A look at skills

[Dauntless]

I've been beating my head against a wall trying to decide exactly how I want to implement skills in my game.  First off, let me introduce what I call the problem domain:

Skills vs. Talents-  Many games list as skills more what I would consider talents.  For example, Streetwise, Persuasion (Con, Manipulation, etc), Seduction (Charm), Brawling (totally instinctual and untrained fighting ability), Notice (Perception, Recon, Search, etc), et al.  The reason I need to seperate these into distinct groups will become apparent in a bit.

Skill/Talent Depth vs. Breadth-  How narrowly focused are the skill sets?  For example, let's say you decide to create a skill called Math.  What exactly does a skill level of Math 5 (on a scale of 1-10) mean?  Does it mean that you are average in your Math skills?  Or does it mean that you have gone up through intermediate level math skills (say for example, up to trigonometry or pre-calculus).

The Default problem-   Some skills simply can not be defaulted to.  If you don't know Quantum Physics, then you can't repair the FTL drive.  If you don't know how to program, then you can't use your native intelligence to debug a program.  These academic or knowledge-based skill are the easiest to consider.  Some skills are more intuitive or creative based.  For example, most artistic skills and/or talents are based upon ingenuity.  But again, it's possible for some of these skills to have no defaults.  If you don't know how to play the violin, you're not going to be able to play it (well), even if you have the dexterity of a god.  OTOH, some talents are almost purely unlearned or learned subconsciously (seduction, persuasion).  Physical skills/talents are also a grey area.  Some games have jumping or swimming as a skill.  Jumping is just something innate, though you can train yourself to be better.  Swimming is a learned skill however, and if you never learned, you're virtually guaranteed to drown (not necessarily, afterall, someone somewhere in human history had to learn how to swim).  Brawling is another example.  we all have an instinctive ability to fight, but you can definitely train yourself to be better....ditto for gymnastics.

Proficiency vs. Mastery-  This is an extremely subtle variation of the Depth vs. Breadth problem.  Going hand in hand with the Depth vs. Breadth problem is accounting for the difference between, "How well do you know/perform what you know" and "How much do you know?".  For example, let's say that I have a Rifle skill of 7.  That means I'm fairly skilled at using the business end of it.  But does that mean that I can field strip it and do repairs and cleaning?  For all you know, I may just be a really good shot.  The problem gets deeper with academic or artistic skills.  Say for example that I have a Calculus skill of 5.  Does that define how good I am at Calculus, what problems I can solve with calculus (perhaps I don't know integration...only derivation techniques), or a combination of the two?  See how this goes hand in hand with the Depth vs. Breadth problem?  It's uncertain if you only have one quantifier with a skill if you are measuring your performance capability, the depth/breadth of the skill, or a combination of the two.

Depth vs. Breadth is an extremely important consideration, because it determines how broad your area of knowledge is.  For example, do you simply create a Math skill or do you create a Calculus skill?  Do you just have a Biology skill, or do you have an Organic Chemistry skill?  Do you have a Small Arms skill, or do you have a Rifle, Shotgun and Support Weapon skill?  Depending on the fine grained level of detail you choose for this, what does the skill number mean?  For example, say I choose to simply create a skill called Math.  If I have a skill of 8 (out of 10), what does that mean?  Does it mean that I can solve Calculus problems?  If so, is the difficulty the same as if I tried to solve a simpler Geometry problem?  Can you see the problem I'm getting at?  If you have it defined as a broad area skill, then you might want to impose a skill penalty by saying "This is a calculus problem, so you have a -4 to your roll (or it's gone from an average problem, to a difficult proble)".  But then you have to say...what if it's a difficult calculus problem?  Do you just keep bumping up the difficulty level?  The problem with defining a skill as too broad or generic is that it opens up the problem of whether you actually have the knowledge/capability of performing the task (the Proficiency vs. Mastery problem).

 Some games try to get around this by having Subskills.  First off, it's possible to not cover all subskills, and secondly, it's possible to be worse at a subskill than the general skill (and trying to figure out the cost in such a case can get ugly if you have several subskills which are worse than the general skill).  So they can work to a degree, especially if the skills are already at a fairly fine grained level of detail.  But I think you can still have problems with this method if the top-level skill is still too broad.

Solution?-  The only thing I could think of to solve the issues I've been having is to have Skills/Talents qualified by more than one simple number.  At a minimum, skills should be qualified by:

Skill Definition- This determines the range of capabilities offered by the skill.  For example, if you define a Small Arms skill, then it explains what weapons you can use with this.  This is supplemented by subskills.  This isn't actually listed next to the skill on the character sheet, but rather defined in the game rules.  However, this is must be well defined.

Proficiency-  A measurement of the native controlling factors that influence a skill/talents ability.  For example, Acrobatics is primarily reliant on one's Agility, so Agility plus possibly some other factors determine this number.  The most common other factor that will help your native ability is experience.  For example, if you have great dexterity but have only been firing guns for a week, whereas I have an average dexterity but have been shooting them for years, we will wind up having the same value here.  You can therefore think of Proficiency as an average of Native Attribute + Experience + (talents or advantages).

Mastery- This solves how much you know within the range of the Skill Definition.  In effect, it becomes what we normally think of as Skill Levels for most games.  The problem with this is that Talents (and a few skills) don't normally have a mastery component to it....either you're good at it or you're not....but there's no breadth/depth problem associated with them.

Skill Type-  This is a broad catchall that tries to solve the Default To Problem.  If a skill is listed as Academia, then it can not be defaulted to.  If it is listed as Creative or Physical, then it can be defaulted to one of the primary attributes or aptitudes.  Note that it is possible to have Creative + Academia.  This means that there is a learned component to the skill (a pure creative skill might be Drawing or Dancing, while a Creative + Academia would be playing most musical instruments)

First off, any ideas on other ways to solve the issues in my Problem Domain?  Secondly, my game system is semi-universal in the sense that the character creation process is setting and genre specific (I have a lifepath system which is half character class, half choose your own skill).  I'm working on two game settings, one a militaristic gritty, hard sci-fi campaign, the other a low fantasy, pseudo-historical pan asian setting.  I'm wondering how finely grained I should make these campaign settings in regards to the skill sets?  I definitely don't want to make them corase grained, because I believe that a more narrowly constrained Skill Definition will provide more color and flavor than having a broad-based generic skill.  And of course, determining this is necessary to determine skill costs.

[Andrew Martin]

Have a look at "Vincent's Standard Rant: Power, Credibility and Assent", here: http://www.indie-rpgs.com/viewtopic.php?t=3701 Point number 4 is the most important.

Do you realise that game system descriptors, like skills and talents, merely indicate what's important about the game? Is it worthwile attempting to make a completely accurate and "realistic" skill and attribute system that accurately models what humans can or can't do in reality, when your game settings are: "one a militaristic gritty, hard sci-fi campaign, the other a low fantasy, pseudo-historical pan asian setting"?

Surely it would be easier and better to create descriptors of what's important in each setting?

=============================================

If you are intent on pursuing this kind of path, one solution I've found is to have three sets of very broad skills. In my "generic" game system/s, I have these three very broad skills: (Species), (Culture) and (Occupation), which are named with the specific species, culture or occupation. These very broad skills are set to Average for most PCs and NPCs, and contain all the skills that most members of that particular species, culture or occupation have. So, for example, a average human blacksmith from France would have (using Fudge descriptor levels, where Fair = average) these descriptors:

Code Select Expand

Species - Human: Fair
Culture - France: Fair
Occupation - Blacksmith: Fair

If you're using a percentile system instead, then use some thing like:

Code Select Expand

Species - Human: 50%
Culture - France: 50%
Occupation - Blacksmith: 50%

I usually assume about 10 - 12 broad skills are included in each very skill, together with lots of narrow skills which aren't important, or that we (players and I) don't know about.

After that, I allow a few (10 - 12) other individual skills and descriptors to further differentiate this particular character from other members of the same species, culture and occupation. Lately, I've been tempted to just have a fourth very broad skill, which includes all the above, which is simply called "Me!" (or something like that) which covers all the stuff that's important to the individual character/player.

I've found in practice with players that this kind of arrangement of grounding characters in a specific specific species, culture and occupation stops player paranoia that they've forgotten to get some necessary and important skills, and hugely speeds up character generation, by eliminating unnecessary detail, but still allowing detail if required by players.

I also have attributes to cover important things about character speed or initiative and common stuff that most characters have. All these attributes, again, default to average, so that only if player wants, is there any need to specify a different descriptor level.

[Andrew Martin]

I definitely don't want to make them corase grained, because I believe that a more narrowly constrained Skill Definition will provide more color and flavor than having a broad-based generic skill.  And of course, determining this is necessary to determine skill costs.

Do you know what I found when I had skill and/or attribute costs in a RPG? I found that players min-maxed their skills to get those that most important to them and got skills and other descriptors that made little sense. When playing games we had characters with enormously powerful enemies that logically had to deal with the entire party instead of just the one PC who had the enemy, we had strange and very weird characters with descriptors and disadvantages that made no sense and made a mockery of the game setting. I found that these problems occured in Fuzion, GURPS, WW Storyteller, and in many other game systems that had a points cost/buy system. All these game systems, and your system too, has this problem that the game system rewards players for behaving in this way, not in the way that the setting would have.

My suggestion is to remove the points cost/buy system entirely. Instead, start off the PCs with average attributes and skills, and give the players what ever they need to make a competent character that the player likes. Then simply review the character and say "No!" for those characters that don't fit the setting (but first make sure that the players know the setting!). Then have the system make the outcomes for success and failure of the PC's goal like those in Ron's Trollbabe, where the GM narrates PC success, and the player narrates PC failure. GM narration of PC success means that success doesn't go over the top, and player narration of PC failure means the players aren't railroaded by the GM.

[Andrew Martin]

I have a lifepath system which is half character class, half choose your own skill.

That would be an ideal thing to add in helping to create the individual "Me!" very broad skill/descriptor container of descriptors. Instead of having:

Code Select Expand

Me!: Fair

Have the life path generate the individual historial events for the players that are stuck for inspiration.

[Dauntless]

Let me start on my design philosophy...whether it be for designing pen and paper roleplaying games or for designing computer programs.

It's all about modeling.  Now the very first question we have to ask ourselves is what exactly it is we're trying to model (simulate).  I call this the "Problem Domain".  It's the set of questions that we're trying to answer.  Once we've figured what it is we're trying to model, we start to determine the essential qualities and aspects of the.  This is perhaps the second most important step of the modeling process (the foremost being figureing out exactly what it is you're trying to model...sounds silly, but many people don't think through this portion thoroughly enough).  Once we figure out the essential qualities of the system, we try to capture how the interdependencies between them work (i.e. how an aspect of an object influences or is influenced by other aspects of other objects).

As you said, having skills and talents makes them important in a game by the sheer virtue that they exist in the game.  It's very possible to create a game in which there are no skills or talents and rely totally on attributes if skills or talents were not an important essence of the system you're trying to model.  But modeling is essentially about whittling away at the unnecessary until you get to the core essential qualities that represent the problem you're trying to provide a solution for.  Different game systems try to provide resolutions different problems, and hence what they model will be different from system to system.  To look at it another way, games do two things, they ask different questions, and they provide different means of giving answers to those questions.  Hence, system matters because system determines how we're going to model the problem domain, and what we're trying to model determines the implementation of what we're trying understand/solve/integrate with (modeling in a game design sense isn't solving a problem per se, or trying to increase our understanding, but rather to provide a contextual vehicle in which we can explore various situations).

So skills and talents are important in my model because I'm trying to simulate "what if" scenarios.  Therefore reproducing cause-effect relationships are very prominant in the model I'm trying to create.  The more detail I put forth the more accurate these what-if scenarios will be.  By putting the player in the shoes of the character the player will understand his limits and capabilities better.  In a way, any system (model) that you create is a constraining system on your inputs (your choices) and therefore your outputs (what happens in the game).  The trick is in determing what to constrain and how those constraints influence the outputs.  Constrain the system too much, and you limit player choice (and ergo, limiting the possible outcomes).  Constrain it too little, and you lose an anchor to plausibility while also makign arbitration exceedingly more difficult to be consistent and fair.

As for the min-maxing abuse I have within my game-mechanics several conventions designed to reduce this.  For example, emotional, mental and awareness states will affect likelihood of success.  Hidden variables which are unknown to the player but which are known to the GM will provide a level of non-determinism...i.e., the player could memorize all the rules and factors which would help/hinder him, but he still won't be able to come up with the exact odds of said success or failure.  Moreover, min-maxing abuse is very prevalent in situations which the player can exploit context sensitive advantages.  I see this more as a weakness of a GM's inability to provide scenarios which negate the min-maxing abuse.  I also think min-maxing abuses occur more prevalently in point-cost based systems where players hunt for the most cost-effective solutions rather than have a character concept in mind.

[Dr. Velocity]

Since I'm more a fan of rules lite, this may sound biased, but how much difference should it TRULY make, being able to have a skill of 7 with a rifle but not knowing how to field dress it - to me this seems obsessive and definitely unnecessary.

Unless the player is creative and wants it differently, I assume anything related to a skill is also around that level, because its just easier for everyone. The only other alternatives are to have some fraction of your skill represent the 'maintenance' and such, so you'd have a 3 or 4 in field dressing and care, or the last alternative, to have multiple categories for each skill like 'use', 'maintenance', 'specs' etc etc - I want to play a game, not bookeep things that, overall, should really have no effect on the game itself, so I myself lean more toward having a 'guns' skill and a 'fight' skill and such so the players can describe and decide whatever they want, there's still a system and structure but the GAME itself, and not the system or the mechanics, becomes the focus.

[Dauntless]

I just read the Power, Credibility and Assertion post, and I have to say that in a way I agree with it, and in a way I don't.

The author presumes that Credibility is solely within the jurisdiction of the players in the group.  However, to me, this is exactly what rules are for, to help enforce credibility.  He says that all systems assign credibilty to various aspects, but that ultimately its up to the players to make the call as to whether to surrender credibility and hence give power to the author of an action.

Frankly, I find that an odd statement.  While it's true that I see rules as a guideline rather than a fiat truth, if we delegate to players the authority to determine whatever they want to accept, then there's no point in having rules whatsoever.  Fundamentally, it is the role of the GM to arbitrate issues and if players don't agree, then they can either swallow it or find another GM.  To hold such power is tantamount to the belief that, "The customer is always right"...the customer in this case being the player.  To give players the final ability to give credibility and hence power is misguided I would say.

Now, if we say that internally the player has this privilege, then I agree wholeheartedly, but if we extend this to say that players can effect the game outcome is I believe very wrong depending on the system you're trying to model.  Now, let's say the player's are all gods, then such a system design might make sense, but otherwise to allow this is tantamount to allowing players (and hence characters) to alter reality at their whim.  And for all practical intents and purposes, game rules are reality.

So now we get back to what players can and can't do in "reality" (the game world).  This is exactly why modeling how skills work in my system is crucial because they help determine the capabilities and limitations of what the characters can perform.  The credibility and power lies within the rules of the definition of skills and how they are implemented.  Now, if the rules aren't very credible, say for example that you have one skill called Athletics that lets you do everything from gymnastics to swimming to figure skating to pooldiving that would personally lose credibility to me as a player.  However, I wouldn't have the capacity to tell another player...."sorry, but that's just bogus that you're a world class athlete in every sport in the Olympics...I don't buy it, so you're not able to do that fancy acrobatics maneuver to get a dodge bonus".

I happen to believe that immersion into a game world is helped by the level of detail of a system.   If the system is too abstract and generic, then our interpretations become more ambiguous (especially between the shared imaginings of the other players and GM).  I believe this is one of the flaws of Universal systems.  Universal systems lack a certain flavor that help define the genre of setting that we're trying to model and hence lend a generic feel to them. I also believe that judicious use of limitations and constraints are also helpful (given the problem you are trying to model...if your game is about god-like wizards, then constraints and limitations will hurt the sense of immersion).

[Dauntless]

Some skills have more of a use of Mastery than others.  The rifle example may seem out of place, but does have its merits.  As an example, imagine that your character has never fired a gun before in his life, but as the zombies are all around, a cop gives you a spare pistol.  Unfortunately, he neglects to tell you how to chamber a round, set the safety or release the magazine.  It's easy enough to actually shoot a pistol especally if you have good hand-eye coordination.  In fact, a brand new shooter may be a better shot than someone with average to poor dexterity who has lots of experience.  But what happens if you get a gun jam or you have to reload?

By having the skill system designed as such, it gives you immediate visual feedback that says, "This character basically only knows how to pick up a gun with the safety off and shoot it".  This way the GM doesn't have to arbitrarily arbitrate (I've always wanted to say that...) a solution, "Hmmm, well, your character hsa no law enforcement or military background, but you have a really high dex and a decent intelligence....maybe you can figure it out."

Where Mastery really comes in handy are academic type skills however.  Simply by looking at the Mastery level, you can skip a task roll by determining that the character doesn't have sufficient knowledge to even attempt it.  

So while it may seem that the additional numbers are a burden on play, they are actually there to help and in some cases will actually save game time rather than waste it.  Of course there will be other situations where there may be some decision making required by the GM to figure out what the Mastery level should even be.  So now you have two components to worry about...a difficulty (which proficiency covers) and a requirement (which is what mastery tries to resolve).

[Dr. Velocity]

Hmm. Ok well I don't get into the nitty gritty of these forums enough to recall other similar posts or points or direct you to threads, but let me just give you my own observation on your idea, and the things that automatically come to mind for me, just so you have one more person's off-the-cuff genuine opinion on this.

Modeling things of course is what we do when we pursue any virtual or vicarious activity, and gaming is the best example. But looking at general statements, how and where do you come up with a baseline for modeling things? You say the idea of possibly having a jammed gun when the zombies attack could be a critical juncture in a game, and I agree, its very common and expected and exciting and just GIVES you the feel of a zombie movie. This however, could also be modeled by simply having a fumble or even a failure-but-not-quite-a-fumble be a jammed gun, or dropping the bullets or what-have-you, or even rolling against your Cool or Sanity or other statistic - this would give you the potential to simulate 'freezing' up in combat but not have it necessarily based on not having the ability to use the weapon correctly - you can CALL it whatever you want, but then thats the arbitration which you are obviously trying to avoid when possible, ideally.

Now move on to magic or unusual ability or power - how do you model that, effectively, and 'realistically'? How do you model 'how much' a player knows about throwing a fireball, or using The Force, summoning demons, or any of a million other things, even detecting lies. If you cannot consistently model everything in your system the same way, should you really be nitpicking so specifically?

I myself (like a lot of others probably) am almost intuitively good at English but know almost nothing more than the absolute basics about the WAY the English language works - I can diagram a sentence if pressed and given time to study, but normally it means nothing to me, I just KNOW what sounds "right" and what doesn't - I'm not perfect and my intuition is sometimes wrong, but it can't be argued that I don't have a good grasp of the structure of the language, much like people who can play music by ear but not read a lick of sheet music. To me, this is more simply a variation of a skill, rather than a different aspect of it, at least game-wise, it would make more sense to model it that way - I could say I'm good at English or at music, and unless it was a common and vitally important part of the system, it wouldn't matter HOW I was good at it or exactly what it entailed, unless I as a player want to make distinctions for the sake of color.

Back to your example of the zombie and the jammed gun, the reason we can see this event happening is because we've read a story or seen a show about that sort of situation, its dramatic. Since there ARE no zombies, this means that the situation and therefore the type of writing which its included in is undoubtedly written with a cinematic bent - and the cinematic medium is by its very definition, "not realistic", so I am unclear on attempting to model something realistically, to the point where you need to know not only what you know about a main skill, but how much you know about its peripheral and associated sub-skills, but the thing in this example cannot be modeled from reality but only from interpretations of someone else's imagination...

I realize the zombie example may be a bad one since you may be working on a game which can be played without any paranormal aspects, but my point is here that I feel you may be missing a point: the gun may jam when confronted with the zombie because it HAS TO - its necessary for the movie/story, for the character to have to run away, for the cop to get eaten, etc., because the author/director is trying to tell a story and make a point, NOT because the character really failed his 'use gun' skill. This again though, involves GM arbitration, so to me, what you're wanting is a system mechanic which has built-in arbitration to CREATE dramatic moments, BUT to me, unless this particular dynamic is going to be a commonly recurring and possible theme in each game, I think you may be trying to include a mechanic for the sake of one or two good thrills, but which will otherwise simply serve to complicate and bog down normal play.

[TonyLB]

I came upon this thread.  It has, in itself, quite a few good points on Realism.  It also has the following section of reference to yet more links:

Hello,

I'm afraid this is a broad-spectrum smattering of older threads, but I really do recommend reviewing them, for this thread's discussion to be maximally productive.

Purpose of rules - this thread is mainly about GM authority, but the realism issue crops up subtly more than once, especially on page 3
Hard Time (prison RPG)
Simulationist reality and Narrativist reality
Transparency again - again, this is not directly about realism as a term, but everything in the thread relates to the topic
Playing The Riddle of Steel (a little)
The psychology of combat
Failure=Advancement - a good example of avoiding the realism-debate for purposes of a better discussion
The FooFoo Factor
What is the most realistic RPG? - a seminal thread. I strongly recommend not posting to this (current) thread until you really read and process this older one, especially Ralph's (Valamir's) post at the end of page 1. I also recommend skipping over my objections to the issue of suspension-of-disbelief, as it kind of derailed the discussion.
"Realism" valued in both G and S?
Realism in RPGs
Realism in RPGs II
Using realism in RPGs part 2

Best,
Ron

These folks do a daunting (heh) job of pointing out the many ways in which the act commonly referred to as 'modelling the real world' is in fact a highly subjective undertaking.  I found it (as usual) good reading, and a good eye-opener.  I hope you find it equally useful!

[Dauntless]

Dr. Velocity-
"Reality" is dependant on the context of your game world.  It's possible to create a "realistic" world that includes magic or super-powers.  Reality is really just another way to say, "This is how the world works".  The key is making sure that there is consistency in what you are trying to do, and that (unless the reality of your game world is really whacky), if I'm given two sets of inputs and initial conditions which are exactly the same, then I should get exactly the same results or rather, the exact same probability of getting said results.

Why is it necessary to delve into such a level of detail when you can simply give it an off the cuff description from the GM?  Because the level of detail helps the GM determine how he's going to describe the scene.  I've always stuck to the philosophy that rules are guidelines that help the GM decide how the flow of action will proceed, or help him resolve how to arbitrate  something if the answer is neither obvious nor if the flow of the story is dependant on it.  As another game designer once said, "It's easier to ignore the rules and go with your own description of events, than it is to have to decide the outcome of an event with no help of the rules whatsoever".

Now the achilles heel of detailed simulationist style games is that the rules become a crutch rather than an aid.  Instead of letting the dramatic flow of the story unfold the way a GM envisions, he relies on the rules to provide the story flow.  Sometimes this is interesting just to see where it goes, but it lacks the direction that only the guiding hand of a human player (and not the impartial non-sentient rules of the game)  can provide.  When this happens, you've gone from having a roleplaying game to playing a wargame.  If there is no need to weave a story, then there's nothing wrong with this, and it's what appeals to many players.  However, humans like to have a sense of continuity and meaning that having a direction or purpose provides.  This is only possible if the GM is allowed to "tailor" the outcomes to a certain degree.

In my philosophy, detailed rules are there when the GM doesn't know what to do or requires a level of detail that abstracted generic systems can't provide.

[Dauntless]

Tony LB-
I just read the post: Purpose of Rules and I think my previous post above answered his question.

What is the purpose of the rules in regards to actually dictating what occurs? Are they designed for player vs. player conflict? An accepted means of limiting Player control within a game? With the GM as an "above the law" participant in a game, what purpose do the rules serve to them?

As I stated above, the ability of the GM to fudge the rules is to allow for the events to take on a direction or purpose.  If the GM lacks this ability he in essence becomes a referee for a wargame, doing nothing more than judging certain ambiguous factors (does this unit have a line of sight to the target or not for example).

However, in contrast to the Power and Credibility thread, the suggestion that players should have the right to "fudge" the rules by giving credence and hence authority to an action doesn't create direction.  Rather it creates a "I want things to go my way according to my beliefs or wants".  This is usually a prescription for chaos.  

Now, if only one person has this authority (the GM) it's fine, if everyone has this, there's only one option.....there's no GM at all.  From what I understand there are some game systems that do this (I believe one of them is called Universalis or maybe Mythus).  In this style of play what happens is more akin to an ad-libbed or improv drama with no director to give pointers.  It's an intriguing idea, but if it's anything like improv music or improv comedy, its extremely difficult to pull off properly.  With the GM acting as the conductor or director, he provides a sense of continuity and direction that often is lacking when several different people all have different ideas about what direction to take the story in.  Communal vision is almost never as profound or evocative as having a single visionary weave a story.

I'm reading some of the other posts now though....thanks for all the links!

[Andrew Martin]

Communal vision is almost never as profound or evocative as having a single visionary weave a story.

Are you sure about this? I used to think the same way, until I tried out communal vision where players are highly informed of the game setting and I found that it was better with communal vision. Have a look at my actual play experience here: http://www.indie-rpgs.com/viewtopic.php?t=11749 Based on this experience, I'm making sure that for my future system designs, all players can manipulate the flow of the story within the setting context. I've decided to junk the rollercoaster ride. :)

[Dauntless]

If the group synchs together right, then what it produces can be extremely beautiful and one is all the more amazed that the group was able to integrate all the seperate parts together into a unified creation.  However, with the limited experience I have with such things (improv drama and listening to improv music) it's very rare to experience this.

That said, I think that a group can work together to improve how they work together to improve their chances at synchronizing themselves.  This team effort has some side benefits as well I would imagine by being able to better read what the other players have in mind.

When I do martial arts or when I dance with a new partner, this "sensitivity" is very important.  When you dance with a new partner, you have to firmly establish a leader and a follower, but once you've practiced with them for awhile, the distinction becomes less and less clear and both contribute to the synthesis of something new.

So I'm not saying it's impossible, but it requires time, effort, committment and a maturity on the parts of all the gamers in the group.  If even one of them is too headstrong then it will ruin it for everyone else.  In other words, your group has to consist of all team players.  If you have such a group...lucky you :)

[Andrew Martin]

So I'm not saying it's impossible, but it requires time, effort, committment and a maturity on the parts of all the gamers in the group.  If even one of them is too headstrong then it will ruin it for everyone else.  In other words, your group has to consist of all team players.  If you have such a group...lucky you :)

I totally agree. So if all the players and the GM are agreeing to play by the rules of the game, why not have the rules of the game encourage and reward this kind of play? :)