Conflict vs Task Resolution to facilitate Narrativism

[John Kim]

Since whether or not to make a Task Resolution roll is decided based on current conditions as they exist right now...with total disregard for what impact success or failure might have on the A+B+C relationship...using Task Resolution makes Nar play difficult because you're constantly running into situations where a roll winds up making addressing the premise at that time impossible.

Hmmm.  Valamir, as far as I can tell, this is an old argument against dice-rolling resolution -- the idea that a particular result rolled on the dice will destroy the drama (i.e. your A+B+C).  For example, Theatrix approached this by suggesting that there be a step at the start of each resolution: "Does the story require a particular result?"  If so, then you skip the remainder of the process.  Your argument for Conflict Resolution is that by picking the stakes carefully, you can avoid threatening this.  In other words, if a possible outcome would destroy the drama, then you don't roll for it.  

In practice, I found bad task results to not bother me.  In my experience, if I didn't pre-plan particular plotlines, but instead created dense and complex situations, then particular task outcomes almost never harm what I feel is the drama.  If you want to talk about examples of where a particular result threatens to tear apart A+B+C, then I'd be interested -- because I have found it to be exceedingly rare in my games.  (I do generally have some level of script immunity against PCs dying -- this is sometimes built into the system such as in the Buffy RPG.)  An example of where I talk in particular about using Task Resolution is http://www.indie-rpgs.com/viewtopic.php?t=9367">Spirit Tests in Vinland (Jan 2004).  

On the other hand, you also hint at another potential problem, it seems:

The normal course for Task Resolution is to allow the current situation as it exists at this moment in time to detemine what Task Resolution is called for.  It doesn't look ahead to see what the player's ultimate goal for his character is.  If a character is sneaking into a persons room the GM will call for a sneaking Test <insert your game equivalent here>.  It doesn't matter whether the player is sneaking into the room in order to leave a surprise gift and a rose in order to make the person feel happy, or if the player is sneaking into the room in order to leave a bomb in order to kill the other person.  That ultimate goal is immaterial to the task at hand which is sneaking into the room and therefor requires a sneak roll.

Now if you as a GM start casting your eyes ahead..."keeping the conflict in mind" and use that to start picking and choosing which of several possible Task Resolutions to use and which not to...based on that sense of the overall conflict...then you are NOT using "Task Resolution"  you are using text book Conflict Resolution sensibilities to try to adapt the Task Resolution system to better focus on resolving conflicts.

This seems strange to me.  Every game that I've played in has had a distinction about when you need to roll and when not.  i.e. So if you spend the afternoon playing hide-and-seek with some kids, you don't have to make a thousand Perception and Stealth rolls.  If you go rabbit hunting for a week, you don't roll for each shot.  This is explicitly stated in many if not most rulebooks.  The advice varies in what the edge cases are and how do you decide them.  But the gist is that if it isn't important, don't roll it.  

This sounds a bit like MatrixGamer's suggestion that Task Resolution inherently means lots of trivial rolls (roll, roll, roll).  But I thought that was debunked -- i.e. Task vs Conflict is not a matter of scale.

[paulkdad]

I see the argument now. Let me summarize the main points:

[*]Task resolution, when done well, is not arbitrary.
[*]Conflict resolution, when done poorly, is arbitrary.
[/list:u]
Fine. I'll go along with that. It's not much of a comparison, but I won't disagree with the conclusion. But the next point:

[*]It is possible to do task resolution while keeping the conflicts in mind, and this does not change the fact that it is task resolution.
[*]It is not possible to do conflict resolution while keeping the tasks in mind, because doing so would change it into task resolution.
[/list:u]
Sorry. I'm not going along with that one. I stated the perceived double standard as clearly as possible, and I hope it makes sense.

What seems to me looking through the different threads is that there IS no clear meaning to them. The definitions have at all times been spongy and frayed at the fringes.

So, if our understandings differ, we need to discard the entire concept? No. I've read what you've said, and I thank you for the discussion. I've learned a lot about CR and TR by participating in it. But at this point all I see is an attempt to "rescue" task resolution, and personally I don't see it threatened in any way.

[Irmo]

[*]It is possible to do task resolution while keeping the conflicts in mind, and this does not change the fact that it is task resolution.
[*]It is not possible to do conflict resolution while keeping the tasks in mind, because doing so would change it into task resolution.
[/list:u]
Sorry. I'm not going along with that one. I stated the perceived double standard as clearly as possible, and I hope it makes sense.

'Perceived' is the thing here. I never referred to "conflict resolution while keeping the tasks in mind" as task resolution. I believe I gave a quite succinct definition of what's what earlier in this thread.

So, if our understandings differ, we need to discard the entire concept? No.

If we can't come to a common understanding of a theoretical concept, then it has failed in its task. Because that's what they exist for.

I've read what you've said, and I thank you for the discussion. I've learned a lot about CR and TR by participating in it. But at this point all I see is an attempt to "rescue" task resolution, and personally I don't see it threatened in any way.

And all I see is quasi-religious mumbo-jumbo suggesting one has to have passed rites of initiation to even come close to understanding what's supposed to be a THEORETICAL concept but which here means whatever one wants it to mean today in order not to have to revise one's opinion. And uttered in such a fashion that one has every room to backpedal should one have missed the attack. ("Oh, you did that? Well, that doesn't really count, because....")

Sorry. Doesn't fly. The issue is not the rescue of anything, but the verification of a claim that was made. A claim, I might add, for that precious little evidence has been presented. Both John and I have asked (repeatedly now) for concrete examples. All we have been given is hypothetical situations that are practically impossible to arrive in actual play.

When -as was the case earlier in this thread- people refuse to stand up for their own definitions, just to avoid being shown that their own arguments aren't in line with their own definitions, the discussion is becoming increasingly dishonest. Also, I deliberately slept over Ralph's last post, because past midnight last night, I saw it as from the bottommost drawer of "discussion" styles and would likely have posted a rather harsh retort. Frankly, my view of it hasn't improved by much.

John Kim holds a Ph.D. in physics, I hold two master's level degrees, in chemistry and molecular biology. I don't know John personally, but from his credentials doubt he is someone easily baffled by complex theoretical constructs. We both have quite some experience RPing. So if all some of you can come up with is "You just don't understand" it means no more and no less that you've made an extremely poor job at making your case. Which likely means that there is none.

[Eero Tuovinen]

Hey, Irmo: I'm all for good fun, but you seem to be taking the discussion as an argument, and taking the argument very personally to boot. Consider:
1) it's possible that you are right and folks here have been deluded about task vs. conflict resolution. Wouldn't it be great if you could point it out? Why get mad about it? And how will getting all snarky help in convincing us of our errors?
2) this shouldn't be an argument to begin with, and in no case should you be trying to win it. I know, because I'm not arguing, I'm discussing. So arguing back is like hitting an unarmed man, and it's not virtuous at all. So lay back, will you?
3) getting mad over Ralph is like kicking Jehovah's Witnesses on the street. Satisfying, perhaps, but ultimately demonstrating a lack of perspective. Don't you have better things to do?

If we can't come to a common understanding of a theoretical concept, then it has failed in its task. Because that's what they exist for.

Between us, that is. People who can come to agreement might still find an use for it. If you don't, then just don't use it. If you see through the veil of illusion that is conflict resolution, then it shouldn't be difficult to just ignore it for now. Come back to it later with an essay, actual play report or a game design, perhaps. Marshal arguments that point out your position in a way that cannot be ignored. That's much more sensible than trying to argue the matter here, now. (As arguing isn't clearly convincing anybody.)

[paulkdad]

People who can come to agreement might still find an use for it.

Well said. For my part, I view RPGs as an art form, not a science. And it certainly is true that definitions in the arts can look like "quasi-religious mumbo jumbo" compared to definitions in the sciences. But it's not about credentials, backpedaling, dishonesty or rites of initiation. For me, it's about what works for me, for the moment, in my work.

So, even if we could apply the same degree of objectivity and precision to RPGs as is possible (for example) in Physics, would I care? Probably not. This is because, in my opinion (or delusion, if you prefer), artistic expressions are enriched by diversity of thought, not uniformity.

[Irmo]

Hey, Irmo: I'm all for good fun, but you seem to be taking the discussion as an argument, and taking the argument very personally to boot. Consider:
1) it's possible that you are right and folks here have been deluded about task vs. conflict resolution. Wouldn't it be great if you could point it out? Why get mad about it? And how will getting all snarky help in convincing us of our errors?

The critical points have been pointed out time and again. There's no use pointing things out any more when people chose to tell someone who dares to differ with them they have no idea what they are talking about instead of adressing any argument.  And I'm not the one who introduced snarkiness here.

2) this shouldn't be an argument to begin with, and in no case should you be trying to win it. I know, because I'm not arguing, I'm discussing. So arguing back is like hitting an unarmed man, and it's not virtuous at all. So lay back, will you?

Main Entry: ar·gue
Pronunciation: 'är-(")gyü
Function: verb
Inflected Form(s): ar·gued; ar·gu·ing
Etymology: Middle English, from Middle French arguer to accuse, reason & Latin arguere to demonstrate, prove; Middle French arguer, from Latin argutare to prate, frequentative of arguere; akin to Hittite arkuwai- to plead, respond
intransitive senses
1 : to give reasons for or against something : REASON
2 : to contend or disagree in words : DISPUTE
transitive senses
1 : to give evidence of : INDICATE
2 : to consider the pros and cons of : DISCUSS
3 : to prove or try to prove by giving reasons : MAINTAIN
4 : to persuade by giving reasons : INDUCE


Also, discussing requires actually working with the other side's arguments. As far as I can see, that hasn't been done.  A sensible discussion also requires that people stand by their word and admit that when they said A=B yesterday and say A!=B today, they're contradicting themselves, rather than complaining about being measured with their own standards.

Telling people they don't have any idea what they are talking about is NOT discussing.

Between us, that is. People who can come to agreement might still find an use for it.

Only if the agreement is real, and not based on personal sympathy for the person making a statement. As the fact that people denounced their own definitions shows, there IS no agreement.

Come back to it later with an essay, actual play report or a game design, perhaps. Marshal arguments that point out your position in a way that cannot be ignored. That's much more sensible than trying to argue the matter here, now. (As arguing isn't clearly convincing anybody.)

ANY arguments can be ignored if someone is unable to admit to not having a case. I posed concrete questions here, they were ignored. John Kim reported experiences directly contrary to the claims raised here. They were ignored.

[Irmo]

People who can come to agreement might still find an use for it.

Well said. For my part, I view RPGs as an art form, not a science. And it certainly is true that definitions in the arts can look like "quasi-religious mumbo jumbo" compared to definitions in the sciences.

If that were true, there would not be any academies of arts or music.

We know what an adagio is, or a fugue, or pizzicato. We can distinguish cubism from impressionism etc. We can tell a bust from a relief. We can tell romanesque from gothic architecture. We know what a vanishing point projection is.

The POINT of theory is that it serves to communicate concepts. For that, it has to be clear and obvious to someone experienced in the field. Otherwise, it fails.

[Adam Cerling]

If you want to talk about examples of where a particular result threatens to tear apart A+B+C, then I'd be interested -- because I have found it to be exceedingly rare in my games.

A situation is on my mind right now that I think might serve as an example.

I'm part of a Mind's Eye Theater LARP that uses Task Resolution. It's a based on Werewolf: the Apocalypse. Veins of dysfunction riddle the game, as is typical of most MET LARPs; with 20+ players, differing Creative Agendas are almost inevitable. I lean Narrativist, while the GM leans Gamist.

A wonderful opportunity arose for me to address Premise with my character, a werewolf trickster. Tensions have long been running high at our Sept (community of werewolves) between hard-liner and moderate werewolves. Emblematic of this tension is a blatant rivalry between my trickster (a moderate) and a werewolf judge (a hard-liner).

The premise I want to address is that this tension is self-destructive: it will destroy our Sept from the inside. To address this Premise, using Task Resolution, I secretly prepared a way to return from the dead -- and then I goaded my rival into attacking and killing me. Now, according to common player knowledge, my character is dead, when in fact he recovered after he was given a water burial. Let's call this Task A.

To completely address the Premise, I will reveal myself as being alive after they've had time to think. I'll ask the Sept if this is really the kind of community they want to be. Then I will hand myself over to my rival. If he and the Sept have learned from their mistakes, he'll let me live: otherwise, he'll kill me again, and this time I don't have a trick up my sleeve. All this is Task C.

But to get from A to C, I need to give the Sept time for my death to sink in. They need to reflect about how they've come to a place where they kill their own. To give them that time, I need to go undetected, observing them, long enough to choose the right moment. I have prodigious supernatural powers of stealth with which to do this, but I could still get unlucky. This is task B.

If I am discovered before the Sept is ready -- if I fail at task B -- then C is undermined. If the Sept has no time to reflect on my death, my return just looks like a cheap trick. If the pain of what they have become doesn't have time to set in to their bones, my attempt to address Premise fires a blank.

I suppose if the GM were sympathetic to my attempt to address my Premise, he might waive B entirely or interpret a failure in a nonthreatening way. But that doesn't change the fact that B exists only because we use task resolution, and B is also an unnecessary potential point of failure.

[John Kim]

The premise I want to address is that this tension is self-destructive: it will destroy our Sept from the inside. To address this Premise, using Task Resolution, I secretly prepared a way to return from the dead -- and then I goaded my rival into attacking and killing me.

If I am discovered before the Sept is ready -- if I fail at task B -- then C is undermined. If the Sept has no time to reflect on my death, my return just looks like a cheap trick. If the pain of what they have become doesn't have time to set in to their bones, my attempt to address Premise fires a blank.

I suppose if the GM were sympathetic to my attempt to address my Premise, he might waive B entirely or interpret a failure in a nonthreatening way. But that doesn't change the fact that B exists only because we use task resolution, and B is also an unnecessary potential point of failure.

It seems to me that you want to have a particular plot outcome here, which is linear up to the final point.  i.e. You wanted from the start for Task A + Task B + Task C to happen. You must succeed at all three or your plan is broken.  From my point of view, I agree that randomized Task Resolution is destructive to this.  Randomized Task Resolution means that neither you nor anyone else will be able to enforce that the sequence will come about.  

On the other hand, I don't think that being able to mandate such a sequence is necessary for Narrativism.  I would need to know more about the situation, but I would think that it is still possible that a Premise is addressed even if your plan fails.  It seems to me that even in most Conflict Resolution systems, it is possible for your plan to fail.  Nor is the chance of failure necessarily greater in Task Resolution.  

By comparison to my style, you are investing very heavily in plans going your way.  But I think that addressing of some Premise is possible even if you can't address exactly the issue you want in the way that your want.  

So, if our understandings differ, we need to discard the entire concept? No. I've read what you've said, and I thank you for the discussion. I've learned a lot about CR and TR by participating in it. But at this point all I see is an attempt to "rescue" task resolution, and personally I don't see it threatened in any way.

Well, I don't know where you came up with the idea of "rescuing".  I mean, what would it be being threatened by?  The topic for the thread was about whether Task Resolution is inherently worse than Conflict Resolution for facilitating Narrativism.  There were some pretty clear statements that it is by many posters (starting with Christian/xenopulse, Adam/WhiteRat, and Eero).  

Now, as I stated, I generally use and prefer Task Resolution.  However, if I also accept the statements that Task Resolution is opposed to Narrativism, I'm left with a couple logical conclusions: (1) I am not very interested in Narrativism; (2) what I am doing is not Task Resolution; or (3) my techniques actually impede my style of play without my realizing it.  I'm fine with either #1 or #2, but I don't think #3 is true.

[paulkdad]

Sorry guys, but I'm changing my mind here (to some degree). I'm also dumping the issue of "arbitrariness" because I think it is strictly a GNS concern* and is not central to the idea of using TR to address Narrativist Premise. Whiterat presented an excellent example, and while there is only so much that you can infer from anecdotal evidence, in my opinion it scores points for Irmo and John.

The premise I want to address is that this tension is self-destructive: it will destroy our Sept from the inside. To address this Premise, using Task Resolution, I secretly prepared a way to return from the dead -- and then I goaded my rival into attacking and killing me ... To completely address the Premise, I will reveal myself as being alive after they've had time to think. I'll ask the Sept if this is really the kind of community they want to be.

The Narrativist Premise here had to do with addressing the tension of self-destructive behavior. But did this Premise demand a linear A+B+C construct? I don't think so, and this is just enough to make me "convert" on the central issue here. I'll explain why by first referencing a statement from Ron's essay, Story Now:

Narrativist Premises vary regarding their origins: character-driven Premise vs. setting-driven Premise, for instance. They also vary a great deal in terms of unpredictable "shifts" of events during play. The key to Narrativist Premises is that they are moral or ethical questions that engage the players' interest. The "answer" to this Premise (Theme) is produced via play and the decisions of the participants, not by pre-planning.[my emphasis]

While I wouldn't want to create a commandment (Thou Shalt Not Pre-Plan!), this excerpt would suggest that Nar Premise does not demand a predetermined outcome. If this is the case, then why would task resolution make addressing Narrativist Premise more difficult? It wouldn't. The problem is not that task resolution interferes with Whiterat's ability to address an excellent Narrativist Premise, but that the Premise has been addressed through pre-planning, and has been crafted with one right outcome in mind.

It seems to me that part of the issue of whether TR interferes with the ability to address Nar Premise boils down to this: if pre-planning is used to address NP, then task resolution is more likely to interfere.

How to apply the definitions of CR and TR seems to be the biggest source of disagreement. Falling back on the Provisional Glossary is but one source of information. Personally, I prefer descriptive definitions to prescriptive ones, so I'd prefer to see one arising out of usage. Correct me if I'm wrong, but usage here defines the difference between TR and CR in terms of Stake. Are you resolving what's being done or are you resolving what's at stake? Admittedly, to the degree that "what's being done" and "what's at stake" overlap, this definition may create more problems than it solves. I take it for granted that usage may eventually render these terms obsolete. If this discussion is just about nailing down prescriptive definitions of TR and CR, then say so and I will bow out now.

So, getting back to the topic, I think it would be helpful to focus on Irmo's idea, and forget Daredevil's initial post. If I say, "Prove to me that TR does not facilitate addressing Nar Premise," then not only am I asking for a negative proof, but all I am going to get is a lot of anecdotal evidence, which I will refute with my own anecdotal evidence. That leads nowhere. On the other hand, I could paraphrase Irmo's question to read, "Theoretically, under what conditions would TR interfere with one's ability to address Nar Premise?" and that definitely leads somewhere.

So, Whiterat, do you still think the problem was with task resolution, or was it a problem of pre-planning?

*Logical outcomes are always in relation to the Premise. Applying Narrativist logic, decisions that derail the story become arbitrary. Applying Gamist logic, decisions that trivialize the system become arbitrary. Applying Simulationist logic, decisions that shatter the illusion become arbitrary.

[Valamir]

Paul you're confusing yourself.  Whiterat's example while well intentioned is NOT in anyway an illustration of either narrativist play nor why Conflict Resolution is beneficial.

Conflict Resolution is not prefered because it generates a Linear Outcome.  Far from it.  Any sort of linear outcome is completely anti Nar and Conflict Resolution is specifically designed to not permit it by making "what's at stake" determined randomly.

Also don't allow yourself to be confused by the careless way certain posters have started using A+B+C as if that were a linear track.  I linked to Vincent's Blog so that people could actually read it.  A+B+C is simply a short hand notation for Issue + Character + Situation.  Not Event #1, Event #2, Event #3.

This is what it boils down to:

Current State:  A+B+C.  Thematically charged situation.
Conflict Resolution:  Conflict may be resolved 1 way, it may be resolved another way.  No matter which way its resolved it either preserves, resolves, or leads to another Thematically charged situation.  It does this because "what's at stake" was defined so that no matter which way the dice fall premise is addressed.


Contrast this to:



Current State:  A+B+C.  Thematically charged situation.
Task Resolution:  No effort is made to make sure that the results of the resolution preserve, resolve, or lead to another Thematically charged situation.  Task Resolution is concerned only with identifying a particular action at a point in time that is deemed difficult enough to not be automatic and require a task roll.  The result of that roll MAY preserve, MAY resolve, or MAY lead to another Thematically charged situation.  But, since the roll was not engineered with that purpose in mind, it also may not.

Task Resolution MAY wind up eliminating the Issue (A).  e.g. the Issue is whether the captain will choose to return to port to save his wounded friend in violation of his orders or risk his friend's life while completeing his mission and pursuing the enemy ship.  The friend is seriously wounded.  The logic and causality of Task Resolution demand he make a "health check" or "saving throw" or whatever.  The friend dies.  No more issue.  The captain now doesn't have any choice to make.

Task Resolution MAY wind up eliminating the Character (B).  The issue is to confront ones brother about his evil ways and give him one more shot at redemption.  In a fight with a guard the hero dies before ever approaching his brother or confronting him.  No more character.  The issue doesn't get resolved (yes he's redeemed, no he isn't) because it never even happens. The involved parties simply die before they find out.

Task Resolution MAY wind up eliminating the Situation (C).  Returning to the Captains choice issue.  The captain must choose between his friend and his mission and the situation is that the enemy ship is about to escape.  The rules of the game call for a "sailing check" or the like whenever a ship is trying to escape.  The GM abides by that rule and certainly by the logic of causality the situation warrants it.  The enemy ship fails the check and suffers a mandated critical hit.  It loses a mast, gets thrown in irons and is now helpless.  Boom, no more thematically charged moment.  The captain no longer has to choose between getting his friend to port or pursuing the enemy for an extended period during which his friend will almost certainly die.  Because the Task Resolution system just rendered the enemy ship helpless.  The captain can sink or capture that ship with minimal delay and still get his friend to port in time to save his life.  The whole premise of pitting Friendship vs. Duty has just been shot to hell


Now from a SIM perspective...none of those outcomes are particularly bad. They are all logical, they are all reasonable, they don't in anyway violate the dream.  If the potential thematic issue just shriveled up and blew away...so what...Sim play views thematic issues as a nice bonus when they happen but not the primary point of play.

From a Nar perspective those outcomes just crapped all over the entire purpose of why those players were sitting around the table playing that game at that time.  It MAY not have happened like that.  The friend may have survived his health saving throw thereby preserving the issue.  The hero may have survived the guard and gone on to confront his brother.  The enemy ship may have escaped again leaving the captain to choose between friendship or duty.  But there's no guarentee of that.  Using Task Resolution for Nar play leaves you perpetually in danger of having the wrong die roll at the wrong time dismantle the entire purpose of the game.

Conflict Resolution doesn't work like that  In Conflict Resolution you set the stakes specifically to capture the key issue "What's at Stake is do I capture the enemy ship in time to save my friends life"  If you manage to resolve that to the positive than you accomplished both yes...but you also addressed the premise.  The captain chose Duty over Friendship.

If the stakes were resolved to the negative there are a couple of possible outcomes.  Different CR systems will deal with them in a different way, some like PTA or Dust Devils will leave it entirely open to narration as to how the stakes were lost.  Others like DitV will fill in the nature of the loss through the various raise and see steps.


I submit that this is probably the closest thing to a Nar litmus test you're going to find.  If you can imagine yourself in one of the above examples, and not minding that the premise just got dismantled because the roll that was made was perfectly reasonable and appropriate and you'd have been more upset if the GM had skipped or fudged it...than you probably aren't playing Nar.

Does that make the critical difference between the two any clearer.

[Eero Tuovinen]

Paul: I for one am just a little bit frustated by this conversation, because it seems to me that my arguments are just not being considered in any way. Perhaps I'm just so stupid that I'm talking besides the point, but in that case I'd like it if somebody explained that to me. It does, however, seem to me that my take on the matter should be sufficient to answer the questions bandied here. At least, I don't garner any illumination from what others are writing.

In the interest of clarity, what about if I restate my points in simple and straightforward form, without any reasoning or trying to slot the issue into the discussion so far? I think I'll do that, perhaps it helps somewhat.

Eero talks about conflict vs. task resolution

So-called "conflict resolution" and "task resolution" are illusions, insofar as discrete game systems are concerned. There is no true conflict or task resolutions, or if there is, they are so vanishingly rare that it doesn't matter for our purposes. Instead, what systems do have is the quality of "resolving tasks" and the quality of "resolving conflicts". The thing is, all rpg systems have these qualities, and thus all rpgs have both "conflict" and "task resolution".

How is that possible? Remember, "system" for our purposes is the means we use to introduce stuff to the SIS. (Check those terms from the Glossary, if they're not familiar! You won't understand this otherwise!) Any method of resolving both tasks and conflicts is part of system. Now, do we all accept that in narrativist play, both conflicts and tasks exist? If they do, what happens to them? I mean, if there are conflicts and tasks, what needs to happen for play to go on? That's right, the tasks and conflicts have to get resolved. Play won't progress if both tasks and conflicts are not resolved. (Or rather, narrativist play won't. You could have play in which no conflicts are resolved, but that won't allow addressing Premise.)

Assuming that we have narrativist play, we can also assume that tasks and conflicts are both getting resolved. How does it then happen in practice? I direct you back to what I wrote earlier: there are three conceivable approachs a given game design could take to narrative conflict resolution.
1) having a conflict resolution system, but being rules-light about task resolution (Dust Devils, PTA, Fastlane, most modern formalistic nar games)
2) having a task resolution, ensuring that the tasks carry conflict significance (Sorcerer, TRoS)
3) having both task and conflict resolution side by side (Dogs in the Vineyard)

As you can see, there are various ways of doing this. However, before we talk more about them ways, I'll have to define 'conflict' and 'task' somehow, so you understand what I'm talking about:
Conflict: a decision-point in a narrative where the different ways the story could go carry thematic meaning. In nar terms, the conflict is a vector for a Premise, and resolving it answers the premise.
Task: events of a story, in general.

Now, many people seem to define these terms in different ways, so you might not agree with me. Significantly, they are definitions based on the narrative qualities of play, and not on system matters at all. I think that the above definitions are similar to those fine definitions in their narrow bands of consideration, and are furthermore the only logical (meaning, true) way to define the distinction in general. Neither conflict or task hold any meaning for rpg theory unless they're successfully differentiated and the meaning of conflict vis-a-vis narrativist play explained.

Understanding those definitions is crucial. Note that they hold for traditional narratives as well, if you wish to use them. The key property of those definitions is that while tasks happen because of SIS congruity ("realism", in other words), conflicts are caused by the artist (players) perceiving premise-weight in a task. That premise-weight is our problem, and the reason for the need to differentiate between task and conflict resolution: to create a good story (that is, to play satisfying narrativism), you need to be able to create and resolve those conflicts. A story is good if it has good conflicts, it's that simple. Conflict resolution rules, as they are called, allow us to introduce and resolve conflicts, and thus create a good story. Without conflict resolution rules there is no guarantee that conflicts even happen, much less that they get resolved.

Short sidetrack: I'm saying above that conflicts are just tasks with Premise-weight attached. This means that "winning the fight" or "impressing the lady" or whatever are always only tasks, and become conflicts only when they have that meaning. Like, "winning the fight to end the war" or "impressing the lady to love her and leave her" are conflicts. This is probably clear to you all, but I emphasize it anyway: the distinction between conflict and task is not about f***ing scale or stakes or anything at all, it's about story meaning only. Every task can potentially carry conflicts, and no task is always a conflict task. (Hmm... protagonist suicide is a conflict in almost all cases...)

Now, back to practice. As the keen mind should already understand, the above definitions mean that resolving tasks and conflicts are separate issues. They are not totally meaningless for each other, because conflicts are ultimately resolved (that is, brought to SIS) through narration of task resolution, but neither are they connected in any strong sense. A game could have a little bit of conflict resolution rules and a whole bunch of task resolution, or vice versa. It's a matter of degree, not of classifying games into two categories of method. This I will now prove through exactingly detailed examples:

Primetime Adventures

How does PTA resolve tasks and conflicts? First, most tasks are resolved through simple narration, the player just says how he wants the task to resolve. He can resolve success, or he can resolve failure. only when other players disagree is any more complex system used. Still, that also is a task resolution system, even when it's about as simple as could be.

PTA assumes that players will never disagree about a task unless success in the task predicates success in an accompanying conflict. In other words, characters always succeed in tasks, unless there's a conflict in there. The point about PTA task resolution is this: when a conflict is recognized, task and conflict resolutions are combined:
The task resolution part: before anybody knows what's going to happen, the players can opt to use edges, which represent task resources of the character. This corresponds with "choosing the appropriate skill" in traditional task resolution systems. The player chooses if he wants to use edges, and is responsible for only using them when he could conseive a task narration where the edge is used.
The conflict resolution part: all the rest of the PTA conflict prosedure concerns itself with the conflict: it's recognized, the stakes under conflict are defined, and so on. The conflict is resolved, the player with the most successes wins. The player with the highest die narrates, which becomes important soon.
The task resolution part second: task resolution proper happens only when the conflict winner is already chosen. What's more, the task resolution is handled through a method very common in formalistic nar games: the narrator chosen by the conflict resolution has free hand in deciding whether any given task in the situation succeeds. This is task resolution. Some games have only "GM decides" task resolution, and it's still task resolution. It's task resolution in PTA, too, even when the main attention is turned to the conflict winner.

PTA has a Drama-based task resolution system, insofar as a given narrator decided any task based on it's merits. On the other hand, the conflict resolution system is Fortune-based. The two are affected by each other, if in no other way then in the fact that tasks in conflicts are resolved in a little different way than tasks outside of conflicts (different narrator). Point: PTA has relatively lots of rules for conflict resolution, and few rules for task resolution. Regardless, both get resolved.

Sorcerer

(Let me note that IMO Sorcerer is extremely complicated and unique among narrativist games, so I'll probably hash this analysis.)

In Sorcerer task resolution is paramount mechanically. Everything you'll ever resolve in Sorcerer is tasks. (Humanity rolls are uncertain about this, due to different interpretations of what the rolls mean.) Look at what you're doing in that game: it's very much about whether the character succeeds in a given task or not, and about the tools he uses (Stamina vs. Cover, for instance). How are conflicts resolved, then?

Sorcerer has a play structure and emphasis on situation that tends to create conflicts. For example, the demon relationships are full of conflicts, and the GM is instructed to create more conflicts (each Bang is a conflict by the above definition). This makes it extremely likely that the task resolutions also resolve conflicts. What's more, the GM is instructed to encourage conflict resolution through not preplanning adventures and driving play towards tasks with significant stakes. Functional Sorcerer play very much interprets these GM responsibilities as rules to follow, because otherwise the rules are comparatively easy to bend towards other priorities from nar (as has been proven by Sorcerer's fame as a difficult game to play correctly).

So Sorcerer is a nar-faciliating game due to the way it invests task resolution with conflict resolution. The system does double duty in this regard. The very same task resolution system is used whether there's a conflict there or not (temporary bonuses might differ in practice, though). Point: Sorcerer has lots of rules for task resolution, and little rules for conflict resolution. Specifically, it's conflit resolution rules are the kind that people are not used to following (disguised as GM advice). So you could say that Sorcerer uses task resolution. Still conflicts get resolved.

Dogs in the Vineyard

(I already wrote about Dogs, but because it's the only example of concurrent task and conflict resolution I can think of, I'll include it again. Probably there are lots of these, starting with friggin' Amber or something, but I can't care enough to think about it right now.)

In DiV there are two different ways of resolving tasks:
1) any character task succeeds, unless another player wants to contest the success. This is the rule called "Say yes or roll the dice".
2) If a task is contested, a conflict is tructured out of it. Like in PTA, it's assumed that nobody wants to contest unless there's a conflict. During conflict, a player may narrate any task success for his character, assuming that
a) he raises during it, or reverses/blocks/takes the blow, and the task is compatible with the requirements of the rules action in question.
b) the other player won't reverse or block, if it was a raise.

As you can see, the second task resolution method is rather more intricate. What's more, during conflict those same actions that are used to resolve task resolution are used to resolve the conflict itself. The two systems are separate, but connected, and used concurrently. What's more, there are multiple levels of task and conflict resolution going in a very intricate manner. For example, damage: tasks may cause damage, the amount depending on narration, but damage may also hold conflict meaning, even for a conflict completely disassociated with this one. It's frequently difficult to really distinguish between task and conflict resolution in DiV, but ultimately the two are different, and the connections are all crafted by the designer, not grown in any natural manner.

--

Before returning to theory, let's also check out traditional solutions. This is important, because the above games (even Sorcerer) strive to resolve conflicts in above-the-board manner, with the players negotiating with some degree of No Myth thinking. However, there are radically different methods of conflict resolution in traditional games, games that are typically called "task resolution systems".

Firsty, one typical way of conflict resolution is to try to disarm conflict beforehand. (To understand this, it's imperative that you remember what conflict we're talking about here.) D&D does this to a great degree: the prime directive for the GM is to build a functionally cooperating party and to encourage out-of-game negotiation of any disagreements about story stuff (will we sell out to the evil overlord, for instance). This is intentional disarming of conflict.

Another important point is controlling player ability to initiate conflict. This is done on the social level pretty frequently (meaning, the GM decides what adventure comes next), but games have rules for it, too... at least they should... criminy, it seems I can't think of any examples. Wonder why is that? Anyway, the main way of controlling ability to initiate conflict is the ability to control scene framing and adventure choice. Pretty impossible to make a bid for the overlordship of the city when the adventure of the day concerns looting orc caves.

If conflict happens, and it does because some games want it, a typical approach is for the GM to resolve the conflicts through illusionistic technique. WW games are like this, pretty much. (A pre-resolved conflict, by the by, means preresolved premise, and thus theme. That's why these games can be characterized as theme-focused sim, in case you didn't know.) Such "illusionistic techniques" mean mainly control of task initiation (I demonstrated this earlier in the thread) in it's different forms, really. Stake definition plays a part, too. All of these games leave defining stakes for the GM, and he is instructed to do it only after he knows the results, which pretty much means that the conflicts always go the GM's way.

The point of the above paragraphs is to note that these games have conflict resolution rules, too. They're just pretty undefined, and predicated on the role of the GM. This is natural for a game that doesn't strive to give players conflict control. The only point here was to drive home the idea that all games have conflict resolution systems, in one way or another.

--

At this point, let's go back to the theoretical question. We have demonstrated that separating rpg systems into conflict and task resolution classes is a faulty and inexact practice. On the other hand, it should also be clear that tasks and conflicts can be usefully differentiated, and furthermore that the two use very independent means of resolution. These are prime points of understanding in themselves, and should help in laying this thread to rest.

However, more angles: even if we agree that there's no point in talking about solely task or conflict resolving systems, it's still true that we can talk about system leaning in this regard. We could say that a system uses "conflict resolution" if the central means of resolving conflicts is mechanical (or even if such a mechanical option was available to the players). We could say that a system uses "task resolution" otherwise. That's fine with me, I can live with that.

But when we have cleared up what a conflict resolution system actually is, the original question of the thread remains: why, exactly, conflict resolution supports narrativism? Keen minds have already deduced the answer, but let's lay it out anyway:

To faciliate addressing Premise (the definition of a narrativist game), the game has to make sure that
1) conflicts (premise-answering points of narrative) are formed and
2) conflicts are resolved.
This is because of how Premise is defined: if you read Ron's essay about it, you'll note that Premise itself is defined as a question that hypothetically comes to being on some symbolical level when fictional situations are considered. (I'll assume here that nobody doubts the existence of Premise and narrativism itself, so we get somewhere today.) To be exact, Premise is the question that is being answered by the events of the narrative. Now, to have the answer, you need degrees of freedom in the narrative. There has to be points of the story where it could go the other way, because any other kind of point isn't an answer to the Premise. Point: this is how we defined conflict to begin with. Conflict resolution is addressing Premise, and addressing Premise is conflict resolution, because conflicts are simply points at which Premise is addressed. Not necessarily the same Premise all the time, but still some Premise.

Now, in the view of the above, it's quite reasonable to claim that I'm full of shit about addressing Premise, because actually Premise is frequently addressed outside conflict resolution. Most examples of addressing Premise come from outside conflict resolution situations, when you think about it. "Will I kill him?" is a typical Premise-situation you could expect to see, and in almost any system there won't be conflict resolution here. What gives?

The answer to the conundrum is that actually, conflict is being resolved. The great majority of rpg conflicts are internal, and the great majority of rpg rules systems resolve those conflicts through free player choice (MLwM is a strong exception). Only external conflicts are usually resolved mechanically. However, this doesn't affect the nature of the conflict, or nature of Premise-answering. It just illustrates what we already know, that rpgs are a different type of narrative art from the classical type. Internal and external conflicts are different, and are dealt differently, because of notions designers have about Character. It might be that rewarding roleplaying in the main requires player autonomy over character, but that's not our topic today.

OK, now we have our answer: reliable conflict resolution is necessary for narrativist play because narrativist play is conflict resolution. Anything that diminishes the reliability of conflict resolution in a game is anti-nar.

The Point

My point to all of you: "Task resolution can be just as good for narrativism as conflict resolution" is IMO not even a sensible claim, considering what those terms mean. What you can demonstrate is that task resolution can be used to resolve conflicts. This only means that your system has a conflict resolution that uses task resolution as a method. Sorcerer does it, it's nothing weird or unexpected. Whether such a system is better than another kind of conflict resolution (like, say, PTA) is a complex design question, and not about theory at all: it depends on what your specific goals are in the design, and the specific forms of task and conflict resolution you're considering.

[xenopulse]

There's little I have to add to Ralph's and Eero's impressive analyses, except for the following:

Look carefully at PtA as an example of conflict resolution. Even if you resolve a task with the rolls, the rolls are not based on the difficulty of the task, the circumstances, or any other causal relations--the rolls are based mainly on screen presence (importance of the character for the current episode) and fan mail (feedback mechanism from other players). With screen presence ranging from 1-3, the edge, which might be coinsidered a skill as Eero wrote, might only have as little as 25% influence; add fan mail dice, and the skill relevance diminishes significantly.

Compare and contrast this with regular task resolution games. In D&D, the rolls are based on skill and difficulty mechanisms that model something. They are not influenced by how important the character is, how much interest the player has in that conflict, the feedback other players give, or the character's stake (contrary to TRoS' spiritual attributes).

So, while Eero is right that all games somehow address tasks and conflicts, there is a fundamental difference that I see between the basic design approach to task or conflict resolution regarding the facilitation of Nar play, which is what we're talking about, after all.

Mechanics that model a resolution system on causality/realism or challenge/play balance support Nar play less than those that model it on player and character involvement. The first two are usually focused on resolving tasks. The latter is usually focused on resolving conflicts. Not a 100% corelation (see Eero's points on Sorcerer), but close.

(edited for typo correction)

[MatrixGamer]

While I wouldn't want to create a commandment (Thou Shalt Not Pre-Plan!), this excerpt would suggest that Nar Premise does not demand a predetermined outcome.

Pardon my asking, but as a novice, does the narrative agenda require a set outcome or does it call for a set plot? A set outcome would be, well... just that. At the beginning of the game I as GM would know what the players would find regardless of what they do. I done this in games ans it works but it feels forced. If the agenda instead picks a plot then the players merely have to accomplish some very broad steps to reach a plot outcome, it doesn't matter what the outcome is if the steps have been made.

I use this second approach in Matrix Games. If the game is a murder mystery, the the game starts with a description of a crime scene. The plot supposes that they find clues, figure out who could have done it, arrest them and put them on trial. It doesn't matter who they pin the charge on as long as it flows from the clues they make up in the game.

Are there any old threads that will answer this question? (You can see Ron has been molding my behavior - I'm anticipating the reply that this has already been discussed so I'm just cutting to the chase and asking what the thread is.)

Chris Engle
Hamster Press

[John Kim]

Task Resolution MAY wind up eliminating the Situation (C).  Returning to the Captains choice issue.  The captain must choose between his friend and his mission and the situation is that the enemy ship is about to escape.  The rules of the game call for a "sailing check" or the like whenever a ship is trying to escape.  The GM abides by that rule and certainly by the logic of causality the situation warrants it.  The enemy ship fails the check and suffers a mandated critical hit.  It loses a mast, gets thrown in irons and is now helpless.  Boom, no more thematically charged moment.  The captain no longer has to choose between getting his friend to port or pursuing the enemy for an extended period during which his friend will almost certainly die.

Ralph, I know you don't think this has to do with pre-planning, but it certainly seems like it to me.  As I see it, here you are depending on a very specific situation to be set up or your thematic charge is destroyed.  The friend has to be gravely wounded enough to require land to survive but not so far gone as that recovery is impossible; and furthermore the enemy ship has to be just barely out of reach -- possible to catch but only if many days are taken in the chase.  My reaction to this is to blame the narrow planning rather than the Task Resolution.  

Yes, randomized Task Resolution will dismantle these sorts of plans.  They will also create opportunities which you hadn't planned on.  Using it successfully in-game means running with what happens rather than sticking with the plan.  In each of these cases, the thematically-charged moment you have plan is derailed, but that doesn't eliminate the possibilities which are offered in what does happen.  

Task Resolution MAY wind up eliminating the Character (B). The issue is to confront ones brother about his evil ways and give him one more shot at redemption. In a fight with a guard the hero dies before ever approaching his brother or confronting him. No more character. The issue doesn't get resolved (yes he's redeemed, no he isn't) because it never even happens. The involved parties simply die before they find out.

I generally play in low-mortality games where PC death essentially has to agree to.  However, that is not because mortality lacks thematic charge.  Your example was a hero who is killed by a guard.  Realism may kill the character -- but realism also provides a lot of fallout from that event.  When a real, living character dies there are issues galore which comes out of that.  In your specific example, I would pick up with the hero's friends who when executing his will and funeral realize that they need to contact his brother about the death.  So now we don't have the planned-for brother-to-brother talk, but we instead have the funeral scene.  

Paul: I for one am just a little bit frustated by this conversation, because it seems to me that my arguments are just not being considered in any way.

My point to all of you: "Task resolution can be just as good for narrativism as conflict resolution" is IMO not even a sensible claim, considering what those terms mean. What you can demonstrate is that task resolution can be used to resolve conflicts. This only means that your system has a conflict resolution that uses task resolution as a method. Sorcerer does it, it's nothing weird or unexpected. Whether such a system is better than another kind of conflict resolution (like, say, PTA) is a complex design question, and not about theory at all: it depends on what your specific goals are in the design, and the specific forms of task and conflict resolution you're considering.

Eero, as I see it, the problem is that you've introduced yet another set of definitions into an increasingly crowded field.  Also, I'm not sure how I should take your arguments, since I can't tell what you're arguing for.  As far as I can see, you've punted on the conclusion which most people are arguing.  You state that the choice for Narrativist games -- i.e. between (1) resolving-conflict-through-task and (2) resolving-conflict-through-conflict -- is a complex question that you don't address.  

To a degree, this can be seen as arguing against arguments by Ralph/Valamir and Christian/xenopulse that Task Resolution is inherently inferior for Narrativism.  You apparently disagree, claiming that it may be inferior or superior depending on the details of the design.  However, the general tone seems to be arguing against myself and Irmo.  

So, while Eero is right that all games somehow address tasks and conflicts, there is a fundamental difference that I see between the basic design approach to task or conflict resolution regarding the facilitation of Nar play, which is what we're talking about, after all.

Mechanics that model a resolution system on causality/realism or challenge/play balance support Nar play less than those that model it on player and character involvement. The first two are usually focused on resolving tasks. The latter is usually focused on resolving conflicts. Not a 100% corelation (see Eero's points on Sorcerer), but close.

Ack.  So this is taking it from yet another angle -- the question of the inputs, which as you say doesn't directly translate to Task/Conflict.  I'm glad to talk about this, but it is yet another can of worms.  Can you give an example of how causality/realism hinders Narrativism?