| 5 people marked this as a favorite. |
Hmm, interesting. I think it's currently ambiguous. My preferred answer right now would be a different order:
Degree-Conditional Effect whenever applies, otherwise Generic Step Shift
That is, if you are ever at a degree of success where an ability wants to shift your degree of success, do so at once (the ability has now done its thing). Once you have done this, you can apply any "treat this as one degree better/worse" stuff, and then if that causes you to use a degree-conditional effect, then do that. Of course, this doesn't help if you have two degree conditional effects at the same time ("Treat a success as a critical success" and also "Treat a success as a failure") but we at least don't have much (any?) situations where we specifically treat all successes as failure or all failures as successes right now.
| 3 people marked this as a favorite. |
I think I have a paradox.
Tempt Fate (divination, Fortune) Power 2
Casting [[F]] Somatic free action; Trigger You or an ally within range attempts a saving throw.
Range 120 feet; Target you or a willing ally in rangeIf the triggering saving throw's result is a success, it counts as a critical success. If it's a failure, it counts as a critical failure, and the critical failure can't be reduced by abilities that usually reduce critical failure, such as improved evasion. If the triggering ability did not have both a critical success and critical failure condition, tempt fate fails and your Spell Point is refunded.
A monk stunning fist criticals you.
You have Tempt Fate in effect.
You roll normal success before abilities.
What happens?
1: Normal success > critical = normal fail > Tempt Fate = critical fail (damage, flat-footed, stupefied 2, stunned)
2: Normal success > Tempt fate = critical success > critical = normal success (damage)
If we use what I said above, it'd be 2, since tempt fate is degree-conditional.
| 2 people marked this as a favorite. |
Specific over general sounds like a good first ordering, but you might still need a tie breaker -- either active player/creature or target player/creature. Probably doesn't matter which is picked, but should be formalized.
Ambiguity like this is what game-breaking exploits are made of, and D&D 3.0 / 3.5 had a good number of them. So unless you can promise to never ever release abilities that interact like this - and force third party companies not to either, which seems unlikely - then establishing a clear order of priority is a good idea.
Personally I'm not fond of abilities turn a success into an auto-critical, or that turn a failure into a critical failure. I feel like the critical effect is supposed to be something special and less common, so that players feel good (or bad) when they achieve it; having it happen regularly and reliably cheapens it. This feeds into game design as well: the description of a given spell or ability's critical success can do something awesome (and its critical failure can do something terrible), but only if that condition is rare enough. If the designers writing spells and abilities expect criticals to happen all the time, then they're held back from making them awesome. It makes the game less heroic.
I'd do the sensible thing, and factor any applicable shift in degree as +/- 1(or 2, but I don't think there are ability that makes shift more than 1 step), tally them all, then apply.
This should also prevent applying multiple times the same ability (like, trat succes as critical, shift back, treat the now succes again as critical)
|
| 1 person marked this as a favorite. |
This whole building context sensitive critical success and failure results into specific abilities is already a complete mess. Every ability that has this (And it seems like they're putting it ALL OVER) has its wordcount DRAMATICALLY raised, and leaves the door wide open to poor interpretation.
Critical Success and Failure being determined by the difference in the the DC vs Total Result is neat and all but the bookkeeping nightmare it's going to create is going to drive GMs absolutely MAD trying to track what condition or effect I should be using.
If we have to keep the the Treat X as a Critical Success/Failure Abilities, can we PLEASE PLEASE PLEASE encode a rule that the Degree of Success or Failure can only be moved in each direction ONCE per Check/Attack? Pretty please?
| 3 people marked this as a favorite. |
I could see one ability that turns failures into successes being paired with another that turns successes into critical successes.
It'd be important to know which order to apply those in.
I think maybe only the initial roll's status should matter for triggers which change the success/failure status.
Then, once the roll's been shifted to that status, one checks what the result is.
But already I can see an issue if the attacker scores a critical (which shifts the target's results) and the target also has an ability which shifts her results.
-I turn your success into a failure!
-Hah, but I can turn my success into a critical success!
-No, it was a failure first, so you cannot do that!
So, with Defender rolling
Attacker applies best effect which changes status of the roll based on the roll itself, but no further effects based on the new status gained (even if they have two separate abilities).
Defender then applies best effect which changes the status of the roll, but again with no secondary effect based on the new status gained.
Check last status for its results (and apply any Attacker/Defender abilities that don't change the status, but do amplify/mitigate effects.)
Seems as simple as possible, even if a bit lawyer-y.
So in the example dialogue, the success gets turned into a failure first, then the ability to turn a success into a critical success does not activate.
Cleaner version:
Roll occurs, a status comes up (crit success/success/etc.)
Initiator: Apply one effect that changes that status (if applicable)
Recipient: Then applies one effect that changes status (if applicable)
Initiator: Apply abilities to that result (i.e. x1.5 damage)
Recipient: Apply abilities to THAT result (i.e. Resistance 10)
Note: This can sometimes work out in favor of the recipient, if say the initiator has an ability that nullifies critical successes and the recipient has an ability that turns success into critical successes.
So... (sigh) maybe the initiator should choose order of operations???
| 2 people marked this as a favorite. |
I like Castilliano's initiator/recipient order of operations both for determining the degree of the roll as well as the result of the subsequent impact. It makes sense.
It also makes sense to restrict the change in degree of success to one move each, chosen by each party.
So if the initiator has two abilities, one to to treat the roll as one degree better and one to turn a success into a critical success, and the result of a roll is a fail then they don't get to use both abilities to go from fail -> Crit success. They have to choose one of their abilities (which in this case should be the ability to treat the degree as one better).
Then if the defender has equivalent defensive abilities to treat a crit success as a success and one to make the degree one worse then they get to choose which they want to use. In this case they should choose the make one degree worse ability and the result of the roll would go back to fail.
This would restrict the movement of degree to a max of two per action. It's more than one move but it lets both sides play with their toys.
| 1 person marked this as a favorite. |
On the idea of an initiator/recipient formal order:
Fantasy Flight's X-Wing miniatures game has rules about modifying dice roles (using their weird symbol dice that I strongly dislike) during attacks. An individual attack or defense die has essentially three tiered outcomes on it, with a "best" result, a "not as good/conditional" result, and a blank result.
Basically, the attacker rolls, modifies their own dice, the defender has an opportunity to modify the attack dice, and that's it. Then the defender rolls dice, modifies their own dice, the attacker has an opportunity to modify the defender's dice, and that's it. Results are compared between the two dice pools at that point.
It works okay, but there's a big difference with that game: it uses dice pools. Because there are usually multiple dice involved, the ability to change one die's result is typically less powerful than changing the results of a save (which is considerably more all-or-nothing). It would result in some abilities always trumping other abilities, which could be unfortunate. It would also, at least, be consistent and therefore could be designed around.
Hmm, interesting. I think it's currently ambiguous. My preferred answer right now would be a different order:
Degree-Conditional Effect whenever applies, otherwise Generic Step Shift
That is, if you are ever at a degree of success where an ability wants to shift your degree of success, do so at once (the ability has now done its thing). Once you have done this, you can apply any "treat this as one degree better/worse" stuff, and then if that causes you to use a degree-conditional effect, then do that. Of course, this doesn't help if you have two degree conditional effects at the same time ("Treat a success as a critical success" and also "Treat a success as a failure") but we at least don't have much (any?) situations where we specifically treat all successes as failure or all failures as successes right now.
This seemed a little confusingly worded at first, in that the short version seemed completely unclear (because Degree-Conditional Effect did not immediately mean anything to me) and the long version was fine (because the term was explained and retroactively felt obvious). It reminded me of a rule on modifying numbers in the card game INWO, which had a defined order of operations for changing the attack/defense values of cards (and didn't use play order, like most card games).
When several cards modify Power or Resistance, changes to a specific value come first, then effects that
multiply or divide, and then effects that add or subtract. For instance, if Grassroots Support (increases Power to 6) and The Big Prawn (doubles Power) are linked to England, and the NWO card Law and Order (gives all Straight Groups +2 Power) is in play, first change
England’s Power to 6, then double it to 12, and then add the +2. Its final Power is 14.
I've always thought that was a fairly elegant way to handle that, since the numbers needed to be consistent regardless of whose turn it was. Mark's quick rule appeals for much the same reason, but as he points out, leaves the awkward space for two degree-conditional effects. Attacker/defender (or vice versa) means one will always trump the other, and potentially gets more confusing (imagine a bard doing a saving finale type reaction that modifies the result of an ally). Perhaps conditional positive shifts trump negatives (or vice-versa)?
If we have to keep the the Treat X as a Critical Success/Failure Abilities, can we PLEASE PLEASE PLEASE encode a rule that the Degree of Success or Failure can only be moved in each direction ONCE per Check/Attack? Pretty please?
I'm not sure if I like or dislike this idea. Having something that makes failure a success, and stacking it with something that makes success a critical success...seems like it would be cool? But also it seems like it would almost certainly be exploitable in a not-fun way, so limiting this movement is probably a good idea and seems in the spirit of the overall effort to keeping the d20 result relevant across all levels.
Wow, this is getting really messy. And the more such abilities exist, the more unforseen interactions there are going to be.
I think the examples above illustrate the danger inherent in any "treat any success as critical" type abilities. Even though they fit in with the historical behaviour of abilities like Evasion, they interact badly with the new +/-10 critical roll mechanism.
Compare this to the simplicity of the thing D&D 5e got really right, the advantage/disadvantage mechanism. Each side in a conflict may have multiple abilities that affect the roll, but they're managed by a unified system. Advantages and disadvantages cancel out, and the result is that you either have advantage or disadvantage or you don't. You then roll one or two dice, add whatever modifier, and that's your final result.
Now, I like the +/-10 mechanism, and I don't want Pathfinder to mimic 5e, either in specific rules or in being oversimplified. But we really need clarity in these rules to prevent broken interactions.
A restriction that says "initiator gets to pick one ability to affect it; then recipient gets to pick one ability to affect the outcome" sounds like a fairly healthy way to limit abilities. After all, by high levels people will have picked up enough feats to chain them in stupid ways, turning a critical failure into a critical success.
But I think my preferred solution would be to cut these result-changing effects out of the game entirely. Find some other way to affect the result - either a bonus modifier, or extra dice. Sure, you'd have to put up with grumbling from everyone who's nostalgic for the way Evasion used to work, but it's not like people won't grumble anyway.
| 2 people marked this as a favorite. |
One thing I like about the 'Initiator picks up to one modifier followed by Recipient picking up to one modifier' is that it brings a bit of a card game feel to the combat without slowing things down. I can easily imagine a quick exchange like:
GM: The boss rolls a success against you AND it has a special ability where one success can turn into a Critical Success. It uses it take...
Player: Hold on! I burn this expensive trinket I bought that turns a Critical Success against me into a Fail. HA! You just activated my trap card!
GM: Haha Dammit! Ok Party up.
| 1 person marked this as a favorite. |
Its only getting messy because we're throwing tons of alternatives out there and the developers need time to think through all the issues and formulate a definite plan, coupled with any hard-and-fast rules they need to add to their meta-rules manual for their in-house team to avoid writing new things that break the assumptions their proposed order of operations requires to be consistent.
I'm fairly confident they can come up with a 2-3 step list of rules that'll be easy in practice to remember/use once they decide on some some primary principles.