Task resolution/Conflict resolution

[Grover]

I was thinking about the distinction between task resolution and conflict resolution, and noticing how conflict resolution is never used in computer role playing games, and here are some ideas I came up with, which I put forward to find out if I'm on to something, or off my rocker.

 First, it should be noted that the distinction between task and conflict resolution is a fuzzy line, and a matter of taste.  There are far more than 2 levels of abstraction.  For example, consider the following hierarchy:  Find an artifact > Explore a dungeon > Defeat the orc tribe on the second level > Ambush an orc patrol > Make a Hide check.  I would categorize something into conflict vs. task resolution based on whether the group is primarily interested in the outcome on it's own, or if they are primarily interested in it as an aspect of resolving something larger.

 So, supposing that you are doing task resolution - why do it if everyone is primarily interested in whats happening at a higher level?  I see these reasons:
 1) Boardgame style fun.  In a combat situation, this is the fun of trying to find a tactically sound solution to the current problem
 2) A chance to highlight some attribute of your character.  I would break this down into 2 subcategories:
   a) A chance to show off a skill or ability.  i.e. look how strong Thrudd is!  He killed 3 orcs in a round
   b) A chance to show a relationship or attitude.  Does your fighter consider it more important to guard the mage, or pursue the fleeing orcs.  You can also show how much you character hates something by engaging in overkill (I flamestrike the kobold)
 3) An opportunity to automatically add color.  The conflict resolution of a fight will tell you which side won, and maybe how badly they were hurt, but it won't tell you that someone risked their life to take down the orc leader, or that someone ran away to drink a healing potion, or any other odd bit of detail that could be thrown in.

 I was thinking about conflict resolution in terms of computer games, and I think the third reason is the compelling one in that arena (although the first one is also important in computer games).  I haven't been able to think of a reasonable conflict resolution mechanism which is suitable for a computer game.  Does anyone else have any ideas?

Steve

[John Kim]

I was going to make a new thread on conflict vs task resolution, in reply to Simon Hibbs, split from http://www.indie-rpgs.com/viewtopic.php?t=13331">The state of play, or refute-o-matic.  However, since Grover started this new thread, I thought I would post here instead.  So, as background, I recently ran about a dozen sessions of James Bond 007.  You can read about the campaign at http://www.darkshire.net/~jhkim/rpg/jamesbond007/

So, supposing that you are doing task resolution - why do it if everyone is primarily interested in whats happening at a higher level?  I see these reasons:
 1) Boardgame style fun.  In a combat situation, this is the fun of trying to find a tactically sound solution to the current problem
 2) A chance to highlight some attribute of your character.  I would break this down into 2 subcategories:
   a) A chance to show off a skill or ability.  i.e. look how strong Thrudd is!  He killed 3 orcs in a round
   b) A chance to show a relationship or attitude.  Does your fighter consider it more important to guard the mage, or pursue the fleeing orcs.  You can also show how much you character hates something by engaging in overkill (I flamestrike the kobold)
 3) An opportunity to automatically add color.  The conflict resolution of a fight will tell you which side won, and maybe how badly they were hurt, but it won't tell you that someone risked their life to take down the orc leader, or that someone ran away to drink a healing potion, or any other odd bit of detail that could be thrown in.

I agree with all these.  The common thread among these three is that it isn't true that "everyone is primarily interested in what is happening at a higher level".  Often, the details are interesting and indeed the meat of the scene.  See below for discussion with Simon Hibbs concerning the James Bond genre.  

That's what the problem is with the munchkin argument. Gamers who take that stance assume that the kicking ass in combat or always getting the hot sex is the goal of every gamer and the point of every game.

Exactly right, traditional task based game systems fail here because game mechanicaly they are such blunt instruments. Their focus on simulating physical action drasticaly limits the kinds of conflicts they can even attempt to resolve.

In  traditional game design the game system is there to answer questions like 'does the character seduce the agent', 'does he shoot the guards' and in the process we find out if the character is cool or not. Is he like bond or not.

In modern games we know the character is cool - it's a Bond game and he's a secret agent. How could he not be cool? The question is, does he distract the enemy agent with his kiss for long enough to plant a radio beacon on her skirt. He shoots the guards, but does he do it well enough to get to the reactor room before Henchman X.

In a traditional game he'd have to shoot each guard, and do enough damage to take them out, and have enough movement points left to get to the reactor room first. None of the game mechanics directly address the central question which is, does he make it first? Whether he makes it or not is an emergent product arrived at indirectly from the individual game mechanics. In a modern narativist game, it's the central conflict the game mechanics are resolving. The guards being shot is merely a means to an end and therefore isn't itself what we're trying to resolve.

You judge here that task-based systems are a "failure" here, which I don't agree with.  Personally I think task-based is often more appropriate.  In particular, I feel that for the Bond superspy genre, task-based is more appropriate.  In this genre, the broader external conflict really isn't in doubt.  We know that Bond is going to defeat the villain and save the day.  The question is how he is going to do it.  That's what task-based system is for.  In other words, the interest is not in the big picture, because that's already known.  What's interesting is the the details along the way, i.e. the "how".  

So a task-based system like James Bond 007 focuses on the steps along the way, and in particular on quality of success.  A conflict-based system like you describe focuses on success or failure.  i.e. You keep trying to force failures on Bond by rolling for, say, whether he gets there first.  In contrast, in James Bond 007, Bond almost always succeeds.  With most skills over 20, he has guaranteed 100% chance of success on many tasks.  Instead, you roll for how well he succeeds (in JB007 terms, "Quality Rating").  So instead of a single conflict roll which shows failure, you have Bond roll several successes -- showing off just how good he is, but perhaps the successes aren't quite enough.  

Now, this might seem like quibbling (i.e. if there aren't enough successes, isn't that really a "failure").  But I think this is importance.  The JB007 system shows off the superhuman competance of the double-ohs, which is a vital part of the feel (in my opinion).

[M. J. Young]

So, supposing that you are doing task resolution - why do it if everyone is primarily interested in whats happening at a higher level?  I see these reasons:
 1) Boardgame style fun.  In a combat situation, this is the fun of trying to find a tactically sound solution to the current problem
 2) A chance to highlight some attribute of your character.  I would break this down into 2 subcategories:
   a) A chance to show off a skill or ability.  i.e. look how strong Thrudd is!  He killed 3 orcs in a round
   b) A chance to show a relationship or attitude.  Does your fighter consider it more important to guard the mage, or pursue the fleeing orcs.  You can also show how much you character hates something by engaging in overkill (I flamestrike the kobold)
 3) An opportunity to automatically add color.  The conflict resolution of a fight will tell you which side won, and maybe how badly they were hurt, but it won't tell you that someone risked their life to take down the orc leader, or that someone ran away to drink a healing potion, or any other odd bit of detail that could be thrown in.

I think your reason #3 hides a reason #4 which is particularly applicable in computer games: resource management.

If you have a simple conflict resolution mechanic incorporating some form of relative success (the Multiverser term for it; "Quality Rating" in Bond terms) with people at the table, you can look at the dice and say, you did it, but it was close, and it cost you--and then you can decide what it cost, how it was done, whether the character is injured, used up ammo, drank that healing potion, or what have you. Those are subjective judgments regarding what it seems reasonable for the character to have consumed in achieving that goal with that level of success.

The computer can't really make that kind of subjective judgment. Further, in most situations it can't let you make that kind of subjective judgment, either.

Off the top of my head, fixes for this:

[*]The computer determines the level of your success, then deducts from your various resource scores amounts randomly distributed among them. This means that the display might not have shown you getting hit, but you took damage, or you might not have fired your gun but you used up ammo, or whatever other resources the character has.[*]It would be more difficult for the computer to limit how it distributes its reduction of resources to cover only those resources which you "used"; it might also lead to some odd results, if for example using your gun always means you take less damage.[*]The computer could calculate the number of resource points the character had to have spent to succeed at that level, and then lock up the game while the player determines where those points were lost. This puts things in a different light, as for example a character who depends on his arrows in combat and never uses magic might always spend away his magic points so that he still has maximum resources in arrows. Of course, this also requires that specific resource use makes a difference in combat.[*]You could set it up so that the player determines what resources to use in what ratios (I will fire more bullets so as to lose fewer hit points), and the computer then reduces each resource according to the instructions given by the player at the beginning of the encounter--90% from ammo, 10% from life, that sort of thing. I think in this case, "life" would always have to be included at a minimum threshold, such that if the calculation eliminated all of the other resource, life would take the rest of the loss. This, though, is so close to task resolution at this point that it would be easier for the computer to use task resolution to determine the resource management questions anyway.[/list:u]
Those are some thoughts, at least.

--M. J. Young

[Marco]

I came here from John's post in the refute-o-matic thread which dealt with the question of cool characters as relevant to Narrativisim. I think that's a red-herring. I don't see why a system that lets characters be competent (but neither super-competent nor completely incompetent) would be antithetical to narrativism.

That isn't to say that I'm not aware of the protagonization argument (that a character failing horrifically when 'he shouldn't' makes identification with the character harder from an audience perspective)--it's that I think that's a sub-set of Narrativism that gets mistaken for the whole. I don't see why it should be integral to that play across the board.

Like a lot of things, this seems to me to be a matter of preference which gets stated as fact from time to time.

In even a most stridently narrativist James Bond game the premise is not 'how cool is Bond?'

I think that the examples given here:

In modern games we know the character is cool - it's a Bond game and he's a secret agent. How could he not be cool? The question is, does he distract the enemy agent with his kiss for long enough to plant a radio beacon on her skirt. He shoots the guards, but does he do it well enough to get to the reactor room before Henchman X.

(emphasis added)

Have nothing to do with premise--and, in fact, they do, IMO, ask the question of what happens--specifically how well--Bond's level of dominance. The idea that there could or could-not be failure is, again, more story-plot-vs-simulation issue (which, while it sounds like Nar vs Sim is, IMO/IME really a technique-level question).

Sure, the question may be if Bond reaches the reactor room first--but it's the content of the obstacles that makes the question interesting or exciting. It think that what level of abstraction players like and what degree of risk they are willing to accept will differ vastly across any CA.

Finally: Modern games? I released one less than a month ago that uses task based resolution. While I realize there have been some trends in some places (here) away from task resolution, I don't think I'd use the term 'modern' to describe those games.

-Marco

[simon_hibbs]

I use the term 'Modern games' because I'm not sure if 'Narrativist games' is the right term in the GNS sense. Narrativism in GNS is about adressing premise.

Now some games such as HeroQuest do directly enable us to adress premise in game mechanical terms. The premise 'Loyalty trumps Honour'  can be addressed because loyalty and honour can be assigned ratings and used directly in the resolution system. I'm not sure that all the 'narrativist' games that have been mentioned are narrativist in this narrow GNS sense.

In a more general sense, I think narrativist game systems can be taken to mean those in which the resolution system can address whatever conflict is most imprtant to the scene. For example, at one point in the Amber novels Bleys fights his way up a narrow mountain path through literaly hundreds of opponents. He took on each one in turn, cutting his way through them. In the book, it's narrated something like this: "Through much of the afternoon he fought them, killing each in turn, slowly advancing up the mountain."

To me a narrativist game is one which can resolve this scene in a way that maps well to the way I would want to actualy narrate it as a story. The game mechanics map directly across so that an action in the narrative "he fought past a hundred men, yada, yada..." maps to one resolution in the game system. HeroQuest can do this, as can ADRPG and in fact that scene from the novels is used as an example in the rules. A game like D&D does not offer the same narrative flexibility because if we map each action resolution in D&D to a statement in the story that one scene could take a whole book to narrate.

Both are narrativist to the extent that both can generate a narrative, but D&D, Runequest, Traveller, GURPS, etc force a fixed pace and structure on the narrative, restricting your creative controll.


Simon Hibbs

[Marco]

I'm not sure that all the 'narrativist' games that have been mentioned are narrativist in this narrow GNS sense.

I'm not sure either--but there's definitely a difference between Narrativist (which deals with the address or premise) and a variety of techniques for dramatic narration (Theatrix, I think, qualifies as the latter but not the former).

I think on The Forge Narrativist is going to need to be used in the GNS sense if we're going to get anywhere with it.

What I took issue with was the description of games as 'modern' or 'traditional'--I think there is no such division. There's traditional and maybe 'alternative'--but not 'modern.'

Conflict based vs. task-based resolution isn't an evolution, just an alternative.

-Marco

[John Kim]

In a more general sense, I think narrativist game systems can be taken to mean those in which the resolution system can address whatever conflict is most imprtant to the scene. For example, at one point in the Amber novels Bleys fights his way up a narrow mountain path through literaly hundreds of opponents. He took on each one in turn, cutting his way through them. In the book, it's narrated something like this: "Through much of the afternoon he fought them, killing each in turn, slowly advancing up the mountain."

To me a narrativist game is one which can resolve this scene in a way that maps well to the way I would want to actualy narrate it as a story. The game mechanics map directly across so that an action in the narrative "he fought past a hundred men, yada, yada..." maps to one resolution in the game system. HeroQuest can do this, as can ADRPG and in fact that scene from the novels is used as an example in the rules.

I agree with Marco that you should use a different term than "Narrativist" here.  For example, in GURPS, this would be handled through its mass combat system.  So GURPS handles this, though I suspect you don't classify it with HQ or ADRPG.  GURPS is also the origin of the HeroQuest method of having different scales of skill contests.  GURPS has quick contests, simple contests, and I believe extended contests.  

A game like D&D does not offer the same narrative flexibility because if we map each action resolution in D&D to a statement in the story that one scene could take a whole book to narrate.

Both are narrativist to the extent that both can generate a narrative, but D&D, Runequest, Traveller, GURPS, etc force a fixed pace and structure on the narrative, restricting your creative control.

Really, here GURPS and HQ and ADRPG have multiple levels of resolution.  Traveller, D&D, and RuneQuest do not.  But then again, My Life With Master also doesn't have multiple levels of resolution.  You're stuck with only the one level of conflict resolution.  My point is that multiple levels of resolution (i.e. quick vs extended) is different than the type of resolution (task vs conflict).