GNS, Intent and Motivations

[Marco]

Using these labels for clarification I stand firmly by my assertion that Motive (i.e. Game Motive) is completely inseperable from GNS.   Or to put it another way.  GNS does not exist in a meaningful way apart from this motive.

Which does not make GNS less valuable to a game designer--who can use their own motives as Ron pointed out. It does make it a poor tool as a way to rate games or describe play other's play (which gets done a lot "that's drift" when the person in question isn't the player).

One of the reasons I think 'Drift' is not an especially useful term (in the way it gets used--if speaking from one's own perspective then it describes what *I* did to the game system to make it play that way) then it's verifiable. Otherwise it's as easy as ascribing GNS motivations to someone else.

-Marco

[Seth L. Blumberg]

I must be some kind of freakin' idiot to jump into this discussion.

Mike Young's example was perfect. "If I see a player kill the baby kobolds, because they aren't worth any EXP alive, then I will assume that this is a Gamist decision." We assume this is Gamist because prioritizing EXP is a Gamist mode.

If you remove the because and say "If I see a player kill the baby kobolds, then I will assume that this is a Gamist decision" you're left with a statement that just doesn't hold up any more. Without the intent behind the action, we cannot easily assign it to a mode.

Note that the examples of Gamist behavior that Ron (who is the primary proponent of leaving motive out of the question) supplied are quite different.

From the Ying in the Yang thread:

When I see Bob the Player say, "Yeah! You suck!" and eagerly grab the dice for his turn as his fellow player laughs ruefully, gazing at his failed saving throw, then I recognize one of the many kinds of Gamism in action ... or at least I'm alerted to keep an eye on how the group reacts over the course of the whole session to this sort of behavior. Three rounds later, the two players cooperate like fiends to double-team the troll wizard and a good roll saves their bacon - they high-five each other, and Sam the Player says to the GM, "Yeah! You suck!" and they all laugh, delighted. OK, I say, point 2 for Gamism goin' on here, and keep watching.

In other words, the behaviors that indicate in which mode someone is playing in a given instance are not, repeat, not the actions of their character. They are metagame behaviors, and they exist in the social sphere surrounding the game.

(This was a revelation to me, and yet again transformed my understanding of GNS.)

[damion]

contracycle:

Voter A opposes racism, and therefore votes for a Democrat who advocates affirmative action.
Voter B opposes racism, and therfefore votes for a Republican who opposes affirmative action.

How are we to tell them apart based on motivation? And if there motivations are identical, should not their decision be?

I think there are several reasons we are disconnecting.
1)I think it's obvious that people can make the same decision with different modivations. GNS specific examples are the the previous posts.

2)You seem to be confusing the granularity of modivation. Your example is like saying all players have the same modivation becuase they are trying to have fun while gaming.  GNS cares about the outcome a person is
striving for in their decisions. It doesn't care why they want that outcome. Your right, that's irrelevent and unknowable.

A republican may think that affermative action encourages racism, so they want to ban it.
A democrat may think affermative action discourages racism, so they are in favor of it.
GNS would be like classifying them as pro/con affermative action, which is pretty easy from their actions.
 
We classify GNS by actions, because that is what we can see, we can't see goals, reasons, modivations, priorities, or anything like that.  Say an observer see's a decision and says it's Narrativist. This decision is Narrativist because the observer sees that behavior as indicating the person had a Narrativist goal when they made the decision . Yes, there is error because we are trying to get inside people's heads, but there's not much you can do about.  Classifying behavior is meaningless, it's like saying a behavior is happy or sad, it makes no sense.

[edit]: Actually, valamir said what I meant pretty well. Their modivations relative to the game are what matter.

[Marco]

I must be some kind of freakin' idiot to jump into this discussion.

Actually, that's a real good point--but while the "chicken dance" may indicate gamist behavior and someone saying "now ... to further the plot I do THIS!" will indicate Narrativist behavior, in the absence of such a motivation-explaining cue how does one judge?

(What I'm saying is that a victory dance is about the same as saying "because they're worth no XP alive ... ")

-Marco

[deadpanbob]

In other words, the behaviors that indicate in which mode someone is playing in a given instance are not, repeat, not the actions of their character. They are metagame behaviors, and they exist in the social sphere surrounding the game.

(This was a revelation to me, and yet again transformed my understanding of GNS.)

Seth,

I absolutely agree that, in particular with Gamist behaviors, there are meta-game behaviors that the players engage in that help to determine a preffered mode of play (again if observed over a long enough series of Instances of play).

However, and at the risk of sounding like a broken record, even these meta-game behaviors (the high fives, the name calling) could represent other modes of play.  With Ron's example, it may be unlikely that we are seeing Narrativist behavior in action - but its possible.  It is probably impossible that Ron's example represents Simulationist behavior - probably.

And, even if we restrict ourselves to only making a longer term series of observations about meta-game or OOC behavior on the part of the players, all we can really say with the model sans motivation is:

"These players typically engage in gamist behavior.  Therefore, these players typically engage in gamist behavior."

Not germaine to this thread - but in designing a game, it would be nice to have, although unlikely, a much clearer definition of what rules/premises/color/themes etc. support which types of behavior.

They only way I as a game designer can apply GNS to my game, AFAIK, is through second-hand observation of playtesting.  That is to say, only by asking others not myself to play test and then either watching or reading a journal can I apply the theory to my game.

If I want to design a game that heavily supports/encourages/rewards Narrativist and Gamist modes of play, what am I to do?

In my mind, in my opinion, based on my experience (i.e. probably not relevant to you, your milage may vary, you are likely going to disagree) - I will need to impune the Game Motives of the play testers to figure out if they don't like the game because specific components don't support Gamist or Narrativist modes of play - especially because they themselves prefer Gamist or Narrativist modes of play.

Without either an explicit stipulation of the playtesters perfered mode(s) of play (which necessarily requires an explicit understanding of GNS) and/or specific and explicit statements of motivation for why they played in a certain fashion - the playtest will not be worth much to me, and I won't be able to repair or redesign my game in such a way as to meet my goals.

Cheers,

Jason

[C. Edwards]

I would like to throw something into the mix.

According to modern behavioral science, our feelings don't actually do anything.  Consciousness - subjective experience,sentience - has zero behavioral manifestations; it doesn't do anything.

In technical terms: consciousness, subjective experience, is "epiphenomenal" - it is always an effect, never a cause.

I think that renders motivation, in the emotional sense, invalid as a determiner of GNS goals.  Valamir's use of "game motivation" seems much more like a response to physical stimuli and that makes it, in my opinion, sound more plausible.  Then again, I'm no scientist.

-Chris

[deadpanbob]

I think that renders motivation, in the emotional sense, invalid as a determiner of GNS goals.  Valamir's use of "game motivation" seems much more like a response to physical stimuli and that makes it, in my opinion, sound more plausible.  Then again, I'm no scientist.

Much as my gamist tendencies and my joy at being the Devil's Advocate might pull me in the other direction...

I must admit that I like this concept too.  While I do think that, for my own personal uses, GNS requires something deeper than an analysis of behavior, Valamir's coining of the term Game Motivation makes a lot of sense to me too.

Frankly, I don't want to think about the deeper "psychological" conditions that drive my players to kill orc children.  Don't want to know.

In any case, this has been a very good discussion so far - IMO.  I appreciate everyone's comments.

Cheers,

Jason

[jdagna]

When I see Bob the Player say, "Yeah! You suck!" and eagerly grab the dice for his turn as his fellow player laughs ruefully, gazing at his failed saving throw, then I recognize one of the many kinds of Gamism in action ... or at least I'm alerted to keep an eye on how the group reacts over the course of the whole session to this sort of behavior. Three rounds later, the two players cooperate like fiends to double-team the troll wizard and a good roll saves their bacon - they high-five each other, and Sam the Player says to the GM, "Yeah! You suck!" and they all laugh, delighted. OK, I say, point 2 for Gamism goin' on here, and keep watching.

In other words, the behaviors that indicate in which mode someone is playing in a given instance are not, repeat, not the actions of their character. They are metagame behaviors, and they exist in the social sphere surrounding the game.

Granted, that metagame behavior may be more revealing than in-character or in-game behavior.

On the other hand, in Ron's own example, he admits that even a seemingly obvious case of Gamism "at least I'malerted to keep an eye on how the group reacts over the course of the whole session."  So even in a metagame context, we're still not sure what mode is represented by a given action.  We still need a larger sample of activity to base our judgements on.  Add in the complexity of drift, and you're still left with a situation that I feel is too complex to deal with strictly in observable terms.  We can limit our discussion to what is observable, but I feel that it's like studying a tree by only looking at its shadow.

[Seth L. Blumberg]

We can limit our discussion to what is observable, but I feel that it's like studying a tree by only looking at its shadow.

Actually, it's more like the situation faced by experimental physicists, which has been compared to trying to figure out how to make a Swiss watch by slamming two watches together at high speed and examining the debris.

It is impossible to determine someone's mode of play by examining their motives, because you cannot ever observe their motives. You can only observe their behavior. (Saying "I am motivated by X" is a behavior, too, and--as has often been observed in this forum--not always the most significant behavioral indicator of actual mode of play. Sometimes people aren't clear on their own priorities.)

On the other hand, "prioritizing" is not an observable behavior; it is a conclusion about someone's value system, reached by induction. (I have never seen anyone prioritize anything. Have you?) Thus, GNS involves consideration of what could loosely be called "motive," if only as a hypothesized and epiphenomenal determinant of behavior.

To sum up: you're both right. Now go talk about something productive. :)

[M. J. Young]

I actually see three sides to this debate.

The side that make the most sense to me is that which maintains that conduct is only relevant as it implies motive or intent. Gamism isn't about what you do, but about why you do it. The same is true for simulationism and narrativism.

The side which opposes this maintains that we can't know motive or intent, we can only know actions. This seems to have the stench of Hume about it--we can't know reality, only perception. To take this seriously, it would seem that there are no gamist motives, only gamist actions; no narrativist intents, only narrativist choices.

But every time this side attempts to illustrate what it means, it slips into impugning motive.
In response to me,

Again, you are using the abstracted nature of the model as an excuse to lump what would almost certainly, in practive, be very distinct behaviours into a single category. If we were to FAIL to come up with at least a tentative "why" from the "what" we saw, then the only answer would be that we do not know. But how probable is it that these two behaviours are really, over time, indistinguishable?

Notably agreeing that we must "come up with at least a tentative why", which, I insist, is motivation or intent.
Similarly, this has been caught by others,

If I see a player kill the baby kobolds, because they aren't worth any EXP alive, then I will assume that this is a Gamist decision. I can't see how it prioritizes verisimilitude, or narrativist premise. What I can't assume is what motive caused the player to do what he did.

Yet clearly he has either inferred or been told that the player killed the baby kobolds from the motivation of gaining experience points. Again, he could have done it because his character recognizes baby kobolds are irredeemably evil and will only grow to be wicked killers; or because his character background states that his baby brother was killed by kobolds giving him a hatred for the entire race; or because knights of his order are the sworn enemies of kobolds and do not spare the young. It is only that motivation--the experience points--that makes this a gamist decision.

Incidentally, it could equally be a gamist decision not to kill these kobolds, if for example the player was assessing his resources for future combat in which he might gain more experience or more treasure, or in which his character might be at risk.

I have yet to read a post here that explains how behavior can lead to classification without passing through the step of impugning motive or intent; in fact, it seems to me that the classification itself is defined by motive or intent. The majority seem to agree with this; the minority can't seem to explain otherwise without falling into some hidden trap of letting motive or intent in through the back door. It is there in all the posts on both of these sides.

The third side is represented solely by Ron. He maintains that motive and intent have nothing to do with this whatsoever, as he understands them. They are a taxonomy, like observing the spot patterns on the wings of butterflies. Gamists are gamists because they act like gamists, not because they have gamist motives, and so for simulationists and narrativists. He never slips into letting motive in through the back door.

Yet at the same time, elsewhere, he has consistently denied that anything within the context of the game can be pigeonholed as exclusively belonging to one genus. All mechanics can be used for any mode. Stances are not strictly relevant. Immersion is not a guaranteed tell. There is not a single behavior of which it can be said that this belongs exclusively to one, or even that it is inherently excluded by any one. It is only in the totality of the behaviors that we find the pattern. Yet what is that pattern? What is it that we are able to deduce from the totality of behaviors, that thing that causes them to be related to each other--that suggests victory dances are like killing baby kobolds, and both of these like gloating over successes and cursing failures?

Ron says that the connecting thread is not motive, and not intent.

I could suggest that the connecting thread is attitude; but to my mind, I don't see how that is different from motive or intent--and I don't see any reason why attitude would be acceptable where motive and intent are not (the same objections all apply).

In the end, what is the quality of gamist behavior that makes it gamist, that can be defined without creating a tautology?

--M. J. Young

[contracycle]

Actually, that's a real good point--but while the "chicken dance" may indicate gamist behavior and someone saying "now ... to further the plot I do THIS!" will indicate Narrativist behavior, in the absence of such a motivation-explaining cue how does one judge?

You can't - you can only observe the behaviours in search of such of clues.  So the subject of our investigation must be the actual behaviour of the participants; without such observation any proposal as to the motive of the players is pure speculation.  Motives are multiple - maybe a given player is living out a power fantasy or working out some parental issue - what drives people to undertake actions is seldom simple.  If, however, there are patterns of behaviour which are roughly universal across the activity, then it doesn't particularly matter what the motives are, or how complex they are.  You do you not need to know what motivates the use of a hamemr - knowing the behaviour for which the hammer will be employed, you can design accordingly.

In the case of someone who is gaming socially, cited above, and is a "true X" but playing in a game generally "Y", the example I feel is abused.  Nobody is proposing - indeed everyone strenuously objects to - attempting to impose a given play style on anyone.  But identifying that there is a disparity in stylistic preference allows us to assess how profound a problem it is, and what it is that constitutes the problem.  Otherwise all we would be able to do is complain.

We might for example propose that this player might be better off in another game of type X, if the quality of the play experience is important enough to them.  If the socialising aspect is more important to them, we could at least try to design a quality game of type Y which is at least tries to avoid being a bad and particularly annoying game of its type, this might make the experience less painful for the player.   Perhaps games can be built which explicitly contain elements which cross from one mode to another - this has not been much explored yet but given that the model accepts co-existance of modes, albeit at differentiated priorities, its possible that such a design might be feasible.  

At least with such a model of behaviour we can start looking at the options and engaging in constructuive behaviour.  Without such a model all we would be able to do is say "too bad" and pat them on the back.  To an extent that might be all we can do anyway, but at least we have something to work with now - or at least, those that think the model is worth adhering to do.  It gives us a place to go looking for cues, and a theory as to how interpret those cues.  Even if the theory is wrong in some siginificant way, going out and accumulating the data which disproves it would lay the groundwork for whatever model, if any, should succeed it.  We will necessarily have to speculate on the motivations of players in this process; we will have to ask questions, discuss our experiences, explore our own motives and those of our players - but to priviliege those aspects above what is observable would be to use the least reliable available data.  When we see behaviour, we are seeing motive expressed - but it is the behaviour which is accessible.

[contracycle]

Similarly, this has been caught by others,

If I see a player kill the baby kobolds, because they aren't worth any EXP alive, then I will assume that this is a Gamist decision. I can't see how it prioritizes verisimilitude, or narrativist premise. What I can't assume is what motive caused the player to do what he did.

Yet clearly he has either inferred or been told that the player killed the baby kobolds from the motivation of gaining experience points. Again, he could have done it because his character recognizes baby kobolds are irredeemably evil and will only grow to be wicked killers; or because his character background states that his baby brother was killed by kobolds giving him a hatred for the entire race; or because knights of his order are the sworn enemies of kobolds and do not spare the young. It is only that motivation--the experience points--that makes this a gamist decision.
[/quote]

Any of those might be a "true" motive, and any of them may well be claimed by the player if asked.  But by examining the intentionally fuzzy "instance of play" we can develop an opinion as to which of the behavioural modes that player is likely to adopt in the future.

If for example, this behaviour was most prevalent when the player was nearing level changes, such that small numbers of XP were worth the hassle of pursuing, I would suggest this would reinforce the interpretation that the goal of the player is related to the mechanical expression and capacity of their character.  All of these may be incorrect perceptions; nothing absolves the observer of the capacity for error.  But at least it holds the potential for independant verification - you could pass on your analysis of a players mode to other observers, who may agree or disagree with your analysis.  In my experience, players do not play significantly differently with different GM's, and to me this supports the models argument.

I have yet to read a post here that explains how behavior can lead to classification without passing through the step of impugning motive or intent; in fact, it seems to me that the classification itself is defined by motive or intent. The majority seem to agree with this; the minority can't seem to explain otherwise without falling into some hidden trap of letting motive or intent in through the back door. It is there in all the posts on both of these sides.

Not exactly.  As I see it, the claim is that ALL expressions of A motive will fall into one of three forms; we can concretely address those forms whereas we cannot concretely address motivations anywhere nearly as easily.  One of the opportunities this offers is to reverse engineer motiviation in those circumstances where none is apparent, or none that makes sense.  This, again, is opbviously unreliable IMO based as it is on speculation - but if it can be argued that, indeed, the behaviour is consistent with a certain observable mode then it is at least not baseless speculation.  This has allowed two things to date: people have said "aha, now I understand what so and so was thinking better" and as someone has recently posted, understanding a given games mode expectations helped them enjoy a game for what it was better than they had done previously.  I consider both of these successes based uppon the fact that the model is deliberately limited to actual behaviour rather than the motivations which are claimed to lead to such behaviour.

Lots of people assert that the motive for gamism is competition; I disagree as to this motive.  I think it is based on "doing" rather than competing, but that is an argument for other times.  But in proposing an alternate motivation for gamism, or challenging an existing perception of gamism, and whether rightly or wrongly, I am not challenging the EXISTANCE of gamism.  Conflicting interpretations may well produce differing mechanical structures, but I think we'd have a better chance of a constructive discussion.

[deadpanbob]

Even if the theory is wrong in some siginificant way, going out and accumulating the data which disproves it would lay the groundwork for whatever model, if any, should succeed it.  We will necessarily have to speculate on the motivations of players in this process; we will have to ask questions, discuss our experiences, explore our own motives and those of our players - but to priviliege those aspects above what is observable would be to use the least reliable available data.  When we see behaviour, we are seeing motive expressed - but it is the behaviour which is accessible.

Contra,

I don't think the model is wrong.  I also agree that the model provides us with a much more stable common language for and better tools for productive discussions about roleplaying instances, and how/why certain game designs may or may not support particular modes of play (behavior) for a given instance of play (or more likely over a series of Instances).

But, the behavior that is exhibited by RPGers in my experience isn't that easy to classify - even over long periods of observation.  I've been playing with the same group for almost ten years now, and I'll be darned if I know exactly what their behaviors indicate about their preferred mode(s) of play.

Sure, I haven't had much current experience observing them play filtered through my newfound (yet strangely limited) understanding of GNS.  But still, I have distinct memories of how they behave, and more than that a general sense of what they'll react well to and what they won't.

The problem is that their behavior exhibits pretty equal characterisitcs of all three modes of play.  Is this even possible?  I find it much easier, through behavior alone, to mark their stance preferences.

Perhaps I should start another thread to discuss what behaviors correspond to which modes of play.  Because if the model can be applied using only behavior - then we should all be able to come to a consensus agreement about which behavior maps to which mode.  Because it sounds to me like the 'behavior only' camp is arguing that behavior is a more rigorous method of assessing GNS procilivities.  But this is where I have the problem - I suspect that we can't come to any agreement about more than just the most obvious behaviors.

My point of view on this is that the behavior is important (perhaps even more so than motive/intent), but that the model works better, for me, when I filter through the necessarily arbitrary imputation of my players motives.  I think, again just for me, that asking my players about their specific Game Motives and getting an explicit understanding of them makes the model work even better.

My biggest problem is: I'm designing a game, and I'd like to be able to get some playtesters outside my circle to try it out.  Chances are I won't be there to observe their behavior - most likely I'll only get a written journal about the game (or at best a voice recording of the session).  From that scant evidence, how do I determine that my game is meeting one of its core design goals: Encouraging/supporting Narrativist and Gamist modes of play (in almost equal measure)?

Cheers,

Jason

[Mike Holmes]

I've made a great mess with my baby kobolds, haven't I. I should have left it to Ron. My point has been all along that I think Ron is correct. I hold no third position. If I've done a poor job of it, then my comments should be ignored, not disected.

That said, and at the risk of further complicating things, the baby kobolds example, is, to me all behavior. The definition of Gamism, is making decisions based on in game metrics or some sort of striving by the player. Or something to that effect. So, doing something for EXP is the behavior known as Gamism. If Ralph wants to call that a "Game Motive" then great. But that's just relabeling something we've already got a label for. Motives, I assumed would be what caused the player to "Make a decision based on in game metrics or some sort of striving".

This is entirely semantic. I see making a decision that follows certain criteria as a behavior (in fact it seems the very definition of a behavior). Ralph sees it as a motive. The point of our "side" has been and always will be that looking at the specific decision, what is observable most offten, and most importantly, is the behavior itself. And that we don't have to look any deeper than that. And in that, Ralph seems to agree. We don't need to look at what he calls "Psychological" motives.

So, since we seem to only disagree on terminology, I can only attack it sematnically. To which I'd say that unless we adopt Ralphs coventions, and even possibly then, calling the isms motives will be confusing to people who will think that we are looking beyond the "game motives". calling it behavior would certainly serve to maintain the distinctions between what certainly is notable behavior, and what are most certainly motives (Ralph's "psychological motives").

So that's my vote on what terminology to use.

Mike

P.S. Is it just me, or does it seem that in a roundabout way we're just rehashing the age old debate between Behavioralism, and Personality theory? And if so, isn't that a good reason to drop the debate (I certainly wouldn't claim to be an expert on any of this).

[Marco]

I think the crux of the debate for me is: "How can you identify, say, Narrativist behavior without a dead-giveaway (the player goes "mmm ... this will be good for the story")?"

If you can't then using GNS to analyze anyone else's play is an exercise in seeing what you want to see.

If you can, I'd like to know how.

If you're pretty sure you can most of the time, what are the objective-tells?

-Marco