Sage Advice Compendium Update 1/30/2019

[FrogReaver]

It absolutely is, there is nothing special about the Disengage action in my interpretation. As you assemble your turn, the Disengage action is just one more discrete element in the ordered list. It gets processed in that order. So, movement elements before the Disengage action provoke OAs, movement elements after it do not. No element in the list needs to have a duration applied for the effect of Disengage to happen.

1) Move
2) Disengage
3) Move

(1) provokes OAs, (3) does not provoke OAs as expected. The duration of any entry on this list has no bearing on that outcome, and my contention is that duration is the wrong way to be thinking about all of this.

Very good that's my opinion of the disengage action as well. It's a single sequential discrete event. The dodge action is as well. The ready action is as well. It seems to me that your saying that all actions except the attack action are sequential discrete events. Do you have a good reason to believe the attack action is different than all the other actions in that regard?

 

[FrogReaver]

Well, as I’ve said, I think the answer is really simple. The PHB says the Attack action means making an attack. There is no mention of a declaration phase. There are exceptions to the general rules that allow for splitting your Attack action with movement, implying that your Attack action now has two or more discrete elements, assuming you have multiple attacks from Extra Attack.

Why does it need to be any more complicated than that? Like all other actions, the Attack action starts as one discrete element, and can be split by specific things as documented in the rules (e.g. movement, bonus actions that are triggered by a single attack, bonus actions with no trigger, etc). If the Attack action said something about being able to make one or more weapon attacks between now and the end of your turn, then I would agree that the duration or effect of the Attack action matters. It doesn’t say that, though, which points at the Attack action being all your individual attacks (unless you split the action via the specific rules allowing that).

The duration of the attack action matters because we need to know how long it lasts in order to determine when we have met the condition to bonus action shove. Hasn't that always been the obvious reason to care about duration in this conversation?

What your saying isn't simple. The same arguments about duration can be made in your system, its just duration is now duration is concerned with a single sequential element vs multiple sequential elements. Instead of talking of duration as number of seconds we can talk of it as number of discrete elements. So there's no escaping the concept of duration. It's just as present in your formulation as in the other, whether or not you want to talk about it.

 

[Maxperson]

So 2nd most important thing.....

The logical non-mathematical argument is called reducto ad adsurdum. It's based on the law of excluded middle (that something must either be true or it must be false). The mathematical concept of proof by contradiction is the same thing. It's like you talk about things you don't understand and it's obvious you don't have the faintest clue.

Keep ignoring it if you like, but it's not going to make your proof valid. Another possible true or false proposition exists, that being that actions are divisible. That invalidates your proof. You don't get to just say you are ignoring it, so it doesn't affect you and your proof. That's not how it works.

 

[Asgorath]

So then to use the buff all you as a player do is simply "declare you re taking the disengage action and you've taken it". So then the discrete sequential event that is the disengage action is simply a player declaration is it not?

There is no declaration needed, you just take the action. It's a discrete operation, ordered with other discrete operations in your turn. If the action applies a lasting effect, then the effect applies for the duration. So, if your turn is "move, Disengage, move" then the movement before the action provokes OAs while the movement after the action does not. There are many other effects in the game, most of them from spells, that work in a similar manner. We're not arguing about how Bless works, or that you have to declare that you're going to cast the spell, you just use the Cast a Spell action and then the effect applies for the duration, right?

Very good that's my opinion of the disengage action as well. It's a single sequential discrete event. The dodge action is as well. The ready action is as well. It seems to me that your saying that all actions except the attack action are sequential discrete events. Do you have a good reason to believe the attack action is different than all the other actions in that regard?

Yes, because there are explicit rules that say you can break your Attack action into smaller pieces (of one or more attacks each) and insert movement between those pieces. The individual pieces still make up the Attack action as a whole. There are also bonus actions like TWF that trigger off an individual piece of the Attack action. That doesn't turn the Attack action into a declaration and effect with a lasting duration, it just means your Attack can be multiple discrete elements, with rules about what can go between those elements. Again, there are several examples of actions that effectively say "if you take this action, then X happens until the end of your turn" (e.g. Disengage, Dodge). The Attack action does not say that, so it must not work like that. If the rules wanted the Attack action to work in this manner, then surely it would use the same kind of language that the Disengage and Dodge action use, right?

 

[Maxperson]

If the player tells you he is doing X that is a declaration that you are doing X.

Okay, but so what. It has no game validity whatsoever. A player declaring to you that he is making an attack when his turn comes up has the same game validity as that same player declaring that he is pulling a Black Hawk Helicopter out of his pocket, flying into the air inside it, and shooting missiles at the bad guy. Nothing can trigger off of either declaration, and neither declaration can be forced on the player without a house rule, because there is no game validity to either one.

You assume that a declaration of action can only exist in a phase that occurs before the action is taken. That simply isn't the case.

I'm not assuming anything. It's a fact that declarations are completely informal and have no game validity. Declare any action you like. Declare it any time you like. Declare it months in advance, or months after the fact. It's all the same to the game.

 

[Asgorath]

The duration of the attack action matters because we need to know how long it lasts in order to determine when we have met the condition to bonus action shove. Hasn't that always been the obvious reason to care about duration in this conversation?

What your saying isn't simple. The same arguments about duration can be made in your system, its just duration is now duration is concerned with a single sequential element vs multiple sequential elements. Instead of talking of duration as number of seconds we can talk of it as number of discrete elements. So there's no escaping the concept of duration. It's just as present in your formulation as in the other, whether or not you want to talk about it.

The Attack action starts as one discrete element. The rules say you can break this action into separate pieces and insert other (edit: specific types of) discrete elements in between. The Attack action is still just those N pieces, and is complete once those N pieces have been processed or resolved.

Example: Move, Attack, Move. If you have 3 attacks from Extra Attack, then your turn might be moving your miniature, making 3 attack rolls against your target, and then moving again after you kill it. The Attack action is clearly over once you've made those 3 attacks.

Example: Move, Attack, Move, Attack, Move. If you have 3 attacks from Extra Attack, then your turn might be moving your miniature, making 2 attack rolls against a target, moving to a new target since you killed that one, making 1 attack roll against a new target, and then moving some more after you kill that one as well. The Attack action is clearly over once you've made those 3 attacks. We don't care that you moved in between, your Attack action is still those 3 attacks in 2 discrete elements on your turn.

The "If you X, you can Y" triggering translates to this concept in a very straight forward manner. Before you add a Y element to the list, X must be completed. If X is the Attack action and you split your Attack action into 2 discrete elements so you could move between attacks, then Y must come after those 2 Attack action elements in the list. Example: Move, Attack, Move, Attack, Move, Shield Master Shove. In a nutshell, if something triggers from the Attack action, then you just don't get to do the Attack action (or pieces thereof) after you've done the triggered bonus action.

Or, if we want to really simplify this: No part of the triggering condition can come after the triggered event. If the trigger is the Attack action, then it makes no sense that you can perform parts of the Attack action after the triggered bonus action.

 

[FrogReaver]

Keep ignoring it if you like, but it's not going to make your proof valid. Another possible true or false proposition exists, that being that actions are divisible. That invalidates your proof. You don't get to just say you are ignoring it, so it doesn't affect you and your proof. That's not how it works.

I have a rule that shows actions are not divisible by movement. Therefore your proposition doesn't disprove my proof.

 

[Maxperson]

And it's also why I'm advocating that we simply stop thinking about this in terms of "the duration of an action". There is no rule that allows you to move while taking the Cast a Spell action, implying you have to stand there and perform the V/S/M components as needed. You can move before this action, and you can move after this action. The only thing that matters here is that the spell's casting time is 1 action.

Actually, I think it's important. This discussion has moved yet another opinion!! I now believe that actions are divisible. They are just not all divisible by all things. They are divisible, because of the duration is longer than instant, which makes the duration important. How long the actions last helps me figure out what might or might not be able to be used inside of it.

By default, you start your turn with your movement and your action. You might also have a bonus action, or you might do something on your turn that triggers a bonus action. You might move, Cast a Spell, then move some more. The only thing that matters here is that you no longer have an action this turn, why do we care how many seconds the "Cast a Spell" action took? That action is simply resolved in the correct order, relative to the rest of the things you do on your turn.

The length of an action is not important to you at this point in the discussion. It is, however, greater than instant. The Cast a Spell action and Bonus Action spells prove that, as does the Attack action.

Your turn might be move, cast Healing Word, move, cast Sacred Flame, move. Why does it matter how long Healing Word or Sacred Flame took to cast? The rules say that if you cast a spell as a bonus action, you can only use your action to cast a cantrip:

"A spell cast with a bonus action is especially swift. You must use a bonus action on your turn to cast the spell, provided that you haven't already taken a bonus action this turn. You can't cast another spell during the same turn, except for a cantrip with a casting time of 1 action."

That cantrip has a casting time just as long as the Wish spell. I think that the reason you can't cast anything other than a cantrip in a round that you use a bonus action spell is game balance. It's certainly not a function of time since as far as time is concerned, Action(cantrip) + Bonus Action(spell) = Action(Wish) + Bonus Action(spell).

 

[FrogReaver]

The Attack action starts as one discrete element. The rules say you can break this action into separate pieces and insert other (edit: specific types of) discrete elements in between. The Attack action is still just those N pieces, and is complete once those N pieces have been processed or resolved.

Example: Move, Attack, Move. If you have 3 attacks from Extra Attack, then your turn might be moving your miniature, making 3 attack rolls against your target, and then moving again after you kill it. The Attack action is clearly over once you've made those 3 attacks.

It's clearly over by the time you've made 3 attacks, it's not clearly over because you made 3 attacks. There is an important distinction there.

Example: Move, Attack, Move, Attack, Move. If you have 3 attacks from Extra Attack, then your turn might be moving your miniature, making 2 attack rolls against a target, moving to a new target since you killed that one, making 1 attack roll against a new target, and then moving some more after you kill that one as well. The Attack action is clearly over once you've made those 3 attacks. We don't care that you moved in between, your Attack action is still those 3 attacks in 2 discrete elements on your turn.

As previously stated, I agree with you that we know the attack action has at least ended by the time you've made those 3 attacks. That doesn't mean it didn't end before though. It just means we have 100% certainity no matter our disagreement that it's at least ended by then.

The "If you X, you can Y" triggering translates to this concept in a very straight forward manner. Before you add a Y element to the list, X must be completed. If X is the Attack action and you split your Attack action into 2 discrete elements so you could move between attacks, then Y must come after those 2 Attack action elements in the list. Example: Move, Attack, Move, Attack, Move, Shield Master Shove. In a nutshell, if something triggers from the Attack action, then you just don't get to do the Attack action (or pieces thereof) after you've done the triggered bonus action.

There are 2 possibilities for the attack action
Possibility 1: the action consists of N discrete sequential elements
Possibility 2: the action itself is a discrete sequential element and provides the ability to perform an additional N discrete elements.

My evidence for possibility 2 is that other actions are a single discrete sequential element that provide an effect that last through N discrete sequential events. Why should the attack action be any different?

 

[FrogReaver]

There is no declaration needed, you just take the action. It's a discrete operation, ordered with other discrete operations in your turn. If the action applies a lasting effect, then the effect applies for the duration. So, if your turn is "move, Disengage, move" then the movement before the action provokes OAs while the movement after the action does not. There are many other effects in the game, most of them from spells, that work in a similar manner. We're not arguing about how Bless works, or that you have to declare that you're going to cast the spell, you just use the Cast a Spell action and then the effect applies for the duration, right?

How does anyone know you are taking the disengage action if you don't declare you are taking it?