The Forge Archives

GNS and "Congruency"

This idea was inspired by all the recent GNS discussion using examples of individual game decisions that are ambiguous in terms of revealing the GNS decision-making preferences of the participants who made them. Instead of regarding this common situation as merely an inconvenience for GNS analysis, I believe it could be a phenomenon useful to focus some attention on in its own right.

Quote from: Mike HolmesJust as a point of clarification (which may, ironically, confuse the issue some), quite often the same decision can be made for different reasons. I know I point this out all the time, but it is important here. It is one of the reasons that it is hard to determine the nature of a decision by observation. Quite often through coincidence the same decision could have been motivated by more than one of the three motives. Not always, sometimes it's very obvious. But quite often.

Such observations in recent threads have been particularly helpful in developing this concept, which I've chosen to term congruency.

Definition and Usage:

A congruent decision is a decision made by a participant (GM or player) during play that cannot, on the basis of the visible behavior resulting from the decision, be categorized as belonging to a specific mode of decision-making enumerated by the underlying model. In the context of the GNS model there are exactly four possible congruencies, representing the four combinations of two or more modes for which a decision may be ambiguous.

This definition refers only to visible behavior, not underlying motive or any other unknowable quality. Thus, such questions as whether or not every individual decision "really is" purely G, N, or S in nature, or whether it can be "43% this, 57% that," are irrelevant for purposes of discussing congruency.

To specify a particular congruency, we could say that a decision is "congruent with respect to" or simply "congruent with" any two (or all three) modes. For shorthand we might use (and I will use) e.g. "S-N congruent" for "congruent with respect to Simulationism and Narrativism." Congruency is reflexive; there is no distinction between e.g. "G-S congruency" and "S-G congruency." However, for a further shorthand, when speaking in the context of discussion of one of the three modes, it should be OK to say that a decision is e.g. "S congruent" meaning congruent between the mode specified (in this case, S) and the mode under discussion.

Keep in mind, however, that even when speaking of congruency between two specific modes, the third mode cannot be completely disregarded. For example, a decision that is unambiguously Gamist is not S-N congruent. To be S-N congruent, the decision has to be either unambiguously not Gamist, or ambiguous with respect to Gamism (in which case it could be G-S-N congruent, which also counts as S-N congruent).

A congruent decision, by definition, does not influence the long-term judgment of whether the participant's overall pattern of decision-making falls into one or the other of the modes with which the decision is congruent.

By analogy with the GNS modes themselves, we could also describe a system or practice as congruent (or more specifically, as e.g. G-S congruent) as a verbal shorthand meaning that that system or practice promotes or rewards decision-making by participants that is congruent in the way specified. This purely definitional and is not meant to assume that such systems or practices exist.

Incongruency

The opposite of congruency can only be termed incongruency, which sounds like a bad thing but, in and of itself, isn't. An incongruent decision is nothing more or less than a decision whose visible expression provides evidence that the decision-maker's pattern of decisions conforms to a specific one of the three GNS modes. (It doesn't have to be conclusive evidence. If a player were to make a series of incongruent decisions, some clearly simulationist and others clearly narrativist, we might be unable to characterize his overall pattern of decision-making, but each individual decision would provide some evidence in favor of one or the other possibility.)

In practice, incongruent decisions appear usually to be ones where an observer can tell that there were other options open to the player which, if chosen, would have advanced different GNS goals than the choice the player actually made. Thus, what makes a decision incongruent (or congruent) is not just the decision itself, but also the circumstances (especially, other choices that were available) under which the decision is made. For that reason, it's acceptable and understandable to describe a decision situation in which the decision has not yet been made as congruent or incongruent, as long as it's understood that this is a loose usage that really means "likely to result in a congruent/incongruent decision."

Congruency versus Coherency

Congruency and coherency are not the same thing, though they are related. Definitionally they exist on different levels: coherency is an emergent quality that applies to an entire system, group, or large unit of actual game play, while congruency applies to individual decisions. Nonetheless, in the application of theory, the concept of coherency/incoherency "reaches down" into the realm of individual game mechanisms or practices, while the concept of congruency/incongruency "reaches up" into patterns of decision-making and the in-game situations in which they occur. It's likely they're going to meet and coexist somewhere in the middle.

Examples of poor play resulting from an incoherent system, and examples of excellent play in a coherent system, all seem to begin with a player making an incongruent decision. In the former case, the decision conflicts with the goals of other participants, representing dysfunction. In the latter case, all the participants have compatible goals so there is no dysfunction. In fact, the others generally take pleasure in the fact that the decision-maker made a choice that advances everyone's goals even though other choices were available.

A useful concept?

If congruency is outside the control of system designers or game participants, then it is a useless concept.

If congruency cannot be altered independently of coherency, then it is functionally equivalent to coherency and therefore useless as a separate concept.

I believe that neither of these is the case. I believe that certain techniques and design considerations bear directly upon congruency, and that they are useful in two ways:

1. To attenuate the dysfunction caused by an incoherent system or a group of participants with incoherent goals, by reducing the occurrence of incongruency during play.

2. To adjust the level of metagame or other forms of self-conscious decision-making in coherent play, in either direction, by controlling the occurrence of incongruency during play.

I further hypothesize that various forms (and especially, the most successfully functional forms) of vanilla play, abashed play, and drift will prove to be characterized by rules and practices promoting congruency.

The interesting questions, I believe, are whether there are other techniques to be discovered, or if the known techniques can be applied in new ways such as embodying them in system designs.

Example 1: OOC knowledge and G-S congruency

A GM running a fantasy game plays out an encounter with vampires. A player-character recognizes the obvious vampiric demeanor of the vampires, and immediately acts upon the strategy of fleeing until sunrise, then finding the vampires' coffins and staking them. The GM is displeased because he believes that the player-characters, being from outside the European-style portion of the world in which the vampires exist, should not be able to recognize the vampires for what they are, let alone already know the best way to destroy them.

Clearly there are GNS coherency problems at work here. Why does the GM not want the player-character to act on OOC knowledge? Because he believes that the goal of play is to create verisimilitude, and besides, the vampires were supposed to be difficult opponents that would hook the players into exploration of a really cool storyline. Why does the player want to act on OOC knowledge? Because the game system rewards him for dispatching enemies in the most effective possible way with the least risk.

But even though the specific problem stems from lack of coherency, it could be solved on the level of congruency. If the vampires in the encounter have to be straight out of Hollywood, then the GM can easily make sure the player-characters are aware of the same vampire legends the players know. (He would also allow for that knowledge when designing the difficulty of the encounter.) Then, the player's decision to try to stake the vampires would be a completely G-S congruent one. Or, if the GM doesn't want the characters to know anything about the enemy, then he should invent a new creature with different habits and weaknesses than a standard vampire. The players will try their best to figure out how to survive and destroy the enemy, just as their characters would do in that situation, so again, their decisions will be G-S congruent.

Note that neither of these solutions, for better or worse, affects the underlying GNS issue. The GM still has Simulationist expectations, and the player is still chasing those Gamist rewards. But the specific instance of dysfunction that was making the participants unhappy has been averted.

This is only one example of dozens if not hundreds of known practices and system rules that appear to bear directly on G-S congruency. Many of these, as in the example, take the form of "don't-do" constraints, suggesting that congruency does come at a cost in design freedom. A GM who wants G-S congruency, and wants Hollywood vampires, and wants the player-characters to not already know how to kill said vampires, is just SOL.

S-N congruency

As Mike Holmes pointed out in the same post the quote above was taken from, a player acting entirely on narrativist motivations would still make decisions compatible with "what the character would do" much of the time, because a good story has to be plausible. In other words many non-Gamist decisions are S-N congruent.

What we should be interested in are the exceptions. When does S-N incongruency arise?

- When a participant makes a decision that is consistent with verisimilitude but detracts from the story (deprotagonization).
- When a participant makes a decision that is consistent with the creation of Story or exploration of Premise, but detracts from verisimilitude in some way. It might do so by being inconsistent with the character, or by invoking an explicitly metagame mechanism.

Note that opting not to make use of an available Narrativist-oriented metagame mechanism is not proof that a simulationist decision was made, as long as the option chosen instead was also consistent with the goal of creating story, so such a decision is not usually S-N incongruent. However, use of a metagame mechanism is almost always S-N incongruent.

This immediately suggests that a dichotomy of taste could arise, between those who prefer the self-conscious decision making of Narrativist metagame mechanisms and those who prefer to create story within the constraints of S-N congruency. I believe this dichotomy is already known in practice as the distinction between vanilla and explicit Narrativism. While there is a continuum between them, the distribution of preferences along that continuum appears to be bimodal, and taking the concept of congruency into account explains why. This also explains why so many examples of vanilla Narrativism are difficult to distinguish from Simulationism. Quite simply, the practitioners want it that way.

Besides metagame mechanics, other elements that could increase or decrease the prevalence of S-N congruity in a game include character design (traditional heroes will face fewer potentially S-N incongruent decisions than anti-heroes or more complex characters), choice of Premise, and the degree of realism in a setting. Attempts to portray settings that "simulate the world of movies instead of the real world" can be characterized as attempts to foster S-N congruency (though how successfully they accomplish it is unclear).

Example 2: S-N congruency and Illusionism

A now-wiser but still self-confident neonate Vampire leads her companions into the betrayer's lair for the inevitable confrontation that both sides hope will be final. Oops, the enemy just make a spectacular success roll with the Possession discipline. Looks like our vengeful vampess is going to be working for the opposition for the rest of the scene. How deprotagonizing, especially since the major Premise happens to be about the nature of loyalty.

The GM faces a decision that clearly has potential to be S-N incongruent: apply the effects of the roll, or fudge the roll to reduce the enemy's success to an intermediate result that will play into the Premise rather than override it. If he chooses to let the roll stand, it's clearly an S-N incongruent Simulationist decision. If he lies about the results of the roll, it's clearly an S-N incongruent Narrativist decision. Or is it? Let's look again at the definition of congruency: "...the visible behavior resulting from..." An omniscient observer can clearly see the decision's incongruency, as can the GM himself. But the GM isn't telling, and there are no omniscient observers on the scene. As far as the players are concerned, no incongruency has occurred. Since congruency is based on visible behavior, there is no difference between the effective illusion of congruency and congruency itself.

Thus, I believe, the concept of congruency gives us a framework for understanding illusionism in a straightforward way that is free of speculation about motivation. Other forms of illusionism appear to be similarly associated with other applications of congruency.

G-N congruency?

I think this occurs, but perhaps only in limited circumstances, such as, for example, when the Premise is about risk or fortune or fate. It's certainly possible to create excellent Story with deep Premise about a group of people who go out and hunt a big powerful monster to take its treasure (ask the hundreds of Herman Melville buffs who live in my town), and under such circumstances, at least, the possibility of promoting G-N congruency appears promising. The actual techniques would probably be similar to those of S-G congruency (including illusionism) in many cases, with more severe constraints.

It also appears that competitive storytelling games can be G-N congruent, through an entirely different approach.

As for G-N-S congruency, it would have to incorporate G-N congruency as a start. It is certainly possible for an individual decision to be G-N-S congruent. But as for attempting to maintain such congruency consistently in a real game, my suspicion is that it's theoretically possible, but the necessary constraints would be so severe that it would rarely be worth it.

----------

That's as far as I've got. I believe congruency could be a useful concept, not because it says very much that's new, but because it gives us new language to use in applying the GNS model to real-world issues of coherency, drift, and taste. This could also help in the understanding and acceptance of the GNS model, because it addresses in a GNS-consistent way many of those in-between cases and tricky examples that people keep offering up as challenges to the validity of GNS.

One more note: even as I've been writing this, others have been adding posts that are getting at the same idea. Just a few minutes ago Mytholder posted:

QuoteI'm well aware decisions are key here. I know a dramatist can make Sim decisions. I just don't think the majority of a players' decisions are "significant" in terms of GNS. It really doesn't matter if I chose to eat in that inn to ensure I don't suffer from fatigue-related penalities, or because it's a logical thing to do in terms of the simulation, or because I'm deliberately providing a plot opening for the GM. All three play styles are fully compatible with the action. It's only when the play styles are IN CONFLICT that GNS comes into it.

Congruency gives us a more rigorous definition of "significant in terms of GNS," and also finds a way to apply GNS to the remaining "ground" (in the figure-vs.-ground sense) of decision, by specifying which modes those decisions are congruent with.

- Walt

While the idea of adding more jargon to an already jargon-heavy body of communication personally fills me with dismay, I can't argue with your logic Walt and I agree that this would be useful to do in this instance.   What you are posting makes sense to me.  

Laurel

Hi Walt,

I finally, finally managed to get to this thread. I think you've posted something well worth attention. I'll begin by saying that in all practical, observational points, I agree in full.

The real question, as you know, is as you stated: If congruency cannot be altered independently of coherency, then it is functionally equivalent to coherency and therefore useless as a separate concept.

My own take on the matter is that I have been, all along, thinking of Coherence in such a way that it includes your concept of Congruency. It's still pretty hard for me to separate them, for a couple of reasons.

1) I have tried to stress that compatibility of goals, in practice, is the defining feature of Coherence. The fact that goals are most compatible when they are similar-to-identical, as well as the fact that I tend to prefer such play situations personally, are not relevant; any compatible mix of different goals ("convergence?") is Coherent too.

2) I tend to include the entire spectrum from "atomic" GNS decision, to "molecular" GNS activity (observable), to "substance" GNS profile (very observable), all the way to "group" or "object" GNS profile when I discuss these things. Or more accurately, I tend to encourage discussion at the upper end and let the lower/finer end take care of itself. (I'm working on some material to clarify this issue, to myself as well as to anyone who's interested. I'll be posting that when it's done.)

Now the real question is whether my own ease of combining techniques with outcomes in #1, as well as my inclination to discuss mainly the upper-end (observable, functional) of the spectrum in #2, have been causing problems in discussion. If so, then Congruency as a concept would be the perfect solution.

What I need to nail down is, if we use Congruency as you've defined it, what need is there for Coherence, as a term? I'm kind of chewing it over, personally, not goin' one way or another. Can you clarify that for me, or provide more examples of how the two terms might interrelate in practice?

Best,
Ron

Quote from: Ron Edwards
2) I tend to include the entire spectrum from "atomic" GNS decision, to "molecular" GNS activity (observable), to "substance" GNS profile (very observable), all the way to "group" or "object" GNS profile when I discuss these things. Or more accurately, I tend to encourage discussion at the upper end and let the lower/finer end take care of itself. (I'm working on some material to clarify this issue, to myself as well as to anyone who's interested. I'll be posting that when it's done.)

Now the real question is whether my own ease of combining techniques with outcomes in #1, as well as my inclination to discuss mainly the upper-end (observable, functional) of the spectrum in #2, have been causing problems in discussion. If so, then Congruency as a concept would be the perfect solution.

Speaking for myself I think, you've hit on the source of the problem I had for a long time.  You understood that GNS was formulated on the atomic level, and could freely translate the concepts up to the "molecular" and "subtance level".  But for me (and I suspect based on my own observations, many of us) I didn't see the translation going on and thought that GNS was formulated at the "substance" level...which in turn led me skim over and pay less attention to those times when you brought up "decisions".

Mytholder and I had a very extended thread which basically boiled down what Walt skillfully summarized as Congruency.  If we'd had the vocabulary of Congruency at the time of the thread as it relates to GNS, we probably could have spent less time circling each other, and more time moveing the thread forward.

Quote
What I need to nail down is, if we use Congruency as you've defined it, what need is there for Coherence, as a term? I'm kind of chewing it over, personally, not goin' one way or another. Can you clarify that for me, or provide more examples of how the two terms might interrelate in practice?

The way I see it from Walt's description is that they operate at different scales.

Congruence, as he's defined it, is strictly an atomic level phenomenon.  It allows us to identify Incongruent decisions that can be identified as occupying a GNS decision from the congruent decisions that could be more than one position and which can't be determined.  This is what Mytholder was calling "significant" and we were representing as blanks in the decision maps we were drawing (the blanks being congruent and the non blanks being incongruent...of course, in Walts model we'd need to use something other than blanks to represent the 4 different types of congruency).

Coherency on the other hand looks at the pattern of incongruent decisions over the period of the game.  If the pattern of Incongruent decisions is dominantly one position, N----N----N--N, for instance, then the game is N Coherent.  If the pattern of Incongruent decisions is a mix of GNS positions, N---G----N----S---S---S---G, for instance, then the game is Incoherent.

Whether an Incoherent game is dysfunctional or a functional hybrid would need to be determined using other tools.

Needless to say, I'm a BIG fan of this method of analysing GNS.  Walt managed to summarize alot of concepts I was batting around with Mytholder very succinctly.

Hey, wait.

Isn't coherency at the game-text level, I mean at the V:tM or Sorcerer or Smurfs: the Smurfing level, and doesn't it mean that all the game's rules plus its hype work together well to drive the game?

Like an N-coherent game is one where you can make consistently N decisions without having to compromise with the rules, right?  The game supports you in your consistently N decision-making.

If so, congruency and coherency aren't that closely related at all.  I think congruency is more related to (but opposite to) 'perviness,' as in "I'm a pervy Narrativist."  

A highly congruent N-S game is one that supports decisions that might be N, might be S.  Sorcerer, I'd say as an example, and having never played it, is more congruent than The World, the Flesh, and the Devil, because it's damn hard to make a (game-rule significant) decision in the WF&D that isn't Narrativist.  Sorcerer on the other hand has all kinds of support for Sim decisions, like stats with numbers and things.  Yes you can/must/will use them to drive a story, but any given Lore roll (again just talking out my butt) might be Sim instead, hard to say.  I fail my Lore roll -- is what happens because it makes for a good story, or because it's consistent with the world-sim?  Who knows?  Hence: congruent.

('Pervy' is kind of the opposite of 'vanilla,' right?  Which would make 'highly congruent' and 'vanilla' into cousins, which sounds right to me.)

-Vincent

Sorry about this, but here's what I mean.

A coherent, incongruent ('pervy') game like the WF&D looks like this:

N--N-NNNN-N--N-NN-N-

A coherent, congruent ('vanilla'?) game like Sorcerer (I speculate) looks like this:

N----N--N---N---N----N

An incoherent, congruent game looks like this:

------N--S-GG---N----

And an incoherent, incongruent game looks like this:

G-NNS--GGS-SN--G-NNS

Where - is a decision that an observer can't tell by looking at it alone whether it's S, G, or N.

Oh and which, rereading it, is almost exactly what you said, Valamir.

-Vincent

Hi Vincent,

Coherence, like nearly all the terms in the essay except for system-tools, refers to play. Just like "Gamist RPG" refers to "game whose rules facilitate Gamist play," "coherent RPG design" refers to "game whose rules facilitate coherent play."

From the essay: By "coherence," I mean the degree to which a group of people can hit upon and sustain a shared Premise ... - and by definition, continue to enjoy the social role-playing activity consistently.

It so happens that I claim, in practice, that GNS-focused game designs are more reliably coherent, but that's not a definition.

So if I'm not mistaken, coherence exists as an end-product - the point is whether Congruency exists as a means to it that can be identified in some useful way or level, or whether it's a grab-bag, possibly-unnecessary synonym for "play which facilitates Coherence."

I agree that techniques for that kind of play, which successfully resolve potential incompatibilities among (a) within-mode differences (N vs. N'), or (b) among-mode differences (G vs. N vs. S), deserve a lot of attention. But do they deserve any name but Coherence-preserving or Coherence-creating techniques? Considering that that's what they actually do?

Just to be clear, I'm not trying to scrub out Walt's suggestion regarding the term Congruence. I want to dissect out the topic with great care, because it's important.

The most important thing, of course, is what it seems we all agree on: that the level of GNS application between one person's decision and successful group play does need to be brought more into the light, both in the essay and during discussions.

Best,
Ron

Um, I'm having problems with the terms selected. Congruent means something like "Similar to" or parallel. Having some qualities in common. While Incongruent means the opposite. Having dissimilarities.

So did I misread the definitions above, or are these terms being used in a really odd fashion? This seems totally counterintuitive to me.

Wouldn't it be congruent = behavior that adheres to one GNS style, and Incongruent = behavior that does not adhere to a single style?

If this is the case then I can see using congruency to say something like "the players' deisions being congruent with Gamism led to a coherently Gamist experience." In this case indicating the atomic level discussion of behavior as it relates to an entire game experience (a "molecular" level event). As Ralph and others have intimated it might be used.

Or am I just confused?

Mike

At this point I'm seeing them as 2 completely seperate ideas trying to define 2 seperate (but related) concepts.

Whether an individual decision is Congruent or Incongruent does not tell you whether the resulting game with be Coherent or Incoherent, therefor I don't think we can see Congruency as being "Coherent Preserving" or not.

That said, I think the very definition of Coherency needs to be tightened up.  For instance in the part of your essay you quote above you say "By 'coherence,' I mean the degree to which a group of people can hit upon and sustain a shared Premise ... - and by definition, continue to enjoy the social role-playing activity consistently"

There are two weakness with this definition I see.

First, the last part very strongly implies that a game needs to be coherent in order for players to enjoy the activity consistently.  This is at odds with the comments you made in another thread of mine regarding Coherency, where you acknowledge the potential for Functional Hybrids.  

Second, I have difficulty with tieing the concept of Coherency back to Premise.  One reason is that there are several different types of Premise discussed in the essay (a seperate pet peeve of mine), but there is no indication of which form of premise leads to Coherency.  Another is that in practice, Coherency has been used to refer to a game which targets a specific GNS position consistently.  Thus, I fail to see the purpose of tieing Coherency back to Premise at all, when what it appears to be is an aspect of GNS positions.

Note:  I'm not saying the definition of Coherency needs to be scrapped or changed.  Just that it could use a good bit of clarifying.  It is difficult to evaluate whether "Congruence" is really the same thing as "Coherent" when I can't fully understand what is meant by "Coherent".

Ron,  

Huh.

So I've been thinking too narrowly about coherence, then.  You're saying that this:

NSS--NNN-S--NSSN---S

Might well be a coherent NS game, not necessarily an incoherent game with N and S in conflict.

Yes?

-Vincent

Mike,

Walt is using the term in the technically correct manner which seems odd on first blush.  

If a decision can be used to support both a G position and an S position than the decision is Congruent between G and S.  Thus a decision which only supports a single position in not congruent with any other decision and is therefor considered incongruent.

What I am hearing so far makes it sound like 'congruent' decisions (I agree with Mike, the term is counter-intuitive) are decisions at the atomic level that cannot be determined to be Particles G, N, or S and function as any of them.  (Perhaps a GNS uncertainty principle at work?)

The reason Ron seems to be having trouble settling them in terms of coherence is because he's looking for a relationship (much like the many people who try to connect Director Stance to Narrativism).  There isn't any.

What we seem to be talking about here is what O-type blood is once it's in the blood stream.  You infuse O-type into a person with AB blood and what do you get?  A living person with AB blood.  The same goes for all the GNS modes and hybrids.  An O-type decision (one that is congruent) disappears when you consider the overall flow of the game.  It's the AB blood cells that tell you what blood-type a person has, no matter how much O-type blood has been infused.  (Note; the patient is dying if they have too many different blood-types mixing in their veins - that is incoherency - the amount of O-type blood makes no difference.)

So basically O-particles work as any of the types as needed and they are not considered in the search for coherency.  If a molecule has O-particles and G-particles, then it is a G-molecule; the same goes for S-particles and N-particles.  This continues to carry forward all the way up to the 'substance level.'  If a substance is made primarily of G-molecules, no matter how many O-particles it has, it's the G-element fully coherently.

The only reason I see this as an asset is that it allows us to say, "Oh that's an O-type decision, we cannot determine GNS-state from that" - a mechanism to agree to disagree more readily and move on.  A patch if you will to the 'crunchiness' of the GNS.

Fang Langford

A few quick points:

1. I will hereinafter try to stick to "congruence" and "coherence" instead of the alternatives ending with y.

2. The thread with Mytholder that Valamir refers to was what gave me the idea to pursue congruence as a concept. I found that discussion very interesting despite (and in part, because of) the communications difficulty it revealed.

3. This is what... we were representing as blanks in the decision maps we were drawing (the blanks being congruent and the non blanks being incongruent...of course, in Walts model we'd need to use something other than blanks to represent the 4 different types of congruency).

Probably not necessary in most real-world cases. Any type of congruence means you cannot characterize the decision specifically as G, N, or S; the only difference between the types is whether one of the three possibilities (and if so which one) is ruled out.

Now, to the main issue, which Ron has not surprisingly homed in on. It's irrelevant that coherence is defined differently from congruence if in practice they amount to the same thing. So, do they?

Quote from: Ron EdwardsI have tried to stress that compatibility of goals, in practice, is the defining feature of Coherence. The fact that goals are most compatible when they are similar-to-identical, as well as the fact that I tend to prefer such play situations personally, are not relevant; any compatible mix of different goals ("convergence?") is Coherent too.

"compatibility of goals, in practice"... but where's the emphasis? Does this mean compatibility of goals with each other in the context of a particular practice, or does it mean compatibility of the goals with the practice itself? I always read it as the latter. (It's almost but not quite the same thing... one single goal must always be compatible with itself, but could be incompatible with the practice.)

If it's the latter, the congruence has no direct definitional relationship to coherence, but a logical connection can be made that congruence implies coherence. If it's the former, then congruence is by definition one form of coherence. In either case, though, congruence is far from synonymous with coherence itself.

From what I've seen, advice stemming from GNS theory usually emphasizes achieving coherence in ways other than promoting congruence (e.g., pick a single goal and focus on it).

Also, there is an "idealized" concept of coherence that may not be implied in the original definition but is, I believe, often read into it. Idealized coherence raises the bar from compatibility of goals with each other and/or with the practice, to active promotion of the desired goals by the practice. (Especially when one primary goal is identified, compatibility of goals no longer appears the issue; compatibility of practices with the goal becomes paramount, and that bar can easily be raised to "promotion of" that goal.) For that idealized form of coherence, coherence and congruence become completely disjoint at the atomic level (an individual decision either promotes the main goal specifically, or it is congruent, or it promotes a different goal and is incoherent; it cannot be any two simultaneously). At higher levels this translates into a linear trade-off between the prevalence of congruent decisions versus coherent decisions, represented by the continuum between vanilla and pervy.

Phew, very abstract, all that. In actual practice I see the following differences:

- A group of players can be loosely described as coherent or incoherent based on how similar their goals are, without taking into account their practices. This may not be strictly proper, but it's done. Congruence has no meaning in that context.

- A game system can be characterized as coherent or incoherent, without taking into account the players, based on whether it promotes a compatible set of goals or not. My hypothesis is that a game system can also be characterized as congruent or incongruent based on whether it promotes congruence between different goals or not. A game might be coherent by virtue of it clearly promoting a single GNS goal type, or it might be coherent by virtue of promoting multiple cross-GNS goals but also promoting congruency between those goals. Also, it might (in theory) be coherent, despite promoting multiple cross-GNS goals, through some other means, but I don't believe any such means are known.

- A relatively small unit of game practice can be characterized as promoting congruence or incongruence, before the fact, at a level where the concept of coherence appears difficult to apply. Usually this comes in the form of realizing that a certain practice unnecessarily promotes incongruence. For example, a system might offer metagame rewards for a character to behave in certain protagonistic ways without regard for whether that behavior makes any sense for the character (unnecessary S-N incongruence), or a setting might introduce a puzzle that is easy for the players to solve but unreasonable for the characters to be able to do so (unnecessary G-S incongruence). A game system rule promoting G-S incongruence could be a definition of what a "loophole" is.

- The concept of coherence appears to be most "at home" at the level of evaluating a game system. The concept of congruence appears to be most "at home" at the level of evaluating a specific rule, encounter, scene, or other relatively small unit of play.

- Specific gamemastering practices can often easily be described as promoting congruence or incongruence. I haven't seen much discussion here of effects of gamemastering practices on coherence or incoherence. (That doesn't mean it's not a viable concept; it's the nature of this forum to look at things from a system designer's point of view.)

- Again, the main point: consistent congruence throughout a system might imply coherence at least of a sort. Incongruence, even when pervasive, does not imply incoherence. A gamist game might be filled with rules loopholes and might despite (or because of) that be a good coherent gamist game. A narrativist game might reward players for changing the character's nature during play for purely noncausal reasons, and still be a good coherent narrativist game.

- However, I would suggest that a system that promotes goals from different GNS modes, and does not promote congruence, must be incoherent.

- Walt

Fang's blood-type analogy reminds me of a point I forgot to make: since congruence applies to a particle, only a pervasive pattern of congruence has any meaning at higher levels. So when I say something like "congruence is by definition one form of coherence," I mean that pervasive congruence throughout play, not some small number of individual instances, would result in coherence.

An open question is, how pervasive can congruence be? To borrow Fang's analogy, do any "Type O" people exist, or does everyone inevitably have enough "A", "B" etc. particles that they must fall into some other type? If pervasive congruence is not a real phenomenon, then as Fang says the applicability of the concept is very limited.

I believe pervasive congruence is real and in fact fairly common, which is why specific occurrences that introduce additional unnecessary incongruence (such as a particular scene where OOC knowledge suddenly becomes an issue) cause noticeable problems in play.

- Walt

Quote- Specific gamemastering practices can often easily be described as promoting congruence or incongruence. I haven't seen much discussion here of effects of gamemastering practices on coherence or incoherence. (That doesn't mean it's not a viable concept; it's the nature of this forum to look at things from a system designer's point of view.)

Based on an initial, quick reading - "Robin's Laws" is all about congruent GM techniques.  I'll reread carefully for anything that could be brought to bear on coherence . . . just like I'll have to re-read this thread carefully to make sure I understand the concepts.

FWIW, I agree with Valamir that something about this issue is one big thing folks "trip" over regarding GNS.  Without an understanding of the subtleties around atom/molecule/substance (or decision/style/prefernce, etc.), it's easy to become convinced that "GNS doesn't apply to how I play games, so it's bunk."

Gordon