Hybrids - looking for an understanding of

[Stuart DJ Purdie]

I've been thinking about what a hybrid is lately, particularly after
Decision Merging: An Appoach to Hybridization. What it mostly inspired in me is confusion over what a hybrid is.

My original working defintion was:  A game that, without modification, may be played without Drift, in more than one metagame GNS mode.

Checking the original sources, I found the Ron's 'System doesn't matter' doesn't mention hybrids.  In 'GNS and other matters ...', a definition of hybrid is given as:

two modes can exist side by side in the design, such that differently-oriented players may play together

This is in full agreement with the definition implicitly used in the recent thread mentioned above.  My working definition was a subset of the above, and thus some of the problems start to be seen.

However, in the recent Simulationism eassy, the definition is given as

A game whose rules include facilitating elements for more than one mode of play. Observed functional hybrids to date include only two GNS modes rather than all three, and one of the modes may be considered primary or dominant, with the other playing a supportive role.

At this point, I'm rather confused.  If a hybrid allows two players with differing metagame priorities to play together [0], then how can one be subordinate to the other?  Surely that must mean that the player playing with the subordinate metagame priority will be de-protagonised compared to the player playing with the dominant meta-game priority, assuming each player has a fixed metagame priority?

Additionally, if in a hybrid a player's metagame priority is allowed to wander between the two GNS modes, then, if the supportive mode is Sim, meaning that expolration of some aspect is prioritised,  how does that differ from non-hybrid play with either a Gamist or Narrativist goal, where it is well accepted that exploration is integral to all RPG's, and constitutes an important part?

Stuart

[0] Aside: I'm not convinced that this is a good situation.  But that's for another thread.

[clehrich]

My original working defintion was:  A game that, without modification, may be played without Drift, in more than one metagame GNS mode.

This seems to me the correct definition.  In the thread Pan Reward-ism there was an alternate definition proposed, at least implicitly, which is of a game that allows the players to select an appropriate dominant GNS mode.

If a hybrid allows two players with differing metagame priorities to play together ..., then how can one be subordinate to the other?

Seems to me this is again the question of "instance."  If at one instance one player's preferences are dominant, at another they will be subordinate.  It's not throughout the entirety of the game, but rather instance by instance.

Additionally, if in a hybrid a player's metagame priority is allowed to wander between the two GNS modes, then, if the supportive mode is Sim, meaning that expolration of some aspect is prioritised,  how does that differ from non-hybrid play with either a Gamist or Narrativist goal, where it is well accepted that exploration is integral to all RPG's, and constitutes an important part?

Ron will (I hope) correct me here, but my sense is that in Sim play, the ideal is such that no meta-game priorities enter the equation, only Exploration; while this cannot be functionally true, it is a goal which can be striven for, whereas in Gamist and Narrativist play one must recognize the non-exploration meta-game priorities.

Still, I do think that the question of hybrids raises a number of serious difficulties for the model.  I don't think hybrid games break the model, but they do demand a certain amount of precision that is ordinarily lacking, particularly between the issue of coherence at the design level and the selection of Creative Agenda at the instance level.

Chris

[MK Snyder]

Sailing off into another direction quickly, it's my observation that most players have a hybrid style of play; for them, the corresponding game that supports their style meets their needs.

Since one dimension of prioritizing is the frequency of choices made in a session by a player (the choices of choices, sorry) two players with differing "hybridizations" can play together, both drawing on the choices offered from the same game.

That is, the "Sim/Nar" player will make "Sim" choices more frequently than the "Nar/Sim" player from a game with oh, 50/50 split between Sim and Nar elements.

I don't think "hybrid" should be at all interpreted as perjorative.

[Wormwood]

So far, in what I've seen there are three main ways a game can support multiple modes coherently:

1) A game can manage drifting, allowing different groups to adjust fluidly to different modes. Essentially this is facilitated drift.

2) A game can permit multiple modes, and is design (intentionally or accidently) such that an arena of decision catagories do not interfere with each other. This seems to be the standard idea of congruence.

3) A game can permit multiple modes and is designed (intentionally or accidently) such that some arenas of decision catagories are indistinguishable in terms of decisions reached.  This is the idea I mentioned before about decision merging, it's also a form of congruence, although one that didn't seem mentioned in particular.

The sense I got was that technically the later two are functional hybrids, but I'm unclear if the first is considered such or not. In fact it seems there is a general confusion on that matter.

The idea of subordination seems directly related to the scope of decision classes (i.e. the portions of the general class of decisions which fit a given mode, which are subsequently supported by the game). A game which subordinates one mode but permits both to be played has a larger scope of decisions classes belonging to the dominant mode, and fewer belonging to the subordinate one.  In both cases it is vital to realize that almost no games support a given mode fully. Rather a given game supports a variety of decision classes which are predominantly of it's primary mode.

(As I write this, it occurs to me that I may be using the term decision class for what is typically called Creative Agenda, blame the mathematician in me.)

Well I hope that's food for thought,

   -Mendel S.

[clehrich]

3) A game can permit multiple modes and is designed (intentionally or accidently) such that some arenas of decision catagories are indistinguishable in terms of decisions reached. This is the idea I mentioned before about decision merging, it's also a form of congruence, although one that didn't seem mentioned in particular.

Are you referring to some particular thread?  This is an interesting conception, and one I'd like to follow up on.  Can you give a reference?  Regardless, can you elaborate a bit, especially on the distinction between this and #2 (which I think I understand in general, but I'd love more detail about how you formulate this)?

Chris

[Mike Holmes]

MK, definitely Hybrid should be seen as positive. Incoherent means that due to poor design more than one mode is supported in such a way as to make it difficult to impossible to play due to the supporting rules interfering with each other. That's the pejorative term.

I think people have misinterpereted caveats that attempting hybrid designs is difficult and often becomes incoherent, as saying that they're bad. On the contrary, from one POV, that being that presumably they are more able to support a wider variety of players, they are a great goal. Hence the love of TROS. This isn't to say that "single mode" games aren't a good idea as well, just that it's a design choice.

I've said, and most agree, I think, that most players shift between modes with great regularity. Rare is the "mono-mode" player, IMO. In fact, I'd go so far as to say that all players can enjoy all modes, and that dislike of certain modes is mostly due to player history encountering modes that were either presented dysfunctionally, or in such a way as to quash other favored modes.

Chris, Mendel, I'd say that Congruence, and similar tactics like Mendel has in the other recent thread, are tools to achieving good Hybrid designs.

Mike

[Wormwood]

Chris,

I was referring to the reference on the first post on Decision Merging. In that thread, I have at least one example of a game design which fufills this idea, at least significantly.

Implicit in some of the discussion of congruency is the sense that there are two dangers in inconsistent modes (or decision classes in general). First is the failure of the game to manage the different approaches, and so a break down of the systems that help support each character. Second is the meta-game discord of incoherency. This is possible even in a congurent system. Put simply, even if two different decision classes do not conflict in actual play the holders of the two classes may view each other's actions as problematic regardless. Convincing the simulationist that the die-hard gamist isn't going to tear down the setting's consistency, and convincing the gamist that the simulationist will be an interesting competitor are both non-trivial.

The distinction I see between 2 and 3 is that the later is a subclass of congruency which:

a) is easier to design and fully implement (i.e. it can permeate a design, rather than just be an option in a larger design goal).

b) helps address meta-game conflict between decision classes.
c) attempts to bring to light a form of trivial congruency where approaches do not conflict due because they cannot be distinguished on the play level.

In some sense, I view this as designable congruency, although it is certainly possible to design games with other forms of congruency.

I hope that helps,

   -Mendel S.

[Jason Lee]

If memory serves...

#3 is actually the definition of Congruent - as defined in the GNS and Congruency thread.  

Grr...which thread?  This one or Pan Reward-ism?  Well, I'm already typing.

The idea I threw out in the Pan Rewardism thread was an attempt at something Congruent.  I think the key to Congruent play is priority non-interference.  I think any mechanics designed to force another playing into a mode will make it go splat.  I cannot recall where (maybe the Star Wars threads in Actual Play), but I do recall some discussion of Congruent play from the GMing angle and how Robin's Laws basically amounted to techniques to accomplish that.  I think this is also key to actually getting Congruent play - GMing senarios specifically with that intention.  The book can encourage this kind of behavior, but a lot of it rests in the group's approach.

#2 is the common definition of a hybrid game - the combat system might be Gamist, the narration mechanics are Nar, etc.  Sort of like individual game elements are mini-games/wrappers that don't overlap with each other if they conflict in priority...or something like that.  I think hybrid is generally used more loosely though.

#1 seems like the definition of Abashed.


Correct me if my definitions are off.

******

I've been having this brewing idea that system level conflicts can be defined as either and Exploration or Stance conflict.

An Exploration conflict is where one Exploration element is railroaded as variable control in order to facilitate exploration of another.  The conflict  occurs when a player wants to explore the controlled element.  As you cannot explore something being railroaded a player/player power struggle insues over the element.

A Stance conflict is just a power struggle over something, like two players thinking they both have Director power over the Cheetos.

Every other GNS conflict that occurs to me seems to be some sort of taste preference, a more social conflict.  Sam doesn't like it when you're inconsistent, Ned hates your lack of theme, Gary just wants to kill the damn orcs, etc.

Which all makes me think GNS actually sits above Exploration, closer to the Social level....but I'm veering way off course now.

*****

Anyway, I think the system can design itself such that Exploration and Stance conflicts to do occur, but other conflicts are kinda left needing to be encouragement about how to structure the social contract.

Who knows if this holds water...just a thought.

[Ron Edwards]

Hi there,

It's always helpful to go back to the real variables at hand, which are features of actual play.

QUICK SIDE POINT
1. Number of modes per game group, given a single instance of play to look at. For present purposes, let's call an instance a session, with the proviso that stuff happened during that session and people got into it. So, number of modes per game group means different people with different goals/priorities of play.

2. Number of modes per person, again, given the same amount of time and content of play. Here we're looking just at one person.

I'm happy with letting #1 vs. #2 above just be an irrelevant variable; if Bob's Narrativism aids and abets Sam's Gamism, then cool, and similarly, if Bob goes on a Gamist toot which sets up and abets his later Narrativism, then also cool. I took the time to clarify the difference, though, so it doesn't become a red herring regarding my next points.

In both cases above, I'm not interested in little piddly flashes or details - if for thirty seconds, someone feels kind of odd that the Schmeisser submachine-gun fired for X combat rounds without a chance to jam, I really don't care, given that the rest of the session he or she was completely jazzed about the chance to level up faster than the person to the left.

TO THE POINT
So what makes hybrid play? When we look at that one instance and see that Mode X and Mode Y were firing on identifiable cylinders in some kind of functional (= fun) way. Hybrid play is therefore a subset of Coherent play, just as single-mode play is. I make no apologies for considering Coherent play good, and Incoherent play bad; the terms are inherently judgmental in my articles.

Now, what's all this about one mode being subordinate to the other? A lot of people forget that I am very uninterested in internal states, i.e., the GNS inside one person's head (if any). I'm interested in the communicative, social interaction among people; that's the only place I look for GNS in action. Therefore ...

I suggest that when zones or spheres of GNS-incompatibility arise during play, then one has to give way, in order to preserve the social fun. I also suggest that when hybrid play is going on, these chances for incompatibility crop up pretty regularly. If we're talking about Sam and Bob, side by side, when their characters are dealing with the same stuff in a way which affects them seriously, then either Sam's priorities will have to give way to Bob's, or vice versa. If we're talking about Bob by himself, then the same applies to his own different modes.

See, that "Bob by himself" bit is important. He's not really by himself. He's playing with others. If he just pixie-leaps about among Gamist and Narrativist priorities, it's confusing to everyone else, and in many cases will screw up their respective participations, if they can't tell which is the "real" goal.

Here's my claim, then: that over the course of that instance of play, and remember that it probably includes dozens of decisions and interactions among everyone at the table, play is functional insofar as everyone knows what the hell everyone else is up to, in GNS terms. I am talking about all play, including and perhaps most especially including play among people who never heard of GNS and wouldn't care if they did. So the idea is that if more than one mode is identifiably operative during the instance, it's not because people do a 180 every few minutes, but because what looks like a bit of a 180 is actually serving the interests of the "larger" direction.

AND NOW ABOUT TEXTS
Looking over my essays, I'm seeing that I made a mistake in restricting most of the hybrid discussion and definitions to design, rather than play. With any luck, I just corrected that.

Let's take a look at The Riddle of Steel. We have some hefty Simulationist elements of play, both in terms of "realism" and in terms of the player getting into the character's head (specifically offense and defense during combat). Yet numerically, the Narrativist elements of play so greatly override nearly any possible decision made in line with the Sim mode, that there's no way, without Drifting, for play actually to climax based on Sim priorities. Those Narr elements, the Spiritual Attributes, can only serve the player priority of whatever it is play is supposed to be about, thematically. They don't even operate as a constraint, as the player may change them almost at will.

The Simulationist elements of this game provide a chassis of gritty causality which is then rather brutally kicked and beaten, through the sheer weight of dice bonuses, into a thematic engine for players' "statements" regarding what's important, in ethical and moral terms ("values," if you will, Aidan). The Sim serves the Narr. The Riddle of Steel promotes, in my view, remarkably good Coherent play, specifically Narrativism with strong Simulationist hybrid support.

Again, though, this is all design. It doesn't speak to those whose standards for increasing the Spiritual Attributes drop below the recommendations made in the text, or to those who limit how frequently they may be spent to 0 for character improvement (which also opens the door for changing the SAs' content). It doesn't speak to those who simply throw the SAs out, when they rightly recognize that these rules practically cannot help but give Narr play a constant headlock on the Sim play.

Before I go on, Stuart, is this making any sense? Addressing any points or issues that led you to start the thread? Is there anything I'm failing to recognize about what you're asking?

Let me know so my next post can stay on the track that you're on.

Best,
Ron

[Walt Freitag]

Hi Mendel,

Technically, only #3 is a description of congruence. The definition of congruence is precisely the indistinguishability of which decision-making priority is being used to make a decision or a series of decisions.

Consistent congruence throughout play would therefore in theory obviate any internal or external failure to handle or accept a player's decisions, based on the simple principle that if no observer can tell what the decision-making priority was, the decision cannot be in any meaningful way incompatible with some other expected priority.

If decisions are identifiable as to their priority, but they manage to somehow not interferere with other players' different priorites, then something other than congruence is going on -- most likely a form of segmentation of play, where each players' imaginative space is shared primarily with the GM and only minimally if at all with the other players'; but other possibilities have been discussed, such as deliberately crafted "asymmetrical" play where one player's e.g. Simulationist agenda actually supports or enhances another player's e.g. Narrativist agenda and vice versa.

Other than this terminological fix (which makes your point simpler; what you're talking about as a "subclass of congruence" is in fact the entirety of congruence and a subclass of hybridization techniques in general), I agree with everything you've said, except that pervasive congruence is easier to design than other, mostly undefined, forms of functional hybridization. I haven't seen enough congruence-promoting design to come close to being able to gauge its relative difficulty.

On the previous thread I mentioned two examples of pervasive Gamist-Simulationist congruence from actual practice: "Save the world" scenarios, and traditional dungeon crawling with typical traditional "chaotic greedy" characters. In both cases, character behavior exhibiting maximum fidelity with the imagined game space will be all but indistinguishable from character behavior supporting a Gamist priority on mastering the challenge. In the first case this is due to a specially constrained situation (no sane character given the need and the ability to save the world would act other than to attempt to do so to the best of his or her ability); in the second, a special constraint on character personality.

While in practice these examples of congruence are generally achieved by adding constraints within a more general system, there's no doubt that systems could be designed with the same or similar constraints built right in. But I haven't yet seen any mechanisms suggested for promoting pervasive congruence that don't amount to significant constraints on character or situation. Even constraints in other elements of exploration might not suffice. A prison setting, for example, is highly constrained but could easily support incongruent G, S, or N decision-making in the absence of further constraints on character and/or situation, such as it being a WWII prison camp where everyone's cooperating to dig an escape tunnel.

- Walt

[Wormwood]

Jason and Walt,

Thanks for clarifying my misuse of congruence. (Yet another reason to suggest a lexicon is needed.)

I'm quite wary that the discussion has shifted to several topics which did not lie within the intent of the original author. As such perhaps a new thread or two would be preferrable. In expectation of such a thread, I'll keep my response short. (I'd post my own split, but I've already started a thread on this matter)

The two issues as I see it:

Jason, the idea of classifying conflict seems interesting, but I'm not sure if those catagories are really descriptive or complete. Certainly both stance and exploration provide locations for player discord (a term I prefer since it is less ambiguous in terms of in game conflict). But there also seems to be a variety of other areas of conflict, such as system disagreements and agenda interpretation. Obviously the most any design can enact is a reduction in discord, not it's removal. Second, in what way is a stance conflict and an exploration conflict distinct, other than where they occur? Admittedly I'm playing devil's advocate here, but it seems that conflict analysis is sufficiently important that it deserves careful inspection.

Walt, the key reason why we disagree about the ease of producing congruence is that most game design exploits congruence, rather than designing it. Most games are designed quite accidently with some region of congruence. Taking that game and constraining the play to just the arena of congruence will suffice to make what is likely an unplayable mish-mash into a functional hybrid, albeit quite limited. Actually designing the support for different decision classes to be indistinguishable seems to have no such limitation. Very little work has been done in intentionally designing such games, however.

I hope that is food for thought,

  -Mendel S.

[Stuart DJ Purdie]

Hybrid play is therefore a subset of Coherent play, just as single-mode play is.

Just so that I'm clear on this, Coherent is taken to mean that all elements of the play agree on how things should be - i.e. it's best defined by the lack of anything that feels out of place, or contradictory?  That's more or less the obvious meaning, but I want to be explicitly clear, as that's (regretably) not always the case (c.f. Simulationism)

Going from that then, Hybrid seems to be a catch-all term for functional play that does not correspond to a single one of the GNS metagame proiorities.

I suggest that when zones or spheres of GNS-incompatibility arise during play, then one has to give way, in order to preserve the social fun.

I'd agree, and actually go further.  If mode X and Mode Y give way at random, then that will not help - it needs to be (at least for the most part) predictable by all the players, else confusion will arise.  By example, I could play so that during a combat, Sim will give way to Gamism, but other than that, the Gamism will give way to the Sim.  That's something that operates at the social contract level.

Here's my claim, then: that over the course of that instance of play, and remember that it probably includes dozens of decisions and interactions among everyone at the table, play is functional insofar as everyone knows what the hell everyone else is up to, in GNS terms.

Ok, here's one of the confusion points.  You have excluded Congruence there, (possible intentionally, possibly not - that's where I'm confused).  With congruent play, it's not possible to understand precisly why a player makes certain decisions (as two (or more!) possibilities map to the same outcome), making it not possible to determine what the other players are up to.

I fully agree with the principle you gave, by the way.

Let's take a look at The Riddle of Steel. We have some hefty Simulationist elements of play, both in terms of "realism" and in terms of the player getting into the character's head (specifically offense and defense during combat). Yet numerically, the Narrativist elements of play so greatly override nearly any possible decision made in line with the Sim mode, that there's no way, without Drifting, for play actually to climax based on Sim priorities. Those Narr elements, the Spiritual Attributes, can only serve the player priority of whatever it is play is supposed to be about, thematically. They don't even operate as a constraint, as the player may change them almost at will.

I read what you've written, and I can only see that as a pure Nar game.  When the premise is not being explored, the Exploration common to all RPG's is clearly seen.  Realism is implemented through system, but Narrativisit priorities overide that.

Prehaps I can illustrate this by pointing out that I can easily accept a Nar/Gam hybrid, with either one being subordinate to the other.  I can also accept a simulationist hybrid, with either narativist or gamist suboridate.  All four of those have clear features that would not be seen in single mode play.

The problem is how can one tell the difference between straightforward Narrativism, (which 'floats' on a sea of Exploration, as do all modes), and Nar - subordiante Sim hybrid (which is Nar, with some elements of Exploration proioritised, except where it conflicts with the Narrativisim).  The both seem to be exactly the same thing to me - meaning either virtually every game is hybrid, with subordinate Sim, or none are.

Stuart

[Ron Edwards]

Hi Stuart,

Excellent - looks like we're making sense to one another.

You wrote,

Hybrid seems to be a catch-all term for functional play that does not correspond to a single one of the GNS metagame proiorities.

I'd amend this to exclude the "none" category, such that instead of "not corresponds to a single one," I'd say, "corresponds to two or more."

Issue #1 would seem to be the congruence thing. As far as I can tell, and I urge Walt to correct me if necessary, congruence operates at a finer scale than my "instance." It seems to me that if we have to invoke congruence in order to explain a set of behaviors, that's basically a tip-off that we haven't gained enough data to go to GNS-talk yet.

Is that a punt on my part? Maybe - unless you agree with me, as many people apparently do not, that GNS does not operate on a "particulate" model, with a G "instance" being composed of G particles that are simply too fine to see. I really think that GNS operates and exists at a level of analysis that is coarser than the individual decision, and below that level, it's not really very interesting per unit time.

A good analogy is found in biology. For instance, the interior components of a cell are not alive - the membranes, the plasma, the DNA, the proteins, etc, none of it (well, I'll spot you the chloroplasts and mitochondria, but nothing else). The cell is alive - which is to say, specific sets of interactions are occurring among these non-living parts. "Living" can only be referenced to those interactions (= metabolism), not to a substance or a subset-component among them.

Now watch Walt come along and kick all of this into a bunch of puff-dust ...

Issue #2 is that business about Simulationism in playing The Riddle of Steel. This game, as written, includes phenomenal detail regarding (e.g.) variables in the damage delivered by different weapons. I consider the attention necessary to bring these rules into play to be significant, in terms of game time and attention - you simply have to get into the mode of thinking that enjoys the consideration that a poniard and a dirk are two extremely different objects relative to human tissue. You could Drift the rules, I suppose, to ignore this consideration, but my take on the game is that a large amount of attention to such things, to a degree that I think is fairly called Simulationist, is necessary - such that their quantitative violation by the Spiritual Attribute system takes on all the more emotional impact when it does occur. This is a tricky, tricky design feat, and that's why TROS gets the accolades it does, and deserves them.

Recently, I summarized my thinking on hybrid game design, which of course should be stated in terms of play first. Briefly to recap ...

I think Simulationist play may provide excellent subordinate hybrid features of both Gamist and Narrativist play. (respective examples: Shadowrun for S/G and The Riddle of Steel for S/N)

I think Gamist and Narrativist play, absent Simulationism (very important), have an exclusive but very easily tranposed relationship. (example: Tunnels & Trolls)

I think both Narrativist and Gamist play make for terrible subordinate hybrids to Simulationist play; the chances for dysfunction of many kinds to get rolling are very high. (examples await discusion) This issue can grade over into discussions of Abashedness and Incoherence.

All subject to much debate, I'm sure.

Best,
Ron

[Valamir]

Can you define what subordinate means in this context?

[Ron Edwards]

Hi Ralph,

By "subordinate," I mean that the mode of play in question can be identified through observation, but we can also identify that another mode is in action as well (could be sequential or almost simultaneous, I think) - and that the features of that second mode are amplified or focused by the presence of the first. The second mode is dominant; the first is subordinate.

I'm kind of stuck for a better example than The Riddle of Steel ... but maybe the G/S examples might help. Think of a role-playing situation in which the primary goals at the table are Gamist, and in which the "stakes" involve character survival, character profit, and completing a given mission, more or less in that order. I'm thinking of Shadowrun, and an interesting detail is that inter-player competition is probably relatively low. In this case, the Gamist "weight" or "confirmation" comes from the players' ability to employ the in-game details like equipment, mobility, firepower, and deployment, both in terms of prep for the mission and of the inevitable unknown factors that will crop up during the mission.

So in this case, up to a certain extent, the more in-game Sim is involved in contributing to those details (the weight of the SMG, the effects of a bullet wound to the thigh, the movement rate of the corporate assassin), the more intensity is imparted to the Step On Up. What keeps it from becoming wholly Sim is the Step On Up itself, as communicated across the people in play almost all the time (and which would be absent in Sim play). I'd also expect a certain amount of metagame mechanics in there in order to allow for metagame agenda to override the Sim-looking mechanics on occasion, especially in terms of burning resources for immediate bonuses, and similar.

All this is distinct not only from Sim by itself, as I mention above, but also from Gamist play by itself, in which the Explorative elements aren't expected to accord with anything except the ability of the in-game Situation (Challenge, in my new essay) to promote Step On Up, at all times. Tunnels & Trolls play remains the best example - the weapon weights and damage ratings, for instance, have literally nothing to do with any conceivable in-game consistency or cause; a weapon weighs more because it does more damage, as a direct means of effectiveness trade-offs for the player to strategize about, and that's that.

Best,
Ron