Roll for Effect or Intent?

[pemerton]

Hit points (and equivalent) are in part a measure of mental health and exhaustion. You can take psychic damage from magical mockery. 'I skip a stone across the lake like we used to do when we were kids, to cheer my sister up [and thus heal her a bit]'.

This came up a few times in discussions of 4e: what's the DC for a Diplomacy check to allow another character to spend a healing surge? The logic of hit points suggests it should be possible; but there is the risk of intruding on the role of the Healing skill.

One possibility is that, on the failed Diplomacy roll, the other person loses hp as the putatively cheering remark actually leads them to reflect on their current dire circumstances.

Having players express the desired goal and then have that be the expected outcome of a successful roll. How is that not player determined stakes?

There's a ton of player-defined stakes games out there (including one I wrote) and they all tend to have a big discussion about this stuff, including guidelines for what may or may not be permissible stakes, how to navigate different expectations, etc. This solves the 'I persuade the king to give me the crown and a million gold pieces' issue.

4e D&D handles this in a variety of ways: treasure parcels per level (for feasible gifts/extortion outcomes/results of searching etc); p 42 (for damage consequences); spending a healing surge in combat (for recovery consequences); etc. It helps to have a good working knowledge of the range of effects available from the typical powers available at a given level of play.

Marvel Heroic RP builds the limits on stakes into the core resolution mechanics: all outcomes are die-rated, and the rating of an outcome is determined by the effect die of the action resolution roll used to bring it about.

Torchbearer 2e has a list of possibilities (and corresponding difficulties) associated with each skill.

Burning Wheel, on the other hand, is pretty open-ended. And Prince Valiant even more so. We use a mixture of "common sense" and "negotiation" to set limits and possibilities where there's uncertainty or disagreement.

my understanding of systems that use that principle, like Burning Wheel in which it is phrased as "Roll the dice or say, 'yes'", for an action declaration to be permissible, it needs to pass the following two tests, both of which are decided (informally) by table consensus: a) it has to be within genre, and b) it has to have whatever fictional positioning is required. Depending on the game, your "shooting the moon" example might fail the first test, unless it's a mythical type setting where characters do things like that, and would probably fail the second too, unless the character has an ultra long range moon-shooting arrow or something. Notice I said "table consensus". This is not a power the GM can exercise unilaterally.

This fits my experience of BW pretty well.

The Adventure Burner/Codex does have a discussion of when the GM can say "no", mostly in the context of Wises and similar abilities being used to establish backstory, where that bit of backstory is already an (as yet unrevealed) part of the GM's "big picture". This is an area where the game is a bit unclear (or, if you like, inconsistent) on the distribution of backstory authority. Though in my experience it hasn't been a genuine problem in play.

Torchbearer 2e is pretty similar to BW in many ways, but is much clearer in giving the GM the bulk of the backstory authority. So it's version of "roll the dice or say 'yes'" (which is called the Good Idea rule) works without issues, in my experience at least.

 

[pemerton]

Here's an example of the difference that happened to me once:

The group has infiltrated the enemy encampment and we are in the enemy leader's tent killing him in his bed.

An enemy guard hears a bit of noise and pokes his head inside the tent. I say that my barbarian tries to intimidate him into forgetting all about us with a throat-slitting gesture and a 'you saw nothing'.

I had a decent bonus and also rolled a critical success.

The GM agreed it was a critical success and said, oh, he is so scared of you, and therefore of how big a threat to his mates you are, that he immediately runs away in terror and raises the alarm.

Task resolution: this is acceptable.
Conflict resolution: this is not acceptable

I think this illustrates how task resolution, in combination with no sense that the GM is obliged to honour the intent, is a recipe for railroading.

 

[soviet]

I think this illustrates how task resolution, in combination with no sense that the GM is obliged to honour the intent, is a recipe for railroading.

Yes. It doesn't have to. The GM in my example has been my regular GM for a long time and isn't normally like that (hence why I remember it). But for sure task resolution/rolling for effect privileges the GM's vision/editorial control in a way that helps/enables some playstyles but hurts/blocks others. Moments of resolution create jigsaw pieces for the GM to assemble into an overall outcome rather than simply determing that outcome directly.

 

[pemerton]

The player wants to throw a rock. It might or might not hit the tree. Everything else is GM-determined.

If I as a person in the real world throw a stone at a tree, and I hit, I might hope that it distracts the nearby guard, but I cannot guarantee that it will.

In Burning Wheel, on a successful roll the task is successful and the intent achieved. On a failed roll, the intent is not achieved - and the GM narrates what happens, which may or may not include the task succeeding.

So there are three possibilities:

• Task succeeds, and intent succeeds (requires a successful roll) - the stone hits the tree and the guard is distracted;
  
• Task succeeds, but intent fails (GM narration of a failed roll) - the stone hits the tree but the guard ignores it (because they are strong-willed; because they don't hear it; because it disturbs a beehive and the guard doesn't want to move closer to the angry bees; etc);
• Task fails and intent fails (alternative GM narration of a failed roll) - the stone misses the tree.

What is not possible, in BW, is that the task fails but the intent succeeds - eg it's not possible that the stone misses the tree but strikes a gong sitting behind the tree in the backyard, and the guard goes to investigate the gong. A lot of "event-based" scenarios rely on this sort of possibility though - eg the PCs fail their attempt at interrogation, and so find a clue on the corpse instead - and one well-known set of principles for running those sorts of scenarios (the three-clue rule and node-based design) are expressly intended to produce this sort of result.

Sticking to failure in intent-based resolution requires a type of discipline that is not compatible with a certain, fairly typical, sort of scenario.

A roll is called for to resolve uncertainty in whether the approach succeeds in achieving the intent or not.

Another, slightly different, principle I try to follow is "Say yes or roll the dice."

I think they're not just different, but actually contraries, or close to.

"Say 'yes' or roll the dice" - in its DitV and BW formulation (and I don't really know of any other) - means that we first work out, by reference to player intent, whether or not there is anything at stake, and then on that basis decide whether or not the dice need to be rolled, and hence whether or not there is uncertainty.

Whereas the 5e-style "roll where there's uncertainty" first calls upon the GM to make a determination about the fiction (which can include having regard to secret/unrevealed backstory and setting elements), before then deciding whether or not to make a roll. Thus it is primarily the GM who gets to decide what is at stake in the situation.

To build on what I said above in this post, "roll where there's uncertainty" is quite consonant with techniques like the three clue rule. Whereas "say 'yes' or roll the dice" is not, because of the way that it generates finality on a failure.

 

[soviet]

In Burning Wheel, on a successful roll the task is successful and the intent achieved. On a failed roll, the intent is not achieved - and the GM narrates what happens, which may or may not include the task succeeding.

So there are three possibilities:

* Task succeeds, and intent succeeds (requires a successful roll) - the stone hits the tree and the guard is distracted;​

• Task succeeds, but intent fails (GM narration of a failed roll) - the stone hits the tree but the guard ignores it (because they are strong-willed; because they don't hear it; because it disturbs a beehive and the guard doesn't want to move closer to the angry bees; etc);
• Task fails and intent fails (alternative GM narration of a failed roll) - the stone misses the tree.

What is not possible, in BW, is that the task fails but the intent succeeds - eg it's not possible that the stone misses the tree but strikes a gong sitting behind the tree in the backyard, and the guard goes to investigate the gong. A lot of "event-based" scenarios rely on this sort of possibility though - eg the PCs fail their attempt at interrogation, and so find a clue on the corpse instead - and one well-known set of principles for running those sorts of scenarios (the three-clue rule and node-based design) are expressly intended to produce this sort of result.

Sticking to failure in intent-based resolution requires a type of discipline that is not compatible with a certain, fairly typical, sort of scenario.

I understand all of this. I was trying to characterise what strict TR looks like in response to someone who broadly denied it existed.

 

[pemerton]

Yes. It doesn't have to. The GM in my example has been my regular GM for a long time and isn't normally like that (hence why I remember it). But for sure task resolution/rolling for effect privileges the GM's vision/editorial control in a way that helps/enables some playstyles but hurts/blocks others. Moments of resolution create jigsaw pieces for the GM to assemble into an overall outcome rather than simply determing that outcome directly.

John Harper had a good diagram to illustrate this "jigsaw pieces" approach: The Mighty Atom

 

[pemerton]

I understand all of this. I was trying to characterise what strict TR looks like in response to someone who broadly denied it existed.

Yes, I know. But in the context of the thread is seemed like it might be useful to spell it out.

 

[Charlaquin]

Okay, I think I was a bit confused because you said hitting the tree with a thrown rock (the task/approach) is trivially easy and wouldn't require a roll. By that, I understood you to mean the task would automatically succeed, no roll required, and it does appear that you are having the rock hit the tree on both success and failure of the die roll. Likewise, you're calling for the roll to determine if the intent is accomplished, so I'm not sure what about my post you're objecting to except that I'm talking about task and intent as two things that can succeed or fail independently, which is what it looks to me like they're doing in your description of your adjudication, but after reading your post I think I was wrong about that.


What I mean by "maintain the relationship between task and intent" is that task and intent both either succeed or fail together, similar to what you're saying above. Task resolution only cares about the success or failure of the task, so you can succeed on the task but fail on the intent, or fail on the task but succeed on the intent, "breaking" the relationship between the two. I thought that's what you were describing happening in your adjudication, but I now think I was mistaken.


I'm asking two things or maybe the same thing in two different ways. It has to do with you having said the stone hits the tree on both success and failure. I've been calling that the task, but it's an oversimplification or shorthand for hitting the tree in such a way that would prove successful in distracting the guard, which I think has led to our miscommunication. In an earlier post, you had basically dismissed that task as trivially easy, leading me to believe you were ruling it an auto-success. My questions were an attempt to understand why then you were calling for a roll and if it was because you were subbing in a different task (hitting the tree just right, or something else) rather than the one you had said you'd let succeed automatically (just plain hitting the tree), which I now think is probably what you meant.


I think you missed the "If not" at the beginning of my question which referred to my previous question which you don't seem to have understood. My bad for not having been more clear. From your response here, though, especially that you'd call for a Dex check, I'd say you do seem to be using the die roll to resolve whether the task (throwing the rock) is performed in such a way as to achieve the intent (distracting the guard), thus maintaining the relationship, in this ruling anyway, between task and intent. I'd say our miscommunication here highlights the difficulty of knowing what GMing principles might be in play from moment to moment that would allow a task resolution system like 5E's ability check system to possibly result in something that looks more like conflict resolution.

You’re confused because you’re trying to treat “the task” and “the intent” as two different things that can be resolved independently, but I am not. There is only the action. What you’re calling “the task” (what I usually call the approach) and what you’re calling “the intent” (what I usually call the goal) are both part of it. Without both, there can be no action and no resolution (in my games, that is). I’m not calling for a roll to see if the rock hits the tree. I’m not calling for a roll to see if the rock hits the tree in a particular way that would be distracting. I’m not calling for a roll to see if the guard gets distracted when the rock hits the tree. I’m calling for a roll to see if the PC successfully distracts the guard by throwing the rock at the tree.

I don’t know what these “task resolution” and “conflict resolution” principles you’re talking about are, but neither of them sound like what I’m doing. If I had to give what I’m doing a name like this, it would be “uncertainty resolution.” The roll is not to resolve what happens in the fiction, it’s to resolve uncertainty in the outcome of the PC’s specific attempt to accomplish a goal by the means they described. It’s to determine if the approach succeeds in achieving the goal, or fails to do so, and incurs a cost or consequence as a result.

Let me unpack our working example. The player declares “I try to distract the guard by throwing a rock at a tree,” or something along those lines. I’d say that’s reasonably specific and includes both goal and approach, so it’s a complete action declaration. Now, I need to determine the results, so I ask myself, “can I reasonably imagine the character throwing a rock at a tree successfully distracting the guard?” Yes, I can. If I couldn’t, I would just rule that it fails and move on, because there’s no point rolling if it can’t succeed. Next, I ask myself, “can I reasonably imagine the character throwing the rock at the tree and failing to distract the guard?” Yes, I can. If I couldn’t, it would just succeed because there’s no point rolling if it can’t fail. But I can imagine it failing in a few different ways, actually: the throw could miss the tree, or it could hit the tree but the guard might not hear it, or it could hit the tree and the guard could hear it, but instead of being distracted, he might recognize that there’s probably someone sneaking around throwing rocks and become more alert rather than distracted. Next, I ask myself, “would this action have a cost or a consequence for failure?” Another way to think about this is, does anything stop the player from trying over and over again until they succeed? Now, in the case of the rock missing the tree, I don’t really think there is. I guess the player could run out of rocks to throw, but we probably haven’t established how many rocks are on the ground in the character’s immediate vicinity, and even if they did run out, they could move on to throwing twigs, they could throw random objects from their pack… at some point, this becomes functionally indistinguishable from a task that can’t be failed, which is why I said earlier that hitting the tree with the rock is trivially easy. However, in the case that the rock does hit the tree, there is a potential consequence for failing - the guard might become more alert. So, now we have a goal, an approach that could reasonably achieve it or fail to achieve it, and a cost or consequence deterring repeated attempts. So, we need a die roll to resolve this uncertainty and determine which outcome we move forward with.

After breaking that all down, I do kinda think that a better roll might be a Wisdom (Insight) on the guard’s part rather than a Dexterity (Deception) on the player’s part.