Immediate actions can interrupt

[Elves]

I've seen debate over when immediate actions can be used. When someone attacks you, can you abrupt jaunt away to foil their attack? Or do you have to declare the jaunt before they make the attack, since there's no discrete space between declaring the attack and making it?

Immediate actions were introduced in Miniatures Handbook, which includes this important bit that later definitions of immediate action, for some reason, omitted.

This action may interrupt other actions, taking effect just before they do. The last immediate action declared takes place first.

So there you have it. Even though these sentences weren't carried over to RC, they elaborate on what RC says rather than contradicting it, so I don't think they're overruled by primary source.


What I'm not 100% clear on is whether an interrupted action would be wasted. It sounds like the immediate action is really pre-empting, not interrupting, since it takes effect "just before" and "takes place first". You could argue that would allow the interrupted creature to take a different action in response, because their action never occurred. On the other hand, that would lead to "feint wars" where people declare actions in order to get their enemies to pop immediates, then do something completely different in response. That's not especially well-functioning gameplay so I think you do have to conclude that interrupts can waste actions, and that abrupt jaunt is indeed OP.

[thatothersting]

This all seems correct. Although, consider the following: any creature adjacent to you with a 15' reach that is making a melee attack (or one that is making a ranged attack whose range includes the square that you teleport to) should still have their attack resolve normally. They are targeting you, not your square; this might be a bit of a fiddly interpretation for ranged attacks (if the immediate action "interrupts" the attack despite naturally being declared after the one attacking has announced their attack then we have to wonder if the arrow is launched "just before" the attack completes, or if it's only in the air after the roll has been made, etc etc) but it seems obvious enough for things in melee. Attacks in D&D are abstractions in much the same way HP and AC are, after all - characters don't just stand there and get slapped in the face when their turn is over, and they don't take a single swing when they're trying to smack someone else, either. So, as silly as it may seem, teleporting within an aggressor's threatened area looks to me like it isn't enough to interrupt the attack, it merely changes which square you're standing in when you get your teeth knocked out. It foils the melee attacks of most humanoids and other medium sized creatures, but anything large that's wielding a spiked chain (or that is bigger than that or which has an abnormally long reach like a roper) simply won't care.

Still a bit more powerful than it probably should be, but not quite so bad as it looks at first.

[Gruftzwerg]

It's about the right timing between the "declaration of an action" and the "resolve of an action".

E.g. an enemy want to attack you.
1. he declares his attack
2. you have the option to interrupt with an immediate action (if you have one)
3. the attack gets resolved. (hit/miss.. damage..)

Which means, you have to use the immediate action before the attack & damage rolls in this chase. If the enemy already rolled, it's to late by RAW.

Most of the times people do step 1+3 in one go because no one has immediate actions to interrupt at all. But if someone has the option at the table, the players and the DM should try to declare everything and not directly jump on to the dice rolls.

[Elves]

Which means, you have to use the immediate action before the attack & damage rolls in this chase. If the enemy already rolled, it's to late by RAW.
I agree, you can't interrupt something that already happened.
What you can do, according to Miniatures Handbook, is interrupt the action when it's declared but before it's occurred. Without this quote from Miniatures Handbook, there's a case that that isn't possible (eg, the first answer in this thread (https://rpg.stackexchange.com/questions/35921/immediate-actions-during-other-actions)).

[KillianHawkeye]

I believe that the best usage scenario is for the user of an immediate action to always declare their interrupt of an impending action before knowing the result of said action. This obviously requires that everyone state what their combat actions will be before announcing the results of said actions, to give other combatants a chance to announce their interrupts.

Even when interrupt actions are not in play, this is a good general practice to follow to make sure the DM or the target player is prepared to receive an attack or other combat action before it's actually carried out. I also find it's a good habit to be descriptive in combat situations. People can usually see an attack coming before it hits them, for example, so as a DM I try to tell the player when an enemy is going to attack and then describe the enemy beginning their attack before rolling to determine the outcome. Ideally, I can get my players to do this as well (most of the time).

[Darg]

An immediate action works like a free action that works out of turn.

That said, actions only represent an amount of time. If a creature has a speed of 50, moves 40, and wants to attack the wizard who jaunts away; the creature has yet to attack because they haven't rolled and therefore still has that 10' movement to use.

At least that is how we do it. Spell casting is interruptable by the book, an attack has no rule for it. So we assume that the action isn't expended because the book doesn't say it is.

The concept of interruption happening before the action in question is not actually a concept of 3.5. Take bull rush for example, the AoOs happen after you have already moved into your opponent's square and therefore have a chance to hit the wrong person. The trigger has already happened. There is no rule to retroactively return to a state before the trigger happens. The feat stand still does not prevent a creature from moving at least one square. If the wizard abrupt jaunts because they were attacked, they were attacked and then jaunt away. If the wizard jaunts because they think they are going to be attacked, they jaunt before the action is expended.

[claypigeons]

An immediate action works like a free action that works out of turn.

That said, actions only represent an amount of time. If a creature has a speed of 50, moves 40, and wants to attack the wizard who jaunts away; the creature has yet to attack because they haven't rolled and therefore still has that 10' movement to use.

At least that is how we do it. Spell casting is interruptable by the book, an attack has no rule for it. So we assume that the action isn't expended because the book doesn't say it is.

The concept of interruption happening before the action in question is not actually a concept of 3.5. Take bull rush for example, the AoOs happen after you have already moved into your opponent's square and therefore have a chance to hit the wrong person. The trigger has already happened. There is no rule to retroactively return to a state before the trigger happens. The feat stand still does not prevent a creature from moving at least one square. If the wizard abrupt jaunts because they were attacked, they were attacked and then jaunt away. If the wizard jaunts because they think they are going to be attacked, they jaunt before the action is expended.

If you hate Abrupt Jaunt so much, just ban it.

[Gruftzwerg]

At least that is how we do it. Spell casting is interruptable by the book, an attack has no rule for it. So we assume that the action isn't expended because the book doesn't say it is.


Seems 100% RAW to me.

IIRC we have rules (don't ask me where...I'm still searching) that a declared action ain't spend until the action is resolved and may be re-declared if any interruption should occur. And as you said, we have rules that you can lose a spell due to interrupt.


The real question that remains is, do you lose the standard action with the spell together or is the action preserved. That's what I would like to know xD
Can the caster just cast another spell?^^

[Darg]

If you hate Abrupt Jaunt so much, just ban it.

Why do I have to hate abrupt jaunt? I think it is a perfectly usable ability. Jaunt behind cover, out of range of a full attack, make room for an ally to get in between, ruin a charge, etc. It has a lot of uses and is still an extremely strong ability in its own right. I just don't see the need to inflate the utility by fabricating a scenario that the rules or the ability itself does not cover.

By RAW, AoOs are legal melee attacks even though a creature has already moved out of your threatened area. I personally don't see a mechanical issue of doing the same thing for abrupt jaunt if one wants to use it in the middle of a swing.

[Elves]

A problem with having it not waste the action, as I mentioned in OP, is the possibility of "feinted" action declarations. Say a conjurer and a cleric are next to each other and you want to atk the cleric: why not first declare that you're attacking the wizard, hoping to get him to waste his immediate, then when he does so, just attack the cleric like you always wanted to?

Sure, maybe the wizard doesn't take the bait -- but in a conversational setting, what's to stop you from declaring a "feinted" action in hope of baiting (wasted) immediates and if no one takes the bait just saying "no, actually I'll do this instead". So it necessitates some strict rules about declared actions being final, but that interferes with the fact that in many cases players do genuinely rethink their actions at the last moment. Reflecting, I think it's more practical to treat the action as genuinely interrupted, not just pre-empted.

[Gruftzwerg]

By RAW, AoOs are legal melee attacks even though a creature has already moved out of your threatened area. I personally don't see a mechanical issue of doing the same thing for abrupt jaunt if one wants to use it in the middle of a swing.

It all boils down to when the Immediate action is taken and how many AoO the enemy has. Have a look at the following two scenarios:

A: Caster declares Abrupt Jaunt for some reason. > any enemy in melee range gets a AoO

B. Caster causes an AoO. > Enemy declares AoO. > Caster declares Abrupt Jaunt. > Enemy would need another AoO (Combat Reflexes) to attack the caster.

Multiple Immediate actions are handles the way trading card games handle em. You build up a "stack" of immediate effects on a regular action and handle em by the "Last in, first out!"-rule. This is the reason why the enemy in scenario B needs 2 AoO and can't re-declare his initial AoO since it is still on the "stack". This is still true even if the outcome of the stack would preserve his AoO.

edit: note that an AoO is by common sense an immediate action, but not as 3.5 defines immediate actions (regarding the swift action that you have to spend next turn). But AoO go on the "stack" (or build a stack) like Immediate actions.

[Asmotherion]

Can't seem to find a rule to back this up, but I think teleportation effects don't cause attacks of oportunity. Or that could be how I DMed it... In any case, don't take this as hard evidance. If anyone could contribute on this would be nice.

[Darg]

A problem with having it not waste the action, as I mentioned in OP, is the possibility of "feinted" action declarations. Say a conjurer and a cleric are next to each other and you want to atk the cleric: why not first declare that you're attacking the wizard, hoping to get him to waste his immediate, then when he does so, just attack the cleric like you always wanted to?

Sure, maybe the wizard doesn't take the bait -- but in a conversational setting, what's to stop you from declaring a "feinted" action in hope of baiting (wasted) immediates and if no one takes the bait just saying "no, actually I'll do this instead". So it necessitates some strict rules about declared actions being final, but that interferes with the fact that in many cases players do genuinely rethink their actions at the last moment. Reflecting, I think it's more practical to treat the action as genuinely interrupted, not just pre-empted.

I don't see why it has to be a game of feints. Per RAW there isn't a declaration phase and it's possible to declare the target after rolling the attack or even while rolling. If the DM is using their meta knowledge to "challenge" you, call them out on it. They should be playing the characters as if they were in the world. Besides making actions in anticipation is tactical. Making actions in response is simply reactionary.

Moving a character to threaten 2 targets at once has other reasons and implications as well, i.e. threaten attacks of opportunity or increase the number of targets possible. Going out of the way to create a situation not covered by the rules when it's perfectly capable of being within the purview of the rules is jumping down a rabbit hole.


Can't seem to find a rule to back this up, but I think teleportation effects don't cause attacks of oportunity. Or that could be how I DMed it... In any case, don't take this as hard evidance. If anyone could contribute on this would be nice.

They do not as it isn't movement as used by the rules. By the same token things done as part of movement can't be done while teleporting. Dimension Door uses the word "transfer" and teleport uses "transported."


Multiple Immediate actions are handles the way trading card games handle em. You build up a "stack" of immediate effects on a regular action and handle em by the "Last in, first out!"-rule. This is the reason why the enemy in scenario B needs 2 AoO and can't re-declare his initial AoO since it is still on the "stack". This is still true even if the outcome of the stack would preserve his AoO.

edit: note that an AoO is by common sense an immediate action, but not as 3.5 defines immediate actions (regarding the swift action that you have to spend next turn). But AoO go on the "stack" (or build a stack) like Immediate actions.

As per the PHB, an AoO happens after the trigger which then interrupts the flow of actions by happening immediately.

~~~~~~~~~~~~~~~~~

Something to note as to why WotC changed the wording of immediate actions: it wasn't a separate action. In the miniatures handbook it is a swift action that can happen at any time. The book states that feather fall is a swift action. Based on this it can be reasonably assumed that the rules for swift and immediate actions weren't fully fleshed out and the removal of the line in the OP was a deliberate change.

[Elves]

Based on this it can be reasonably assumed that the rules for swift and immediate actions weren't fully fleshed out and the removal of the line in the OP was a deliberate change.
How they work didn't change. You can still use an immediate as a swift on your turn, and your immediate still consumes your next swift. The only difference is that immediates aren't called a type of swift. That doesn't seem significant.

Eliminating the MH-only text doesn't result in a change, it results in less clarity as to the definition. By definition then it can't be read as a clear statement of intent.


This is different from the debate we had viz. PRCs about whether the invocation of primary source requires explicit contradiction. All MH does is give a more precise definition of what "can be taken at any time" means. More thoroughly defining a rule that exists is different from adding a new rule that isn't fundamentally incompatible with the existing one.

[Darg]

How they work didn't change. You can still use an immediate as a swift on your turn, and your immediate still consumes your next swift. The only difference is that immediates aren't called a type of swift. That doesn't seem significant.

Eliminating the MH-only text doesn't result in a change, it results in less clarity as to the definition. By definition then it can't be read as a clear statement of intent.


This is different from the debate we had viz. PRCs about whether the invocation of primary source requires explicit contradiction. All MH does is give a more precise definition of what "can be taken at any time" means. More thoroughly defining a rule that exists is different from adding a new rule that isn't fundamentally incompatible with the existing one.

The problem with the MH text is that immediate actions act like a swift action on your turn and swift actions do not happen just before the event that caused the response for it's use. The way it is now, immediate actions aren't initiative changing actions like a readied action, therefore you aren't acting at the initiative you take the action. At best I would say that immediate actions would resolve simultaneously like how the Counterfire ability works. Meaning you get the bonuses from Urgent Shield and Glimpse Peril, but Abrupt Jaunt would still have you hit by the attack roll (though it would prevent further attacks from a full attack possibly.)

[Elves]

The problem with the MH text is that immediate actions act like a swift action on your turn and swift actions do not happen just before the event that caused the response for it's use.
Immediate actions count as swift actions when taken on your turn, so that doesn't apply. I suppose you could argue that "counting" as a swift doesn't make it abide by the rules of one but at that point you're spinning oodles for the sake of it, as the intent is clearly stated.

the baseline is: there's been debate about whether immediates let you interrupt if they don't explicitly say so. Here, in the original source for immediate actions, is a statement that they do.

[hamishspence]

"Counters" in Book of Nine Swords, rely on the idea that an immediate action can go off after an opponent "makes an attack" but before the attack actually lands.

For the Wall of Blades (Iron Heart) counter - the trigger is the opponent making an attack, you make an attack roll - and you may use the higher of your attack roll or AC, against the incoming attack. And yes, you can decide to do so after you learn the result of the opponent's attack. Not all counters take place this "late" though - some take place slightly earlier, before you know the result of an attack.

So, it's

"Opponent makes a melee or ranged attack, rolls dice, then martial artist uses Wall of Blades counter to change their AC (if they get lucky) then the opponent's melee attack dice roll is compared to the martial artist's new AC, to see if it hits".

If it works for a Counter, it should work for an "Immediate action spell" as well - the same principles are in play.

[Railak]

My favorite immediate action thing comes from the shadowcaster 3rd level mystery Flicker. You cast the mystery as a standard action, then for a number of rounds based off level you can immediately teleport 5ft/2 levels every round as an immediate action. It actually has if used in reaction to an attack it grants 50% miss chance.

When I look at that it tells me that your guys argument that the attack isn't made just because you interrupt the attack to teleport away, and therefore are no longer a valid target is wrong. Now if you teleport before the attack is declared that's a different story, they aren't mid swing, they haven't even raised their weapon yet.

[Gruftzwerg]

My favorite immediate action thing comes from the shadowcaster 3rd level mystery Flicker. You cast the mystery as a standard action, then for a number of rounds based off level you can immediately teleport 5ft/2 levels every round as an immediate action. It actually has if used in reaction to an attack it grants 50% miss chance.

When I look at that it tells me that your guys argument that the attack isn't made just because you interrupt the attack to teleport away, and therefore are no longer a valid target is wrong. Now if you teleport before the attack is declared that's a different story, they aren't mid swing, they haven't even raised their weapon yet.
It's a specific ability and thus can't set any global rules. It can at best create an exception for its own niche.
Thus, sadly can't be taken as evidence for a global rule.

[hamishspence]

The global rule for immediate actions is that they can take place at any time.

As such, this can include "in the middle of another being's action".

They aren't required to take place "before" or "after".

[Darg]

It can happen at any time sure. There is also no rule that you need to declare your target before your attack roll or that an attack roll can be negated without specific exception. Flicker gives a specific exception, Abrupt Jaunt does not. So we are left with the question of the chicken and the egg with one side believing one over the other.

I believe the evidence hardline points to that both actions resolve simultaneously. The caster gets their AC rolled against and hit if the attack succeeds while porting out of the way of further attacks/AoOs.

At best we can say that flicker provides a penalty to it's specific scenario without it stating as such or we can say it is actually providing a benefit for the caster instead which is the normal paradigm.

[Elves]

The global rule for immediate actions is that they can take place at any time.

As such, this can include "in the middle of another being's action".

They aren't required to take place "before" or "after".
Unless an ability says it can do that, there's arguably no discrete period between the result of an attack and the damage roll. If the attack hits, the damage takes place.

Going by the MH quote, ("This action may interrupt other actions, taking effect just before they do") you can actually use an immediate to pre-empt an action at any point, even after damage has been dealt, as long as the action hasn't ended yet.

But for most standard actions, the damage roll is the end of the action, so the question of whether you can always or only by exception interrupt between attack roll and damage roll is relevant.


An alternate reading of "This action may interrupt other actions, taking effect just before they do" would be that if the action has already taken effect, the immediate can hardly take effect beforehand. In this reading you could only interrupt between declaration and execution of an action, and all other uses are by exception only.

That's probably wrong, but it does highlight the curious part of the MH quote which is it only provides for an immediate taking effect before an action, not during it. For example, someone with a 30 ft speed moves 20 ft next to you and you abrupt jaunt away. Previously we would have seen this as happening in the middle of the other creature's move action but per MH it actually happens before the other creature's move action. If we assume that this doesn't mean we have to replay things, it "takes effect" retroactively (so that, for example, if the moving creature has an aura that deals fire damage, the creature who jaunts away takes no damage even if they didn't jaunt away before being caught in the aura).

A different reading is that each space of movement counts as a discrete action for the purpose of interrupts, which isn't provided for in the text but may be the cleanest reading and lines up with conventional use.

[Darg]

I think the reason the MH text was left out was because then they would have to define "taking effect before they do" would be. It's pretty easy to say someone did something in response to an action. It's a lot harder and unfair to say they did it in response to a specific mechanical motion that is intrinsically part of an action.

I doubt anyone would allow a readied movement take full effect in the time between a raised arm and the downward swing of an attack even though it says it happens just before the trigger. A normal person wouldn't have the reaction speed to close a door on an arrow in flight originating 10' away even if they were prepared for it. The logical thing to do would be to close the door before the arrow even left the bow, meaning before the attack.

A readied action also doesn't negate the fact that the character can decide to change the result of the action they are taking. Take the readied arrow at the creature coming through the door example from the PHB. They might have planned to run up to the group of foes. After taking that critical to the face, such a plan of action would now be suicide so they decide to retreat instead of pressing forward. If they were forced to continue with the exact parameters of the action after getting suddenly attacked, the combat would not be as lively as that creature might have to move the next 50ft with the dread of knowing that their fate left their control even though if it were reality they'd be able to retreat to safety.

[Tzardok]

Don't the rules for readied actions say that you must finish the action interrupted by a readied action unless it has become impossible to do so?

[Darg]

Don't the rules for readied actions say that you must finish the action interrupted by a readied action unless it has become impossible to do so?

Nope.


Assuming he is still capable of doing so, he continues his actions once you complete your readied action.

Nothing about continuing the action exactly as initially intended and the plurality of "actions" implies a generality of the ability to continue making actions such as not being killed or disabled in some way to prevent the further taking of actions. It's not like walking into an invisible boulder would allow you to continue your movement nor would it prevent you from moving off in a different direction. The boulder would stop a charge but not normal movement. And it's not like a stopped charge would stop you from being able to continue taking actions either. A full attack isn't declared beforehand meaning you can attack, see how it goes, then decide how you want to proceed by making more attacks or take a move action. I don't see a reason why this wouldn't apply for any type of full round action that doesn't have something that makes it impossible to emulate the start of. If you charge, you don't have to attack at the end of it making it a double move without the -2 penalty. If you withdraw, the benefit happens at the start so if you negate the AoO it is the withdraw action.

Also, the full attack action says this:


You do not need to specify the targets of your attacks ahead of time. You can see how the earlier attacks turn out before assigning the later ones.

To me that seems like the declaration of a target is done as part of an attack, not the action itself.

[hamishspence]

A normal person wouldn't have the reaction speed to close a door on an arrow in flight originating 10' away even if they were prepared for it. The logical thing to do would be to close the door before the arrow even left the bow, meaning before the attack.

A martial artist in D&D can, with the right Counter, block an arrow with a blow from their sword. That doesn't take place before the arrow leaves the bow - that takes place as the arrow reaches them.

[Darg]

A martial artist in D&D can, with the right Counter, block an arrow with a blow from their sword. That doesn't take place before the arrow leaves the bow - that takes place as the arrow reaches them.

But that isn't a normal person now is it. The character requires training, signified by being a feat. IRL, such a feat requires the ability to anticipate and predict the flight path of the arrow. Then again, in game, deflecting an arrow doesn't happen just before the attack is made, it happens as the attack is made and is already going to hit you. Closing the door happens before the attack is made. The scenarios are different.

[Elves]

In the context of a readied action, it's nonsensical to let the interrupted creature choose a different action afterward, because then you have a readied action being triggered by an event that never takes place.

The same applies to immediate action abilities that are triggered by a certain event, but many aren't, so it's a little trickier.


A martial artist in D&D can, with the right Counter, block an arrow with a blow from their sword. That doesn't take place before the arrow leaves the bow - that takes place as the arrow reaches them.
I'd caution against extrapolating from specific immediate action abilities, because many have unique text detailing how they work -- for example, wings of cover is explicit that it doesn't waste the triggering attack, and wall of blades is explicit that it can be used after attack roll but before damage. But you could just as validly write immediate action abilities that say the opposite. In another example, close wounds "effectively prevents" damage the target has just taken, but that doesn't mean that's true of all immediate action healing. The use parameters of a particular ability don't necessarily generalize.

[Darg]

In the context of a readied action, it's nonsensical to let the interrupted creature choose a different action afterward, because then you have a readied action being triggered by an event that never takes place.

The opposite is also true though. If you teleport away before an attack, that attack never happened unless the declaration of the target of an attack happens at the same time as the attack roll giving you the ability to still hit another nearby target.

[Elves]

There are definitely still some inclarities in the use of immediate actions and to what extent "interrupted" actions are wasted.

What the MH quote does tell us is that they can interrupt and save you from other actions, and that they can be used when an action is declared but hasn't happened yet, contrary to arguments like this (https://rpg.stackexchange.com/questions/35921/immediate-actions-during-other-actions).

It also tells us that they take effect before the action they interrupt. That could be read to mean that they take effect retroactively, or it could be read to mean that they must be used before the interrupted action "takes effect" in order to have an effect.

How this interacts with discrete movement and multi-step actions like full attacks is sort of thorny; if willing to be a little flexible you could take the use of "action" in that quote as colloquial and treat individual decisions like each space of movement or each attack made as an "action" for this purpose, which isn't the strictest RAW but lines up better with conventional implementation.

There are a lot of specific immediate action abilities that have specific parameters for how they work; like I said I don't think those should all be treated as indications of how immediate actions work in general.

You could also argue that this text from MH is invalidated by the later definition in RC. I don't think that's necessarily the case, because this omitted sentence simply defines in more detail the meaning of what RC says.

Even if you don't agree, we don't know why the text was later omitted, but we do know that the text was written in the first place because they believed it. That's a stronger evidence of intent than the forerunning, which means it does provide a margin of positive evidence of intent.


TLDR while more clarity would still be good on immediates, this bit of text at least clarifies that they can be used to pre-empt declared actions, whereas some people had previously argued that there's no discrete "declaration" step -- that declaring an action is the same as beginning to execute it.

[Darg]

There are definitely still some inclarities in the use of immediate actions and to what extent "interrupted" actions are wasted.

What the MH quote does tell us is that they can interrupt and save you from other actions, and that they can be used when an action is declared but hasn't happened yet, contrary to arguments like this (https://rpg.stackexchange.com/questions/35921/immediate-actions-during-other-actions).

It also tells us that they take effect before the action they interrupt. That could be read to mean that they take effect retroactively, or it could be read to mean that they must be used before the interrupted action "takes effect" in order to have an effect.

How this interacts with discrete movement and multi-step actions like full attacks is sort of thorny; if willing to be a little flexible you could take the use of "action" in that quote as colloquial and treat individual decisions like each space of movement or each attack made as an "action" for this purpose, which isn't the strictest RAW but lines up better with conventional implementation.

There are a lot of specific immediate action abilities that have specific parameters for how they work; like I said I don't think those should all be treated as indications of how immediate actions work in general.

You could also argue that this text from MH is invalidated by the later definition in RC. I don't think that's necessarily the case, because this omitted sentence simply defines in more detail the meaning of what RC says.

Even if you don't agree, we don't know why the text was later omitted, but we do know that the text was written in the first place because they believed it. That's a stronger evidence of intent than the forerunning, which means it does provide a margin of positive evidence of intent.


TLDR while more clarity would still be good on immediates, this bit of text at least clarifies that they can be used to pre-empt declared actions, whereas some people had previously argued that there's no discrete "declaration" step -- that declaring an action is the same as beginning to execute it.

This thread got linked in a comment on that Q&A site at the bottom. Their comment mentioned that the definition of the immediate action was part of a different rule set. So I happened to find a book to look at and it was true. The definition for immediate actions are found in the glossary for skirmishes along with definitions for other things that do not match up with normal play. As this is so, the definition and rules for its use in MH are quite invalid because it doesn't pertain to normal D&D rules.

We now know why it was omitted in future rulebooks.

[Elves]

This thread got linked in a comment on that Q&A site at the bottom. Their comment mentioned that the definition of the immediate action was part of a different rule set. So I happened to find a book to look at and it was true. The definition for immediate actions are found in the glossary for skirmishes along with definitions for other things that do not match up with normal play. As this is so, the definition and rules for its use in MH are quite invalid because it doesn't pertain to normal D&D rules.

That's a good catch, and Hey I Can Chan, feel free to pipe in if you're reading this -- but given that immediate actions were carried over to the main rules with the exact same functionality, you can't deny that it's a strong statement of intent about how they're meant to work.


We now know why it was omitted in future rulebooks.
I disagree. Hey I Can is assuming:

a) that the sentence not being included in RC means the writers disagreed with it. But in many cases, RC or other sources omitted examples and explanatory text from the original source of a rule. That's why we sometimes go back to where rules first appeared for more context.

b) that its omission in RC provides any information at all. It doesn't -- it just removes information, creating more vagueness. That makes the positive evidence from the original appearance informative, regardless of source, because it's all we have to go on in resolving this ambiguity.

[Darg]

That's a good catch, and Hey I Can Chan, feel free to pipe in if you're reading this -- but given that immediate actions were carried over to the main rules with the exact same functionality, you can't deny that it's a strong statement of intent about how they're meant to work.


I disagree. Hey I Can is assuming:

a) that the sentence not being included in RC means the writers disagreed with it. But in many cases, RC or other sources omitted examples and explanatory text from the original source of a rule. That's why we sometimes go back to where rules first appeared for more context.

b) that its omission in RC provides any information at all. It doesn't -- it just removes information, creating more vagueness. That makes the positive evidence from the original appearance informative, regardless of source, because it's all we have to go on in resolving this ambiguity.

The problem is the the rules in the MH are different. The skirmish rules are not the rules for playing D&D. They share some similarities, but they are fundamentally different. Some actions are missing and others don't work the same. Examples are that you can't bull rush and charging doesn't provide the -2 penalty to AC nor the +2 to bull rush attempts because it doesn't exist.

Just because it is the first iteration of the rule, doesn't mean it is the first iteration for non-MH play. This means that the text not being a part of the rule first presented for 3.5 play is more than likely how they wanted it worded. Especially when you realize that the rules for swift and immediate actions were reprinted many times in the same way before its mention in the RC.

[Elves]

Our problem is RC's definition of immediate action doesn't provide enough information to tell when they can or can't be used. It leaves open several possibilities.

The skirmish rules aren't RAW for the main game. But there's an ambiguity in the immediate action rules, and the MH definition is the most authoritative source we have for disambiguating it.

If Hey I Can Chan wants to say "technically, my reading is viable", he can -- it's one of multiple viable readings under RC. But if we're making a good-faith effort to figure out how immediate actions should work given all the evidence at hand, MH is evidence.


This means that the text not being a part of the rule first presented for 3.5 play is more than likely how they wanted it worded.
We don't know why this sentence wasn't included. You can suggest that they wanted to change how it worked, but you could just as well argue that they thought the ability to interrupt was already implied in the free action rules (usable at any time). So the omission doesn't tell us anything in particular. OTOH, we do know that the MH wording is how they intended it to work in one particular context.



Imagine there's a new variant of chess called skirmish chess, which includes a new rule. Normal chess decides to incorporate the new rule, but doesn't include a clear definition. People argue over multiple ways the rule could work. Which would win? The best way is to refer to the definition from skirmish chess. So even though the skirmish chess rules aren't RAW for normal chess, they effectively end up being so.

[Darg]

The term interrupt in the PHB is used to describe 2 scenarios: interrupting a spell cast and to describe the immediate nature of an AoO as it interrupts the flow of actions in a round. The first is quite obvious on how it works. The second interrupts the flow of actions by happening the instant after the trigger goes off instead of waiting turns.


An attack of opportunity "interrupts" the normal flow of actions in the round. If an attack of opportunity is provoked, immediately resolve the attack of opportunity, then continue with the next character's turn (or complete the current turn, if the attack of opportunity was provoked in the midst of a character's turn).

Interrupting actions doesn't mean that it retroactively happens prior to an event in 3.5e. AoOs and readied action had to be given explicit permission to interrupt spell casts because the first happens after the action happens and the latter happens before the spell cast.

At this point there is little else to debate. One can use a rule for skirmish or one can use the rule provided for normal play. How one wants to rule it is up to them. I personally find little sense in the argument that using the skirmish rule prevents feint paradoxes. The paradox also happens when the event that causes the use of the action is prevented from happening, whether by killing the target or negating the action in it's entirety. Yes, I'm going to wait until they choose to attack me to simply use celerity to cast finger of death to kill them before they actually choose to attack me. It's silly. If you choose to decide that at the quickest they resolve simultaneously and they do nothing more than what is written, you remove all paradox. You don't waste any character's action based on speculation.

[Elves]

At this point there is little else to debate. One can use a rule for skirmish or one can use the rule provided for normal play.
And what is the rule provided for normal play? The whole reason we're discussing this is it's not clear how exactly it works.

You use an either-or phrasing, but it's not as if these two statements are in contrast to each other; they can function in complete agreement.

[Darg]

And what is the rule provided for normal play? The whole reason we're discussing this is it's not clear how exactly it works.

You use an either-or phrasing, but it's not as if these two statements are in contrast to each other; they can function in complete agreement.

It's usable at any time. It is that simple. It does not roll back time for themselves only. If the caster wants to jaunt before getting hit they can. All they have to do is pick a moment before the attack takes place and the creature continues from that moment to finish their turn, possibly doing something different because the jaunt happened prior to the event they wanted to prevent. You are right, they can work in agreement. I'm really just against any interpretation that causes a paradox mechanically. If it happens in response to an action taken it shouldn't prevent the action from taking place, nor should it require that the action continue even though the stimulus for the action no longer existed in the first place.

[hamishspence]

The idea for "Immediate Actions Interrupting" appears to be "in response to an attack beginning to take place".

Just as Feather Fall is "in response to the character beginning to fall".

So, just you can fall but take no damage (thanks to your Immediate Action) so you can have someone begin swinging their weapon at you but take no damage (thanks to your Immediate action allowing you to teleport out of the path of the swing).

No paradox in not taking Falling damage thanks to using an Immediate Action to interrupt the fall - no paradox in not taking melee damage thanks to using an Immediate Action to interrupt the attack.

[Darg]

The idea for "Immediate Actions Interrupting" appears to be "in response to an attack beginning to take place".

Just as Feather Fall is "in response to the character beginning to fall".

So, just you can fall but take no damage (thanks to your Immediate Action) so you can have someone begin swinging their weapon at you but take no damage (thanks to your Immediate action allowing you to teleport out of the path of the swing).

No paradox in not taking Falling damage thanks to using an Immediate Action to interrupt the fall - no paradox in not taking melee damage thanks to using an Immediate Action to interrupt the attack.

You're coming up with something not supported by rules. Attacks are resolved with attack rolls which are opposed by AC. If there is no AC to oppose the roll then there is no attack to miss. There is no "starting to attack."

Flicker is probably the best example. It either alters the rules to make immediate action teleport weaker or stronger. The likely outcome is that it was meant to be a benefit, not a penalty.

[Elves]

In the end, it's kind of a stupid discussion because we don't have enough information to definitely say which is right.

To start from the basics, one argument is that because an immediate action can be taken "at any time", you can use it at literally any point during a creature's turn, including when it's about to take an action.

Hey I Can Chan's counterpoint is that because the game lacks a codified declaration phase, declaring an action is the same as performing it -- rendering most actions un-interruptable.

I think his argument has merit, but considering the evidence we have, it seems significant that the original description of immediate actions does provide that any action can be interrupted.

His argument is also based on the fact that action declaration isn't a defined gameplay step the way it is in some games -- but since in practice it's something that occurs and that some mechanics (including some immediates) do interact with, it may not be true that it's mechanically insignificant.

Our choices are to either contact the old devs and try to get a firmer answer, or choose the readings we prefer.