JC Update on bonus actions

[Tanarii]

I don't want to say any of the things your pointing out are wrong. But, the enemies in our games don't seem to lend themselves to situations where these considerations are "very high".

I'm sorry to hear that.

But I can see why you might not value the defensive side of a Shove.

[Elric VIII]

Has anyone pointed out that you can get nearly the exact same attack sequence using PAM with a 1-handed quarterstaff and shield. You use your first attack to attempt the shove action, then you get your remaining attacks with the bonus PAM hit making up for the shove.

The only difference is the flavor is changed and the damage dice are now lowered to 1d6/1d4 compared to the max 1d8. Of course, I'd use a glaive and GWM instead.

This seems like a knee-jerk nerf reaction to something that annoyed him in game.

[ProsecutorGodot]

Has anyone pointed out that you can get nearly the exact same attack sequence using PAM with a 1-handed quarterstaff and shield. You use your first attack to attempt the shove action, then you get your remaining attacks with the bonus PAM hit making up for the shove.

The only difference is the flavor is changed and the damage dice are now lowered to 1d6/1d4 compared to the max 1d8. Of course, I'd use a glaive and GWM instead.

This seems like a knee-jerk nerf reaction to something that annoyed him in game.

The key difference being that you aren't interrupting an action in a way that isn't allowed in the rules when you use PAM in this example. You're using one of your extra attacks as a shove, which is allowed as part of the attack action. This way you can still use the remaining attacks you might have and then the bonus action attack.

You can't nest actions (bonus actions are actions) between your attacks as JC has clarified, this is supported by RAW since movement is the only thing given specific mention as something you can do between the attacks in your action. As unpopular as this ruling is, it's now perfectly consistent with similar effects and rulings. You're free to treat is exactly as War Magic is in the Sage Advice Compendium where he says before or after attack action is fine.

I also think it's unfair to call this some sort of knee jerk moment when people have demanded a clear answer to this constantly for years. I'm just happy that there's now a clear and consistent answer that you are free to ignore at your own table. The only true loss is that if this makes it into the compendium or an errata, AL characters are going to be affected by it tremendously.

[JoeJ]

Has anyone pointed out that you can get nearly the exact same attack sequence using PAM with a 1-handed quarterstaff and shield. You use your first attack to attempt the shove action, then you get your remaining attacks with the bonus PAM hit making up for the shove.

The only difference is the flavor is changed and the damage dice are now lowered to 1d6/1d4 compared to the max 1d8. Of course, I'd use a glaive and GWM instead.

The problem with using a glaive is it's a two-handed weapon so you can't shove with a shield; you'd have to shove with your hand. Mechanically, yeah, it's the same, but it doesn't look as cool.

[Kaliayev]

Honestly, I think the true disease at work here is the "plain language" obsession the 5e rules have combined with trying to run something as mechanically tight and restrictive as AL. Writing your rules with DMs making calls at the table in mind doesn't work in a situation where multiple different DMs need to make the call for the same character and even have people looking over their shoulders while they do it. If the rules were written in a more legalistic way, then a lot of these unofficial and annoying twitter clarifications wouldn't need to happen, and all the symptoms like this would be much more limited.

Really, I think this isn't making a Mountain out of a Molehill at all, this sort of fundamental breakdown in communication and fairness is pretty much a worst case scenario of the problem and illustrates why it was a bad idea to begin with. If Crawford is so annoyed with having to make rulings in tweets he should've written rules in sharp ink instead of smudged pencil. And the worst part is, since this reversal is in a neutral zone between official and unofficial, AL doesn't really have motivation to issue a rebuild to people when their mechanics have been more or less yanked out from under them.

Well said.

I remember Mearls recently saying he wished they had designed the game without bonus actions.

Now I definitely wish they had.

At this point, I'm also thinking Mike Mearls is right and the bonus action system warrants a clear redesign that accommodates everyone. The approach the team is taking is increasing confusion and enmity. If you've been paying attention to the UA releases, you've noticed that they've actively moved toward what JC argues for in his tweets. For example, the UA order cleric's voice of authority and the XGtE racial feat orcish fury contain the words "immediately after" for specificity in timing. This shift ultimately acknowledges that this level of specificity in timing wasn't actively pursued with regard to bonus actions in PHB. In the PHB, this level of specificity only appeared for the bonus actions involved in flurry of blows and the charge feat (damage bonus). Now, one can approach the PHB's rules in one of two ways. One can approach with a formal logic structure where you have a strict, internal order of operations that can't be ignored. X then Y then Z. This is what the team is currently pushing. The other way uses an informal logic/RAF approach. This is the approach that your average player/table is going to take. Since the aforementioned specificity has limited application, the rules regarding user's choice in timing give players leeway with the timing of their shield shove and other such bonus actions.

A substantive element of this mess is the disagreement between 2016 SAC and 2017 SAC. The former claimed that the informal logic/RAF approach was intended. The latter reverses this stance, while making no mention of intention. Uh, okay? SAC effectively reached a point where it apparently disagreed with its own framework of RAW, RAI, RAF. The approach the team is currently taking makes the system clunkier for the average table, so it's effectively less enjoyable, and arguably less manageable, going forward. That's not good for the game overall.

P.S. I'm curious how all the formal logic structure advocates feel about the internal order of operations around component casting. I know it was brought up and quickly dismissed in this thread.

[RSP]

You can't nest actions (bonus actions are actions) between your attacks as JC has clarified, this is supported by RAW since movement is the only thing given specific mention as something you can do between the attacks in your action. As unpopular as this ruling is, it's now perfectly consistent with similar effects and rulings. You're free to treat is exactly as War Magic is in the Sage Advice Compendium where he says before or after attack action is fine.

What makes you think you can't have a BA in between attacks? From the same thread as the SM tweet, by JC:

"My tweet below was addressing bonus actions and reactions that have triggers. A bonus action that has no trigger—such as Cunning Action and the misty step spell—can take place whenever you want on your turn (PH, 189). #DnD"

The "tweet below" was referencing not taking BA's during an attack if they required a trigger (specifically a trigger of the Attack Action), and he says you could still take them in between attacks if the trigger is an attack. JC seems to specifically address that Misty Step works between attack as well, though the layout of multiple tweet threads is tough to put in a coherent layout.

Just wondering if I missed a tweet.

I also think it's unfair to call this some sort of knee jerk moment when people have demanded a clear answer to this constantly for years. I'm just happy that there's now a clear and consistent answer that you are free to ignore at your own table. The only true loss is that if this makes it into the compendium or an errata, AL characters are going to be affected by it tremendously.

I don't get this. His previous tweet was from 2015 and specifically addressed the SM bonus action Shove. Why were people demanding an answer if one had already been given (one that has the same exact authority as the current answer: also a tweet)?

[ThePolarBear]

Just wondering if I missed a tweet.

Having being confused for a bit i've come to the realization that both the first ("actions have coherence") and the second ("Misty Step can go between attacks") are intended and can work together coherently: The first is the baseline assumption, the second is a specific exception.

In short, the usual specific > general thing.

BUT, if the above is the correct interpretation, it means that a fighter5/whateverwithspellsX can't take the Attack action and attack once, cast a spell with action surge, then attack again, since the exception is made for bonus actions only, for what i can see. But i might be missing something somewhere.

[Elric VIII]

The key difference being that you aren't interrupting an action in a way that isn't allowed in the rules when you use PAM in this example. You're using one of your extra attacks as a shove, which is allowed as part of the attack action. This way you can still use the remaining attacks you might have and then the bonus action attack.

You can't nest actions (bonus actions are actions) between your attacks as JC has clarified, this is supported by RAW since movement is the only thing given specific mention as something you can do between the attacks in your action. As unpopular as this ruling is, it's now perfectly consistent with similar effects and rulings. You're free to treat is exactly as War Magic is in the Sage Advice Compendium where he says before or after attack action is fine.

I also think it's unfair to call this some sort of knee jerk moment when people have demanded a clear answer to this constantly for years. I'm just happy that there's now a clear and consistent answer that you are free to ignore at your own table. The only true loss is that if this makes it into the compendium or an errata, AL characters are going to be affected by it tremendously.

The issue is that if he thinks it's cheesy to shove first with SM then he must also have a problem with this.

I assume it was a knee-jerk response some he completely reversed his position in a way that ended up nerfing a feat that wasn't op, while still allowing that mechanic to function in other ways.

The problem with using a glaive is it's a two-handed weapon so you can't shove with a shield; you'd have to shove with your hand. Mechanically, yeah, it's the same, but it doesn't look as cool.

Precisely my point. One can still do the thing he calls a "cheese factory." You just use a different feat and mechanic.

[alchahest]

And the best part is you have to be one-handing a quarterstaff to do it, which while not game breakingly powerful is a super cheesey. you want to get rid of cheese, start with the one handed extra attacks PAM and a quarterstaff give you, hah

[RSP]

Having being confused for a bit i've come to the realization that both the first ("actions have coherence") and the second ("Misty Step can go between attacks") are intended and can work together coherently: The first is the baseline assumption, the second is a specific exception.

In short, the usual specific > general thing.

BUT, if the above is the correct interpretation, it means that a fighter5/whateverwithspellsX can't take the Attack action and attack once, cast a spell with action surge, then attack again, since the exception is made for bonus actions only, for what i can see. But i might be missing something somewhere.

After writing my prior post, I was wondering about Action Surge and if you'd be able to use it to interject and action within another action.

Correct me if I'm wrong, but there isn't anything that actually states that actions cannot occur together?

I actually don't think the system works if action cannot overlap:

A Monk who takes the Dodge Action as a BA using Patient Defense can then overlap that Action with the Attack Action. That is, the Dodge Action isn't completed; in fact, it lasts until the start of the Monk's next turn. (Compare this to Shield, which is a split second Reaction and completed, though the Magic defense remains for a similar duration.)

Likewise, a fifth level fighter using Action Surge can use the Attack Action to make one attack, then Disengage, move another 15' and complete their attack action on a different target.

Again, these all seem to be RAI and I'm unaware of anything in the RAW which contradicts these. The only tweet I'm aware of that even suggests it are the ones we're discussing and JC specifies the restriction is with BAs that have a condition.

[ThePolarBear]

After writing my prior post, I was wondering about Action Surge and if you'd be able to use it to interject and action within another action.

Correct me if I'm wrong, but there isn't anything that actually states that actions cannot occur together?

I actually don't think the system works if action cannot overlap:

A Monk who takes the Dodge Action as a BA using Patient Defense can then overlap that Action with the Attack Action. That is, the Dodge Action isn't completed; in fact, it lasts until the start of the Monk's next turn. (Compare this to Shield, which is a split second Reaction and completed, though the Magic defense remains for a similar duration.)

Likewise, a fifth level fighter using Action Surge can use the Attack Action to make one attack, then Disengage, move another 15' and complete their attack action on a different target.

Again, these all seem to be RAI and I'm unaware of anything in the RAW which contradicts these. The only tweet I'm aware of that even suggests it are the ones we're discussing and JC specifies the restriction is with BAs that have a condition.

I can only repost a tweet from JC, again.

I'm not aware of any rule that prevents it, but apparently that is the point. There's no rule, so you can't do it. But, again, action surge might be intended to be an exception and the whole lawyering we are doing wasn't really expected.

The fighter is doing 2 different actions, and those apparently do not have the "nesting" proprieties of certain reactions and bonus actions. Mind you, i can't think of a difference in allowing a fighter to do the above or taking the Disengage Action and then taking the Attack Action, but i can quite clearly see a difference between doing the above and having to attack a creature AGAIN, since that creature didn't die with the first attack, before having the ability to disengage.

Your Monk example is appropriate, since it is a bonus action, after all, even if it is also an action. (usual "as" use, like the "are monk fists magical" ruling? Reading a little bit too much into it again and even "as" is not meant to represent a particular meaning in the rules?)

It could be fun to ask "Does the "trigger" prerequisite also extend to "proper actions"? Is a fighter, using action surge, allowed to take the attack action, attack once, action surge, take the cast a spell action/disengage/, move, attack again with Extra Attack?"

Obviously i'm not sure of what to think of it or if it is in any form correct, but...

[Tanarii]

I assume it was a knee-jerk response some he completely reversed his position in a way that ended up nerfing a feat that wasn't op, while still allowing that mechanic to function in other ways.

There is nothing knee jerk about it. The base ruling on bonus action timing, that made this change to Shield Master required, was made least 6 months ago, as can be seen by the similar trickle-down effect changing the War Magic ruling in the SA compendium in Oct 2017.

This is not a ruling specific to shield master. Its a general ruling on bonus action months ago requiring a change to a previously made ruling on shield master.

[RSP]

I can only repost a tweet from JC, again.

I'm not aware of any rule that prevents it, but apparently that is the point. There's no rule, so you can't do it. But, again, action surge might be intended to be an exception and the whole lawyering we are doing wasn't really expected.

The fighter is doing 2 different actions, and those apparently do not have the "nesting" proprieties of certain reactions and bonus actions. Mind you, i can't think of a difference in allowing a fighter to do the above or taking the Disengage Action and then taking the Attack Action, but i can quite clearly see a difference between doing the above and having to attack a creature AGAIN, since that creature didn't die with the first attack, before having the ability to disengage.

Your Monk example is appropriate, since it is a bonus action, after all, even if it is also an action. (usual "as" use, like the "are monk fists magical" ruling? Reading a little bit too much into it again and even "as" is not meant to represent a particular meaning in the rules?)

It could be fun to ask "Does the "trigger" prerequisite also extend to "proper actions"? Is a fighter, using action surge, allowed to take the attack action, attack once, action surge, take the cast a spell action/disengage/, move, attack again with Extra Attack?"

Obviously i'm not sure of what to think of it or if it is in any form correct, but...

If you look down in that tweet string, JC references this prior tweet of his:

"My tweet below was addressing bonus actions and reactions that have triggers. A bonus action that has no trigger—such as Cunning Action and the misty step spell—can take place whenever you want on your turn (PH, 189). #DnD"

Which to me infers "only BAs or Reactions that have a trigger need concern themselves with nesting rules." So any other BA can occur anytime. Reactions are all inherently based on triggers, therefore would always follow the nesting rule. Actions might be suggested that they follow a similar rule, but that seems against how the game is supposed to work.

Would be nice, as suggested in that tweet string, if JC would just make an errata/Sage Advice to define Actions and when they occur/end, if that's his intent.

I just find it very difficult to definitively have beginning and ends in terms of Actions when the game has overlapping turns in each round: character A is still Dodging during character B's turn as they occupy the same in-game time period. Likewise, character B is still Attacking during character C's turn, though the action isn't being resolved at that point in real life time, it is still occurring in game.

If the default rule is you have to complete an action prior to doing anything else (with the specific rule of moving in between attacks being a specific vs general exception), then you can't Dodge and then move; you can't Disengage and then move; you can't Ready and then move.

Perhaps it's just me and how I look at the game though. I guess if you play it as every other character is patiently waiting doing nothing when it's another character's turn, it works, but that's not what the rules state about turns:

"A typical combat encounter is a clash between two sides, a flurry of weapon swings, feints, parries, footwork, and spellcasting. The game organizes the chaos of combat into a cycle of rounds and turns. A round represents about 6 seconds in the game world. During a round, each participant in a battle takes a turn."

This certainly seems to indicate, to me at least, that everything happens together in-game (and suggests attacks aren't just the one mechanical attack).

[Chaosmancer]

I'm not aware of any rule that prevents it, but apparently that is the point. There's no rule, so you can't do it.

I find this an incredibly worrying statement. Wasn't part of 5e's design philosophy to move away from this idea that things are impossible unless there is a specific rule to allow them?

I think we need to be very careful about allowing that logic back into the game, it seemed to do more harm than good in the past.

[PhoenixPhyre]

Is there a difference between actions that you take once (do the in-universe thing once) but have lasting effect and those that are an atomic set of actions?

As I see it,

Dodge, Disengage, and Dash are modal abilities. You do activate <mode> and it lasts until <condition>. Once you've taken the action, you're free to do other stuff without any other restrictions.

Attacking is doing a thing (making an attack roll) that then ends. Extra attack says that you can move between those attack rolls (as a specific exception). These "atomic" actions can't be broken up unless a more specific rule intervenes.

That's my head-canon, anyway.

[RSP]

Is there a difference between actions that you take once (do the in-universe thing once) but have lasting effect and those that are an atomic set of actions?

As I see it,

Dodge, Disengage, and Dash are modal abilities. You do activate <mode> and it lasts until <condition>. Once you've taken the action, you're free to do other stuff without any other restrictions.

Attacking is doing a thing (making an attack roll) that then ends. Extra attack says that you can move between those attack rolls (as a specific exception). These "atomic" actions can't be broken up unless a more specific rule intervenes.

That's my head-canon, anyway.

What differentiates Dodge and Attack in the RAW in this way? As I read JC's tweets, it looks like Attacking isn't a thing (unless there is no Extra Attack available) as the Attack Action isn't completed until all attacks are done (not sure how this translates when you don't take an available Extra Attack). So Attacking isn't the Attack Action, but rather completing all Attacks is the Attack Action. Likewise, Dodge doesn't complete as an Action after the first attack against you but continues through until your next turn. Dodge only ends when your next turn begins or combat ends, whichever comes first.

If combat is "a flurry of weapon swings, feints, parries, footwork, and spellcasting," that description doesn't really translate to "your ~6 second Action is 1 weapon attack."

[Kaliayev]

The issue is that if he thinks it's cheesy to shove first with SM then he must also have a problem with this.

I assume it was a knee-jerk response some he completely reversed his position in a way that ended up nerfing a feat that wasn't op, while still allowing that mechanic to function in other ways.

It's not just shield shove he finds "cheesy." He also dislikes how the shove attack was presented in the PHB. For him, contested ability checks don't belong in combat. Instead of a player rolling athletics vs a creature's athletics/acrobatics, he thinks there should be a saving throw. You can see this being implemented in the latest UA, where the minotaur's reactionary shove requires a saving throw.

I find this an incredibly worrying statement. Wasn't part of 5e's design philosophy to move away from this idea that things are impossible unless there is a specific rule to allow them?

I think we need to be very careful about allowing that logic back into the game, it seemed to do more harm than good in the past.

Yes, that's why some of us are more concerned with the team's approach toward the community's broad interpretation than the semantics of the bonus actions rule.

[ProsecutorGodot]

What differentiates Dodge and Attack in the RAW in this way? As I read JC's tweets, it looks like Attacking isn't a thing (unless there is no Extra Attack available) as the Attack Action isn't completed until all attacks are done (not sure how this translates when you don't take an available Extra Attack). So Attacking isn't the Attack Action, but rather completing all Attacks is the Attack Action. Likewise, Dodge doesn't complete as an Action after the first attack against you but continues through until your next turn. Dodge only ends when your next turn begins or combat ends, whichever comes first.

If combat is "a flurry of weapon swings, feints, parries, footwork, and spellcasting," that description doesn't really translate to "your ~6 second Action is 1 weapon attack."

Dash resolves immediately, all dashing does is double your allowed movement in a turn. As soon as you take the dash action, it resolves and you are allowed to double your movement. Dodge likewise resolves immediately, as soon as you declare the dodge action it takes effect and all attacks against you until the start of your next turn are made at disadvantage and you gain advantage on Dex saves. Actions don't stretch, you take the action and it's done.

This is where it's important to keep in mind that not all classes get an Extra Attack. Extra Attack is an optional feature that allows you take more attacks as part of your attack action. The Attack Action is only resolved after you make all of your attacks(reference Making an Attack for why it has to be resolved instead of just declared). You don't have to use all of your attacks granted by Extra Attack, but if you resolve one attack and then use a bonus action that was conditional on you taking the attack action, you are deciding not to use any possible extra attacks you had.

I made a mistake saying "you can't nest actions" as there are cases where you are allowed to put non- conditional bonus actions (Misty Step, Cunning Action) between attacks. You simply cannot put any action conditional on taking the attack action inside your attack action.

[PhoenixPhyre]

I made a mistake saying "you can't nest actions" as there are cases where you are allowed to put non- conditional bonus actions (Misty Step, Cunning Action) between attacks. You simply cannot put any action conditional on taking the attack action inside your attack action.

That jibes with my reading as well.

It makes me think that it's the Attack action that's special here. The Attack action specifies that when you take that action, you make an attack (or more in case of Extra Attack). Thus, to have "taken the Attack action" === to have "made the attack(s)". You can't do anything that's conditional upon "taking the Attack action" until that action is complete because the condition isn't met yet, but you can do anything that isn't conditional on it because they're independent things.

As to the main topic, I doubt any table I'm at will care one way or another. I certainly let monks decide to flurry for a 4th attack after they took the normal BA unarmed attack (so the total stays the same but the declaration comes later). I've never seen Shield Master in play, so :shrug:

[RSP]

Dash resolves immediately, all dashing does is double your allowed movement in a turn. As soon as you take the dash action, it resolves and you are allowed to double your movement. Dodge likewise resolves immediately, as soon as you declare the dodge action it takes effect and all attacks against you until the start of your next turn are made at disadvantage and you gain advantage on Dex saves. Actions don't stretch, you take the action and it's done.

This is where it's important to keep in mind that not all classes get an Extra Attack. Extra Attack is an optional feature that allows you take more attacks as part of your attack action. The Attack Action is only resolved after you make all of your attacks(reference Making an Attack for why it has to be resolved instead of just declared). You don't have to use all of your attacks granted by Extra Attack, but if you resolve one attack and then use a bonus action that was conditional on you taking the attack action, you are deciding not to use any possible extra attacks you had.

I made a mistake saying "you can't nest actions" as there are cases where you are allowed to put non- conditional bonus actions (Misty Step, Cunning Action) between attacks. You simply cannot put any action conditional on taking the attack action inside your attack action.

I obviously disagree with how you view the descriptions of Dash and Dodge (I'd say the Actions continue while mechanical benefits are in place), but let's look at them your way:

Dash lets you potentially double your movement. Nothing says you have to take all that movement. If you are correct, and Dash is resolved as soon as you declare it, rather than when you actually use the movement, how is that different than declaring you're taking the attack action and then waiting on the actual attacks? In other words, what let's you resolve Dash with a simple declaration, while holding off on the actual part of the action that has a mechanical effect (the movement); yet the Attack Action operates on a completely different model: every possible instance of its potential must be fulfilled before another action can be made?

Would a fighter be able to Dash in between attacks, using Action Surge, as you see it?

Again, it may be you're right and this is JC's intent, but I can't see anything RAW that states this.

[PhoenixPhyre]

I obviously disagree with how you view the descriptions of Dash and Dodge (I'd say the Actions continue while mechanical benefits are in place), but let's look at them your way:

Dash lets you potentially double your movement. Nothing says you have to take all that movement. If you are correct, and Dash is resolved as soon as you declare it, rather than when you actually use the movement, how is that different than declaring you're taking the attack action and then waiting on the actual attacks? In other words, what let's you resolve Dash with a simple declaration, while holding off on the actual part of the action that has a mechanical effect (the movement); yet the Attack Action operates on a completely different model: every possible instance of its potential must be fulfilled before another action can be made?

Would a fighter be able to Dash in between attacks, using Action Surge, as you see it?

Again, it may be you're right and this is JC's intent, but I can't see anything RAW that states this.

Actions that are not contingent on each other are trivially rearrange-able. So attack - AS Dash - move - attack is identical to AS Dash - attack - move - attack, which I think we all agree is legal (because movement has a particular exception from the "must complete the attack action" part). SM BA prone - attack - attack or attack - SM BA prone - attack aren't legal because the Attack action isn't complete until all the attacks are made (and thus the contingent bonus action can't be used yet).

[ThePolarBear]

If you look down in that tweet string, JC references this prior tweet of his:

"My tweet below was addressing bonus actions and reactions that have triggers. A bonus action that has no trigger—such as Cunning Action and the misty step spell—can take place whenever you want on your turn (PH, 189). #DnD"

Which to me infers "only BAs or Reactions that have a trigger need concern themselves with nesting rules." So any other BA can occur anytime. Reactions are all inherently based on triggers, therefore would always follow the nesting rule. Actions might be suggested that they follow a similar rule, but that seems against how the game is supposed to work.

Tthe "tweet below" is not the one i linked, it's the one that you can see being quoted if you look at the actual tweet. Which is different and might give you a different perspective.

Would be nice, as suggested in that tweet string, if JC would just make an errata/Sage Advice to define Actions and when they occur/end, if that's his intent.

Yep.

I just find it very difficult to definitively have beginning and ends in terms of Actions when the game has overlapping turns in each round: character A is still Dodging during character B's turn as they occupy the same in-game time period. Likewise, character B is still Attacking during character C's turn, though the action isn't being resolved at that point in real life time, it is still occurring in game.

The character is still doing something, and i agree. But it's not "taking an Action", which should be a mechanical abstraction. Before anything mechanical might happen again, if there's something that triggers something else, the trigger has generally to happen and complete first. Apparently, "taking an Action" means exausting it and reap as many as the benefits one wants, before declaring it concluded.

If the default rule is you have to complete an action prior to doing anything else (with the specific rule of moving in between attacks being a specific vs general exception), then you can't Dodge and then move; you can't Disengage and then move; you can't Ready and then move.

Yes that you can. A character can dodge and move. A player cannot however declare the Dodge action, move, then receive the benefits of having taken the Dodge Action. I mean, its character will start benefitting from the effects of the Dodge Action as soon as he declares it, and the Action comes to pass. Diggy the Dodger now is focusing on defence. All attacks bla bla disadvantage. Now, Diggy the Dodger can move.

Perhaps it's just me and how I look at the game though. I guess if you play it as every other character is patiently waiting doing nothing when it's another character's turn, it works, but that's not what the rules state about turns:

"A typical combat encounter is a clash between two sides, a flurry of weapon swings, feints, parries, footwork, and spellcasting. The game organizes the chaos of combat into a cycle of rounds and turns. A round represents about 6 seconds in the game world. During a round, each participant in a battle takes a turn."

This certainly seems to indicate, to me at least, that everything happens together in-game (and suggests attacks aren't just the one mechanical attack).

Narratively, yes. But you do not turn back the clock when the wizard that is just after the rogue chooses to disintegrate the rogue's target, which he could just barely reach by double Dashing, and let the rogue do something else, since the rogue just happen to reach a pile of dust? Or do not tell the fighter the effects of said disintegrate until the "end of round", when everything is resolved at once?

Turns are an abstraction to handle quick paced, real time situations in a way that's orderly and simple. It's obviously not going to be completely realistic until you atomize time to a level where combat becomes a sequence of infinitesimal movements. What characters do narratively is necessarily different from what mechanically happens, even if the two things are intrisecally linked.

I find this an incredibly worrying statement. Wasn't part of 5e's design philosophy to move away from this idea that things are impossible unless there is a specific rule to allow them?

I think we need to be very careful about allowing that logic back into the game, it seemed to do more harm than good in the past.

I think it's more of a necessary evil. Just think about all the problems that can happen with the adjudication of the triggers for the Ready action. Setting a limit on how much one can slice a turn, an action, everything... is simply something that it's already in game. Nothing that is not a specific exception can happen between the casting of a spell and its effect being manifest, even when there are 120' of distance between caster and point of origin of fireball, when there's a streak of fire flying into the air. Saving Throws are there to mimic reactions, but there's no codified "launch rock to intercept streak of fire and make detonation happen in mid air" concept that isn't fully on the DM shoulders to adjudicate.

[ProsecutorGodot]

I obviously disagree with how you view the descriptions of Dash and Dodge (I'd say the Actions continue while mechanical benefits are in place), but let's look at them your way:

Dash lets you potentially double your movement. Nothing says you have to take all that movement. If you are correct, and Dash is resolved as soon as you declare it, rather than when you actually use the movement, how is that different than declaring you're taking the attack action and then waiting on the actual attacks? In other words, what let's you resolve Dash with a simple declaration, while holding off on the actual part of the action that has a mechanical effect (the movement); yet the Attack Action operates on a completely different model: every possible instance of its potential must be fulfilled before another action can be made?

Would a fighter be able to Dash in between attacks, using Action Surge, as you see it?

Again, it may be you're right and this is JC's intent, but I can't see anything RAW that states this.

When you say "I dash" it's done, your movement is double. Moving is not part of Dashing, doubling your movement is all it does, the action is done. When you take the attack action it says "With this action, you make one melee or ranged attack. See the "Making an Attack" section for the rules that govern attacks. Certain features, such as the Extra Attack feature of the fighter, allow you to make more than one attack with this action."

You'll note that in "Making an Attack" it says, in 3 steps "1: Choose a Target, 2: Determine Modifiers, 3: Resolve the Attack"

Extra attack allows you more than one attack as part of the same action, attacks made as part of that action must be resolved to complete the attack. You can't say "I take the attack action" and then not attack, the action involves making at least one attack and you must resolve the attack. Extra Attacks must be part of the same action, which means they also have to be resolved before the action is considered taken. You can choose not to take extra attacks but that doesn't mean your action carries over until you do, it means as part of that action you only used one attack. Extra Attacks are optional.

On the note of Action Surge to dash, absolutely. I thought I was clear that when I said you're not allowed to nest actions that I only meant actions conditional on completing the attack action.

Frankly the argument that Dash remains in effect until your next turn is asinine. You use the action, it doubles your movement, movement is not an action, it's a resource. You don't have to use all your movement on your turn, not using all your movement on your turn doesn't mean your turn lasts forever, it means you didn't use all of it. There is nothing to resolve when you choose to dash, it doubles your movement, that's all it does.

This is exactly the same as the Attack Action, all it does is make an attack. Extra Attack gives you the option but does not require you to make more than one attack as part of the same action. When you make an attack, the attack must be resolved. It's written plain as day in the rules.

[Arial Black]

If 'taking the Attack action' is EXACTLY THE SAME THING as 'executing the attack', and you MUST complete the whole Attack action before you can take the bonus action that it generates, then if you don't actually attack then you have not completed your Attack action, because they are the same thing!

This means that the 3rd level battlemaster can never use Commander's strike!

"Commander's Strike: When you take the Attack action on your turn, you can forego one of your attacks and use a bonus action to direct one of your companions to strike...."

So, you have one attack. You want to use Commander's Strike to get your rogue friend to attack with Sneak Attack. So you 'take the Attack action'....but as soon as you do you are ALSO actually attacking, according to JC, because 'taking the Attack action' is the same thing as actually executing the attack! So, tough noogies.

Alright, take the Attack action but 'forego your attack' and then take the bonus action...oh dear; no can do. You must complete the Attack action before you can take the bonus action it generates, and you can only complete it by executing the attack (because 'taking the action' is the same thing as 'executing the attack'). Since you haven't made the attack then, by definition, you have not 'taken the Attack action' and have therefore not generated that bonus action.

What's that you say? "Ah, but they are actually different things"? In that case, as soon as I 'take the action' but BEFORE I 'execute the attacks' I have generated that bonus shield bash and can use it when I like.

If 'taking the Attack action' IS synonymous with 'executing the weapon attack allowed by the Attack action', then a creature with one attack declares/actually attacks simultaneously. That makes sense.

But, if this IS the rule, then it is unavoidable that if the Attack action allows you to execute two attacks (because Extra Attack) then 'declaring the Attack action' and 'executing BOTH attacks' are one and the same thing!

It means that IF the interpretation is that declaring/attacking are the same thing, then BOTH attacks MUST be resolved instantly!

The ONLY way that you can attack/move/draw another weapon/attack someone else 30 feet away later in the round is under the interpretation that 'declaring the Attack action' is NOT one and the same thing as actually 'executing the attacks allowed by the Attack action'!

Since Extra Attacks do not have to be taken all at the same time, and the fact that BOTH attacks must 'happen at the same time' as taking the Attack action, his quote that they are the same thing is nonsense.

Also, his sudden insistence that any action, including the Attack action, is 'indivisible', is given the lie by....well...the many, many things in the game which literally can and do occur during other Actions, including (but not limited to): Readied actions, bonus actions that are not triggered by things that happen in a specific order, free object interactions, as well as movement.

Even in this new 'indivisible' interpretation, the Ftr 5/Wiz 3 can move, 'declare' his Attack action (but only ONE of his two attacks 'happen at the same time'!), shoot an arrow, drop his bow, move, draw his sword, cast misty step as a bonus action, move, and then complete his so-called 'indivisible' Attack action by executing his second attack.

Yeah, his cries of "Actions are indivisible!" as the reason that you have to complete your action before you can shield bash are not credible.

[JoeJ]

If 'taking the Attack action' is EXACTLY THE SAME THING as 'executing the attack', and you MUST complete the whole Attack action before you can take the bonus action that it generates, then if you don't actually attack then you have not completed your Attack action, because they are the same thing!

This means that the 3rd level battlemaster can never use Commander's strike!

"Commander's Strike: When you take the Attack action on your turn, you can forego one of your attacks and use a bonus action to direct one of your companions to strike...."

This is a case of specific beats general. Not being able to take a bonus action shield bash between attacks seems kind of dumb to me, but this example doesn't help the argument.

[ProsecutorGodot]

SNIP

You take the Attack Action, choose to give up your attack, and use your bonus action to use Commanding Strike. This is a different circumstance to people trying to insert a bonus action into their attack action and still make use of all their attacks.

The attack in this case is being used as a resource given up to enable your bonus action Commanding Strike. It's more comparable to making a standard Shove than comparing it to a Shield Master shove.

It has nothing to do with the attack action being indivisible, it has to do with actions conditional on taking the attack action happening after.

Deliberately running away in a direction that doesn't make any sense doesn't help the case either. Movement, as repeatedly mentioned, is given express permission to go between your Extra Attacks. It's in the rules. Additionally, bonus actions without the condition of happening when you take the attack action can happen between your extra attacks because unless otherwise specified you choose the timing. Also in the rules. The only "new" clarification that's important here is how conditional bonus actions are handled. It's not even a new clarification as the Sage Advice Compendium explains this with War Magic.

Not a single thing you had that Fighter do in your example actually broke sequence in their attack action. Using the shield master shove before* or between any extra attacks would break sequence.

You can dislike the ruling, I even agree with the general consensus that it's dumb, but it's consistent and makes sense within the rules which is what I think people wanted more than anything.

*Sage Advice Compendium mentions that nothing would technically be wrong with doing it first, which is why this is primarily an AL issue.

[ThePolarBear]

[...]

While this is smart and funny, and worth sending to JC as a question, there is a problem: renouncing one of your attacks is not a bonus action - or any action, really.

So, while you are correct that as soon as you take the action you have generated your right to bash, you still cannot spend the bonus action to bash just as much as you can't spend the bonus action to command a strike from one companion until all that is involved with the attack action has passed. Commanding Strike, however, allows for single reduction on number of attacks, and if 0 attacks remain you have completed your action, and can therefore use your bonus action immediately. So a non-extra attack fighter can reduce its own attacks to 0, ending the attack action, prompting the ability to use the triggered bonus action. The action has not been segmented - no other action/bonus action/reaction has happened inside the attack action - so there is no problem taking into consideration the tweets from JC as everything is still consistent with what has been written.

Actions are not indivisible full stop. Actions are not allowed to be nested with other actions or reactions that do not have specific exceptions. In fact, actions are by themselves already divided: for example the resolution of an attack action has many steps, each with their own resolution process and flow.

"Declaring" is also not the verb used anywhere. The verb is "take", which is quite different. I can declare i will pick up a bottle of wine, but taking it has an extra step to it, even just considering semantics.

Now, i've not seen the tweet you are directly referring to, or if i did i forgot about it, so i can't really go over something i have no memory of in detail. It would help to have it linked as a reminder.

[Xetheral]

When you say "I dash" it's done, your movement is double. Moving is not part of Dashing, doubling your movement is all it does, the action is done.

(Emphasis Added.) Wait, what? How would a character double their speed without moving? If the dash action doesn't involve moving at your table, what is the character doing when they take the dash action?

[ProsecutorGodot]

(Emphasis Added.) Wait, what? How would a character double their speed without moving? If the dash action doesn't involve moving at your table, what is the character doing when they take the dash action?

From a game mechanics standpoint taking the dash action doubles your speed on that turn. No part of taking the dash action actually involves using that movement. You can double your speed and not move your full speed, or at all, that doesn't mean you didn't take the dash action.

You would almost always use your movement on a turn you dash, but movement is distinctly separated from the action. Anything requiring you to have made the dash action (Charger) only requires that you take the action and double your speed for that turn, not that you move. The second part of Charger however does specify that you need to move AND take the dash action for the additional benefit.

[ThePolarBear]

(Emphasis Added.) Wait, what? How would a character double their speed without moving? If the dash action doesn't involve moving at your table, what is the character doing when they take the dash action?

Well, moving is part of dashing, but the Dash Action doesn't imply movement. Unless some movement is spent, there's really no going anywhere in any particular way. Dash is called that way because it is presumed that when someone takes it it's because that movement is needed. But by itself the action isn't really anything more than an increase in allowance of movement. It's like the Dodge Action when there's no one in sight. The character doesn't suddendly start breakdancing and evading imaginary arrows :D

I find easier to describe someone "Dodge"-ing while there's no real threat ("The character seems on its toes and twitchy, reflexively reacting and observing any potentially hostile patron of the tavern") then someone "Dash"-ing ("the creature seems to think of running, but than changes idea"?). Not that it would make sense to have actions tracked (or taken) while outside of a quick paced situation like combat anyway.