But anyway, last action to me covers the activity and not its subordinate actions.
An activity typically involves using multiple actions to create an effect greater than you can produce with a single action, or combining multiple single actions to produce an effect that's different from merely the sum of those actions. In some cases, usually when spellcasting, an activity can consist of only 1 action, 1 reaction, or even 1 free action.
The text makes it clear that activity-contained subordinate actions are still the base actions in their own right. For every consideration except the one text override, subordinate actions are their "raw" actions.
The text chooses to keep "activity" separate and isolated from ever being an "action" by default.
This means that the text requires and provides specific and narrow instruction to consider the initialization of an activity as an action for the sake of preventing activity chains from being used in place of their first subordinate action. That's not allowed.
Where I'm going with that is just to say that "activity" is a very narrow and specific term. There is no context in which you read "action" and are supposed to translate that into "action & activity."
If the text of Drink, etc, intended for activity action chains to be incompatible, it would have done that in the text, with something to the effect of
"Your previous action was a Strike, and it was not a subordinate action."
Because yes, subordinate actions are still their respective actions. They are only modified specifically as their activity instructs.
An activity can choose to *not* invoke or chain a "Strike action" if it wants to avoid being compatible with effects that trigger from Strike specifically.
When the devs think "hey, it would be a bad idea for this damage action to be compatible with Strike effects" they do that, and simply never invoke Strike as a subordinate action.
The most player-known example is Elemental Blast.
The new swarm eidolon does the inverse, and makes for a better example here.
When dispersed, your eidolon scatters its component bodies across a wider area, increasing its size by one category (to Large for most swarm eidolons). It has a reach of 0 feet, can’t make Strikes, and can occupy the same space as other creatures. While dispersed, it can use the Swarming Assault activity.
If you mis-apply the subordinate action text to mean that activities are wholly separate in concept, and not a chain of their subordinate actions, guess what?
That would mean that this text only blocks raw-Strike. All subordinate Strikes would be valid to a swarm eidolon and only that "base Strike action" would be blocked. Tandem Strike, SMN's version of reactive strike, and all other subordinate strikes would be valid.We already know that's not how the game works; the subordinate actions inside activities really are their "raw actions" done in sequence for all considerations except for the single override put into the text. The initialization of an activity is *not* the same as starting it's first subordinate action, effectively injecting a "ghost action" at the start of every activity before the subordinate actions.
______
Letting subordinate actions trigger things like Sneak Attack, etc, is the same rule-thread that results in "previous action was ___" not caring if that action was a subordinate action or a "base" one. If some ability cares, it needs to call out "no subordinate actions."
If you try to reinterpret the text so that activities "box" all their sub actions away by default, that affects a whole lot more than "previous action" considerations; no outside ability remains able to trigger from nor modify subordinate actions.
My last activity was Swipe, but my last action was a Strike. If both are true, then Drink of My Foes has had its requirements met and should work, without violating the "activities and their subordinate actions are not the same thing" rules.
"ghost action" is a way to describe the emergent effect of that rule upon the game in a more intuitive manner that does not trigger this "symmetry presumption"
It can be genuinely confusing to say that
"Starting an activity is not the same thing as starting a subordinate action that exists inside an activity for considerations of spell effects like Haste ..."
Instead, the same outcome can be translated as:
"When you start an activity, consider that activity to itself be the first action of that activity sequence, before any subordinate actions. This prevents spell effects like Haste from ..."
As far as I can tell, the game logic from that translation creates an identical result.
__________
One of the key confusions caused by the text's wording is that of assumed symmetry. The readers translate the odd wording (of example effects, not even the rule itself!) into something they understand, and add extra rules/restrictions that don't exist in the text.
If this has the effect of blocking "your next action" abilities, what about "your previous action?" abilities.
Without a visual metaphor like "activities start with a ghost action before the sub action chain," it's harder to understand that no, that proposed logic does not follow.
The text of subordinate actions is saying that "Squargs(activities) are not Blorps(actions). As an example, if you've got an ability with a "your next Blorp" requirement, you cannot use a Squarg.
Players are reading that, and then applying what is a rather limited logical statement in ways far outside the text, because they presume the reverse is also true.
___________
Translating the subordinate action text again:
A is not B.
This means that, when you get a free B via Haste, etc, you cannot use A. This also means that using A does not trigger effects with "if your next B is ____."
(activities are not actions. The rest of it are examples of what effect this rule has.)
_______________
Players are incorrectly stating another unrelated relationship, that Bs contained inside A no longer qualify as Bs. To be valid Bs, they must be alone, outside of As.
That easy misinterpretation is completely unrelated to the correct "A is not B" rule. It comes from people working backward to justify a "your prev action" block, which they assumed was there to be symmetrical with the real "your next action" block. Once they notice the text doesn't actually instruct that, they seek another way to create the same result.
But logical effects are not always symmetrical like that.
That reader invented rule contradicts what the text instructs about sub actions. The text outright says sub actions are still actions, except however the activity modifies them.
If some activity text says "this Strike doesn't count as a Strike for other abilities or effects," then it would not qualify for Drink, Twist the Knife, etc.
I'll repeat myself that imposing that restriction that "actions cannot match / qualify if it's a subordinate action" would break a whole lot more than people are presuming.
That Swarm eidolon example is the rare case of the reverse, where that invented rule would enable activities clearly intended to be unavailable.
It really would prevent even core features like Sneak Attack from proccing off subordinate Strikes.
Activities are not actions, and cannot be used in place of actions. This is a restriction placed upon activities, not actions.
This is incredibly easy to rule consistently.Your previous action is your previous action. By default, sub actions are actions.
There is no textual reason to re-write that default into the opposite, where "previous action" instead actually means "previous non-subordinate action". Note that this is impossible to change just for "previous" actions, and will affect all action triggers & requirements with the same "non-subordinate" requirement.
Being unable to use activities in place of actions does not change sub-actions into being non-actions hidden inside their activity.
_____
Please attempt to articulate why "activities != actions" translates into the last sub action in an activity chain being disqualified from meeting "previous action was ___" criteria.
I'm fairly certain, based on how they describe spellcasting being 'a single action, reaction or free action', that the 'action' used in activity is the literal action cost. You're kind of borrowing the wrong definition here.
If the text wanted to use that requirement, it would have been worded to do so. Something like "The previous action you spent was to make a Strike"
That would be a clever way to disqualify sub-actions, as you spend the action point cost on activities, and not on their sub actions. (though the text could/would also just use their own term of Subordinate Actions)_____
Activities are chains of 1 or more subordinate actions, can custom modify those subordinate actions, and often have a different action point cost than their sub actions.
Outside of special cases like spellcasting, which is always treated as an activity, (not sure what mechanical effect this creates, but they are clear on that odd rule), it is not possible for an activity to exist without a sub action.
An activity with no sub action would itself just be a custom action. Activities are defined by having actions performed inside of them.
An activity typically involves using multiple actions to create an effect greater than you can produce with a single action, or combining multiple single actions to produce an effect that's different from merely the sum of those actions. In some cases, usually when spellcasting, an activity can consist of only 1 action, 1 reaction, or even 1 free action.
An activity might cause you to use specific actions within it. You don't have to spend additional actions to perform them—they're already factored into the activity's required actions. (See Subordinate Actions.)
When an activity instructs you to perform a sub-action, you are still performing that action, and all abilities that watch for ____ action trigger still occur. This includes things like Sneak Attack.
A Strike being a sub action does not loose it's status as "an action" you have performed.
Basically me and another GM have different opinion when you look at the "damage dealt".
Requirements Your last action was a successful Strike with the barrow’s edge Effect Your blade glows as it absorbs your foe’s vitality. You regain Hit Points equal to half the damage dealt.He looks at what damage the target took, applying both resistance and weakness, block and also accounting fore hit-points left, so if the target only have 2hp you get only 1 healing.
While i would rule it as half of the total damage of the attack before resistance, weakness, and block.
Basically he checks hp lost at Step 4: Reduce Hit Points
And i argue its should be at Step 2: Damage Type: "Once you've calculated how much damage you deal, you'll need to determine the damage type."How do you/Would you rule it? or do you have any other insight?
It all depends on what your definition of “deal” means, but only if you don’t read the name of the ability.
Because originally I was with you, Step 2 - you deal damage, so you get healed.
But the ability is called *Drink* of My Foes. So you are sucking or subsuming or drinking the vitality of the damage you have inflicted. Thus, you can “deal” 90 bajillion damage, but if - after blocks and resistances and windsheer and definitions of what “is” is - the foe only *loses* 2 hps, then you are only “drinking” 1 hp.
No need for subordination. Just some simple drinking. Except a lot of folks seem to get stupider…when…they drink. Don’t drink and subordinate. It gets messy.
I would very much like to see some more examples in which abilities fail to function as intended under each of the two interpretations.
Then I'll likely settle for the interpretation that appears to cause the least headaches.
|
| 3 people marked this as a favorite. |
"Using an activity is not the same as using any of its subordinate actions."
So, if you use Sudden Charge, you're not using Strike.
After finishing Sudden Charge, your last action was a Sudden Charge, not a Strike.
Doubly so, since "You can use only one single action, activity, or free action that doesn't have a trigger at a time. You must complete one before beginning another. "
And a RAW definition for last :
"There are four types of actions: single actions, activities, reactions, and free actions."
The Strike you do as part of Sudden Charge does not count as an action. Though it is still a Strike.
| 1 person marked this as a favorite. |
I would very much like to see some more examples in which abilities fail to function as intended under each of the two interpretations.
Then I'll likely settle for the interpretation that appears to cause the least headaches.
Skirmish Strike, a 1A activity to Step & Strike, in either order.
______________
"Action agnostic" or "sub actions are actions you perform" version:
Activities are not Actions, so when things call for "actions," performing activities don't qualify. Means that Skirmish Strike cannot be used with Haste, because even if the first sub-act is Strike, the activity always has to go before its sub-acts, so your next action is not Strike, and therefore a hasted action is ineligible for Skirmish Strike.
Note that there is 0 restriction or conditional applied to sub-actions, which is as the text instructs. This is only a restriction upon activities.
After the activity starts, you are still doing those sub-actions as if they were normal raw actions. For these sub-acts to *not* qualify/trigger because of their status as sub-actions, the special text of the activity or trigger would need to specify that it cares about sub-or-not.
In this version, "Your previous action was __" can be the last sub-action of an activity, there is no exclusion. Skirmish Strike (with Strike as last) qualifies for Drink, etc.
This is also why "If you ___" action triggers fire when sub-actions happen. The conditional does not care if that's a raw Strike, or if the Strike is a sub-action inside something like Skirmish Strike.
Sneak Attack procs inside Skirmish Strike, because your sub-Strike counts as a "Strike."
______________
"You never perform sub-actions, only the parent activity is the 'action'":
Like the other version, this also blocks things like hasted actions from using activities, because the one (and now only) action you perform is the activity-action.
"Your previous action was ____" does not match, because the check ignores all sub-actions, to instead point at the activity that invoked the sub-acts.
Skirmish Strike != Strike, so Drink is invalid.
The issue is that the change in rule is not limited to "your prev ___" conditionals.
All abilities that call out a specific action to match no longer work off sub-actions.
A few abilities do still work by coincidence of wording. Precision Ranger triggers bonus dmg via: "The first time you hit your hunted prey..." which does not specify "Strike," do doing damage via "Skirmish Strike" matches the conditional.
But most are like Sneak Attack: "If you Strike a creature..."
If you as a player perform Skirmish Strike, deal dmg with the sub-strike, then the ask to do Sneak Attack damage:
The ability Sneak Attack checks if you have or have not performed a Strike. But sub-acts are not actions you perform in this version, the only 'action' you did was Skirmish Strike. Which is not Strike. So Sneak Attack cannot not proc.
"Sneak Attack still works!"
"It's not broken, You just have to use a raw, non-subordinate Strike to meet "If you Strike" conditionals."
"That's obviously how Rogue was balanced. It would be way too OP if you could get that bonus damage with sub-Strikes. Flurry of Blows could get 2 procs inside 1A, that's ridiculous. Obviously too-good-to-be-true."
_____________
Real quick, there's also the inverse example of swarm eidolon which is imo easier to see "yeah, that's BS"
... It has a reach of 0 feet, can’t make Strikes, ...
The normal "sub-or-not agnostic" version:
Any time an activity calls for a strike, the eidolon cannot perform the sub-strike while this restriction applies.The "only the activity is the action performed" version:
If you refuse to allow sub-actions to match against checks for the action, then you can bypass the "can't make strikes" restriction with activities. You can now have the swarm of rats make Strikes via Tandem Strike or other activities.
All checks for "is this action a strike?" get redirected to the activity, which is not "Strike." Even when those checks are restrictions to prevent "Strike".
Also would mean some nonsense around familiars, or any other place where something is blocked specifically from making named actions like Strikes. Just do an activity with the banned action as a sub-action instead.
"I'm not doing a Strike, I'm doing ___ activity!" becomes not a restriction, but a huuuge loophole ripe for abuse.
The Strike you do as part of Sudden Charge does not count as an action. Though it is still a Strike.
I do need to call out that this is an impossible foundation that's creating an issue.
Action is both a resource, but also a label of categorization and identity. Strike "costs 1 action" to do. And, Strike "is" an action. You cannot remove its status as an action without breaking it conceptually.
If you make a "Strike" you are performing the action called Strike. It's not possible to Strike, without that being an action you do.
Strike "is" one of the basic actions, trying to redefine that identity is like trying to claim that a wolf is sometimes still "a wolf" but simultaneously not being "a creature." It breaks the law of identity as a concept.
Considering that Strike is printed inside the basic "actions" in the book, you would need one crazy super specific of a rules override to say "in this context, actions loose their categorization as actions." In reality, the text says nothing of the sort.
____________________
Instead, it's full of things like:
Activities involve "using multiple actions to create an effect greater than you can produce with a single action,"
and
"An activity might cause you to use specific actions within it. You don't have to spend additional actions to perform them—they're already ..."
(this one is a perfect example, because it uses both sides of "action." Both action as a "thing you do" categorically, and "action" as resource points you spend)
You cannot take away Strike's status as "an action" when it's inside an activity.
Activities modify their sub-actions exactly as they say they do. One of the default edits is the action cost, which is always edited to 0, because the activity pays it instead. This does not remove sub-action's categorization as actions.
I have no idea how that claim can be textual, it really seems to be grasping at straws for a post-hoc justification for a vibes based ruling.
That makes sense to me. Thanks!
|
|
| 1 person marked this as a favorite. |
I have no idea how that claim can be textual, it really seems to be grasping at straws for a post-hoc justification for a vibes based ruling.
With all due respect, it is what I feel about your reasoning too.
So, agree to disagree it is.
And if people play PFS, best to bet on the worst interpretation possible.
How do you rule it?
With an iron fist!
I try to avoid making appeals to popularity, and prefer to show the evidence upon which my claims are built, and the logic then applied.
It's a pain in the ass to be so wordy, but I do it so that if there is a fault in my evidence or logic, someone can notice and call out that error.
My goal is to maximize my own fallibility whenever I make claims like this.
_______________
In this case, I have run out of angles trying to explain that there is no textual restriction, nor disqualifier, placed upon sub-actions. Zero.
The "Subordinate Actions" text instead does the opposite, it puts the restriction upon the activity-action and clarifies that sub-actions have all their normal traits and effects, unless the activity-action text removes them.
(Yes, paizo states that activities are actions too, btw. IDK if that clears up any confusion or makes it worse, but everything you "do" in pf2 seems to be actions. "...[the sub-action is normal], but it's modified in any ways listed in the larger action" <- that "larger action" is the activity)
There is no text saying that sub-actions don't count for certain things, in certain contexts, while they do count for others.
That kind of selective discretion around sub-actions is required if some things are compatible, like Sneak Attack, but not others, like Drink.
_________
There *is* room to RaI that activities are not just "first actions," but instead count as a container around their sub-actions. While there is no text instructing that, there is no RaW text that conflicts.
The issue is that the effect of that interpretation is not being applied evenly, and that version of "using activities is not the same as using their sub actions" means disabling anything that keys off those named sub-actions, Sneak Attack only works w/ raw Strikes, etc.
Or, as many people have mentioned, the intention is that the activity-action is a wrapper around the subactions, meaning anything referring to the action as a whole (which can be three, two, one, free or re-action) only refers to the activity-action, which logically is the case for things referencing next or previous actions, while anything referring to an individual action can reference the actions within, like anything giving a flat benefit to any action of a specific type.
Tbh, that kinda sounds like:
when the text says "your previous action" it actually meant to say "your previous activity, or action if there was no activity"
Yeah, I cannot agree with that at all.
I have to trust the devs to understand their own system, and that the devs knew that "action" could mean "sub action."If they wanted to default to that suggested logic, such a rule needed to be put somewhere. Such as in the Subordinate Actions text.
_________
Oddly enough, the logical effect that your post seems to be dancing around is very similar to the programming concept of scope.
That while inside the activity, you have the scope for sub-Strikes to count as Strikes and proc matching effects, but once outside the activity, there has been a change of scope and sub-Strikes no longer count as Strikes.
(Again, without that "temporarily compatible" scope loophole, Sneak Attack, etc, are rendered incompatible by any simple "wrapping".)
I do not see any textual instruction to support it, but people can do whatever they want with the ruleset.
_____
As far as I can tell, the system is set up so that actions are all done in a flat sequence or chain. An activity kicks of a chain of actions, where the first in the sequence is the activity-action.
That helps explain the wording on Simultaneous Actions, which doesn't allow you to pause the activity chain partway through to start another activity or base action, you have to play that original activity to completion first, and can only pause it partway for Reactions or triggered free actions.
Not sure what this is in response to, and apologies that I do a lot of post-post editing, so you may want to extract some quote text that wont change next time you refresh.
That says that outside of resolving the subordinate action itself you have to treat it as if you have done the activity and not the separate subordinate actions.
Ah, ok.
If you treat the activity-action as the first action leading a chain of sub-actions, that possible interpretation disappears.There is no need for a "closer" or "end wrapper" action, nor to put the sub-actions on some deeper hidden level underneath the activity action.
__________
And I'll restate that no, the text does not instruct as you claim. Saying you cannot use X in place of Y does not change how you treat Y. Zero modifications to Y have been made. You still "do the Y", even in that context.
And there has been no instruction to the effect of "X contains invoked Ys, hiding them from other effects."
[if you have Haste] an activity that includes a Strike wouldn't count, because the next thing you are doing is starting an activity, ...
This directly reveals that there is a chain-of-actions behavior. If there was a "wrapping" or "container" behavior, then the first sub-actions would not matter. The "because" would be radically different. In this text, the reason it doesn't work is because you cannot get to that *matching* sub-action without hitting the incompatible activity-action.
This implies that the sub action *does* match the criteria, but still doesn't work because the first action in the chain is the activity-action.
We just went over that activities are actions themselves. Back to the "ghost action" metaphor.
The "activities are the first action in a flat chain of sub-actions"
reading would not allow one to use a Hasted action for activities, because activity-actions have names, and always go in front of sub actions.
(This is why the simultaneous rules are directly above the sub-action rules. You cannot break a chain of activity & sub-actions to start another ___;
the two rules are paired because that rule is required for this "never compatible with 'next action ___'" text to work.)
'As another example, if you used an action that specified, “If the next action you use is a Strike,” an activity that includes a Strike wouldn't count, because the next thing you are doing is starting an activity, not using the Strike basic action.'
That text is written as if to say, ~"Yes, the first sub-action may be a correctly matching Strike for Haste, etc, but you still have to perform the activity-action Skirmish Strike first. Therefore, not compatible."
Again, the text says "Not compatible, because" your next action is the activity, never the sub-Strike. In this "because" it's the order of events that is the "why."
If there *was* a problem with the sub-Strike because of it's identity as a subordinate action, the "because" explanation would have been very different.
| 2 people marked this as a favorite. |
If I use an activity with manipulate, does this mean that I also take a reactive strike at the end of the activity? And this could happen, I an ally could readied friendly push me, there couple be some triggered free action movement I have, whatever. I could end that activity next to an enemy with reactive strike.
If there’s a “ghost blocker action” why doesn’t it trigger reactive strike? It should right? But that’d be stupid.
Additionally, to add some fuel to this fire, there are numerous monster abilities dependent on your last subordinate action counting. Zelekhut’s double attack activity into knockdown is a prime example, as written it literally can’t be activated if your last action wasn’t a strike. But it has a special ability specifically for this case!
There are other examples too, and as I think was mentioned earlier in thread bloodrager is pretty screwed if they can’t drain off metastrikes.
The only interpretation that’s consistent with all printed abilities is that both the activity and last subordinate action coun, as weird as that is.
If I use an activity with manipulate, does this mean that I also take a reactive strike at the end of the activity?
This part is still tripping up my attempt to parse it. Are you trying to R.Strike your own manipulate to hit yourself?
After checking the trigger... that is actually valid. You are a creature within your reach, so in theory you can R.Strike yourself, potentially disrupting your own ___.As odd as it sounds, you can trigger Reactions to your own actions, and many are intended for that. It doesn't matter if your trigger is an action, or is a sub-action inside an activity. If the trigger gets a name match, it's valid.
As an example, Sense the Unseen is a Reaction with the Trigger of "You fail a check to Seek"
It doesn't matter if that Seek is a sub-action, stand alone 1A, etc. The moment the Trigger occurs, you hit "pause" and perform the Reaction. And there are a fair number of weird cases where Seek gets bundled and modified inside of some other action, and Sense the Unseen would work with all of them.
_____________________________
That Zelekhut is a great find, and is a rather irrefutable case.
That text is even phrased with an "if," and there's no wording to indicate it's some special permission / specific override. ("... the zelekhut *may* perform a Knockdown ...)
Keying off the sub-action of an activity for "last action" abilities must be normally valid thing.
Though it might be possible to find contradicting text, imo this topic has a rock solid answer.
Double Attack [two-actions] The zelekhut makes two chain Strikes, each targeting a different creature. Each Strike counts toward the zelekhut's multiple attack penalty, but the penalty doesn't increase until after it has made both attacks. If the zelekhut subsequently uses the Knockdown action, it affects all creatures it hit with Double Attack.
If there’s a “ghost blocker action” why doesn’t it trigger reactive strike? It should right? But that’d be stupid.
The "ghost action" is just the named activity. I'm only describing it as a ghost for those that don't yet realize the activity is an individual action in and of itself.
If that activity-action itself has the [manipulate] trait, then it would trigger R.Strike the moment it is initiated, before reaching sub-actions.
But, a lot of the time, it is those sub-actions that will trigger things like R.Strike. In that case, R.Strike cannot jump ahead and go early, you have to wait for the trigger to actually fire during the sub-action.
To craft an example that should make this seem reasonable, I'll lean on Skirmish Strike again.
[Skirmish Strike] --> [Step] --> [Strike] is the action chain.
3 actions total: one is an activity (the "ghost action" that "blocks" Haste, etc,), and two are sub-actions.
It makes no sense for the Activity to inherit the "is a ranged attack" from the sub-action Strike. That would mean triggering R.Strike before the Step could even happen.
Same principle applies for [manipulate], etc. The activity-action is not the sub-action, and does not inherit from or get changed by the sub-action.
__________________
does this mean that I also take a reactive strike at the end of the activity?
Are you asking if it's valid to intentionally wait and delay the R.Strike to some extent?
I don't think that's valid. Most triggers like "A creature within your reach uses a manipulate action or a move action ..." indicate each trigger moment is a single point in time, where you can pass or take that Reaction.
Wait a bit and then swing doesn't seem to be an option.
You might get multiple trigger matches throughout a single activity, but I don't think you can intentionally delay your Reactions like that.
I’m saying that if there’s an activity action also happening at the end does that also trigger reactive strike just like the start of the activity.
Apologies for my garbled phrasing, my autocorrect was acting up and I didn’t notice in time to edit.
I’m saying that if there’s an activity action also happening at the end does that also trigger reactive strike just like the start of the activity.
Nope. There is no "ghost ender action" that the activity creates.
If that happened, then "last action was ___" abilities would become incompatible. Just gotta remember the named activity is itself an action. There are no containerization worries, nor scope.It's all one flat chain of actions.
It really is kept as simple as that Skirmish Strike 3 action chain example.
[Skirmish Strike (activity-action) ] --> [Step (sub-action #1)] --> [Strike (sub-action #2)]
That is how you stay in agreement with the textual instruction for
Using an activity is not the same as using any of its subordinate actions. For example, the quickened condition you get from the haste spell lets you spend an extra action each turn to Stride or Strike, but you couldn't use the extra action for an activity that includes a Stride or Strike. As another example, if you used an action that specified, “If the next action you use is a Strike,” an activity that includes a Strike wouldn't count, because the next thing you are doing is starting an activity, not using the Strike basic action
and also retain compatibility with the "last action was Strike" abilities: Zelekhut's Double Attack --> Knockdown ... Exemplar's ___ --> Drink of My Foes ... etc.
___________________
Also grabbed some text on 0A free actions / Reactions while I was there:
When its trigger is satisfied—and only when it is satisfied—you can use the reaction or free action, though you don't have to use the action if you don't want to.
| 1 person marked this as a favorite. |
This was a good thread necro'ed. After reviewing the logic cases, each with scenarios to disprove them as the correct action phrasing "theory of everything," I think the only thing that's clear is that the writers weren't consistent in their wording.
I'll have to rethink what to allow based on what's too good or too bad to be true. That's not an issue. It's interesting to see these flaws uncovered though. Thanks for the thoughtful discussion.
| 2 people marked this as a favorite. |
It’s evidence that pushes towards subordinate actions counting as the last action in general. If it actually said in the rules the at they didn’t that’d be one thing, but if we’re inventing a restriction that’s not actually printed instances where that breaks things is evidence against the restriction.
I think the only thing that's clear is that the writers weren't consistent in their wording.
After almost 2 years of digging into pf2 text, and learning that there's a whole swarm of temp authors credited in their books, I'm impressed the system is not more of a mess.
While I do want to +1 and agree that the wording isn't anywhere as consistent as it could / should be, I'd like to throw an * on that.
While the wording can vary, once you translate the text into formal logic, that sort of ~game code instruction, that formal logic is incredibly consistent, and near bulletproof.
Issues with the actual logic are super rare, to the point that I don't have an example on deck to use.
Something like no one actually following the Imm/Wk/Res text's actual instructions is as close as I can think of.
The "real RaW" never actually resulted in broken logic, and worked in 100% of possible cases. But the community invention did genuinely create contradictions in logic, where you needed to use the single rules text to create at least two conflicting logical procedures under the hood to reach the outcome of the "pop every weak/res once" community norm.
And yeah, while not an error of the text, Paizo still should have been aware of the RaW v community contradiction and done something years ago, but better late than never.
That rock-solid formal logic under the hood is why Foundry is able to code the ruleset into a VTT to begin with. And why an indie dev can manage to make a full computer game like Dawnsbury Days where there's no GM to handle exceptions.
(and that solid logic is why Foundry needing to code special exemptions / cheat the Imm/Wk/Res rules based on context was such a red flag for me to look closer and learn that "nope, the community is just wrong on this one")
_____________________
If anyone has text to point at that breaks the logic of the whole
[activity-action] --> [sub-action] --> [sub-action]
idea, please share it. As far as I know, running it with that notion of a flat sequence of actions genuinely does work 100% of the time.
Uh... why was my post deleted?
I don't think it's normal for a rule-abiding response to get shadow-removed like that.
I'm honestly surprised that errenor's post was worthy of removal, and I am now unable to re-read it in order to agree/disagree with that judgement, lol. But that's beside this particular quibble of mine.
If the reply to the offending post didn't break a rule, and it contributed to the discussion (such as by elaborating on a prior point), then by all accounts it should remain, yes?
It seems rather strange for it to be branded guilty by association and removed.
| 1 person marked this as a favorite. |
We have several interpretations re: activities composed of actions:
-We all agree that activities aren't just their actions, otherwise Spellshape would work for Spellstrike.
-Some actions (i.e. Flensing Slice) require that a previous action be an activity, so the end of an action isn't just its subordinate action.
-Some common actions for monsters (i.e. Knockdown) key off of Strikes, yet the creature's obvious attack routine involves an activity with subordinate Strikes. There's no mention of these actions being exceptional or breaking any rules.
I imagine most of us would let a lion Pounce, Strike w/ its jaws, and (if successful) then Grab too, right? I suppose it'd be balanced if both teams were hamstrung by saying no, but that runs counter to PF2's everyday-language vibe IMO.
While the rules draw a clear delineation at the beginning of an activity, there's no matching rule re: the end of an activity. So it seems clear that Scooter's right: "The only interpretation that’s consistent with all printed abilities is that both the activity and last subordinate action coun(t), as weird as that is." That's the interpretation for smoothest play given all the examples we've seen. So Barrow's Edge should work after a non-basic Strike, if the Strike was the last action in that activity.
With Swipe, I'd reference the damage to the second person struck. The language is in the singular "foe's", though I could see counting both as the sentence naturally speaks of common instances, isn't meant to exclude (or include) corner cases.
Wow, there are a lot more "previous action was ___" abilities than I realized.
And if sub-Strikes inside of activities don't count, all of these become incompatible with a whole lot of normal use-cases, many of which are inside their own archetypes. The idea of hitting a Strike, then doing a follow-up action is super common in pf2, and uses that "previous action" wording.
And I'll repeat the inverse problem it causes if you don't allow sub-actions to count as matches.
That ruling means that bans on specific actions like "Strike" can no longer stop Strikes that happen to be sub-actions inside an activity. A loophole a mile wide.
Swarm Eidolon's "cant Strike" becomes incredibly narrow, and even an in-class activity like Tandem Strike enables them to loophole the Strike ban and sub-Strike.
Imo, that's the even larger elephant looming over the discussion that rather definitely ~answers the question as far as any GM running the game needs.
Option A: Activities are a flat chain of actions that go in sequence. All are "actions"
[activity-action] --> [sub-action #1] --> [sub-action #2]
[Skirmish Strike] --> [Step] --> [Strike]Knock on result: "Previous Action was __" text is compatible with the final sub-action chained by its activity.
_____________________Option B: Activities are the only action, sub-actions are contained inside the activity, and are not "real actions"
Knock on result: only the named activity is valid for a match.
The sub-Strike inside Skirmish Strike does not match or trigger as "Strike." Boosts like Sneak Attack no longer apply to sub-Strikes.
And bans on actions like "Strike" can be loopholed via activity sub-Strikes.
________________
And all that searching was to look for another example of that "ugh" case of Flensing Slice, and I finally found another like it in Cratering Drop from Winged Warrior.
The "ugh" is that both of these say "previous actions was" and specify the named activity, not the last sub-action of that chain.
Requirements Your previous action was Pluck from the Sky using a melee Strike, and the target failed or critically failed its saving throw.
I don't blame anyone whose understanding of this topic was screwed up by those, as Paizo could/should instead use their own "activity" word instead of "action" to avoid the confusion.
I'd almost rather be able to call them dev inconsistencies / errors and leave it there, as that would be way easier to talk about. But by the rules of logic, that wording is still valid.
It is not at all intuitive, but in formal logic, including computer logic, thing A can have multiple identities that both return a match.
The natural impulse is to make identity exclusive, where defining "A is a blarg" automatically strips it from having any other identity. But that is NOT default in formal logic; identities are only removed if specifically instructed to do so.
Values ARE exclusive in that way, which is where that natural intuition comes from.
Computer pseudo-code:
variable Integer x; x = 5; x = 10; x = 9;
Each time you set the value, that erases any previous relevance of old values.
Identities are more like a list you keep adding to.
In computer pseudo-code:
variable Object A; A.is("car"); A.is("machine"); A.is("sedan");
By the end of that logic, A has gained three identities.
____________________
The unintuitive result is that sub-actions retain their base identity as a "Strike" or whatever, and also gain another via the activity. Because the activity is the thing chaining the sub-actions together, all sub-actions gain that "is activity" identity. And that activity "is" all of its component actions.
Crazy levels of academic / pretentious pedantry needed to even talk about that.
What's annoying is that this 2 for 2 phrasing isn't even usable as ammo for a counterargument.
"Your previous action was [activity name]" just doesn't make natural sense to anyone, and I'm presuming it's only technically correct due to a fluke of chance instead of a self-aware writing choice (though maybe I shouldn't underestimate the kind of person who writes an entire ttrpg ruleset).
That phrasing isn't useful if you're trying to insist that activities have some special off-page containerization mechanic going on. From that PoV, one would much prefer activities be wholly not actions.
_______________
Another case where activity = action is reinforced, with a very confusing result:
... Some rules will refer to the Cast a Spell activity, such as “if the next action you use is to Cast a Spell.” Any spell qualifies as a Cast a Spell activity, and any characteristics of the spell use those of the specific spell you’re casting.
Flip flopping on activity vs action like that is just nuts when writing for a general audience.
At the very least, it is useful to confirm that, yes, activities are actions, and text calling for "actions" is indeed compatible with / is a match for activities.
But at least for me, the absolute RAW of "'next action used' doesn't qualify" given as an example, I can safely rule that "this also applies to 'previous action used'" is equally valid to not work.
Using an activity is not the same as using any of its subordinate actions. For example, the quickened condition you get from the haste spell lets you spend an extra action each turn to Stride or Strike, but you couldn't use the extra action for an activity that includes a Stride or Strike. As another example, if you used an action that specified, “If the next action you use is a Strike,” an activity that includes a Strike wouldn't count, because the next thing you are doing is starting an activity, not using the Strike basic action
"Next action" doesn't qualify because the activity-action is first in line.
That is not mirrored. The first action may be the activity-action, but final action is the final sub-action.
The monster meta-Strike Pounce + Knockdown style abilities was a great example made by Castilliano, and it's worth repeating.
There are a lot of "previous action was a Strike" style abilities once you start looking for them. All sorts of grabby tentacles, injection stingers, etc. A lot more player-obtainable ones than I had thought, too.
And I'd wager 95%+ of tables have already been playing with that "previous action" compatibility being totally fine and dandy, they just never formalized the under the hood logic.