Mirror Images, AoE damage, Magic Missile, and such

[BurgerBeast]

@Erys:

Fireball is, in your view, incapable of warming, moving, or heating anything, right? Because, RAW, fireball only causes damage to creatures and ignites flammable objects?

So, if a wizard cast fireball at a metal table, with one hundred metal cups full of ice-cold water, inside a typical dungeon room...

And then a character immediately sat down and had a drink of water, he’d find that the table was still cold, the water was still ice-cold, and the glasses were still ice-cold, right?

And any DM who ruled that the cups warmed up, even a few degrees, would be breaking the RAW, right? This is clearly houserule territory, right?

Because, nowhere in the description of the fireball spell does it say that fireball can warm objects. Indeed, fireballs are not even hot, by RAW. They can ignite flammable objects, but the fireball itself is not hot and cannot effect nonflammable objects in any way, right?

[Rakoa]

Yes, such a description of Fireball would be a houserule. However, it is one that most people would agree with. Can you say that most people are agreeing with your houserule?

[BurgerBeast]

Are you for real?

All rules in D&D work this way. They do what they say; nothing more and nothing less.

Does mirror image say that it creates three illusory duplicates?

When something is poorly written, which does happen, JC clarifies. You can take it or leave it.

“Poorly written” is a matter of interpretation. And what JC says is interpretation.

What he says is RAI, which reinforces RAW.

But RAI is still not RAW. We are talking about RAW.

When you don't, you are house ruling.

You’re conflating this by introducing RAI. When you don’t follow RAW, you’re houseruling. I think we agree on this.

In the cause of Mirror Image, it says there are 3 duplicates that occupy your space and proceeds to tell you exactly how those duplicates interact with the game rules.

Yes. But you’re going farther than “exactly.” You’re saying that it is a complete explanation. It’s pretty obvious to me that it is not a complete explanation. It explicitly states the conditions under which the exact explanations apply.

Simply having illusory duplicates present intrudes on game rules. Before you get into the specifics of attacks, you have interaction with the rules.

Magic missile is not one of the conditions that is explained.

You say this definitively means that the spell has absolutely no effect on magic missile. You say this. JC says this. The rules do not. Never.

[BurgerBeast]

Yes, such a description of Fireball would be a houserule. However, it is one that most people would agree with. Can you say that most people are agreeing with your houserule?

I say it’s not a houserule. I say that when the text of the spell says: “an explosion of flame,” then the spell creates an explosion of flame.

The DM is not houseruling when he rules that explosions can move things, or that flame is hot. Explosions can move things. Flame is hot. When the spell says it creates an explosion of flame, I say it actually does create an explosion of flame. Because it says it does. And a spell does what it says.

I bring this up because, as this makes apparent, the reason that Erys and I do not agree on this particular issue is grounded in the fact that we don’t even agree on what RAW means, what it means to go against RAW, and whether fluff is even a thing.

[Tanarii]

For all the people in favor of Magic Missile ignoring Mirror Image, how would you rule the spell's interaction with another kind of illusion? If you're generous with minor illusion, "I create a lifelike poster of myself (slightly crouched if you're above 5 ft) and I stand behind it. Can someone directly on the opposite side of that illusion target you with Magic Missile? They need to see their target and until they discern the illusion, all they can see and target is the false image.

Silent Image explicitly prevents sharing a space with the caster, but Major Illusion certainly doesn't have that caveat. Is everyone on the same page that you would not be able to target the caster in this case? They would be fully obscured by an illusion.

You can create illusions that conceal a caster, in many cases provided they don't physically interact with the caster and automatically reveal them to everyone. That's fine. If you cannot see the target you cannot target them.

Mirror Image doesn't do that. It creates a shifting set of images (that must be passing through each other and the caster), making it impossible to tell which is which. But that doesn't prevent you from seeing the creature at all times, allowing you to target "that guy over there with Mirror Image up".

Remember, nothing in Mirror Image indicates it is in any way creating independent self-targetable images for an attacker to pick amongst. Even attackers with attack rolls must be able to target the creature first.

When you house rule to say that you must pick among the images first, the process becomes:
1) pick an image. If it's not the caster, too bad.
2) if you pick the caster, check if it's an attack roll, then roll to redirect to image.

You've seriously buffed the spell with this house rule, obviously. It does double duty against attacks.

Another simpler house rule if you don't the specific case is:
Magic Missile inflicts individual Hits as if you had successfully made an attack roll.

That has various implications for the Magic Missile spell, of course.

[Hadoken]

All rules in D&D work this way. They do what they say; nothing more and nothing less.

I agree with you that 5e writes spells in a circumscribed way. RAW is to apply the specific rule of the spell description to override the general rule. In the absence of a specific rule in the spell description, the general rule applies. Mirror Image helpfully provides mechanical support to adjudicate overriding the general rule for targeting a creature with an attack in its description. Unfortunately, it doesn't provide a specific rule for how to target a creature with a non-attack effect that is under the effects of the spell. I think this must be due to one of two reasons. Either the designers didn't consider how non-attacks would interact with mirror image, in which case I think it's only prudent to conclude there is no definitive RAW answer for the issues posed to spells like magic missile brought up in this thread, or the designers considered mirror image as written to have already provided the answer for how it interacts with non-attack effects.

In the first case, oversight by WotC, Jeremy Crawford has been clear on his RAI interpretation. In the second case, presumed clarity, I think the more elegant answer for RAW is the one that requires no additional adjudication by the DM, namely that mirror image does not interact with non-attack effects. They bypass it entirely. If the mirror images were potential targets, it would be up to the DM to design their own rules for how to select a target from among the remaining images. That's certainly possible, and I wouldn't begrudge anyone doing it in their game. I wouldn't even mind playing in a game where that was the case since it would provide a rather nice buff to this second level spell. However, I think because the spell doesn't provide a specific ruling on this issue, RAW falls on the side of the general rule; non-attacks are unaffected.

As far as why mirror image acts this way and other illusion spells don't, I would say it's because, by the RAW, we (the omnipotent "we") know which duplicate is real and which is illusory, whereas mirror image makes it clear that its duplicates are indistinguishable. In the case of a spell caster duplicating themselves with major image, we know which one is real and which one is not (the one on the left, not the right, or whatever) so that it's possible to target the illusion and not the spell caster. For mirror image, we can't tell which is which. They share the same space, are constantly in motion. It's simply an unknown, even for the DM, in a way that other spell effects are not. This is what I think makes it a special case, and I believe is also why the spell itself tells you how to go about determining which image is hit when the spell caster is targeted by an attack. For the major image duplicate, the specific rule applies; you can target it because the spell tells us it can be interacted with. The nature of that interaction is left open-ended rather than circumscribed as mirror image is.

[BurgerBeast]

You can create illusions that conceal a caster, in many cases provided they don't physically interact with the caster and automatically reveal them to everyone. That's fine. If you cannot see the target you cannot target them.

The idea that an illusion is revealed as an illusion if anyone interacts with it is not at all clear. I read the rules to say that the illusion is revealed to the specific person who interacts with it.

The usual case is that it would be revealed, but this doesn’t have to be the case. So if you’re using this as the basis of an argument about mirror image I’d say you’re not necessarily on solid ground.

Mirror Image doesn't do that. It creates a shifting set of images (that must be passing through each other and the caster), making it impossible to tell which is which. But that doesn't prevent you from seeing the creature at all times, allowing you to target "that guy over there with Mirror Image up".

But it doesn’t let you target him. It forces a randomized roll to determine the target. So you actually can’t target him.

Remember, nothing in Mirror Image indicates it is in any way creating independent self-targetable images for an attacker to pick amongst.

Nothing except that they do get targeted (which implies they can be targeted), they get hit, and hits destroy them.

Even attackers with attack rolls must be able to target the creature first.

Right, but they can’t. Because when the player declares “I attack Bob,” Bob is not targeted. A roll decides what is targeted. The attacker cannot.

When you house rule to say that you must pick among the images first, the process becomes:
1) pick an image. If it's not the caster, too bad.
2) if you pick the caster, check if it's an attack roll, then roll to redirect to image.

You've seriously buffed the spell with this house rule, obviously. It does double duty against attacks.

This was addressed in the other thread. Nobody runs it this way. This is a straw man erected by people who are misrepresenting my argument.

The character fails to select a target, because the character cannot distinguish between them. The spell description provides a method for randomly determining the target.

[Xetheral]

Do you agree that people often (mis)use conditional statements to refer to biconditional relationships?

Yes.

So you agree that people often (mis)use conditional statements to refer to biconditional statements. So if n is the incidence of people using conditional statements to refer to biconditional relationships, we agree that n>0.

If so, why would it be reasonable to conclude that a conditional statement can never refer to a biconditional relationship? Isn't that inherently contradictory?

No. It’s always an invalid form of reasoning. If you are uncertain, you cannot claim certainty. Basically, if you’re not sure you need to ask.

But you're also claiming that conditional statements never refer to biconditional relationships, which requires that n=0.

Does not compute. The incidence can't both be greater than zero and equal to zero.

[BurgerBeast]

@Hadinen: thanks for the thoughtful response you. I disagree with you, as you already know, but you’ve articulated your point well.

I have this question:

I agree with you that 5e writes spells in a circumscribed way.

I could concede that this is generally true. The problem is, if you want to say that 5e writes Spell X in a circumscribed way, doesn’t that require you to look for evidence that it is circumscribed, first?

Because otherwise you’re voluntarily reading with bias, no?

I think that mirror image, based on its spell description, is clearly not circumscribed. If I were to enter this discussion with the assumption that it must be circumscribed because all 5e spells are, then I might see circumscription where it is not.

And this is essentially what I think is happening.

Either the designers didn't consider how non-attacks would interact with mirror image, in which case I think it's only prudent to conclude there is no definitive RAW answer for the issues posed to spells like magic missile brought up in this thread, or the designers considered mirror image as written to have already provided the answer for how it interacts with non-attack effects.

You are absolutely right here, except that I would say that even if the designers considered mirror image as written to have already provided the answer for how it interacts with non-attack effects, that doesn’t change what they wrote.

We have to read what they wrote and ignore what they meant to write, or thought they wrote, when discussing RAW. (Don’t we?)

So, from where I sit, I would say you are absolutely right to consider the two possibilities, but the very act of considering them is an exercise in RAI. Because whatever the designers think or thought, the words are the same in either case. The RAW do not change.

[BurgerBeast]

So you agree that people often (mis)use conditional statements to refer to biconditional statements. So if n is the incidence of people using conditional statements to refer to biconditional relationships, we agree that n>0.



But you're also claiming that conditional statements never refer to biconditional relationships, which requires that n=0.

Does not compute. The incidence can't both be greater than zero and equal to zero.

You’ve misrepresented me, here. I never said that conditional statements never refer To biconditional statements. I said it is always wrong to assert that a conditional statement refers to a biconditional statement. You need evidence of the second conditional to assert a biconditional.

So, even if you happen to be right, you were wrong to assert that you were right because you couldn’t have known.

- - -

So if four math students say:

2 x 2 = 4 because 2 groups of 2 are 4.

2 x 2 = 4 because when you skip-count by 2s, the 2nd number is 4.

2 x 2 = 4 because the square root of 4 is 2

2 x 2 = 4 because 2 + 2 = 4

The fourth student is wrong. It doesn’t matter that he is right about his answer and right about both claims. He’s wrong about causality, and he doesn’t know what the hell he’s talking about.

Does that make sense?

[Pharaon]

But it doesn’t let you target him. It forces a randomized roll to determine the target. So you actually can’t target him.

If you cannot target the caster, how is mirror image triggered?

Each time a creature targets you...

[BurgerBeast]

If you cannot target the caster, how is mirror image triggered?

This is precisely the heart of the matter. [edit: well more accurately, it's a different way of highlighting the heart of the matter]

In my view, there isn't anything to trigger. The duplicates are already there. When the player declares his intent to target the caster, this tells the DM that the character is trying to attack the caster. But the character cannot meaningfully select the caster, so this tells the DM to use the d20 mechanic to resolve the target.

I personally think this is more elegant.

Some of my opponents have the idea that something (presumably some form of magical defense) is triggered by attacks. I do not hold this view. And I see no evidence for this view in the spell.

[Erys]

@Erys:

Fireball is, in your view, incapable of warming, moving, or heating anything, right? Because, RAW, fireball only causes damage to creatures and ignites flammable objects?

So, if a wizard cast fireball at a metal table, with one hundred metal cups full of ice-cold water, inside a typical dungeon room...

And then a character immediately sat down and had a drink of water, he’d find that the table was still cold, the water was still ice-cold, and the glasses were still ice-cold, right?

And any DM who ruled that the cups warmed up, even a few degrees, would be breaking the RAW, right? This is clearly houserule territory, right?

Because, nowhere in the description of the fireball spell does it say that fireball can warm objects. Indeed, fireballs are not even hot, by RAW. They can ignite flammable objects, but the fireball itself is not hot and cannot effect nonflammable objects in any way, right?

Yes, such a description of Fireball would be a houserule. However, it is one that most people would agree with. Can you say that most people are agreeing with your houserule?

I say it’s not a houserule. I say that when the text of the spell says: “an explosion of flame,” then the spell creates an explosion of flame.

The DM is not houseruling when he rules that explosions can move things, or that flame is hot. Explosions can move things. Flame is hot. When the spell says it creates an explosion of flame, I say it actually does create an explosion of flame. Because it says it does. And a spell does what it says.

I bring this up because, as this makes apparent, the reason that Erys and I do not agree on this particular issue is grounded in the fact that we don’t even agree on what RAW means, what it means to go against RAW, and whether fluff is even a thing.

So you want to claim the fireball, by RAW, is an explosion that will heat the cups... but by such reasoning -> why are the cups still on the table? Shouldn't the explosion knock them to the ground and possibly tip the table over? Why are you limiting the spells description, it says "explosion"? If I want to knock someone off a bridge, an explosion should at least have a chance to do that... right?

Of course not.

The spell doesn't say it does any of that. So, by RAW, it doesn't. You are adding layers to the spell that are not part of the spell; ergo, you are house ruling when you say the fireball does anything more than X damage to creatures in Y area and catches unattended flammable items on fire. The same is happening with your interpretation of Mirror Image and its interactions with spells that do not have Attack rolls, such as Magic Missile or Finger of Death.

It is not to say some of these house rules are unreasonable, but they are still outside the scope of the RAW for the spells in question.

I agree with you that 5e writes spells in a circumscribed way. RAW is to apply the specific rule of the spell description to override the general rule. In the absence of a specific rule in the spell description, the general rule applies. Mirror Image helpfully provides mechanical support to adjudicate overriding the general rule for targeting a creature with an attack in its description. Unfortunately, it doesn't provide a specific rule for how to target a creature with a non-attack effect that is under the effects of the spell. I think this must be due to one of two reasons. Either the designers didn't consider how non-attacks would interact with mirror image, in which case I think it's only prudent to conclude there is no definitive RAW answer for the issues posed to spells like magic missile brought up in this thread, or the designers considered mirror image as written to have already provided the answer for how it interacts with non-attack effects.

In the first case, oversight by WotC, Jeremy Crawford has been clear on his RAI interpretation. In the second case, presumed clarity, I think the more elegant answer for RAW is the one that requires no additional adjudication by the DM, namely that mirror image does not interact with non-attack effects. They bypass it entirely. If the mirror images were potential targets, it would be up to the DM to design their own rules for how to select a target from among the remaining images. That's certainly possible, and I wouldn't begrudge anyone doing it in their game. I wouldn't even mind playing in a game where that was the case since it would provide a rather nice buff to this second level spell. However, I think because the spell doesn't provide a specific ruling on this issue, RAW falls on the side of the general rule; non-attacks are unaffected.

As far as why mirror image acts this way and other illusion spells don't, I would say it's because, by the RAW, we (the omnipotent "we") know which duplicate is real and which is illusory, whereas mirror image makes it clear that its duplicates are indistinguishable. In the case of a spell caster duplicating themselves with major image, we know which one is real and which one is not (the one on the left, not the right, or whatever) so that it's possible to target the illusion and not the spell caster. For mirror image, we can't tell which is which. They share the same space, are constantly in motion. It's simply an unknown, even for the DM, in a way that other spell effects are not. This is what I think makes it a special case, and I believe is also why the spell itself tells you how to go about determining which image is hit when the spell caster is targeted by an attack. For the major image duplicate, the specific rule applies; you can target it because the spell tells us it can be interacted with. The nature of that interaction is left open-ended rather than circumscribed as mirror image is.

Well said Hadoken. I think you are spot on with this assessment.

***

All that said, I think Pharaon just hit the home-run with that.

If you cannot target the caster, how is mirror image triggered?

This is precisely the heart of the matter.

In my view, there isn't anything to trigger. The duplicates are already there. When the player declares his intent to target the caster, this tells the DM that the character is trying to attack the caster. But the character cannot meaningfully select the caster, so this tells the DM to use the d20 mechanic to resolve the target.

I personally think this is more elegant.

Some of my opponents have the idea that something (presumably some form of magical defense) is triggered by attacks. I do not hold this view. And I see no evidence for this view in the spell.

If the caster of Mirror Image requires being targeted for his protection to work, and you are saying the spell prevents him from being targeted -> you just made the caster immune to all non-area effect spells that require saves. So, in order to correct this issue you are forced to disregard the very language of the spell that requires the caster to be targeted. Ergo, house rule.

[Pharaon]

When the player declares his intent to target the caster, this tells the DM that the character is trying to attack the caster.

The spell doesn't say "each time a creature declares you the intended target," it says "each time a creature targets you."

By "trigger" I mean that the player only knows to roll for mirror image after being targeted.

[RSP]

I still see a problem with your view, Burger. The RAW tells us what happens when the caster of MI is targeted, not what happens when the illusions are targeted. And, in general, the RAW doesn’t cover what happens when you try to target something not allowed by the spell description.

So you are houseruling how to handle that aspect (that is, it’s a houserule by you deciding the interaction is the same as if you target the caster of MI). Really, if you disregard the official ruling by JC and the RAI, MM should do whatever you’ve ruled is the result of selecting invalid targets (I think you posted earlier that the spell should fail and the slot be wasted, but it’s a long thread and I could be wrong).

Selecting an illusion shouldn’t interact with the MI at all because nothing in the spell description covers that.

Further, and I can’t believe this hasn’t been mentioned here (though perhaps I missed it), but RAW, if you allow non-attack spells to target the illusions created by MI, they wouldn’t be effected, as “If an attack hits a duplicate, the duplicate is destroyed. A duplicate can be destroyed only by an attack that hits it.” So anything targeting the duplicates that isn’t an attack wouldn’t destroy it.

I’m not sure if that’s been taken into account by Burger’s take on the spell.

[ThePolarBear]

The fourth student is wrong.

No, they aren't. Multiplications can be expressed as a string of additions, and that is actually how multiplication is defined as an operation.

On the topic:

RAW, for general targeting purposes, there's no requirement ANYWHERE to be able to distinguish, identify, or even know that what you want to target exists or is there to begin with. There's no method given to identify, choose, indicate what and where the target is.

Saying that you cannot distinguish between images and real does nothing, RAW, to prevent targeting a MI-affected caster via MM, since the only RAW rules to govern such targeting are:

1) it must be seen.
2) there must be a clear line traceable.

That's it. If you can't distinguish, but can see, then there's the ability to target. Anything else is adding to rules OR ruling as a DM, both things are possible by RAW, but not RAW themselves. Limiting targeting to a creature perception is something that is commonly done, but actually not required ANYWHERE in the rules.

edit: range notwithstanding

[EggKookoo]

RAW, for general targeting purposes, there's no requirement ANYWHERE to be able to distinguish, identify, or even know that what you want to target exists or is there to begin with. There's no method given to identify, choose, indicate what and where the target is.

I think you can hit an invisible creature with MM if you know it's nearby (per RAW). You can even hit an invisible creature whose location you're not even sure of with a regular attack; you just have disadvantage on the roll.

[Tanarii]

I think you can hit an invisible creature with MM if you know it's nearby (per RAW). You can even hit an invisible creature whose location you're not even sure of with a regular attack; you just have disadvantage on the roll.

You must be able to see a target for Magic Missiles.

The requirement to be able to see your targets is specific to individual spells. Each one that has that requirement states it in the spell description, like Magic Missile.

There's a few non-attack-roll cases where it should probably have been added. Acid Splash springs to mind.

[EggKookoo]

You must be able to see a target for Magic Missiles.

Ok, right, missed that.

[Lunali]

So if four math students say:

2 x 2 = 4 because 2 groups of 2 are 4.

2 x 2 = 4 because when you skip-count by 2s, the 2nd number is 4.

2 x 2 = 4 because the square root of 4 is 2

2 x 2 = 4 because 2 + 2 = 4

The fourth student is wrong. It doesn’t matter that he is right about his answer and right about both claims. He’s wrong about causality, and he doesn’t know what the hell he’s talking about.

Does that make sense?

I disagree with your example, the fourth is literally how multiplication was described when I first learned it, to expand on it a bit:
2 x 3 = 6 because 2 + 2 + 2 = 6

[Arial Black]

Do you aim magic missile?

I say you aim magic missile and you aim hold person. And then they unerringly hit the thing you aimed them at.

You say: you don’t aim it. It just always hits.

I say: you aim it, and then it always hits what you aimed it at.

Then you are provably wrong.

Whether you are striking with a melee weapon, firing a weapon at range, or making an attack roll as part of a spell, an attack has a simple structure.

1. Choose a target. Pick a target within your attack's range: a creature, an object, or a location

2. Determine modifiers. [snip]

3. Resolve the attack. You make the attack roll. On a hit, you roll damage, unless the particular attack has rules that specify otherwise. Some attacks cause special effects in addition to or instead of damage.

The reason I'm quoting this is not to prove that MM is not an attack, because this is beyond doubt and even you aren't trying to argue that it is.

I'm quoting it because you falsely assert that 'choosing a target' is one and the same thing as 'aiming at a target'.

What the PHB quote clearly demonstrates is that 'choosing a target' (step 1) is a completely different thing than 'aiming' at a target (step 3), being able to hit or miss.

You cannot 'miss' choosing your target! You cannot 'miss' at step 1. You can only 'miss' at step 3, and only if there IS a step 3!

For spells which require a target to be chosen, then they still have step 1, but they have no need for a step 2 or 3.

MM is such a spell. You (step 1) 'choose your target(s)'; no chance of 'missing' and choosing the wrong target!

IF the 'choosing your target' part could possibly miss, it would require an attack roll to resolve! Then step 1 would require its own attack roll just to see if you are allowed to make another attack roll at it later on at step 3!

[EggKookoo]

You cannot 'miss' choosing your target! You cannot 'miss' at step 1. You can only 'miss' at step 3, and only if there IS a step 3!

I think BurgerBeast is just saying when you 1) choose your target, you have no way of knowing if the target you chose is the actual creature or one of the illusions. Therefore your missile could unerringly strike an illusion. It's not an outrageous position to take and is sensible on the face of it.

I would still rule that MM just strikes the creature and ignores the effects of Mirror Image mainly because I think it makes MM cooler, and the spell is probably balanced to work that way. It gives the spell a little oomph, and its counter-intuitive targeting kind of adds to its supernatural quality.

As for how it manages to do that in-game, who knows? It's magic. Somehow it "knows" whats an illusion and what's a viable target.

[BurgerBeast]

Well, this conversation has been valuable for me. Granted, there will not ba any agreement reached anytime soon, but the conversation has drawn attention to the types of different assumptions that lead to the different thought processes.

So you want to claim the fireball, by RAW, is an explosion that will heat the cups... but by such reasoning -> why are the cups still on the table? Shouldn't the explosion knock them to the ground and possibly tip the table over? Why are you limiting the spells description, it says "explosion"? If I want to knock someone off a bridge, an explosion should at least have a chance to do that... right?

Of course not.

The spell doesn't say it does any of that. So, by RAW, it doesn't.

It’s not simply a matter of “of course” not, to me. Seriously.

The spell says that it creates an explosion of fire. And you repeatedly say that “a spell does what it says.” And I agree with that. This does not automatically mean that a spell creates overly dramatic effects, and especially not mechanical ones. But it does mean, absolutely, that an explosion of fire is created.

So, when you say that “a spell does what it says, nothing more, and nothing less,” this can be spliced into three claims:

1. A spell does what it says.
2. A spell does nothing more than what it says.
3. A spell does nothing less than what it says.

I think you’re failing to honour 1 and 3.

Also, I think you only consider mechanical effects in this, but that is not explicitly or implicitly a part of the caveat heuristic. You seem to be adding a restriction that spells descriptions should only be read mechanically and everything else ("fluff") should be ignored.

You are adding layers to the spell that are not part of the spell; ergo, you are house ruling when you say the fireball does anything more than X damage to creatures in Y area and catches unattended flammable items on fire.

Yeah, we disagree here. Because this is less than what the spell says. That’s a failure to honour “nothing less.” The spell does say other things, which you appear to ignore.

It is not to say some of these house rules are unreasonable, but they are still outside the scope of the RAW for the spells in question.

Essentially, this. You think my view of RAW is overly permissive. I think your view is overly narrow.

]If the caster of Mirror Image requires being targeted for his protection to work, and you are saying the spell prevents him from being targeted -> you just made the caster immune to all non-area effect spells that require saves.

I don’t think the caster requires being targeted for the spell to work. I think the spell starts working the moment it is cast until the moment it expires, even if the caster is never attacked.

In a nutshell, it appear to me that you view the game in a way that is almost exclusively mechanical. I do not. Everything that happens when the caster is targeted is purely mechanical, and largely arbitrary. For example, it was an arbitrary choice of the designers to have the caster roll the d20 (as opposed to the DM or the attacker). There is no insight to be gained from the fact that the wizard's player rolls the d20 to determine the target. In my view, people are mistakenly using this fact (that the wizard's player rolls) to attribute qualities to the spell that do not exist. Most obviously, this idea that the spell redirects attacks. That’s ridiculous, in my view. The player who rolls the d20 redirects the attack. This has no analog in the fiction. In the fiction, the attack hits one randomly selected target, regardless of how that is mechanically determined at the table.

So, in order to correct this issue you are forced to disregard the very language of the spell that requires the caster to be targeted. Ergo, house rule.

But then you are guilty of the same charge, because you disregard every aspect of the spell except the damage and the ignition property. Depending on how far you take this, it may not even be a fireball (in the way I would define a fireball) in your world. For example, by your method, it would be wrong to attribute any heat to the fireball at all. By my method, I find justification in the RAW: “an explosion of flame.” That is written in the spell. That’s a fact. If you ignore it, you are claiming that the spell does less than it says it does.

The spell doesn't say "each time a creature declares you the intended target," it says "each time a creature targets you."

By "trigger" I mean that the player only knows to roll for mirror image after being targeted.

Yes, we went through this in the other thread. I am confident that this confusion is based around the semantics of what it means to “target you.”

Feel free to explain what the difference between “targetting X” and “intending to target X” from the fictional creature's point of view. There isn’t any distinction to be made.

However, when you consider that there are players sitting around a table, it makes sense to differentiate between “intending to target X” and “targeting X.” Particularly in the case of indistinguishable targets, such as occur with mirror image.

Hadoken referred to this quite eloquently.

Spoiler: Hadoken’s eloquent explanation

Show

As far as why mirror image acts this way and other illusion spells don't, I would say it's because, by the RAW, we (the omnipotent "we") know which duplicate is real and which is illusory, whereas mirror image makes it clear that its duplicates are indistinguishable. In the case of a spell caster duplicating themselves with major image, we know which one is real and which one is not (the one on the left, not the right, or whatever) so that it's possible to target the illusion and not the spell caster. For mirror image, we can't tell which is which. They share the same space, are constantly in motion. It's simply an unknown, even for the DM, in a way that other spell effects are not. This is what I think makes it a special case, and I believe is also why the spell itself tells you how to go about determining which image is hit when the spell caster is targeted by an attack. For the major image duplicate, the specific rule applies; you can target it because the spell tells us it can be interacted with. The nature of that interaction is left open-ended rather than circumscribed as mirror image is.

I still see a problem with your view, Burger. The RAW tells us what happens when the caster of MI is targeted, not what happens when the illusions are targeted. And, in general, the RAW doesn’t cover what happens when you try to target something not allowed by the spell description. (emphasis added)

It’s not “not allowed by the spell description.” It’s never stated that the images cannot be targeted. Also, it is specifically stated that images can be targeted by the mechanic. So they can, in principle.

So you are houseruling how to handle that aspect (that is, it’s a houserule by you deciding the interaction is the same as if you target the caster of MI).

I’m not houseruling. I’m saying that since the spell text doesn’t say anything, the DM has to use a general rule or make a ruling.

I’m not deciding that the interaction is the same as if you were attacked. If I did, that would be a houserule – I agree. I’m saying that the DM has to make a ruling. Somehow the DM has to determine what happens when someone tries to target the caster with a spell, because the spell does not say. The DM can do this in any way that is appropriate.

Really, if you disregard the official ruling by JC and the RAI, MM should do whatever you’ve ruled is the result of selecting invalid targets (I think you posted earlier that the spell should fail and the slot be wasted, but it’s a long thread and I could be wrong).

Yes, if the invalid target is selected. That still has to be determined, because it’s possible to select the creature and not a duplicate. But then: yes. Exactly. This is exactly my point.

Selecting an illusion shouldn’t interact with the MI at all because nothing in the spell description covers that.

Again, I would say this is incorrect. If nothing in the spell description covers it, then you should use the general rules, because the specific rules don’t tell you what to do.

Further, and I can’t believe this hasn’t been mentioned here (though perhaps I missed it), but RAW, if you allow non-attack spells to target the illusions created by MI, they wouldn’t be effected, as “If an attack hits a duplicate, the duplicate is destroyed. A duplicate can be destroyed only by an attack that hits it.” So anything targeting the duplicates that isn’t an attack wouldn’t destroy it.

I’m not sure if that’s been taken into account by Burger’s take on the spell.

It has, in the other thread. It’s a can of worms, but the part of the text that claims to clarify what an attack is fails to do so. It just so happens to be the same sort of error that was made in the way people read mirror image.

No, they aren't. Multiplications can be expressed as a string of additions, and that is actually how multiplication is defined as an operation.

Okay, good. You’re right about this. But it drives even harder at my point. Whether or not the person is right still depends on what they meant. If they intended to represent a repeated multiplication then they are definitely correct. But you can’t be sure that it’s what they meant. If they intended to assert that multiplication and addition are the same thing, then they are incorrect. This can be tested by asking other questions, such as what is 2 x 3? If they say it’s five because 2 + 3 = 5, then they are obviously wrong in their thinking, even on the question of 2 x 2.

And this is my point. Invalid thinking is always invalid, even when the conclusion is correct.

On the topic:

RAW, for general targeting purposes, there's no requirement ANYWHERE to be able to distinguish, identify, or even know that what you want to target exists or is there to begin with. There's no method given to identify, choose, indicate what and where the target is.

Targets (PHB 204): “A typical spell requires you to pick one or more targets to be affected by the spell’s magic.”

The caster must pick a target. Not simply see a target.

Mirror Image (PHB 257): "Each dart hits a creature of your choice that you can see within range."

"Of your choice" = you choose from what you see.

Saying that you cannot distinguish between images and real does nothing, RAW, to prevent targeting a MI-affected caster via MM, since the only RAW rules to govern such targeting are:

1) it must be seen.
2) there must be a clear line traceable.

That's it. If you can't distinguish, but can see, then there's the ability to target. Anything else is adding to rules OR ruling as a DM, both things are possible by RAW, but not RAW themselves. Limiting targeting to a creature perception is something that is commonly done, but actually not required ANYWHERE in the rules.

edit: range notwithstanding

See above. If there are four identical “people,” and you want to hit one of them. You have pick which one. I’m not sure we really need to discuss this, though, because your proposition is absurd.

Player: “I cast hold person on the orc,”
DM: “Which orc? There are three.”
Player: “Nothing in the rules say I must select one! I see an orc. That is enough. I cast hold person on the orc!”
DM: …

The DM needs to know what to trace the line to, right? That has to come from the player.

Maybe I'm misunderstanding you?

[Arial Black]

I think BurgerBeast is just saying when you 1) choose your target, you have no way of knowing if the target you chose is the actual creature or one of the illusions. Therefore your missile could unerringly strike an illusion. It's not an outrageous position to take and is sensible on the face of it.

Your body becomes blurred, shifting and wavering to all who can see you.

So, when I 'choose my target' for my magic missile/hold person/power word kill I might aim at the wrong blurry outline and 'miss' my target?

How do we adjudicate that? By 'natural language'? Or by the rules?

RAW, spells do exactly what they say they do, for the purposes of how they interact with other RULES elements (as opposed to fluff).

We KNOW what effect blur has on other game rules; it says so in the spell description:-

For the duration, any creature has disadvantage on attack rolls against you.

That's what the spell does say. But wait! It doesn't say ANYTHING about how it interacts with spells that don't use attack rolls, so RAW it works against magic missile/hold person/finger of death too?

That is, of course, absurd. But it is the exact same 'justification' that BurgerBeast is using.

[BurgerBeast]

I disagree with your example, the fourth is literally how multiplication was described when I first learned it, to expand on it a bit:
2 x 3 = 6 because 2 + 2 + 2 = 6

Yep. I addressed this above.

Then you are provably wrong.

You’re wrong, bro. There’s a jch easier proof. What do you learn from the attack roll? Do you learn who you are aiming at, or do you learn if it hits or misses? You learn if it hits or misses.

The attack roll determines a hit or a miss. Nothing about what you’ve said is correct on this. At all. Not in the least.

I'm quoting it because you falsely assert that 'choosing a target' is one and the same thing as 'aiming at a target'.

That’s not exaclty what I’m claiming. I’m just saying you’re wrong, really.

I’m claiming that rolling an attack roll does not represent aiming. It represents hitting or missing. How do you know what you hit or missed? Well, that depends on what you aimed at before you rolled the attack roll.

You have to aim at something first, then you attack it. I’m not sure what is getting lost in translation, here.

What the PHB quote clearly demonstrates is that 'choosing a target' (step 1) is a completely different thing than 'aiming' at a target (step 3), being able to hit or miss.

Yes. Step 1 is deciding what to aim at.

You cannot 'miss' choosing your target! You cannot 'miss' at step 1. You can only 'miss' at step 3, and only if there IS a step 3!

Except that if you are trying to hit Bob, and you aim at Joe, you defintievely miss Bob, right? What are you playing at?

For spells which require a target to be chosen, then they still have step 1, but they have no need for a step 2 or 3.

Right. So if you want to hit Bob (who has mirror image up), but you select a duplicate as your target, then you can’t hit Bob, right?

In my view, mirror image interferes with step 1. The d20 roll effectively does step 1. It selects the target at which the attacker aims. If it comes up “duplicate” then it is a a guaranteed miss of Bob. But does it hit the duplicate? Roll an attack roll to find out.

So, when I 'choose my target' for my magic missile/hold person/power word kill I might aim at the wrong blurry outline and 'miss' my target?

Precisely.

How do we adjudicate that? By 'natural language'? Or by the rules?

By the rules. And this is the topic of the thread. Which rules? The spell doesn’t offer any.

That's what the spell does say. But wait! It doesn't say ANYTHING about how it interacts with spells that don't use attack rolls, so RAW it works against magic missile/hold person/finger of death too?

Well, that depends on the rules, doesn’t it? We have to think about the situation and the rules to figure out what happens when someone casts a targeted spell at a trget that he can’t distinguish from three others (if there are any). This has been my point all along. The spell doesn’t say, so it would be wrong to make an absolute claim about what the spell says when the spell doesn’t say ANYTHING.

So, the DM has to make a ruling.

If you think it is so wrong to conclude that the spell definitely works against magic missile (which I never did claim, by the way), then how is this any worse than claiming that the spell absolutely does not work against magic missile? – In light of the fact that the spell says NOTHING, both conclusions seem equally wrong to me.

That is, of course, absurd. But it is the exact same 'justification' that BurgerBeast is using.

If it is so obviously absurd, then feel free to tell me why.

[Arial Black]

1. A spell does what it says.
2. A spell does nothing more than what it says.
3. A spell does nothing less than what it says.

I think you’re failing to honour 1 and 3.

Also, I think you only consider mechanical effects in this, but that is not explicitly or implicitly a part of the caveat. You seem to be adding a restriction that spells descriptions should only be read mechanically and everything else ("fluff") should be ignored.

Add all the fluff you want, but in terms of how spells interact with other RULES elements, it is not a choice of 1, 2, or 3. It's all three. 'The truth, the whole truth, and nothing but the truth'. It is not 'The truth, the whole truth, OR nothing but the truth'.

I don’t think the caster requires being targeted for the spell to work. I think the spell starts working the moment it is cast until the moment it expires, even if the caster is never attacked.

Sure, the spell 'works' (ie is in existence) for its entire duration, just like a sword exists from the moment it is forged until the moment it is destroyed. But the spell only interacts with other game rules in the way that it says it does, WHEN it says it does. In the case of the sword, this is mainly when it actually hits a creature after a successful attack roll, but it might also be when it is the target of a spell etc. For MI, it interacts with other game rules exactly like it says it does, no more and no less, WHEN it says it does. And that is when the creature is the target of an attack.

In a nutshell, it appear to me that you view the game in a way that is almost exclusively mechanical. I do not. Everything that happens when the caster is targeted is purely mechanical, and largely arbitrary. For example, it was an arbitrary choice of the designers to have the caster roll the d20 (as opposed to the DM or the attacker). There is no insight to be gained from the fact that the wizard's player rolls the d20 to determine the target. In my view, people are mistakenly using this fact (that the wizard's player rolls) to attribute qualities to the spell that do not exist. Most obviously, this idea that the spell redirects attacks. That’s ridiculous, in my view. The player who rolls the d20 redirects the attack. This has no analog in the fiction. In the fiction, the attack hits one randomly selected target, regardless of how that is mechanically determined at the table.

Following your 'logic', then you must rule that the shield spell works like this (based on the spell description which says that it is cast as a Reaction when the caster is HIT by an attack):-

The orc throws a javelin. It HITS me. Javelins to damage, and that damage is enough to kill me because I only have 1hp left.

Then, AFTER it hits (and kills) me, I cast shield(!), and the javelin slides out of my head and I come back from the dead.

Or...could it be...that the game mechanics of that spell (triggers AFTER being hit) represent something else: the javelin would've hit me, but I managed to cast shield just in time to deflect the javelin so that it NEVER hit me at all!

But, how can something that never hits me at all trigger something that only gets triggered after it HITS?

Because the game mechanics actually ret-con the hit into a miss. No-one has any problems running this or conceptualising this. So why do you have such difficulty getting your head around mirror image?

Just like shield ret-cons a hit into a miss, mirror image ret-cons the choice of target. And it does so exactly the way it says it does!

The PLAYER targets the caster of MI, the SPELL EFFECT re-directs the target. The player themself NEVER targets an image.

It’s not “not allowed by the spell description.” It’s never stated that the images CANNOT be targeted.

There it is! The spell doesn't say I can't, so I can! Brilliant! Blur never says it doesn't protect me against power word kill, so it does! Prestidigitation NEVER says that it doesn't make wishes come true, so it does!

Also, it is specifically stated that images can be targeted by the mechanic. So they can, in principle.

Yeah. They can be targeted by the mechanic. The individual images cannot be targeted any other way, they cannot be affected by ANYTHING except what the spell says they can. The spell description says, "A duplicate can be destroyed ONLY by an attack that HITS it. It ignores ALL other damage and effects." Such as, y'know, being the target of ANYTHING, including attacks, because the attack targets the real creature.

If there are four identical “people,” and you want to hit one of them. You have pick which one.

No, you see ONE creature which has mirror image on. Just like you might see one creature with blur on. You don't have to choose which blurry outline you are casting power word kill at, and you don't have to choose which duplicate image you are attacking, you just choose the creature.

Player: “I cast hold person on the orc,”
DM: “Which orc? There are three.”
Player:There is one orc. The one with mirror image on.

FIFY.

[BurgerBeast]

Sure, the spell 'works' (ie is in existence) for its entire duration, just like a sword exists from the moment it is forged until the moment it is destroyed. But the spell only interacts with other game rules in the way that it says it does, WHEN it says it does.

Or in the way other rules say it does, when they say it does... no?

In the case of the sword, this is mainly when it actually hits a creature after a successful attack roll, but it might also be when it is the target of a spell etc.

But the rules for swords don't say they interact with spells. The rules for spells would say that. So you appear to agree that if other rules interact with swords, then they apply. I would add that if other rules interact with mirror image, they apply.

For MI, it interacts with other game rules exactly like it says it does, no more and no less, WHEN it says it does. And that is when the creature is the target of an attack.

So, I was right to say that you apply this "no more and no less" idea to mechanics only. Because the spell says it creates an explosion of fire. But you say that it does not create an explosion of fire, presumably because you write this off as fluff.

Following your 'logic', then you must rule that the shield spell works like this (based on the spell description which says that it is cast as a Reaction when the caster is HIT by an attack):-

The orc throws a javelin. It HITS me. Javelins to damage, and that damage is enough to kill me because I only have 1hp left.

Then, AFTER it hits (and kills) me, I cast shield(!), and the javelin slides out of my head and I come back from the dead.

Or...could it be...that the game mechanics of that spell (triggers AFTER being hit) represent something else: the javelin would've hit me, but I managed to cast shield just in time to deflect the javelin so that it NEVER hit me at all!

We don't disagree about how shield works. "An invisible barrier of magical force appears and protects you." It lasts until the start of your next turn. The +5 to AC is the result of the protection.

Mirror image creates three illusory duplicates. Nothing about protecting you. The d20 re-targetting mechanic is the result of the duplicates.

Pretty simple.

But, how can something that never hits me at all trigger something that only gets triggered after it HITS?

It doesn't. You're confusing the mechanics with the fiction. The hit result rolled at the table triggers the mechanical option to cast shield. If the wizard doesn't cast shield, or use some other ability to negate the hit, then it is confirmed that he is hit. If the wizard casts shield, then we realize that the hit never happened, neither mechanically (it was overwritten) nor in the fiction.

Because the game mechanics actually ret-con the hit into a miss. No-one has any problems running this or conceptualising this. So why do you have such difficulty getting your head around mirror image?

I don't. I just don't think mirror does this, because the spell doesn't say it does. Mirror images interferes with target selection. It happens before you roll the attack.

You can mechanically represent this as a ret-con of the act of actually targeting the caster, but that can only happen if the attacker actually has the ability to target the caster. He does not. He does not have the ability to pick a target.

Just like shield ret-cons a hit into a miss, mirror image ret-cons the choice of target. And it does so exactly the way it says it does!

No, it doesn't. You never have to cast shield unless you are hit. But mirror image is already cast, and it makes you difficult to target. Mirror image does nothing if you are successfully targeted. It makes you hard to target. From start to finish.

The PLAYER targets the caster of MI, the SPELL EFFECT re-directs the target. The player themself NEVER targets an image.

I disagree. I don’t think the spell re-directs the target. I think the spell is what causes the targetting difficulty. If anyone treis to target the caster with an attack, this is mechanically resolved at the table by rolling a d20.

There it is! The spell doesn't say I can't, so I can! Brilliant! (emphasis added)

I never said “I can.” I said nobody knows if I can, without looking to other rules. You, and Erys, and anyone else, cannot know the answer to whether I can or can’t. Because the spell doesn’t say, either way.

To contrast this, you are saying:

The spell doesn’t say I can’t, so I can’t!– sorry. This is also untrue. If you don’t know whether you can or can’t do something, you have to find out. You can’t make up the answer.

Blur never says it doesn't protect me against power word kill, so it does! Prestidigitation NEVER says that it doesn't make wishes come true, so it does!

Again, I never said it does. I said you can’t claim that it certainly does not. There are other rules in play, and they need to be considered.

Yeah. They can be targeted by the mechanic. The individual images cannot be targeted any other way, they cannot be affected by ANYTHING except what the spell says they can.

Where does the spell say that the images can only be targeted by the spell itself? Why couldn’t a player intentionally try to strike an illusory duplicate? Where in the rules is this option excluded?

The spell description says, "A duplicate can be destroyed ONLY by an attack that HITS it. It ignores ALL other damage and effects." Such as, y'know, being the target of ANYTHING, including attacks, because the attack targets the real creature.

There is so much wrong with this that it’s hard to unravel it. But:

Why is this sentence even necessary if only attacks can hit it? It’s not. So apparently other things can hit them. Other things just can’t destroy them.

No, you see ONE creature which has mirror image on.

No. You see one creature and three duplicates. That’s what the spell says. And you don’t know whether any one of them is a duplicate or the caster.

Just like you might see one creature with blur on. You don't have to choose which blurry outline you are casting power word kill at, and you don't have to choose which duplicate image you are attacking, you just choose the creature.

Blur doesn’t create duplicates. A spell does what it says, remember?

FIFY.

No, you really didn’t. In my example there were actually three orcs. No mirror image spell, at all.

So, it remains unanswered.

[Erys]

OK BurgerBeast, this is going exactly how it did last time. Nowhere.

The spell doesn't say "each time a creature declares you the intended target," it says "each time a creature targets you."

By "trigger" I mean that the player only knows to roll for mirror image after being targeted.

Yes, we went through this in the other thread. I am confident that this confusion is based around the semantics of what it means to “target you.”

Feel free to explain what the difference between “targetting X” and “intending to target X” from the fictional creature's point of view. There isn’t any distinction to be made.

However, when you consider that there are players sitting around a table, it makes sense to differentiate between “intending to target X” and “targeting X.” Particularly in the case of indistinguishable targets, such as occur with mirror image.

/Sigh...

Arial Black answered this with the PHB:

1. Choose a target. Pick a target within your attack's range: a creature, an object, or a location

2. Determine modifiers. [snip]

3. Resolve the attack. You make the attack roll. On a hit, you roll damage, unless the particular attack has rules that specify otherwise. Some attacks cause special effects in addition to or instead of damage.

It says black and white: choose a target. Not, 'intend to', it says "choose". As in, "I choose the creature using Mirror Image." It is that cut and dry.

By trying to argue semantics you are changing the core rules, parsing the language, and reinterpreting them to suit your stance.

And you know what, this is fine if you would own up that these are house rules. The fact you make these changes and claim you are in RAW is what irks me (and probably others).


***
Since you are dissecting people post line by line I want to revisit something you conveniently ignored:

You want to claim since the spell Mirror Image does more than protect against Attacks because it says 'the images are indistinguishable'; so lets go back to the Fireball example. The spell specifically says it makes 'an explosion': if I want to knock someone off a bridge- using the pretense that it makes an explosion and explosions create concussive force- do you allow it?

[BurgerBeast]

Arial Black answered this with the PHB:

It says black and white: choose a target. Not, 'intend to', it says "choose". As in, "I choose the creature using Mirror Image." It is that cut and dry.

But it’s not that cut and dry. Hadoken very eloquently explained why it is not cut and dry, in the case of mirror image. You praised him for the post.

By trying to argue semantics you are changing the core rules, parsing the language, and reinterpreting them to suit your stance.

I’m not arguing semantics.

And you know what, this is fine if you would own up that these are house rules. The fact you make these changes and claim you are in RAW is what irks me (and probably others).

I don’t care what irks you. I care about making sense. In my opinion you don’t know the difference between a house-rule and a ruling. That doesn’t bother me. I suggest you stop being irked and actually address my points.

***
Since you are dissecting people post line by line I want to revisit something you conveniently ignored:

You want to claim since the spell Mirror Image does more than protect against Attacks because it says 'the images are indistinguishable'; ...

No, I do not want to claim that. I am claiming that mirror image does less than protect from attacks. Mirror image does not protect against attacks. It creates duplicate targets.

...so lets go back to the Fireball example.

Sure, keeping in mind that mirror image does not have any additional function other than that it creates duplicates.

The spell specifically says it makes 'an explosion': if I want to knock someone off a bridge- using the pretense that it makes an explosion and explosions create concussive force- do you allow it?

No. The only reason you would even think this is because you are making up arguments that are not my arguments, and attributing them to me.

I even specified, in this in this very thread, that I would not insist on mechanical effects. But you can’t just pretend the spell does not create an explosion of flame. Not especially if your justification for denying the explosion of flame is: the spell does what it says. It says it creates an explosion of flame.

[Kane0]

And to think I've been using MM wrong all this time... I thought it was an excellent Mirror Image stripper, split three bolts between targets and you either get rid of all the images or two and hit the target. Pretty good trade for spell slots if you ask me.

But that's just my table, they like the flow of play and counterplay.