Bill who has heard of Bogo's act and kind of walked by it once but never watched it intently walks up to a silent image of Bogo doing his whole shebang. The illusion bags the drums, has it's mouth open side in a scream and the little bells shake too and fro.

http://viralgifs.com/wp-content/uploads/2014/03/chimp_drummer.gif?fb43fe

Does bill have to interact with the illusion physically or use his action for an investigation check to determine that this ain't real?
He is free to assume anything he sees is "not real." However, he doesn't get the benefit of seeing through it the way somebody who KNOWS it's an illusion does unless he makes the Investigation check or physically interacts.

Otherwise, how does he know whether this is a silent image or is actually Bogo under the effects of a silence spell cast by a Bard who was tired of the little guy disrupting his performance?

He doesn't, without taking an Investigate action or attempting physical interaction.

This freedom to assume anything you like is "not real" up until interaction with it proves you wrong is why illusions can be pretty useless if GMs won't play along. GMs know when the PC's illusions are, well, illusions, and could just have his NPCs ignore them if he wanted. Whether he's justified IC or not is really a matter of his judgment, with only his respect for his players' enjoyment and fairness to them to guide him. Well, that, and an honest effort to figure out how convincing a given illusion might be, given its limitations.

smcmike

2016-08-15, 03:34 PM

http://viralgifs.com/wp-content/uploads/2014/03/chimp_drummer.gif?fb43fe

Does bill have to interact with the illusion physically or use his action for an investigation check to determine that this ain't real?

Fun!

Bill has to either use an action for investigation or interact physically to determine that this is certainly an illusion. He may, however, theorize that it is an illusion without any check. The best alternative theory is, of course, that the local bard wanted a bit of peace and quiet and Silenced Bogo.

IShouldntBehere

2016-08-15, 03:39 PM

Fun!

Bill has to either use an action for investigation or interact physically to determine that this is certainly an illusion. He may, however, theorize that it is an illusion without any check. The best alternative theory is, of course, that the local wizard wanted a bit of peace and quiet and Silenced Bogo.

So, now let us say both the Illusionary Bogo & The Real Bogo are side by side similar to our two men in the OP.

Suppose bill is having a bad day and wishes to take out Bogo and thus declares I declares "I cast Disintegrate on Bogo". What would the proper/your response be as DM?

smcmike

2016-08-15, 03:44 PM

So, now let us say both the Illusionary Bogo & The Real Bogo are side by side similar to our two men in the OP.

Suppose bill is having a bad day and wishes to take out Bogo and thus declares I declares "I cast Disintegrate on Bogo". What would the proper/your response be as DM?

Which one, the silent one or the noisy one?

IShouldntBehere

2016-08-15, 03:59 PM

Which one, the silent one or the noisy one?

Interesting.

So now to go back to original example. The man and the silent image of a man standing next to each other, can you hear the difference between them? After all the real man will actually be breathing and producing normal being-alive noises, which the invisibility threads around here have taught me by RAW humans can pinpoint with great precision unless the real man is doing something to make a stealth check. :smallwink:

However, leaving aside the sillier bits of RAW for a moment this cuts at an issue, where does the line lay? I personally think it is fair to say that in the OP's example one cannot tell. Therefore choosing the target randomly or as a no-information choice is fair play. That said I'm not sure a test the scenario really provides a good model for what the OP & the discussion around it are trying to get at:

When is a illusion so obvious as to be useless even without passing a save against the magic? I think we all know there is a line and that drawing it isn't unfair. My Bogo example was only to have a discussion to highlight that point for non-participating observers of the conversation.

I think that generally outside the OPs specific scenario, a Silent Image is probably a very poor stand in for a person or other animate object in most situations. It works great for a wall, or a vase, or some other thing would be silent. So while it works in the very narrow scenario defined in the OP I don't think we can use that to draw any particularly meaningful insights for real play.

I generally wouldn't expect anyone to mistake a Silent Illusion for a real creature at anything but a great distance, and I wouldn't expect it would take touching it or investigating for this to be true.

Theodoxus

2016-08-15, 04:47 PM

I generally wouldn't expect anyone to mistake a Silent Illusion for a real creature at anything but a great distance, and I wouldn't expect it would take touching it or investigating for this to be true.

And this is where our table would shout "Get your physics/science out of our game." We play a very 'it's magic, that's why' game. When logic makes game mechanics fail, especially those of a magical nature, like illusions, we kick logic in the nuts.

The rule might not make sense on the surface, but that's true for a lot of mechanical necessities. But illusions are very clear - it requires an action to determine if they are real or not. That's why I'm fine with the spell fizzling, but the action being spent to determine that the person is actually illusory.

In your Bogo example, it would still require someone to look at the vocal monkey to the silent monkey and really perceive why one is silent... an action taken to determine reality.

IShouldntBehere

2016-08-15, 05:44 PM

And this is where our table would shout "Get your physics/science out of our game." We play a very 'it's magic, that's why' game. When logic makes game mechanics fail, especially those of a magical nature, like illusions, we kick logic in the nuts.

The rule might not make sense on the surface, but that's true for a lot of mechanical necessities. But illusions are very clear - it requires an action to determine if they are real or not. That's why I'm fine with the spell fizzling, but the action being spent to determine that the person is actually illusory.

In your Bogo example, it would still require someone to look at the vocal monkey to the silent monkey and really perceive why one is silent... an action taken to determine reality.

...and it's where at my table we'd happy to take the world in a way that makes sense to us as we process it. Though I don't think we'd really frame that as "kicking" anything "in the nuts" but I suppose to each their own in the way the like to view the game world and in how they like to address the viewpoints of others that don't share their tastes.

Aeson

2016-08-15, 06:22 PM

Suppose bill is having a bad day and wishes to take out Bogo and thus declares I declares "I cast Disintegrate on Bogo". What would the proper/your response be as DM?
Unless Bill has already determined which Bogo is the real Bogo or unless determining which Bogo is real is trivially easy, he has to choose one of the two Bogos that he sees. Bill can attempt to determine which Bogo is the real Bogo however he chooses or he can just pick one as the target, but saying "I target Bogo" when he sees two things which could be Bogo and does not know which one is real will not cut it.

Which one, the silent one or the noisy one?
If the real and illusory Bogos are close together and Bill is not so close to them as to make it trivially easy to determine which one is producing the noise, that may not help so much as you think it would. Bogo can still be irritating if he's on the other side of the square, and maybe Bill can hit either Bogo with a Disintegrate from his current position, but if Bill is much further from either Bogo than the Bogos are from one another, determining which one is the noisy one and which one is the silent one is going to take some investigation on Bill's part.

Now, if Bill's standing right next to the two Bogos, that's another story, and I'd probably let it go as a scenario where it's trivially easy to distinguish between the two Bogos.

Regarding the discussion about whether or not you lose the spell when you cast it at an invalid target, how about we take a look at the costs and benefits of performing the actions in a (potential) combat scenario?

Assume that I am a solitary spellcaster and have a spell which cannot affect illusions, and further assume that if I attempt to cast a spell on an illusion of which I am not aware I will perceive the illusion and be refunded the spell. Say that I come upon a room which appears to contain a single hostile monster which can be neutralized by my spell and is worth expending the spell upon; for simplicity's sake, assume that my spell cannot fail to affect the target and I cannot fail to pass a check to perceive illusions should I take an action to attempt to do so. Say that I take an action to attempt to perceive whether or not the monster is real. If the monster is real, then I need to take another action to neutralize it with my spell. If the monster is illusory, the illusion is revealed and the encounter is over. If I instead attempt to cast my spell on the monster and the monster is real, my spell takes effect normally and I expend a spell and an action to neutralize the monster. If I attempt to cast my spell on the monster and the monster is illusory, I spend an action and reveal the illusion but do not lose the spell. To summarize:
- If the monster is real: It takes two actions and a spell to neutralize if I open with a perception check, or one action and one spell if I open by attempting to cast the spell.
- If the monster is false: It takes one action and no other resources to end the encounter regardless of how I choose to proceed.

Attempting to cast the spell is therefore strictly better if the spell is refunded whenever I choose a target which, unbeknownst to me, is invalid due to being illusory. The worst outcome for attempting to cast the spell is the same as the best outcome for attempting to perceive the illusion. Things become worse for attempting to perceive the illusion if there are multiple real monsters in play or if I can fail the check, and marginally better if I can perceive the existence of multiple illusions off of a single check and multiple illusions are present. I would therefore say that attempting to cast the spell needs to have a higher opportunity cost than attempting to perceive the illusion; casting the spell is in general the better option, unless the potential target is not worth the spell in the first place. As a result, I would rule that if you attempt to cast a spell on an invalid target, you will lose the spell.

Except Silent Image can't overlay a person. It'll automatically be revealed as an illusion due to physical interaction.
Say you walk into a room and see what appears to be a real, corporeal man and a real, corporeal chair, and you see the man walk through the chair. Assuming that things which appear to be real and corporeal must be either illusory or both real and corporeal, there are three possible situations - the man is real and the chair is illusory, the man is illusory and the chair is real, or both the man and the chair are illusory - but you do not have sufficient information to determine which of these cases is correct. Seeing the man walk through the chair reveals that something in the room is not real (which could give advantage on the check to perceive illusions), but until something that you know to be real attempts to interact with the illusion, merely seeing impossible interactions should not reveal the illusion.

I might allow something which your character strongly believes to be real to reveal an illusion as though it were real, if you can convince me that the character's belief in the reality of the one is significantly stronger than the character's belief in the reality of the other, but that's getting dangerously close to allowing illusions the power to warp a character's perception of reality so strongly that it could be argued that a character who believes strongly enough in an illusion will perceive real things to be illusory if they interact in impossible or implausible ways with the illusion, and that probably breaks too many things to be worth it except maybe as something that happens to NPCs.

Tanarii

2016-08-15, 06:28 PM

And this is where our table would shout "Get your physics/science out of our game." We play a very 'it's magic, that's why' game. When logic makes game mechanics fail, especially those of a magical nature, like illusions, we kick logic in the nuts.Exactly. It's a set of abstract rules providing a mechanical game resolution for a game-play situation. Not a set of rules for simulating another reality.

The rule might not make sense on the surface, but that's true for a lot of mechanical necessities. But illusions are very clear - it requires an action to determine if they are real or not. That's why I'm fine with the spell fizzling, but the action being spent to determine that the person is actually illusory.IMO that's the way to rule it. For any given definition of 'fizzle' and 'determine that the person is illusionary'. In other words: The spell may cost a slot or it might not, in the process of fizzling; It could be automatic determination, or require the Investigation check to succeed. Either seems like a good ruling, within possible interpretations of the rules. But either way, I think the action should be lost.

But allowing a determination to be made without a check or physical interaction that a creature is an illusion is definitely house-ruling illusions. Someone might suspect from the lack of sound (in the case of Silent Image) that something is an illusion, but they can't determine it for sure.

smcmike

2016-08-15, 08:02 PM

The rule might not make sense on the surface, but that's true for a lot of mechanical necessities. But illusions are very clear - it requires an action to determine if they are real or not. That's why I'm fine with the spell fizzling, but the action being spent to determine that the person is actually illusory.

In your Bogo example, it would still require someone to look at the vocal monkey to the silent monkey and really perceive why one is silent... an action taken to determine reality.

That depends on the situation. In an otherwise quiet space in which you are close enough to easily locate sound, there's no reason that your character should have difficulty figuring out which figure the noise is emanating from.

This is not the same as using an Investigation check to determine the reality of an image. It's simply the obvious sensory information provided to your character at first glance. Regardless of how dumb my character is or how poorly he rolls on his Investigation check, he will never hear sound coming from a Silent Image.

Also, the consequences are quite different from a check: knowing where the sound is coming from does not tell me which is real or fake (if any), and it certainly doesn't turn anything transparent.

Personally, I'm for stretching the bounds of non-action chicanery to solve puzzles like this. Crack a joke and see if one of them laughs: not an action.

As a result, I would rule that if you attempt to cast a spell on an invalid target, you will lose the spell.

Well said. I'm convinced.

Say you walk into a room and see what appears to be a real, corporeal man and a real, corporeal chair, and you see the man walk through the chair. Assuming that things which appear to be real and corporeal must be either illusory or both real and corporeal, there are three possible situations - the man is real and the chair is illusory, the man is illusory and the chair is real, or both the man and the chair are illusory - but you do not have sufficient information to determine which of these cases is correct. Seeing the man walk through the chair reveals that something in the room is not real (which could give advantage on the check to perceive illusions), but until something that you know to be real attempts to interact with the illusion, merely seeing impossible interactions should not reveal the illusion.

Here I think I disagree. The rule says physical interaction reveals the illusion. I think this means that when the interaction occurs, even between two objects of uncertain reality, it becomes immediately apparent which is real and which is fake. When illusion runs into reality, reality wins. Its like an optical illusion where, once you figure out what is going on, you suddenly can no longer fool your brain into seeing it the other way.

Exactly. It's a set of abstract rules providing a mechanical game resolution for a game-play situation. Not a set of rules for simulating another reality.

But allowing a determination to be made without a check or physical interaction that a creature is an illusion is definitely house-ruling illusions. Someone might suspect from the lack of sound (in the case of Silent Image) that something is an illusion, but they can't determine it for sure.

Right - but that suspicion can be all you really need. "I'll attack the noisy one."

Xetheral

2016-08-15, 09:44 PM

BurgerBeast

2016-08-15, 09:48 PM

You know my answer here: Joe rolls for his saving throw.

Why?

Because the Mirror Image text regarding attacks represents the only mechanical effect of the spell. This assertion is supported by the following:

1. Argument from authority: Crawford said so.

2. Argument from absence: if there were any other mechanical effects, I would expect them to be discussed in further detail. Specifically, targeted save spells would have been included in the attack mechanic if they were meant to be affected.

3. Argument from fluff: the spell creates untrackable shifting duplicates. Hero wouldn't be able to pick one out to target even if he tried. If Hero was using Firebolt, he would just say "I cast Firebolt at Joe."

4. Argument from fluff (2): Hold Person, and other targeted save spells (other than DEX saves) don't care about physical position very much. You can't dodge them, and partial cover doesn't help avoid them. The basic effect of Mirror Image is to make your physical location slightly unclear, which is not enough to foil targeted spells, just as Displacer Beast's displacement does not foil targeted spells.

5. Extension of argument from fluff (my own fluff): I think Mirror Images work like mirror images, and that a targeted non-dex spell could be cast based upon viewing the target within a mirror, so long as there is line of effect.

For me all of these arguments fail.

Spell targets Joe, unless the Hold Person spell caster decides otherwise. Because Hold Person is not an attack, Mirror Image has no affect on target selection.

I don't think this holds. Mirror Image has no effect on target selection, I agree. But when you cast Hold Person and there are four potential targets, Hold Person doesn't pick one of them by itself.

Why is it different from the Silent Image scenario? Because that's the mechanical rule built into Mirror Image. If they wanted it to work for all target selection, attacks and non-attacks, they should have said so.

This logic doesn't hold up to scrutiny. If they wanted it to work in the same way, they should have said so. If they wanted it to work in a different, specific way, they should have said so. But if they wanted it to work under the same conditions it always works, they wouldn't need to say anything (I am arguing RAW here).

The target's validity is irrelevant in regards to the action; Hold Person takes an action to cast. If you're targeting a valid target, the spell takes effect; if you're not targeting a valid target, it doesn't.

Yep.

Joe makes a saving throw against Hold Person, assuming that Joe himself is a valid target for Hold Person, because Mirror Image only affects attack rolls.

To this I would say: Nope. Mirror Image only imposes a targeting mechanic on attack rolls. It doesn't impose anything on save spells, but the caster of a save spell is still looking at four potential targets and has to choose one.

Personally, I tend to think of single-target spells that don't require attacks rolls as being more similar to AoEs than attacks, just with a tiny area.

Except that they can't target areas. They usually target "creatures" or the like.

Long story short, the reason why Silent Image can cause Hold Person to be wasted, but Mirror Image can't, is because the Silent Image of Joe isn't where Joe is, so it can't target Joe, but the Mirror Image is where Joe is, so Joe gets targeted for being the targeted Joe in the targeted square.

So what if Joe shared his square with a silent image?

And what if Joe shared his square with another humanoid?

Note: this means that if, in the earlier example, the Silent Image of Joe was in the same place as the real Joe, you would only see one Joe (since the illusion would be fully covering him), and the real Joe could be targeted by targeting Hold Person at the image of Joe (although that leads into a debate about line of sight and illusory cover).

Well I don't think this is RAW. I think you've said as much already, though.

My ruling comes from interactions in other games (let's take World of Warcraft, in this instance).

When you attempt to cast a spell in WoW against an invalid target (healing an enemy combatant, casting a damaging spell at a friendly guard), the game notifies you that it's an invalid target (without allowing you to cast/waste resources), and you've essentially wasted the action (it's not quite the same, as everything happens in real time, so the second or two it takes to register you're casting at an invalid target isn't quite as detrimental as wasting an entire action).

However, since this is specifically interacting with an illusion, and it requires an action to interact with it - I would be perfectly happy to have that include an attempt to cast a spell at the target illusion.

So, in effect, the player designates a target, then attempts to cast the spell. I would notify the player that they have detected an illusion, and that the specific spell (Hold Person, in this instance) has no effect and hasn't spent resources (no slot loss), but that their action isn't wasted, as they now see through the illusion.

So how do you deal with Scenario 2?

Scenario 2

Joe has cast Mirror Image.

Hero declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?

Since you're polling for individual rulings:

Ruling: Hero attempts to cast Hold Person, but the magic fails to take effect. He feels a curious emptiness around the creature, and there is nothing for his magic to anchor itself to. He ends his action without expending any spell points but it will take a fresh action to re-cast the spell on a new target.

Reasoning: This ruling maintains a balance between the usefulness of illusions and player fun (not wasting important resources like spell slots on nothing). It's also easy to adjudicate and it fits easily into the logic of how magic works in my world.

I'm cool with this.

Scenario 2

Joe has cast Mirror Image.

Hero declares to the DM that he wishes to cast Hold Person on Joe. What happens, and why?

Overall, I'd say this is really in the "GM call" territory. It's one of the areas I think 5e could have been a little more explicit on, at least giving advice/guidelines in the DMG, but the question boils down to what you're "really doing" when you cast a spell.

Are you performing a closed loop that can't have its beginning without its ending, such as trying to close a circuit which, if a load is missing from its appointed place, will not complete and thus nothing happens? Or are you firing a metaphorical gun, and whether your target is there or not doesn't matter to the fact that you did, in fact, pull the trigger?

I personally go with the latter. You perform the gestures, you pour out the magical energies, but you happened to do it to something that isn't there. Woops. Nothing happens. Just like you don't get a fueled car if you pour gasoline into an illusory gas tank while the real car is ten feet away. You get gasoline all over the ground (and quite the fire hazard).

But a DM can rule the other way, too. 5e is not explicit on this.
Eh, different circumstances. Spells aren't name-locked, unless they say so. "I'm targeting that guy, who I think is Fred," is how I would read a player saying he's targeting a particular creature that I know is actually Frieda disguised as her brother, Fred. The creature is there. She's a valid target for the spell. The spell affects her, because she's the creature at which you aimed it.

The question here is whether aiming the spell at a "creature" that isn't really a creature does anything. If you cast hold monster at a statue because you think it's a gargoyle (you paranoid adventurer, you), does that instantly reveal it's not a creature?

I think it weakens illusions and strengthens spells into free detection abilities to treat targeting as a necessary condition to even attempt the action. "Sorry, you can't do that" tells the savvy player "ah, that's an illusion; I'll hit the other one, instead."

Thanks for your thoughts. I went through similar process. I'm guessing you don't feel like posing an answer?

Mirror image specifically says in its own description that it only works on attacks.

I contend that it does not.

phb 194, "if there's ever any question weather something counts as an attack, the rule is simple: if you're making an attack roll, you're making an attack. As such, mirror image does nothing against hold person, or any other save-based spell for that matter.

Certainly Hold Person is not an attack. In fact I even agree that mirror image "does nothing" against Hold Person unless you think that Hero, for some reason, cannot see the four "Joes." I contend that, because there are four Joes, Hero must choose one of them as his target, and this has nothing to do with the mechanics of Mirror Image that are specific to attacks.

If you're having trouble picturing in your head how that plays out, note in this case the school of hold person: Enchantment. There's no physical interaction with the creature, the spell tricks the target's mind into being unable to move.

But how does Hero select a target, if he can see four Joes?

If Hero casts a Sleep spell covering both the SI and Joe, Joe falls asleep and the SI vanishes when Joe loses concentration (assuming a high enough roll for HP affected by Sleep etc). SI has zero HP, after all.

If Hero casts Fireball and blasts both Joe and SI, Joe gets hurt and the SI does not. Joe might be able to change the SI to show damage, but not as a reaction to the fireball, so the interval between BOOM and the start of Joe's turn would reveal the truth to observers.

To me, the targeting mechanisms for these spells make them irrelevant to the conversation.

If Hero casts Firebolt and targets SI, it passes right through SI and reveals it as an illusion.

It seems that most spells affect Joe differently than SI.

Therefore, I'd DM this as a single-target spell targeting SI reveals it as an illusion, including Hold Person. If Joe puts the SI in the same space as he so that the two cannot be distinguished then Hold Person would have a 50-50 chance of hitting Joe,...

We're good up to here.

...but Firebolt would hit both at the same time.

Doesn't firebolt specifically say You hurl a mote of fire at a creature or object within range, and doesn't this rule out the possibility to hitting two things?

Well unless the DM tells them it's an illusion, all they know is it's an illegal target. Maybe Joe isn't actually a humanoid and instead is just a vampire in makeup? Maybe he is in an anti-magic sphere? Maybe he is immune to spells of that level? All that a player definitely knows is that the target is not a valid one.

I agree here. Even if someone saw a real person intersecting an illusion of a person in space, this would not reveal (make transparent) the illusion. It would grant the knowledge that there is an illusion at play, but without knowing any of the "people" to be real, you wouldn't know which of the two was an illusion, nor whether both were illusions. You would only know that they are not both people (because two people can't intersect, but every other combination can).

General Comments

It seems abundantly clear to me that Mirror Image does not apply only to attacks (despite what others are claiming).

In the first place, I would nit-pick and say that Mirror Image doesn’t actually apply to attacks, because it doesn’t affect them. It applies to targeting, and to targetting with attacks only. But let’s assume that’s what is meant.

The special d20 mechanic certainly only applies to (targeting with) attacks, but the rest of the spell description does not. The rest of the spell description applies to everything. I know that many are refuting this, so I offer this in an attempt to illustrate how a conditional works: Scenario 3.

Scenario 3

Let’s consider a new spell. It’s called Burgerbeast’s Zany Image. The text is exactly the same as Mirror Image except for one small change. Instead of saying each time a creature targets you with an attack, it says each time a dwarf targets you with an attack.

Joe casts Burgerbeast’s Zany Image.

An elf (whom we will refer to as Heroine) declares a melee attack on Joe. What happens?

BurgerBeast

2016-08-15, 10:02 PM

I'm trying to come up with a real world example that parallels the language in the text for Mirror Image, with the purpose of showing that: asserting something which applies only under certain conditions does not assert anything when those conditions are not met.

If someone can come up with a better analogy, to prove me right or to prove me wrong, please do. In the mean time, here's the best I've come up with.

Burgerbeast's Terrible Party Analogy

I am having party. It will take place in Canada where the legal age of majority is 19 in my province. I will tend the bar. I will always add or remove cups to the bar to make sure that there are always four cups on the bar. I will add cups in such a way so that one cup will contain beer and three will contain water, always. The guests will be able to see the cups but not their contents.

If a person between the ages of 19-25 asks for a drink, I will roll a d20. On a 1-5 I will give them a beer. On a 6-20 I will give them a water.

So go ahead and tell me what I am going to do if someone over 25 asks for a drink.

I posit that you don't know, because the description doesn't mention it.

Yet it seems to me that many in this thread use a different form of logic, in which it obviously must be true that anytime someone over 25 asks for a drink I will always give them a beer. Apparently, the reason why this must be true is because I only specified what I would do if the person was 19-25, therefore anyone over 25 must automatically receive beer. I personally don't understand this and see it as a leap in logic.

Kane0

2016-08-15, 10:16 PM

Tanarii

2016-08-15, 10:38 PM

Also, the consequences are quite different from a check: knowing where the sound is coming from does not tell me which is real or fake (if any), and it certainly doesn't turn anything transparent.

<snip>
Right - but that suspicion can be all you really need. "I'll attack the noisy one."absolutely I should have been clear, the consequences of determining its an illusion is it goes transparent, and you know for sure it's an illusion. But it's entirely possible for a bad illusion to be pretty obvious something is up. What conclusion will be drawn as to 'something' would probably vary from character to character. But an educated guess is totally possible.

It's equally fair to say that the system is an abstraction of the game world designed to simplify and quantify the resolution of contested or uncertain actions.Did we say the same thing, or the opposite? :smallconfused:

I don't think this holds. Mirror Image has no effect on target selection, I agree. But when you cast Hold Person and there are four potential targets, Hold Person doesn't pick one of them by itself.

This logic doesn't hold up to scrutiny. If they wanted it to work in the same way, they should have said so. If they wanted it to work in a different, specific way, they should have said so. But if they wanted it to work under the same conditions it always works, they wouldn't need to say anything (I am arguing RAW here).Mirror Image doesn't indicate the duplicates are discrete non-overlapping targets in such a way that they can be chosen amongst as a target. There is nothing to indicate there is more than one actual target. Just a rule for how the spell affects targeting with
Attacks.

It seems abundantly clear to me that Mirror Image does not apply only to attacks (despite what others are claiming).How did you come to this conclusion, exactly? What makes it abundantly, I mean? :smallamused:

(Edit: If you mean your follow up paragraphs about attacks, yes, it applies to targeting with attacks. That's two conditions: targeting, and with attacks.)

Scenario 3

Let’s consider a new spell. It’s called Burgerbeast’s Zany Image. The text is exactly the same as Mirror Image except for one small change. Instead of saying each time a creature targets you with an attack, it says each time a dwarf targets you with an attack.

Joe casts Burgerbeast’s Zany Image.

An elf (whom we will refer to as Heroine) declares a melee attack on Joe. What happens?Joe is attacked. Mirror Image doesn't come into play.

Edit:fixed quotes

BurgerBeast

2016-08-16, 12:05 AM

Hero: "I cast hold person at Joe"
DM: "There are two Joe's, which one do you want to target?"

Hero A: "I suspect trickery! Would an arcana check or something help?"
DM: "Certainly, but it'll use your action to try."

Hero B: *Picks one at random*
(Targets Joe) DM: *Rolls save* "Joe is held/avoids the spell!"
OR
(Targets Illusion) DM: "You cast Hold person at Joe's Illusion, revealing his deception! The Image is also held."

Hero C: "Silly DM, I'm upcasting my Hold Person so I can target all the Joes!"
DM: *Rolls save* "Joe and his illusion are both held/avoid the spell!"

That's usually the gist of things at my table.

Looks pretty good to me. My only contention is that you're the first person to suggest that the duplicates can be held. I don;t interpret it this way because I think this violates the targeting restrictions of Hold Person (it targets creatures) and the text of Mirror Image, which says that the duplicates "ignore all other damage and effects" (other than attacks, that is).

Mirror Image doesn't indicate the duplicates are discrete non-overlapping targets in such a way that they can be chosen amongst as a target. There is nothing to indicate there is more than one actual target. Just a rule for how the spell affects targeting with Attacks.

I honestly don't know how you can figure this. The "real" target has a different AC from the images. The d20 mechanic chooses the specific target only, but an attack still has to be rolled. What's it rolled at? The one target that was determined by the [edit: d20] roll, and is decidedly not one of the others. If one duplicate is hit, it disappears while the others remain.

How did you come to this conclusion, exactly? What makes it abundantly, I mean? :small amused:

I'm sorry to sound so flippant, but reading the text and understanding logic. It's abundantly clear because of the abundant number of times I've re-read it and ran it past others to my satisfaction.

(Edit: If you mean your follow up paragraphs about attacks, yes, it applies to targeting with attacks. That's two conditions: targeting, and with attacks.)

I see what you are driving at, but it's one condition. It's just a specific condition. There's not much point in arguing this though, if you can't see my side of it.

Joe is attacked. Mirror Image doesn't come into play.

Well, at least you're consistent. So, what does the elf see?

Kane0

2016-08-16, 12:21 AM

Looks pretty good to me. My only contention is that you're the first person to suggest that the duplicates can be held. I don;t interpret it this way because I think this violates the targeting restrictions of Hold Person (it targets creatures) and the text of Mirror Image, which says that the duplicates "ignore all other damage and effects" (other than attacks, that is).

I said what I did for two reasons:
1: I did that all without cracking open the PHB, much like would be done in game. If it were actually me casting the spell I'd check it quickly in the book before my turn rolled around, but if I didn't have that time or it was another person in the group casting they probably would cast based off the information "Hold Person", ala 'Thats a person, I hold him'.
Idiosyncrasies such as 'Technically hold person targets a creature, and an image isn't a creature so now what?' just get skimmed over in favor of game flow, especially if the player doesn't have that metagame knowledge.
2: My group has a slightly odd sense of humor sometimes. Holding an illusion is exactly the kind of thing that would get a stupid grin from some of us.

I did not consider Mirror Image at all, so if something written within contradicts what would happen then we'd have to take that into account too. If Mirror image ignored spells and effects other than attacks then the DM would probably just say it was unaffected by the spell (Either by way of immunity or a successful save), or it automatically gets rid of the image targeted, or the real creature is affected anyways.

georgie_leech

2016-08-16, 01:38 AM

BurgerBeast

2016-08-16, 02:49 AM

Did you read what I wrote?

Actually, it's a leap in logic on your end. 'If X, then Y;' 'if in the range of ages from 19-25, roll a die to determine if they get a beer.' That tells us information on that specific range. We have no reason to assume you wouldn't give someone over 25 a beer, just as we don't assume you're obligated to instead punch them in the face, or recite Shakespearean Sonnets at them, or jump on a table, do a handstand, and cluck like a chicken.

Precisely. That's why I said:

So go ahead and tell me what I am going to do if someone over 25 asks for a drink.

I posit that you don't know, because the description doesn't mention it. (emphasis added)

We have no reason to assume anything. We seem to agree here. I'm not sure how you read it any other way.

'If you wear a blue shirt, you get a beer.' That does not mean you get a beer if you have a red shirt.

Yes! Nor does it mean you don't get a beer, nor does it mean...

Now here's the thing:

Mirror image says each time you are targetted by an attack, you roll a d20...

But if you are targeted by hold person you don;t do any of this stuff...

Therefore, for exactly the same reason, we cannot conclude anything about what happens if you cast hold person.

The test text in the spell Mirror Image that is conditional on being targeted by an attack does not apply.

But there is other text in the Mirror Image description that is not conditional. That text comes in the first paragraph (nd the last, but that bit is not relevant here).

So the only thing Mirror Image says that is relevant to Hold Person comes in the first paragraph, and it's this: Three illusory duplicates of yourself appear in your space. Until the spell ends, the duplicates move with you and mimic your actions, shifting position so it’s impossible to track which image is real.

So, for the caster of Hold Person, the only relevant effects of Mirror Image are that he sees four "people" in the same square, all making the same actions, and shifting positions do that they are impossible to track."

So how does someone target the true target using Hold Person? You don't just get to ignore the sections of Mirror Image that are relevant! You must acknowledge that the caster sees four "people" in the same square, all making the same actions, and shifting positions do that they are impossible to track.

[edit: Why the Terrible Party analogy Works (not sure how easy it is to follow):

Burgerbeast's (Terrible Party Analogy = Mirror Image Argument)

I am having party. It will take place in Canada where the legal age of majority is 19 in my province. I will tend the bar. I will always add or remove cups to the bar to make sure that there are always four (cups = "Joes") (on the bar = visible). I will add (cups = "Joes") in such a way so that one (cup = "Joe") will (contain beer = be the real Joe) and three will (contain water = be duplicates), always. The guests will be able to see (the cups = the "Joes") but not their (contents = true identities).

If a (person = character) (between the ages of 19-25 = wishing to attack Joe) (asks for a drink = tries to target Joe), I will roll a d20. On a 1-5 I will (give them a beer = determine the target to be Joe). On a 6-20 I will (give them a water - determine the target to be a duplicate).

So go ahead and tell me what I am going to do if (someone = a character) (over 25 = casting hold person) (asks for a drink = tries to target Joe).

I posit that you don't know, because the description doesn't mention it.

Yet it seems to me that many in this thread use a different form of logic, in which it obviously must be true that anytime someone (over 25 = casting hold person) (asks for a drink = tries to target Joe) I will always (give them a beer = determine the target to be Joe). Apparently, the reason why this must be true is because I only specified what I would do if the person was (19-25 = attacking), therefore anyone (over 25 = casting hold person) must automatically (receive beer = target Joe). I personally don't understand this and see it as a leap in logic.

Zalabim

2016-08-16, 03:38 AM

If you really want a head scratcher, eldritch blast only targets creatures and can't be used to target objects. So you can shoot an enemy to death but then can't keep blasting their dead body.

I will edit in additional comments/replies if I need to, but I needed to say this as soon as I read about Eldritch Blast. Clearly, Eldritch Blast can only be used to attack their life points directly. Once their life points are depleted, the game is over and you can't use it any more. I assume their soul also departs to some shadowy realm.

Alternatively, this is why you read the fine print.

[Add 1]

Option 3.

3.A"You find yourself unable to cast your spell. Do you want to try something else?"

3.B"You find yourself unable to cast your spell. <Next player>"

1.B"You finish your spell. <dice clatter> The figure is unmoving."

[Add 2]

I generally wouldn't expect anyone to mistake a Silent Illusion for a real creature at anything but a great distance, and I wouldn't expect it would take touching it or investigating for this to be true.

It shares this visual-only aspect with Mirror Image.

smcmike

2016-08-16, 05:07 AM

It seems abundantly clear to me that Mirror Image does not apply only to attacks (despite what others are claiming).

I will note here, again, that this is not how Crawford sees it, though of course you don't have to follow him.

Scenario 3

Let’s consider a new spell. It’s called Burgerbeast’s Zany Image. The text is exactly the same as Mirror Image except for one small change. Instead of saying each time a creature targets you with an attack, it says each time a dwarf targets you with an attack.

Joe casts Burgerbeast’s Zany Image.

An elf (whom we will refer to as Heroine) declares a melee attack on Joe. What happens?

The attack resolves normally.

Zalabim

2016-08-16, 05:08 AM

Burgerbeast's Terrible Party Analogy

I am having party. It will take place in Canada where the legal age of majority is 19 in my province. I will tend the bar. I will always add or remove cups to the bar to make sure that there are always four cups on the bar. I will add cups in such a way so that one cup will contain beer and three will contain water, always. The guests will be able to see the cups but not their contents.

If a person between the ages of 19-25 asks for a drink, I will roll a d20. On a 1-5 I will give them a beer. On a 6-20 I will give them a water.

So go ahead and tell me what I am going to do if someone over 25 asks for a drink.

That isn't quite parallel to what's happening here. The question is what happens when someone over 25 asks for a beer. To which the answer is that your rule about handing out water 75% of the time to people of the specified ages does not apply.

When the Hero asks to attack Joe, the rules of mirror image apply and that attack may be redirected to an image. When the Hero asks to cast Hold Person on Joe, the rules of mirror image do not apply and the spell resolves without consideration for the existence of mirror image.

BurgerBeast

2016-08-16, 05:29 AM

That isn't quite parallel to what's happening here. The question is what happens when someone over 25 asks for a beer. To which the answer is that your rule about handing out water 75% of the time to people of the specified ages does not apply.

When the Hero asks to attack Joe, the rules of mirror image apply and that attack may be redirected to an image. When the Hero asks to cast Hold Person on Joe, the rules of mirror image do not apply and the spell resolves without consideration for the existence of mirror image.

This is the same distinction you made in the other thread, and in the context of most of what you have said, this makes sense. Now I just have to resolve how this context avoids bigger problems in terms of targeting issues that have arisen in both threads.

If you trace some of the conversations in this thread back, I think you'll find that my analogy does apply to what other people have said, even though it does not apply to what you said.

smcmike

2016-08-16, 05:52 AM

[edit: Why the Terrible Party analogy Works (not sure how easy it is to follow):

Burgerbeast's (Terrible Party Analogy = Mirror Image Argument)

I am having party. It will take place in Canada where the legal age of majority is 19 in my province. I will tend the bar. I will always add or remove cups to the bar to make sure that there are always four (cups = "Joes") (on the bar = visible). I will add (cups = "Joes") in such a way so that one (cup = "Joe") will (contain beer = be the real Joe) and three will (contain water = be duplicates), always. The guests will be able to see (the cups = the "Joes") but not their (contents = true identities).

If a (person = character) (between the ages of 19-25 = wishing to attack Joe) (asks for a drink = tries to target Joe), I will roll a d20. On a 1-5 I will (give them a beer = determine the target to be Joe). On a 6-20 I will (give them a water - determine the target to be a duplicate).

So go ahead and tell me what I am going to do if (someone = a character) (over 25 = casting hold person) (asks for a drink = tries to target Joe).

I posit that you don't know, because the description doesn't mention it.

Yet it seems to me that many in this thread use a different form of logic, in which it obviously must be true that anytime someone (over 25 = casting hold person) (asks for a drink = tries to target Joe) I will always (give them a beer = determine the target to be Joe). Apparently, the reason why this must be true is because I only specified what I would do if the person was (19-25 = attacking), therefore anyone (over 25 = casting hold person) must automatically (receive beer = target Joe). I personally don't understand this and see it as a leap in logic.

Another analogy. The following is posted on a telephone pole in town:

"It's Critical Hit Night at Crazy Larry's Saloon!

Roll dice for free drinks! We won't go to bed til the last Orc is dead!

Any time a lady orders a drink, she will get to roll our GIANT d20 for a chance at a free drink or other fabulous prizes!

Must be of legal drinking age."
------------

What happens when a dude orders a drink?

AvatarVecna

2016-08-16, 06:29 AM

To this I would say: Nope. Mirror Image only imposes a targeting mechanic on attack rolls. It doesn't impose anything on save spells, but the caster of a save spell is still looking at four potential targets and has to choose one.

And you are welcome to have that opinion. What you are not welcome to do is tell me what my ruling is, so you can **** right off with your "Nope". Silent Image and Mirror Image are different spells that have different effects, so nobody in the freaking universe is required to declare that they work exactly the same way.

So what if Joe shared his square with a silent image?

And what if Joe shared his square with another humanoid?

I have addressed both of these already. If he shares his square with the Silent Image, it targets him. If he shares his square with both another valid target and a Silent Image that doesn't look like him, I will either roll randomly to see which valid target is affected, or I will have the spell affect whoever cast the Silent Image they're hiding in.

Well I don't think this is RAW. I think you've said as much already, though.

Of course it's not RAW, we're talking about targeting Illusion spells. You can do whatever the flying **** you want to do in regards to illusion spells in your game, but you can **** off with telling me that my ruling is wrong.

Tanarii

2016-08-16, 09:17 AM

1) I honestly don't know how you can figure this. The "real" target has a different AC from the images. The d20 mechanic chooses the specific target only, but an attack still has to be rolled. What's it rolled at? The one target that was determined by the [edit: d20] roll, and is decidedly not one of the others. If one duplicate is hit, it disappears while the others remain.

2) I'm sorry to sound so flippant, but reading the text and understanding logic. It's abundantly clear because of the abundant number of times I've re-read it and ran it past others to my satisfaction.

3) I see what you are driving at, but it's one condition. It's just a specific condition. There's not much point in arguing this though, if you can't see my side of it.

4) Well, at least you're consistent. So, what does the elf see?1) so what if they have separate ACs? That doesn't make them discrete non-overlapping individual targets that you can willingly pick amongst. That just means an illusion doesn't have armor. Nothing indicates that you can choose a single illusion as a target intentionally. You can choose to target the creature. That's it. If the thing you are targeting it with is an attack, then you follow the mechanic. If not, you don't. Simple rules resolution. Done and done.

2) Nah nah it's all good I was just teasing you about it being 'abundantly clear' because it was funny, the dictotomy of you being so sure it was clear, in the face of pretty much everyone arguing against you and the details of all those arguments. But I get what you mean. You're certain you're interpreting it correctly. That's totally cool by me, I'm sure I'm interpreting things correctly all the time.

3) I see your side of it. But your side isn't 'logic', it's extending fluff to mechanics. You're adding something that isn't in the mechanics, because you envision the spell executing itself a certain way, based on your interpretation of the fluffy non-machanical parts of it. You the description of 'multiple' images, then extrapolate that and end up at 'multiple discrete individually targetable images outside the context of the spell mechanics'.

4) The elf sees the same thing as everyone else, a swirling mess of illusionary images around a creature that he can't pick and choose between. But for whatever reason, when he tries to target the creature with an attack, there's no chance he'll his an image instead of the creature. Because magic.

5) Bonus: New Joke!: Due to participating in D&D forums for two decades, I automatically translate 'logical' as "my prefered way of seeing/doing things, that I will now rationalize to myself, then justify as logical". :smallamused:

Edit: fix spoiler tags this time /facepalm

Edit2: to be clear, my 'logical' position I'm rationalizing and justifying stems from viewing magical spells as a kind of programming function. They do what they say. You don't extrapolate from what they say to allow them to be used differently, unless what they say explicitly allows you to somehow. If something targets a creature and does fire damage, you can't also target an object and set it on fire unless the spell says so. If something redirects targeted attacks to images (which are potentially destroyed based on mechanical resolution via the attack mechanic) then that's what happens. You don't extrapolate that to mean that targeting with a non-attack must be chosen amongst them.

Not that targeting with a non-attack makes any damn sense in the first place. I feel stupid just typing it. Stoopid rules.

But how do I know Im doing some rationalizing somewhere and not everything is really 'logical'? Because that position means my view of allowing someone to attempt to target Silent Image seems (and probably is) contradictory, because it's not a creature. But everything I've said so far seems perfectly logical to me, despite that, because reasons I've explained upthread.

Xetheral

2016-08-16, 12:43 PM

Exactly. It's a set of abstract rules providing a mechanical game resolution for a game-play situation. Not a set of rules for simulating another reality.
It's equally fair to say that the system is an abstraction of the game world designed to simplify and quantify the resolution of contested or uncertain actions.Did we say the same thing, or the opposite? :smallconfused:

Neither. I was agreeing with you as to the abstract nature of the rules, but trying to point out that just because the rules are abstract doesn't mean that everyone agrees with your view of how those rules should be used. I'll try to give a more concrete explanation of the difference between our two (equally valid) statements in my reply to your most recent post:

(Responding to Burgerbeast) I see your side of it. But your side isn't 'logic', it's extending fluff to mechanics. You're adding something that isn't in the mechanics, because you envision the spell executing itself a certain way, based on your interpretation of the fluffy non-machanical parts of it. You the description of 'multiple' images, then extrapolate that and end up at 'multiple discrete individually targetable images outside the context of the spell mechanics'.

Under your approach to the abstract nature of the rules, above, what Mirror Image does is provide the caster with an explicit mechanical benefit relating to incoming attacks. Everything else is window dressing or "fluff", and thus not part of the abstraction. From the competing perspective I described above, what Mirror Image does is create illusionary duplicates, whose effect on attacks is abstracted by the rules to provide a convenient resolution mechanic. Under the approach you advocated, the abstraction is the game. Under the approach I described (which I'm not advocating because I view them as equally valid), the abstraction is a tool to help run the game.

----------

BurgerBeast, assuming I'm understanding your position correctly, you're interpreting Mirror Image under the approach I described above, where the effect of the spell is to create illusory duplicates. Because you envision those duplicates as a collection of Silent Images, you feel the rules in the spell are incomplete regarding how to abstract the effect of those images. It's perfectly fine to fix that problem by filling the hole in the rules (either as a houserule expansion to Mirror Image or a general ruling that basic mis-targeting rules should apply, if anyone could agree on what those basic rules are). But might I suggest, instead, that given the scope of the rules hole created by your interpretation of the effect of Mirror Image, it might be simpler (and possibly more in line with RAI) just to re-interpret that effect so that the provided targeting rules become adequate?

For example, if the duplicates created by Mirror Image were more analogous to the visual effect of shooting at an Agent in The Matrix (except always-on), between the overlapping nature of the images and their fleeting, constantly-shifting movements it would make sense that the individual images are not independently-targetable at all. This interpretation is still just as fully in line with the text of the Mirror Image, but permits the accompanying rules to provide a complete resolution mechanic. (This does depart from previous edition's conception of Mirror Image, but given that the spell is somewhat notorious in 3rd edition for the headaches it produces, I posit that the change was intentional.)

Theodoxus

2016-08-16, 01:44 PM

I'm AFB, but it seems to me, if you're casting a spell that requires a saving throw on someone protected by Mirror Image, you'd use the same mechanic as if you were casting an attack spell. If you roll above the necessary number "to hit" Joe, you affect Joe, and he gets to save. Otherwise, you've targeted an illusion, and your spell (Hold Person in this case), fizzles.

Upgrading Hold Person to include all the images and Joe, will cause it to affect Joe. Increasing it to only cover 3 Joe's, would have the same implication as if there was only 1 image remaining.

It seems the only fair way to adjudicate the interaction. IMO, of course.

Tanarii

2016-08-16, 02:15 PM

Neither. I was agreeing with you as to the abstract nature of the rules, but trying to point out that just because the rules are abstract doesn't mean that everyone agrees with your view of how those rules should be used.Oh for sure. I wouldn't be here if I expected everyone to agree with me. That'd be boring as hell. :smallbiggrin:

Under your approach to the abstract nature of the rules, above, what Mirror Image does is provide the caster with an explicit mechanical benefit relating to incoming attacks. Everything else is window dressing or "fluff", and thus not part of the abstraction. From the competing perspective I described above, what Mirror Image does is create illusionary duplicates, whose effect on attacks is abstracted by the rules to provide a convenient resolution mechanic. Under the approach you advocated, the abstraction is the game. Under the approach I described (which I'm not advocating because I view them as equally valid), the abstraction is a tool to help run the game.You've got that back to front. Under my view that the rules are an abstraction, said abstraction is a tool to help run the game. That's it's entire purpose, to provide a tool when a method or resolution is needed. It's when you expect the rules to be a simulation of the in-game universe, the underlying physics of the underlying universe, that the rules abstraction is the game.

I'm providing an alternate explanation to BurgerBeasts interpretation of what's happening in-game. He's got one in-game reality in mind, that it creates discrete illusions that can be targeted. Then he's trying to work backwards from that to interpret the rules abstraction via that in-game reality, making them simulation rules that are the game. I provided an alternative in-game explanation, to point out that's what he's doing because the way he's doing it is circular: Envisioning how in-game reality works, then using that to interpret the rules, then claiming that means it's how the rules say in-game reality must work.

You can either go abstraction --> possible in-game realities, or in-game reality --> interpreting abstraction. They're both fine. But if you go in-game reality --> interpreting abstraction --> justifying in-game reality, it's circular.

Edit: To be clear:
Gamist: abstraction --> possible in-game realities. Thus for any given in-game reality you desire to have, you can select the same abstraction for resolution. The abstraction is a tool for resolving (often multiple) in-game realities.
or
Simulationist: in-game reality --> interpreting abstraction. For a given-in game reality that you have, you interpret or change the abstraction so that it matches the given in-game reality. Thus, the abstraction is the game.

Segev

2016-08-16, 02:18 PM

Thanks for your thoughts. I went through similar process. I'm guessing you don't feel like posing an answer?
I did answer in that quote: I would rule that casting a spell is a process you perform, and that the target restrictions are what the spell can affect, not a prerequisite to even being able to perform the process of casting. Therefore, for a single-target spell, you'd have to specify the target unambiguously with information your PC has, and that target is "hit." Whether its affected or not depends on whether it really was valid or not. Hit Joe's mirror image with hold person, and you spend the slot, but since the image isn't a valid target, nothing happens.

It can be ruled the other way - that you can't even cast the spell without having a valid target, so no spell slot is expended, etc. - but I think that's nerfing illusions too much and giving mages a needless leg up.

As to your party-in-Canada analogy, the problem you're running into is that we don't have a default case in your party-in-Canada.

If you didn't have your random rolling method for 19- to 25-year-olds, we still wouldn't know if you'd give a 19- 25-year-old a beer or a water.

In D&D 5e, we know what happens if there's no mirror image up and Bob attacks Joe: Bob attacks Joe and things resolve normally. We also know what happens if Bob casts hold person on Joe, absent illusory doubles: Joe makes a save and is either held or not.

If mirror image has no rules for what effect its illusory duplicates have in a particular situation, the default is that the same thing happens as if the spell weren't in effect at all. So, if we assume that "Joe" is just something Bob can declare as his target, and hold person is not an "attack," then mirror image does nothing to protect Joe from Bob's spell.

Where I think that argument fails isn't on the "we don't know what happens" front, but on the targeting front. Bob doesn't know where "Joe" is. Just as he couldn't cast hold person on an invisible Joe who he hadn't figured out the location of (perhaps due to a hide check), Bob can't cast hold person on "Joe," because that's not an unambiguous definition of his target. There are 4 "Joe"s there. He has to pick one. The spell doesn't provide a strict rule on how to make that determination at random, if you assume "attack" means "anything you roll a d20 vs. AC" and only that. But the determination must still be made, and it's the DM's task to adjudicate how.

I'd probably still just roll. But you could do it with figurines, cards, or whatnot, letting Bob's player pick one of several after you secretly determine which is the real one, for instance.

smcmike

2016-08-16, 02:39 PM

Just as he couldn't cast hold person on an invisible Joe who he hadn't figured out the location of (perhaps due to a hide check),

Actually, Bob can't cast Hold Person on Joe if he can't see him, period. He could be holding Joe's invisible hand, and he still wouldn't be able to cast it on him if he can't see him.

On the other hand, I don't see any rule that says he had to know Joe's location. It's hard to think of a scenario where he can see Joe but can't figure out where he is (other than the "multiple Joes" scenario), but there is no rule that Bob must target the correct square with Hold Person.

Segev

2016-08-16, 03:39 PM

Actually, Bob can't cast Hold Person on Joe if he can't see him, period. He could be holding Joe's invisible hand, and he still wouldn't be able to cast it on him if he can't see him.

On the other hand, I don't see any rule that says he had to know Joe's location. It's hard to think of a scenario where he can see Joe but can't figure out where he is (other than the "multiple Joes" scenario), but there is no rule that Bob must target the correct square with Hold Person.

Okay. Then as long as he can see a line of Joe-impersonators, all of whom are in range, he can say "I target the real Joe" and the spell will pick the real one out, even if Bob had no clue which one it was?

And he can, as long as he can see everyone in the crowd, specify "whoever just picked my pocket," even though he didn't see who that was? As long as he can see that person now, even though he doesn't know which one of the many people he can see is the right one?

Just how little information can he get away with to have the spell sniff out exactly the right person amongst a crowd? Could he use this method to identify the murderer in a murder mystery, as long as he gathered them all together in a room?

"Alright, everyone, take a seat. I now cast hold person on the person who committed the murder we're investigating. Everybody who's innocent, please stand up...thank you. Arrest that man as soon as he can move."

Tanarii

2016-08-16, 03:45 PM

No, he just picks a creature he can see as a target. If there are multiple apparent creatures he can see to pick as a target, he chooses one. (Edit: what happens if he picks an invalid target is a separate but valid question.)

The problem is that you're assuming that Mirror Image creates multiple discrete "creatures he can see as a target" things to pick from. It doesn't explicitly do that. That's an assumption on your part.

What it does is create multiple things that, when you target the creature with an attack, there is a resolution mechanic for what happens next.

BurgerBeast

2016-08-16, 04:13 PM

Clearly, Eldritch Blast can only be used to attack their life points directly. Once their life points are depleted, the game is over and you can't use it any more. I assume their soul also departs to some shadowy realm.

(I assume that by life points you mean that points.) This strikes me as interpretation that goes beyond the RAW of the spell. For example, what if the "creature" is a PC who is at zero hit points and dying. This PC's hit points are depleted, but the PC is alive, and is therefore a creature. It would seem to me that you can drain hit points until the PC is dead.

I will note here, again, that this is not how Crawford sees it, though of course you don't have to follow him.

I will note here, again, that I know. I'm not arguing RAI, I'm arguing RAW.

The attack resolves normally.

And what does the elf see?

Another analogy. The following is posted on a telephone pole in town:

"It's Critical Hit Night at Crazy Larry's Saloon!

Roll dice for free drinks! We won't go to bed til the last Orc is dead!

Any time a lady orders a drink, she will get to roll our GIANT d20 for a chance at a free drink or other fabulous prizes!

Must be of legal drinking age."
------------

What happens when a dude orders a drink?

Well, first of all, the fact that it is posted on a telephone pole implies nothing. There may not even be a party. But I'm going to assume that the party is real and the ad is truthful, so the ad basically functions as RAW in this case.

Answer: It's impossible to know. Nothing that applies to the ladies applies to the dude. So, how does one (the DM) decide what happens when the dude orders a drink? Well, the DM ignores the specifically described method for choosing a drink for a lady, but he still has to come up with a reasonable method for choosing a drink for a dude. With no guideline to follow, he must resort to a general method: maybe come up with a menu and let the dude pick.

All that is important is that the DM:
(1) ignores "will get to roll our GIANT d20 for a chance at a free drink or other fabulous prizes!" because it doesn't apply to the dude (in the case of Mirror Image, the DM ignores all of the d20 mechanics for a save spell attack)
(2) considers all of the rest of the relevant information: It is still "Critical Hit Night at Crazy Larry's Saloon!", the dude is still there, and it is still possible for a dude to order drinks. (In the case of Mirror Image, the DM still considers that Joe is still there, and there are still 3 illusory duplicates that are indistinguishable)

Thus, a DM could rule in a number of ways, such as "the bartender refuses to give you a drink," "the bartender gives you a beer," "the bartender lets you choose between four different drink options" and "the bartender lets "roll [the] GIANT d20 for a chance at a free drink or other fabulous prizes!" (Likewise, in the case of Mirror Image, the DM can rule in any number of ways)

My point has always been that there is no logical to exclude any of these possibilities in the case of the man. (and it is the same in the case of Mirror Image)

Beyond this, there is no way to say "since he is a dude, none of the text described applies applies to him, therefore he's not even at Critical Hit Night." Nor "Since he's a man, he automatically wins a beer on the wheel - he doesn't even have to roll the spin. (thus, there is no way to say "since the attack is a save spell attack, the caster effectively can't even see that there are duplicates." Nor "since it's a save-spell attack, it automatically strikes the true target - the caster doesn't even have to pick one."[/QUOTE]

Now, I will emphasize that there is no reason given in the text of Mirror Image to explain why Hold Person can simply ignore the general effects (first paragraph) of the Mirror Image spell.

However, I will concede straightaway that anyone who advocates the position, as Zalabim does, that the spell Hold Person, itself, by virtue of the fact that it targets a creature, does allow Hold Person to effectively ignore the general effects (first paragraph) of Mirror Image, has a strong case.

The problem as I see it, with such a view, is that it robs illusions of much of their function.

So, if you advocate this position, you can do two things:

(1) acknowledge that this is in fact a limit of illusions that they are not targetable

(2) point to something specific that makes the illusory duplicates of Mirror Image significantly different from a typical illusion, specifically in terms of targetability

And you are welcome to have that opinion. What you are not welcome to do is tell me what my ruling is, so you can **** right off with your "Nope". Silent Image and Mirror Image are different spells that have different effects, so nobody in the freaking universe is required to declare that they work exactly the same way.

You misunderstand. You can rule however the f**k you want. I am saying that your ruling is purely invented and has nothing to do with the text (and, as mentioned, you can do this if you f**king want). "Nope" = "nope, I don't agree." Not "nope, you can't do that."

You'll notice that I previously said "Yep" to one of your rulings. It meant "Yep. I agree." It didn't mean "Yep. I will let you do that." But you didn't interpret my words unfavourably in that case. You didn't get mad at me for presuming to "allow you do anything," so apparently you're inconsistent in your readings.

So get that f**king knot out of your panties.

I have addressed both of these already. If he shares his square with the Silent Image, it targets him. If he shares his square with both another valid target and a Silent Image that doesn't look like him, I will either roll randomly to see which valid target is affected,...

Good. I like you consistency.

...or I will have the spell affect whoever cast the Silent Image they're hiding in.

Wait, whuh? Why?

Of course it's not RAW, we're talking about targeting Illusion spells. You can do whatever the flying **** you want to do in regards to illusion spells in your game, but you can **** off with telling me that my ruling is wrong.

Did you not read the second sentence? The one that said "I think you've said as much already"? Did I say your ruling was wrong? I simply stated that it isn't RAW. Holy f**k, pal. You feeling all right?

1) so what if they have separate ACs? That doesn't make them discrete non-overlapping individual targets that you can willingly pick amongst.

It makes them discrete. Maybe we have different Ideas about what discrete means.

Also, nothing in the text says that they overlap. (Am I misinterpreting "space" as a game-term meaning a 5'x5' grid space? Is "space" just the usual word in this context as opposed to a game-term?)

That just means an illusion doesn't have armor.

But the image does not. Hence they are different. Hence, discrete.

Nothing indicates that you can choose a single illusion as a target intentionally.

Agreed. However, you can strike a single illusion. Hence, it's a different (i.e. discrete) illusion from the other two, and it's different than than the true target.

You can choose to target the creature. That's it.

Except that "it is impossible to track which image is real," so how do you know that what you chose is not an illusion. (This might be semantic, on the basis of what it means to target, or it might be a real problem. I'm trying to get to the bottom of it. It's hard because I know how I am using the word target , but it's harder to determine what others mean when they use the word)

If the thing you are targeting it with is an attack, then you follow the mechanic.

Yes. The mechanic is there to resolve the targeting issue. The targeting issue arises from the first paragraph of the spell description, which applies generally, though. So the targeting issue arises and exists for everything. What the mechanics does is resolve this problem for attacks only. It's a solution for attacks only. What I am saying is that the problem exists for Hold Person, but the spell doesn't offer a solution.

If not, you don't. Simple rules resolution. Done and done.

Exactly. So, in my view, the targeting problem still exists, but no solution is presented in the text.

3) I see your side of it. But your side isn't 'logic', it's extending fluff to mechanics. You're adding something that isn't in the mechanics, because you envision the spell executing itself a certain way, based on your interpretation of the fluffy non-machanical parts of it. You the description of 'multiple' images, then extrapolate that and end up at 'multiple discrete individually targetable images outside the context of the spell mechanics'.

I've been charged with this before. I'm innocent until proven guilty. Please tell me which fluff I've extended to which mechanics, and I will answer the charge. I contend that the mechanics support the interpretation that the images are discrete and independently targetable (but targetable might mean different things to us. Differentiating this can only help the conversation, so I'm willing to try to get to the bottom it.)

4) The elf sees the same thing as everyone else, a swirling mess of illusionary images around a creature that he can't pick and choose between. But for whatever reason, when he tries to target the creature with an attack, there's no chance he'll his an image instead of the creature. Because magic.

So, if an elf sees "a swirling mess of illusionary images," you think she should automatically have the ability to target the real person in there with a melee attack, even though the real person is not distinguishable from the three images? And, even though there is no specific ruling on the matter? (Remember, the spell does not provide a mechanic for the elf just because it provides a mechanic for the dwarf. There is no mechanic for the dwarf elf. None.)

5) Bonus: New Joke!: Due to participating in D&D forums for two decades, I automatically translate 'logical' as "my prefered way of seeing/doing things, that I will now rationalize to myself, then justify as logical". :small amused:

Lol. Fair enough. I try to not throw the word around lightly.

Edit2: to be clear, my 'logical' position I'm rationalizing and justifying stems from viewing magical spells as a kind of programming function. They do what they say.

I charge you with not doing what the spell says, but with one condition. I say that in the case of Hero casting Hold Person, you do not do what the spell says, because the spell says that it "is impossible to track which image is real," yet you let Hero target (track and choose) Joe (real). This is supposed to be impossible, according to the spell. Unless you concede that Hold Person functions in such a way that it allows Hero to make this distinction.

You don't extrapolate from what they say to allow them to be used differently, unless what they say explicitly allows you to somehow. If something targets a creature and does fire damage, you can't also target an object and set it on fire unless the spell says so.

Yes. I agree. But in the case of setting fire, you might find that there is a general rule that applies (in D&D it doesn't). So, if there was a general rule that fire attacks always give a percentage chance of setting fire, then all fire attacks get this whether their description says so or not. If you wanted the spell to have an effect such as a greater percentage chance, then the spell must specify it. I think we are in complete agreement here.

If something redirects targeted attacks to images (which are potentially destroyed based on mechanical resolution via the attack mechanic) then that's what happens.

Yes, but there is a general rule for targeting. So before you worry about re-directing the attack, you have to determine whether the thing is targeted to begin with. In the case of Mirror Image, note that the text says "Each time a creature targets you with an attack" as opposed to Each time creature [B]makes an attack roll. I contend that this is important. The spell would not change, by your interpretation, if the spell said the latter. But I contend that this difference is intentional because it triggers before the attacker actually attacks. It redirects the targeting process, not the attack directly. After it redirects the targeting process, the attack is then resolved.

I contend that if the spell operated like some other abilities (the Shield Spell or the ability Illusory Self), then the mechanics would support the explanation that the attack itself is re-directed. But Mirror Image does not interfere with the attack roll. It interferes with the targeting process. After this interference, the attack roll is made normally. The designers could have made it so that the attacker rolled the attack, and then the attack was redirected. In this case, the roll (that was already made) would just be applied to the appropriate AC, after the spell moved the attack to a target.

You don't extrapolate that to mean that targeting with a non-attack must be chosen amongst them.

And I never did. I took the information given in paragraph 1, in combination with the Hold Person text, to determine that a target must be chosen. This is why I do not contend that the d20 method applies to save-spell attacks. But this doesn't mean that save-spell attacks get to ignore the information in paragraph 1 of Mirror Image. No one gets to ignore the text in paragraph 1.

Not that targeting with a non-attack makes any damn sense in the first place. I feel stupid just typing it. Stoopid rules.

And this is a big part of the problem.

But how do I know Im doing some rationalizing somewhere and not everything is really 'logical'? Because that position means my view of allowing someone to attempt to target Silent Image seems (and probably is) contradictory, because it's not a creature. But everything I've said so far seems perfectly logical to me, despite that, because reasons I've explained upthread.

This is a great question. Without getting to douchy, you stay true to yourself, never accept anything without reason, and be willing to abandon a position when it stops making sense. (I can already anticipate the number of members who will take umbrage with this :smallbiggrin:)

BurgerBeast, assuming I'm understanding your position correctly, you're interpreting Mirror Image under the approach I described above, where the effect of the spell is to create illusory duplicates.Because you envision those duplicates as a collection of Silent Images, you feel the rules in the spell are incomplete regarding how to abstract the effect of those images.

I do not envision those duplicates as Silent Images. I think Silent Image has no bearing on the discussion about Mirror Image. I do, however, think that the rules of illusions, in general, apply to both spells. The only reason I brought up Silent Image was to illicit the general rules around illusions. Anyone who applies a particular general rule to some illusions, but not all, must either point to justification for such a discrepancy or admit their error.

(snipped because irrelevant)[QUOTE]

[QUOTE]For example, if the duplicates created by Mirror Image were more analogous to the visual effect of shooting at an Agent in The Matrix (except always-on), between the overlapping nature of the images and their fleeting, constantly-shifting movements it would make sense that the individual images are not independently-targetable at all.

This strikes me as what you charged me with. This seems to me to have no justification in the text, and it seems add more to the spell than what is actually there. I also have no problem with you choosing to do this, but I don't like it because I think it changes the spell. What I am trying to do is not change the spell.

I know that this probably sounds funny, given that you think I am trying to equate Silent Image to Mirror Image, but I am not trying to make this equation and I oppose the idea of equating them (as you do).

This interpretation is still just as fully in line with the text of the Mirror Image,

If by this, you mean that both explanations unjustifiably add to the spell, but roughly to the same extent, then I agree. There's no reason to prefer one over the other. But I reject both because I don't wish to change the spell.

...but permits the accompanying rules to provide a complete resolution mechanic. (This does depart from previous edition's conception of Mirror Image, but given that the spell is somewhat notorious in 3rd edition for the headaches it produces, I posit that the change was intentional.)

Yeah. This might sound weird, but I think the Mirror Image spell description is complete. I think the general rules for targeting with spells are incomplete, and they are what cause problems here. I contend that the Mirror Image text clearly leaves anything that is not an attack in the realm of the general rules.

I'm AFB, but it seems to me, if you're casting a spell that requires a saving throw on someone protected by Mirror Image, you'd use the same mechanic as if you were casting an attack spell. If you roll above the necessary number "to hit" Joe, you affect Joe, and he gets to save. Otherwise, you've targeted an illusion, and your spell (Hold Person in this case), fizzles.

And I contend that while the spell Mirror Image does not direct you to do this, this is a perfectly acceptable way to resolve the problem. According to Jeremy Crawford, this is not RAI. I contend that it is acceptable under RAW because RAW doesn't cover it.

Upgrading Hold Person to include all the images and Joe, will cause it to affect Joe. Increasing it to only cover 3 Joe's, would have the same implication as if there was only 1 image remaining.

This raises other problems, surrounding how Hold Person functions. But I agree that this is more-or-less (the last one seems off mathematically to me) acceptable according to RAW (but again, RAW says nothing).

It seems the only fair way to adjudicate the interaction. IMO, of course.

I suggest that there are other fair ways. I just happen to be interested in resolving the RAW, and hearing what others have to say.

BurgerBeast

2016-08-16, 04:32 PM

It sounds to me like we rule more-or-less the same way in this case, then, and for more-or-less the same reasons. Thanks for the clear explanation. I will ignore most of it just to raise one distinction that I think is relevant.

As to your party-in-Canada analogy, the problem you're running into is that we don't have a default case in your party-in-Canada. If you didn't have your random rolling method for 19- to 25-year-olds, we still wouldn't know if you'd give a 19- 25-year-old a beer or a water.

Yes, on the face of it, this is a problem. But it's not really a relevant problem because my whole contention is that there is no rule, not that my specific rule is right. So, in the case of the party-in-Canada, we are forced to use or judgment based on what we do know. I contend that this is what happens in the D&D example, too.

...If mirror image has no rules for what effect its illusory duplicates have in a particular situation, the default is that the same thing happens as if the spell weren't in effect at all.

I disagree here. The spell is in effect. The spell has effects that do not depend on being attacked. I contend that the same thing happens as if the spell is in effect and the caster is not attacked.

The entire first paragraph of Mirror Image is in effect for the entire duration of the spell, whether the caster is attacked or not. These effects are "Three illusory duplicates of yourself appear in your space. Until the spell ends, the duplicates move with you and mimic your actions, shifting position so it’s impossible to track which image is real." Just because the caster is not attacked is not sufficient reason to say that these effects are also not in effect. These effects do not depend on being attacked.

So, if we assume that "Joe" is just something Bob can declare as his target, and hold person is not an "attack," then mirror image does nothing to protect Joe from Bob's spell.

In my view, I'd change "does nothing to protect Joe from Bob's spell" to "does the same things is generally does: there are three illusory duplicates of Joe in Joe's space. Until the spell ends, the duplicates move with Joe and mimic his actions, shifting position so it’s impossible to track which image is Joe."

Where I think that argument fails...

I may fail there, too. But would you mind shedding some insight on where you think I've gone wrong, above?

IShouldntBehere

2016-08-16, 04:37 PM

Mellack

2016-08-16, 04:53 PM

There is no evidence that I see that says the mirror images are individually targetable. The spell says they are constantly shifting and impossible to track. I would interpret that to mean you cannot target separately. This is supported in that if using an attack, it specifies how you have to roll to determine which that attack is actually at. Nothing is mentioned about having to first pick one of the four first. Indeed, in burger's reading you would be forced first to make a choice, and then also have the mechanical roll that possibly sends it to an image. It would seem to me that such a reading would obviously be too much for such a spell.

Tanarii

2016-08-16, 05:02 PM

Hey BurgerBeast I don't have time to digest and respond properly until later, but I just wanted to say thanks for not getting bent out of shape. Not many people can be told that their logical isn't logical without regarding it as a personal attack. (And I don't take any of your responses as such either.)

AvatarVecna

2016-08-16, 05:26 PM

You misunderstand. You can rule however the f**k you want. I am saying that your ruling is purely invented and has nothing to do with the text (and, as mentioned, you can do this if you f**king want). "Nope" = "nope, I don't agree." Not "nope, you can't do that."

You'll notice that I previously said "Yep" to one of your rulings. It meant "Yep. I agree." It didn't mean "Yep. I will let you do that." But you didn't interpret my words unfavourably in that case. You didn't get mad at me for presuming to "allow you do anything," so apparently you're inconsistent in your readings.

So get that f**king knot out of your panties.

Good. I like you consistency.

Did you not read the second sentence? The one that said "I think you've said as much already"? Did I say your ruling was wrong? I simply stated that it isn't RAW. Holy f**k, pal. You feeling all right?

My problem is that I have never been talking about RAW, I've only ever been talking about how I would rule things - and you seem to agree with that statement. So the first time I gave my ruling, you said (as you've stated above) "Yep", as if stating that you agree. While your agreement is cool, I also really don't care about it, because I did not come into this thread to argue with you or anybody else, I came into this thread to present my ruling of the situation. At this point, you asked for my thoughts on a similar-but-decidedly-different situation, which I presented...at which point you responded with "Nope, that's not RAW compliant". Of course it's not RAW compliant, it's my ruling; rulings are, by their very definition, not rules, so they can't be RAW. The statement "your ruling isn't RAW" is as meaningless as the phrase "elves are not dwarves".

You asked what I would say happens in that situation; when I told you, you dismissed my ruling for not being RAW, never stated that you were dismissing it because it was RAW, and instead gave a strong impression that you were dismissing it because you just thought it was stupid. If, instead of just saying "Nope" in regards to me stating what my ruling was, you had said "I don't agree with that ruling", there would be no swearing, but because you presented your dismissal in a misleading way that gave a strong impression of "what a stupid ruling", you get cussed at.

Wait, whuh? Why?

I should explain that part better, since it seems to have confused you. Earlier, when discussing the idea of multiple people hiding under an illusion they had cast, I stated I would likely default to randomly determining which hiding humanoid got targeted; alternatively, if one of the two people hiding under the Image was the caster of the image, I would say that they were the ones targeted, rather than randomly determining the target (if you need to justify it, we can just call it that the magical energy of the Hold Person spell followed the channel of magical energy that was leading from the Silent Image to the caster). However, if the caster is not hiding under the Silent Image, and the multiple people hiding under it are not the casters of the Image, I would not redirect the Hold Person to the caster because the caster isn't where the spell was sent.

smcmike

2016-08-16, 06:47 PM

Okay. Then as long as he can see a line of Joe-impersonators, all of whom are in range, he can say "I target the real Joe" and the spell will pick the real one out, even if Bob had no clue which one it was?

No. I understand your point, but I think Mirror Image works differently from your examples. Mirror image makes it hard to figure out exact where the target is, but doesn't create real problems in terms of identity - you are still targetting Joe, who you can see. You aren't sure which position he's in, but apparently that doesn't matter.

I will note here, again, that I know. I'm not arguing RAI, I'm arguing RAW.

And what does the elf see?

Yeah I'm not really trying to convince you. Mostly I'm trying to show everyone else why you are wrong. Though, of course, I'd love to convince you. The elf sees a person who cast a spell that would only fool a dwarf.

Well, first of all, the fact that it is posted on a telephone pole implies nothing. There may not even be a party. But I'm going to assume that the party is real and the ad is truthful, so the ad basically functions as RAW in this case.

Haha. This is wonderful. Thank you for assuming that my hypo was not a trick designed to make you think there was a party when there was not. The fact that it was posted on a telephone pole is fluff.

Answer: It's impossible to know.

Of course it isn't impossible to know. Crazy Larry's Saloon is a Saloon. Saloons serve alcoholic drinks for money. This is context.

With no guideline to follow, he must resort to a general method: maybe come up with a menu and let the dude pick.

Yeah, that'll work.

Thus, a DM could rule in a number of ways, such as "the bartender refuses to give you a drink," "the bartender gives you a beer," "the bartender lets you choose between four different drink options" and "the bartender lets "roll [the] GIANT d20 for a chance at a free drink or other fabulous prizes!" (Likewise, in the case of Mirror Image, the DM can rule in any number of ways)

My point has always been that there is no logical to exclude any of these possibilities in the case of the man. (and it is the same in the case of Mirror Image)

Wait, the DM can rule in any number of ways in response to MI? Are you now simply arguing that Mirror Image is ambiguous, and your targetting requirement is simply one possible valid interpretation? I thought your claim was harder than that. I would agree with the claim that the spell text is somewhat ambiguous.

Beyond this, there is no way to say "since he is a dude, none of the text described applies applies to him, therefore he's not even at Critical Hit Night." Nor "Since he's a man, he automatically wins a beer on the wheel - he doesn't even have to roll the spin. (thus, there is no way to say "since the attack is a save spell attack, the caster effectively can't even see that there are duplicates." Nor "since it's a save-spell attack, it automatically strikes the true target - the caster doesn't even have to pick one."

Ok, so it's ambiguous. Given this ambiguity, what would be the most reasonable assumption for the man to make when deciding whether to head over to Larry's? Should he head in assuming that he will get a chance to roll the Giant D20? I would say no. He might point to the first sentence from the flier, but management would likely direct his attention to the later clause, which defined what the first sentence really meant.

The problem as I see it, with such a view, is that it robs illusions of much of their function.

So, if you advocate this position, you can do two things:

(1) acknowledge that this is in fact a limit of illusions that they are not targetable

(2) point to something specific that makes the illusory duplicates of Mirror Image significantly different from a typical illusion, specifically in terms of target ability.

I will try to return to this question in a bit. It deserves some thought, and I want to make sure I'm clear.

In the meanwhile, I do have three questions for you:

1. The clause concerning attacks is triggered "each time a creature targets you with an attack." Why don't you apply your general rule regarding targetting prior to this point? If a creature targets one of your images, it isn't targetting you with an attack, right?

2. In your interpretation, is there any meaningful difference between how attacks and non-attack targeted spells play out? It seems like the proposed resolutions are just variations of the D20 method for attacks.

3. If not, why did the spell provide a specific mechanics for one, but not the other?

TurboGhast

2016-08-16, 08:02 PM

-Snip-
Before we even arrive at your examples, it seems we have to determine what is targeted. How does the DM do that?

And, just to be clear: does Hero lose the spell slot and/or the action?

Again, this seems reasonable to me.
-Snip-

Target is determined by the regular targeting rules used at the table. Presumably, both the target and the image are potential targets that appear identical, but differ in location. If the player indicates the illusion as the target of Hold Person, the caster would lose their spell slot and action via casting it at an illusion.

Thanks for liking my ruling!

BurgerBeast

2016-08-16, 10:07 PM

My image of Mirror Image was something that was rather unlike a freestanding image like Silent Image or another illusion.

In the OP the Silent Image is 10ft away creating two distinct "Bobs". Mirror image I always imagined creatures blurred indistinct mass of Bobs. They pass through one another, never quite wholly coming out of contact with each other or the caster. It is as though the Mirror Images are in a very close, wild dance with one another including the "Real Person" and they can all pass through each other.

If we consider then someone coming up with say a Bow & Arrow and approaching the two situations we can see how they are unlike. The bow & arrow user seeing the two silent image bob's standing 10 feet apart has two obviously distinct targets. Neither target is ambiguous, nor is particularly obvious (putting aside hearing for the moment) that anything is different about them. Each target is distinct and normal. They are to his senses equally bob, and he can aim one way or another.

Sure. So you're distinguishing between the two examples here based either on difficulty to resolve individuals on proximity. So I suggest working through more examples, in such a way as to discover exactly when and why proximity affects targeting and when and why resolution affects targeting. For example, can you target a sword in someone's hand? An apple on someone's head? One of two grapplers? One of seven gnomish grapplers who for some reason are arranged in a tangled ball of flesh?

Now we consider the Mirror Image case. Our bow wielder no longer has two distinct targets: He has one target. The mass of dancing, intersecting bobs all sharing the same space.

Excuse the fact that I will break your text down and nitpick, but it's intended to serve the conversation. Is it the intersecting or the sharing the same space that affects targeting? To me it seems to be neither, because there is nothing in the rules to prevent targeting an individual item on a character's person, as an example of sharing a space (here I am interpreting "space" to mean a 5'x5' square). Likewise, intersecting, for me, doesn't make an individual target ambiguous. Three intersecting illusions are still three illusions. It is possible to hit just one.

It's obvious there is magical illusion bullcrap going on after all, when else do 4 clones of each other move randomly through each other at a rapid rate?

I can agree to this without agreeing that the four "duplicates" collectively comprise one target.

The challenge here is less choosing the correct target and more figuring out where the target begins and ends so that he can place the pointy bits of his arrow there.

Agreed, and recognizing this, the designers provided rules for doing so. The rules they provided apply to attacks.

He must through his aim tell his arrow exactly where to strike, he uses his arms and the laws of physics to tell the arrow "Go to that exact point, because thats where I think Bob is in that mess of movement" and he may or may not wind up right.

This part: the "Go to that exact point, because thats where I think Bob is in that mess of movement" and he may or may not wind up right is resolved by the attack roll (the d20 roll against the AC of the "duplicate" that is located where I think Bob is in that mess of movement). It is the attack roll that tells us whether the attack, as targeted, hits or misses.

But before we reach the attack roll, a target must be selected. Bob must tell the arrow "Go to that exact point, because thats where I think Bob is in that mess of movement". The spell steps in and does this, via the spell-specific d20 mechanic.

Our spell caster comes into the silent image problem, not with a bow but with a vague piece of magic that "targets a creature" and he more or less has the same problem as our bow wielder, he has two distinct targets and must choose which one he thinks is a the creature. He might do this by seeing through the illusion with a save, at random, by using his action, or just because one is playing the drums and screaming but suspiciously making no noise. Either way he can choose a right and wrong target for our spell.

Yes. Nothing in the spell description interferes with the choice. It proceeds according to the general rules of the game. Then the spell itself resolves via the rules of the spell (the spell rolls a saving throw to determine if the spell more-or-less "hits" or "misses").

Our spell caster comes into the mirror image problem now with the vague piece of magic that "targets a creature" and his situation is a bit different than the bow user. He has no projectile, he needn't place anything in the exact physical space it is occupying at the time. In this he can select as his one target the creature at the center of that tightly confined dance of limbs and bodies merging, intersecting and separating.

In saying "the creature at the centre" you are identifying one target, which may or may not be the real Joe.

We know that he needn't place something with precise physical precision since it doesn't require an attack roll, how nimble or armored the target is. It's fair to point at this odd mass of swirling things and say "That Creature, the one in that mass of tangled magical fluffery" because the mirror images aren't wholly distinct from bob in the way his freestanding silent image is. (emphaisis added)

This is fine in the context of your fluff. But this contradicts the spell description which says that "it’s impossible to track which image is real." Your fluff makes it possible to track which image is real. Cool with me if this is your ruling, but it doesn't match up to the spell description.

Of course none of this applies if you see Mirror Image as merely as producing clones that stand cleanly shoulder to shoulder with each other, freestanding with their own independent spaces. I guess I can can kind of see that interpretation working, but I struggle a bit to find it in the text exactly.

This description is not in the text. When discussing RAW, I am open to varied fluff, so long as the fluff doesn't change the way the spell works or contradict the spell description. I think your fluff contradicts the spell description.

There is no evidence that I see that says the mirror images are individually targetable. The spell says they are constantly shifting and impossible to track. I would interpret that to mean you cannot target separately.

Except, as we are about to see, any attack only occurs against one of the images. Thus, before the attack occurs, it targets one "duplicate." Therefore it does target separately. Therefore is possible to target separately, at least in the case of attacks.

This is supported in that if using an attack, it specifies how you have to roll to determine which that attack is actually at.

Yes. This is the act of picking a target. In the case of an attack, the spell text hijacks the player's ability to do so, and does it for him. Presumably this is to simplify what could otherwise be a complicated resolution system.

Nothing is mentioned about having to first pick one of the four first.

Not directly. But notice that you don't roll and attack first and determine the AC, and then determine the target, and ten resolve whether it hits or misses. Instead you roll to determine the target first, and then roll the attack, and then resolve whether it hits or misses. This seems significant. Whether the player or the spell mechanics are responsible for "targeting," targeting is clearly occurring before the attack roll.

Indeed, in burger's reading you would be forced first to make a choice, and then also have the mechanical roll that possibly sends it to an image. It would seem to me that such a reading would obviously be too much for such a spell.

This (the first sentence) is incorrect.

My problem is that I have never been talking about RAW, I've only ever been talking about how I would rule things - and you seem to agree with that statement. So the first time I gave my ruling, you said (as you've stated above) "Yep", as if stating that you agree. While your agreement is cool, I also really don't care about it, because I did not come into this thread to argue with you or anybody else, I came into this thread to present my ruling of the situation. At this point, you asked for my thoughts on a similar-but-decidedly-different situation, which I presented...at which point you responded with "Nope, that's not RAW compliant". Of course it's not RAW compliant, it's my ruling; rulings are, by their very definition, not rules, so they can't be RAW. The statement "your ruling isn't RAW" is as meaningless as the phrase "elves are not dwarves".

You asked what I would say happens in that situation; when I told you, you dismissed my ruling for not being RAW, never stated that you were dismissing it because it was RAW, and instead gave a strong impression that you were dismissing it because you just thought it was stupid. If, instead of just saying "Nope" in regards to me stating what my ruling was, you had said "I don't agree with that ruling", there would be no swearing, but because you presented your dismissal in a misleading way that gave a strong impression of "what a stupid ruling", you get cussed at.

Except you have repeatedly mentioned that you don't care what I think, so why get upset if you think that I think you are stupid? :smallwink:

Anyway. Let's move on (You can get in a last word if you want, and I'll leave it).

I should explain that part better, since it seems to have confused you. Earlier, when discussing the idea of multiple people hiding under an illusion they had cast, I stated I would likely default to randomly determining which hiding humanoid got targeted; alternatively, if one of the two people hiding under the Image was the caster of the image, I would say that they were the ones targeted, rather than randomly determining the target (if you need to justify it, we can just call it that the magical energy of the Hold Person spell followed the channel of magical energy that was leading from the Silent Image to the caster). However, if the caster is not hiding under the Silent Image, and the multiple people hiding under it are not the casters of the Image, I would not redirect the Hold Person to the caster because the caster isn't where the spell was sent.

Thanks. Makes sense.

Yeah I'm not really trying to convince you. Mostly I'm trying to show everyone else why you are wrong. Though, of course, I'd love to convince you. The elf sees a person who cast a spell that would only fool a dwarf.

So, just the person. No duplicates?

Or the duplicates, too? (They just don't fool him.)

Haha. This is wonderful. Thank you for assuming that my hypo was not a trick designed to make you think there was a party when there was not. The fact that it was posted on a telephone pole is fluff.

Hey, I'm sorry to have to make that clear, but you never know what you'l get back as a "gotcha" on an online forum. Just covering my bases. :smallbiggrin:

Of course it isn't impossible to know. Crazy Larry's Saloon is a Saloon. Saloons serve alcoholic drinks for money. This is context.

You have enough context to limit the possibilities, but you have no context to assume what will happen. There is no way to say any individual event will happen with any degree of certainty without more information.

Wait, the DM can rule in any number of ways in response to MI? Are you now simply arguing that Mirror Image is ambiguous, and your targetting requirement is simply one possible valid interpretation? I thought your claim was harder than that.

Now and always. Feel free to go back to my very first post in the other thread. I have never wavered in my stance.

I would agree with the claim that the spell text is somewhat ambiguous.

I wouldn't call it ambiguous. I'd say it makes it clear that there are no RAW to deal with how Mirror Image interacts with save spells.

Ok, so it's ambiguous. Given this ambiguity,...

Well, we'll roll with it... :smallwink:

...what would be the most reasonable assumption for the man to make when deciding whether to head over to Larry's? Should he head in assuming that he will get a chance to roll the Giant D20? I would say no. He might point to the first sentence from the flier, but management would likely direct his attention to the later clause, which defined what the first sentence really meant.

He should not make any assumption. He should only expect what is written. (1) He should expect it to indeed be "Critical Hit Night at Crazy Larry's Saloon!" (2) He should expect to Roll dice for free drinks (but not necessarily with the same conditions described specifically for ladies), because the sign said as much. (Note: if he is assay to marketing tricks, he might ask for clarification about whether he gets to roll free drinks, since he's not a lady - but as advertised he should get to, and if they didn't let him, he'd have good resonate complain). (3) He should expect the place to stay open "til the last Orc is dead!" (whatever that means). (4) He should expect to see ladies "get to roll [their] GIANT d20 for a chance at a free drink or other fabulous prizes!" any time "the lady orders a drink."

In the meanwhile, I do have three questions for you:

1. The clause concerning attacks is triggered "each time a creature targets you with an attack." Why don't you apply your general rule regarding targetting prior to this point? If a creature targets one of your images, it isn't targetting you with an attack, right?

I would, but the text of Mirror Image over rides it. To explain it better: "targeting" is the act of resolving the target. It is a process even though it isn't necessarily obviously so. It begins when a player declares the intended target and it ends when the target is understood by the DM. In most cases it resolves simply: "I attack that orc (points at the min)" -> DONE or "I attack the kobold with the battleaxe" -> DM locates the kobold with battle-axe and makes there aren;t any others with battleaxes, maybe even confirms "this one?" -> player nods -> DONE or "I attack the nearest goblin" -> DM looks at table and makes makes a determination OR the DM resolves it his own mind in TotM -> DONE.

Targeting Joe, it seems to me, almost always starts as "I want to attack the wizard" -> DM knows that Mirror Image is active -> DM interrupts targeting process with Mirror Image mechanic -> DM resolves d20 mechanic -> DONE.

Note that if a player specified "I want to attack one of the duplicates" it would resolve in exactly the same way, including how the d20 resolves.

2. In your interpretation, is there any meaningful difference between how attacks and non-attack targeted spells play out? It seems like the proposed resolutions are just variations of the D20 method for attacks.

In terms of targeting? No, no difference unless specified. They pretty much are. But since the spell doesn't suggest any one of them in particular, the DM is free to resolve it however he sees fit. Not so if it is an attack. If it is an attack, he must use the mechanic given in the text.

3. If not, why did the spell provide a specific mechanics for one, but not the other?

I don't know, and I don't intend to speculate. I only know what they wrote.

Target is determined by the regular targeting rules used at the table. Presumably, both the target and the image are potential targets that appear identical, but differ in location. If the player indicates the illusion as the target if Hold Person, the caster would lose their spell slot and action via casting it at an illusion.

Thanks for liking my ruling!

Cool. This all looks good.

Mellack

2016-08-16, 10:38 PM

My quote:Indeed, in burger's reading you would be forced first to make a choice, and then also have the mechanical roll that possibly sends it to an image. It would seem to me that such a reading would obviously be too much for such a spell.

Burger's response: This (the first sentence) is incorrect.

How so? You have repeated said that the first step is to choose a target. You have also said that all the images of mirror image are possible targets. So you must therefore require the attacker to choose a target from the choices. Then, if they are lucky enough to choose the caster, the mechanics of the spell would require you to roll to see if it gets shifted to one of the images if they are using an attack. Is that not exactly the chain of events you say RAW requires?

AvatarVecna

2016-08-16, 10:54 PM

Except you have repeatedly mentioned that you don't care what I think, so why get upset if you think that I think you are stupid?

I repeatedly mentioned not caring one way or another for your opinion on my ruling, because everybody plays the game differently, so whatever. I don't appreciate being treated like I'm stupid for having an opinion that differs from someone else's, purely on the basis that my opinion is apparently less factual than there's despite how our respective opinions are on a thing where the entire point of this thread is that there aren't really any facts to base our opinions on, so we go with our feelings.

BurgerBeast

2016-08-16, 11:56 PM

georgie_leech

2016-08-17, 12:28 AM

The first step is to choose a target, but in the case of making an attack, you don't actually get to choose one. Mirror Image interrupts your choice and effectively chooses for you, via the d20 mechanic. In any other case, Mirror Image does not interrupt.

So, no. It is not.

Where's it say that? If the duplicates are independently targetable, clearly the part about attacks can only be applied if the attack is in fact targeted on the caster. After all, it wouldn't trigger if an enemy attacked a Silent Image next to the caster, right? The clause isn't overriding the general targeting rules, it acts in addition to them. 'If X, then Y,' not 'if X, Y instead and not X.'

Sabeta

2016-08-17, 12:54 AM

Hold Person
Choose a humanoid that you can see within range.
The target must succeed on a Wisdom saving throw
or be paralyzed for the duration. At the end of each of
its turns, the target can make another Wisdom saving
throw. On a success, the spell ends on the target.

I apologize if this has been brought up before, but there's four pages of huge walls that I don't feel like sifting through. Hold Person specifies "Humanoid", not "Humanoid Creature", so in theory it could work on anything with a humanoid shape; be that a person, a statue, or a Silent Image. As such, I would allow Hero to pick a target, and if he guesses wrong then the Spell Slot is consumed, but at least he reveals the image for what it is.

I'm not terribly in the mood for arguing with many of the people in this thread, so I won't respond beyond this unless I think it's necessary. People with poor reading comprehension who miss the point of my post entirely will just be ignored.

BurgerBeast

2016-08-17, 02:32 AM

Where's it say that?

It's hard to know which part of what I said you want evidence for, so I'll try to cover it all here.

The first step is to choose a target,... (PHB 194)

...but in the case of making an attack, you don't actually get to choose one. Mirror Image interrupts your choice and effectively chooses for you, via the d20 mechanic. In any other case, Mirror Image does not interrupt.
(this is an unavoidable consequence of the process of a player trying to describe one of four possible targets that are "shifting position so it’s impossible to track which image is real" to the DM. It's literally impossible.)
(It also gains support from the spell text for mirror image, which states "[e]ach time a creature targets you with an attack during the spell’s duration, roll a d20 to determine whether the attack instead targets one of your duplicates" [emphasis added]. Note that when a creature targets you, it is in step (1) of attacking [PHB 194], and will not move to step (2) until a target is chosen. So, when this section of the Mirror Image text resolves, it marks the end of step (1) and then the DM may proceed to step (2)).

If the duplicates are independently targetable, clearly the part about attacks can only be applied if the attack is in fact targeted on the caster. After all, it wouldn't trigger if an enemy attacked a Silent Image next to the caster, right?

Correct. But is literally impossible for a player to describe one of four identical targets, which are impossible to track, in a way that distinguishes it from the others, to a DM. (As a challenging exercise, you could try it.) So try as he may, the player cannot complete step (1) of making an attack (see PHB 194). The Mirror Image text created this problem, and its no coincidence that the Mirror Image texts provides a solution. It instructs the DM to employ the d20 mechanic to select a target. After this, the player and DM are free to continue on to step (2) of Making an Attack.

The clause isn't overriding the general targeting rules, it acts in addition to them. 'If X, then Y,' not 'if X, Y instead and not X.'

X and Y are both impossible without the mechanic. The character cannot select a target in any meaningful way, thus the target must be random. The mechanic handles this problem.

Hold Person
Choose a humanoid that you can see within range.
The target must succeed on a Wisdom saving throw
or be paralyzed for the duration. At the end of each of
its turns, the target can make another Wisdom saving
throw. On a success, the spell ends on the target.

I apologize if this has been brought up before, but there's four pages of huge walls that I don't feel like sifting through. Hold Person specifies "Humanoid", not "Humanoid Creature", so in theory it could work on anything with a humanoid shape; be that a person, a statue, or a Silent Image. As such, I would allow Hero to pick a target, and if he guesses wrong then the Spell Slot is consumed, but at least he reveals the image for what it is.

I'm not terribly in the mood for arguing with many of the people in this thread, so I won't respond beyond this unless I think it's necessary. People with poor reading comprehension who miss the point of my post entirely will just be ignored.

Humanoid is a technical game term, describing a particular type of creature (also a technical game term).

An illusion is not a creature.

(I'm having trouble finding a source on any of this but I am fairly confident in my answer. I did a brief scan of all three books and failed to find justification.)

smcmike

2016-08-17, 06:52 AM

(On the subject of elves responding to spells that only fool dwarves)

So, just the person. No duplicates?

Or the duplicates, too? (They just don't fool him.)

It doesn't matter.

You have enough context to limit the possibilities, but you have no context to assume what will happen. There is no way to say any individual event will happen with any degree of certainty without more information.

Similarly, I don't know whether Joe will pass his save or not.

He should not make any assumption.

Assumptions are always necessary. We live on assumptions.

(2) He should expect to Roll dice for free drinks (but not necessarily with the same conditions described specifically for ladies), because the sign said as much. (Note: if he is assay to marketing tricks, he might ask for clarification about whether he gets to roll free drinks, since he's not a lady - but as advertised he should get to, and if they didn't let him, he'd have good resonate complain).

No, he shouldn't. He has no complaint, because the sign clearly stated who actually gets to roll the dice for free drinks.

(3) He should expect the place to stay open "til the last Orc is dead!" (whatever that means).

This is known as "mere puffery." This is an actual legal term, and it means that you aren't justified in taking every word in a piece of advertising literally.

(4) He should expect to see ladies "get to roll [their] GIANT d20 for a chance at a free drink or other fabulous prizes!" any time "the lady orders a drink."

Yes.

Now and always. Feel free to go back to my very first post in the other thread. I have never wavered in my stance.

I wouldn't call it ambiguous. I'd say it makes it clear that there are no RAW to deal with how Mirror Image interacts with save spells.

I see what you mean now. You think the rules are extremely clear right up until the point where there are no rules - you think that it is clear RAW that one must somehow pick one of the four images to target with a spell, but there are no rules for how to do so, correct?

I don't know, and I don't intend to speculate. I only know what they wrote.

And yet, here we have an ambiguity - your method of interpretation can't explain this. Mine does.

But is literally impossible for a player to describe one of four identical targets, which are impossible to track, in a way that distinguishes it from the others, to a DM. So try as he may, the player cannot complete step (1) of making an attack (see PHB 194). The Mirror Image text created this problem, and its no coincidence that the Mirror Image texts provides a solution. It instructs the DM to employ the d20 mechanic to select a target. After this, the player and DM are free to continue on to step (2) of Making an Attack.

And yet you say Mirror Image created the SAME problem for targeted spells, but does not resolve it. This is odd, right?

Let's give the designers the benefit of the doubt. They didn't make a mistake. They intended the mechanic to define the fluff.

Mellack

2016-08-17, 08:58 AM

I
...but in the case of making an attack, you don't actually get to choose one. Mirror Image interrupts your choice and effectively chooses for you, via the d20 mechanic. In any other case, Mirror Image does not interrupt.
(this is an unavoidable consequence of the process of a player trying to describe one of four possible targets that are "shifting position so it’s impossible to track which image is real" to the DM. It's literally impossible.)
(It also gains support from the spell text for mirror image, which states "[e]ach time a creature targets you with an attack during the spell’s duration, roll a d20 to determine whether the attack instead targets one of your duplicates" [emphasis added]. Note that when a creature targets you, it is in step (1) of attacking [PHB 194], and will not move to step (2) until a target is chosen. So, when this section of the Mirror Image text resolves, it marks the end of step (1) and then the DM may proceed to step (2)).

Correct. But is literally impossible for a player to describe one of four identical targets, which are impossible to track, in a way that distinguishes it from the others, to a DM. (As a challenging exercise, you could try it.) So try as he may, the player cannot complete step (1) of making an attack (see PHB 194). The Mirror Image text created this problem, and its no coincidence that the Mirror Image texts provides a solution. It instructs the DM to employ the d20 mechanic to select a target. After this, the player and DM are free to continue on to step (2) of Making an Attack.

SO it is literally impossible to choose which image to attack, but then you demand anyone casting a non-attack spell choose which image to attack? You say the spell choses for you if it is an attack, but they somehow left out what happens otherwise? I fail to see how this double standard supports your interpretation.

Tanarii

2016-08-17, 09:21 AM

But is literally impossible for a player to describe one of four identical targets, which are impossible to track, in a way that distinguishes it from the others, to a DM. (As a challenging exercise, you could try it.) This is exactly what I meant when I said the spell doesn't create discrete separate individually targetable images.

What we know: It creates identical images. It creates images that move around. It creates them in such a way that no one can tell which is which. And most importantly that they exist in the same space.

There's a couple of ways to interpret that. You appear to have settled on one, and declared it to be the right way to interpret that. That they are non-overlapping separate instances of an image of the creature. (Simulationist thinking & ruling: your ruling)

Btw the biggest flaw in that is that there isn't enough space in the creatures space for multiples of the creature to fit.

But one alternative is that they swirl around, moving through each other and regularly overlapping each other and the creature. They aren't necessarily even constantly solid. Something between Blur and what you're envisioning. (Simulationist thinking & ruling: It can pull attacks off target slightly causing them to miss the creature, but non-attacks just affect the entire mess.)

Meanwhile, gamist ruling is just by the book, and figure out possible explanations after about what the hell Mirror Image looks like and why it works in game. I've provided one above. I'm sure there are others someone else can think of.
1) When you pick a target, you target the creature.
2) non-attacks just work on the target as usual.
3) attacks trigger the Mirror Image rules.

georgie_leech

2016-08-17, 10:52 AM

My copy of the PHB has it as 'whenever a creature targets you with an attack' as the first part of the attack clause. Thus, that part only has relevance if someone successfully targets them in the first place. By RAW, Mirror Images doesn't make enemies target randomly, it redirects attacks to the images some percent of the time. Supported by the phrase 'instead attacks.' The attack was directed at the caster, but the spell made it target one of the images instead.

So if the spell first requires you to select the correct target from among the duplicates, only then does the die roll to determine target come into play. Your reading of the spell gives it two layers of defence against being targeted by attacks, not one. Either attacks and spells can just select the caster, or neither can.

BurgerBeast

2016-08-17, 12:40 PM

It doesn't matter.

Of course it does. Answer the question and we'll see that it does. I call shenanigans on you.

Similarly, I don't know whether Joe will pass his save or not.

If by similar, you mean analogous, then no, not at all, but nice try.

Assumptions are always necessary. We live on assumptions.

I think you know that I was not saying "Never make assumptions." What I was saying was "it would be wrong to make any assertions about what will happen because there isn't sufficient reason to assert anything, in this case."

No, he shouldn't. He has no complaint, because the sign clearly stated who actually gets to roll the dice for free drinks.

The sign made the general claim: "Roll dice for free drinks! We won't go to bed til the last Orc is dead!" Before it got into specifics. There is nothing to link those claims to "ladies only."

This is known as "mere puffery." This is an actual legal term, and it means that you aren't justified in taking every word in a piece of advertising literally.

If you want to use "this is an ad" to justify why the first few sentences should not be taken literally, then my response is: the analogy doesn't hold because the spell description for Mirror Image in The Player's Handbook is not an advertisement.

Note that this does not mean the analogy cannot be used. It just means that if the analogy is used in a way that doesn't apply to the intended case, it's being improperly used.

I see what you mean now. You think the rules are extremely clear right up until the point where there are no rules - you think that it is clear RAW that one must somehow pick one of the four images to target with a spell, but there are no rules for how to do so, correct?

Yes. Precisely.

And yet, here we have an ambiguity - your method of interpretation can't explain this. Mine does.

This isn't ambiguity. A straightforward reading of the text shows that there are no rules to solve the situation. Your reading, which goes beyond a straightforward reading, does.

I have no problem with what you are doing. I have a problem with anyone claiming that it is RAW. That's all.

And yet you say Mirror Image created the SAME problem for targeted spells, but does not resolve it. This is odd, right?

Very.

Let's give the designers the benefit of the doubt. They didn't make a mistake. They intended the mechanic to define the fluff.

I have no problem with giving the writers the benefit of the doubt. I have no problem assuming they didn't make a mistake. But I have a problem identifying the mistake, so I won't presume to know what it was. I am okay with anyone making any assumptions they want, and ruling on that basis, as long as they admit that it is an interpretation, and not RAW.

SO it is literally impossible to choose which image to attack, but then you demand anyone casting a non-attack spell choose which image to attack?

I don't. The rules do. PHB 194.

You say the spell choses for you if it is an attack, but they somehow left out what happens otherwise?

Yes. This is the source of the entire discussion.

I fail to see how this double standard supports your interpretation.

I don't remember identifying a double standard, nor using it as support.

This is exactly what I meant when I said the spell doesn't create discrete separate individually targetable images.

Well, taken literally, these are very different. The spell does create "discrete separate individually targetable images" and it is possible for the character to attack one. The problem is that it's impossible for the player to unambiguously identify one of them to attack, even though the character can, because the character can see them.

What we know: It creates identical images. It creates images that move around. It creates them in such a way that no one can tell which is which. And most importantly that they exist in the same space.

Agreed.

There's a couple of ways to interpret that. You appear to have settled on one, and declared it to be the right way to interpret that.

No, I have not. I have only claimed that the other (matrix-like) explanation is not necessarily true.

That they are non-overlapping separate instances of an image of the creature. (Simulationist thinking & ruling: your ruling)

Not my ruling. I haven't ruled on the matter. I am open to any interpretation that doesn't violate the RAW.

Btw the biggest flaw in that is that there isn't enough space in the creatures space for multiples of the creature to fit.

Not relevant, since I don't advocate the position.

But one alternative is that they swirl around, moving through each other and regularly overlapping each other and the creature.

I'm cool with this. I always have been.

They aren't necessarily even constantly solid.

But the real Joe is "constantly solid," and the duplicates are impossible to distinguish, so they can't have a (visual) property that is different than he does. At this point you violate the spell description, unless I am misunderstanding you.

Something between Blur and what you're envisioning.

I can accept identical illusions that overlap the caster. No problem. Let's just ignore the "blurry" bit.

(Simulationist thinking & ruling: It can pull attacks off target slightly causing them to miss the creature, but non-attacks just affect the entire mess.)

Again, I reject any ruling that violates the text. I think this does. Conversely, I accept any ruling and any fluff that doesn't violate the text.

Meanwhile, gamist ruling is just by the book, and figure out possible explanations after about what the hell Mirror Image looks like and why it works in game. I've provided one above. I'm sure there are others someone else can think of.

And I'll accept any of them, so long as they rare reconcilable with the RAW.

1) When you pick a target, you target the creature.

This violates my understanding of what it means to target. Although not many people have clearly defined what it means to "target," it is clear form the discussions that many others have identified this problem, too.

2) non-attacks just work on the target as usual.

I think everyone can get behind this. This is the most fruitful path to pursue. And this is the path I've tried to stay on. So the question becomes "how does Hold Person usually work?" Or more generally "How dos targeting usually work?"

This is what I think needs to be discussed, and will lead to consistent rulings.

3) attacks trigger the Mirror Image rules.

If you mean paragraphs 2-4, then there is no doubt not disagreement here. If you mean to include paragraphs 1 and 5, then this is clearly not what is written - hence, a problem.

My copy of the PHB has it as 'whenever a creature targets you with an attack' as the first part of the attack clause. Thus, that part only has relevance if someone successfully targets them in the first place.

Yes, I alluded ages ago to the idea that "target" may be a source of the problem. What is meant by "target" influences how one reads the text.

By RAW, Mirror Images doesn't make enemies target randomly, it redirects attacks to the images some percent of the time. Supported by the phrase 'instead attacks.' The attack was directed at the caster, but the spell made it target one of the images instead.

Yes, I can break down why I think this hinges on the meaning of target, if you like. (I did so recently in one of the two threads).

So if the spell first requires you to select the correct target from among the duplicates, only then does the die roll to determine target come into play.

But it is literally impossible for the player to select a target and communicate which one of the four he selected to the DM. So the game freezes. And it's not possible to resume until the player achieves the impossible.

Your reading of the spell gives it two layers of defence against being targeted by attacks, not one.

Only using your stipulations on targeting. I don't use those. So my resolution system never supplies "two layers" of defence.

Either attacks and spells can just select the caster, or neither can.

A player making an attack must select a target (but can't).
A player casting a save-spell must select a target (but can't).

So, neither can.

Segev

2016-08-17, 01:07 PM

Yes, on the face of it, this is a problem. But it's not really a relevant problem because my whole contention is that there is no rule, not that my specific rule is right. So, in the case of the party-in-Canada, we are forced to use or judgment based on what we do know. I contend that this is what happens in the D&D example, too.
Except that there is a default case. I reiterate this even though you respond to it below for clarity. I'll elaborate after your response.

I disagree here. The spell is in effect. The spell has effects that do not depend on being attacked. I contend that the same thing happens as if the spell is in effect and the caster is not attacked.Er... I'm not sure this really makes sense.

We have default rules for what happens if you attack somebody without mirror image in place. Mirror image explicitly overrides them with the d20-determination mechanic, declaring that you don't get to pick which of the 4 Joes you want to attack, and instead determine randomly whether you are attacking the real Joe or an illusion.

We also have default rules for what happens if you cast hold person without mirror image in place: you choose a target, and they attempt a saving throw.

Mirror image does not explicitly override this. Therefore, the default rule remains that the caster (who we've been calling "Bob") chooses his target.

The entire first paragraph of Mirror Image is in effect for the entire duration of the spell, whether the caster is attacked or not. These effects are "Three illusory duplicates of yourself appear in your space. Until the spell ends, the duplicates move with you and mimic your actions, shifting position so it’s impossible to track which image is real." Just because the caster is not attacked is not sufficient reason to say that these effects are also not in effect. These effects do not depend on being attacked.Agreed. But that's why I make the comment later about where the argument that hold person ignores mirror image entirely fails.

I may fail there, too. But would you mind shedding some insight on where you think I've gone wrong, above?I was stating that the argument that hold person ignores mirror image fails. It fails because, as you correctly note, the first part of mirror image is still in effect: there are still 3 illusory duplicates of Joe that are indistinguishable from Joe himself.

Hold person still requires a target to be selected. Therefore, since there is no overriding mechanic declaring how you determine whether hold person hits Joe or an illusion, it is up to the DM to determine a way for Bob's player to select his target when Bob cannot know which of the 4 is the real Joe.

There is no evidence that I see that says the mirror images are individually targetable. The spell says they are constantly shifting and impossible to track. I would interpret that to mean you cannot target separately. This is supported in that if using an attack, it specifies how you have to roll to determine which that attack is actually at. Nothing is mentioned about having to first pick one of the four first. Indeed, in burger's reading you would be forced first to make a choice, and then also have the mechanical roll that possibly sends it to an image. It would seem to me that such a reading would obviously be too much for such a spell.The rules for "picking" are given in the overriding d20 mechanic. You're right, mirror image removes the choice from the player entirely once the player declares he wants to attack Joe. Even if the player wants to attack an image, the same mechanic is used.

There is nothing in the spell that says that Joe can be picked out from amongst the images, nor that he can be targeted specifically by simply saying "I target Joe." Everything about the spell's description suggests exactly the opposite. The trouble we're having is that we have an explicit choice-making mechanic specified in the event that Joe is "attacked," but lack such should he be targeted in any other way.

Nothing in the spell suggests that you can say "I target Joe with a single-target spell." To the contrary, it suggests that the shifting images are, in fact, distinct, since it gives a specific number of them and specifies that they vanish one by one as they're hit with attacks.

No. I understand your point, but I think Mirror Image works differently from your examples. Mirror image makes it hard to figure out exact where the target is, but doesn't create real problems in terms of identity - you are still targetting Joe, who you can see. You aren't sure which position he's in, but apparently that doesn't matter. No. Quite the contrary. You're thinking of blur, which makes it hard to tell where somebody is by blurring them and making them seem to fill more space than they do. This spell gives disadvantage to attackers. Mirror image, on the other hand, quite explicitly creates multiple visual copies of Joe when he casts it. The d20 mechanic provided even states clearly that it's used to determine if Joe is the one you actually attack, or if it's one of his images. The images even have a distinct AC from Joe's own, calculated differently than Joe's might be.

The spell is absolutely creating a problem of identifying whether you're targeting Joe or an illusion of him. It doesn't make it hard to tell where the target is; it makes it hard for you to tell whether your target is the target you think it is.

The first step is to choose a target, but in the case of making an attack, you don't actually get to choose one. Mirror Image interrupts your choice and effectively chooses for you, via the d20 mechanic. In any other case, Mirror Image does not interrupt. Precisely.

Where's it say that? If the duplicates are independently targetable, clearly the part about attacks can only be applied if the attack is in fact targeted on the caster. After all, it wouldn't trigger if an enemy attacked a Silent Image next to the caster, right? The clause isn't overriding the general targeting rules, it acts in addition to them. 'If X, then Y,' not 'if X, Y instead and not X.'It actually interrupts the standard targeting rules when you make an attack with that d20 mechanic. It changes them. It acknowledges, for attacks at least, that you WANT to attack Joe, and says "but to tell whether the 'Joe' you actually attack is the one you WANTED to attack, roll a d20."

For all other targeting situations, it provides no such acknowledgement. Instead, it simply has 4 copies of Joe which are explicitly indistinguishable. You (or Bob) must still pick one of them as the target of hold person. That mirror image provides no rules for how to determine whether Bob happened to pick the right 'Joe' or not doesn't change that Bob has to pick one.

SO it is literally impossible to choose which image to attack, but then you demand anyone casting a non-attack spell choose which image to attack? You say the spell choses for you if it is an attack, but they somehow left out what happens otherwise? I fail to see how this double standard supports your interpretation.No, it is literally impossible to tell which image is the real Joe and which images are fakes. You can, at any given freeze-frame moment of time, point to 4 individual images that constantly shuffle their positions. At the moment you target your spell (at whatever point in casting it you metaphorically "pull the trigger"), you specify one of the 4 images.

Because there is no defined means within the RAW that you must use to determine whether you picked the right image or not, the DM is free to invent whatever mechanism he chooses, though it should probably still have the correct probability of success. (In fact, the d20 version does NOT have the correct probability of success in the case of just 2 surviving illusions, and is thus inferior to many that a DM might invent.) The DM is even free to use the d20 method used for attacks. But he has no means commanded by the RAW.

All the RAW gives him for targeting spells on mirror image is that there are 4 indistinguishable potential "Joe"s. Only one of which is real. How he determines whether Bob picks the right one is up to him.

smcmike

2016-08-17, 01:16 PM

Of course it does. Answer the question and we'll see that it does. I call shenanigans on you.

Call shenanigans all you want. I say it doesn't matter, in the exact same way that I don't think it matters what color a fireball is.

The sign made the general claim: "Roll dice for free drinks! We won't go to bed til the last Orc is dead!" Before it got into specifics. There is nothing to link those claims to "ladies only.".

Sure there is. They are on the same flier, describing the same event, and it is common practice to start with one statement setting an outline, followed by further clarification precisely defining the original statement.

"I'll be working all weekend. I'll be tending bar on Saturday from noon until close, and I've got two shifts at the Hospital on Sunday."

Will the speaker be working on Saturday morning?

This isn't ambiguity. A straightforward reading of the text shows that there are no rules to solve the situation. Your reading, which goes beyond a straightforward reading, does.

I have no problem with what you are doing. I have a problem with anyone claiming that it is RAW. That's all.

Your argument is that the RAW are terminally broken, and cause the game to freeze until the DM steps in and fixes it. My argument is that they are not, and that you need to read the first section of the spell in light of the second section.

I have no problem with giving the writers the benefit of the doubt. I have no problem assuming they didn't make a mistake. But I have a problem identifying the mistake, so I won't presume to know what it was. I am okay with anyone making any assumptions they want, and ruling on that basis, as long as they admit that it is an interpretation, and not RAW.

If this is a debate about the definition of RAW, count me out. As far as I'm concerned, the RAW is literally the rules as they appear in the book, and any application of them whatsoever is an interpretation. So, yes, I am discussing an interpretation, by far the most reasonable interpretation, and one followed by the designers of the game.

Tanarii

2016-08-17, 01:25 PM

Well, taken literally, these are very different. The spell does create "discrete separate individually targetable images" and it is possible for the character to attack one. The problem is that it's impossible for the player to unambiguously identify one of them to attack, even though the character can, because the character can see them.We don't know that. The spell doesn't say that the creature still appears to look real. It doesn't even say that the illusions appear real. All it says is they are "illusionary duplicates", and that it is "impossible to track which which image is real".

No, I have not. I have only claimed that the other (matrix-like) explanation is not necessarily true.In that case, why are (were?) you insisting that you have to pick a target from among the images? There's no RAW reason to do that except a specific simulation-based interpretation of the Mirror Image non-resolution part of the text. And it's an interpretation that creates consistency contradictions.

Not my ruling. I haven't ruled on the matter. I am open to any interpretation that doesn't violate the RAW.Having to choose targets once, then again as part of the Mirror Image spell, certainly seems contradictory. Having to choose a target if not an attack, but resolve per a random roll if an attack, also seems contradictory.

But the real Joe is "constantly solid," and the duplicates are impossible to distinguish, so they can't have a (visual) property that is different than he does. At this point you violate the spell description, unless I am misunderstanding you.You do misunderstand, per my first response in this post. Nothing in the spell text says that the "real" Joe must continue to appear constantly solid and individually distinguishable any more than the images must.

Again, I reject any ruling that violates the text. I think this does. Conversely, I accept any ruling and any fluff that doesn't violate the text.
And I'll accept any of them, so long as they rare reconcilable with the RAW.
This violates my understanding of what it means to target. Although not many people have clearly defined what it means to "target," it is clear form the discussions that many others have identified this problem, too.Oh. Well in that case, that's the problem. Your understanding of what it means to target is the problem. :smalltongue:

I think everyone can get behind this. This is the most fruitful path to pursue. And this is the path I've tried to stay on. So the question becomes "how does Hold Person usually work?" Or more generally "How dos targeting usually work?"Easy. The player picks the thing to be affected. That thing is affected. If special rules apply, apply them

Silent Image = Creates a separate and discrete image not bound to any creature. If you target the image, resolve as usual. (Edit: Also, provides a mechanic for revealing the target was an illusion.)
Mirror Image = Binds spell to a creature. Doesn't specify it creates separate and discrete images that can be individually targeted. Instead provides a mechanic for targeting the creature with an attack. No special rule for targeting with something that's not an attack. Resolve thing that's not an attack as usual.

Segev

2016-08-17, 01:49 PM

Your argument is that the RAW are terminally broken, and cause the game to freeze until the DM steps in and fixes it.

If "the game freezes until the DM makes a ruling" is your definition of "terminally broken," then 5e is terminally broken by design. "Rulings, not rules," is a fundamental design paradigm of the system. A controversial one, but nevertheless it is one.

Tanarii

2016-08-17, 01:53 PM

If "the game freezes until the DM makes a ruling" is your definition of "terminally broken," then 5e is terminally broken by design. "Rulings, not rules," is a fundamental design paradigm of the system. A controversial one, but nevertheless it is one.
This is a misapplication of "rulings not rules", which means the system is designed to be flexible so that the DM can rule how he needs to in his specific campaign. See: the entire skill system.

It doesn't have to do with forcing the DM to rule on contradictory rules. (Edit: Or intentionally interpreting the rules in a fashion that causes a contradiction.)

Segev

2016-08-17, 01:54 PM

The player picks the thing to be affected. That thing is affected. If special rules apply, apply them.Agreed! However, I posit that the player must actually identify a target by a means his PC could do so, and not in a way that GENERATES information the PC does not have.

So, for instance, Bob cannot gather the suspects of a closed-room murder mystery in the same room, have them all sit down, cast hold person on "the one who committed this murder," then know that anybody who stands up when he asks the innocent to do so is, in fact, innocent of that murder. Because Bob doesn't know WHO the real murderer is, so the spell can't pick that target out for him. Bob has to KNOW which of the targets he sees he is targeting.

Silent Image = Creates a separate and discrete image not bound to any creature. If you target the image, resolve as usual. (Edit: Also, provides a mechanic for revealing the target was an illusion.)
Mirror Image = Binds spell to a creature. Doesn't specify it creates separate and discrete images that can be individually targeted. Instead provides a mechanic for targeting the creature with an attack. No special rule for targeting with something that's not an attack. Resolve thing that's not an attack as usual.Bolded for emphasis. You're wrong here. It explicitly specifies that it creates discrete images. If they were not discrete, there wouldn't be a defined number, they wouldn't be defined to act exactly as the real caster, and you couldn't hit individual ones with attacks.

smcmike

2016-08-17, 01:57 PM

Bolded for emphasis. You're wrong here. It explicitly specifies that it creates discrete images. If they were not discrete, there wouldn't be a defined number, they wouldn't be defined to act exactly as the real caster, and you couldn't hit individual ones with attacks.

You can hit individual ones with attacks, but you cannot purposefully target ones with attacks. You target Joe, and the text kicks in.

Segev

2016-08-17, 01:59 PM

You can hit individual ones with attacks, but you cannot purposefully target ones with attacks. You target Joe, and the text kicks in.

A case of a specific exception-rule overriding the general rules. The d20 mechanic is a resolution mechanic provided for attacks.

I would, if I gambled, bet money that the original writer didn't even think of "spells that are not attacks" when he wrote the spell. Neither to expressly exclude it (because if he had, he would have), nor to expressly include it (because if he had thought that hard about it, he'd have made sure it counted as an attack).

smcmike

2016-08-17, 02:04 PM

A case of a specific exception-rule overriding the general rules. The d20 mechanic is a resolution mechanic provided for attacks.

I would, if I gambled, bet money that the original writer didn't even think of "spells that are not attacks" when he wrote the spell. Neither to expressly exclude it (because if he had, he would have), nor to expressly include it (because if he had thought that hard about it, he'd have made sure it counted as an attack).

I'm glad someone is willing to discuss intent of the writer.

My theory is that the writer assumed that everyone would understand that the attack clause was the only real mechanical effect of the visual effects described in the first paragraph. This is, perhaps, an oversight. Yours is that they were simply mistaken.

smcmike

2016-08-17, 02:05 PM

Tanarii

2016-08-17, 02:06 PM

Agreed! However, I posit that the player must actually identify a target by a means his PC could do so, and not in a way that GENERATES information the PC does not have.

So, for instance, Bob cannot gather the suspects of a closed-room murder mystery in the same room, have them all sit down, cast hold person on "the one who committed this murder," then know that anybody who stands up when he asks the innocent to do so is, in fact, innocent of that murder. Because Bob doesn't know WHO the real murderer is, so the spell can't pick that target out for him. Bob has to KNOW which of the targets he sees he is targeting.Right. The PC targets 'that creature over there'.

If that creature has Mirror Image on it, you resolve as follows: If the action used is not an attack, you've targeted the creature; If it's an attack, you resolve per the Mirror Image spell. (Edit: added not to the first clause.)

If that creature doesn't have Mirror Image on it, you've targeted the creature.

You can visualize it how you need for this to work. If you want to say the PC targets "that creature over there with Mirror Image on it" go for it, it doesn't assume any visualization. If you want to say "that creature over there that appears as a swirling mess of images that I can't clearly see and get a fix on" do that instead.

Bolded for emphasis. You're wrong here. It explicitly specifies that it creates discrete images. If they were not discrete, there wouldn't be a defined number, they wouldn't be defined to act exactly as the real caster, and you couldn't hit individual ones with attacks.No. It doesn't. It doesn't specify there is a way to tell them apart. It DOES specifically say they shift and move around in a way so that you can't tell which one is real. This means they are not necessarily discrete, as in visually identifiable as separate from the caster and each other, images.
(Edit2: added necessarily to the last sentence)

Sabeta

2016-08-17, 02:09 PM

Silent Image = Creates a separate and discrete image not bound to any creature. If you target the image, resolve as usual. (Edit: Also, provides a mechanic for revealing the target was an illusion.)
Mirror Image = Binds spell to a creature. Doesn't specify it creates separate and discrete images that can be individually targeted. Instead provides a mechanic for targeting the creature with an attack. No special rule for targeting with something that's not an attack. Resolve thing that's not an attack as usual.

I more-or-less agree with this statement.

As I said earlier, Hold Person allows you to target Humanoids. I still stand by this because Burger completely failed to refute it. (Side Note: If he cites the book or SA regarding this I may be willing to change my mind, but it's unlikely) If Hero believes in the Illusion, then there would be nothing preventing him from making the attempt. Just like there's nothing stopping you from trying to cast Hold Person on a Young Dragon. The spell will fail, you lose the slot, and you move on. If Hero doesn't want to waste his slot he can either investigate it as normal, try throwing something at it, delay his spell until an ally confirms it, or raise the spell slot and just target both.

Mirror Image reacts differently. All of the images here still occupy the same 5-foot space, and therefore anything that is targeting one is capable of targeting all. A Dragon's Breath Attack, Fireball, Cone of Cold, Thunder Wave, and even Hold Person. It's not an "AoE" effect, but it does operate of a Saving Throw. You're not making an attack, you aren't being forced to choose a target, you simply trigger a Saving throw and hopefully land it. If you want another way to look at this, Hold Person simply doesn't interact with any of the Mirror Images because there's nothing to hold. It automatically attempts to grasp the person whose there; so just like how it fails against Silent Image it fails to grasp any of the Mirrors in Mirror Image.

BurgerBeast

2016-08-17, 02:49 PM

Except that there is a default case. I reiterate this even though you respond to it below for clarity. I'll elaborate after your response. (snip)

By default case, you meant just one target.

By default I case, I meant the case of four targets.

The reason I defaulted to this is because the case of one target is not necessarily informative. By this I mean that someone might overlook the fact that targeting is necessary when attacking one target when only one target is available. So they may conclude that targeting is not necessary, by default.

No, it is literally impossible to tell which image is the real Joe and which images are fakes. You can, at any given freeze-frame moment of time, point to 4 individual images that constantly shuffle their positions. At the moment you target your spell (at whatever point in casting it you metaphorically "pull the trigger"), you specify one of the 4 images.

Because there is no defined means within the RAW that you must use to determine whether you picked the right image or not, the DM is free to invent whatever mechanism he chooses, though it should probably still have the correct probability of success. (In fact, the d20 version does NOT have the correct probability of success in the case of just 2 surviving illusions, and is thus inferior to many that a DM might invent.) The DM is even free to use the d20 method used for attacks. But he has no means commanded by the RAW.

All the RAW gives him for targeting spells on mirror image is that there are 4 indistinguishable potential "Joe"s. Only one of which is real. How he determines whether Bob picks the right one is up to him.

Well explained. Thanks.

Call shenanigans all you want. I say it doesn't matter, in the exact same way that I don't think it matters what color a fireball is.

I do call shenanigans. And in light of this, this whole exchange:

Originally Posted by BurgerBeast
Well, first of all, the fact that it is posted on a telephone pole implies nothing. There may not even be a party. But I'm going to assume that the party is real and the ad is truthful, so the ad basically functions as RAW in this case.
Haha. This is wonderful. Thank you for assuming that my hypo was not a trick designed to make you think there was a party when there was not. The fact that it was posted on a telephone pole is fluff.

Seems like it was actually a bit of foreshadowing.

I'll tell you why it's relevant:

If you answer no, then you are ignoring the fact that the first paragraph applies generally. This bears directly on the MI discussion, because the analogy is that casting mirror image only creates illusory duplicates that are visible to attackers - and this gets so bizarrely convoluted and pre-emptive that it is obviously problematic.

If you answer no yes, then you have the problem of explaining how, upon seeing the a mirror-imaged Joe, an elf is able to attack as if he just sees Joe. How you resolve this points directly to how you resolve targeting. This is relevant to the rules of targeting with an attack (PHB 194), which are relevant to the discussion as a whole.

You have made small evasive maneuvers in the discussions before, but none that were blatant enough or significant enough to call out. This one is. You are within your rights to not answer the question, but I am within my rights to claim that you are not arguing in good faith.

Sure there is. They are on the same flier, describing the same event, and it is common practice to start with one statement setting an outline, followed by further clarification precisely defining the original statement.

You can infer the specific from the general all you want. I am saying it is an error of logic. If I want to know what a sentence means, I read the sentence. If context can provide clues, great. That doesn't mean that context is always right.

"I'll be working all weekend. I'll be tending bar on Saturday from noon until close, and I've got two shifts at the Hospital on Sunday."

Will the speaker be working on Saturday morning?

The answer is: I don't know. Maybe the speaker works also works at the gas station and works Saturday morning at the gas station. You don't get to say that he will not work Saturday morning just because he said nothing about Saturday morning. You don't get to say that he doesn't work Saturday morning at the bar or the Hospital either, just because he didn't say so. You might happen to know that the bar is closed on Saturday mornings, which would allow you to conclude he is not working the bar Saturday morning, but that's precisely because you have more information.

I had this exact conversation with a friend yesterday.

If someone says "I will take an umbrella if it rains."
It is an error of logic to conclude that "He will not take an umbrella if it is sunny."
This is a common fallacy in Logic 101 courses worldwide.

Your argument is that the RAW are terminally broken, and cause the game to freeze until the DM steps in and fixes it. My argument is that they are not, and that you need to read the first section of the spell in light of the second section.

No, my argument is that the rules are not terminally broken. My argument is that the rules of MI exist to fix what would otherwise be a broken circumstance.

If this is a debate about the definition of RAW, count me out. As far as I'm concerned, the RAW is literally the rules as they appear in the book, and any application of them whatsoever is an interpretation.

If those are your definitions, then this isn't a debate. We agree.

So, yes, I am discussing an interpretation, by far the most reasonable interpretation, and one followed by the designers of the game.

No. you do not get to claim that your interpretation is followed by the designer of the game. Short of new information, you have no way to make that claim. You can claim that you and the designers reach the same conclusion.

Quote Originally Posted by BurgerBeast
Well, taken literally, these are very different. The spell does create "discrete separate individually targetable images" and it is possible for the character to attack one. The problem is that it's impossible for the player to unambiguously identify one of them to attack, even though the character can, because the character can see them.We don't know that. The spell doesn't say that the creature still appears to look real. It doesn't even say that the illusions appear real. All it says is they are "illusionary duplicates", and that it is "impossible to track which which image is real".

Even if Joe doesn't look real, and the illusions don't look real, it is "impossible to track which image is real," this refutes nothing of what I said.

(Although I should make it clear that I don't think the spell alters Joe's appearance. I'm cool with it, as long as it's consistent with the rest.)

In that case, why are (were?) you insisting that you have to pick a target from among the images? There's no RAW reason to do that except a specific simulation-based interpretation of the Mirror Image non-resolution part of the text.

PHB 194. There is no simulation-based interpretation going on.

...And it's an interpretation that creates consistency contradictions.

Not that I can see.

Having to choose targets once, then again as part of the Mirror Image spell, certainly seems contradictory.

Well it's not contradictory. If you mean contradictory to the text, then you're right, it is. This is why I do not advocate choosing targets twice. I advocate chosen targets once.

...Having to choose a target if not an attack, but resolve per a random roll if an attack, also seems contradictory.

It may seem like a lot of things. It's still RAW.

You do misunderstand, per my first response in this post. Nothing in the spell text says that the "real" Joe must continue to appear constantly solid and individually distinguishable any more than the images must.

Okay, we'll get rid of constantly solid, then. He is distinguishable because he can be hit when the duplicates are not. That is a distinction. He is distinguishable because ehe is not the same thing as a duplicate. That is a distinction. He is distinguishable because ehe has a different AC (in most cases). He is distinguishable because he is a creature and they are illusions.

Oh. Well in that case, that's the problem. Your understanding of what it means to target is the problem. :small tongue:

In all seriousness, this is probably a conversation worth having.

Easy. The player picks the thing to be affected. That thing is affected. If special rules apply, apply them

Silent Image = Creates a separate and discrete image not bound to any creature. If you target the image, resolve as usual. (Edit: Also, provides a mechanic for revealing the target was an illusion.)
Mirror Image = Binds spell to a creature. Doesn't specify it creates separate and discrete images that can be individually targeted. Instead provides a mechanic for targeting the creature with an attack. No special rule for targeting with something that's not an attack. Resolve thing that's not an attack as usual.

Mirror Image is specify separate and discrete duplicates that can be targeted. I'm not sure how you can claim otherwise. It provides a special mechanic for targeting the creature or a duplicate with an attack.

Is it possible to target both a duplicate and the real Joe with one attack? No.
Why not? Because they are two different things. They are discrete.
Is it possible to hit one of them but not the other? Yes.
Why? Because they are individually targetable.

georgie_leech

2016-08-17, 02:53 PM

BurgerBeast

2016-08-17, 03:12 PM

I'm glad someone is willing to discuss intent of the writer.

My theory is that the writer assumed that everyone would understand that the attack clause was the only real mechanical effect of the visual effects described in the first paragraph. This is, perhaps, an oversight. the oversight you suggest is much larger.

Interestingly, not only am I willing to answer in terms of intent, I gave more-or-less the same answer in one of the two threads.

Right. The PC targets 'that creature over there'.

And then the DM has to interpret what that means, in light of there being four potential targets that are not distinguishable. According to PHB 194, you cannot proceed until a target is selected.

If that creature has Mirror Image on it, you resolve as follows: If the action used is not an attack, you've targeted the creature; If it's an attack, you resolve per the Mirror Image spell. (Edit: added not to the first clause.)

If that creature doesn't have Mirror Image on it, you've targeted the creature.

This goes back to he discussion about defaults. There are four potential targets. When there are four potential targets, "you haven't targeted the creature" by saying "that creature over there." The DM must ask you to specify which one.

You can visualize it how you need for this to work. If you want to say the PC targets "that creature over there with Mirror Image on it" go for it, it doesn't assume any visualization. If you want to say "that creature over there that appears as a swirling mess of images that I can't clearly see and get a fix on" do that instead.

Not if you care about RAW. You can't ignore that there are four potential targets, RAW. You can't arbitrarily pick one of the four for the player, RAW. So you shouldn't just violate RAW and "visualize it however you need for to work."

No. It doesn't. It doesn't specify there is a way to tell them apart.

None said it does. We said that they are discrete and cannot be told apart.

It DOES specifically say they shift and move around in a way so that you can't tell which one is real. This means they are not necessarily discrete, as in visually identifiable as separate from the caster and each other, images.
(Edit2: added necessarily to the last sentence)

I don't think you are using discrete in the same way that we are. You can't four of something if it is not discrete, as Segev said. Even if four illusions overlap each other, so you can't tell if it is one illusion or four, this doesn;t change the truth of the situation. There are four, not one. They are discrete, not continuous. They overlap, they don't merge together.

Mirror Image reacts differently. All of the images here still occupy the same 5-foot space, and therefore anything that is targeting one is capable of targeting all.

So, is [I]Hold Monster (cast at 5th-level to keep the example consistent) capable of targeting four frogs because they are all in the same 5-foot space?

A Dragon's Breath Attack, Fireball, Cone of Cold, Thunder Wave, and even Hold Person.

The spell description says "Choose a humanoid..." That's one. The target is one humanoid.

It's not an "AoE" effect, but it does operate of a Saving Throw. You're not making an attack, you aren't being forced to choose a target, you simply trigger a Saving throw and hopefully land it.

The text of hold person requires you to choose a target.

If you want another way to look at this, Hold Person simply doesn't interact with any of the Mirror Images because there's nothing to hold. It automatically attempts to grasp the person whose there; so just like how it fails against Silent Image it fails to grasp any of the Mirrors in Mirror Image.

This is the auto-selecting problem which has been discussed at length.

Tanarii

2016-08-17, 03:20 PM

Even if Joe doesn't look real, and the illusions don't look real, it is "impossible to track which image is real," this refutes nothing of what I said.

(Although I should make it clear that I don't think the spell alters Joe's appearance. I'm cool with it, as long as it's consistent with the rest.)Okay. So let's leave behind the visualitation for a second.

Oh no wait, we can't. Because your interpretation requires a specific visualization of the spell as individually identifiable and intentionally targetable (ie discrete) images.

PHB 194. There is no simulation-based interpretation going on.
<snip>
It may seem like a lot of things. It's still RAW.Shenanigans. No, page 194 nor RAW require reading & interpreting Mirror Image as individually identifiable & intentionally targetable (discrete) images. you don't get to claim RAW for your Rules-as-interpreted due to your visualization of the spell. Because that's what you're doing, despite claiming you're not. How do I know? Because next you claim.

Okay, we'll get rid of constantly solid, then. He is distinguishable because he can be hit when the duplicates are not. That is a distinction. He is distinguishable because ehe is not the same thing as a duplicate. That is a distinction. He is distinguishable because ehe has a different AC (in most cases). He is distinguishable because he is a creature and they are illusions.

<snip>

Mirror Image is specify separate and discrete duplicates that can be targeted. I'm not sure how you can claim otherwise.
And I'll just quote my response to Segev unthread:

No. It doesn't. It doesn't specify there is a way to tell them apart. It DOES specifically say they shift and move around in a way so that you can't tell which one is real. This means they are not necessarily discrete, as in visually identifiable as separate from the caster and each other, images.
(Edit2: added necessarily to the last sentence)

Because they are two different things. They are discrete.
Because they are individually targetable.RAW doesn't require a Why. It tells us What. Your Why answers are not in the RAW and they are your personal visualization and interpretation of how the spell works.

BurgerBeast

2016-08-17, 03:23 PM

Let's throw some more logs on the fire. A Ranger with Whirlwind uses it next to someone with Mirror Image up. Do they automatically attack all three images plus the actual creature, or do they make an attack that has a chance of being redirected to the images as normal?

The Ranger's Whirlwind Attack opens a whole ugly can of worms. I had a pretty good discussion about it yesterday. You can interpret it it in one of two ways: since the Ranger "sees" four creatures, he can make an attack at each one; or since the ranger truly only sees one creature, he can make one attack and it resolves as normal.

Unfortunately for me, I tend to see it the first way. Assuming this, then: Problems do not arise if the ranger, by chance, targets and hits each duplicate before he targets the real Joe. Any other case causes problems because:

(1) If he targets a duplicate and misses, he's not technically supposed to be allowed to attack it again, even though the random determination may bring it up.
(2) If he targets the real Joe, he's not supposed to be allowed to target him again, even though the random mechanic may bring it up.

In this case, I would house rule that the ranger gets to make 1 + n attacks, where n is the number of duplicates at the start of the Whirlwind Attack. I know that this goes against RAW, but I think just about any decision does.

Segev

2016-08-17, 03:27 PM

I'm glad someone is willing to discuss intent of the writer.

My theory is that the writer assumed that everyone would understand that the attack clause was the only real mechanical effect of the visual effects described in the first paragraph. This is, perhaps, an oversight. the oversight you suggest is much larger.Not really. "Didn't even consider spells-that-weren't attacks" can encompass "assumed that any hostile effect targeting the caster is an attack." This is especially true if rumors of how little communication there was between authors of various sections of the book is true; the definition of "attack" being restricted so sharply may not have been known to the writer of the spell.

In any event, the "visual effects" are, in fact, part of the RAW, unless the spell says otherwise. It's silly to assume that there's something magical (pun intended) about "an attack" vs. "an action that is very like an attack but doesn't happen to roll a d20 to hit" that makes you suddenly able to distinguish which of the images is the real caster in the latter case, but not the former.

Right. The PC targets 'that creature over there'.

If that creature has Mirror Image on it, you resolve as follows: If the action used is not an attack, you've targeted the creature; If it's an attack, you resolve per the Mirror Image spell. (Edit: added not to the first clause.)

If that creature doesn't have Mirror Image on it, you've targeted the creature.Not according to the RAW. The RAW don't include a clause that lets you automatically identify which if the indistinguishable images is the real caster just because you're not using an attack. The default rule for casting a spell is that you must specify a target. You can't tell which image is the real caster. You must pick ONE. You don't get to say "the real one" any more than Bob gets to target "the real murderer" in the closed room murder mystery.

It doesn't specify there is a way to tell them apart. It DOES specifically say they shift and move around in a way so that you can't tell which one is real. This means they are not necessarily discrete, as in visually identifiable as separate from the caster and each other, images.
(Edit2: added necessarily to the last sentence)Except that they ARE discrete or they would NOT be individual images capable of being counted.

Since they CAN be counted and you CAN (and in fact do) hit only one at a time when you attack, they MUST be discrete. There is no other option without redefining "discrete" to mean something other than what it does.

As I said earlier, Hold Person allows you to target Humanoids. I still stand by this because Burger completely failed to refute it.No, it allows you to affect humanoids. Or are you claiming that having hold person prepared allows you to freely determine, without an action or spell slot, whether a given thing is really a humanoid or merely looks like one? Because that's the consequence of claiming you can't even target a non-humanoid with that spell, to the point that you can't even attempt the casting and have it fail because of an invalid target.

This is also the only way that interpretation would let you auto-target the real caster: you'd try targeting images until you successfully were allowed to do so, consuming no actions and no spell slots until you were allowed to because you chose the right target.

The proper interpretation, I contend, is that if you try to use hold person on a non-humanoid, you do in fact cast the spell (using the action and spell slot), but it does nothing for the same reason that throwing a fireball at a fire elemental does nothing.

Let's throw some more logs on the fire. A Ranger with Whirlwind uses it next to someone with Mirror Image up. Do they automatically attack all three images plus the actual creature, or do they make an attack that has a chance of being redirected to the images as normal?

Now that's an interesting question! It's definitely a DM's call, but I'd have it roll against each image and the real caster's AC. Good way to clear out a mirror image spell all at once if you can hit the mage's images! (I base this on the fact that upcasting hold person to hit 4 targets would let you hit the caster by virtue of targeting all 4 of "him.")

Tanarii

2016-08-17, 03:42 PM

Not according to the RAW. The RAW don't include a clause that lets you automatically identify which if the indistinguishable images is the real caster just because you're not using an attack. The default rule for casting a spell is that you must specify a target. You can't tell which image is the real caster. You must pick ONE. You don't get to say "the real one" any more than Bob gets to target "the real murderer" in the closed room murder mystery.Yes, according to RAW. The spell's RAW doesnt specify that the duplicates are in any way targetable individually, only as part of the redirect mechanic. In fact, the spell requires you to target the creature first, then redirects. So even with a melee attack, you must first be able to target 'that creature with Mirror Image on it'.

Except that they ARE discrete or they would NOT be individual images capable of being counted.

Since they CAN be counted and you CAN (and in fact do) hit only one at a time when you attack, they MUST be discrete. There is no other option without redefining "discrete" to mean something other than what it does.
individually separate and distinct.
"speech sounds are produced as a continuous sound signal rather than discrete units"
synonyms: separate, distinct, individual, detached, unattached, disconnected, discontinuous, disjunct, disjoined
"discrete units of sound"

None of those are required by RAW mirror image being able to count sub-portions of the spell, ie duplicate images. It can just as easily be considered a single and non-discrete visually inseparable (and confusing) single thing, that redirects targeting attacks to a then counted number of 'duplicate image' sub-portions of the single entity.

BurgerBeast

2016-08-17, 03:43 PM

Okay, let's slow down. Between you and me, nobody is intentionally arguing in bad faith. Let's go through this, step by step.

Okay. So let's leave behind the visualitation for a second.

Oh no wait, we can't. Because your interpretation requires a specific visualization of the spell as individually identifiable and intentionally targetable (ie discrete) images.

My interpretation requires a visualization that fits RAW. It does not require a specific visualization. I will adopt your visualization and continue to discuss this in terms of your visualization, as long as your visualization does not violate RAW.

Shenanigans. No, page 194 nor RAW require reading & interpreting Mirror Image as individually identifiable & intentionally targetable (discrete) images.

194 does not (and I never claimed it did). Also, your claim has changed. If you define discrete as intentionally targetable, then we need to get into that. Discrete does not carry that implication, for me. For me discrete means "not continuous" or "capable of being considered one thing." It is possible to be discrete and not be "intentionally targetable." Also, the phrase "intentionally targetable" raises problems in the particular discussion, because, if we enter the fiction, the character has the ability to swing his sword at one of the images. However, outside the fiction, there is no way for the player to indicate one specific image to the DM.

...you don't get to claim RAW for your Rules-as-interpreted due to your visualization of the spell. Because that's what you're doing, despite claiming you're not. How do I know? Because next you claim.

I contend that I am not, but we'll see what follows.

And I'll just quote my response to Segev unthread:

Quote Originally Posted by Tanarii
No. It doesn't. It doesn't specify there is a way to tell them apart. It DOES specifically say they shift and move around in a way so that you can't tell which one is real. This means they are not necessarily discrete, as in visually identifiable as separate from the caster and each other, images.
(Edit2: added necessarily to the last sentence)

Quote Originally Posted by BurgerBeast
Because they are two different things. They are discrete.
Because they are individually targetable.

RAW doesn't require a Why. It tells us What. Your Why answers are not in the RAW and they are your personal visualization and interpretation of how the spell works.

Here's my full quote (why did you change it and not indicate that you did?):

Is it possible to target both a duplicate and the real Joe with one attack? No.
Why not? Because they are two different things. They are discrete.
Is it possible to hit one of them but not the other? Yes.
Why? Because they are individually targetable.

So I said:
(1) they are, RAW, discrete, because you cannot target two at once.
(2) they said they are, RAW, individually targetable because you can hit one of them and not the other.

These answers do come from the RAW. You cannot target more than one creature with an attack unless a specific rule says so. So (1) comes form the attacking rules. The Mirror Image text shows you how to resolve the one target that is struck by an attack, and never are two targets struck. So (2) comes form the MI rules.

Other than this, you are right to say that RAW doesn't require a Why. However, any Why that you provide must comply with the What that is given in RAW.

If you contend, for example, that the duplicates are not discrete, then that comes into conflict with their behaviour as described RAW.

Segev

2016-08-17, 03:58 PM

Yes, according to RAW. The spell's RAW doesnt specify that the duplicates are in any way targetable individually, only as part of the redirect mechanic. In fact, the spell requires you to target the creature first, then redirects. So even with a melee attack, you must first be able to target 'that creature with Mirror Image on it'.No, it requires you to attack the creature, in which case it specifically overrides the normal need to pick your target specifically and replaces it with the d20 resolution mechanic.

individually separate and distinct.
"speech sounds are produced as a continuous sound signal rather than discrete units"
synonyms: separate, distinct, individual, detached, unattached, disconnected, discontinuous, disjunct, disjoined
"discrete units of sound"

None of those are required by RAW mirror image being able to count sub-portions of the spell, ie duplicate images. It can just as easily be considered a single and non-discrete visually inseparable (and confusing) single thing, that redirects targeting attacks to a then counted number of 'duplicate image' sub-portions of the single entity.
If they are not individually separate and distinct, then you cannot remove just one without removing some or all of the others, in whole or in part.

Regardless of how you choose to interpret it, you don't have one creature you can point to and say, "That's the creature I want to target." If I have four playing cards and constantly shuffle them about, but only one of them is the Ace of Spades, you don't get to tell me "I pick the Ace of Spades" when I've asked you to find it while I'm shuffling them about. You have to point to one of them, not to the cluster.

They're not "part of" the caster. They're distinct. Indistinguishable, but distinct. Otherwise there wouldn't be 3 images created; there'd be an illusion covering him. The entire point of the spell is that you can't tell which of the 4 is the caster. You don't get to bypass that by saying "I target the caster" any more than Bob gets to bypass the murder mystery by saying "I cast hold person on the person who committed this murder."

BurgerBeast

2016-08-17, 04:04 PM

Yes, according to RAW. The spell's RAW doesnt specify that the duplicates are in any way targetable individually, only as part of the redirect mechanic. In fact, the spell requires you to target the creature first, then redirects. So even with a melee attack, you must first be able to target 'that creature with Mirror Image on it'.

Okay. In the multitude of lines of argument I forgot some of your assumptions.

"The spell's RAW doesnt specify that the duplicates are in any way targetable individually" - this is true, but the spell does specify that duplicates are in fact there in the square.

The question of whether they are targetable can only be answered by the question whether an "illusory duplicate" is targetable, generally. There is no RAW on this, you are right. So, if you contend that the duplicates are not targetable at all, then it follows that they are not targetable by hold person. If you contend that they are targetable, generally, but not by hold person because they are not humanoids, then you have RAW on your side. But if you contend that they are not targetable because of the text MI, you are mistaken. The text of MI is silent on the ability of save-spells to target Joe.

"So even with a melee attack, you must first be able to target 'that creature with Mirror Image on it'." - this place player's declaration can only be resolved on the prior assumptions about targeting. If you hold those assumptions about targeting, then I concede this point and you are correct.

But then we have to discuss whether you apply these same targeting guidelines in other circumstances, because if you don't, it's inconsistent, and if you do, it presents bigger problems.

individually separate and distinct.
"speech sounds are produced as a continuous sound signal rather than discrete units"
synonyms: separate, distinct, individual, detached, unattached, disconnected, discontinuous, disjunct, disjoined
"discrete units of sound"

None of those are required by RAW mirror image being able to count sub-portions of the spell, ie duplicate images. It can just as easily be considered a single and non-discrete visually inseparable (and confusing) single thing, that redirects targeting attacks to a then counted number of 'duplicate image' sub-portions of the single entity.

You seem to be confusing whether particular individual is capable of discerning that something is discrete with whether something is, in principle, discrete. If you can say that there are four of something, that something is discrete. Period. It may be true that someone can't tell - they may think it is one thing - but that does not make it continuous, in principle.

The very fact that the spell says there are three duplicates means that,in principle, they are discrete. This is supported by the text that redirects attacks to one of them. No one is contesting that they may appear continuous in your fluff. We are contesting that your fluff can change them so that they are not, in principle, discrete.

Tanarii

2016-08-17, 04:31 PM

Okay, let's slow down. Between you and me, nobody is intentionally arguing in bad faith. Let's go through this, step by step.Never thought you were. You don't seem like the kind of person to argue in bad faith. You seem like the kind of person who wants to niggle away at something until it's resolved, especially when it doesn't make any sense to you. :smallwink:

My interpretation requires a visualization that fits RAW. It does not require a specific visualization. I will adopt your visualization and continue to discuss this in terms of your visualization, as long as your visualization does not violate RAW.Perhaps visualization is a bad word. I mean: your specific interpretation of the spell requires reading the RAW (ie the actual text of the rule), which is the "What" of the spell, and deciding that a single specific underlying "Why", is the correct answer. That's a visualization of how the spell works.

194 does not (and I never claimed it did). Also, your claim has changed. If you define discrete as intentionally targetable, then we need to get into that. Discrete does not carry that implication, for me. For me discrete means "not continuous" or "capable of being considered one thing." It is possible to be discrete and not be "intentionally targetable." Also, the phrase "intentionally targetable" raises problems in the particular discussion, because, if we enter the fiction, the character has the ability to swing his sword at one of the images. However, outside the fiction, there is no way for the player to indicate one specific image to the DM.Sorry, in attempting to be clear, I was more confusing. Grabbing the google definition suffices for me: "individually separate and distinct." Mirror Image doesn't require that the duplicates must be considered separate and distinct as opposed to a portion of the single Mirror Image 'entity' for the spell to work as written.

Here's my full quote (why did you change it and not indicate that you did?):Was on a tablet briefly, and trying to cut down to the meat of the post, the "Why" portions. My bad.

So I said:
(1) they are, RAW, discrete, because you cannot target two at once.
(2) they said they are, RAW, individually targetable because you can hit one of them and not the other.
(1a) RAW, you cannot target two at once. You can't intentionally target one image. You must target the creature. The spell gives a resolution process to determine if it redirects attacks to duplicates. Since it does not redirect non-attacks, and (per the spell) you still must target the creature, non-attacks affect the creature as normal.
(1b) RAW, the spell does not give you a mechanic to redirect to two simultaneously. You have decided to interpret that to mean they are discrete.
(2) RAW, they are NOT individually targetable. See 1a above. You target the creature. It then redirects to the duplicates.

Edit:
(2b) Rereading my post, I see I misunderstood your second point. I assume you mean since a redirect only (potentially) eliminates one target, and not another target, that means they are discrete. Sure, they may be so in context off resolving how many times the spell works. But that doesn't mean they are a separate entity from 'caster with Mirror Image on him' . Or even just 'Mirror Image'.

smcmike

2016-08-17, 04:32 PM

I'll tell you why it's relevant:

Thanks! This is good practice, generally, when trying to convince someone of something.

If you answer no, then you are ignoring the fact that the first paragraph applies generally. This bears directly on the MI discussion, because the analogy is that casting mirror image only creates illusory duplicates that are visible to attackers - and this gets so bizarrely convoluted and pre-emptive that it is obviously problematic.

If you answer no yes, then you have the problem of explaining how, upon seeing the a mirror-imaged Joe, an elf is able to attack as if he just sees Joe. How you resolve this points directly to how you resolve targeting. This is relevant to the rules of targeting with an attack (PHB 194), which are relevant to the discussion as a whole.

The reason it isn't relevant is that it just leads back to the same issues we've been discussing, and leads to me restating my position. The Elf can target Joe as if he just sees Joe because the mechanical part of the spell text effects Dwarves.

For what it's worth, I agree - it would probably be too weird to say he can't see them. So, he must be able to see them, but for some reason that doesn't interfere with his targetting process. The reason isn't relevant, though.

You have made small evasive maneuvers in the discussions before, but none that were blatant enough or significant enough to call out. This one is. You are within your rights to not answer the question, but I am within my rights to claim that you are not arguing in good faith.

And you would be wrong. Let's not go back to accusing each other of things.

You can infer the specific from the general all you want. I am saying it is an error of logic. If I want to know what a sentence means, I read the sentence. If context can provide clues, great. That doesn't mean that context is always right.

I'm not even sure what "context is always right" is supposed to mean. Maybe give me some context. Regardless, context is very frequently determinative of meaning. Tone, too.

The answer is: I don't know. Maybe the speaker works also works at the gas station and works Saturday morning at the gas station. You don't get to say that he will not work Saturday morning just because he said nothing about Saturday morning. You don't get to say that he doesn't work Saturday morning at the bar or the Hospital either, just because he didn't say so. You might happen to know that the bar is closed on Saturday mornings, which would allow you to conclude he is not working the bar Saturday morning, but that's precisely because you have more information.

I agree that you can't completely rule out the possibility that the speaker is working Saturday morning. The statement is ambiguous. Here are some possible interpretations;

1. The speaker is working every hour of the weekend, but for some reason chose to highlight only portions of that work. This is the most literal reading, which I see as equivalent to your position about Mirror Image.
2. The speaker's description of her shifts at the bar and the hospital constituted an explanation or definition of her original statement, and she is not working beyond the shifts specified. This is equivalent to my position.
3. Something in between: it is possible that she didn't mean the first statement literally, and also that she didn't mean the list to be exhaustive.

If I had to guess, without knowing anything about the person, and even assuming that some people do, in fact, work 48 hours in a weekend (at a hospital, perhaps), I would go with interpretation (2). It's ambiguous, but it is very common to say one thing and then say another thing to define the first.

If someone says "I will take an umbrella if it rains."
It is an error of logic to conclude that "He will not take an umbrella if it is sunny."
This is a common fallacy in Logic 101 courses worldwide.

The reason it is a common fallacy is that people do not speak in formal logic, and in every day use this is a reasonable interpretation of the sentence. 5e attempts to write in everyday language.

No, my argument is that the rules are not terminally broken. My argument is that the rules of MI exist to fix what would otherwise be a broken circumstance.

And yet they do not fix the circumstance in which the caster is targeted with Hold Person. Hence, you are saying they are broken.

No. you do not get to claim that your interpretation is followed by the designer of the game. Short of new information, you have no way to make that claim. You can claim that you and the designers reach the same conclusion.

Interpretation, conclusion, whatever.

smcmike

2016-08-17, 04:41 PM

In any event, the "visual effects" are, in fact, part of the RAW, unless the spell says otherwise. It's silly to assume that there's something magical (pun intended) about "an attack" vs. "an action that is very like an attack but doesn't happen to roll a d20 to hit" that makes you suddenly able to distinguish which of the images is the real caster in the latter case, but not the former.

I agree. The action that is like an attack but doesn't need to roll a d20 to hit does not let you distinguish which one is real. It just doesn't care. It's not physically precise - which is why it doesn't need you to roll an action to hit.

Sabeta

2016-08-17, 04:52 PM

SNIP

Read past the first line of my post. I've said twice (three times now) that you are allowed to use your slot to target the illusion, and the spell would fail. You lose a slot, bit identify the image. Likewise, you could try to cast the spell on a monster, but the spell would be similarly wasted.

Segev

2016-08-17, 05:22 PM

I agree. The action that is like an attack but doesn't need to roll a d20 to hit does not let you distinguish which one is real. It just doesn't care. It's not physically precise - which is why it doesn't need you to roll an action to hit.It still, however, requires you to choose at whom you're targeting it. It is not so "physically precise" that it can select the target you WANTED even if YOU are unable to distinguish said target.

Again: see Bob casting [I]hold person to solve a murder mystery.

Read past the first line of my post. I've said twice (three times now) that you are allowed to use your slot to target the illusion, and the spell would fail. You lose a slot, bit identify the image. Likewise, you could try to cast the spell on a monster, but the spell would be similarly wasted.

Ah. I have no problem with that. Though in the case of mirror image, you'd sadly lose track of which image you cast it on, so it doesn't net you anything other than a lost slot and action. (Because mirror image specifically doesn't let you keep track of which one is which.)

Use it like that on a silent image of Joe, though, and... well, I don't think you technically should be able to say, "Oh, that's an illusion," but you can tell that your spell failed. (At least, I think you can; do you automatically know if something failed or succeeded a save? Or would you be unable to determine if the silent image of Joe made a save vs. was an invalid target?)

If you rule that it auto-reveals the illusion, that's a house rule, I think, but it certainly isn't overpowered. You spent an action and a spell slot to get what you could have with an action and an Investigation check.

Sabeta

2016-08-17, 05:52 PM

In the hypothetical I assumed Hero knew one of the targets was an Illusion, since he could have the option to Investigate to reveal it's identity. Likewise, interacting with it reveals it's nature, and so it stands to reason that your spell failing would offer some feedback as to its nature.

You're right about Mirror Image. Although you could hold one, you would end up with a paralyzed blur of four guys. Unlike the displacer beast it says nothing about what happens if you're paralyzed or your movement becomes 0.

smcmike

2016-08-17, 07:25 PM

It still, however, requires you to choose at whom you're targeting it. It is not so "physically precise" that it can select the target you WANTED even if YOU are unable to distinguish said target.

Again: see Bob casting [I]hold person to solve a murder mystery.

As I've explained, I'm working backward from the idea that the further text of the spell is an attempt to fully define the function of the duplicates. It provides a mechanic for interaction with attacks and none for interaction with non-attacks. Therefore it only protects against attacks.

Targeting real separate individuals does not work like this, and therefore the murder mystery example is distinguishable. The rule established by Mirror Image only applies to Mirror Image, which is a unique circumstance.

What this means for the fluff of the spell is up to you. I think Picasso-Joe is a valid fluff interpretation, or Neo-Joe, or Hall-of-Mirrors Joe. Whichever fluff you choose, it isn't something that baffles spells, unlike the murder mystery.

Foxhound438

2016-08-18, 03:42 AM

Certainly Hold Person is not an attack. In fact I even agree that mirror image "does nothing" against Hold Person unless you think that Hero, for some reason, cannot see the four "Joes." I contend that, because there are four Joes, Hero must choose one of them as his target, and this has nothing to do with the mechanics of Mirror Image that are specific to attacks.

But how does Hero select a target, if he can see four Joes?

This is getting into the realm of the kind of "which way do you try to open the door" DM munchkinnery that I despise. Even if I play along with this ****, after I say "the closest one", how do you determine that the closest one isn't the real one? it always ends up being you trying to give a spell an effect that it doesn't have, based on how you fluff the spells, in your own imagination of them.

(Forgive me for being abrasive, but it really does remind me of the time that a dm literally asked me which way I tried to open a door, and it irritates me on a visceral level whenever I see something similar.)

For your "how do I chose", I could easily fluff it that as I cast a non-attack spell, I'm "feeling" out the target, through non-physical mystic magic fluff ways, and "grasp" the mind of the target. even if I don't see which of these four Joe's are real physically, I can grasp the mind of the one that matters, because my spell takes effect on the "mind" level, which has nothing to do with the visual representation of the creature or anything around it. I could run around this fluff war all day with another player at my table, or we could stick to the explicit effects of game features, follow the specific rules where they apply, and follow the general rules where no specific rule applies.

nb4 the whole "that brings us back to the murder mystery dillema", my response would be that you know the identity of your target. Identity in this case not necessarily meaning name, or details about the person, but more broadly "that guy", and having four images you're pointing to, or ninteen, or seven hundred billion five hundred and twelve doesn't matter, because it's still "that guy" that you're targeting. Basically, see the whole "they aren't discretely targetable" string of this thread.

BurgerBeast

2016-08-18, 03:47 AM

Summary of the Argument Thus Far

Situation:

I. Joe, who is a room by himself, casts mirror image.
II. Hero enters the room and declares “I cast hold person on Joe.
III. As DM, what do you do?

1. Consult the text for hold person (PHB 251)

1.a. Choose a humanoid that you can see within range.

1.a.i. Already we have our first problem. We will call it the invalid target problem (resolved that there is no RAW on the matter, and therefore this is not a RAW issue. However, any stance you take on this issue must be coherent, consistent, and should not raise any issues that are clearly undesireable)

1.a.ii. We also have our second problem. We’ll call it the semantics of targeting problem. (currently unresolved)

1.b. The target must succeed on a Wisdom saving throw or be paralyzed for the duration.

1.b.i. There is no actual disagreement here. The only issue that arises here is due to 1.a.i.

1.c. Nothing else in the text has really entered the argument.

2. Consult the text for mirror image (PHB 260)

2.i. There is some disagreement over what is fluff and what is mechanics. We’ll call this the fluff or mechanics problem. (currently unresolved)

2.ii. There is disagreement over the conditional: each time a creature targets you with an attack and to what it applies. We’ll call this the entirety or exception problem. (currently unresolved)

2.b. Three illusory duplicates of yourself appear in your space.

2.b.i. There is some disagreement over wether the duplicates are discrete. We’ll call this the discrete duplicates problem. (currently unresolved)

2.c. Until the spell ends, the duplicates move with you and mimic your actions, shifting positions so it’s impossible to track which image is real.

2.c.i. There is some disagreement here, but it arises mostly because of 2.i, 2.ii, and 2.b.i. There is room for intrusion from 1.a.i and 1.a.ii as well.

2.d. You can use you raction to dismiss the illusory duplicates.

2.e. Each time a creature targets you with an attack during the spell’s duration, roll a d20 to determine whether the attack instead targets one of your duplicates.

2.e.i. This has been the subject of much debate, but I posit that the debate is not really about what this sentence means. It is about 1.a.ii, 2.ii, and perhaps 1.a.i.

2.f. Nothing else in the text has really entered the argument.

3. The Invalid Target Problem.

The rules for Targets under Spellcasting can be found on PHB 204. They say nothing about what happens when a spell is cast at/on an invalid target.

Since they say nothing, there is no RAW, so a DM must houserule. Some relevant questions are:

Is it even possible to cast a spell at an irrelevant target? (Or does the DM just say, that's not possible, what do you want to do instead?)

If it is possible, then what happens? Does the spell take effect, or do nothing, or does something else happen? Does the player use up a spell slot? Does the player use up their action?

This problem is important because how you rule on this particular problem will carry through the rest of the argument, and you will need to check for inconsistencies.

4. The Semantics of Targeting Problem

What does it mean to target a thing?

Is targeting a discrete event or a process?

The argument for discrete: Target: a player declares a target. Done.

The argument for process: Target: the process that begins when a player declares an intended target and ends when the DM has unambiguously determined the target. In between there may be a number of methods employed by the DM to facilitate the process.

This problem is important when we discuss 2.e, because it determines the point at which the mirror image text becomes relevant to attacks.

5. The Fluff or Mechanics Problem

There is some debate over whether particular sentences are “fluff” or “mechanics.”

In the first place you will need to decide if you think this is a true dichotomy. In other words, must a particular bit of text be either “fluff” or “mechanics,” but not both? Or can a bit of text be both.

Secondarily, can a bit of “fluff” text intrude on mechanics?

The answer to this is important to challenges that attempt to exclude some of the text from the PHB that it is “fluff” and therefore not relevant to RAW discussion.

6. The Entirety or Exception Problem

There is disagreement over whether paragraphs 2, 3, and 4 of the spell mirror image describe the entire mechanics of the spell, or only the case when the target is attacked.

Those who advocate the “entirety” claim that paragraphs 2, 3, and 4 comprise the entire mechanical effect of the spell. As such, the entire spell is relevant only to attacks. This position often includes the contention that paragraph 1 should be ignored because it is simply “fluff” that is provided to supplement the mechanics (which are found in paragraphs 2, 3, and 4 only.)

Those who advocate the “exception” claim that paragraphs 2, 3, and 4 describe what happens specifically in the case of an attack, but there is more to the spell. They contend that the spell’s effects, described in paragraph 1, take (mechanical) effect regardless of whether the target is attacked.

7. The Discrete Duplicates Problem

Are the illusory duplicates discrete?

The yes side: contends that the duplicates are discrete and individually targetable.

The no side: contends that the duplicates are not discrete and cannot be individually targeted. (I am sorry if I haven’t got this right. This one is the hardest for me to understand)

BurgerBeast

2016-08-18, 04:06 AM

Gwendol

2016-08-18, 06:26 AM

Counterspell interrupts the spell while being cast. By that reasoning, the slot wouldn't be used up because the spell casting wasn't actually completed.

I don't think Counterspell explicitly interrupts the spell while cast. It causes the spell to fail, thus implying the casting of it may indeed be completed.

Gwendol

2016-08-18, 06:29 AM

My Stances:

The Invalid Target Problem: A player can target an invalid target. The spell is cast (the slot is lost), and the action is lost.

The Semantics of Targeting Problem: targeting is a process.

The Fluff or Mechanics Problem: It's not a true dichotomy because sometimes fluff intrudes on mechanics.

The Entirety or Exception Problem: Paragraph 1 applies, regardless. Paragraphs 2, 3, and 4 describe what happens specifically in the case of an attack, but there is more to the spell.

The Discrete Duplicates Problem: the duplicates are discrete.

I have another problem here: if a wizard or cleric faces an enemy with mirror image up, why would they choose to use hold person, or any similar targeted spell (with only one target)? Especially if Hero is a wizard with likely 16 INT or more.

smcmike

2016-08-18, 07:55 AM

I think that's a fair summary, Burger.

A couple of points -

"Entirety" is an ok word for my stance, but maybe not the best possible word. Paragraph one does have meaning, but that meaning is limited by the mechanical definitions of the later paragraphs.

This is the root of my argument, and either informs or renders irrelevant all of the other points of contention that you have identified. To the extent that I've entered into debates regarding the other points, it was in an attempt to demonstrate that models other than my own create significant problems, in the hopes of convincing you that it would be simpler to follow my model.

Here is the basic outline:

1) It is possible for later sections to limit or define earlier sections, because we are using natural language, not formal logic.

2) Using this approach leads to a good outcome.

2a) The balance of the spell is better if it doesn't prevent targeting with non-attack spells. Compare with Blur.

2b) Interactions with spells and effects are easy to adjudicate. I don't need to make up a house rule for targeting with spells, and I don't see any significant inconsistencies arising from this interpretation.

2c) This interpretation is internally consistent with the fact that non-attack spells have no effect on the images. Eldritch Blast can destroy an image, but Sacred Flame cannot. This makes more sense if Sacred Flame never targets the images.

2d) I can use the simple everyday definition of "target" without any further consideration. Your "targeting is a process" argument is rather tortured.

2e) My outcome is consistent with the clarifications offered by Crawford.

Edit- I previously included "other than dispel magic" in 2c. After thinking about it, this is interesting - how do you see Dispel Magic interacting with Mirror Image, BurgerBeast? In my mode, it's easy - you cast dispel magic on the creature who cast Mirror Image, and the Mirror Images go away, as do any other low-level spells on the target. How does it work for you?

Tanarii

2016-08-18, 08:48 AM

Summary of the Argument Thus Far

Situation:

I. Joe, who is a room by himself, casts mirror image.
II. Hero enters the room and declares “I cast hold person on Joe.
III. As DM, what do you do?Good summary. The only niggle I'd have (and it's a minor one) is that you should make it clear you're using Joe as a variable. As in, since there is a Semantics of Targeting problem, it should be clear that saying "I cast hold person on Joe" is (potentially) not a valid statement. Besides, there are (and should be) many ways a player will choose to identify their target. "I cast hold person on $Joe" is how I would write it in a script, showing it's a variable. Or perhaps "I cast hold person on [Joe]"?

/endniggle

Segev

2016-08-18, 11:29 AM

This is getting into the realm of the kind of "which way do you try to open the door" DM munchkinnery that I despise. Even if I play along with this ****, after I say "the closest one", how do you determine that the closest one isn't the real one? it always ends up being you trying to give a spell an effect that it doesn't have, based on how you fluff the spells, in your own imagination of them. The RAW are silent on how you make that determination, other than this: when the confusion is due to mirror image, the images are explicitly, per the RAW indistinguishable from the caster, to the point that even if you determine "the closest one" is the caster for the time it took to swing a sword at him once, you can't be sure of that for a second swing in the same Fighter's turn.

Because the RAW expressly state that you cannot determine which is the real one by such descriptors as "the closest one," that leaves a requirement for some means of randomly determining whether "the closest one" is the real Joe or an illusion.

The RAW provides this for "attacks." It remains silent on it for anything else. It is argued by one side that this means that things like hold person, which are not "attacks," ignore mirror image entirely and magically pick the real Joe out of the cluster of 4 "Joe"s.

I point out that the only way to be consistent with this ruling is for Bob to be able to solve murder mysteries by casting hold person on "the suspect who actually committed the murder I'm investigating." Because Bob has no more information about which of the suspects is the murderer than he does about which of the images is the real Joe.

If you do not like that result, then you can instead argue (as I do) that the effect of the spell includes the fact that there are, in fact, 3 illusory images of Joe. This is not fluff with some strange not-really-representing-a-number-of-actual-images counter set to "3." This is what the spell does.

With that interpretation, the RAW being silent on how you determine if "the closest one" is the real Joe or not simply means that the DM has to come up with a method.

Given 5e's absolute passion for DM-rulings, this should surprise nobody as a possible outcome of a spell written in 5e.

(Forgive me for being abrasive, but it really does remind me of the time that a dm literally asked me which way I tried to open a door, and it irritates me on a visceral level whenever I see something similar.)Condolences. I can contrive situations (mostly games like Paranoia) where that's justified - usually for humor value - but I can also see why in most games that would be arduous.

For your "how do I chose", I could easily fluff it that as I cast a non-attack spell, I'm "feeling" out the target, through non-physical mystic magic fluff ways, and "grasp" the mind of the target. even if I don't see which of these four Joe's are real physically, I can grasp the mind of the one that matters, because my spell takes effect on the "mind" level, which has nothing to do with the visual representation of the creature or anything around it. I could run around this fluff war all day with another player at my table, or we could stick to the explicit effects of game features, follow the specific rules where they apply, and follow the general rules where no specific rule applies. You could, but that's not what the spells say they do. And it still leaves you with Bob being able to "feel out" the target, based on only one of the potential targets being the real guilty murderer. You're slicing that line awfully thin when you start to say, "No, see, it can't feel out THAT, but it can feel out THIS, because I want two different results in these two different scenarios."

Or you can just do what Crawford does and spot-rule without any care in the world for consistency or the ability of readers of the rulebooks to be able to judge what the rules actually do between two similar situations without having to write in to Crawford (or ask their DM) to define the effects. Which is why I find Crawford's rulings useless - their lack of consistency means they fail to provide guidelines to DMs to extrapolate to other rules, and thus they are useful only in the sense that they let theoreticians claim Crawford is the default DM for RAW discussions. Which I reject because it's appeal to invalid authority, since Crawford isn't running any games that most of us are in.

nb4 the whole "that brings us back to the murder mystery dillema", my response would be that you know the identity of your target. Identity in this case not necessarily meaning name, or details about the person, but more broadly "that guy", and having four images you're pointing to, or ninteen, or seven hundred billion five hundred and twelve doesn't matter, because it's still "that guy" that you're targeting. Basically, see the whole "they aren't discretely targetable" string of this thread.But you don't. You know that in that mass of guys is the guy you want to target. You may suspect that only one of them isn't an illusion. But you don't know which one. If you can pick "that guy" out of a pile of illusory "that guys," you can pick "that guy" out of a pile of real impersonators of "that guy," and you can pick "the real murderer" out of a pile of suspected murderers of which only one really is guilty.

2a) The balance of the spell is better if it doesn't prevent targeting with non-attack spells. Compare with Blur. I disagree. "Screws over rogues and fighter-types primarily" isn't better balance than "screws with everybody." If anything, the only part of mirror image as written that is too good is that fighter-types' preferred means is the only way to hit the spell's weakness compared to blur: that its protection goes away as it's used.

2b) Interactions with spells and effects are easy to adjudicate. I don't need to make up a house rule for targeting with spells, and I don't see any significant inconsistencies arising from this interpretation. Other than the fact that, for this one particular instance, you can ignore the need to pick a target, I suppose.

2d) I can use the simple everyday definition of "target" without any further consideration. Your "targeting is a process" argument is rather tortured. Nonsense. While I wouldn't say "targeting is a process," myself, if you argue that all you have to do is "name" your target by some means other than essentially pointing at it and saying "that one," you're back to Bob solving the murder mystery.

2e) My outcome is consistent with the clarifications offered by Crawford.That's nice. Pity Crawford's clarifications aren't typically consistent with themselves; he rules one way in one question and another way in the same question on other rules. (See: conjure woodland beings vs. fabricate, both of which don't specify that the caster gets to choose SPECIFICALLY what the spell creates, but which Crawford has said the first therefore lets the DM make the final choice, while everybody I've brought this up with insists fabricate lets the caster choose. Of course, that's not Crawford, I suppose; maybe he WOULD be consistent and rule that fabricate lets the DM decide what is actually made from the materials the player selects.)

Edit- I previously included "other than dispel magic" in 2c. After thinking about it, this is interesting - how do you see Dispel Magic interacting with Mirror Image, BurgerBeast? In my mode, it's easy - you cast dispel magic on the creature who cast Mirror Image, and the Mirror Images go away, as do any other low-level spells on the target. How does it work for you?Fairly easy, actually: if you target an image or the caster, you still are targeting the SPELL, so dispel magic works the way you want it to, here, smcmike.

smcmike

2016-08-18, 12:03 PM

I disagree. "Screws over rogues and fighter-types primarily" isn't better balance than "screws with everybody." If anything, the only part of mirror image as written that is too good is that fighter-types' preferred means is the only way to hit the spell's weakness compared to blur: that its protection goes away as it's used.

I'm assume that by "Mirror Image as written," you are referring to your model of how the spell works? For clarity's sake, maybe refer to it as "my interpretation" or some other clear signifier.

Honestly, I don't understand your balance argument at all.

I am contending that your version is too powerful, and using Blur as a comparison point. "Protects against one class of threats" is less powerful than "protects against multiple classes of threats." Hence, my version is less powerful, and more in line with Blur in power. In most situations, my version is still significantly better than Blur, I think.

Also, it has a nice balance - the sort of threats that it protects against are also the best way, apart from dispel magic, to get rid of that protection.

Other than the fact that, for this one particular instance, you can ignore the need to pick a target, I suppose.

You ignore the need to pick the target when attacking - you go from "I target Joe" to "Joe rolls to make me target an image." There is no need to pick a target between mirror images in any circumstance. This is convenient, because it would be impossible to do so.

Nonsense. While I wouldn't say "targeting is a process," myself, if you argue that all you have to do is "name" your target by some means other than essentially pointing at it and saying "that one," you're back to Bob solving the murder mystery.

This just doesn't follow from my argument, which is that Mirror Image only affects targeting per the specifically described mechanic. If you are looking for a fluff description of why you don't need to pick a specific image, I've offered several. On the other hand, I'm not convinced that you or BurgerBeast solved the two-step targeting problem, and you seem now to be rejecting his solution.

That's nice. Pity Crawford's clarifications aren't typically consistent with themselves; he rules one way in one question and another way in the same question on other rules. (See: conjure woodland beings vs. fabricate, both of which don't specify that the caster gets to choose SPECIFICALLY what the spell creates, but which Crawford has said the first therefore lets the DM make the final choice, while everybody I've brought this up with insists fabricate lets the caster choose. Of course, that's not Crawford, I suppose; maybe he WOULD be consistent and rule that fabricate lets the DM decide what is actually made from the materials the player selects.)

I'm sure there's another thread about this somewhere. I don't know anything about the nuances of the argument, but the language of the two spells are quite different. This doesn't seem like a particularly potent attack on Crawford to me. Regardless, I give his opinion some weight. If you don't, that's ok with me. It isn't necessary for my argument.

Fairly easy, actually: if you target an image or the caster, you still are targeting the SPELL, so dispel magic works the way you want it to, here, smcmike.

No, the way it works in my interpretation is that you can target the caster, thereby removing Mirror Image and also any other effects.

Segev

2016-08-18, 12:24 PM

No, the way it works in my interpretation is that you can target the caster, thereby removing Mirror Image and also any other effects.

Then Bob can target "the real murderer," and bypass any other effects (like not knowing who that is).

Unless you choose to be inconsistent and just rule on a case-by-case basis, OR inject a house rule that mirror image expressly does not count as multiple targets, despite saying it creates the images which are, in fact, distracting targets.

smcmike

2016-08-18, 12:26 PM

Then Bob can target "the real murderer," and bypass any other effects (like not knowing who that is).

Unless you choose to be inconsistent and just rule on a case-by-case basis, OR inject a house rule that mirror image expressly does not count as multiple targets, despite saying it creates the images which are, in fact, distracting targets.

This just doesn't follow from my argument, which is that Mirror Image only affects targeting per the specifically described mechanic. If you are looking for a fluff description of why you don't need to pick a specific image, I've offered several. On the other hand, I'm not convinced that you or BurgerBeast solved the two-step targeting problem, and you seem now to be rejecting his solution.

Segev

2016-08-18, 12:32 PM

smcmike

2016-08-18, 12:37 PM

Ah, right.
Your argument's problem, for me, is that it reduces my immersion in the game and thus reduces my enjoyment, because somehow having 4 "Joe"s standing there makes it hard for me to tell which one is the real one if I'm throwing something at him, but doesn't make it hard for me to tell which one is the real one if I want to pick his pocket, cast hold person, or tap him with a touch-ranged buff. Somehow, if I'm picking his pocket, I know exactly which one is Joe, but if I'm swinging my fist, I can't tell.

Right, yours is an argument based upon visualization. And do you know what? I agree that the pickpocket example is troubling. I don't have the same problem with the magic examples, because magic.

Segev

2016-08-18, 12:44 PM

Right, yours is an argument based upon visualization. And do you know what? I agree that the pickpocket example is troubling. I don't have the same problem with the magic examples, because magic.

Fair enough. I have the problem with the magic examples because I still can't justify why it would work for mirror image but not the murder mystery.

And it's not pure visualization; rather, it's that the visualization is literally what the spell does.

Strip away resolution mechanics. What Joe, who isn't reading the 5e PHB, knows his spell does is create 3 illusory duplicates of himself. Brad the fighter sees 4 "Joe"s and has to pick one of them to swing his sword at. Ben the rogue sees 4 "Joe"s and should have just as much trouble picking out which one's pockets to pick. Bob the mage sees 4 "Joe"s and should have just as much trouble picking which one to hold.

Add the resolution mechanic to allow Brad to figure out if he attacked the right "Joe" or not, and that shouldn't somehow make Ben and Bob automatically able to pick out the right one. Not when the mechanic provided for Brad is simply silent on Ben and Bob's situations.

I get that your argument is that the resolution mechanic for Brad is restrictive, saying that ONLY Brad is affected, because Ben and Bob aren't triggering the resolution mechanic and thus ignore any effect of mirror image. But that doesn't make sense in the game world. In the game world, nobody's rolling d20s. Brad's swinging a sword at 1 of 4 images of Joe; Ben is attempting to pick the pockets of one of 4 images of Joe; Bob is casting a spell at one of four images of Joe.

Tanarii

2016-08-18, 12:45 PM

This just doesn't follow from my argument, which is that Mirror Image only affects targeting per the specifically described mechanic. If you are looking for a fluff description of why you don't need to pick a specific image, I've offered several. On the other hand, I'm not convinced that you or BurgerBeast solved the two-step targeting problem, and you seem now to be rejecting his solution.

Ah, right.

Your argument's problem, for me, is that it reduces my immersion in the game and thus reduces my enjoyment, because somehow having 4 "Joe"s standing there makes it hard for me to tell which one is the real one if I'm throwing something at him, but doesn't make it hard for me to tell which one is the real one if I want to pick his pocket, cast hold person, or tap him with a touch-ranged buff. Somehow, if I'm picking his pocket, I know exactly which one is Joe, but if I'm swinging my fist, I can't tell.

This makes zero sense.

And the "fluff" readings which try to justify it wind up violating the RAW description of what the spell does, which is create 3 images of Joe. Not "blur Joe's position," but actually create specific numbers of images which can be hit individually.

It's violating your immersion, and you think it's violating RAW, because you are insisting on making two assumptions about "actually create specific numbers of images which can be hit individually":
1) That being able to be hit individually via a redirect mechanic within the Mirror Image spell after the creature is targeted = images being able to be individually targeted outside of that mechanic.
2) An in-game interpretation that they must be physically separate, or visually identifiable as separate, images.

Neither of those is required by the spell.

Segev

2016-08-18, 12:48 PM

It's violating your immersion, and you think it's violating RAW, because you are insisting on making two assumptions about "actually create specific numbers of images which can be hit individually":
1) That being able to be hit individually via a redirect mechanic within the Mirror Image spell after the creature is targeted = images being able to be individually targeted outside of that mechanic.
2) An in-game interpretation that they must be physically separate, or visually identifiable as separate, images.

Neither of those is required by the spell.

The latter actually is, by the very definition of being an image of the caster. And given that the spell expressly states the images move around so that you can't tell from moment to moment which one is which, any argument that they are not distinctly recognizable individual images that you can discretely count must ignore the text of the spell.

georgie_leech

2016-08-18, 12:56 PM

The latter actually is, by the very definition of being an image of the caster. And given that the spell expressly states the images move around so that you can't tell from moment to moment which one is which, any argument that they are not distinctly recognizable individual images that you can discretely count must ignore the text of the spell.

Not necessarily. Have you ever crossed your eyes enough to get the images of what you're looking at to partially overlap? The duplicates can be overlapping without breaking the text of the spell. If anything, it's probably harder to tell them apart that way; it's hard to track the images separately if you can't tell where one ends and the other begins.

Foxhound438

2016-08-18, 01:28 PM

Segev

2016-08-18, 01:35 PM

georgie_leech

2016-08-18, 01:51 PM

Can you still identify them as "that's one image; that's another?"

If "yes," then you still have to choose at which one you're targeting your effect.
If "no," then how do you hit one when you roll and get an image?

'You try to hit Joe, but it turns out he was a little to the left of where you thought he was. You instead cause an image of Joe to fade away as you slash it's arm.' I really don't see why that's so unreasonable for 'no, they're not independently targetable.'

Tanarii

2016-08-18, 01:57 PM

Not necessarily. Have you ever crossed your eyes enough to get the images of what you're looking at to partially overlap? The duplicates can be overlapping without breaking the text of the spell. If anything, it's probably harder to tell them apart that way; it's hard to track the images separately if you can't tell where one ends and the other begins.
Exactly. That's a perfectly valid interpretation of the RAW, that doesn't violate it in any way.

I'll say I don't think it MUST be visualized that way. Any visualization that meets the requirements of RAW is fine: That you target the creature first, and in the case of an attack, it is potentially redirected to an image and resolved as a hit or miss.

Edit: Sorry, that's making an assumption about RAW that's at the center of the discussion. I'll rephrase:

Any visualization that meets the requirements of RAW is fine: That you target the creature first with an attack, then it is potentially redirected to an image and resolved as a hit or miss.

This removes the assumption that for a non-attack, you must target the creature first. But for an attack, you must target the creature first, then it redirects to an image. That's RAW.

BurgerBeast

2016-08-18, 07:07 PM

As I've explained, I'm working backward from the idea that the further text of the spell is an attempt to fully define the function of the duplicates. It provides a mechanic for interaction with attacks and none for interaction with non-attacks. Therefore it only protects against attacks.

So then, we really need to address the question of whether the spell's mechanics are restricted to paragraphs 2, 3, and 4. Why do you suppose they even included paragraph 1? They could have just cut it and we'd all be in agreement.

What this means for the fluff of the spell is up to you. I think Picasso-Joe is a valid fluff interpretation, or Neo-Joe, or Hall-of-Mirrors Joe. Whichever fluff you choose, it isn't something that baffles spells, unlike the murder mystery.

And you regard paragraph to be purely fluff, as well. In my view this is baseless (I am aware of your basis, but I don't accept). Even some of the claims that all spells begin with "fluff" text seems baseless, in this same sense, to me, upon a light skim of the spells section of the PHB.

----------

I think that's a fair summary, Burger.

A couple of points -

"Entirety" is an ok word for my stance, but maybe not the best possible word. Paragraph one does have meaning, but that meaning is limited by the mechanical definitions of the later paragraphs.

This is the root of my argument, and either informs or renders irrelevant all of the other points of contention that you have identified. To the extent that I've entered into debates regarding the other points, it was in an attempt to demonstrate that models other than my own create significant problems, in the hopes of convincing you that it would be simpler to follow my model.

Yep. You and I hold irreconcilable views. I know what you think and why you think it, and I think you're wrong. End of story. I'm sure you would say the same about me and my stance. Essentially, the conversation's been fun, but it's over. Unless you want to continue to debate what essentially amounts to "how to read."

Thanks for engaging.

Here is the basic outline:

1) It is possible for later sections to limit or define earlier sections, because we are using natural language, not formal logic.

2) Using this approach leads to a good outcome.

2a) The balance of the spell is better if it doesn't prevent targeting with non-attack spells. Compare with Blur.

2b) Interactions with spells and effects are easy to adjudicate. I don't need to make up a house rule for targeting with spells, and I don't see any significant inconsistencies arising from this interpretation.

2c) This interpretation is internally consistent with the fact that non-attack spells have no effect on the images. Eldritch Blast can destroy an image, but Sacred Flame cannot. This makes more sense if Sacred Flame never targets the images.

2d) I can use the simple everyday definition of "target" without any further consideration. Your "targeting is a process" argument is rather tortured.

2e) My outcome is consistent with the clarifications offered by Crawford.

Edit- I previously included "other than dispel magic" in 2c. After thinking about it, this is interesting - how do you see Dispel Magic interacting with Mirror Image, BurgerBeast? In my mode, it's easy - you cast dispel magic on the creature who cast Mirror Image, and the Mirror Images go away, as do any other low-level spells on the target. How does it work for you?

Yep. I refute 1. It's a book of rules, not a book of loose guidelines. A book of rules is expected to be clear and unambiguous, and this includes logical. In the same way that a sport's Book of Laws is not written in natural language, so neither should a rulebook for a game be written in natural language. Aside form whether it should be written in natural language, I think there is evidence to show that it is not, in fact, written in natural language. So, it at least attempts to follow logical guidelines beyond a typical book written in natural language.

I can concede that there is evidence that the authors attempted to use more "reader-friendly language" or "natural language" as a means of clearly communicating, but the fact remains that what is being communicating is a set of rules, for the purpose of applying them to a game - just as a rugby referee applies the laws of rugby.

I think 2, 2b, and 2e are irrelevant.

I think 2a is wrong.

I think 2c is interesting because if the spell can never strike the duplicates, they wouldn't need to specify that it has no effect. This can be read as evidence against your view.

2d is worth discussing. I explain my "tortured view" (I think in the other thread) and why it matters. I think everyone agrees with what I meant to say, the problem is that I didn't say it well.

TL;DR We clearly disagree on 1. Either we try to resolve it or acknowledge that we don't agree.

----------

I'm assume that by "Mirror Image as written," you are referring to your model of how the spell works? For clarity's sake, maybe refer to it as "my interpretation" or some other clear signifier.

No. What is meant is literally the words of mirror image, as written. Specifically, the words that you claim should be ignored because different words, further down the page, are more important.

There is no need to clarify, and there is no interpretation involved. The words are here,and they say what they say. The default assumption is that words are in a book for a reason. To say that words in a book (a book of rules, in fact, in a section that deals with almost exclusively rules) are meant to be meaningless, is your interpretation.

I am contending that your version is too powerful, and using Blur as a comparison point. "Protects against one class of threats" is less powerful than "protects against multiple classes of threats." Hence, my version is less powerful, and more in line with Blur in power. In most situations, my version is still significantly better than Blur, I think.

Your argument fails because it's not necessarily so. "Protects against two classes of threats" can be weaker than "protects against one class of threats."

Also, it has a nice balance - the sort of threats that it protects against are also the best way, apart from dispel magic, to get rid of that protection.

I could (but I don't) agree that your version is superior to the PHB version of the spell. This still would not make it correct, which is what I am concerned about.

(snipped because irrelevant until the text disagreement is resolved)

I have another problem here: if a wizard or cleric faces an enemy with mirror image up, why would they choose to use hold person, or any similar targeted spell (with only one target)? Especially if Hero is a wizard with likely 16 INT or more.

Fair question, but beyond the scope of what we're discussing (except in terms of the balance arguments). We are discussing the case where they do cast hold person. The reasons aren't relevant.

Good summary. The only niggle I'd have (and it's a minor one) is that you should make it clear you're using Joe as a variable. As in, since there is a Semantics of Targeting problem, it should be clear that saying "I cast hold person on Joe" is (potentially) not a valid statement. Besides, there are (and should be) many ways a player will choose to identify their target. "I cast hold person on $Joe" is how I would write it in a script, showing it's a variable. Or perhaps "I cast hold person on [Joe]"?

/endniggle

Yeah, I'm looking for a better way to explain this problem. I'm open to suggestions. This approach might be fruitful. The whole reason I said that targeting is a process is to deal with this by showing that ultimately targeting hasn't occurred until the DM knows what the target is. Yet there essentially infinite ways that player might try to convey the intended target to the DM. If the player makes invalid declarations, then it seems to me that: (1) by the discrete view, targeting is yet to happen. (2) by the process view, the DM gets to help out by recognizing that the player is trying to select target, and work to resolve the intended target.

----------

It's violating your immersion, and you think it's violating RAW, because you are insisting on making two assumptions about "actually create specific numbers of images which can be hit individually":
1) That being able to be hit individually via a redirect mechanic within the Mirror Image spell after the creature is targeted = images being able to be individually targeted outside of that mechanic.
2) An in-game interpretation that they must be physically separate, or visually identifiable as separate, images.

Neither of those is required by the spell.

These reasons do not hold up in our context. I suspect that we disagree much earlier in the thread, in the same place that I disagree with smcmike: over whether all of the text is relevant.

Not necessarily. Have you ever crossed your eyes enough to get the images of what you're looking at to partially overlap? The duplicates can be overlapping without breaking the text of the spell. If anything, it's probably harder to tell them apart that way; it's hard to track the images separately if you can't tell where one ends and the other begins.

There is one thing. But there are two images. Two images. If you disagree that there are two images, then please explain to me how one image overlaps with itself.

If this doesn't drive the point home, then maybe we disagree about whether it is possible to target an image.

----------

'You try to hit Joe, but it turns out he was a little to the left of where you thought he was. You instead cause an image of Joe to fade away as you slash it's arm.' I really don't see why that's so unreasonable for 'no, they're not independently targetable.'

By the rules, there is no way for what you describe here to happen. What you try to hit is decided based on what you target. Whether you hit is decided by the attack roll. If you "try to hit Joe" that means that you targeted the real Joe. If you "miss," it's because your attack roll was low. There's no way, mechanically, for a missed attack against Joe to strike an image.

In the interest of reading your word fairly, though: it is possible to strike a duplicate when "trying to strike Joe," but this isn't in fact what happens. What in fact happens is that you target an image instead of Joe, and hit it. This should be described as "confronted with four Joes, you pick one and swing at it. You hit it! But it turns out you picked a duplicate, and not the real Joe!"

RickAllison

2016-08-18, 07:18 PM

So then, we really need to address the question of whether the spell's mechanics are restricted to paragraphs 2, 3, and 4. Why do you suppose they even included paragraph 1? They could have just cut it and we'd all be in agreement.

And you regard paragraph to be purely fluff, as well. In my view this is baseless (I am aware of your basis, but I don't accept). Even some of the claims that all spells begin with "fluff" text seems baseless, in this same sense, to me, upon a light skim of the spells section of the PHB.

----------

Yep. You and I hold irreconcilable views. I know what you think and why you think it, and I think you're wrong. End of story. I'm sure you would say the same about me and my stance. Essentially, the conversation's been fun, but it's over. Unless you want to continue to debate what essentially amounts to "how to read."

Thanks for engaging.

Yep. I refute 1. It's a book of rules, not a book of loose guidelines. A book of rules is expected to be clear and unambiguous, and this includes logical. In the same way that a sport's Book of Laws is not written in natural language, so neither should a rulebook for a game be written in natural language. Aside form whether it should be written in natural language, I think there is evidence to show that it is not, in fact, written in natural language. So, it at least attempts to follow logical guidelines beyond a typical book written in natural language.

I can concede that there is evidence that the authors attempted to use more "reader-friendly language" or "natural language" as a means of clearly communicating, but the fact remains that what is being communicating is a set of rules, for the purpose of applying them to a game - just as a rugby referee applies the laws of rugby.

I think 2, 2b, and 2e are irrelevant.

I think 2a is wrong.

I think 2c is interesting because if the spell can never strike the duplicates, they wouldn't need to specify that it has no effect. This can be read as evidence against your view.

2d is worth discussing. I explain my "tortured view" (I think in the other thread) and why it matters. I think everyone agrees with what I meant to say, the problem is that I didn't say it well.

TL;DR We clearly disagree on 1. Either we try to resolve it or acknowledge that we don't agree.

----------

No. What is meant is literally the words of mirror image, as written. Specifically, the words that you claim should be ignored because different words, further down the page, are more important.

There is no need to clarify, and there is no interpretation involved. The words are here,and they say what they say. The default assumption is that words are in a book for a reason. To say that words in a book (a book of rules, in fact, in a section that deals with almost exclusively rules) are meant to be meaningless, is your interpretation.

Your argument fails because it's not necessarily so. "Protects against two classes of threats" can be weaker than "protects against one class of threats."

I could (but I don't) agree that your version is superior to the PHB version of the spell. This still would not make it correct, which is what I am concerned about.

Fair question, but beyond the scope of what we're discussing (except in terms of the balance arguments). We are discussing the case where they do cast hold person. The reasons aren't relevant.

Yeah, I'm looking for a better way to explain this problem. I'm open to suggestions. This approach might be fruitful. The whole reason I said that targeting is a process is to deal with this by showing that ultimately targeting hasn't occurred until the DM knows what the target is. Yet there essentially infinite ways that player might try to convey the intended target to the DM. If the player makes invalid declarations, then it seems to me that: (1) by the discrete view, targeting is yet to happen. (2) by the process view, the DM gets to help out by recognizing that the player is trying to select target, and work to resolve the intended target.

----------

These reasons do not hold up in our context. I suspect that we disagree much earlier in the thread, in the same place that I disagree with smcmike: over whether all of the text is relevant.

There is one thing. But there are two images. Two images. If you disagree that there are two images, then please explain to me how one image overlaps with itself.

If this doesn't drive the point home, then maybe we disagree about whether it is possible to target an image.

----------

By the rules, there is no way for what you describe here to happen. What you try to hit is decided based on what you target. Whether you hit is decided by the attack roll. If you "try to hit Joe" that means that you targeted the real Joe. If you "miss," it's because your attack roll was low. There's no way, mechanically, for a missed attack against Joe to strike an image.

In the interest of reading your word fairly, though: it is possible to strike a duplicate when "trying to strike Joe," but this isn't in fact what happens. What in fact happens is that you target an image instead of Joe, and hit it. This should be described as "confronted with four Joes, you pick one and swing at it. You hit it! But it turns out you picked a duplicate, and not the real Joe!"

Just for your response to georgie_leech, we do have an exact mechanic to describe what happens when you "miss" Joe and hit a duplicate: the rest of the Mirror Image spell. Literally, the whole mechanic of rolling the d20 is for exactly the situation Georgie suggests. The person is baffled and looking at two overlapping images, but he can't tell which is the real one and which one he is seeing in his head. For something that can be pointed in its vague direction (Hold Person), it doesn't matter. A slash or a projectile, however, need to be far more accurate. The HP just aims for the middle of all the images, while an attack doing the same will likely miss.

smcmike

2016-08-18, 07:43 PM

So then, we really need to address the question of whether the spell's mechanics are restricted to paragraphs 2, 3, and 4. Why do you suppose they even included paragraph 1? They could have just cut it and we'd all be in agreement.

This is silly, considering that I noted below that paragraph 1 does have function.

And you regard paragraph to be purely fluff, as well. In my view this is baseless . . . Even some of the claims that all spells begin with "fluff" text seems baseless, in this same sense, to me, upon a light skim of the spells section of the PHB.

I don't think I said "all" spells. If I did, I was clearly wrong. If you'd like examples of spells I find similar in structure, I can provide them.

Yep. I refute 1. It's a book of rules, not a book of loose guidelines. A book of rules is expected to be clear and unambiguous, and this includes logical. In the same way that a sport's Book of Laws is not written in natural language, so neither should a rulebook for a game be written in natural language. Aside form whether it should be written in natural language, I think there is evidence to show that it is not, in fact, written in natural language. So, it at least attempts to follow logical guidelines beyond a typical book written in natural language.

I can concede that there is evidence that the authors attempted to use more "reader-friendly language" or "natural language" as a means of clearly communicating, but the fact remains that what is being communicating is a set of rules, for the purpose of applying them to a game - just as a rugby referee applies the laws of rugby.

Sorry, but here you are wrong. It is more than a book of rules. It is a handbook, and includes a lot of information beyond rules. In the case of spells, it often includes a fair bit of fluff or description for the purpose of immersion.

(1) doesn't state that this is the only reading of the spell, only that it is a possible reading. Considering that the purpose of language is to communicate, and the majority of readers on these threads did read it this way, I'm pretty confident it passes that threshold.

I think 2, 2b, and 2e are irrelevant.

Right. Good outcomes don't matter, and it isn't important that a spell creates easy-to-adjudicate interactions.

I think 2a is wrong.
. . .
Your argument fails because it's not necessarily so. "Protects against two classes of threats" can be weaker than "protects against one class of threats."

I could (but I don't) agree that your version is superior to the PHB version of the spell.

I'm glad you circled back here, though you haven't really advanced the argument that your interpretation is balanced.

When you say "'Protects against two classes of threats' can be weaker than 'protects against one class of threats,'" are you speaking in the abstract, or are you actually talking about our two interpretations of the spell? Because if you do have an argument that protecting against both spells and attacks is weaker than just protecting against attacks, I'd love to hear it.

As for what arguments you could (but don't) agree with, state them or (don't).

I think 2c is interesting because if the spell can never strike the duplicates, they wouldn't need to specify that it has no effect. This can be read as evidence against your view.

No, it can't, unless you ignore the existence of area of effect spells. You're flat wrong on this point.

2d is worth discussing. I explain my "tortured view" (I think in the other thread) and why it matters. I think everyone agrees with what I meant to say, the problem is that I didn't say it well.

Right, you did not convince me that this isn't a problem for your argument. I agree that your concept of targeting is a possible reading, but I don't think it's the simplest or most natural reading. Hence "tortured" as opposed to "wrong."

There is no need to clarify, and there is no interpretation involved. The words are here,and they say what they say. The default assumption is that words are in a book for a reason. To say that words in a book (a book of rules, in fact, in a section that deals with almost exclusively rules) are meant to be meaningless, is your interpretation.

Your arguments were more convincing when you were focusing on the issue instead of making stupid side-points. I understand that you and Seg think you are reading the spell the only way it can be read. I disagree. I'm merely suggesting that it would be more clear if we use neutral signifiers about our arguments. I was honestly confused by what Seg was trying to say here. You haven't illuminated it in this response.

BurgerBeast

2016-08-18, 07:48 PM

Thanks! This is good practice, generally, when trying to convince someone of something.

The reason it isn't relevant is that it just leads back to the same issues we've been discussing, and leads to me restating my position. The Elf can target Joe as if he just sees Joe because the mechanical part of the spell text effects Dwarves.

Yeah the same issues but a different conclusion form you. I suggest that is why you don't want to answer. I suggest that "it's not relevant" is an excuse to not answer a clearly analogous case because your answer will differ and you will be forced to offer an explanation.

This drives at the problem of ignoring some of the text.

For what it's worth, I agree - it would probably be too weird to say he can't see them. So, he must be able to see them, but for some reason that doesn't interfere with his targetting process. The reason isn't relevant, though.

Too weird? Some reason? The reason is entirely relevant.

The reason presents the question: why is it that the elf can't target these particular images. This raises the question of why they are different from illusions in general, which raises questions about targeting in general.

In the mirror image case, it was easy to argue that attacks are subject to the targeting issue but save-spells by-pass the targeting issue precisely because (for many others, at least, if not you) they could fall back on magic as the reason for an otherwise nonsensical view. The claim led to "homing-spells" and other contrived explanations that people felt free to posit because magic was involved.

By giving the exact same example, but changing the condition to "being a dwarf," this presents the same issue but illuminates the red herring that magic provided (and everyone latched on to). In this context, the situation makes no sense for precisely the same reason, but there's no "magic" escape hatch (pun intended). Well, there is still a magic escape hatch, but it's not to be found in hold person.

Aside from this, and specifically in your case, you are face with the challenge of explaining why two attacks lead to different mechanics. There is no attack/saving throw distinction.

So, since the elf is not under any magical influence, why does the elf get to just say "I target Joe" when there are 4 "Joes"? (Note I am not asking why is his attack not re-directed by mirror image. I am asking why he doesn;t have to choose a target.)

I'm not even sure what "context is always right" is supposed to mean. Maybe give me some context. Regardless, context is very frequently determinative of meaning. Tone, too.

Either am I, so long after writing it. Basically, searching for context in a text is more likely to lead you to the wrong conclusion that the right one. The same is true (only moreso) for tone. (See a gajillion arguments on internet forums, for example).

So, if you were to say "context is very frequently determinative of meaning." I would reply that "context is even more frequently misconstrued, which causes more confusion about meaning." And trying to decipher tone is even worse.

I agree that you can't completely rule out the possibility that the speaker is working Saturday morning. The statement is ambiguous.

We must have different definitions of ambiguous. Unspecified is not the same as ambiguous. Ambiguous, of assertions, means that they can be taken in more than one way (usually two ways). Unspecified means that they can be taken to mean anything.

The question of whether the speaker is working on Saturday morning is not ambiguous. It is clearly unspecified.

Here are some possible interpretations;

1. The speaker is working every hour of the weekend, but for some reason chose to highlight only portions of that work. This is the most literal reading, which I see as equivalent to your position about Mirror Image.
2. The speaker's description of her shifts at the bar and the hospital constituted an explanation or definition of her original statement, and she is not working beyond the shifts specified. This is equivalent to my position.
3. Something in between: it is possible that she didn't mean the first statement literally, and also that she didn't mean the list to be exhaustive.

If I had to guess, without knowing anything about the person, and even assuming that some people do, in fact, work 48 hours in a weekend (at a hospital, perhaps), I would go with interpretation (2). It's ambiguous, but it is very common to say one thing and then say another thing to define the first.

The point is, you are wrong to guess at all. Nothing was specified so you have no basis to form a guess.

This is entirely different from an ambiguous statement, in which an assertion is being made, but it's meaning is not clear. In this case, it is still wrong to choose one over the other, but it is right to exclude all possibilities that are not within the range of possible intended assertions.

The reason it is a common fallacy is that people do not speak in formal logic, and in every day use this is a reasonable interpretation of the sentence. 5e attempts to write in everyday language.

The reason it's a common fallacy is because people don't understand English. It may be true that they speak it sloppily or incorrectly in their everyday lives, but this just means that it is commonly used incorrectly. It doesn't make the common use correct.

The 5e PHB may attempt to write in everyday language. There is plenty of evidence to say that it does, but there is also plenty of evidence to show that it doesn't. There is plenty of game-specific vocabulary, game-specific jargon, and rules that are written in ways that show an attempt to be clear (beyond everyday language).

The fact remains that the Player's Handbook is a rule book for a game.

And yet they do not fix the circumstance in which the caster is targeted with Hold Person. Hence, you are saying they are broken.

No, I am saying there are no rules for targeting. I am saying that having no rules for targeting can raise problems when considered along with the text of mirror image. I think the rules for mirror image were most likely written to clarify any confusion, but accomplished the opposite.

Interpretation, conclusion, whatever.

You can't be serious. It is a critical distinction and makes all the difference in this case. Consider:

Ed thinks global warming is real because he saw a future vision of earth in his cornflakes this morning.
Sally thinks global warming is a real think because the majority of scientists say so.
Samantha thinks global warming is real because she has looked at the evidence and was convinced.

Yeah... interpetation, conclusion, whatever...

smcmike

2016-08-18, 08:09 PM

We must have different definitions of ambiguous. Unspecified is not the same as ambiguous. Ambiguous, of assertions, means that they can be taken in more than one way (usually two ways). Unspecified means that they can be taken to mean anything.

This is entirely different from an ambiguous statement, in which an assertion is being made, but it's meaning is not clear. In this case, it is still wrong to choose one over the other, but it is right to exclude all possibilities that are not within the range of possible intended assertions.

No, we do not disagree about the definition of ambiguous. As you describe below, we disagree about how much meaning can be placed into a particular set of statements. Suggesting that this is also a definitional dispute smells of cheap point-scoring.

The reason it's a common fallacy is because people don't understand English. It may be true that they speak it sloppily or incorrectly in their everyday lives, but this just means that it is commonly used incorrectly. It doesn't make the common use correct.

This sums up our disagreement in its entirety. Language comes from common usage. If language does what the speaker wants it to - for example, communicate something to someone, it is correct.

You can't be serious. It is a critical distinction and makes all the difference in this case. Consider:

Ed thinks global warming is real because he saw a future vision of earth in his cornflakes this morning.
Sally thinks global warming is a real think because the majority of scientists say so.
Samantha thinks global warming is real because she has looked at the evidence and was convinced.

Yeah... interpetation, conclusion, whatever...[/SPOILER]

Do you really think Crawford is drawing his conclusion based on corn flakes? He said mirror image has no effect on magic missile, because it doesn't involve an attack. You are being silly.

BurgerBeast

2016-08-18, 08:25 PM

This is silly, considering that I noted below that paragraph 1 does have function.

Oh, sorry. I suspect we still disagree on what that functon is and to what extent it applies.

I don't think I said "all" spells. If I did, I was clearly wrong. If you'd like examples of spells I find similar in structure, I can provide them.

No, I'm not even sure if you're the one who said it. It's been mentioned that the trend is significant enough that assumptions about the first few sentences being fluff are justified.

Sorry, but here you are wrong. It is more than a book of rules. It is a handbook, and includes a lot of information beyond rules. In the case of spells, it often includes a fair bit of fluff or description for the purpose of immersion.

(1) doesn't state that this is the only reading of the spell, only that it is a possible reading. Considering that the purpose of language is to communicate, and the majority of readers on these threads did read it this way, I'm pretty confident it passes that threshold.

No, sorry. You are. Being a book of rules does not exclude the possibility of being more than a book of rules. Not that it's particularly relevant, but the 5e spell descriptions are probably contain the least fluff of any edition of the game. Infer from that what you will.

On the topic of "possible" readings, I'm not particularly sorry to say that I took this to also mean "plausible." I refute the plausibility.

Right. Good outcomes don't matter, and it isn't important that a spell creates easy-to-adjudicate interactions.

This is the same logical error that you continue to make. (If A then B implies if not A then not B. This is incorrect.)

All I was driving at is that my ruling is equally good on all counts. So it is irrelevant because on this point I consider both interpretations equal.

I'm glad you circled back here, though you haven't really advanced the argument that your interpretation is balanced.

It's pretty simple: Even if mirror image is capable of thwarting save-spell attacks, this doesn't increase it's power to the point that I consider "broken." Most characters have an alternative (weapon attacks, attack spells) that will work as expected given the intended function of the spell, and some have the ability (AOEs) to can ignore it completely. There is no real is of the spell ever thwarting more than one save-spell, because people will learn how to deal with the spell pretty quickly. I just don't think it has anything near the "broken" levels of power that have been ascribed to it. Even magic missile is still pretty effective by my ruling.

When you say "'Protects against two classes of threats' can be weaker than 'protects against one class of threats,'" are you speaking in the abstract, or are you actually talking about our two interpretations of the spell? Because if you do have an argument that protecting against both spells and attacks is weaker than just protecting against attacks, I'd love to hear it.

In the abstract, as you were. You said "classes of threats," not attacks and spells. Of course I wouldn't argue that. I was contending what you said.

As for what arguments you could (but don't) agree with, state them or (don't).

I did: "[Y]our version is superior to the PHB version of the spell."

No, it can't, unless you ignore the existence of area of effect spells. You're flat wrong on this point.

Agreed. I concede this point. (Technically what I wrote is correct, but in spirit I was wrong.)

Right, you did not convince me that this isn't a problem for your argument. I agree that your concept of targeting is a possible reading, but I don't think it's the simplest or most natural reading. Hence "tortured" as opposed to "wrong."

What I mean is that it is of no consequence because even DMs who do not call it a process follow the "process." And even DMs who follow the process must make the discrete action within it.

It's of no consequence, except for one corner case (that still leaves room for interpretation).

Your arguments were more convincing when you were focusing on the issue instead of making stupid side-points.

Ditto. But you've been making "stupid side-points" for a lot longer than I have. I guess you're starting to understand what it's like.

That's why I tried to do it, despite knowing it would eventually fail.

I understand that you and Seg think you are reading the spell the only way it can be read. I disagree.

We're not saying it's the only way it can be read. It can be read in a (probably infinite or near-infinite) number of ways that are wrong. There are probably a fair numbers of ways that it can be read correctly, too.

BurgerBeast

2016-08-18, 08:38 PM

Just for your response to georgie_leech, we do have an exact mechanic to describe what happens when you "miss" Joe and hit a duplicate: the rest of the Mirror Image spell. Literally, the whole mechanic of rolling the d20 is for exactly the situation Georgie suggests.

No, it is not. The d20 mechanic determines what you will target (i.e. try to hit). The attack roll determines if you hit or miss.

The person is baffled and looking at two overlapping images, but he can't tell which is the real one and which one he is seeing in his head.

You've taken your interpretation to a whole new level. Seeing in his head? It's an illusion and it is there. Everyone can see it. It's not in his head, unless you think people just "see in their head" generally.

For something that can be pointed in its vague direction (Hold Person), it doesn't matter. A slash or a projectile, however, need to be far more accurate.

I seem to recall me of using my own personal fluff to justify my own personal house-rules. You've gone a bit "fluff happy" here.

The HP just aims for the middle of all the images, while an attack doing the same will likely miss.

How, in your visualization, is attacking the middle not the very best strategy?

----------

The spell's targeting dynamics do not allow for you to target one image, miss and strike another, any more than you are allowed to miss with your sword and then roll to see if you hit something else.

The specific determines the target. The attack roll determines whether the attack hits or misses. It ends there. There is no extra mechanic to determine if your miss ends up striking something else.

You are welcome to interpret an outcome however you wish, in retrospect and for the purpose of making a combat more interesting, but mechanically, that's how it works.

Tanarii

2016-08-18, 08:58 PM

The spell's targeting dynamics do not allow for you to target one image, miss and strike another, any more than you are allowed to miss with your sword and then roll to see if you hit something else.

The specific determines the target. The attack roll determines whether the attack hits or misses. It ends there. There is no extra mechanic to determine if your miss ends up striking something else.
right, but for an attack, you target the creature first, and it is possible redirected. At which point you determine a hit or miss by the normal rules.

What's important here is that an attack (at least) targets the creature. Initially. it's explicit.

Now, I agree it's an extension to say that this, which is the default rule for targeting anyway, also applies to non-attacks. But since it is the default rule to target the creature under normal circumstances, that makes sense to me.

smcmike

2016-08-18, 11:46 PM

No, sorry. You are. Being a book of rules does not exclude the possibility of being more than a book of rules. Not that it's particularly relevant, but the 5e spell descriptions are probably contain the least fluff of any edition of the game. Infer from that what you will.

Ah, so you agree that the spell descriptions include fluff. This is why there is a meaningful distinction between being merely a book of rules and being more than just a book of rules - the text is serving multiple purposes.

This is the same logical error that you continue to make. (If A then B implies if not A then not B. This is incorrect.)

This is incorrect in formal logic, but quite correct in a variety of situations. I am not making a logical error, I am denying that the rules are written in formal logic.

In fact, there's a nice Latin phrase for the situation: expressio unius est exclusio alterius. This is a canon of textual interpretation which means that to express one thing is to exclude another. Here, by expressing the situation in which Mirror Image has a mechanical effect, the rules exclude other situations.

It's pretty simple: Even if mirror image is capable of thwarting save-spell attacks, this doesn't increase it's power to the point that I consider "broken." Most characters have an alternative (weapon attacks, attack spells) that will work as expected given the intended function of the spell, and some have the ability (AOEs) to can ignore it completely. There is no real is of the spell ever thwarting more than one save-spell, because people will learn how to deal with the spell pretty quickly. I just don't think it has anything near the "broken" levels of power that have been ascribed to it. Even magic missile is still pretty effective by my ruling.

Why is magic missile effective by your ruling? Doesn't it have a chance to hit mirror images and thereby do nothing?

The fact that your interpretation can still be effectively solved with smart tactics doesn't mean that it isn't still much more powerful than other defensive buffs at second level. Also, it can combine with other effects to make a very powerful defensive character - Arcane Tricksters aren't afraid of area of effect spells, and with high Dex scores can keep their images alive too...

I did: "[Y]our version is superior to the PHB version of the spell."

In what way? Feel free to explain this.

bardo

2016-08-19, 12:27 AM

Just to get one thing out of the way. PHB Chapter 9 combat page 194 says: "If there's ever any question whether something you're doing counts as an attack, the rule is simple: if you're making an attack roll you're making an attack."

There is no attack without an attack roll. There is no attack roll without an attack.

If you only read the last part you might get the impression it's a logical P implies Q situation (which would allow for other methods of making an attack without making an attack roll). It's not. It's written with the declared purpose of resolving Q (and we know P implies Q does not resolve Q), so logically it is Q if and only if P. In plain English: No attack without an attack roll. No attack roll without an attack.

Mirror Image mechanics says "Each time a creature targets you with an attack". Hold Person does not call for an attack roll. Therefore Hold Person does not count as an attack. Therefore Hold Person has no mechanical interaction with Mirror Image.

But what about the first paragraph with the duplicates and the shifting the fading? Hold Person targets a humanoid creature that the caster can see within range. Target is humanoid? Check. Can see the target? Check. Target within range? Check. But the duplicates?!?!? The caster knows all the duplicates are just one creature. The caster can't pin-point the physical creature in the shifting illusion but the caster doesn't need to pin-point anything. Pin-pointing things is for people making an attack, which Hold Person is not. As far as the caster is concerned the mirror images are definitely just one creature. Humanoid. Within Range. Roll saving throw.

Bardo.

Segev

2016-08-19, 09:27 AM

'You try to hit Joe, but it turns out he was a little to the left of where you thought he was. You instead cause an image of Joe to fade away as you slash it's arm.' I really don't see why that's so unreasonable for 'no, they're not independently targetable.'

So, your answer is, "No, you can't point to an image and say 'I swing at that one,' but you can hit one and only one image when you attack the cluster of them?"

That is self-contradictory. It's like saying you can't pick out one card from the deck because they're all neatly piled on top of each other, but when you jab a stiletto into the deck, you will hit only one of the cards.

Honestly, that sounds like a premise for a cool close-up magic trick, but it should be obvious why that is rather nonsensical.

It is for this reason that I stand by my post to which you were responding: If you can individually identify the images (even if you can't track which is which) such that, in a freeze-frame moment, you can point to one of them and say "that one," then they're individual targets. That's...kind of the definition. It's the whole reason the spell works, in theory. (In practice, of course, you can make arguments that it works because mechanics and rules say it does, but I reject such 3.5-style arguments in the 5e forum, at least when they require that you ignore what the spell's DOING in order to have their mechanical effect.)

If you cannot individually identify the images, such that it's impossible to point to one, even with a freeze-frame moment, and say "that one, but not those others," then there is no way to swing your sword or stab your spear such that you will hit one and only one of the images. It's too intermingled with the others; you'd hit multiple and probably all of them.

The reason blur works differently is because it literally is expanding the apparent space the single creature occupies. You can point to the blur and say "that creature." And you're talking about the blurred creature. With mirror image, if you point to an image and say "that creature," you're talking about a specific image.

You don't get to point to John and Joe standing shoulder to shoulder and say "I target Joe" and expect the spell to pick which of the two is Joe and which of the two is John.

RickAllison

2016-08-19, 10:01 AM

bardo

2016-08-19, 10:46 AM

The reason blur works differently is because ...

Nope. Blur and Mirror Image are quite similar. They both create a visual effect that hinders attack rolls against the caster. Neither one of them hinders casting spells against the caster (unless the spell calls for an attack roll).

Maybe "target" is a poor choice of word to describe spells because it fools you into thinking casting a spell involves aim. But no, there's no aiming involved. Spells that involve aiming would specifically call for a spell attack roll. Not the case with Hold Person.

I don't see the relevance of asking can or can't you target an individual mirror image, but anyway the answer is no. A creature surrounded by mirror images is still just one creature. You target it wholesale because there are no sub-targets in 5e. You can't say "I aim for the mirror image on the left" any more than you can say "I am for the eyes". I mean, you can say that if you think it's fun, it just doesn't affect how your action will be resolved.

Bardo.

smcmike

2016-08-19, 10:54 AM

Nope. Blur and Mirror Image are quite similar. They both create a visual effect that hinders attack rolls against the caster. Neither one of them hinders casting spells against the caster (unless the spell calls for an attack roll).

Maybe "target" is a poor choice of word to describe spells because it fools you into thinking casting a spell involves aim. But no, there's no aiming involved. Spells that involve aiming would specifically call for a spell attack roll. Not the case with Hold Person.

I don't see the relevance of asking can or can't you target an individual mirror image, but anyway the answer is no. A creature surrounded by mirror images is still just one creature. You target it wholesale because there are no sub-targets in 5e. You can't say "I aim for the mirror image on the left" any more than you can say "I am for the eyes". I mean, you can say that if you think it's fun, it just doesn't affect how your action will be resolved.

Bardo.

Your reasoning is so nice and clear. Get out of this thread while you still can.

bardo

2016-08-19, 11:02 AM

Grapples and Shoves also are described as special melee attacks, but it seems they actually set up how the book treats attacks that lack attack rolls, where they are explicitly called out as such. So while it is possible to have attacks without attack rolls, they must be called out as such by the text.

Good catch. So more precisely: The general rule is no attack without an attack roll and no attack roll without an attack. Then a specific rule can say this specific action is an attack without an attack roll, or say this specific attack roll is not an attack. So as always, specific rule over-rides general rule.

No doubt on Hold Person, though. It's not an attack.

Bardo.

Tanarii

2016-08-19, 11:20 AM

So, your answer is, "No, you can't point to an image and say 'I swing at that one,'That is correct. Per RAW, you cannot point to an image and say 'I swing at that one'.

You must point to the creature, say 'I swing at him', and then the spell redirects you to swinging at an image instead.

Edit: I'm assuming 'swing at' is short hand for making an attack in this case.

Segev

2016-08-19, 11:43 AM

Nope. Blur and Mirror Image are quite similar. They both create a visual effect that hinders attack rolls against the caster. Neither one of them hinders casting spells against the caster (unless the spell calls for an attack roll).You're presuming your conclusion, here. Whether mirror image hinders targeting the caster with a spell is specifically what is under contention.

Maybe "target" is a poor choice of word to describe spells because it fools you into thinking casting a spell involves aim. But no, there's no aiming involved. Spells that involve aiming would specifically call for a spell attack roll. Not the case with Hold Person.No, I'm not assuming any sort of aiming. I'm assuming that you have to be able to state unambiguously who or what is going to be affected by the spell, and do so without presuming the spell has knowledge about the entity so affected that you or your PC lack.

I don't see the relevance of asking can or can't you target an individual mirror image, but anyway the answer is no. A creature surrounded by mirror images is still just one creature. You target it wholesale because there are no sub-targets in 5e. You can't say "I aim for the mirror image on the left" any more than you can say "I am for the eyes". I mean, you can say that if you think it's fun, it just doesn't affect how your action will be resolved.You aren't aiming for "the mirror image" in the sense of the spell. You're selecting one of 4 images to affect. If you use "an attack" to do it, mirror image specifies how you pick between the 4 images to determine if you hit the real one or not. If you use something that is not "an attack," it provides no overriding rules; the DM must figure out how to make the determination despite it being impossible to track which image is the real one.

"A creature surrounded by mirror images is still just one creature" is no more or less true than "a creature surrounded by 3 silent images of itself is still just one creature." And since Bob, who is casting hold person, cannot tell which of the 4 images of Joe (who is either under mirror image or has 3 friends maintaining silent images of him that behave exactly like mirror image would), also cannot tell if it's mirror image, silent image, or really talented Joe-impersonators doing a complicated dance to make it impossible to track which is which, my contention is that his hold person spell is no more capable of making that determination.

Bob casts hold person on one of the "Joe"s he sees, under all the cases above. He doesn't get to say "I target Joe" and expect the spell to know which one of them is Joe if he does not.

That is correct. Per RAW, you cannot point to an image and say 'I swing at that one'.Which is only true if you are actively making "an attack." If you're just trying to tap him on the shoulder or point at him in a line-up, the rules you're referencing don't apply. In fact, you're hinging your "spells ignore mirror image" argument on the fact that those rules don't apply except and unless you are actively making "an attack."

Let me try to set up a hypothetical to really illustrate what I am talking about:

Let's say that I have a magically talented painter capable of instant, photorealistic paintings (or, alternatively, I have a modern camera with whatever degree of shutter speed is required to take a perfect still-shot). I use the services of this painter (or camera) to obtain an instantaneous picture of Joe while he's under the effects of mirror image.

It is my contention that I could take my finger and point at the 4 "Joe"s in this picture, and identify them individually as "A," "B," "C," and "D." If I had a second picture, an arbitrarily short amount of time after, I could do the same...but I could in no way be certain that I had labeled the same images with the same letters.

Can we at least agree that this is unambiguously true from the description of what mirror image does? Or are we so fundamentally in disagreement that this is not so, to you?

BurgerBeast

2016-08-19, 11:53 AM

Let me try to set up a hypothetical to really illustrate what I am talking about:

Let's say that I have a magically talented painter capable of instant, photorealistic paintings (or, alternatively, I have a modern camera with whatever degree of shutter speed is required to take a perfect still-shot). I use the services of this painter (or camera) to obtain an instantaneous picture of Joe while he's under the effects of mirror image.

It is my contention that I could take my finger and point at the 4 "Joe"s in this picture, and identify them individually as "A," "B," "C," and "D." If I had a second picture, an arbitrarily short amount of time after, I could do the same...but I could in no way be certain that I had labeled the same images with the same letters.

Can we at least agree that this is unambiguously true from the description of what mirror image does? Or are we so fundamentally in disagreement that this is not so, to you?

I think we are more or less here in the discussion.

Another thing that needs to be resolved is any argument for the other side that claims that being in the same space is somehow relevant, and why. It seems to me that at least some of the arguments for why spells do not have to change spell casters do not have to select targets depends on closeness/sameness of Joe and his duplicates, which I reject.

[edit: additionally, but not necessarily related, the "if there is an attack roll, then it is an attack" thing does not in any way exclude the possibility that an attack might not use attack rolls. I totally missed this, and even though it has been used like crazy on this forum, and apparently even invoked by Crawford as justification for rulings, it is actually not what is said in the rules. This has basically shaken the entire foundation of the debate, and left me pretty rattled. In this context, magic missile and other save-spells are not determined to be attacks nor specifically determined to "not be attacks."]

*** reserved space for individual responses which I will add through edits, and try (but probably fail - sorry - to keep shorter and to the point ***

Currently, we disagree about how to read the mirror image spell.

For the purpose of showing inconsistencies in your reading, I invoked the dwarf/elf example. It seems that we are currently at: I contend that my elf/dwarf example is relevant. You contend that it is not. I have given my reasons. You haven't refuted them. There doesn't seem to be much point in moving forward.

Currently, it appears to me that your stance on Mirror Image is that the spell does precisely what the mechanics say because the spell says so, and the general function of the spell doesn't need to make sense because "magic."

You have conceded that in he context of the dwarf/elf example, your explanation does not make sense.

Yet in the original example (the mirror image text in the PHB), you seem to apply the exact same logic but claim you explanation does make sense - a completely different conclusion.

It's the arriving at different conclusions using the same logic that bothers me. This indicates an invalid ("invalid" in the strict philosophical sense) argument.

Tanarii

2016-08-19, 12:07 PM

Which is only true if you are actively making "an attack." If you're just trying to tap him on the shoulder or point at him in a line-up, the rules you're referencing don't apply. In fact, you're hinging your "spells ignore mirror image" argument on the fact that those rules don't apply except and unless you are actively making "an attack."Yes, it's only true if you target with an attack.

I assume that in the case of a non-attack, the normal targeting rules apply: You target the creature. Why? Because nothing says otherwise. The mirror image spell doesn't specify that. You have to make assumptions based on the Mirror Image description of how the spell works, and then further assumptions of how the spell's method of working appears in-game. (Edit: and then work backwards from those assumption to change the default targeting rules. That is simulation ruling: interpreting a rule based on how seperate rule is interpreted to resolve in-game.)

Let's say that I have a magically talented painter capable of instant, photorealistic paintings (or, alternatively, I have a modern camera with whatever degree of shutter speed is required to take a perfect still-shot). I use the services of this painter (or camera) to obtain an instantaneous picture of Joe while he's under the effects of mirror image.

It is my contention that I could take my finger and point at the 4 "Joe"s in this picture, and identify them individually as "A," "B," "C," and "D." If I had a second picture, an arbitrarily short amount of time after, I could do the same...but I could in no way be certain that I had labeled the same images with the same letters. All of your contention is an assumption on your part about how the spell appears in the in-game world. The text of the spell does not require it in any way.

Can we at least agree that this is unambiguously true from the description of what mirror image does? Or are we so fundamentally in disagreement that this is not so, to you?I don't disagree that you can choose to interpret the spell as appearing the way you describe, in-game. I disagree that the text of the spell requires it.

Mellack

2016-08-19, 12:13 PM

You aren't aiming for "the mirror image" in the sense of the spell. You're selecting one of 4 images to affect. If you use "an attack" to do it, mirror image specifies how you pick between the 4 images to determine if you hit the real one or not. If you use something that is not "an attack," it provides no overriding rules; the DM must figure out how to make the determination despite it being impossible to track which image is the real one.

This is where I think people disagree. The spell specifies that you are only redirected if you target the caster. By your overriding rules, you would first have to chose which you are going to attack. Only if you are lucky enough to chose the right one, then you have another step where the spell may change it. That seems to be against the expected power of the spell. I have not seem a justification for why you think the attack gets to start by targeting that caster while a save spell does not.

smcmike

2016-08-19, 12:25 PM

Here is the rule about whether a spell includes an attack: if the spell says "make a attack," it includes an attack. No further inquiry required.

smcmike

2016-08-19, 01:02 PM

For the purpose of showing inconsistencies in your reading, I invoked the dwarf/elf example. It seems that we are currently at: I contend that my elf/dwarf example is relevant. You contend that it is not.

Currently, it appears to me that your stance on Mirror Image is that the spell does precisely what the mechanics say because the spell says so, and the general function of the spell doesn't need to make sense because "magic."

I think I see the issue here, and I don't blame you for this reading of my argument.

I do think a spell should make some sense - if there is no way to translate the rules into some sort of simulationist narrative, it's a bad spell.

The rules of your elf/dwarf spell are significantly harder to translate into a simulationist narrative than the rules of the original, because you wanted to prove a point. It's a bad spell by design.

That doesn't mean that its rules work differently, though, only that there is a fairly serious mismatch between rules and fluff. Here's a counterexample for you:

Mike's Mysterious Attack

You hurl a bubble of acid. Make a ranged spell attack against a creature within range. On a hit, the target takes 1d10 fire damage.

This is a terrible spell. It is very hard to make any sense of it within the simulationist narrative. On the other hand, the rules are very clear - you hurl acid, and do fire damage.

I don't think that Mirror Image is a bad spell. I have little problem reconciling the rules, as I interpret them, with the simulationist narrative.

The way I read your elf/dwarf argument, you are arguing that the spell's effects don't make sense - that they make it hard to construct a simulationist narrative. While I disagree, even if you were to convince me, you would only have shown that it is a bad spell, not that it works differently.

This is why I say your elf/dwarf argument is not relevant.

Yet in the original example (the mirror image text in the PHB), you seem to apply the exact same logic but claim you explanation does make sense - a completely different conclusion.

It's the arriving at different conclusions using the same logic that bothers me. This indicates an invalid ("invalid" in the strict philosophical sense) argument.

I hope you understand how I arrived at different conclusions - same chain of logic, but different inputs. There is vey little difference between an elf and a dwarf upon which to create a simulationist narrative that makes sense. There is a great deal of difference between shooting a bow at someone and casting Hold Person at them.

Segev

2016-08-19, 01:58 PM

All of your contention is an assumption on your part about how the spell appears in the in-game world. The text of the spell does not require it in any way.

I don't disagree that you can choose to interpret the spell as appearing the way you describe, in-game. I disagree that the text of the spell requires it.

The spell explicitly states that there are 3 illusory images of Joe when he casts mirror image. This is indisputable.

Please tell me how there can be 3 illusory images of Joe (for a total of 4 images), but my magical insta-painting does not allow me to point to each of the 4 images of Joe in the picture and label them independently.

Specifically, I am looking for you to explain to me how there can be 4 images of Joe in that painting but I cannot point to each of them and say "here's one image I'm labeling 'A,' here's another image I'm labeling 'B,' here's a third image I'm labeling 'C,' and here's a final image of Joe I'm labeling 'D.'"

You must satisfy both that there ARE 4 images (one real, 3 illusory), and yet that I cannot point to any of them and differentiate them in that single moment in time by their visible displacement from each other.

georgie_leech

2016-08-19, 02:06 PM

The spell explicitly states that there are 3 illusory images of Joe when he casts mirror image. This is indisputable.

Please tell me how there can be 3 illusory images of Joe (for a total of 4 images), but my magical insta-painting does not allow me to point to each of the 4 images of Joe in the picture and label them independently.

Specifically, I am looking for you to explain to me how there can be 4 images of Joe in that painting but I cannot point to each of them and say "here's one image I'm labeling 'A,' here's another image I'm labeling 'B,' here's a third image I'm labeling 'C,' and here's a final image of Joe I'm labeling 'D.'"

You must satisfy both that there ARE 4 images (one real, 3 illusory), and yet that I cannot point to any of them and differentiate them in that single moment in time by their visible displacement from each other.

In point of fact, the wording in my book is 'shifting position so it's impossible to track which image is real.' Which means that movement of the images is what makes them impossible to track. So if you took a single moment in time, the spell absolutely does not say they remain impossible to tell apart; the fact that the images move is a crucial part to them being impossible to identify specifically.

BurgerBeast

2016-08-19, 02:18 PM

As I said earlier, Hold Person allows you to target Humanoids. I still stand by this because Burger completely failed to refute it. (Side Note: If he cites the book or SA regarding this I may be willing to change my mind, but it's unlikely) If Hero believes in the Illusion, then there would be nothing preventing him from making the attempt.

Just to be clear, I did not ever refute that hold person can target humanoids. I refuted that illusions can be considered humanoids. Thus, I never contended that hold person can target an illusion. I contended that hold person can take effect on an illusion.

Nonetheless. I did not provide any evidence that illusions are not creatures. This is pretty important and I'm surprised that no one seems to care.

All of the images here still occupy the same 5-foot space, and therefore anything that is targeting one is capable of targeting all.

Does this mean that if four tiny monsters are in one square, you can target them all with an attack that is capable of targeting one?

Just to get one thing out of the way. PHB Chapter 9 combat page 194 says: "If there's ever any question whether something you're doing counts as an attack, the rule is simple: if you're making an attack roll you're making an attack."

There is no attack without an attack roll. There is no attack roll without an attack.

If you only read the last part you might get the impression it's a logical P implies Q situation (which would allow for other methods of making an attack without making an attack roll). It's not. It's written with the declared purpose of resolving Q (and we know P implies Q does not resolve Q), so logically it is Q if and only if P. In plain English: No attack without an attack roll. No attack roll without an attack. (emphasis added)

I agree that what they meant was if and only if. Unfortunately it's not what they said. Even taking it all in context, "if and only if" is not what it says. It is an interpretation. But no one has ever contested this in this thread.

Mirror Image mechanics says "Each time a creature targets you with an attack". Hold Person does not call for an attack roll. Therefore Hold Person does not count as an attack. Therefore Hold Person has no mechanical interaction with Mirror Image. (emphasis added)

This is what I tried to argue earlier. The problem is, what is the general case for not interacting with mirror image? Does not interacting mean that you do not see the duplicates, for example?

The first paragraph of mirror image is not conditional. It applies always. Thus, the first paragraph applies to hold person, in the sense that the caster of hold person sees four potential targets. The caster of hold person must always choose a target.

But what about the first paragraph with the duplicates and the shifting the fading? Hold Person targets a humanoid creature that the caster can see within range. Target is humanoid? Check. Can see the target? Check. Target within range? Check. But the duplicates?!?!? The caster knows all the duplicates are just one creature.

Knowing they are all one creature does not help in deciding which of the four is the real one.

The caster can't pin-point the physical creature in the shifting illusion but the caster doesn't need to pin-point anything. Pin-pointing things is for people making an attack, which Hold Person is not.

Targeting is required of both spells and attacks. If there were four tiny monsters in the space, then the cater would indeed need to pin-point one of the monsters as a target for hold monster.

As far as the caster is concerned the mirror images are definitely just one creature. Humanoid. Within Range. Roll saving throw.

If it's as simple as "Humanoid. Within Range. Roll saving throw." Then what if there are four humanoids within range?

----------

Nope. Blur and Mirror Image are quite similar.

Evidently not similar enough to employ the same mechanic. Why do you suppose that is?

They both create a visual effect that hinders attack rolls against the caster. Neither one of them hinders casting spells against the caster (unless the spell calls for an attack roll).

If this is your standard, then sure, they are similar. They use completely different mechanics, though, which makes them not very similar. The description of their effects is different, which makes them not very similar.

Maybe "target" is a poor choice of word to describe spells because it fools you into thinking casting a spell involves aim. But no, there's no aiming involved. Spells that involve aiming would specifically call for a spell attack roll. Not the case with Hold Person.

Nobody is bothered about aim. You have to "pick a target." If there is more than one potential target within range, you must pick one.

I don't see the relevance of asking can or can't you target an individual mirror image, but anyway the answer is no.

Well, there's you answer, then. I say the answer is "yes." If the answer is "no," you've got some explaining to do.

A creature surrounded by mirror images is still just one creature.

Agreed. And the three illusory duplicates are still three illusory duplicates.

You target it wholesale because there are no sub-targets in 5e. You can't say "I aim for the mirror image on the left" any more than you can say "I am for the eyes". I mean, you can say that if you think it's fun, it just doesn't affect how your action will be resolved.

You can, actually. You can choose to target a humanoid, the item in his hand, or the magical effect that surrounds him when casting dispel magic, for example. And it does affect how the action is resolved.

----------

Good catch. So more precisely: The general rule is no attack without an attack roll and no attack roll without an attack. Then a specific rule can say this specific action is an attack without an attack roll, or say this specific attack roll is not an attack. So as always, specific rule over-rides general rule.

No doubt on Hold Person, though. It's not an attack.

In this thread, no one is suggesting that hold person is an attack.

No, we do not disagree about the definition of ambiguous. As you describe below, we disagree about how much meaning can be placed into a particular set of statements. Suggesting that this is also a definitional dispute smells of cheap point-scoring.

Well, you said:

I agree that you can't completely rule out the possibility that the speaker is working Saturday morning. The statement is ambiguous. (emphasis added)

And I say that is an incorrect use of the word. So, we do.

This sums up our disagreement in its entirety. Language comes from common usage.

Where it comes from has little bearing on using it properly. Grammar and syntax are real things. If you use them incorrectly, you are using the language incorrectly, regardless of ability to convey meaning (which is completely unreliable since the meaning conveyed will change based on the listener and the time history).

If language does what the speaker wants it to - for example, communicate something to someone, it is correct.

Not correct, useful. They're different. It's possible to do things incorrectly and still achieve the end-goal. Also, in what you have presented here, the "correct" use of language is made contignent on what the listener thinks. There can be multiple listeners, so by your reasoning, one statement can have a million meanings if there are a million listeners. And they can all be correct.

Do you really think Crawford is drawing his conclusion based on corn flakes? He said mirror image has no effect on magic missile, because it doesn't involve an attack. You are being silly.

No, you are. My point remains: just because Crawford and RickAllison reach the same conclusion, this does not mean that their reasons for reaching that conclusion were the same. "ecause it doesn't involve an attack" is all that was said by Crawford. Rick Allison claimed that much more of his interpretation was justified by Crawford.

----------

Again: why isn't the dwarf/elf example relevant?

----------

Ah, so you agree that the spell descriptions include fluff.

I never contended it. I contended that specific examples are or are not purely fluff.

This is why there is a meaningful distinction between being merely a book of rules and being more than just a book of rules - the text is serving multiple purposes.

I never said "merely a book of rules." I said "a book of rules." Just like if I say I am a teacher, that does mean that I am "merely a teacher." You added meaning.

This is incorrect in formal logic, but quite correct in a variety of situations. I am not making a logical error, I am denying that the rules are written in formal logic.

Denying that the rules are written in formal logic is a logical error. Writing in English does fall outside of logic - not even formal logic.

In fact, there's a nice Latin phrase for the situation: expressio unius est exclusio alterius. This is a canon of textual interpretation which means that to express one thing is to exclude another.

It doesn't apply here, at all.

Here, by expressing the situation in which Mirror Image has a mechanical effect, the rules exclude other situations.

This is completely illogical. And "expressio unius est exclusio alterius" does not apply. It can't be used to determine the context in which things apply. It can be used to claim that melee attacks are not ranged attacks, for example. It cannot be used to claim that since melee attacks apply, therefore ranged attacks do not apply. If it worked as you are using here, it would be useless.

Why is magic missile effective by your ruling?

Because upcast as a level 2 spell, it could be used to simultaneously strike all four targets. Thus, with one action and one expenditure of a 2nd level slot, the effect would be entirely negated and Joe would take 1d4+1 damage.

----------

Again: why isn't the dwarf/elf example relevant?

It seems to me that you are of the opinion that the mirror image text (paragraphs 2-4) come into effect only once Joe is targeted with an attack. So how is Joe targeted? Presented with four possible targets, Hero must choose one before Joe can be chosen. And the act of choosing raises the possibility that a duplicate is chosen, in which case the spell will not trigger.

If that is so, then how does one target Joe with an attack? - this is the same complaint that Mellack raised.

It seems to me that your way of interpreting this would lead to double-defense.

----------

right, but for an attack, you target the creature first, and it is possible redirected. At which point you determine a hit or miss by the normal rules.

No, for an attack, if you target the creature, it is redirected. You still have to target the creatures, and there are four choices.

What's important here is that an attack (at least) targets the creature. Initially. it's explicit.

No, it's not. It's only explicit that if you target the creature, it is redirected.

Now, I agree it's an extension to say that this, which is the default rule for targeting anyway, also applies to non-attacks. But since it [B]is the default rule to target the creature under normal circumstances, that makes sense to me.

It is not the default to target the creature. It's the default for the player to choose among all possible targets.

----------

That is correct. Per RAW, you cannot point to an image and say 'I swing at that one'.

This is contended. When there are four potential targets, the player must pick one, and mirror image does not change this.

You must point to the creature, say 'I swing at him', and then the spell redirects you to swinging at an image instead.

Not only is it the case that you must target the creature, you can't target the creature. It's not possible.

You interpret the text to imply that you must target the creature, but it does not say that. It only says what happens if you target the creature.[/QUOTE]

----------

I assume that in the case of a non-attack, the normal targeting rules apply: You target the creature.

The normal targeting rules are not: You target the creature. The normal targeting rules are: pick a creature to target.

Why? Because nothing says otherwise. The mirror image spell doesn't specify that.

We agree here. We just agree on what the default targeting rules are.

You have to make assumptions based on the Mirror Image description of how the spell works, and then further assumptions of how the spell's method of working appears in-game. (Edit: and then work backwards from those assumption to change the default targeting rules. That is simulation ruling: interpreting a rule based on how seperate rule is interpreted to resolve in-game.)

No, you do not need to make any assumptions. You do not to work backward. This is bad reasoning, generally. This particular case is no exception.

All you need to do is read the spell and apply yeah rule as presented.

All of your contention is an assumption on your part about how the spell appears in the in-game world. The text of the spell does not require it in any way.

This is a misrepresentation of my contention.

I don't disagree that you can choose to interpret the spell as appearing the way you describe, in-game. I disagree that the text of the spell requires it.

The text of spell does not require our precise description. But the text of the spell does require that the duplicates are discrete and targetable.

You cannot change the spell so that the duplicates are no longer discrete and targetable, and still claim to be within RAW.

What you might do, for example, is claim that the images are discrete and targetable, but the spell overrides the ability of attacks to target the duplicates.

This prevents obvious problems, though, and I suspect this is why some people decided to ignore the "discrete and targetable" requirement. Ignoring these stipulations in favour of new "fluff" is where they violated RAW.

bardo

2016-08-19, 02:23 PM

You're presuming your conclusion, here. Whether mirror image hinders targeting the caster with a spell is specifically what is under contention.

...

Can we at least agree that this is unambiguously true from the description of what mirror image does? Or are we so fundamentally in disagreement that this is not so, to you?

I think I explained my conclusion rather than presumed it. Here we go again, no assumption left undeclared. Let me know if any of these sound incorrect to you.

A spell description states what a spell does.
A spell does what the spell description says it does.
Some spell descriptions add details about things the spell DOES NOT DO, but we never expect a spell description to list anything (and certainly not everything) that the spell DOES NOT DO.

An proposal along the lines of maybe a spell also does X has to be supported by more than just "because the description doesn't explicitly say it doesn't do X". For example, it would make no sense to suggest maybe Create Water can also create coffee, just because the spell description doesn't explicitly rule out coffee. There has to be something more to support the case or else we're dealing with nonsense.

Now to the case at hand.

Mirror Image describes how the spell affects the appearance of the protected creature. And Mirror Image describes a d20 procedure for resolving attacks against the protected creature.

So we have two matters to discuss. Whether the change in appearance affects casting Hold Person at the protected creature. And whether the d20 procedure applies when casting Hold person at the protected creature. I'll take the second one first, because it's dead easy.

Mirror Image states the d20 procedure is triggered by attacks against the protected creature.
General rule says it's an attack if there's an attack roll.
Hold Person doesn't call for an attack roll.
Hold Person is not an attack.
Conclusion: The d20 procedure described in Mirror Image does not apply when casting Hold Person at a creature protected by Mirror Image.

Now, does the change in appearance affects casting Hold Person at the protected creature?

There is only one general rule about targeting creatures/objects with spells: You can't target a creature/object that has total concealment. The appearance change from Mirror Image does not constitute total concealment.

Hold Person describes three requirements for the spell to work: The creature must be humanoid, the caster must be able to see the creature, and the creature must be in range.

Mirror Image doesn't take the protected creature out of range.
Mirror Image doesn't change the protected creature's type.
Mirror Image doesn't make the caster unable to see the protected creature (as for example Invisibility would). It makes the caster also see duplicates, but the caster still sees the protected creature. The conversation on this item has drifted to scenarios where the caster is fooled into thinking there is more than one creature, but that's another discussion altogether. Mirror Image create additional creatures or fool anyone into thinking there are additional creatures. Everybody knows it's just the protected creature.

I've seen nothing to support Hold Person wouldn't work as usual when targeting a creature protected by Mirror Image. No rule is broken. No requirement goes unfulfilled.

As noted above, the fact Mirror Image doesn't explicitly explain how spell targeting works for the protected creature doesn't mean we need to weave theories, it means we follow the usual rules, and the usual rules say wisdom saving throw.

Bardo.

Tanarii

2016-08-19, 02:38 PM

You must satisfy both that there ARE 4 images (one real, 3 illusory), and yet that I cannot point to any of them and differentiate them in that single moment in time by their visible displacement from each other.It says there are duplicates. They are used in a mechanical fashion individually. Which I'm gathering others wish to interpret as making them 'discrete', although I don't think it does, it doesn't really matter as long as we're on the same page as to where they are individual and where they are not.

And it does NOT say they must be visually identifiable as separate targets at any given point in time.

Joe dirt

2016-08-19, 02:39 PM

The entire reason to cast an illusion is to misdirected an attack.... including spells

smcmike

2016-08-19, 02:42 PM

Regarding Magic Missile:

Because upcast as a level 2 spell, it could be used to simultaneously strike all four targets. Thus, with one action and one expenditure of a 2nd level slot, the effect would be entirely negated and Joe would take 1d4+1 damage.

This is interesting. In the interest of clarity, when you say "negate" the spell, you don't mean that the missiles destroy the duplicates, do you?

Wouldn't it be at least as accurate to say that your version of Mirror Image "negates" the effect of 3/4 of magic missiles? This sounds pretty good to me.

smcmike

2016-08-19, 02:44 PM

The entire reason to cast an illusion is to misdirected an attack.... including spells

Many spells don't include an attack.

Why does the text of the spell only lay out rules regarding attacks?

BurgerBeast

2016-08-19, 02:47 PM

I think I see the issue here, and I don't blame you for this reading of my argument.

I do think a spell should make some sense - if there is no way to translate the rules into some sort of simulationist narrative, it's a bad spell.

The rules of your elf/dwarf spell are significantly harder to translate into a simulationist narrative than the rules of the original, because you wanted to prove a point. It's a bad spell by design.

Nope. If you contend that this is a bad spell design, then it follows that Mirror Image is a bad spell design.

The dwarf/elf fails because there is no good reason to think that a dwarf or elf would see the different things, or it fails because there is no good reason why an elf's attack or a dwarf's attack should target in different ways.

Likewise, mirror image fails because there is no good reason to think that an attacker or a ranged spell caster would see different things, or ir fails because there is no good reason to think that an attack or a ranged save-spell should target in different ways.

Either they both fail. Or they both succeed. (Also, you can;t fall back on the attack/spell difference as an out, because what is important here is the targeting - not what is doing the targeting. Mechanically, targeting is the same in either case: the player picks a target.)

That doesn't mean that its rules work differently, though, only that there is a fairly serious mismatch between rules and fluff.

Only if you contend that "three illusory duplicates appear in your space" is purely fluff. In which case we'd be back to that.

Here's a counterexample for you:

This counterexample is not even close to the same. It directly contradicts itself in its own simulationist narrative, and therefore must be wrong. This does not happen in mirror image. (and, by the way, expressio unius est exclusio alterius does apply here).

I don't think that Mirror Image is a bad spell. I have little problem reconciling the rules, as I interpret them, with the simulationist narrative.

Then you should have no problem reconciling the rules with the simulations narrative of the elf/dwarf spell. But apparently you do have trouble with that one.

The way I read your elf/dwarf argument, you are arguing that the spell's effects don't make sense - that they make it hard to construct a simulationist narrative.

The elf/dwarf example highlights that when the spell makes a specific mechanic for dwarf attacks, this has no bearing on elf attacks, and that the default case is presented to the elf: the elf has four targets to choose from but there are no rules to resolve the target, so the DM must resolve the target using some other means, which are natural in the situation.

The analogy to mirror image is that when the spell makes a specific mechanic for attacks, this has no bearing on save-spells, and that the default case is presented to the caster: the caster has four targets to choose from but there are no rules to resolve the target, so the DM must resolve the target using some other means, which are natural in the situation.

In the dwarf/elf example, only in the case where the elf auto-hits is the simulationist narrative destroyed. If the elf's attack requires a target to be selected, the simulationist narrative fits.

Likewise, with mirror image, only in the case where hold person auto-hits is the simulationist narrative destroyed. If hold person requires target selection, the simulationist narrative fits.

I hope you understand how I arrived at different conclusions - same chain of logic, but different inputs. There is vey little difference between an elf and a dwarf upon which to create a simulationist narrative that makes sense. There is a great deal of difference between shooting a bow at someone and casting Hold Person at them.

There is very little difference between an attacker picking a target and a spell caster picking a target. This is true mechanically and it is also true in a simulationist narrative. Unless you provide a new narrative in which it doesn't make sense.

In point of fact, the wording in my book is 'shifting position so it's impossible to track which image is real.' Which means that movement of the images is what makes them impossible to track.(emphasis added)

Difficult to track, yes. Not difficult to see.

So if you took a single moment in time, the spell absolutely does not say they remain impossible to tell apart; the fact that the images move is a crucial part to them being impossible to identify specifically.

I'm thinking this means you are agreeing with Segev, then?

Everything you've written here seems to be in agreement, but I thought you disagreed with him.

georgie_leech

2016-08-19, 03:06 PM

I was objecting to the requirement that any fluff explanation meet a standard the spell itself doesn't set. It would be similar to requiring a fluff description of Fireball to explain how it melts through everything in its path to create perfect spheres of destruction; that's not what the spell does.

I will frequently object to pieces of an argument without necessarily engaging with it in its entirety or restating my own position, if nothing in the quoted post changes what I've argued.

BurgerBeast

2016-08-19, 03:17 PM

I think I explained my conclusion rather than presumed it. Here we go again, no assumption left undeclared. Let me know if any of these sound incorrect to you.

A spell description states what a spell does.
A spell does what the spell description says it does.
Some spell descriptions add details about things the spell DOES NOT DO, but we never expect a spell description to list anything (and certainly not everything) that the spell DOES NOT DO.

An proposal along the lines of maybe a spell also does X has to be supported by more than just "because the description doesn't explicitly say it doesn't do X". For example, it would make no sense to suggest maybe Create Water can also create coffee, just because the spell description doesn't explicitly rule out coffee. There has to be something more to support the case or else we're dealing with nonsense.

Now to the case at hand.

Mirror Image describes how the spell affects the appearance of the protected creature. And Mirror Image describes a d20 procedure for resolving attacks against the protected creature.

No. Mirror image does not affect the appearance of the creature at all. It creates three illusory duplicates, which you have not mentioned.

So we have two matters to discuss. Whether the change in appearance affects casting Hold Person at the protected creature. And whether the d20 procedure applies when casting Hold person at the protected creature.[/QUOTE

No, we have to discuss whether three illusory duplicates affects casting Hold Person at the protected creature. And whether the d20 procedure applies when casting Hold person at the protected creature.

[QUOTE]I'll take the second one first, because it's dead easy.

Mirror Image states the d20 procedure is triggered by attacks against the protected creature.
General rule says it's an attack if there's an attack roll.
Hold Person doesn't call for an attack roll.
Hold Person is not an attack.
Conclusion: The d20 procedure described in Mirror Image does not apply when casting Hold Person at a creature protected by Mirror Image.

Agreed.

Now, does the change in appearance affects casting Hold Person at the protected creature?

Not the correct question, because mirror image does not change the appearance of the caster. The correct question is: Do three illusory duplicates affect the ability of the caster to select a target for hold person?

There is only one general rule about targeting creatures/objects with spells: You can't target a creature/object that has total concealment. The appearance change from Mirror Image does not constitute total concealment.

Agreed, since there is no appearance change at all.

Hold Person describes three requirements for the spell to work: The creature must be humanoid, the caster must be able to see the creature, and the creature must be in range.

But the general rules of spell casting say that a caster must select a target. There are four targets.

Mirror Image doesn't take the protected creature out of range.
Mirror Image doesn't change the protected creature's type.
Mirror Image doesn't make the caster unable to see the protected creature (as for example Invisibility would). It makes the caster also see duplicates, but the caster still sees the protected creature. The conversation on this item has drifted to scenarios where the caster is fooled into thinking there is more than one creature, but that's another discussion altogether.

The caster is not fooled. He knows that there is one wizard and three duplicates. But he doesn't know which are which. This does not override the rule that a caster must select a target. The caster must select one of the Joes to be the target of his spell.

Mirror Image create additional creatures or fool anyone into thinking there are additional creatures. Everybody knows it's just the protected creature.

True. But no one knows which of the four Joes is Joe.

I've seen nothing to support Hold Person wouldn't work as usual when targeting a creature protected by Mirror Image.

We agree here.

The problem is that this scenario has four targets. If you want to present the "usual case" as the case where there is only one possible target, then you are making a false comparison. The "usual case" to which you ought to compare is the case in which the caster sees four targets and wants to cast the spell. In this case, he picks one.

No rule is broken. No requirement goes unfulfilled.

The rule that you have broken is the rule written in the description of mirror image, which says that the spell creates three illusory duplicates. You broke this rule by changing it. You changes it in two ways:

1. Your version creates zero duplicates.
2. Your version changes the appearance of the caster.

These are rule changes.

As noted above, the fact Mirror Image doesn't explicitly explain how spell targeting works for the protected creature doesn't mean we need to weave theories, it means we follow the usual rules, and the usual rules say wisdom saving throw.

The usual rules say pick a target. If the target is a humanoid, it must make a saving throw.

It says there are duplicates. They are used in a mechanical fashion individually. Which I'm gathering others wish to interpret as making them 'discrete', although I don't think it does,

There are three. If you can say there are three, then they are countable. If they are countable they are discrete. Besides this, they can be struck independently of each other and of Joe. This means they must be discrete.

...it doesn't really matter as long as we're on the same page as to where they are individual and where they are not.

Apparently we are not. Three illusions could literally have identical appearances and occupy the exact same space. They would still be three illusions. Similarly, three ghosts can occupy the same space and still be three ghosts.

And it does NOT say they must be visually identifiable as separate targets at any given point in time.

No, this is just a consequence of being three different things in three different positions at any given point in time.

This is interesting. In the interest of clarity, when you say "negate" the spell, you don't mean that the missiles destroy the duplicates, do you?

Wouldn't it be at least as accurate to say that your version of Mirror Image "negates" the effect of 3/4 of magic missiles? This sounds pretty good to me.

Sorry, I was still reeling from the attacks thing. You're right. The missiles are would do d4+1 to the target but they would not destroy duplicates.

(I do now think that it is within RAW to consider magic missile an attack but this is not the context in which we arguing, and I specifically said so. But just so you know where I was coming from, at my table, magic missile would be considered an attack and would be able to destroy all three duplicates and strike the caster for 1d4+1 damage. This is why I was wrong.)

[edit: I think I would consider magic missile an attack at my table. There might be cases where I would not expect it to behave as an attack. Rather than saying "magic missile would be considered an attack at my table," I prefer to say, I am considering making magic missile "an attack" at my table.]

Many spells don't include an attack.

Why does the text of the spell only lay out rules regarding attacks?

Again, any answer to this is pure speculation. Segev and I have provided answers. The most likely answer in my opinion is that the writer of the spell considered spells such as magic missile to be attacks. The most likely explanation for this is that the writer didn't know that only things with attack rolls count as attacks.

georgie_leech

2016-08-19, 03:23 PM

Why is 'the writers didn't understand the rules' simpler than 'it doesn't work on non-attack spells?'

BurgerBeast

2016-08-19, 03:25 PM

Why is 'the writers didn't understand the rules' simpler than 'it doesn't work on non-attack spells?'

Who said it is simpler?

Tanarii

2016-08-19, 03:25 PM

No, this is just a consequence of being three different things in three different positions at any given point in time.No, it absolutely is NOT a consequence. You just insist on visualizing the in-game resolution of the description this way. And this is exactly what I mean when I say you're insisting on visualizing the in-game reality of the description a certain way. There is no such consequence required by RAW.

And since you say there is only one way to visualize it, and it is a consequence of RAW, and I'm insisting that visualizing it that way is not a consequence of RAW, we're at a stalemate.

Which is fine. I'm always satisfied to reduce an argument to the underlying point of disagreement. At least we know why we're disagreeing. :smallbiggrin:

georgie_leech

2016-08-19, 03:29 PM

Who said it is simpler?

Sorry, that's my own filters kicking in. Why is 'the writers didn't understand their rules' more likely than 'Mirror Image doesn't work against non-attack spells?'

bardo

2016-08-19, 03:34 PM

...

You know what, let's not talk about specific spells and please let's not talk about their interactions. Let's talk about spell targeting in general. So forget the spell Mirror Image and let's talk about ACTUAL MIRRORS.

Two casters are separated by long solid stone wall, 3 feet wide and 20 feet tall. 5 feet over the wall there's a gigantic mirror that enables the casters to see each other. Can they target each other with spells through the mirror?

Two other casters are standing in a hall of mirrors. They both know there isn't anybody else in the hall except the two of them. They both see hundreds of copies of both of them all around, but there's nothing separating them physically. Can they target each other with spells directly or through the mirrors? (we can skip the lol I cast fireball jokes).

Bardo.

smcmike

2016-08-19, 03:45 PM

You know what, let's not talk about specific spells and please let's not talk about their interactions. Let's talk about spell targeting in general. So forget the spell Mirror Image and let's talk about ACTUAL MIRRORS.

Two casters are separated by long solid stone wall, 3 feet wide and 20 feet tall. 5 feet over the wall there's a gigantic mirror that enables the casters to see each other. Can they target each other with spells through the mirror?

Two other casters are standing in a hall of mirrors. They both know there isn't anybody else in the hall except the two of them. They both see hundreds of copies of both of them all around, but there's nothing separating them physically. Can they target each other with spells directly or through the mirrors? (we can skip the lol I cast fireball jokes).

Bardo.

I really like this example, and used something similar somewhere in these endless threads.

One problem with your first version is that they do need line of effect.

The simplest version, which requires that we ignore the fact that there are no rules for facing -

A caster is standing, looking into a large mirror. He sees an enemy approach behind him. Is he able to target that enemy without turning around?

Segev

2016-08-19, 03:48 PM

In point of fact, the wording in my book is 'shifting position so it's impossible to track which image is real.' Which means that movement of the images is what makes them impossible to track. So if you took a single moment in time, the spell absolutely does not say they remain impossible to tell apart; the fact that the images move is a crucial part to them being impossible to identify specifically.I'm going to object to this in two distinct ways which I feel are both valid, but not necessarily dependent on each other:

1) You seem to be implying (if not outright stating) that if they are not shifting (as in my freeze-frame photograph), you can tell them apart. If you mean that it is now possible to say that a given one is the real Joe just by visual inspection, I must question what about their movement changes this. They are stated to be duplicates, so they must be visually mistakable for each other (presumably indistinguishable, but as others have objected to my use of that word, I will refrain). Are you contending that, if you freeze-frame them, the real one stands out obviously? How can this be so but not so if they're shifting?

2) The only reason they are "possible to track" while freeze-framed is that they are not moving in the frozen image/photograph/painting, so the one you label "A" is the same from moment to moment of that photo's existence. The actual moving images shift such that you cannot track which is which from moment to moment.

The nutshell is that I don't see how you can have them be impossible to track due to movement, but possible to tell apart by other than position if they're freeze-framed.

I think I explained my conclusion rather than presumed it. Here we go again, no assumption left undeclared. Let me know if any of these sound incorrect to you.Okay. I'll be quoting you in snippets as much as possible, though, just to cut down on clutter. Sorry if I miss something in my reply.

Mirror Image describes how the spell affects the appearance of the protected creature. And Mirror Image describes a d20 procedure for resolving attacks against the protected creature.

(...)

Mirror Image states the d20 procedure is triggered by attacks against the protected creature.
General rule says it's an attack if there's an attack roll.
Hold Person doesn't call for an attack roll.
Hold Person is not an attack.
Conclusion: The d20 procedure described in Mirror Image does not apply when casting Hold Person at a creature protected by Mirror Image.Your conclusion is not entirely accurate.

The correct way to read it, if you want to get pedantic about parsing the RAW, is that an attack invokes a specified d20 procedure to determine which image (real or fake) is attacked. The RAW require that the d20 method be the means for making this determination, in the event of an attack.

The RAW do not forbid the use of the d20 procedure to determine which image is the target chosen by the caster of hold person. They simply do not require that it be the means used, leaving the DM free to come up with any method, including the option to use the very same d20 procedure as with an attack.

Now, does the change in appearance affects casting Hold Person at the protected creature?

There is only one general rule about targeting creatures/objects with spells: You can't target a creature/object that has total concealment. The appearance change from Mirror Image does not constitute total concealment.

Hold Person describes three requirements for the spell to work: The creature must be humanoid, the caster must be able to see the creature, and the creature must be in range.

Mirror Image doesn't take the protected creature out of range.
Mirror Image doesn't change the protected creature's type.
Mirror Image doesn't make the caster unable to see the protected creature (as for example Invisibility would). It makes the caster also see duplicates, but the caster still sees the protected creature. The conversation on this item has drifted to scenarios where the caster is fooled into thinking there is more than one creature, but that's another discussion altogether. Mirror Image create additional creatures or fool anyone into thinking there are additional creatures. Everybody knows it's just the protected creature.

I've seen nothing to support Hold Person wouldn't work as usual when targeting a creature protected by Mirror Image. No rule is broken. No requirement goes unfulfilled.You're missing a requirement: that the caster specify a target.

You say it doesn't matter, because "everyone knows" it's just one creature. But that's just the point: mirror image makes it impossible to tell which is the real creature.

If you insist that mirror image's illusory duplicates are simply ignored by saying "I target Joe," then you must explain how that differs from claiming that John's disguise is ignored when you say "I target Joe." Or why the trivial fact that Bob doesn't know who the real murderer is out of all the suspects present should prevent Bob from saying "I target the real murderer."

Do you allow your players, when their PCs play the hide-the-ball-in-one-of-three-cups game in the town market, simply say, "I pick the cup with the ball under it?" Or do you require them to choose one of the cups and determine if they chose correctly?

Mirror image does what it says it does. The first thing it says it does is create illusory duplicates of the caster. When Bob's player says, "I target Joe with hold person," the DM and/or Joe's player is perfectly entitled to ask, "Which Joe?"

This is because, as you rightly note, mirror image doesn't specify that the d20 procedure must be used to determine if Bob chose Joe or an illusory duplicate of Joe. Some means of determination must be made, and the d20 method would be a valid choice, but it is not required because the RAW only make it required for "an attack."

It says there are duplicates. They are used in a mechanical fashion individually. Which I'm gathering others wish to interpret as making them 'discrete', although I don't think it does, it doesn't really matter as long as we're on the same page as to where they are individual and where they are not.

And it does NOT say they must be visually identifiable as separate targets at any given point in time.
So... it creates three illusory duplicates which move about such that you cannot distinguish which is the real caster and which are illusions, but somehow you cannot point to a still frame image of them and actually count 4 images?

Tanarii

2016-08-19, 03:56 PM

So... it creates three illusory duplicates which move about such that you cannot distinguish which is the real caster and which are illusions, but somehow you cannot point to a still frame image of them and actually count 4 images?
I actually find it easier to visualize them as NOT being able to take a still frame and actually count 4 images, than I do to visualize being able to count 4 images and not being able to keep track of them. Given they move about in such a way as to not be able to distinguish which is the real caster or keep track of them.

Either are perfectly valid as long as it matches the RAW requirement that attacks target the caster first, then are redirected to an image.

But if you require targeting to first pick a target caster/image for non-attacks, you should be consistent and require it for attacks. And then if they pick the caster, ALSO make the redirect check. (Edit for clarity: .. make the redirect check if it's an attack.)

Otherwise you just do the normal thing: target the caster. If it's an attack, you then invoke Mirror Image for a possible redirect.

Segev

2016-08-19, 04:36 PM

I actually find it easier to visualize them as NOT being able to take a still frame and actually count 4 images, than I do to visualize being able to count 4 images and not being able to keep track of them. Given they move about in such a way as to not be able to distinguish which is the real caster or keep track of them.

Either are perfectly valid as long as it matches the RAW requirement that attacks target the caster first, then are redirected to an image.But... it doesn't. There are 4 images.

I can think of literally no way for there to be 4 images - let alone have them be separate enough that you can strike just one - without being able to point at a still-frame of them and count 4 images.

Think about that again: You are essentially saying there are 4 images but that you can't count 4 images.

What it seems you're reduced to, here, is arguing that there aren't actually images, but that the spell still makes you hit an image that isn't there on a d20 roll with an attack. Not because you're fooled and unable to tell which image is the real one, but because...well, no reason. Mechanics say so. There's nothing happening in the game itself to justify it.

I'm really not trying to be dismissive, but I am unable to see how you're arguing for actually obeying the RAW of there being 3 illusory duplicates. It sounds like you're saying there really aren't for any purpose other than saying there are. Kind of like saying there's a spell which makes water dry. There's still water there, it doesn't have any different effects than normal "wet" water, and anything that touches it still acts in every way as if it were wet, including visually and tactilely, but the spell says it made it dry, so it's dry despite the fact that nobody can tell it's dry and it doesn't act dry. Oh, and people can walk on that water. But it's not water walk; it's dry water. Despite not doing anything related to water being dry.

But if you require targeting to first pick a target caster/image for non-attacks, you should be consistent and require it for attacks. And then if they pick the caster, ALSO make the redirect check. (Edit for clarity: .. make the redirect check if it's an attack.)Addressed multiple times already: the spell explicitly overrides this requirement for attacks, specifying how the selection of one of the four available targets works.

It does not explicitly override this requirement for non-attacks that still require target selection.

Otherwise you just do the normal thing: target the caster. If it's an attack, you then invoke Mirror Image for a possible redirect.You can target one of the things you think is the caster, sure. Which one do you target?

I know it seems like we're going around in circles, here, but that's because you're claiming you can ignore the requirement to pick a target in this one case, but not in others. Even if you won't acknowledge that's what you're doing; you've yet to spell out a difference other than "it's mirror image." Nowhere in mirror image does it say you treat mirror image's created choices differently from other cases where it's hard-to-impossible to tell which of multiple potential targets is the one you want. Well, no, it does, but only if you're "attacking." If you're not "attacking," it spells out no difference.

With no difference spelled out, you do, indeed, do it just like any other situation where you have to specify a target. There are 4 that could be "Joe." Pick one.

georgie_leech

2016-08-19, 04:43 PM

Addressed multiple times already: the spell explicitly overrides this requirement for attacks, specifying how the selection of one of the four available targets works.

Sorry, could you quote the exact text that explicitly overrides the targeting rules? My reading is that the spell changes what happens after being targeted by adding an extra step to resolving the attack, but I could easily be missing something. The d20 I use for Perception Checks seems to have an abnormal number of 1's on it.

Segev

2016-08-19, 04:55 PM

Sorry, could you quote the exact text that explicitly overrides the targeting rules? My reading is that the spell changes what happens after being targeted by adding an extra step to resolving the attack, but I could easily be missing something. The d20 I use for Perception Checks seems to have an abnormal number of 1's on it.

You're free to read it that way. It's a silly way to read it, and a slavish reading of the RAW in such a fashion leads to many other spells having similarly silly implementations, however. Intent is clear: the d20 procedure is a means of selecting which image is "really" targeted when somebody wants to attack the caster.

You're also free to say this doesn't override normal targeting, and instead read it as having an implied "when a creature [intends to] target[] you with an attack," as what is meant.

Maybe I'm injecting a house rule when I don't require you to first pick the right image to target with an attack before using the d20 procedure on top of it. If so, I think it a house rule that prevents a silly outcome. But I don't think I am. I think I'm reading it exactly as 5e intends rules to be read.

But hey, let's play this game in a different way. The rules require that Brad pick his target out of the 4 "Joe"s present. We still have to determine which one Brad chose. If we use the d20 procedure, and Brad turns out to have chosen the real Joe, we've already followed the RAW and used the d20 procedure to determine which one Brad hit. So it still works, in a convoluted self-referential sort of way.

Or we can read it straight-forwardly without trying to take the exception that expresses a specific resolution procedure as somehow making the non-exceptions to the general rule get special permission to ignore the need to pick a target and just be able to say "the one I really wanted to hit, even though I can't tell which one it is."

georgie_leech

2016-08-19, 05:09 PM

You're free to read it that way. It's a silly way to read it, and a slavish reading of the RAW in such a fashion leads to many other spells having similarly silly implementations, however. Intent is clear: the d20 procedure is a means of selecting which image is "really" targeted when somebody wants to attack the caster.

You're also free to say this doesn't override normal targeting, and instead read it as having an implied "when a creature [intends to] target[] you with an attack," as what is meant.

Maybe I'm injecting a house rule when I don't require you to first pick the right image to target with an attack before using the d20 procedure on top of it. If so, I think it a house rule that prevents a silly outcome. But I don't think I am. I think I'm reading it exactly as 5e intends rules to be read.

But hey, let's play this game in a different way. The rules require that Brad pick his target out of the 4 "Joe"s present. We still have to determine which one Brad chose. If we use the d20 procedure, and Brad turns out to have chosen the real Joe, we've already followed the RAW and used the d20 procedure to determine which one Brad hit. So it still works, in a convoluted self-referential sort of way.

Or we can read it straight-forwardly without trying to take the exception that expresses a specific resolution procedure as somehow making the non-exceptions to the general rule get special permission to ignore the need to pick a target and just be able to say "the one I really wanted to hit, even though I can't tell which one it is."

Nah, I'm totally fine with house ruling or interpretations of a spell that different from my own. I've just been discussing this from the perspective of what the RAW is, and how that works. For me, the spell adds a step to the resolution of attacks, wherein the attack can be redirected to one of the duplicates the spell creates. No more and no less. I think an especially RAW reading, if we take the first paragraph as impeding targeting in the first place, is that the spell has two layers of protection against melee attacks, but I don't think anyone is arguing that's intended, balanced, or otherwise how the spell should be run.

Part of my stake in this debate is getting a better understanding of my own position by having counterarguments presented to me, or finding the corner cases. For instance, a strict RAW of how I view Mirror Image and Whirlwind Attack separately would be that one attack gets made, which has the usual redirect chance. However, how I would actually run that in game would be to make a separate attack against all of the images against their 10+DEX AC, and an attack against the actual caster against their normal AC, because the fluff of Whirlwind attack implies to me that the user would be attacking all of the duplicates anyway. However the duplicates are imagined, that particular Ranger ability seems aimed at hitting everything around you regardless.

BurgerBeast

2016-08-19, 05:21 PM

No, it absolutely is NOT a consequence. You just insist on visualizing the in-game resolution of the description this way. And this is exactly what I mean when I say you're insisting on visualizing the in-game reality of the description a certain way. There is no such consequence required by RAW.

And since you say there is only one way to visualize it, and it is a consequence of RAW, and I'm insisting that visualizing it that way is not a consequence of RAW, we're at a stalemate.

Which is fine. I'm always satisfied to reduce an argument to the underlying point of disagreement. At least we know why we're disagreeing. :smallbiggrin:

Discrete literally means countable. The moment you admit there are three, you admit that they are discrete. You can't say there are three and they are not discrete without violating the definition of discrete.

"How many..." applies to discrete things.

"How much..." applies to continuous things.

Do you have some alternate definition of discrete for which this is not the case?

Sorry, that's my own filters kicking in. Why is 'the writers didn't understand their rules' more likely than 'Mirror Image doesn't work against non-attack spells?'

First, I never suggest that "the writers don't understand their own rules." I suggested that "one writer didn't know which rules, that were written by other writers, were (officially) part of the game."

Second, I only reluctantly made the suggestion because someone else suggested their idea was significantly more likely, to the point of being the correct explanation.

Accepting this, the two claims take place in totally different contexts.

The possibility that ""one writer didn't know which rules, that were written by other writers, were (officially) part of the game." Hinges on things like "Who wrote the spell?" "How many writers were in his department?" "How many departments worked on the book?" "How did the departments communicate?" "How was the project divided?" "Were different parts due at different deadlines?" etc. It seems to me that there a number of ways for this error to slip through.

The possibility that "Mirror Image doesn't work against non-attack spells?" as written, is much simpler. No speculation is required. We have the written rules. We know what they say. The number of possible explanations is much more limited, by the meaning of the words on the pages.

In the context of the disagreement, my suggestion did not even need to be likely. It only needed to be likely enough to be a possible explanation. This is because my opponents claimed that they knew the explanation.

You know what, let's not talk about specific spells and please let's not talk about their interactions. Let's talk about spell targeting in general. So forget the spell Mirror Image and let's talk about ACTUAL MIRRORS.

Okay, so I show that you're wrong and we just leave it and move on? :smallwink:

Two casters are separated by long solid stone wall, 3 feet wide and 20 feet tall. 5 feet over the wall there's a gigantic mirror that enables the casters to see each other. Can they target each other with spells through the mirror?

No.

Two other casters are standing in a hall of mirrors. They both know there isn't anybody else in the hall except the two of them. They both see hundreds of copies of both of them all around, but there's nothing separating them physically. Can they target each other with spells directly or through the mirrors? (we can skip the lol I cast fireball jokes).

They can target the other, but only if they are able to pick him out.

I actually find it easier to visualize them as NOT being able to take a still frame and actually count 4 images, than I do to visualize being able to count 4 images and not being able to keep track of them.

Then, it seems to me, you are not acknowledging what it means to be discrete.

Given they move about in such a way as to not be able to distinguish which is the real caster or keep track of them.

Maybe it would help if you could show or describe something that is similar to what you think is happening with the spell. Because, for example, I can imagine a slot-machine, wherein the wheels are spinning so fast that I can't determine exactly when one is lined up with the centre-line. But the fact that they appear blurry doesn't actually make them blurry. We can see that when the wheel stops spinning.

Either are perfectly valid as long as it matches the RAW requirement that attacks target the caster first, then are redirected to an image.

This is not a requirement. This is a logical error.

The spell says: If you attack the target, then the attack is redirected. This does not mean that you must attack the target. You can take an action you want, including attacking something else, and also including casting a spell, eating a goodberry, or whatever.

Just like: If it rains, I will bring an umbrella. This does not mean that it must rain. It might be cloudy, sunny, or whatever.

But if you require targeting to first pick a target caster/image for non-attacks, you should be consistent and require it for attacks.

I'm going to find a way to break this down and get back to you. I get what you're saying but I don't think it's right. It's hard to explain in any case.

BurgerBeast

2016-08-19, 05:58 PM

@Tanarii

So it turns out that what is meant by targeting comes to the fore again:

I. Assume that "targeting" is a discrete event. It has either happened, or it hasn't.

1. The player declares "I __________ Joe." (attack or cast hold person on)

2. The DM states that there are four Joes, and the player must pick one. (targeting has not occurred).

3. The player is now required to specify a target. There are four options and they all look the same.

Nothing the player can say or do at this point can succeed at identifying one of the four possible choices. The game stops, while the DM waits for the player to do the impossible, in either case.

II. Assume that when a player declares an intended target, this is not enough to say that targeting has happened (a target is not selected), but is enough to satisfy the condition "Joe is targeted" because it makes the DM aware that the character is trying to target Joe

1. 1. The player declares "I __________ Joe." (attack or cast hold person on)

2. The DM states that there are four Joes, and the player must pick one. (but now, the DM is within his rights to treat the statement "I attack Joe" or "I cast hold person on Joe" as satisfying the requirement "the player has targeted Joe.").

3. The DM is now within his rights to say that the player has targeted Joe, but there are four options and they all look the same.

The DM is now free to resolve the issue of selecting a target without demanding that the player do it, because it is fair to say that the player has "targeted Joe".

3a. If this is an attack, the mirror image condition is met because "Joe is targeted by an attack" - he DM uses the d20 mechanic to pick a target
3b. If this is a spell, there are no guidelines, but a target still must be selected, so the DM must do something to pick the target (and there are no rules to help him - it seems appropriate to randomize the target)

4. The attack or spell is resolved now that it has a clear target.

Note that assumption I causes the game to grind to a halt in both cases, which is why I think interpreting the word "target" in this way must be wrong because it never works.

Assumption II works in both cases.

So, despite the fact that even I can see some poor word choice within the spell about attacking and redirecting, it actually does make sense if you think that the simple declaration "I attack Joe" can be considered targeting, even though a target hasn't been picked.

It's the only way it works, so far as I can tell.

The idea that spells and attacks both ignore the duplicates and auto-hit Joe does sound good on the face of it, but the problem here is what property of the illusory duplicates makes them not targetable? Some will say that they are only targetable if the spell says so.

To this, I would say:
1. Everything is targetable by attacks, if not explicitly then because they occupy locations in space and locations are targetable. So aiming at the space something is in will resolve (through DM interpretation) into targeting whatever creature or object is there.
2. The illusory duplicates are illusory duplicates. It is not "going beyond the rules" nor it is "adding more to the spell description" to claim that they have all the properties of illusory duplicates. And it is within RAW for a DM to consider illusory duplicates to be targetable.

Telok

2016-08-19, 06:48 PM

Sorry, that's my own filters kicking in. Why is 'the writers didn't understand their rules' more likely than 'Mirror Image doesn't work against non-attack spells?'
Well it fits for me because I have fewer but more positive expectations of the spell and more but rather negative expectations of the company. Simply put I expect the spell to make harming the caster more difficult, specifics depending on the version of D&D used, and I expect the company to screw things up.

So here's a question, say we have a series of rooms and each room contains a flesh golem with an antimatter bazooka and an illusion trap. When someone opens the door they get one action before kablooey, so no time to make investigation checks about the illusion.

In one room the illusion trap makes three identical images of the golem adjacent to the golem. A wizard opens the door and casts 'Vaporize Golem', a single target save or vaporize spell.

In another room the spell trap makes three identical images of the golem in the same square that overlap. Wizard opens the door and casts the vaporize spell.

In the last room the golem in on a Tensors Floating Disk that moves around randomly within the golem's space. The illusion duplicates the golem, the disk, and moves the illusionary duplicates around and through the golem amd each other. Again a wizard opens the door and casts vaporize.

So here are situations with illusionary duplicates that doesn't use Mirror Image and a spell that, like Hold Person, is a targeted save. We could have a similar situation with an animated statue among regular statues, or dopplegangers among humans, or a wizard with apporprately spiffed up simracula of himself.

How do we adjucate the targeting of the spells?

Tanarii

2016-08-20, 09:31 AM

Sorry, could you quote the exact text that explicitly overrides the targeting rules? My reading is that the spell changes what happens after being targeted by adding an extra step to resolving the attack, but I could easily be missing something. The d20 I use for Perception Checks seems to have an abnormal number of 1's on it.As long as the process for targeting with an attack is "target creature, check for (edit: redirect to) target image" it actually doesn't matter if you choose to interpret that as happening in addition to or instead of the normal targeting process for an attack.

It's just not (IMO of course) consistent to choose to do it instead of, since that isn't explicit in the wording. But that's just my opinion. I absolutely can see the viewpoint of those choosing to interpret it as replacing the normal targeting for attacks.

Edit2: especially since their interpretation of the spell requires replacing the normal targeting for non-attacks, causing a conflict if they choose not to interpret it as not replacing 'normal targeting'. In other words, if you already decided the spell changes the way targeting occurs in general, to avoid a conflict between that and the target redirect for attacks, you must also view the attack targeting redirect rule as replacing the newly changed general targeting rule.

So the rule becomes:
1) Mirror Image replaces targeting rules, requiring you to choose among visually identifiable images instead of just directly targeting creature.
2) Mirror image replaces new targeting rule, instead allowing you to directly targeting creature again, then possibly redirecting it to an image.

BurgerBeast

2016-08-21, 11:33 AM

So, after hearing all of the discussion, I'm going to go ahead and say this: This is how the spell should play out, and this is all within RAW:

Player: I cast hold person on the wizard.

DM: He’s got mirror image up. There are four “wizards” and you can’t tell which is the real one. Do you want to just pick one and cast, or try something else?

Player: I want to pick one and cast.

DM: In his head assigns the number 1 to the real wizard. Rolls 1d4. (If he rolls a 1, he then rolls the wizard’s Wisdom save. If the wizard fails, he says: “All four images of the wizard freeze their movement, under the effect of the spell." In any other case, he says: "The spell has no effect.")

Pretty much any other interpretation (including Crawford's) is worse.

----------

From this point on, I'll be more inclined to show that this is the best interpretation and the rest are inferior, if anyone claims otherwise.

Mellack

2016-08-21, 12:14 PM

So why does that have a different targetting mechanic than if I shoot an arrow? That seems to be problem #1.

BurgerBeast

2016-08-21, 12:17 PM

So why does that have a different targetting mechanic than if I shoot an arrow? That seems to be problem #1.

Then you tell me. You use a different mechanic, too.

[edit: added "too"]

Mellack

2016-08-21, 12:30 PM

Your thread, and you are claiming RAW, so please describe your point.

BurgerBeast

2016-08-21, 12:34 PM

Your thread, and you are claiming RAW, so please describe your point.

Why do you use a different mechanic to resolve a spell when it is cast at Joe?

smcmike

2016-08-21, 12:39 PM

Why do you use a different mechanic to resolve a spell when it is cast at Joe?

Dance, dance. You've staked a position, Burger, just defend it.

Mellack

2016-08-21, 12:41 PM

Why do you use a different mechanic to resolve a spell when it is cast at Joe?

I am looking to work through YOUR position. I know how my group works it. So please walk me through both attacks and non-attacks.

BurgerBeast

2016-08-21, 01:53 PM

Maybe you guys don't understand the density of your own questions.

You also claim that attacks and save-spells resolve differently. If we both claim that this is true, then there is no argument. So I don't need to defend my claim. We all agree that the spell mirror image treats attacks and save spells differently.

Does it really change anything if you or I say "because the spell says so"?

Apparently it might for you, so:

Because the spell says so.

smcmike

2016-08-21, 01:59 PM

Mellack

2016-08-21, 02:00 PM

BurgerBeast

2016-08-21, 02:04 PM

So you are unwilling, after having started two threads, to work me through how you run it? I guess so. So what was the point?

I just did. My answer is:

Because the spell said so.

What's next?

[edit: if it helps:

You asked: So why does that have a different targetting mechanic than if I shoot an arrow? That seems to be problem #1.

My answer is: Because the spell says so.

your turn]

RickAllison

2016-08-21, 02:08 PM

I just did. My answer is:

Because the spell said so.

What's next?

[edit: if it helps:

You asked: So why does that have a different targetting mechanic than if I shoot an arrow? That seems to be problem #1.

My answer is: Because the spell says so.

your turn]

No, the question was why targeting works differently between the two. You have to target the caster before Mirror Image's stated effect occurs, this is explicit in the text. So why does pointing a finger at the caster for a Firebolt differ from pointing my finger at them for Disintegrate?

Mellack

2016-08-21, 02:12 PM

You seem unwilling to answer a simple question about what you claim is the RAW way. This does not build confidence in your position. Please take me through each step in how it works with both an attack and a save spell.

BurgerBeast

2016-08-21, 02:21 PM

@RickAllison: Sorry, one at a time. I'll happily explain why you are wrong later, if you care.

You seem unwilling to answer a simple question about what you claim is the RAW way. This does not build confidence in your position. Please take me through each step in how it works with both an attack and a save spell.

I never said my answer is the "RAW way," as though there is only one way. Don;t make false accusations.

I said my way falls within "RAW."

Attack:

Player: I want to attack that wizard.

DM: (Considers this to count as targeting the wizard. But there is no specific target yet.) Uses the d20 mechanic to resolve target. Then, resolves the attack.

Hold Person spell:

Player: I want to cast hold person on that wizard.

DM: (Considers this to count as targeting the wizard. But there is no specific target yet.) Rolls a d4 to resolve the target. Then, resolves the spell.

Mellack

2016-08-21, 02:25 PM

@RickAllison: Sorry, one at a time. I'll happily explain why you are wrong later, if you care.

I never said my answer is the "RAW way," as though there is only one way. Don;t make false accusations.

I said my way falls within "RAW."

Attack:

Player: I want to attack that wizard.

DM: (Considers this to count as targeting the wizard. But there is no specific target yet.) Uses the d20 mechanic to resolve target. Then, resolves the attack.

Hold Person spell:

Player: I want to cast hold person on that wizard.

DM: (Considers this to count as targeting the wizard. But there is no specific target yet.) Rolls a d4 to resolve the target. Then, resolves the spell.

Where does that d4 come from? There is nothing in RAW to state that.

Segev

2016-08-21, 02:31 PM

Actually, nothing in the spell prevents you from using the same mechanic to determine whether the real Joe or a fake "Joe" is hit by the effect in question, regardless of whether the effect is hold person or an "attack."

The RAW require that an attack use the d20 procedure. They do not require that you use the same d20 procedure in the event of hold person, but nothing prevents you from doing so.

So you could use the d20 procedure for both "attack"s and for hold person, if that's how you want to run it.

As for "why are spells treated differently than attacks?" as a question, BurgerBeast saying "because the spell says so" is one of the earliest concessions he made in both threads: that the spell DOES say "attack" in the part of its description that requires you to use the d20 method described to determine whether the real Joe is hit or not.

Spells are treated differently than attacks iff the DM chooses to use something other than the d20 procedure for spells. There are therefore technically two reasons why spells would be resolved differently than attacks:

1) The spell says they can be, and
2) the DM decided they would be.

Mellack

2016-08-21, 02:43 PM

smcmike

2016-08-21, 02:47 PM

Here is a question for Burger and Seg:

You argue that our position is not consistent with RAW. Other than this supposed inconsistency, are there any other reasons to prefer your approach?

Segev

2016-08-21, 02:58 PM

That is a fine position, Segev. Thank you for being forthcoming and clear.

I disagree that that is the best way of running the spell. I believe that makes it far more powerfull than intended. This position also happens to be supported as the intention by the designers. Anyway, any claim to be the best way is nothing other than opinion, and cannot be proven.

Saying something does not go against RAW does not equal that being the RAW way such a thing should work. I could say that every character gets 5 minutes a day where they have to "answer the call of nature" and get disadvantage to perception checks during that time. It certainly does not go against anything written in the books. I can even claim it increases versimilitude. However it is not something that one would claim as RAW. It is a houserule, just as having save spells choose a target is a houserule. If you enjoy it and find it improves your game, please use it. When someone comes here and claims their way is how it "should play out," it smacks of elitism.Eh, I could get down in the weeds here and argue back that the RAW puts those 5 minutes during short- and long rests, where rules about perception and seeing threats apply, but the broader picture is that the analogy is bad.

The spell says it creates duplicates. This obviously is intended to mislead people who are relying on their ability to visually identify the caster as the target of their effects. I back this up by pointing out that, if we remove all mechanics and just try to tell this as a narrative story, all mirror image does is make those duplicates. How do you suppose the fact that 3/4 of attacks will hit an image rather than the real caster is justified, narratively? By the fact that nobody can tell which of the 4 is the real caster, of course.

Regardless of what mechanics the spell gives, suggests, requires, hints at, or remains silent about, the spell creates the 3 illusory duplicates, and they expressly make it impossible to track which image is the real one out of all of them.

While one can call literally ANY ruling a DM makes a "house rule," and be 100% right, there is a definite line between a "ruling" and a "house rule" as we tend to use the terms in this sort of discussion. A "ruling" is something the DM has to determine/adjudicate in order to resolve a situational ambiguity in the RAW, or to come up with a means of handling something the RAW don't spell out explicitly for him. A "house rule" is a change, addition, or subtraction from the RAW which is not strictly required to resolve an ambiguity arising from the RAW.

Your example doesn't resolve an ambiguity. The mirror image decision of how to pick which image is affected (including saying 'it is not an attack, so it automatically hits the real Joe') is a ruling, not a house rule.

I just think the parenthetical ruling violates the RAW by ignoring the part about how it's impossible to track which image is the real Joe.

Here is a question for Burger and Seg:

You argue that our position is not consistent with RAW. Other than this supposed inconsistency, are there any other reasons to prefer your approach?
Mainly that the RAW state there are 3 illusory doubles, and that you cannot track which is the real one out of the 4 you can see. By allowing Bob to hit the real Joe automatically, you're essentially ignoring that part of the RAW, because you're saying Bob can and does track which is the real Joe, if only by virtue of his spell somehow "knowing" which one is the one Bob would have wanted to hit if he could have told which one was Joe.

Mellack

2016-08-21, 03:10 PM

Xetheral

2016-08-21, 03:12 PM

Segev

2016-08-21, 04:15 PM

So the targetting by spells in not a clear part of the RAW, instead requiring a ruling. Such a ruling should presumably be one that makes the game more fun and keeps balance as much as possible. The simpliest way to do that is to ask the designers whose profession is making RPGs, this one in particular, and get their intention on how the spell was expected to work. Crawford has given us that. If you do not like his ruling, then you should make your own based on what you think works best. I think ruling that spells have to choose between images and the caster makes the spell stronger than expected, which lowers the fun level for myself.
Of course rulings are something that can and should vary from group to group. Saying that one way is how it should be done for all groups would be rude.
I mainly find "ask Crawford" to be a lousy method because Crawford doesn't seem to actually put any thought into his answers. He seems to instead simply make a spot-ruling, as I might expect from a DM who doesn't wish to interrupt the flow of a game by stopping to look things up in the books or consider ramifications of his ruling behind that one instance.

And that's a generous reading of it.

I frankly expect better of a game designer. I expect that his rulings on one topic will enable extrapolation to similar topics without having to pretend "no, no, that's totally different even though the wording is ambiguous in exactly the same way, because, um...I don't want the logical conclusion of my spot-ruling to apply here!"

I refuse to have to run to Crawford every time I want to figure out if I'm right in reading the rules. If Crawford provided answers which were illuminating as to a means of accurately reading the rules, as opposed to "in this case, it's this, because that's what I want it to be here," then his answers would be useful.

But his rulings are no more useful than "ask your DM," because they're LESS self-consistent than most DMs' rulings in their home games.

As an example, I don't believe anybody has cited Crawford speaking to hold person, specifically, but has instead referred to Crawford's ruling regarding magic missile. Since magic missile is not hold person, and Crawford has in the past ruled opposite ways on similar wording just because it's two different spells, it is not safe to assume that his magic missile ruling can extrapolate to hold person, nor that a hold person ruling could extrapolate to charm person, etc.

So even this specific appeal to authority fails to be convincing, because there is not sufficient consistency in this authority's rulings to extrapolate from what is cited to what is being done.

And this is why I have so little respect for "but Crawford said..." as a means of interpreting the RAW: he doesn't interpret it so much as spot-rule on it, and it doesn't seem that he considers his spot-rules in light of anything other than the very narrow situation he is asked about.

BurgerBeast

2016-08-21, 04:18 PM

Where does that d4 come from? There is nothing in RAW to state that.

I never claimed there was. I claimed that my view falls within RAW. Not that it is (or is the only) RAW.

Here is a question for Burger and Seg:

You argue that our position is not consistent with RAW. Other than this supposed inconsistency, are there any other reasons to prefer your approach?

Yes. If a player casts mirror image, it should do what it claims to do. There should be four potential targets, and they should be what they are "illusory duplicates." Any action taken by anyone in the game should take this into account.

So the targetting by spells in not a clear part of the RAW, instead requiring a ruling.

Bingo. And my ruling is within RAW.

Such a ruling should presumably be one that makes the game more fun and keeps balance as much as possible.

I just think it should be as fair as possible. "I think save spells auto-target in this one particular instance, but not in any others" is not fair.

The simpliest way to do that is to ask the designers whose profession is making RPGs, this one in particular, and get their intention on how the spell was expected to work.

No, the simplest way is to read the spell and apply it fairly.

Crawford has given us that. If you do not like his ruling, then you should make your own based on what you think works best.

It's not that I don't like his ruling. It's that his ruling violates RAW. It may be RAI, but it is not within the bounds of RAW, which makes it unfair.

I think ruling that spells have to choose between images and the caster makes the spell stronger than expected, which lowers the fun level for myself.

That's fine. But the spell says what it does. Crawford's ruling adds a new function to save-spells that was never there before, applies only to MI, and makes them particularly powerful against MI. In my opinion, this makes all save-spells stronger than expected, which makes the game unfair in this circumstance.

Of course rulings are something that can and should vary from group to group. Saying that one way is how it should be done for all groups would be rude.

I have never said that "one way is how it should be done for all groups." Some people (not you) have illogically arrived at that conclusion despite my being clear about it form the start.

@Xetherai:

Thanks for the detailed explanation. I will try to highlight my disagreements, and give reasons. Earlier in thread, I mentioned that this might all come down to what is meant by targeted. In particular it comes down to what was meant by "targeted" in the spell text.

(snip)

I ran into this problem, too. But then I realized that it is irrelevant because the character cannot distinguish between Joe and his duplicates. The d20 mechanic can be used in either case. If you want to use a d4 when a duplicate is targeted (since the d20 is not stipulated) and the RAW when Joe is targeted, you'll get the same result anyway.

As a side note: You can get bogged down in the difference between "target" meaning "declare an intended target" or "unambiguously select an individual target" but this also is irrelevant if you follow it all the way through.

I will try to write out what I mean as clearly and unambiguously as possible. It will take a while to write it out clearly.

Segev

2016-08-21, 04:22 PM

Edit: I therefore think that, if one wants to be consistent with what appears to me to be the best interpretation of the RAW, one should interpret the visual manifestation of the spell such that the images are both countable and non-targetable.

I ultimately remain in disagreement with this position because it makes zero sense to me from a "modeling something that's really happening in the game narrative" point of view. If the images are not targetable, what is the attack that fails to go for the real one hitting?

Put another way, this interpretation relies too much on "video game logic:" the images aren't really there; they're just a graphical effect prettied into place for the players' benefit. The characters and creatures don't even see them, because they can clearly tell where the real target is; they just aren't allowed to hit it three in four times they attack it. But they can clearly do so if they do anything that isn't an "attack."

5e is not meant to be that computer-code-esq; that was a feature (or bug, YMMV) of 3.5. 5e is meant to be much more a model of something that is "really happening" on some level in the narrative. So when mirror image creates illusory doubles, they're there. The effects of the spell surrounding not hitting the real caster are precisely because the creatures can't tell which is the caster and which are fakes. As long as the one attempting to apply an effect to the caster has to have a target for his effect, then, the images should fool him equally.

It shouldn't matter if Jimmy is trying to stab Joe with a sword, cast a spell on him, or pick his pocket; Jimmy should be equally confused in all three circumstances as to whether the "Joe" he's trying to do this to is the real one or an illusory fake. Moreover, if Jimmy can unerringly pick out the real Joe for pocket-picking or hold person, I contend that you're ignoring the RAW where the spell says that it's impossible to track which image is the real one.

Mellack

2016-08-21, 04:46 PM

It's not that I don't like his ruling. It's that his ruling violates RAW. It may be RAI, but it is not within the bounds of RAW, which makes it unfair.

No, his ruling does not violate RAW. It goes against your interpretation. They are not the same thing.

I have never said that "one way is how it should be done for all groups." Some people (not you) have illogically arrived at that conclusion despite my being clear about it form the start.

You did claim that your way was the way it should work. No disclaimer that it is only for your group. Lets see a quote.

So, after hearing all of the discussion, I'm going to go ahead and say this: This is how the spell should play out, and this is all within RAW:

You even put it in bold. If you want to claim that it is only your opinion, then you should be more careful in how you respond. You seem to have a tendancy to speak in absolutes, such as the quote above. It is off-putting, and distracts from polite discourse.

You have an interpretaion, or ruling, of how you think the spell works. Others disagree with that. Just because yours does not expelcitly break RAW does not make it the best interpretation, the way it "should" work. Best is subjective. It is one interpretation, and one that works for you. Great. Your statements however suggest that others are somehow doing it wrong. Others rulings work better for them, and they are just as correct. Since the text is not explicit, there is no way it "should" be played.

BurgerBeast

2016-08-21, 07:04 PM

No, his ruling does not violate RAW. It goes against your interpretation. They are not the same thing.

Crawford's ruling bestows the ability upon the caster of save-spells the otherwise unheard of ability to auto-target the particular person they want to hit despite seeing four targets and not knowing which is the one they want to hit. That's why it goes against RAW. [edit: and RAI, actually]

You did claim that your way was the way it should work. No disclaimer that it is only for your group. Lets see a quote.

Should work, i.e. by RAW. Who operates on the assumption that everyone should play by RAW?

You even put it in bold. If you want to claim that it is only your opinion, then you should be more careful in how you respond. You seem to have a tendancy to speak in absolutes, such as the quote above. It is off-putting, and distracts from polite discourse.

It's an opinion. Do you think my opinion is going to be how "it shouldn't play out"? What are you playing at? Do I need to say "I think" before everything I write, too? Or are you going to assume that I sometimes write things that I don't think? Get real, pal.

You have an interpretaion, or ruling, of how you think the spell works. Others disagree with that. Just because yours does not expelcitly break RAW does not make it the best interpretation, the way it "should" work.

We are discussing RAW, so there's that. In a discussion about multiple ways to rule, a way that is within RAW is superior to a way that violates RAW. Period.

Best is subjective.

No, it isn't. "Best" is subjective when the subject matter is subjective. "Best" is objective when the subject matter is objective.

It is one interpretation, and one that works for you. Great. Your statements however suggest that others are somehow doing it wrong.

If you can't be bothered to read the thread, then I frankly don't care what you think. If you could be bothered, you'd notice pretty quickly that you are preaching to the choir.

Others rulings work better for them, and they are just as correct.

This is absurd. All interpretations are "equally correct" now?

Since the text is not explicit, there is no way it "should" be played.

The text is explicit. There is a difference between text that is not explicit and text that is explicit but come people can't read.

BurgerBeast

2016-08-21, 08:21 PM

I originally sided with BurgerBeast and Segev (although I also suggested it might be easier to reinterpret the fluff so that the provided mechanics in the spell could offer a complete resolution mechanic). But I've found several of the opposing arguments, particularly Tanarii's, to be convincing. In particular, I think the entire question comes down to whether or not the images are individually targetable, and I'm convinced that they are not, despite being countable. Here's my thought process, for anyone who is curious:

I think it's notable that the spell contains no mechanic for redirecting an attack from an image to the caster.

Don’t you think this is simply because they never expected a player to declare “I want to hit a duplicate – not the real Joe.” No player would ever try to hit the duplicate, and if they did, it would have the exact same effect as trying to hit Joe – it would come down to random chance.

The provided d20 mechanic only triggers when "you" (referring to the caster) are targeted with an attack.

Yes, but the act of trying to target Joe is logically indistinguishable from trying to target a duplicate. If you try to hit Joe, you have a 25% chance to hit him and a 75% chance to hit a duplicate. If you try to hit a duplicate, you have a 25% chance to hit Joe and a 75% chance to hit a duplicate. It’s the same.

Slightly related example: it makes no difference whether you try to roll exactly a 6 or you try to roll under 3 on a 6-sided die. In either case, you have a 1/6 chance of rolling a 6 and a 1/3 chance of rolling under 3. "Trying to roll a 6" doesn't mean much since you have no way to actually target a 6.

So if, in Segev's freeze-frame analysis, the attacker originally targeted Image B, and Image B happens to not be the caster, the d20 mechanic never triggers.

Nor should it. In Segev’s freeze-frame example, the attacker is capable of targeting B specifically by looking at the freeze frame and picking one image, so this takes the place of the d20 resolution mechanic. In a typical game, with no freeze-frame for the player to look at, the d20 mechanic simulates freezing the frame and selecting a target by randomizing A, B, C, and D.

This leaves the provided resolution mechanic woefully (and implausibly) incomplete on it's face.

I disagree.

By contrast, if the caster (as opposed to the image that happens to correspond to the caster at the moment) is the original target of the attack (i.e. the images themselves are not individually targetable) the provided resolution mechanic will always trigger in the case of an attack.

I’m a bit confused by the phrasing in the first sentence, in parentheses. The images do not switch between being Joe or being his duplicate. Joe is one potential target. Each other potential target is a duplicate.

But, yeah, this all boils down to the fact that the player is literally incapable of specifying A, B, C, or D. So the player is incapable of targeting Joe and the player is incapable of targeting a duplicate. If he was capable, there would be no d20 mechanic. The d20 mechanic exists because it simulates the freeze frame. That’s why we have to use either the freeze-frame or the d20 mechanic.

Therefore, it seems to me that the best interpretation of the RAW is that only the caster is targetable by an attack (despite being unidentifiable), and the individual images are not.

The point is that none of the four is specifically targetable. By this I mean that the player cannot intentionally pick out Joe, nor can the player intetionally pick out a duplicate. What is ultimately targeted is up to random chance. But the player can declare the “intent” to target Joe, which triggers the mechanic.

As another example of RAW problems produced if the images are independently targetable, consider what happens if the caster is successfully held. At this point, either all of the images stop moving, or the image corresponding to the real caster stops moving and the others continue to move (the spell is unclear on this point). Either way, if the images are individually targetable, it becomes trivial to circumvent the protection of the spell (either you shoot at the only one not moving, or else once you hit the real one, it's obvious which one it is). This would violate RAW, because, as written, the beneficial effects of Mirror Image do not end if the caster is unmoving. (The spell explicitly says the protection lasts until the spell ends; being helddoesn't end the spell.) By contrast, if the images are not independently targetable, then the spell continues to provide protection to a held caster, in keeping with the RAW.

This is everyone’s problem
This example gives the other explanation the same trouble that it gives me. Regardless of how you “fluff” the spell, it makes no sense that a held target can still benefit from moving illusory duplicates, so it’s irrelvant which side you pick – we share this problem in common. You can’t accuse me of violating RAW in this particular case without also accusing the alternative explanation. There’s no “out” that is exclusive to one explanation.

My ruling:
It's no problem if the caster is held. There are four possible targets and you still don't know which is which. Pick one to hit (or use the mechanics to do it). I don't have a problem with this. The tracking problem is eliminated now, because they are held. Without moving, the duplicates can’t be tracked, but you still don’t know which is Joe and which are duplicates (unless you happen to attack Joe and hit, in whch case it becomes obvious). Once you identify which of the held “Joes” is the real Joe, I would let you auto-target him as long as he remained held. If the hold person effect ends, the duplicates return to their normal behaviour and are impossible to track again for the duration.

My explanation:
In my view, the rules don’t cover this interaction. There is nothing, however, RAW, to prevent it from happening. I don’t necessarily think it contradicts RAW. I, personally, have trouble explaining how Joe can be held and the images can move and yet Joe can be impossible to pick out, particularly because the only way to “shift positions so that it is impossible to track which image is real,” seems to me to be to not shift positions at all.

This problem is only a problem if you “fluff” the held Joe as being stuck in place while the illusiory duplicates move around him. I think that in order to preserve RAW, the duplicates must stop in place. “Until the spell ends, the duplicates move with you (i.e. stop when you stop as well as start when you start) and mimic your actions (freeze in a pose to mimic when you freeze in a pose), shifting position so it’s impossible to track which image is real (i.e. stop shifting when shfting would make tracking possible)” Ironically, I think that the choice of the word “track” as opposed to “identify” actually helps my case here, because technically you still can’t track them when they move.
So instead you have to “fluff” it as the duplicates and Joe all being held. This is the only way for them to “move with you” and yet for it to be “impossible to track which image is real.” They have not stopped doing this, because the moment the effect ends, they will return to normal until the end of the spell’s duration.

You might claim that by allowing the players to auto-target Joe, once he has been identified while held, violates RAW because targeting him specifically means that attack should be redirected. Yes, I am comfortable with it in this situation because it is clear to me that hold person has effectively removed the functionality of mirror image in this case. Also, we only ever arrived at this hypothetical when we violated RAW to say that the duplicates no longer move in the first place.

I would never do the following, but it's worth saying: If you wanted to follow RAW, then when the duplicates must continue to move in such a way that they are still able to "fool" your targeting ability. There is RAW precedent for this in my opinion (see Illusory Self, PHB 118 - which makes no real sense but the intended function is clear). If the illusory duplicate created by Illusory Self can cause an attacker to auto-miss by "interposing" itself between Joe and his attacker, then this seems consistent with the idea that moving illusory duplicates of held Joe can interpose themselves in the same way. Again, I wouldn't do it. My ruling is that hold person eliminates the tracking problem, but the illusions are still indistinguishable. I also, for what it's worth, think this is way cooler (since people have brought fun into it).

Edit: I therefore think that, if one wants to be consistent with what appears to me to be the best interpretation of the RAW, one should interpret the visual manifestation of the spell such that the images are both countable and non-targetable.

I still say this makes no sense. There is simply no case in the context of D&D, where you can count four things, yet not target one of them, unless they are not targetable in principle. Yet illusions are, in principle, targetable.

Aridon

2016-08-21, 08:40 PM

Don’t you think this is simply because they never expected a player to declare “I want to hit a duplicate – not the real Joe.” No player would ever try to hit the duplicate, and if they did, it would have the exact same effect as trying to hit Joe – it would come down to random chance.

Yes, but the act of trying to target Joe is logically indistinguishable from trying to target a duplicate. If you try to hit Joe, you have a 25% chance to hit him and a 75% chance to hit a duplicate. If you try to hit a duplicate, you have a 25% chance to hit Joe and a 75% chance to hit a duplicate. It’s the same.

Slightly related example: it makes no difference whether you try to roll exactly a 6 or you try to roll under 3 on a 6-sided die. In either case, you have a 1/6 chance of rolling a 6 and a 1/3 chance of rolling under 3. "Trying to roll a 6" doesn't mean much since you have no way to actually target a 6.

Nor should it. In Segev’s freeze-frame example, the attacker is capable of targeting B specifically by looking at the freeze frame and picking one image, so this takes the place of the d20 resolution mechanic. In a typical game, with no freeze-frame for the player to look at, the d20 mechanic simulates freezing the frame and selecting a target by randomizing A, B, C, and D.

I disagree.

I’m a bit confused by the phrasing in the first sentence, in parentheses. The images do not switch between being Joe or being his duplicate. Joe is one potential target. Each other potential target is a duplicate.

But, yeah, this all boils down to the fact that the player is literally incapable of specifying A, B, C, or D. So the player is incapable of targeting Joe and the player is incapable of targeting a duplicate. If he was capable, there would be no d20 mechanic. The d20 mechanic exists because it simulates the freeze frame. That’s why we have to use either the freeze-frame or the d20 mechanic.

The point is that none of the four is specifically targetable. By this I mean that the player cannot intentionally pick out Joe, nor can the player intetionally pick out a duplicate. What is ultimately targeted is up to random chance. But the player can declare the “intent” to target Joe, which triggers the mechanic.

This is everyone’s problem
This example gives the other explanation the same trouble that it gives me. Regardless of how you “fluff” the spell, it makes no sense that a held target can still benefit from moving illusory duplicates, so it’s irrelvant which side you pick – we share this problem in common. You can’t accuse me of violating RAW in this particular case without also accusing the alternative explanation. There’s no “out” that is exclusive to one explanation.

My ruling:
It's no problem if the caster is held. There are four possible targets and you still don't know which is which. Pick one to hit (or use the mechanics to do it). I don't have a problem with this. The tracking problem is eliminated now, because they are held. Without moving, the duplicates can’t be tracked, but you still don’t know which is Joe and which are duplicates (unless you happen to attack Joe and hit, in whch case it becomes obvious). Once you identify which of the held “Joes” is the real Joe, I would let you auto-target him as long as he remained held. If the hold person effect ends, the duplicates return to their normal behaviour and are impossible to track again for the duration.

My explanation:
In my view, the rules don’t cover this interaction. There is nothing, however, RAW, to prevent it from happening. I don’t necessarily think it contradicts RAW. I, personally, have trouble explaining how Joe can be held and the images can move and yet Joe can be impossible to pick out, particularly because the only way to “shift positions so that it is impossible to track which image is real,” seems to me to be to not shift positions at all.

This problem is only a problem if you “fluff” the held Joe as being stuck in place while the illusiory duplicates move around him. I think that in order to preserve RAW, the duplicates must stop in place. “Until the spell ends, the duplicates move with you (i.e. stop when you stop as well as start when you start) and mimic your actions (freeze in a pose to mimic when you freeze in a pose), shifting position so it’s impossible to track which image is real (i.e. stop shifting when shfting would make tracking possible)” Ironically, I think that the choice of the word “track” as opposed to “identify” actually helps my case here, because technically you still can’t track them when they move.
So instead you have to “fluff” it as the duplicates and Joe all being held. This is the only way for them to “move with you” and yet for it to be “impossible to track which image is real.” They have not stopped doing this, because the moment the effect ends, they will return to normal until the end of the spell’s duration.

You might claim that by allowing the players to auto-target Joe, once he has been identified while held, violates RAW because targeting him specifically means that attack should be redirected. Yes, I am comfortable with it in this situation because it is clear to me that hold person has effectively removed the functionality of mirror image in this case. Also, we only ever arrived at this hypothetical when we violated RAW to say that the duplicates no longer move in the first place.

I would never do the following, but it's worth saying: If you wanted to follow RAW, then when the duplicates must continue to move in such a way that they are still able to "fool" your targeting ability. There is RAW precedent for this in my opinion (see Illusory Self, PHB 118 - which makes no real sense but the intended function is clear). If the illusory duplicate created by Illusory Self can cause an attacker to auto-miss by "interposing" itself between Joe and his attacker, then this seems consistent with the idea that moving illusory duplicates of held Joe can interpose themselves in the same way. Again, I wouldn't do it. My ruling is that hold person eliminates the tracking problem, but the illusions are still indistinguishable. I also, for what it's worth, think this is way cooler (since people have brought fun into it).

I still say this makes no sense. There is simply no case in the context of D&D, where you can count four things, yet not target one of them, unless they are not targetable in principle. Yet illusions are, in principle, targetable.

TL DR all of the posts, but as a DM I would treat the hold person as an attack and roll the d20 as per the description in Mirror Image. If I roll up the real Joe, then I would give him a saving throw and apply the result to Joe. If I rolled up one of the fake Joes, I would fake roll a saving throw and tell the caster that Joe made his saving throw. This seem the RAW interpretation of this scenario to me.

Erys

2016-08-21, 09:31 PM

RickAllison

2016-08-21, 09:53 PM

The more I read this house rule that requires targeting images when the spell is not an attack, the more I dislike it. It makes a second level spell way too powerful.

Now Magic Missile, which is never supposed to miss- misses 75% of the time. As does higher level spells like Otto's Irresistible Dance, the Powerword's, Maze and more... all thwarted by a 2nd level spell.

At least with Counterspell there is a possible roll (for higher level spells) and it uses one of your own slots to counter these attacks so you cannot really do it willy-nilly. Mirror Image can do it multiply times with one cast! Suddenly Mirror Image so great every one who has access to it must take it or fall behind, PC and NPC alike. And in a bit of irony in the quest to mitigate this second level spell DMs and PCs also need to find a way to reliably defeat illusions or waste valuable resources for no good reason- effectively neutering the entire illusion line of spells.

A vicious, needless cycle in my humble opinion.

Now, I realize every table can make their own rules, but this idea really seems to go against the design intentions of 5th ed. /steps off soap box.

This.

Not to mention that the idea of it being four duplicates as they envision it should be neutralized by one PC with a 10' pole. "I throw it horizontally into the square, noting which image actually makes contact." They can shift around all they like, but a lack of teleportation for the caster means that they just have to stare at the feet of whichever image they just saw the pole bounce off of. Meanwhile, the RAI interpretation would say, "Great, but you are seeing eight feet all in that area. You know his rough position, but you couldn't pinpoint where he actually is."

BB's interpretation creates a shell game, where one just has to keep track of the right image because they must all be apart so they can be targeted. The RAI creates an series of images that can be located and aimed generally at, but a specific form can't be singled out. Thus why only attacks, which can miss due to small faults in aiming, are affected.

Xetheral

2016-08-22, 12:11 AM

I ultimately remain in disagreement with this position because it makes zero sense to me from a "modeling something that's really happening in the game narrative" point of view. If the images are not targetable, what is the attack that fails to go for the real one hitting?

Put another way, this interpretation relies too much on "video game logic:" the images aren't really there; they're just a graphical effect prettied into place for the players' benefit. The characters and creatures don't even see them, because they can clearly tell where the real target is; they just aren't allowed to hit it three in four times they attack it. But they can clearly do so if they do anything that isn't an "attack."

5e is not meant to be that computer-code-esq; that was a feature (or bug, YMMV) of 3.5. 5e is meant to be much more a model of something that is "really happening" on some level in the narrative. So when mirror image creates illusory doubles, they're there. The effects of the spell surrounding not hitting the real caster are precisely because the creatures can't tell which is the caster and which are fakes. As long as the one attempting to apply an effect to the caster has to have a target for his effect, then, the images should fool him equally.

It shouldn't matter if Jimmy is trying to stab Joe with a sword, cast a spell on him, or pick his pocket; Jimmy should be equally confused in all three circumstances as to whether the "Joe" he's trying to do this to is the real one or an illusory fake. Moreover, if Jimmy can unerringly pick out the real Joe for pocket-picking or hold person, I contend that you're ignoring the RAW where the spell says that it's impossible to track which image is the real one.

In the abstract, it's fine to disagree from a "modeling something that's really happening in the game narrative" PoV--in many (most?) cases I'd agree with you! I loathe disassociated mechanics with a passion. But BurgerBeast has been insisting on a RAW analysis, and what you and I stylistically prefer the rules to be should have little bearing on what we each think is the most plausible meaning of the words on the page. (And, incidentally, I think it was 4e, not 3.5, that was known for its overabundance of disassociated mechanics.)

Even if this wasn't a RAW analysis though, the "attacks-only" mechanic of Mirror Image isn't necessarily disassociated. So long as the illusory duplicates are simultaneously countable and non-targetable, it makes perfectly good sense for attacks to be potentially diverted while spells that don't require attack rolls are not. So long as I understand you correctly, you've got a mental image in your mind that the spell creates individually-targetable duplicates, but nothing in the text of the spell requires this to be so. If you re-envision the descriptive text of the spell as creating non-targetable duplicates, the dissonance goes away.

For a specific example of how something can be countable yet non-targetable, consider the spokes on a wheel. They are certainly countable in a freeze-frame, but if the wheel is in motion the individual spokes quickly become impossible to track and not targetable individually by an attack--you have to take your chances. A spell that doesn't require an attack roll, however, can simply target the wheel itself and not need to worry about it.

So no, I'm not ignoring the RAW where the spell says that it's impossible to track which image is real. If anything, I'm taking that even farther by saying that they can't be targeted individually at all. Even someone with True Seeing would be unable to deliberately attack a duplicate, because they simply aren't targetable. (They can end up being the target (noun) of an attack if Mirror Image redirects an attack towards one, but they cannot be targeted (verb) by an opponent.)

----------

Don’t you think this is simply because they never expected a player to declare “I want to hit a duplicate – not the real Joe.” No player would ever try to hit the duplicate, and if they did, it would have the exact same effect as trying to hit Joe – it would come down to random chance.

If your attack bonus or damage is low enough and your enemy's AC is high enough, it's optimal to try to take out the duplicates to make it easier for your teammates (with higher attack bonuses or greater damage) to hit the target. So no, I think the lack of a mechanic to redirect an attack from a duplicate to the real caster is because such a situation is not possible under the RAW.

Yes, but the act of trying to target Joe is logically indistinguishable from trying to target a duplicate. If you try to hit Joe, you have a 25% chance to hit him and a 75% chance to hit a duplicate. If you try to hit a duplicate, you have a 25% chance to hit Joe and a 75% chance to hit a duplicate. It’s the same.

And yet they wrote the RAW in a way that applies only in one of two cases that would be logically indistinguishable under your interpretation that the images are individually targetable. I think this is strong evidence that, under RAW, the cases must indeed be logically distinguishable (by making impossible the case where you target a duplicate).

The images do not switch between being Joe or being his duplicate. Joe is one potential target. Each other potential target is a duplicate.

But, yeah, this all boils down to the fact that the player is literally incapable of specifying A, B, C, or D. So the player is incapable of targeting Joe and the player is incapable of targeting a duplicate. If he was capable, there would be no d20 mechanic. The d20 mechanic exists because it simulates the freeze frame. That’s why we have to use either the freeze-frame or the d20 mechanic.

To the contrary, under Segev's freeze-frame analysis, which image (A, B, C, or D) is real constantly changes. Link to Source (http://www.giantitp.com/forums/showsinglepost.php?p=21119323&postcount=168). Indeed, even which image is (e.g.) "Image A" constantly changes. His idea, assuming I was understanding it correctly, is that the player can target A, B, C, or D, and then, for an attack, the d20 resolution mechanic comes into play to determine if the targeted image happens to be real. (Segev, if I'm misunderstanding something about your analysis, please let me know.) I'm pointing out that, by the text of the spell, the d20 mechanic doesn't apply to attacks that happen to target one of the illusory images, which makes the spell's listed resolution mechanic fail to cover both of the two basic cases (attack's target is real/illusory) that are created by the assumption that images are individually targetable.

By contrast, the RAW is complete if an attacker must target the caster with an attack, despite not knowing exactly where the caster is. (Presumably the attacker does this by picking a point in space equally likely to contain any of the images when the attack arrives, but, as a RAW analysis, the quality of the how/why description doesn't matter much.) The provided d20 mechanic always triggers under this interpretation when there is an incoming attack.

The point is that none of the four is specifically targetable. By this I mean that the player cannot intentionally pick out Joe, nor can the player intetionally pick out a duplicate. What is ultimately targeted is up to random chance. But the player can declare the “intent” to target Joe, which triggers the mechanic.

Just to be clear, you are arguing that the attacker cannot choose to target either 1) the real Joe, or 2) a fake Joe, but that the attacker is still picking an image an attacking that image? In other words, the images are individually targetable, but not "specifically targetable"? Am I correct?

(regarding the case where the caster is successfully held)
This is everyone’s problem
This example gives the other explanation the same trouble that it gives me. Regardless of how you “fluff” the spell, it makes no sense that a held target can still benefit from moving illusory duplicates, so it’s irrelvant which side you pick – we share this problem in common. You can’t accuse me of violating RAW in this particular case without also accusing the alternative explanation. There’s no “out” that is exclusive to one explanation.

I disagree. If the images (both real and illusory) can't be targeted at all, then it doesn't matter if unusual conditions render it obvious which image is which. Sure, the provided explanation for how the spell works suddenly becomes disassociated, but that's not a problem in a RAW analysis, because it's a stylistic complaint (this would matter a lot more in a RAI analysis!). The targeting rules and the mechanic remain consistent (if ridiculous). By contrast, if the images are targetable, knowing which images are illusory is inconsistent with imposition of the (RAW-required) d20 mechanic (and equally ridiculous).

In my view, the rules don’t cover this interaction. There is nothing, however, RAW, to prevent it from happening.

At the table I'd probably treat it the same way you do. But that doesn't matter to a RAW analysis, so yes, there is RAW to prevent it from happening: RAW requires the d20 mechanic to be used whenever the caster is attacked, until the spell ends. There are stated exceptions, and being held isn't on the list. Under formal logic the enumeration of specific exceptions does not preclude the possibility that there are unstated exceptions, but under general principles of textual interpretation, lists of exceptions are assumed to be exclusive absent specific wording to the contrary. RAW, I don't see any way around the conclusion that a held caster is nonetheless protected by Mirror Image.

You might claim that by allowing the players to auto-target Joe, once he has been identified while held, violates RAW because targeting him specifically means that attack should be redirected. Yes, I am comfortable with it in this situation because it is clear to me that hold person has effectively removed the functionality of mirror image in this case. Also, we only ever arrived at this hypothetical when we violated RAW to say that the duplicates no longer move in the first place.

I'm glad you're willing to violate RAW. :) I am too. But you've framed this conversation as a debate over the RAW, and with my hypothetical I've identified an inconsistency between the assumption that images are individually targetable and the RAW that held targets are still protected by Mirror Image. My hypothetical does not in any way depend on concluding that the duplicates never move... it is agnostic on that point. Whether the images move or do not move, assuming the images to be individually-targetable is inconsistent with the RAW; assuming that the images are not individually-targetable is consistent with the RAW. (Although, again, both are equally disassociated in this edge case.)

There is simply no case in the context of D&D, where you can count four things, yet not target one of them, unless they are not targetable in principle. Yet illusions are, in principle, targetable.

Swarms. Swarms are composed of a countable number of creatures, each in principle targetable, but they are not targetable in the context of swarms. Similarly, an illusory creature is, in principle, targetable, but (I assert) is not targetable in the context of Mirror Image.

BurgerBeast

2016-08-22, 12:17 AM

The more I read this house rule that requires targeting images when the spell is not an attack, the more I dislike it. It makes a second level spell way too powerful.

So you still can't read?

Now Magic Missile, which is never supposed to miss- misses 75% of the time.

This is an inaccurate description of 5e magic missile. Nowhere does it say that the spell never misses. If you are referencing an earlier edition of D&D, you might also reference the older version of mirror image.

As does higher level spells like Otto's Irresistible Dance, the Powerword's, Maze and more... all thwarted by a 2nd level spell.

If anyone was dumb enough to waste their spells on a target who has mirror image up, I wouldn't blame anything except the caster's stupidity. I can understand the first time a player does it, but after that - no. Just no.

By the time you're capable of casting Otto's, fighters and rangers can strike the MI caster three times in one round, each, either doing damage or removing the effect.

Besides that (1) you can always cast it yourself, counterspell it, or dispel it.

At least with Counterspell there is a possible roll (for higher level spells) and it uses one of your own slots to counter these attacks so you cannot really do it willy-nilly.

Counterspell is way more powerful because (1) it stops all spells, not just save-spells, (2) it stops spells cast at anyone (not just you) outright, and (3) it is not as easy to anticipate (you can see mirror image, so no surprises).

Mirror Image can do it multiply times with one cast!

Again, only against idiots.

Suddenly Mirror Image so great every one who has access to it must take it or fall behind, PC and NPC alike. And in a bit of irony in the quest to mitigate this second level spell DMs and PCs also need to find a way to reliably defeat illusions or waste valuable resources for no good reason- effectively neutering the entire illusion line of spells.

A vicious, needless cycle in my humble opinion.

Drama, much? Mirror image still gets torn to shreds in a round by martial characters.

Now, I realize every table can make their own rules, but this idea really seems to go against the design intentions of 5th ed. /steps off soap box.

You're totally over-reacting to a spell that still doesn't do much.

This.

Not to mention that the idea of it being four duplicates as they envision it should be neutralized by one PC with a 10' pole.

"Three illusory duplicates of yourself appear in your space" (PHB 260). Yeah. I envision the spell as it says it works. Silly me.

"I throw it horizontally into the square, noting which image actually makes contact." They can shift around all they like, but a lack of teleportation for the caster means that they just have to stare at the feet of whichever image they just saw the pole bounce off of.

Nope. Staring at the image will not work because the duplicates "[shift] position so it’s impossible to track which image is real."

Meanwhile, the RAI interpretation would say, "Great, but you are seeing eight feet all in that area. You know his rough position, but you couldn't pinpoint where he actually is."

You realize that Joe and three duplicates can be described this way, right? You know that this whole false dichotomy you invented is not a direct consequence of acknowledging that "[t]hree illusory duplicates appear in your space", right?

BB's interpretation creates a shell game,...

By my view, you mean the view that "[t]hree illusory duplicates appear in your space", which is what the spell says.

...where one just has to keep track of the right image because they must all be apart so they can be targeted.

Things do not have to be apart so that they can be targeted. This has never been a requirement for being targeted, ever. Also, the rules specifically say "it’s impossible to track which image is real." So even if it was a "shell game," it would be an impossible shell game. So, no. Just no. As in, according to the rules, no.

You continuously jump to irrational conclusions that simply D.N.F. from what I say. You should just stick to making the game fun because logical conversation is clearly not your thing.

The RAI creates an series of images that can be located and aimed generally at, but a specific form can't be singled out.

This fits perfectly into RAW, and into my "interpretation." But when you attack (or cast a spell, you can't hit them all - you have to hit one - which is why a target must be selected randomly for any ability that requires a target [i.e. attacks and spells]).

Thus why only attacks, which can miss due to small faults in aiming, are affected.

This has no basis other than that you pulled it out of thin air. Also, it contradicts RAW.

TL DR all of the posts, but as a DM I would treat the hold person as an attack and roll the d20 as per the description in Mirror Image. If I roll up the real Joe, then I would give him a saving throw and apply the result to Joe. If I rolled up one of the fake Joes, I would fake roll a saving throw and tell the caster that Joe made his saving throw. This seem the RAW interpretation of this scenario to me.

It's that simple.

Erys

2016-08-22, 12:49 AM

Now Magic Missile, which is never supposed to miss- misses 75% of the time. This is an inaccurate description of 5e magic missile. Nowhere does it say that the spell never misses.

/facepalm

No point discussing anything with you if you don't even know the basics. No wonder your RAW is so far off.

Everything I said on that soap box is true. Just like every time I correct you about RAW I am telling the truth. But hey, your game, your nightmare.

BurgerBeast

2016-08-22, 01:32 AM

At the heart of this there is a semantic disagreement over the word targets in the sentence "Each time a creature targets you with an attack during the spell’s duration,...". I hope that it will come to light through this dialogue.

Even if this wasn't a RAW analysis though, the "attacks-only" mechanic of Mirror Image isn't necessarily disassociated. So long as the illusory duplicates are simultaneously countable and non-targetable, it makes perfectly good sense for attacks to be potentially diverted while spells that don't require attack rolls are not. So long as I understand you correctly, you've got a mental image in your mind that the spell creates individually-targetable duplicates, but nothing in the text of the spell requires this to be so.

Nothing in the text of mirror image does. However, the default condition of all things in D&D is that they are targetable.

If you re-envision the descriptive text of the spell as creating non-targetable duplicates, the dissonance goes away.

Mirror image doesn't suggest that duplicates are non-targetable. So they should default to the typical condition of illusory duplicates - they are targetable.

For a specific example of how something can be countable yet non-targetable, consider the spokes on a wheel.

This is a great example and I'm glad you brought it up. I had considered a slot machine or the star-mushroom-fireflower randomizer in Super Mario Bros. 3. Spokes on a wheel are great.

They are certainly countable in a freeze-frame, but if the wheel is in motion the individual spokes quickly become impossible to track and not targetable individually by an attack--you have to take your chances.

Perfect.

A spell that doesn't require an attack roll, however, can simply target the wheel itself and not need to worry about it.

Not if the spell description specifies that it targets "a spoke." Hold Person targets a creature.

So no, I'm not ignoring the RAW where the spell says that it's impossible to track which image is real. If anything, I'm taking that even farther by saying that they can't be targeted individually at all.

Good. But the spell can't effect affect the wheel as a whole. It can only affect one spoke. So how do you pick a spoke?

Even someone with True Seeing would be unable to deliberately attack a duplicate, because they simply aren't targetable.

This just doesn't make sense to me. A creature with true sight should be allowed, but by the RAW, to target an illusory duplicate. In my reading of the RAW, he is. In yours, he is not. "A creature is unaffected by this spell if it... can perceive illusions as false, as with true sight." This particular sentence makes sense by both interpretations, but only my interpretation allows a person under who can perceive illusions as false to still target them. I see this as a point for my team.

(They can end up being the target (noun) of an attack if Mirror Image redirects an attack towards one, but they cannot be targeted (verb) by an opponent.)

I get what you are saying here. I refute it but it's very hard to clearly say what I mean because of the semantics involved. Thank you for bringing in some better vocabulary for this.

So I do think that they can end up being the target of an attack. I also think they can be targeted.

----------

If your attack bonus or damage is low enough and your enemy's AC is high enough, it's optimal to try to take out the duplicates to make it easier for your teammates (with higher attack bonuses or greater damage) to hit the target. So no, I think the lack of a mechanic to redirect an attack from a duplicate to the real caster is because such a situation is not possible under the RAW.

But it is possible in either case. In the case where they are not targetable, the DM just interprets the player's desire as the desire to attack Joe (because he can't distinguish anyway) and resolves it in the same way. My way, if you want to target a duplicate, just state that you want to attack a duplicate. The DM will either roll a d4 or just use the same d20 mechanic.

And yet they wrote the RAW in a way that applies only in one of two cases that would be logically indistinguishable under your interpretation that the images are individually targetable. I think this is strong evidence that, under RAW, the cases must indeed be logically distinguishable (by making impossible the case where you target a duplicate).

I honestly think this goes to what the writer meant by "target." But ignoring this, in my view, either targeting a duplicate or casting a save-spell are both not subject to the d20 mechanic but cannot ignore that here there are four spokes on the wheel. Attacking a duplicate, or casting a save-spell, does not allow the player to target the whole wheel. A single spoke has to be the target.

To the contrary, under Segev's freeze-frame analysis, which image (A, B, C, or D) is real constantly changes. Link to Source (http://www.giantitp.com/forums/showsinglepost.php?p=21119323&postcount=168). Indeed, even which image is (e.g.) "Image A" constantly changes. His idea, assuming I was understanding it correctly, is that the player can target A, B, C, or D, and then, for an attack, the d20 resolution mechanic comes into play to determine if the targeted image happens to be real. (Segev, if I'm misunderstanding something about your analysis, please let me know.) I'm pointing out that, by the text of the spell, the d20 mechanic doesn't apply to attacks that happen to target one of the illusory images, which makes the spell's listed resolution mechanic fail to cover both of the two basic cases (attack's target is real/illusory) that are created by the assumption that images are individually targetable.

I concede straight out of the gate that I was probably misusing Segev's freeze-frame example. I thought it was a single photograph with A B C D labelled left-to-right. Then I thought you could take a second photo in which the spokes changed positions. So since A, B, C, D still refers to left-to-right labelling, the letters can switch as an artifact of the labeling process.

But let's ignore that since we're making good progress. My case is essentially that an attack or spell that target a spoke must ultimately effect a spoke and not the wheel.

By contrast, the RAW is complete if an attacker must target the caster with an attack, despite not knowing exactly where the caster is. (Presumably the attacker does this by picking a point in space equally likely to contain any of the images when the attack arrives, but, as a RAW analysis, the quality of the how/why description doesn't matter much.) The provided d20 mechanic always triggers under this interpretation when there is an incoming attack.

It always triggers under mine, too. In one form or another (and in fact, I'd probably use the d20 mechanic instead of the d4 if I could bring myself to forego mathematical accuracy in the interest of ease of play).

Just to be clear, you are arguing that the attacker cannot choose to target either 1) the real Joe, or 2) a fake Joe, but that the attacker is still picking an image an attacking that image? In other words, the images are individually targetable, but not "specifically targetable"? Am I correct?

I think this is accurate. As long as by attacker you mean the character. The character picks something and hits it. The player has no way of envisioning this choice or conveying this choice to the DM because only the character can see the situation in front of him. Thus, a randomized mechanic is the only sensible way to resolve the final target.

I disagree. If the images (both real and illusory) can't be targeted at all, then it doesn't matter if unusual conditions render it obvious which image is which. Sure, the provided explanation for how the spell works suddenly becomes disassociated, but that's not a problem in a RAW analysis, because it's a stylistic complaint (this would matter a lot more in a RAI analysis!). The targeting rules and the mechanic remain consistent (if ridiculous). By contrast, if the images are targetable, knowing which images are illusory is inconsistent with imposition of the (RAW-required) d20 mechanic (and equally ridiculous).

I don't want you to thin I'm nitpicking here or missing your point. I understand what you are saying. But as soon as you say "it doesn't matter if unusual conditions render it obvious which image is which" you violate RAW, in my view, because "it’s impossible to track which image is real" is RAW, in my view. So RAW is already violated at the point.

For me, it's only worse that in your view the player can now target the real Joe but not the duplicates, because this violates the general rules of D&D, which is a double violation in my view.

I can't quite make the leap to claim and say that once the target is held, the RAW permit the by-passing of the d20 mechanic (I have too much self respect), but the RAW come pretty darned close (much closer than I thought they would). The use of "track" goes a long way to helping in this regard.

At the table I'd probably treat it the same way you do.

Cool.

But that doesn't matter to a RAW analysis, so yes, there is RAW to prevent it from happening: RAW requires the d20 mechanic to be used whenever the caster is attacked, until the spell ends.

Yes. You are right. Again, I think there is still room for me to wiggle her here because of what the word "target" means in the spell text. It cannot meant target in the sense of "ultimately be the subject of the attack roll," because only the mechanic can do that. It must mean "intend to make (the target) the subject of the attack roll" so that the mechanic can make sense.

I then think the DM can use the specific interaction of hold person with mirror image to justify this ruling, nearly within (if not totally within) RAW.

Even still, despite all of this, I think my description violates RAW once, and in the case where it does, your description violates it twice. So I think that your concession (that you'd probably rule the way I do) carries more weight in terms of RAW than you are allowing it. It's not merely shared opinion - the rules actually do back the ruling.

There are stated exceptions, and being held isn't on the list. Under formal logic the enumeration of specific exceptions does not preclude the possibility that there are unstated exceptions, but under general principles of textual interpretation, lists of exceptions are assumed to be exclusive absent specific wording to the contrary. RAW, I don't see any way around the conclusion that a held caster is nonetheless protected by Mirror Image.

[edit: forgot to comment: the italics were added later]
If I understand you, we're in agreement here. I have a feeling that if I got deep in to this we might disagree here or there, but it doesn't seem to be a problem. If it is, let me know, please.

I'm glad you're willing to violate RAW. :) I am too. But you've framed this conversation as a debate over the RAW, and with my hypothetical I've identified an inconsistency between the assumption that images are individually targetable and the RAW that held targets are still protected by Mirror Image. My hypothetical does not in any way depend on concluding that the duplicates never move... it is agnostic on that point.

Absolutely. I think your hypothetical must allow the duplicates to move, otherwise it is not RAW. The reason my description fails is ultimately that I say the images do not move.

Whether the images move or do not move, assuming the images to be individually-targetable is inconsistent with the RAW; assuming that the images are not individually-targetable is consistent with the RAW. (Although, again, both are equally disassociated in this edge case.)

I disagree with basically all of your conclusions, here.

Allowing the images to be individually targetable need not be inconsistent with RAW, it's just that my interpretation is. Someone else who cares more about RAW could handle the held Joe situation by allowing the attacker to target Joe (triggering the d20 mechanic) or target a duplicate, which doesn't trigger a randomizer. This would fall within RAW but be ridiculous.

Furthermore, assuming the duplicates are not individually targetable is not consistent with RAW, on two counts. First, it creates allows for "unusual conditions" to "render it obvious which image is which" when, RAW, "it’s impossible to track which image is real." Second, it violates the default rules of D&D which are that anything is the game is targetable.

Swarms. Swarms are composed of a countable number of creatures, each in principle targetable, but they are not targetable in the context of swarms. Similarly, an illusory creature is, in principle, targetable, but (I assert) is not targetable in the context of Mirror Image.

I think this only helps my case. Swarms are the exception that proves the rule. The context of swarms is precisely what makes the group of creatures into one target, because it is one swarm by the rules.

In fact, swarms do not remove the ability to target a single creature, rather they bestow the ability to target the whole group. I would still allow a PC to try to grab a single unspecified creature or a single specified creature (much more difficult) out of a swarm.

Swarm rules are not there to prevent the targeting of individual creatures, they are there to make it possible to target an entire swarm. Without swarm rules, you would literally have to attack each and every spider in the swarm (which is possible).

BurgerBeast

2016-08-22, 01:36 AM

/facepalm

No point discussing anything with you if you don't even know the basics. No wonder your RAW is so far off.

Everything I said on that soap box is true. Just like every time I correct you about RAW I am telling the truth. But hey, your game, your nightmare.

Read the magic missile text. It doesn't say what you claim it says.

Just to be clear, I believe you are telling the truth about RAW. You're just wrong. In other words, I don't think you're lying. I just don't think you know what you're talking about.

Gwendol

2016-08-22, 03:14 AM

Xetheral

2016-08-22, 03:23 AM

I'm going to reply to the points where I think you want a reply, or I have something useful to say. If I miss any, let me know.

At the heart of this there is a semantic disagreement over the word targets in the sentence "Each time a creature targets you with an attack during the spell’s duration,...". I hope that it will come to light through this dialogue.

Given the sentence structure, I interpret the underlined "targets" in your quote as a verb. Accordingly, I believe it refers to the act of the attacker (whether PC or NPC) selecting the thing they are "attacking" (in the non-technical meaning of the word). The quoted text is ambiguous as to whether it is referring to the caster (i.e. the "wheel") or whether it is referring to the specific image of the caster that at this moment corresponds to the physical location of the caster (i.e. a particular "spoke"). I believe the former is more consistent with the RAW of the spell, while you believe the latter is more consistent with the RAW of the spell. I'm not seeing a semantic difference in our use of the word "target", merely a disagreement over what type of target the RAW anticipates. Can you clarify where you see a semantic disagreement?

Mirror image doesn't suggest that duplicates are non-targetable. So they should default to the typical condition of illusory duplicates - they are targetable.

The spell doesn't explicitly say one way or the other, but that's not the same as being silent on the question. The fact that no rules are provided for attacks targeted on a duplicate, and that no rules are provided for non-attacks targeted on either the original caster or a duplicate, are strongly indicative (but not conclusive) that neither situation was envisioned to arise. Accordingly, I do not think the default rules for illusory duplicates should apply (I'm also not convinced that there are rules broadly-applicable enough to qualify as "default", but since I believe my point stands even if there were default rules, I see no need to further go down that tangent.)

Not if the spell description specifies that it targets "a spoke." Hold Person targets a creature.

Good. But the spell can't effect affect the wheel as a whole. It can only affect one spoke. So how do you pick a spoke?

The wheel in my analogy represents the creature as a whole, including all conditions and effects currently present. The spokes of the wheel represent the visible manifestations of the caster's location, only one of which is accurate at any given moment. So if you're targeting "Joe" as a concept (not the best word, but I can't think of better offhand), you're attacking the wheel. If you're targeting (e.g.) "Joe's Image B", you're targeting a spoke.

To further clarify, in the absence of Mirror Image, the wheel would have a single spoke and be stationary. Effects that are not attacks would target the wheel as a whole. Effects that are attacks need to hit Joe in a vulnerable spot to get past his AC, and so target the spoke specifically. Once there are multiple, spinning spokes, however, you can't target any of them, and have to just aim for the wheel as a whole and hope you get lucky.

This just doesn't make sense to me. A creature with true sight should be allowed, but by the RAW, to target an illusory duplicate. In my reading of the RAW, he is. In yours, he is not. "A creature is unaffected by this spell if it... can perceive illusions as false, as with true sight." This particular sentence makes sense by both interpretations, but only my interpretation allows a person under who can perceive illusions as false to still target them. I see this as a point for my team. I don't see it as a point for either team. Whether or not a creature with True Sight should be allowed to deliberately target an image is a simple consequence of the contested targeting rules, and I don't see either outcome as more compatible with the RAW.

But it is possible in either case. In the case where they are not targetable, the DM just interprets the player's desire as the desire to attack Joe (because he can't distinguish anyway) and resolves it in the same way. My way, if you want to target a duplicate, just state that you want to attack a duplicate. The DM will either roll a d4 or just use the same d20 mechanic.

You're welcome to do so, but your interpretation requires supplementing the RAW with that ruling. My RAW interpretation has the virtue of not requiring an additional ruling to make the spell functional in the basic case of an incoming attack.

I concede straight out of the gate that I was probably misusing Segev's freeze-frame example. I thought it was a single photograph with A B C D labelled left-to-right. Then I thought you could take a second photo in which the spokes changed positions. So since A, B, C, D still refers to left-to-right labelling, the letters can switch as an artifact of the labeling process.

But let's ignore that since we're making good progress. My case is essentially that an attack or spell that target a spoke must ultimately effect a spoke and not the wheel.

Whereas I'd argue that non-attacks never target a spoke, and while attack normally target a spoke, the presence of multiple, spinning spokes makes that impossible.

It always triggers under mine, too. In one form or another (and in fact, I'd probably use the d20 mechanic instead of the d4 if I could bring myself to forego mathematical accuracy in the interest of ease of play).

Not by RAW it doesn't... you have to add a ruling.

I don't want you to thin I'm nitpicking here or missing your point. I understand what you are saying. But as soon as you say "it doesn't matter if unusual conditions render it obvious which image is which" you violate RAW, in my view, because "it’s impossible to track which image is real" is RAW, in my view. So RAW is already violated at the point.

For me, it's only worse that in your view the player can now target the real Joe but not the duplicates, because this violates the general rules of D&D, which is a double violation in my view.

I can't quite make the leap to claim and say that once the target is held, the RAW permit the by-passing of the d20 mechanic (I have too much self respect), but the RAW come pretty darned close (much closer than I thought they would). The use of "track" goes a long way to helping in this regard.

I'm not following something here, but I'm not sure whether you misunderstood my explanation of if I'm misunderstanding your reply. From my perspective, your images-are-targetable interpretation runs into consistency problems the moment unusual conditions render it obvious which image is which. Let me take another shot at explaining why:

In the normal case, an attacker is trying to hit the caster and Mirror Image prevents the attacks from knowing which image is real. Under your interpretation (including a beyond-RAW ruling that the d20 mechanic applies if a duplicate happens to be targeted), the attacker targets (e.g.) "Image A", and then the d20 mechanic is used to resolve whether "Image A" corresponds to the real caster. This is more-or-less functionally identical to my interpretation, which requires the attacker to target the creature in the abstract, and then uses the d20 to determine whether an image might be hit instead. They're functionally identical only because of the inability to discriminate between the images. If conditions change such that it is possible to know which image is real, then if you continue (as RAW requires) to apply to the d20 miss chance, you're effectively no longer permitting attackers to select an individual image to target. It's inconsistent.

By contrast, if you never permit the attacker to target an image in the first place, continuing to follow the (suddenly-ridiculous) RAW after conditions change is perfectly consistent.

And so long as you follow RAW and continue to enforce the d20 miss chance even on held casters, then the letter of the "can't track" clause has been adhered to, because the "can't track" mechanics are being followed, and there is no RAW violation. (Even if the result is headache-inducing.)

Even still, despite all of this, I think my description violates RAW once, and in the case where it does, your description violates it twice. So I think that your concession (that you'd probably rule the way I do) carries more weight in terms of RAW than you are allowing it. It's not merely shared opinion - the rules actually do back the ruling.

I don't see where I'm violating RAW even once, let alone twice. Assuming your opinion hasn't been changed by my re-explanation of the hypothetical above, can you rephrase where you see my interpretation violating RAW? Also, I'd likely rule as you do in the held case simply because avoiding headaches is more important to me than RAW, so I don't see how my concession gives any weight to your RAW analysis.

Allowing the images to be individually targetable need not be inconsistent with RAW, it's just that my interpretation is. Someone else who cares more about RAW could handle the held Joe situation by allowing the attacker to target Joe (triggering the d20 mechanic) or target a duplicate, which doesn't trigger a randomizer. This would fall within RAW but be ridiculous.

If the DM allows an attacker to target the image that corresponds to the real Joe, and then makes the attacker roll the d20 to check for a miss anyway, then they really aren't letting the attacker target the image that corresponds to the real Joe. Hence the inconsistency.

Furthermore, assuming the duplicates are not individually targetable is not consistent with RAW, on two counts. First, it creates allows for "unusual conditions" to "render it obvious which image is which" when, RAW, "it’s impossible to track which image is real." Second, it violates the default rules of D&D which are that anything is the game is targetable.

Count one is not a RAW violation, because I'm continuing to enforce the "impossible to track" mechanic even when a caster is held. Sure, it's visually obvious which one is which (the unmoving one is real) but you still can't track which one is real well enough to target it specifically (hence the headache, but no RAW violation). Sure, I'm going against the literal wording of the fluff, but when the fluff conflicts with the mechanics of Mirror Image and Hold Person, it's not a RAW violation to enforce the mechanics over the fluff. Count two isn't true at all--for example, there are no rules in 5e that say objects are generally attackable: more than one DM on these forums has suggested that such efforts should be modeled as ability checks rather than attacks. (I'm not sure I agree with them, but the point is that whether or not objects are targetable with attacks is up for debate, so you certainly can't say that there is a default rule that everything is targetable).

I think this only helps my case. Swarms are the exception that proves the rule. The context of swarms is precisely what makes the group of creatures into one target, because it is one swarm by the rules.

In fact, swarms do not remove the ability to target a single creature, rather they bestow the ability to target the whole group. I would still allow a PC to try to grab a single unspecified creature or a single specified creature (much more difficult) out of a swarm.

Swarm rules are not there to prevent the targeting of individual creatures, they are there to make it possible to target an entire swarm. Without swarm rules, you would literally have to attack each and every spider in the swarm (which is possible).

Can I get a page citation please on the rule that permits targeting individual members of a swarm? Generally in 5e you can't target specific parts of the opponent, so I think a specific rule would be needed to make an exception in the case of swarms. Without such a citation, I think creatures in a swarm are a perfect example of a countable, non-targetable creature, an example you contended does not exist in 5e.

smcmike

2016-08-22, 05:30 AM

I fail to see the disagreement here. Mirror Image creates three duplicates, making targeting a creature that you see impossible (you can't tell the duplicates apart from the original). Using a spell that requires an attack roll you can simply close your eyes and make the attack with disadvantage, otherwise use magic missile and get rid of the illusions.

Magic missile doesn't get rid of the illusions, unless you call it an attack (it isn't). Magic missile just hits the caster, because he is a creature you can see, and Mirror Image doesn't say anything about prevents spells.

BurgerBeast

2016-08-22, 05:42 AM

I'm going to reply to the points where I think you want a reply, or I have something useful to say. If I miss any, let me know.

Given the sentence structure, I interpret the underlined "targets" in your quote as a verb. Accordingly, I believe it refers to the act of the attacker (whether PC or NPC) selecting the thing they are "attacking" (in the non-technical meaning of the word). The quoted text is ambiguous as to whether it is referring to the caster (i.e. the "wheel") or whether it is referring to the specific image of the caster that at this moment corresponds to the physical location of the caster (i.e. a particular "spoke"). I believe the former is more consistent with the RAW of the spell, while you believe the latter is more consistent with the RAW of the spell. I'm not seeing a semantic difference in our use of the word "target", merely a disagreement over what type of target the RAW anticipates. Can you clarify where you see a semantic disagreement?

Target is definitely a verb in the sentence. I can try to illustrate the semantic problems. Let's imagine we have a single freeze frame of Joe and his three duplicates. For the sake of this example, Joe is A. B, C, and D are duplicates.

1. If the player says: "I attack Joe." - some people consider this targeting. So "declaring an intended target" is one definition.

2. In the context of the fiction, the character has to select a single image from the four identical images he sees - this can be considering targeting.

3. The DM has to resolve the actual target (i.e. the object against which the attack roll is made) - this can be considered targeting and is either very close to or the same as 2.

These can sometimes get complicated especially when people start just firing the word around in sentences. Someone might say, well once the player targets Joe, this activates the mirror image text. And then I might say, wait a minute, how does the player target Joe? Because simply saying "I attack Joe" does not mean that Joe has been targeted. We have mechanic for that.

Another example might be "Can hold person target an object?" Well, the answer is no in the sense that objects are not valid targets. But the answer might be yes in terms of the fact that a character can cast the spell at an object, but to no effect.

And then we get into ridiculous arguments. It only gets worse when negatives are introduced. There are worse sources of confusion that I know have come up, but I have difficulty recalling them and expressing them, because they come up in rare situations that usually already pretty deeply involved.

The spell doesn't explicitly say one way or the other, but that's not the same as being silent on the question.

I fail to see the distinction.

The fact that no rules are provided for attacks targeted on a duplicate, and that no rules are provided for non-attacks targeted on either the original caster or a duplicate, are strongly indicative (but not conclusive) that neither situation was envisioned to arise.

I tend to disagree. If you do hold this view though, how do you explain that having spells cast at Joe was not envisioned to arise?

Accordingly, I do not think the default rules for illusory duplicates should apply (I'm also not convinced that there are rules broadly-applicable enough to qualify as "default", but since I believe my point stands even if there were default rules, I see no need to further go down that tangent.)

I disagree here.

The wheel in my analogy represents the creature as a whole, including all conditions and effects currently present. The spokes of the wheel represent the visible manifestations of the caster's location, only one of which is accurate at any given moment. So if you're targeting "Joe" as a concept (not the best word, but I can't think of better offhand), you're attacking the wheel. If you're targeting (e.g.) "Joe's Image B", you're targeting a spoke.

Okay, well then your wheel analogy is different than I thought. I can't really reconcile it as described, either. I don't know why a caster would be a wheel but the spokes which are inside the wheel and spinning, would represent his location. It just doesn't make sense to me and it seems like there are much better analogies.

To further clarify, in the absence of Mirror Image, the wheel would have a single spoke and be stationary. Effects that are not attacks would target the wheel as a whole. Effects that are attacks need to hit Joe in a vulnerable spot to get past his AC, and so target the spoke specifically. Once there are multiple, spinning spokes, however, you can't target any of them, and have to just aim for the wheel as a whole and hope you get lucky.

But multiple spinning spokes, in the case, are analogous to multiple spinning vulnerable spots. Mirror image does not bestow multiple spinning vulnerable spots. I'm sorry but this makes no sense to me.

I don't see it as a point for either team. Whether or not a creature with True Sight should be allowed to deliberately target an image is a simple consequence of the contested targeting rules, and I don't see either outcome as more compatible with the RAW.

Yeah I would just rule that truesight means you don't see the duplicates at all. But if you rule that a character with truesight can see Joe and the duplicates, but knows the duplicates are false, it's hard to explain why he can't target Joe without triggering the d20 mechanic, and can't target the duplicates at all. In my interpretation he can attack either without triggering the mechanic.

You're welcome to do so, but your interpretation requires supplementing the RAW with that ruling. My RAW interpretation has the virtue of not requiring an additional ruling to make the spell functional in the basic case of an incoming attack.

No, it doesn't. I contend that this is precisely what is meant in the text by "[e]ach time a creature targets you with an attack." I think this means "each time a creature tries to target you with an attack." Or "each time a player declares that he targets you with an attack."

Whereas I'd argue that non-attacks never target a spoke, and while attack normally target a spoke, the presence of multiple, spinning spokes makes that impossible.

This is one of the examples I wanted to bring up. What does it mean to be impossible to target, in this sense? A character can knowingly swing, and fully anticipate that he will hit a spoke, so in this sense he can target a spoke, even though this is not what you mean. What you mean is that he can't say with any degree of certainty which spoke he will pick out of the spinning mass, so he can't target a spoke. In another sense, wherever he chooses to swing his sword, he'll target something. I contend that these are all different and lead to confusion.

I think the analogy is going to take some work to become useful.

Not by RAW it doesn't... you have to add a ruling.

Well, if my players went into a gambling hall and played roulette, I would quickly come up with a mechanic to resolve a 1-in-38 mechanic. In my view this is solution falls within RAW. This is not strictly RAW, but it falls within RAW. In the same way, any time a player does something in which the choice is not distinguishable to the player but will have relevant effects on the game world can be randomized within RAW.

I'm not following something here, but I'm not sure whether you misunderstood my explanation of if I'm misunderstanding your reply. From my perspective, your images-are-targetable interpretation runs into consistency problems the moment unusual conditions render it obvious which image is which. Let me take another shot at explaining why:

In the normal case, an attacker is trying to hit the caster and Mirror Image prevents the attacks from knowing which image is real. Under your interpretation (including a beyond-RAW ruling that the d20 mechanic applies if a duplicate happens to be targeted), the attacker targets (e.g.) "Image A", and then the d20 mechanic is used to resolve whether "Image A" corresponds to the real caster.

On the issue of beyond-RAW: No, it is not beyond-RAW. The situation calls for a randomization mechanic, and none is prescribed, so any mechanic can be used and is within RAW.

My Response: No. The attacker does not target "A" and then the d20 mechanic is used. I never double up on the mechanics. Either choose a letter or resolve with a die.

This is more-or-less functionally identical to my interpretation, which requires the attacker to target the creature in the abstract, and then uses the d20 to determine whether an image might be hit instead. They're functionally identical only because of the inability to discriminate between the images. If conditions change such that it is possible to know which image is real, then if you continue (as RAW requires) to apply to the d20 miss chance, you're effectively no longer permitting attackers to select an individual image to target. It's inconsistent.

It's inconsistent in the sense that the mechanic no longer applies. It's not inconsistent in light of the situation. On the first attack on a held Joe, you would have to use the mechanic. If the real Joe was hit, his position would be given away. You could follow RAW with some wonky explanation (as I said, there are precedents) but they're not to my taste.

By contrast, if you never permit the attacker to target an image in the first place, continuing to follow the (suddenly-ridiculous) RAW after conditions change is perfectly consistent.

Consistent with the mirror image spell rules, but totally inconsistent with the general rules of D&D, which allow people to attack illusions because people can do anything (We can get into the attack/Attack debate but it's not relevant in my opinion). I'd rather violate RAW in the case of one spell than violate a basic principle of the game.

And so long as you follow RAW and continue to enforce the d20 miss chance even on held casters, then the letter of the "can't track" clause has been adhered to, because the "can't track" mechanics are being followed, and there is no RAW violation. (Even if the result is headache-inducing.)

I still think there's an escape hatch here. The images are still not trackable. It's just that right now they are held. The moment they begin to move, they are not trackable. It seems to me that things that are not trackable can still be held.

Nonetheless I concede that your interpretation allows the spell to still function when the wizard is held whereas mine does not.

I don't see where I'm violating RAW even once, let alone twice. Assuming your opinion hasn't been changed by my re-explanation of the hypothetical above, can you rephrase where you see my interpretation violating RAW? Also, I'd likely rule as you do in the held case simply because avoiding headaches is more important to me than RAW, so I don't see how my concession gives any weight to your RAW analysis.

Well, I retract the particular charges that we there in light of your new explanation, but not these two: Not allowing the duplicates to be targeted violates RAW. Illusions, including illusory duplicates, are targetable like everything else. It adds a new function to save-spells in that they can ignore the need to select a target in at least one case (MI).

If the DM allows an attacker to target the image that corresponds to the real Joe, and then makes the attacker roll the d20 to check for a miss anyway, then they really aren't letting the attacker target the image that corresponds to the real Joe. Hence the inconsistency.

Yes. So just using the mechanic solves this. The fluff is easy enough to handle (since we're in the realm of ridiculous anyway) by just letting the duplicates whiz about and fool the attacker.

Count one is not a RAW violation, because I'm continuing to enforce the "impossible to track" mechanic even when a caster is held. Sure, it's visually obvious which one is which (the unmoving one is real) but you still can't track which one is real well enough to target it specifically (hence the headache, but no RAW violation). Sure, I'm going against the literal wording of the fluff, but when the fluff conflicts with the mechanics of Mirror Image and Hold Person, it's not a RAW violation to enforce the mechanics over the fluff.

I think we just disagree on what violates RAW, and perhaps on what constitutes fluff. If something is fluff, it should have no bearing on mechanics. If something intrudes on the mechanics of the game, I don;t consider it fluff.

Count two isn't true at all--for example, there are no rules in 5e that say objects are generally attackable: more than one DM on these forums has suggested that such efforts should be modeled as ability checks rather than attacks.

I'm not saying that they are attackable. I am saying they can be targeted with an attack. Everything in the game can be targeted by virtue of the fact hat it exists in a location, and that location can be attacked. How is resolved is of no consequence.

(I'm not sure I agree with them, but the point is that whether or not objects are targetable with attacks is up for debate, so you certainly can't say that there is a default rule that everything is targetable).

I do. Any location is targetable, and by extension any object is targetable, since it must exist in a location. Again, how it is resolved is of no consequence.

Can I get a page citation please on the rule that permits targeting individual members of a swarm?

No, because there isn't one. I'm not sure we're talking about the same thing though. I can't believe you'd argue with me about whether it's possible to kill one spider within a swarm of spiders.

Generally in 5e you can't target specific parts of the opponent, so I think a specific rule would be needed to make an exception in the case of swarms.

Why? A character specifies that he wants to kill one spider in a swarm. Okay, it's dead. Done.

Without such a citation, I think creatures in a swarm are a perfect example of a countable, non-targetable creature, an example you contended does not exist in 5e.

But swarms are countable and targetable. You can have 1, 2, or 3 swarms, and you can target them separately.

Gwendol

2016-08-22, 06:58 AM

Magic missile doesn't get rid of the illusions, unless you call it an attack (it isn't). Magic missile just hits the caster, because he is a creature you can see, and Mirror Image doesn't say anything about prevents spells.

Interesting. So your position is that Magic Missile, despite requiring a target, is not an attack because no attack roll is required?
But casting Magic Missile requires me to pick a target I can see, and with mirror image up I cannot distinguish between the illusions and the real target. Since Mirror Image lacks the wording of, say, Sanctuary
If the warded creature makes an attack or casts a spell
that affects an enemy creature, this spell ends
we are left guessing. Does mirror image protect against Attacks (requiring attack rolls) or attacks in the common sense, meaning actions meant to harm?

I would rule it's the latter, just because I suspect some weird situations will likely arise from the former.

smcmike

2016-08-22, 07:14 AM

Interesting. So your position is that Magic Missile, despite requiring a target, is not an attack because no attack roll is required?
But casting Magic Missile requires me to pick a target I can see, and with mirror image up I cannot distinguish between the illusions and the real target. Since Mirror Image lacks the wording of, say, Sanctuary
we are left guessing. Does mirror image protect against Attacks (requiring attack rolls) or attacks in the common sense, meaning actions meant to harm?

I would rule it's the latter, just because I suspect some weird situations will likely arise from the former.

It protects against Attacks.

A lot of things could potentially harm someone, but attacks are something specific in the game. Some spells include attacks. You can tell them apart from spells which do not include attacks, because the ones that include attacks say "make a [ranged or melee] attack."

The text of the spell explicitly states that effects other than attacks (Fireball, for example) don't do anything to the mirror images.

The text of Sanctuary is a great example, showing that not all spells which affect enemy creatures are attacks.

Gwendol

2016-08-22, 08:36 AM

smcmike

2016-08-22, 08:42 AM

Or they are just explicitly equating the two. Who knows?

This is definitely a situation where we could have used a little more words, but then again, mirror image has always been a difficult spell to deal with.

It's definitely not equating the two. It's listing two things.

Attack v non-attack spells is an easy distinction.

I think the design goal of this version of the spel was to make it very easy to deal with, compared with the 3.5 version (I don't know enough about 4e or earlier versions to make a comparison). If it works like I think it does, it's very simple, well-defined, and easy to adjudicate.

RickAllison

2016-08-22, 08:52 AM

Or they are just explicitly equating the two. Who knows?

This is definitely a situation where we could have used a little more words, but then again, mirror image has always been a difficult spell to deal with.

They've been pretty clear this edition on what is an attack. "If you are making an attack roll, you are making an attack," with two specific exceptions in that both grappling and shoving are called special melee attacks, but use ability contests.

So the checklist in 5e for an attack is (1) does it have an attack roll and (2) does it explicitly call it out as an attack? Many of the spell attacks actually do both.

Gwendol

2016-08-22, 09:03 AM

Yet, you have to pick a target when using magic missile, that you can see. Going by your view, illusions risk being completely useless as spell bait.

smcmike

2016-08-22, 09:05 AM

Yet, you have to pick a target when using magic missile, that you can see. Going by your view, illusions risk being completely useless as spell bait.

By our view, illusions created by Mirror Image are useless as spell bait (other than attack spells). Illusions created by other effects may be fair game.

Grayfigure

2016-08-22, 09:37 AM

I dont know if anyone has brought this up, but there are two points to consider:

Point 1: Mirror Image shares your square. but your square is a 5 feet by 5 feet block. You, as a player in the world, do not take up the entire square, nor most of it. There is barely enough room for you to share it with 3 other full versions of yourself (I say full versions because it is impossible to tell which one is real or not, because Magic). So there is just enough room for 4 versions of your self to move quickly enough that it is impossible to tell which is real (I say this because it explicitly states this is happening, and that is the effect).

Point 2: Hold Person is not a heat seeking missile. There is nothing in the description that states that it automagically discerns the real target. It is not in the same vein as Magic Missile, which is much closer to a heat seeking missile. It forgoes attack rolls, and goes straight to the target. And even in that case, Mirror Image would work against Magic Missile.....once. If Joe were lucky, 3 images would be blasted instead of him, and he would be bereft of Mirror Image, but still there. Even in this stipulation, Magic Missile has more of a locking effect than Hold Person.

IMO, If Silent Image can baffle Hold Person, so can Mirror Image. Hold Person must be aimed at a humanoid; it does not state if the humanoid has to be real, or can be simply perceived to be real. It does not aim at a square.

So the real question is: How do we perceive Hold Person functioning? Can it be aimed at a square? Can it be aimed at a real-person-as-perceived? Answering THAT question might go a long way toward getting the myriad sides of this debate to common ground.

smcmike

2016-08-22, 09:42 AM

Point 2: Hold Person is not a heat seeking missile. There is nothing in the description that states that it automagically discerns the real target. It is not in the same vein as Magic Missile, which is much closer to a heat seeking missile. It forgoes attack rolls, and goes straight to the target.

Hold person and magic missile use the exactly the same language for targeting - a creature within range that you can see.

Mirror Images are not damaged by Magic Missiles, even if Magic Missiles could hit them (which they can't), because Mirror Images ignore any damage that doesn't come from an attack.

Erys

2016-08-22, 09:52 AM

Again, this is remarkably simple. Spells do what they say. Nothing more.

Mirror Image states it works against Attacks and provides the mechanics for Attacks. It has no mechanics provided for spells/effects, because -by RAW- it does not interact with them. If it did, it would say.

Once you start adding lines of texts, regardless of how right you feel you are, you are no longer in RAW.

Yet, you have to pick a target when using magic missile, that you can see. Going by your view, illusions risk being completely useless as spell bait.

By our view, illusions created by Mirror Image are useless as spell bait (other than attack spells). Illusions created by other effects may be fair game.

To extrapolate the point.

Magic Missile hits the creature you aim for. Mirror Image makes a glamour that creates figments of the caster that move around it, all a few feet from it; this glamour may throw off the eye and cause an Attack to miss, but you know there is a creature there and that does not throw off the magic of MM.

This is vastly different then trying to hit a Silent Illusion that is 20' away from your true enemy.

Segev

2016-08-22, 09:53 AM

Breaking up a conversation with Xetheral for ease of reading:
But BurgerBeast has been insisting on a RAW analysis, and what you and I stylistically prefer the rules to be should have little bearing on what we each think is the most plausible meaning of the words on the page.Where you and I disagree - leaving aside BurgerBeast for the moment - is on which words on the page are "RAW" and which are not.

I contend that the fact that it creates images is RAW. Those images, per the RAW, make it impossible to tell which is the real person and which are just illusions.

You contend that the ONLY actual effect is the d20 check iff an "attack" is declared.

I contend that the RAW provide specific mechanics (the d20 procedure) for determining whether the real person was chosen or not in the case of an "attack," but that the fact they do not provide a specific mechanic for non-attacks doesn't change that there are still 4 images, and that one must be chosen, and that there's no way to make the choice with certainty that you're picking the real one.

I further contend that this is not only consistent with the RAW, but that it makes more sense and avoids dissociating mechanics from what the spell says it does in a narrative sense.

I agree that you CAN read the spell your way and argue that you're holding to the RAW. I don't agree that you are, but it is a very fine point whether the images' existence is RAW or fluff. But since it is at least equally valid to read the images as really being there according to the rules, and thus requiring a means of determining whether a given targeting effect was placed upon the correct target, and that interpreting it that way requires LESS dissociation in mechanics, the way I read it is the superior interpretation of the RAW in all respects.

(And, incidentally, I think it was 4e, not 3.5, that was known for its overabundance of disassociated mechanics.)Oh, it was. But 3.5 was infamous for rules doing illogical things if read strictly compared to what the spell, feat, or whatever SAID it did, descriptively. And for corner cases where otherwise-sensible effects of the RAW could be twisted to weird results.

And I'm way more familiar with 3.5; I enjoy it. I don't like 4e, so I don't discuss it and lack the depth of knowledge required to comment on specifics beyond the reasons I dislike it (which I will not go into here.)

For a specific example of how something can be countable yet non-targetable, consider the spokes on a wheel. They are certainly countable in a freeze-frame, but if the wheel is in motion the individual spokes quickly become impossible to track and not targetable individually by an attack--you have to take your chances. A spell that doesn't require an attack roll, however, can simply target the wheel itself and not need to worry about it.Assuming, for sake of argument, that you really are targeting individual spokes and not the whole wheel...

Let's go ahead and make a silly example to try to create an analog, here.

One of the many spokes on that wheel is actually a lich's phylactery. You have to hit it with the lich-slaying arrow (variant a) or the lich-slaying spell (variant b) to stop the lich from reforming in the next round (because it's been exactly long enough that he'll reform next round).

I find it unpersuasive that you can say, "I target the lich's phylactery-spoke with the spell" without knowing which of those spinning spokes is the right one, but you can't say "I target the lich's phylactery-spoke with the arrow" under the otherwise-identical circumstances.

And I do find the spinning wheel's spokes example to be a good analogy.

You have to pick one of the spokes as your target. You don't know which one is the right one. Determine if you guessed right.

So no, I'm not ignoring the RAW where the spell says that it's impossible to track which image is real. If anything, I'm taking that even farther by saying that they can't be targeted individually at all. Even someone with True Seeing would be unable to deliberately attack a duplicate, because they simply aren't targetable. (They can end up being the target (noun) of an attack if Mirror Image redirects an attack towards one, but they cannot be targeted (verb) by an opponent.)That doesn't make sense to me, either. Nothing in the RAW indicates this, and it would have to because normally, you absolutely can target an illusion deliberately.

Unless you're saying Brad can't swing his sword to pantomime cutting off the arm of an illusion of a troll, even if he knows it's an illusion.

Since Brad CAN do that, and mirror image doesn't say that you can't attack an illusion deliberately if you have some means of telling which is which (e.g. true seeing), the general case still applies and you can.
To the contrary, under Segev's freeze-frame analysis, which image (A, B, C, or D) is real constantly changes. Link to Source (http://www.giantitp.com/forums/showsinglepost.php?p=21119323&postcount=168). Indeed, even which image is (e.g.) "Image A" constantly changes. His idea, assuming I was understanding it correctly, is that the player can target A, B, C, or D, and then, for an attack, the d20 resolution mechanic comes into play to determine if the targeted image happens to be real. (Segev, if I'm misunderstanding something about your analysis, please let me know.)[/quote]This is accurate in broad concept. In essence, my contention is that Brad or Bob are picking A, B, C, or D, but since those labels are instantaneous ones, the colloquial use of "target Joe" means "Yes, we all know you want to target the real Joe, and since you can't tell which is the real Joe, we need to see if you picked the right one."

I'm pointing out that, by the text of the spell, the d20 mechanic doesn't apply to attacks that happen to target one of the illusory images, which makes the spell's listed resolution mechanic fail to cover both of the two basic cases (attack's target is real/illusory) that are created by the assumption that images are individually targetable.I disagree, here, because the d20 mechanic is how you determine if it was really the image you wanted to.

5e is very much not an effort to completely cover all cases and conditions, but rather to provide rules and guides to let you use what things say they do in a mechanical way. Extrapolation and rulings are fundamentally required throughout the game, by design. Like it or hate it, that's the way this edition is written. Intentionally.

The (highly reasonable, honestly) assumption is that Brad wants to attack the REAL Joe if he says "I attack Joe." If there's a legitimate reason Brad wants to target an image, it falls into the same "no RAW provided" situation as Bob wanting to target Joe with hold person. But it is again highly reasonable for the DM to rule that they use the same d20 mechanic to see if Brad hit the image he wanted to. If he didn't, it may be necessary to FURTHER test to see if the image he DID hit was Joe. (If Brad didn't have a specific image in mind, but just wanted to hit "an illusion," I'd personally use the SAME d20 mechanic, because Brad just now has a higher chance to get a desired target and a lower chance of being disappointed.)

Now, all of that is "rulings," because the RAW are silent. But the RAW being silent on how to perform the determination doesn't make that determination unnecessary.

By contrast, the RAW is complete if an attacker must target the caster with an attack, despite not knowing exactly where the caster is. (Presumably the attacker does this by picking a point in space equally likely to contain any of the images when the attack arrives, but, as a RAW analysis, the quality of the how/why description doesn't matter much.) The provided d20 mechanic always triggers under this interpretation when there is an incoming attack.I think "complete" is the wrong term, here, because it requires ignoring the RAW on what the spell says it does: creating images that can't be told apart from the caster.

The RAW are not "complete" in 5e. Their goal was to make it clear enough what things are doing and give guides to the DM as to how to rule on things. Give them room to extrapolate.

Arguing from a position of "read it this way, which doesn't make a lot of sense in connecting description to effects, because it allows us to claim the RAW are complete," is a very 3.5 approach. Maybe 4e, but I can't say either way on that edition.

Just to be clear, you are arguing that the attacker cannot choose to target either 1) the real Joe, or 2) a fake Joe, but that the attacker is still picking an image an attacking that image? In other words, the images are individually targetable, but not "specifically targetable"? Am I correct? I am arguing that each image is individually and specifically targetable, but that it's impossible to know whether the image targeted is the real caster or an illusion. Hence the need to use a resolution mechanic to see if you hit the "right" one.

Again, you're picking a freeze-frame A, B, C, or D, but no matter how many times you've already done this, you do not know whether A, B, C, or D is the same one you targeted last time.

Swarms. Swarms are composed of a countable number of creatures, each in principle targetable, but they are not targetable in the context of swarms. Similarly, an illusory creature is, in principle, targetable, but (I assert) is not targetable in the context of Mirror Image.
As somebody else (I think BurgerBeast) said, swarms are special exceptions to normal rules: they allow you to target the whole, where normally you could not.

I can't speak in depth to 5e swarms, but I know that 3e ones had a rule that single-target effects specifically did nothing to the "swarm-as-creature," because taking out a single critter in the swarm had no appreciable effect.

In short, they're not really analogous because the models used in the mechanics are specifically different under the RAW.