Forum changes: Editing of posts has been turned off until further notice.

Main Menu

Nuances of Conflict Resolution

Started by Halzebier, December 01, 2005, 05:29:54 PM

Previous topic - Next topic


Hi there!

This thread is a sub-thread of Task vs. Conflict Resolution – Saying it for my group wherein we discussed how I could break the concept of Conflict Resolution (CR) to my group. A number of definitional questions came up and I'd like to discuss them here.


I'd say that at its most basic, CR is negotiating how (and THAT) a roll will resolve a conflict. That sounds like a no-brainer, but my point is this: (a) whether the conflict is meaningful, (b) whether the consequences are known in advance, and (c) who calls for a conflict are separate issues.

CR conflicts do not have to be meaningful. "Make a climb check at –4 to see if you can cross the ledge without taking 1 point of damage" is CR. However, formulating a conflict will automatically expose it as meaningful or trivial. As Sean put it in the other thread:

Quote from: Sean[It's] a way of calling bullshit on apparently system-generated and/or habitual conflicts, among other things, as well as getting people out of the 'you do something, you should roll' mentality.

Thus, using CR will naturally lead to meaningful conflicts. Andrew's "Why" from the other thread will be answered and if the answer is not satisfactory, people will find a more interesting and meaningful conflict at hand or just drop the whole thing.

(This has been my personal epiphany when running The Pool, so it's not surprising that I consider this a major mental shift and the reason why the TR-CR distinction is important.)

CR conflicts do not have to be explicit. Eero pointed out Polaris, but as I've only leafed through it so far, I'll point to the The Pool, where the GM is free to invent all sorts of complications after a failed roll. What makes it CR is that it is understood and built into the system (by way of players framing conflicts) that the conflict will indeed be resolved by the roll. CR vs. TR is a complete non-issue here.

CR does not mean that the game is "player-driven" or that conflicts are framed by the players. However, once players (and designers) get a taste of how cool it is to reliably get interesting things done, it's only natural that they'll want to make use of that. So CR can make for a more player-driven game very easily.




Right on, right on, and right on.


Ron Edwards


I have some strong advice: do not do it through email. This must be a face-to-face, social discussion.


Andrew Morris

Here's the definitions of TR and CR from the Provisional Glossary:

QuoteTask resolution
A Technique in which the Resolution mechanisms of play focus on within-game cause, in linear in-game time, in terms of whether the acting character is competent to perform a task. Contrast with Conflict resolution.

Conflict resolution
A Technique in which the mechanisms of play focus on conflicts of interest, rather than on the component tasks within that conflict. When using this Technique, inanimate objects are conceived to have "interests" at odds with the character, if necessary. Contrast with Task resolution.

I differ slightly from this definition. While I make no claims to being an expert, I think the difference between TR and CR can been seen easily by looking at only what the resolution mechanics actually resolve. The thing in game that the players expect will likely lead to the desired result (cause)? TR. The result itself (effect)? CR. Issues like intent and scale are irrelevant for defining TR and CR, though they might relate or indicate. For a good discussion on CR v. TR, check out this thread.


1. A player wants his character to shoot a security camera so that he can get down a hallway without being observed.

1A. The player rolls his Firearms skill to determine whether or not he hits the security camera. He succeeds, and the camera is hit. The likely outcome of this is that the character will now be able to get down the hallway unobserved, but there is no guarantee of it. A patrol could come around the corner, there might be a backup, etc.

1B. The player rolls his firearms skill to determine whether or not he gets down the hallway unobserved. He succeeds, and makes it down the hallway. Most likely, the manner in which he succeeded was by shooting out the security camera, but not definitely. Perhaps he missed the shot, and took out the real monitoring device. Perhaps he missed the camera, which was a decoy anyway, and scared off the guards, who are the only real security.

2. A player wants to find search an NPC's office, so he can find dirt on the NPC.

2A. The player rolls his character's Search ability to search the room, and succeeds. He will find any dirt that's there. There might be dirt, or there might not.

2B. The player rolls his character's Search ability to find the dirt, and succeeds. He finds the dirt, likely from searching the room, but not necessarily.

1A and 2A are Task Resolution, because they are resolving the action (cause), while 1B and 2B are Conflict Resolution, because they are resolving the goal (effect).
Download: Unistat


Andrew, all that's true and good.

Reflect a little and I think you'll see that "you make explicit what the roll resolves before you roll" and "you don't make explicit what the roll resolved until after you've rolled" is another way to say exactly what you just said.

For most purposes, I think, a better way, but that's just me.

John Kim deserves credit for this new formulation, by the way, I believe.


Andrew Morris

Quote from: lumpley on December 01, 2005, 09:58:31 PMReflect a little and I think you'll see that "you make explicit what the roll resolves before you roll" and "you don't make explicit what the roll resolved until after you've rolled" is another way to say exactly what you just said.

I don't think so. What I'm saying is that TR resolves the cause, while CR resolves the effect.

Maybe I'm just not understanding what you mean by "what the roll resolves," since, to my way of thinking, that's what makes the definitional difference between TR and CR. Could you give me some examples of explicit vs. non-explicit, as you see it?
Download: Unistat


You already gave examples.

Explicit: the roll resolves whether I get down the hallway unseen. Nonexplicit: does the roll resolve whether I get down the hallway unseen? Dunno; we left that ambiguous.


Josh Roby

I think we have some crossed wires, and an assumption that characteristics of a thing are the definition of the thing.  Specifically, stakes and resolution systems.  Conflict resolution requires declared stakes, yes.  However, declaring stakes does not make something conflict resolution.

Task Resolution is about, well, resolving a task.  It determines "whether the acting character is competent to perform a task."  It concerns only in-game causality in a relatively narrow window of game time.  It cares nothing for why the character is performing that task, only how competent the character is in tasks of that nature and the resources they can bring to bear on the task.  Stakes can be declared before rolling ("Beat 15 to make the jump without falling into the alley.") or they can be up to the GM after the roll is made ("You failed the roll -- you fall into the alley / you're hanging by your fingertips / you crash through the third-story window.").

Conflict Resolution, on the other hand, is about resolving a conflict.  It arbitrates "conflicts of interests" and it requires the participants of the conflict to have those "interests" even if it requires some anthropomorphism.  That is, everybody comes to a conflict with a desire to be fulfilled, and it's a conflict because there's an obstacle to that fulfillment, often each other.  If no participant has a desire involved, stuff may happen, but it won't be a conflict, and you can not effectively apply conflict resolution.  Conflict resolution determines whose desires get fulfilled and who does not overcome their obstacles. Because CR resolves conflicts, and conflicts consist of interests, CR therefore requires declared stakes in order to define the conflict being resolved.  Otherwise nobody knows what you're rolling dice for.  If the conflict is "I want to kiss Mary-Sue but her father frowns upon that" and I win, we know I kiss Mary-Sue since that's the conflict being resolved.

I like declaring stakes; it empowers and informs players to better address the situation.  It's nifty.  But it's not Conflict Resolution.  CR is a fundamentally different way of approaching a fictional situation, as a narrative instead of as a simulation (not, not, not talking about CA here).  CR deals with characters, interests, and obstacles.  By contrast TR deals with enactors, competencies, and difficulties.  Declaring stakes is a piece of the CR apparatus, but it is not the apparatus itself.
On Sale: Full Light, Full Steam and Sons of Liberty | Developing: Agora | My Blog

Josh Roby

Quote from: Joshua BishopRoby on December 01, 2005, 10:56:17 PMIf no participant has a desire involved, stuff may happen, but it won't be a conflict, and you can not effectively apply conflict resolution.

AKA: Say Yes or Roll Dice.  Don't use the handy Conflict Resolution system to arbitrate a Task.
On Sale: Full Light, Full Steam and Sons of Liberty | Developing: Agora | My Blog

John Kim

Quote from: lumpley on December 01, 2005, 09:58:31 PM
Reflect a little and I think you'll see that "you make explicit what the roll resolves before you roll" and "you don't make explicit what the roll resolved until after you've rolled" is another way to say exactly what you just said.

For most purposes, I think, a better way, but that's just me.

John Kim deserves credit for this new formulation, by the way, I believe.

Well, I discussed it in a blog post: Stakes and Freeform Play.  However, I didn't label it as the definition Conflict Resolution versus Task Resolution at the time.  

I like this definition because it pretty clear.  I'm still a little fuzzy on what changes in a larger sense.  That is, I can see what the difference is, and I've tried them both -- but I have trouble putting a finger on the larger repercussions of the change.  So, suppose I'm rolling a single die.  To determine what it means, say there are two options:

1) You can discuss before the roll and agree what each of the numbers 1 through 6 will mean if they come up.  

2) You can discuss only vaguely what it means, and then when the die is rolled and it comes up "5", agree on the meaning of that number in particular.  

Now, I can see the operational differences between these two.  And I've done some of each.  But I can't completely articulate the differences this makes in a larger sense.  For example, I disagree about Hal's #3 that it necessarily speeds up the game.  You can have fast and decisive cases of #2.  I hate repetitive rolls like the bridge thing, and I almost never have them in my game, but I still do #2 pretty often. 
- John

M. J. Young

What Joshua said.

Seriously, this is very confused, but I think it's largely because we're not meaning the same thing by the same words. I saved a quote from what I take to be the parent thread, because I was hoping that this thread would already exist, and if it didn't I fully intended to start it. Here it is.
Quote from: lumpley on November 29, 2005, 10:26:58 PMI think that "in conflict resolution, we say what's at stake before we roll" is a great way to put it, far better than how I've put it in the past.

From what he says here, I can clearly see why Vincent thinks that's a clear way to put it, but frankly I think it's a terribly murky way to put it.

OAD&D says that the thief trying to climb a wall must roll his percentage against his chance of success, and that if he fails, he falls from the mid-point of the climb (yes, it says that). It also gives very particular details on falling damage, and for monks it provides a skill that can reduce or eliminate damage on a successful roll. Thus if I am playing a thief or monk in OAD&D and I decide to climb the wall to cross a fifty foot deep chasm, I know quite exactly what is at stake--but it is still task resolution, because what the dice determine is whether I make the climb successfully.

If I am just trying to get to the other side to see what's over there, it's simple to apply task resolution and very difficult to apply conflict resolution. However, if I'm trying to get to the other side to catch the other thief before he manages to unlock the door and escape into the minotaur's maze, task resolution becomes the difficult thing and conflict resolution the easy one. To illustrate:
  • Task Resolution: I announce my intention to climb the wall to get across the chasm; the player controlling the fleeing thief announces his intention to unlock the door and flee into the minotaur's maze. I roll to climb walls, he rolls to open locks. If both of us fail, I plunge to my probable death and he is still trapped on that ledge. If I succeed and he fails, I got to the ledge while he was still there, but we don't know what happens next. If he succeeds and I fail, he opened the door and I fell. If both of us succeed, I got across and he opened the door, but we don't know how soon before I got across he managed to get the door open, or if in fact he's just managing to do it when I tackle him. Task resolution has told me a lot, but in half the cases it has not told me what I really wanted to know: did I catch the guy?
  • Conflict Resolution: I announce that I'm going to climb the wall to get to the other side to catch the thief, and the other player announces that he is going to open the locked door and flee into the maze. I roll and he rolls, and our rolls are against each other. If I succeed, I caught him; if he succeeds, he got away. Did I fall? Did he get the door open before I got there? Did I realize I couldn't make the climb and return to my side? Is he still stuck on the ledge on the other side? Those things might be determined, usually by some sort of degree of success concept (by how much did you win/lose) often combined with some sort of assignment of narration/fact creation rights, but the essential question was always about whether I caught him, and that's what the dice decided.
I can see that Vincent is right, that this is in some sense about knowing what is at stake before you roll, but ultimately it is much more about putting at stake the objective and not the task.

The discussion is confused, I think, by the periodic use of event resolution, which I take to mean a collection of tasks bundled into a single roll (Mike frequently mentions a game in which the players can decide how much is determined by a single roll, but if we're talking about combat it's still the difference between whether I landed the blow and whether I won the melee, not about whether I rescued the girl), and of outcome resolution, which I take to be more like conflict resolution without the opposing character interest (Multiverser's GE rolls are of this sort, in which the "conflict" is between what the character wants to have happen and what he wants to avoid, without necessary reference to what anyone else wants or wishes to avoid, such as when natural forces are potentially involved).

It also gets confused (as John has pointed out before) when the accomplishment of the task and the achieving of the objective are indistinguishable, such as the previous example of crossing to the other side of the chasm merely to get there, or winning a duel because a challenge was issued and you are honor-bound to fight. In the latter case it is often difficult to see "what is at stake" as something other than simply the fight itself (although even there conflict resolution can be managed as "do I preserve my honor" rather than "do I win the fight"). The former case is handled in different ways in different games. As mentioned in the glossary, sometimes it is done by anthropomorphizing inanimate objects, and so making the task a conflict between the thief and the wall (or the chasm or the ledge). In other cases, such as Legends of Alyria, such tasks are unimportant, precisely because if they don't involve conflict no one cares. If you want to climb over to the ledge just to see what's there, and you have that ability, you succeed. If you want to do that because it will somehow advantage you over some other character, your effort to reach the ledge is a conflict with the other character.

So thanks for starting the thread, and I hope this illumines more than it obscures.

--M. J. Young


The way I see it is, what folks hereabouts call Conflict Resolution resolves conflicts between players about what's going to happen in the fictional world. Task Resolution by contrast dictates rolls for the declaration of intent of certain types of action or event in the fictional world. (That's not an exhaustive characterization of the difference, but there it is.)

So then the way I break it down is:

Task Resolution, Decided Stakes: Traditional RPG combat (where you've got a solid system and you know if you inflict enough hit points the thing dies and you've got the AC chart in front of you and no-one's pretending that everyone hasn't read the monster books, etc., or at least something that approximates this more than it approximates the next.)

Task Resolution, Open Stakes: "I roll to jump the ravine." "OK, roll." The roll is failed. Now, because the GM didn't specify anything about the ravine, he can do just about anything from insta-death to have the character make a mickey mouse dexterity or saving roll to catch a branch a few feet down and climb back up, nothing happens. This is prime territory for Illusionists. I remember back in the late seventies some D&D players used to think it was cheating if you didn't write down everything in your dungeon as DM before coming to play, and this is the reason why, I think: the default to GM-drama.

Conflict Resolution, Decided Stakes: Dogs in the Vineyard, and Heroquest, if it's run 'right' (which is to say, not the way it's expressed in the book in many cases, but I agree with those around here who claim that the system is best suited to this kind of treatment). "I'm-a kill you." "No, but I don't want to kill you, I just want to make it so you ain't doin' no killin' no more."

Conflict Resolution, Open Stakes: The Pool, but maybe only when the non-GM player wins. Then the player can gain a die and let the GM say what happens, or the player can take a die and say what happens. The constraint on the saying can be pretty darn loose. Universalis complications are kind of like this too, I think, where the player wins the massive pile of chips and then busts out with the new facts all over the place.

Is Donjon's 1 victory = 1 fact rule an example of open stakes TR (where someone besides the GM has credibility over what's introduced into the fictional world) or open stakes CR? I don't know offhand.

I think the CR definition in the glossary, as well as in some of Ron's posts, is slightly misleading. I talked to Ron about this once and I think it's misleading for pedagogical purposes rather than because there's anything Ron misses (I add 'of course' for those of you who won't be put off by my deep respect for the man and his work). But. If I were to re-write that definition I'd explicitly add 'between players' after the first 'interest' and substantially rephrase the bit about inanimate objects, since it still seems to me to reflect 'physics of the gameworld' think. If you can't climb the cliff because there's a bad-ass necromancer on top of it and the GM doesn't want you getting at his necromancer without a fight, then he might pick the necromancer's game value as the cliff's value. On the other hand if he stipulated that mountain as the highest mountain on his map and doesn't want you just to brag about how you climbed it without a fight, he can invoke its mountain rating against you too. Either one is CR. It's only when the system whacks you with climbing rules ("whenever a character climbs, roll...") and the player expects and/or the GM feels compelled to make you roll whether he cares about the mountain or not that it's TR.

Because as it stands it's ambiguous between what I take to be the essentials of the local definition and the kind of stuff e.g. S. John Ross talks about when he talks about Conflict Resolution and which has been around for a long time: doing something against another in-game token vs. rolling against your character's abilities alone. Those are both TR, they're just different kinds (or, rather, they're written as TR and then used as TR or CR depending on the GM's (usually) pattern of invoking them).

Josh Roby

John, the benefit of declaring stakes is increased communication and a larger sense of authorship on the part of the players.  It allows them to accurately gauge their chances of success, failure, and other consequences and armed with that knowledge make informed choices about what statements -- strategic, thematic, celebratory, whatever -- they want to make through roleplaying.  It's the difference between the players pushing unlabeled buttons on the black box GM hoping to create something enjoyable and the players collaborating with the GM in clear terms to create something meaningful.

Or another tack (based on your LJ post you referenced): the act of rolling the die creates an authority to which the participants can refer in your post-roll freeform discussion.  This definitionally constrains the discussion, perhaps in terms that nobody is especially interested in (you fell down the bottomless pit!).  Discussing such things before there are any authorities, which means discussing them without reference to the SiS but by reference to player preferences, prevents most of the situations where you get a roll that unavoidably means something nobody wanted.

Sean, I can't disagree more that Conflict Resolution does not resolve conflicts between players.  If I'm running a game using CR and the Big Bad is about to destroy the world and the PCs are there to stop her, I'm rooting on their side.  There is no conflict between the players.  There are conflicts between characters, however, and that's what CR arbitrates.
On Sale: Full Light, Full Steam and Sons of Liberty | Developing: Agora | My Blog


Well, we certainly disagree.

A question for you: If I'm on their side, then why don't I just punt on the roll. "OK, you save the world, cool!"

The obvious answer to that is: 'that would be cheating!' But why? If everyone wants it to go the same way, just say yes, right? Don't roll?

The word 'want' is multifarious here. The evidence for what you want is what you choose to do, and to a lesser extent what you say and feel about what you want. Human beings have desires that can't all be fulfilled in many situations. You want the players to win, yes, but you don't want them to win in a way that renders the victory empty or hollow. And you don't want the second more than you want the first. Which is why you choose to roll. And that choice is what makes it CR. You want them to win but you want them to work for it more.

What makes it TR is that you roll because you wrote/bought the adventure that way, he's got these stats, the combat rules say this, the players declare actions xyz, and so on.  I agree 100% with the glossary definition of TR - the connection between what I'm saying and it is the belief of some designers that 'the system is the physics of the gameworld', which might be a belief that leads a designer to choose TR, or might be a belief that a designer who uses TR by habit comes to by thinking it through without realizing that there are other possibilities.

(Also, even though I'm usually on the player's side, I'm not always, as GM. I really recommend GMing Dogs (though IIRC you have, so I'm interested if you didn't experience it this way) if you want to see this in action in a functional way. You can totally be all, like, 'fuck you, that's not the way it's going to be' in Dogs as GM, be all in like you are sometimes as a player, because the system doesn't force you to pull your punches. Because the player can win the conflict straight-up even though you put the best pNPC stats you had into the character and pulled out all the stops to make the story go your way.)

Josh Roby

Quote from: Sean on December 02, 2005, 03:07:50 AMThe obvious answer to that is: 'that would be cheating!'

The obvious answer to me is that that would be boring.  Hence I use CR because it elaborates the expression of the conflict.

But even in your example, if I want the players to 'work for it' and that makes a conflict, doesn't that assume the players don't want to work for it?  That's not my experience, either.
On Sale: Full Light, Full Steam and Sons of Liberty | Developing: Agora | My Blog