Friday, September 29, 2006

Conflict and Task Resolution

I am going to write a post called The Ideology of Conflict, and planned this discussion of conflict resolution and task resolution as an aside in it. But I believe its length warrants making a special post out of it, so here you are. The thesis I will defend is this: Conflict Resolution and Task Resolution are useless as analytic terms.

I suggest you take a few minutes to reread Vincent Baker's old post on these two types of resolution. A small quote:

Which is important to the resolution rules: opening the safe, or getting the dirt? That's how you tell whether it's task resolution or conflict resolution.

Task resolution is succeed/fail. Conflict resolution is win/lose. You can succeed but lose, fail but win.

The distinction between CR and TR has been important, historically, because it made people rethink the way roleplaying games worked. But we are now in a position to see clearly that CR and TR are, in fact, identical. There is no difference between the two. As analytic tools, they are useless.

Distinguishing TR and CR went like this. First, you ask what the character is attempting to do. This is the task. Then, you ask why the character is trying to do that; or, almost equivalently, why the player is trying to have the character succeed at the task. This is the stake of the conflict.

Then we define: TR is about resolving the What. CR is about resolving the Why.

So, TR and CR are different precisely in so far as we can distinguish between a What and a Why, between an action and the goal of that action. But when we think about this some more, we will see that this distinction breaks down completely.

The character attempts to open the safe. Why does the character attempt to open the safe? In order to get the dirt on the villain. Ah - so in attempting to open the safe, the character is attempting to get the dirt on the villain. Why does the character attempt to get the dirt on the villain? Because he wishes to blackmail the villain. Ah - so in attempting to get the dirt on the villain, the character is attempting to blackmail the villain. Why does the character attempt to blackmail the villain? Because he wants the villain to release the character's little sister, whom the villain has kidnapped. Ah - so in attempting to blackmail the villain, the character is attempting to free his sister. Why...

You will see the point by now. The structure of task and conflict is not that of a simple duality, but that of an infinite regress. Every action (the What) points towards a larger goal (the Why) which gives the action its value. But this goal immediately furnishes a new description for the action, thus becoming a What. This What points to a new Why. And so forth.

When do we have Task Resolution, and when do we have Conflict Resolution? It would be completely arbitrary to say that the task is opening the safe, and the conflict is getting the dirt on the villain. We could just as well say that the task is to get the dirt on the villain, and the conflict is to free the character's sister. We could just as well say that the task is to make a set of complicated movements with the characters left hand, and the conflict is to open the safe. There is no natural division.

But what if we state the whole thing on the level of the player, instead of the level of the character? (This obviously does not help, since the logic of the situation stays the same; but we will discuss it in order to make things clearer.) TR is about things the player does not care about except in so far as they point to something else; CR is about things the player cares about for themselves. The player doesn't care whether he opens the safe or not, but he does care about whether he gets the dirt on the villain or not.

The counterargument stays just the same. Perhaps it is true that the player only cares about the action (the task, the What) because it is a means to a goal (the Why). But this goal can itself be seen as a What, which points to a further Why. The player does not care about getting the dirt on the villain, he cares about freeing his sister. He does not care about freeing his sister, he cares about being able to think of his character as some who looks after those he loves. And so on, and so forth. The only way to stop the infinite regress is to postulate at some point the "Final Care" of the player. But then the only possible conflict would be whether the player gets his Final Care or not, and at that point the fullness of his caring would reveal itself as total emptiness.

Anything valued points beyond itself, to a value. But that value, being valued, also points beyond itself; and so on. The character of value is transcendence.

So far for a philosophical backing; let's return to game design. If, as I maintain, there is no difference between task resolution and conflist resolution; if any example can only be about choosing a different level of concerns as the one on which resolution takes place; then what was the use of the distinction?

The distinction between TR and CR made people aware that there are different levels of concern on which resolution can take place. This led people to reconsider the level they were using, and finding out that often, it was not the most satisfactory level. It allowed us to see that small physical actions were not always the best level.

That is all. It was enough. It was great. But now we can put those two terms to rest, because the distinction they tried to mark does not exist.


  1. Hello Victor,

    Your reasoning makes sense, but I'm a bit afraid that we're forgeting something on the way.

    It's got to do with the time-scale that's immediately interesting to the player (what he currently is concerned with, in an effort to create an engaging and credible narrative) and getting at what really matters to the player.

    The problem with Task Resolution was that it could be used by a Forceful GM to give the illusion of resolving something, when the really interesting result (say at the level of the scene (timescale)) was completely disconnected.

    "Do I get past that guard?"
    "yes, but you then discover that there are ten others farther on"

    If the player was interested in moving to something else than just stealthing around, he will be disappointed, bigtime.
    And since in a lot of classical rpging, you only take decisions from the character's immediate point of view, it was often difficult to express clearly what one wanted as a player.

    (Of course, if the game is about sneaking past guards and that each instance is interesting to play out by itself, then that's cool.)

    For me, the distinction is useful in so far as it distinguishes "what is important to the player at the current timescale" from what isn't. (That's why Bringing down the Pain is interesting: it lets you zoom in on a particularly interesting part.)
    You're of course completely right that a given fictional event can be a Task or Conflict, depending on the point of view.

    As long as this idea comes along, I'm fine with chucking the terms.

  2. Hi Christoph,

    I absolutely agree with you. I guess there is a double point in your remark.

    1. Conflicts should be chosen at the right scale. Resolving conflicts that are too small (sneaking past a single guard in a game where there are hundreds and the sneaking is not particularly interesting) leads to dissatisfying (because boring) play.

    2. If the player wins a conflict, he should be able to enjoy the victory. For instance, if he successfully rolls for sneaking past the guard, the GM should not immediately give him an even bigger problem to confront, making his success meaningless.

    Both of these are important, and the CR/TR destinction helped us to see them.

    But 'Conflict Resolution' is not a new type of resolution that allows us to always achieve these aims. If I roll for sneaking into the castle instead of rolling for sneaking past this single guard, then (1) the scale is still too low if I have to sneak into 50 castles, this session; and (2) the GM can still take away the taste of victory by saying "aha, and you do not find the grail in this castle!" [evil laugh]. "Then I want to roll whether or not I get the grail!" [Sucess.] "Ok, you find the grail, but King Arthur is no longer interested in it!" "Damn you, I'm going to play with someone else."

    So 1 and 2 are important, but you don't achieve them by adopting a new kind of resolution. (Here, there is no new kind of resolution to adopt.)

  3. Hi,

    I'd like to add that CR and TR might not be so useful to distinguish methods employed in a given instance of play, but rather to look at a set of written rules and what they convey about the usage of the rules in a game.

    Even then, I concede the difference is small, but on a cognitive level, it helped me understand why 'another guard or one hundred' is frustrating, and how to communicate goals and interests clearly as a player and GM.

    Just because a lot of people in the forge diaspora and indie environment are so used to CR and TR, and used to employ implicitly in TR what is an explicit part of CR does not make that definition useless for people new to these concepts.

    Or maybe you just mean it's about time to explore and develop other parts of rpg rules sets and methodologies, because these parts are reasonable well charted?

  4. Wow, I completely agree with this and wrote nearly the same thing on the Forge (see Reply #8 in this thread:

    The only meaningful distinction that I came to between conflict and task resolution centered on when a player must roll the dice. For task resolution, it's when the character's competency is low relative to the complexity of the task at hand, while conflict resolution has the dice kick in when there is a conflict of interest.

    In the example of the safe, dice are rolled, but for different reasons, depending upon whether the game uses task or conflict resolution. Under task resolution, they get rolled because the safe is tough to crack. Under conflict resolution, they get rolled because they might hold the dirt on the villain. By the way, a player's narrative power (i.e., ability to say whether the dirt is in the safe even if the GM doesn't have it there) is a whole separate issue.