Tuesday, March 20, 2018

The Plottist In Westworld: What Simulation Isn't

This is, oh yes, another in a series.
Try to ride a horse? Ponder the nature of reality? Just hang out?
We're taking a detailed look at the original texts of GNS Theory, the Indie theory of RPGs. What, How, Who, When and (big one) Why? are covered at the top of the first one from a few days ago.

That one covers a lot of stuff about what Edwards called simulationism--including its history and relationship to wargames and complex challenge. This one hits on some new stuff.

Contemporary GNS Usage

Two points about the continuing influence of GNS theory today both randomly noticed or re-noticed while doing other things this week (specifically: chasing links people gave me about Frostbitten & Mutilated).

Pop Usage

Some anon dropping some hexcrawl advice on a forum last week:

Q: When should the players find those features that aren't in plain view?

A: Whenever you
a) roll it on a random chart (if simulationist)
b) decide things are boring and need excitement (if narrativist)
Either one works.
The usage of "simulationist" is one GNS would recognize (and bad and confusing), the second is a usage of "narrativist" that is at odds with the original GNS conception by Ron Edwards but which maybe some later narrativist might agree with mayyyybe (I am thinking of a guy who made one popular indie game who said his game provided "a more narrative experience" than D&D but has nothing about the morality/plot relationship that Ron Edwards and first-generation spawn like Vincent Baker and Luke Crane were all about).

Expert Usage 

GNS can only be blamed so much for pop usage. But what about people who could be said to be using it as intended? Here's the first sentence of Ron's essay on simulationism, which we're about to get into: Many thanks are due to Clinton R. Nixon, Paul Czege, Jared A. Sorensen, Ralph Mazza, Christopher Kubasik, and Mike Holmes for comments on the manuscript.  Here's one of those game designers answering a thread recently:

 So, as far as I can tell, the Old School Revolution is about demanding bad game design. Can anyone give me a counter-example?
100% correct. Certain gamers have a strange aversion to admitting they'd rather be playing freeform, so they go out of their way to loudly proclaim their allegiance to a broken rules set so they don't have to. But since the rules set is broken, they have to rely on rulings instead of rules...
D&D is a bad design. And OSR games that try to get the desired play experience by emulating a known bad design are thus themselves bad designs. And therefore OSR gamers who demand D&D-esque clones, are essentially demanding bad design....I could go on for pages re bad design. You have a linear probability curve where only a narrow range of possible results are in play. 
You have a play culture that encourages simulationist thinking and a rules structure that is so abstract there is no stimulative [sic?] value to it whatsoever. You have movement rates traditionally given in inches because the game was meant to played with miniatures...but it didn't bother to give any rules for them... 
[it goes on like this for a long time, not answering any questions or challenges, then ends with the following Get Off My Lawn] 
...no one under the age of 40 gets to lecture someone whose been gaming since the mid 70s on not understanding Old School. Us old timers get lots of laughs listening to 30 somethings try to tell us what gaming was like back in the day.

Let's emphasize: this is usage now --state-of-the-art GNS by someone who was there from day one, thanked in the intro, still designs games, whom nobody associated with the Forge has disavowed in any way, who read and enthusiastically recommended Vornheim and still thinks OSR products are emulators and that rulings not rules is about wanting bad rules instead of simply shorter rules they cn customize.

If a GNS expert would like to come and claim this is fringe usage of GNS theory I'm fine with that. But this is a completely everyday example of where GNS talk tends to go when wielded by someone who actually has read and believed it in 2018--and, again, if you'd like to dispute that: let's talk.

Ron Edwards In Westworld 

We're gonna get into the essay in a minute but first I want to do a little orientation:

Like any kind of jargony speech, it can be hard to keep your bearings in these essays while seeing how individual sentences relate to big conceptual stuff. Here's a paragraph from the middle of the Simulationism essay:

Metagame mechanics  
The term "metagame" is problematic throughout this essay for Simulationist play and rules design. Metagame mechanics, by definition, entail the interjection of real-people priorities into the system-operation. Now, it is foolish to speak of Simulationist play as lacking metagame; that would only apply if the people at the table were themselves rules-constructs as well as the rules, and that's silly. But compared to Gamist and Narrativist play, Simulationist play may be spoken of as lacking metagame [i]interpersonal agenda[/i], like "winning" or "doing well" in Gamism, or addressing a Premise in Narrativism. Its metagame, although fully social, is self-referential, to stay in-game. I recognize that it's a problematic issue and I look forward to some discussion about it. 
I've got a model that, for me, clarifies a lot of what Edwards is talking about throughout these essays. I should emphasize this model isn't something I claim necessarily represents what GNS is or would like to be--it just makes individual sentences Ron is writing make more sense to me. It answers some simple questions (Why three categories? What's 'metagame'? What's a 'story' to Ron?) and, more importantly provides an "anchor" image that I can refer sentences to so they don't creep so far out on the jargony edge that I'm lost:

You're going to Westworld--the one in the show. You are a modern person putting on your vest and cowboy boots and six-shooters.

You arrive, look around at the colorful and dusty streets, what will you do?

You could decide your primary goal was to fight the Westworld robots--and test your sharpshooting against theirs (assuming this is a version where they can hit you)--this is, roughly, what GNS would call a gamist decision. This requires Westworld provide you with opponents--or at least some test (riding a horse, etc) that you might fail at and overcome with skill.

You could decide you were going to organize things so you (pretend-) die tragically in the arms of your robot lover after a series of interactions with the Westworlders, all of which were thematically related to each other and produced some definition of a "meaningful story". You don't want to know what the story is first. This is GNS' narrativist decision.

You could invent a history for yourself (a dentist from Back East looking for new scenery) and hang out, enjoying the variety of both the things sitting in Westworld and what happens when you interact with them, and the interactions you can rig up. "Simulationist" by Edwards' thinking.

(You could do a lot more kinds of things, but that's not important now.)

What Edwards means by "metagame" is this:

All three of these yous would be embedded in a simulated Wild West, but, unlike the simulationist, the gamist and the narrativist are making demands on the plot.

Everyone is in a simulation that does x, y, and z, the gamist and the narrativist want to be in a simulation that does x, y, and z and also provides a or b kind of things in the plot specifically.

This is key.

The Gamist (what they call challenge-based out in the real world) needs the game to provide them with challenging targets (some jewel thieves? or a bank) and the Narrativist needs the game to provide them the raw materials to get an intrigue going on of their choosing. The Simulationist needs nothing. Or, more precisely, they need only the same things everyone in Westworld gets: the hat, the horse that gallops, the chapparal, the talking sex robots and desperadoes.

"Metagame" means, when Edwards uses it, "Specific kinds of things you need the plot to give you beyond just functioning plausibly" (real life can function plausibly for days on end without giving you a target or a plot). This also explains how, yeah, he is trying to say there is some metagame for the Sim player--they need to not be bored punching the piano player or talking to the bartender, they just don't expect the fight to automatically test their skill or for the conversation to automatically lead somewhere. These are demands that the fiction be a certain way (strictly: "metafictional"--above or outside or about the fiction--demands, not really "metagame" but whatever).

This may be a simple or obvious way of thinking (and probably is to people who like GNS), but clarifies a lot of the later weirdness:

This clarifies how Edwards uses the word "exploration" and conflates nearly any detail (no matter how tactically or narratively useful) with simulationist-pleasing--essentially assuming the sim player, having nothing else to do, is going to do a lot more asking the robots random personal questions and poking the cactuses to see if they prick than the other players. That's why he says "Simulationism is expressed by enhancing one or more of the listed elements in Set 1 above (that is Character, System, Setting, Situation, Color, -Z);" basically the cactus-poker is more likely to need the needles to be sharp than the other two, though all of them need cacti as a backdrop to be in the West.

This also clarifies the relation to "premise"--premise is: ideas that are in the game. He talks a lot about "premise" and how narrativism needs it. The whole of Westworld has certain ideas undeniably in it: expansionism is brutal, the world is dangerous and sometimes you gotta resort to violence, women have different roles than men in the old west, etc--these are premises but the narrativist needs them to not just be there but to have to make some big decisions, realizations, choices about how a character interacts with them (ie "address a premise"), like: be like fuck this town I'mma burn it down because imperialism.

This also, I now realize, is why GNS feels comfortable lumping together The Guy Who Wants To Act with The Girl Who Wants To Be Sure The 6 Shooter Has 6 Bullets under "simulationism"--not because of the linguistic convenience that both can be described as "simulating"--but because neither is, in the moment, asking any more of Westworld than to be Westworld. Their story doesn't need to have a plot, even the minimal plot (there was a baddie, I chased him, maybe I caught him) that a gamist requires. That writer character that Anthony Hopkins employs doesn't have to write anyone in for the simulationist this season.

And this is the hidden-in-plain-sight thing: plot is the thing Edwards is so obsessed with that he forgets to say. He is obsessed with plot (all the old narrativists are) and when GNS says "story now" they mean "plot now" and GNS is actually about evaluating games and gamers concerning their relation to plot and so this is why it says all gamers are:

"Need to help make an interesting plot" (Narrativist)
"Need a plot that provides hard stuff to overcome" (Gamism)
"Don't need the plot to be anything in particular" (Simulationism)

...and is also at immense pains to separate that first thing from "Need the plot to be the Dragonlance story" (Edwards would call this simulationist because sim goals are not about needing to help shape the plot, just about wanting to be part of ("Exploration") the story.)

Also his ideas about plot tend to focus or center on plot in the classical (often 3-act) drama sense not the Chandlerian sense or Ian Flemingian sense--basically the kind of stories where every scene is identifiable as contributing thematically (not necessarily tactically) to the final or climax scene. Burning Wheel needs you to confront beliefs, Dogs in the Vineyard needs you to make moral judgments and maybe end them in climactic gunfights.

Nearly every scene in a Sherlock Holmes (challenge-oriented-gamer-extraordinaire) contributes to the climax, but doesn't necessarily address any specific themes. The end doesn't change what you think Holmes is or does--he was smart at the beginning and the end. Hamlet as a character is, on the other hand, revealed and interrogated by his plot. This doesn't have to be fancy: Edwards is narratively interested in FASERIP and its Karma system where being good and being in character and watching out for civilians can get you points that help you win fights. These mechanics beg: who are these heroes? What will they sacrifice to win? The game can be used to frame Marvel drama as drama in the classical sense, though absolutely not in the sense of all good stories (even at Marvel comics).

So, remember: Edwards is very interested in plot, especially plot as revealer of character and especially especially theme. He cares very much who controls the plot, he doesn't give a rat's ass about all the other things in a narrative that also can reveal character or theme. He seems almost blind to them. He forgets to clarify because he assumes his audience also assigns this primacy to plot.

He is a Plottist and we'd all be less confused if, back at the turn of the century when he wrote this, he called himself that.

The anon above in "Pop Usage" who thinks you should reveal detail when the story's getting dull? He's more right than I gave him credit for. That person is worried about the quality of the narrative (that is: the things being narrated, including lots of non-plot story detail). Edwards' narrativist only reveals details because the Plot needs them, not to just spice things up.

Anyway: when you're lost in Edwards, think of Westworld and what these three Westworld players need. More later.

Historical Context

A reaction from someone on G+ to the last essay that might give you a helpful frame of reference:

Two observations I want to share based on where I was when these essays were being published. (In college playing a variety of D&D and WoD and later CoC) 
One is that when I, accepting Ron’s theory at face value, tried to explain the GNS thing to other gamers (that was a long time ago), they came up with the assumption that all three should be balanced in a good game. 
The other is that there was in the online forums that Ron posted in before he went to the Forge and that I was part of back then, “story” had a ton of cachet as did “good role playing” and these were things we were trying to achieve in our games. That’s not immediately obvious because d20 hit immediately at that time and the whole scene changed dramatically.
I think that you nailed the first point (though indirectly) and almost got to the second point when talking about WoD [World of Darkness] and the way that a lot of pro GNS players seem to have been hurt by it.


 Alright, here we go:

Simulationism: The Right to Dream

by Ron Edwards  2003-01-29

Many thanks are due to Clinton R. Nixon, Paul Czege, Jared A. Sorensen, Ralph Mazza, Christopher Kubasik, and Mike Holmes for comments on the manuscript. Several points, key text quotes, and nuances of argument wouldn't be in the article without their input. All inconsistencies or argumentative flaws, on the other hand, may be laid at my door.  
This is the first of three essays about the three GNS modes of role-playing. Each one is about both play and game design, with the former as the basic issue, and each one is intended to develop the points made in my "GNS and related matters of role-playing design" essay. I'm also drawing upon ideas I didn't express in that essay and many, many points of debate at the Forge over the last year. The original essay cleared up a lot of acrimony and misunderstanding...

(They hadn't seen anything yet.)

that had arisen in the previous years and I'm hoping that the current series plays an even more positive role in the current context - not only to remove negative connotations and interpretations (which are now much fewer anyway), but to encourage mutual understanding and appreciation among all role-players about all the available modes of play. 

(And we all smile wryly and feel the bottom edges of our canines with our tongues.)

Each essay isn't a segregated unit only about that one mode. Each will include more general issues, especially if they pertain especially if not uniquely to the mode under discussion, and each one is intended to clarify and develop "GNS and related matters" as a whole. Also, each one concludes with a Hard Question for those who prefer that mode of play. Each Hard Question is supposed to be interesting on its own, but I hope that the three taken together will be much more than merely "interesting."  
Simulationist role-playing has a great deal of power and potential. In the previous essay, I wrote that it "... is expressed by enhancing one or more of the listed elements [Character, Setting, Situation, System, Color]; in other words, Simulationism heightens and focuses Exploration as the priority of play. The players may be greatly concerned with the internal logic and experiential consistency of that Exploration." 

That is: Simulationism in Edwards' estimate is about making some part of the real-seeming parts of Westworld seem extra real -- or at least extra deep. The gambler's text-option tree is 90 levels deep as opposed to the mere 20 that it needs to give that surface patina of plausibility.

Exploration reviewed  
Obviously the thing to do is to get as clear an understanding of "Exploration" as possible. It's our jargon term for imagining, "dreaming" if you will, about made-up characters in made-up situations. It's central to all role-playing, but in Simulationist play, it's the top priority.  
I need to stop the flow for a moment to explain some background, though. My original notions were mainly laid out in "System Does Matter," my first essay about all this stuff, based on my readings about the Threefold Model proposed in the r.g.f.a. discussion group. At the Gaming Outpost, lots of debate ensued about my essay, and eventually a poster called the Scarlet Jester objected to the term Simulationism in terms of its connotations, offering "Exploration" as the replacement - defined as the enjoyment of the "dream" or the imagination as an act in itself. He called his model "GENder" as an alternative to the then-existing GNS.  
GENder made a lot of sense to me, with one exception: Exploration, to me, seemed to be involved in all of role-playing. I decided to modify GNS severely and "float" the three modes on a "sea" of Exploration. In that context, Simulationist play priorities suddenly made more sense - as I saw it and still do, unlike Narrativist and Gamist priorities which are defined by an interpersonal out-of-game agenda, Simulationist play prioritizes the in-game functions and imagined events. 

This is the whole "Narrativism and Gamism demand things from the plot, whereas Sim just needs to be there in Westworld" thing.

Ron then gives some quotes from intro game texts. The problem with all these quotes is that, remember, in-game narrativist and gamist decisions are things you make after you are already simulating (I'm a fighter, I'm a thief, I'm a talking shoe--by playing an RPG I have decided to pretend whether or not I have a plot goal). And so: these introductory texts which are just trying to explain to neophytes and wargamers what an RPG even is are, of course, going to point out that we are going to Westworld and it's like the old West and not real life. So their status as "simulationist rhetoric" is as dubious as the claim that a recipe whose first ingredient is a tomato is "vegetarian":

From the introduction to RuneQuest, second edition (The Chaosium, 1978, 1979, 1980; specific author for this text unknown; game authors are Steve Perrin, Ray Turney, Steve Henderson, and Warren James):

What is a fantasy role-playing game?
A role-playing game is a game of character development, simulating the process of personal development commonly called "life."

[In fairness, later text in the introduction brings in some adversarial GM/player context that sounds more Gamist, but the above quote is reinforced more often throughout the book's rules and text.] 
From the introduction of Skyrealms of Jorune, 3rd edition (Chessex Publications, 1992, author is Andrew Leker):  
Is it possible to win at role-playing? The whole idea of role-playing is to have a good time. Players work toward a common goal, often survival, but sometimes helping a friend in need, or accomplishing a task of unquestioned importance. Although there will be no winner or losers in an absolute sense, you will have the satisfaction of watching your character think through challenges, survive confrontations with other races, grow, and develop new skills.
If you keep in mind all this writing happened pre-internet--before the precision that people like Ron forced onto the discussion (for example, in the AD&D DMG Gygax once wrote "Now and then a player will die" when meaning character) there isn't really anything particularly not-gamist or even not-narrativist here "watch your character think through challenges" could easily be gamey (challenge-oriented) rhetoric, and "grow" could easily be a normal narrativist thing.

Even on his own terms, this chosen-as-illustrative text isn't as strenuously one-way as Ron wants it to be.

From the introduction to Marc Miller's Traveller (1996, author is Marc Miller): 
... the players' enjoyment comes from identifying with the character and vicariously experiencing the situation with that character, just as the reader of a novel and the viewer of a movie identify with the character ...

[The above text is followed by some Impossible Thing Before Breakfast text which will be discussed in the Narrativism essay.] 
(Long story--we'll get to that.)

What's fun or good about that? Simulationist play looks awfully strange to those who enjoy lots of metagame and overt social context during play. "You do it just to do it? What the hell is that?" 

Honestly, this statement seems strange to me unless it's done as a rhetorical exaggeration. I mean: surely even the least imaginative person on earth can see someone pretending to be a duck as, like,  fun for a second. Just being in Westworld would be neat for at least a while--no matter who you are.

However, contrary to some accusations, it's not autistic or schizophrenic, 

I think there's considerable evidence that whoever said that was though.

being just as social and group-Premise as any other role-playing. The key issues are shared love of the source material and sincerity. Simulationism is sort of like Virtual Reality, but with the emphasis on the "V," because it clearly covers so many subjects. Perhaps it could be called V-Whatever rather than V-Reality. If the Whatever is a fine, cool thing, then it's fun to see fellow players imagine what you are imagining, and vice versa. (By "you" in that sentence, I am referring to anyone at the table, GM or player.) To the dedicated practitioner, such play is sincere to a degree that's lacking in heavy-metagame play, and that sincerity is the quality that I'm focusing on throughout this essay.  
Sincere shared creativity: all role-playing has to have it. For some, it's the whole point. 
Is the term fatally flawed? 
More than once, people have called for abandoning the term "simulation" in its entirety. Most of the objections arise from connotations of one sort or another, since it gets used for all sorts of recreational or applied things. If it's Simulationism, then what's it Simulating, and what form does the resulting Simulation take? 
For better or for worse, this issue has never really struck home for me.
My call is that the term is is defined locally and historically, and not really descriptive as such ("simulating") in nearly any application. Here's the variety that I see:  
Simulation in wargaming = historical plausibility ("realism").
Simulation in computer games = rendering, reaction time.
Simulation in behavioral terms = "let's pretend" in terms of our expressions, gestures, and voices.
Simulate in emotional terms = related to lying, as in dissimulate or simulated passion.
Since the term does not carry a single meaning among all the other contexts...
Uh no dude, that's not a variety of meanings at all. In every case those mean "an attempt to imitate" (respectively in those examples: history, continuity of lived experience, other people, people who feel differently than you do, etc).

This is a very old concept in art (the theme of Ron taking the long way around to get to a standard art vocabulary will come up a lot). Mimesis: some art is non-representational (a dot and box pattern for instance) and some is (Pac Man is a symbol representing an imaginary hungry creature). RPGs are representational games (not Uno)--you pretend to be other people (or even things).

Be that as it may, if you have his point of view of plot-obsession, Edwards case for lumping all non-challenge, non-plottist games into one third category makes more sense than he can articulate: all RPGs simulate, and his "simulationist" decisions are the ones where the player wants not much from the plot beyond that. 

That's a good argument for "simulationist" decisions being a decent name (though not for claiming any specific games primarily enable simulationism or that its a major engine for players or designers--that is, not for its status as "a major goal" and the focus of most mainstream games), but anyway let's let Ron finish his bad argument for the name...

Since the term does not carry a single meaning among all the other contexts, assigning a specific meaning for role-playing just seems to be par for the course and not especially or intrinsically confusing. Hastily added: "to me." Maybe I'm just obdurate.  
Taking it role-playing specifically, a new issue arises: it's awfully hard to get at goals of any kind right out of the texts. A good place to start is Advanced Dungeons & Dragons, in just about the first text ever that tried to explain what was going on (Dungeon Master's Guide, first edition, 1979, TSR; the author is Gary Gygax): 

Of the two approaches to hobby games today, one is best described as the realism-simulation school and the other as the game school. AD&D is assuredly an adherent of the latter school. It does not stress any realism ... It does little to attempt to simulate anything either. ADVANCED DUNGEONS & DRAGONS is first and foremost a game for the fun and enjoyment of those who seek to use imagination and creativity.
How to parse this?
Ooh let me. (Take note: you've found Zak Sabbath commenting on a quote by Ron Edwards commenting on a quote by Gary Gygax. Somewhere in the future, someone very boring is masturbating.) Gary is trying to stave off criticism (probably from people like Runequest fans) that the D&D rules don't reflect (most likely) medieval historical warfare realities close enough and/or source material like Tolkien (slightly less likely--though a complaint a lot of GNS people still make, weirdly) and so he's stressing D&D doesn't try to be some other thing. However, being a little less than completely rigorous about language he says "It does little to attempt to simulate anything" which is not exactly accurate but it's Gary who rarely is.

Ron's take is interesting:
It seems unequivocal. However, first, this text is palpably disingenuous regarding "simulates nothing" - the immense efforts devoted in this book to the importance of in-game time and in-game justifications of hit-points, retainer/hireling opinions, costs for castle parts, and much more, do not support his claim.
This is huge but huge regarding a point I went over a lot in the last essay. Ron--and, to this day, GNS in both its pop and expert forms--does not get the gamist/challenge-oriented application of details like this.

Like: you go into a dungeon, you defeat challenges to score gold (like points) you then can trade this gold for castle features (catapults, towers) to fight off sieges and stuff at high levels--another challenge. More gold: more points to fight with. This is pretty close to the actual gameplay of many challenge-style videogames like the original Warcraft (before WoW, just straight Warcraft).

Random details often become important in challenge-oriented play when that play is not solely based on system-specific mechanics.

Lets say you fight a huge giant in a video game. You have a sword that cuts off anybody's head. Here comes the giant--it has and needs a head so...you killed them, unless the programmer writes a special rule about the giant.

Same thing in a card game or an RPG whose rules are sufficiently system-mastery-based: I have the "decapitate" button, you have the creature card. Like scissors vs paper. It's over.

In a traditional RPG, there are some issues though: if the diameter of the giant's neck is way bigger than the length of the sword (a 4' sword decapitating a giant with a neck like the caldera of a volcano?) then it's possible that "slash-->decapitate" doesn't make sense. The rules are forced to either be vague and rely on a ruling (thus: rulings not rules) or go into detail about how long the sword is (which AD&D does) and how wide around the giant's neck is (which AD&D doesn't, but starts going in that direction as time goes on).

So far, same critique: Edwards has never tried to kill a Tarrasque, but...There's a second issue: "Make sense" int he above paragraph is an important phrase here and points to a massive misconception Edwards has about simulation: suspension of disbelief.

Let's say you're back in Westworld--you ask the robot bartender if you might trouble him for some whiskey. He says "No sir". You raise an eyebrow. What? Again he says "No sir". It then becomes obvious this is just a glitch, his little cpu is just saying "No sir" to every question. Then he reboots and gets you your drink.

Now if you're in Westworld to play Boot Hill and shoot some dudes ("gamist"), at the first "No sir"--you're gonna look around, does the bartender have beef? Is there trouble? Is it already time for the shootout? Does this signal have tactical significance?

If you're in Westworld to play Dogs in the Vineyard and makes some moral judgments and experience their consequences ("narrativist"), you'll wonder whether that "No" is a plot twist--the GM introducing some complication that will test your assumptions in some way.

And if you're just playing Deadlands and here to enjoy and poke this bizarre setting  ("simulationist") you're gonna wonder what that tells you about the bartender's relationship to you or someone else.

After a second, you are all going to just realize it's a glitch and go back to what you were doing. You all have--regardless of your GNS goal here experienced a failure of simulation as a distraction.

Again we have a familiar art concept hidden beneath maligned online-RPG arglebargle words ("realism", "verisimilitude", "immersion", etc) that is: willing suspension of disbelief and things that distract from it.

Let's say you're watching a regular romantic comedy. Lisa Kudrow (who will fall in love in implausible circumstances, "unrealistically") 's tissue box suddenly moves from one side of the coffee table to the other. Wait? Did it just become a sci fi? No, the continuity guy on the movie just fucked up.

Even though the tissue box isn't something associated with the goal of watching a Lisa Kudrow romantic comedy, it's distracting when whatever level--or perhaps the word is style--of surface believability that a work of art establishes suddenly isn't there for a second. That's why even comedies and action movies have continuity people and why even cartoons (where everyone has a black line around them) can look wrong even if the point is they're just funny dialogue or morally preachy or whatever.

What's more, everyone has a different threshold at which implausibilities distract them an different thresholds for a bewildering range of different things they see. This isn't something they can do anything about any more about than their height or eye color--I want very much to enjoy the Avengers movie and...Loki on the back of that jeep immediately kicks me out of it. Gamma radiation turning a person into a big green superhero I can believe, but my backbrain rebels at Thor's plasticized costume. This isn't an ideological position--it's just what my particular nature/nurture combo accepts as "plausible for the purposes of temporary suspension of disbelief".

When rules exist simply to make Westworld tick, this isn't necessarily to cater to the participants distinctive goals (and GNS, at least in this iteration, is about goals, even temporary ones) it's simply to avoid distracting them or introducing an element that trips the player's disbelief sensor.

This "Maintenance simulation" is like breathing--rules that exist just so you don't get constantly reminded of the unreality of the situation instead of being able to focus on your own real goals--is another thing Edwards just doesn't get. He's forgetting all the work it took just to make the robot horse he's riding (on his--yeah yeah--desperate quest to prove to his sister that he can protect her despite the fact that...etc) look like a horse and gallop like a horse.

All through this essay we'll see rules that exist for maintenance simulation mistaken for the sole purpose of being there just in order to facilitate a supposed player's desire to just be on a horse. Some people just want to be on a horse--yes--but the rules handle that and more.
Second, and more importantly, Gygax is speaking from a 1970s perspective of role-playing existing as a subset of wargaming. What he calls simulation or realism, I call historical accuracy; what he calls "game" (imaginative, creative), I call Exploration. As an "umbrella point," although D&D and AD&D of this era were procedurally mainly Gamist, all accompanying text by Gygax in any publication represents, I think, very hard-line post-wargame Simulationism as conceived by GNS theory.  
A somewhat lesser issue concerns whether I'm doing great violence to the term Simulationism as proposed in the original Threefold Model. 
In case you forgot, in the original pre-Ron model (the "threefold model") "simulationism" was conceived of as "The thing the guy who says hit points are unrealistic is worried about" (ie the person who had a low threshold for failure of suspension of disbelief) and instead of narrativism there was dramatism.

My answer to this has two parts. (1) The Threefold definitions, for all three modes, tend to benefit in this debate from being moving targets over the years. (2) My set of theorizing, usually called "GNS" although I'm starting to wish for a better umbrella term, explicitly disavows any need for consistency with the Threefold.  
However, although I'm not convinced it's necessary, one possible solution has arisen. Jack Spencer proposed "Emulation" for the goals of play that I currently call Simulationism. If I felt any need for a wholly new term, this would probably be it. 
Baseline Simulationist practice  
The five elements of role-playing as laid out in my GNS essay are obviously where we start. Modelling them is the ideal. My first point about that is that the model need not be static; dynamic characters and settings, for instance, are perfectly valid Simulationist elements. My second point is that different types of Simulationist play can address very different things, ranging from a focus on characters' most deep-psychology processes, to a focus on the kinetic impact and physiological effects of weapons, to a focus on economic trends and politics, and more. I'll go into this lots more later. 

Again this lumping only makes sense if you are coming at it from the POV of "what do they ask of Westworld's plot? Nothing", just as categorizing people into "hot" "not hot" and "boys" only makes sense if you're a straight guy or a lesbian looking for a date. That is: it ignores many salient characteristics and is basically categorizing them toward a specific concern.

The second point is that the mechanics-emphasis of the modelling system are also highly variable: it can handled strictly verbally (Drama), through the agency of charts and arrows, or through the agency of dice/Fortune mechanics. Any combination of these or anything like them are fine; what matters is that within the system, causality is clear, handled without metagame intrusion and without confusion on anyone's part. That's why it's often referred to as "the engine," and unlike other modes of play, the engine, upon being activated and further employed by players and GM, is expected to be the authoritative motive force for the game to "go."
The game engine, whatever it might be, is not to be messed with. It is causality among the five elements of play.
Causality is also king in any kind of complex challenge-play (and also any kind of play oriented toward getting interesting emergent stories out of people who are trying to get to challenge, which Edwards doesn't recognize as a separate goal) so not sure why it's left out.

Whether everyone has to get the engine in terms of its functions varies among games and among groups, but recognizing its authority as the causal agent is a big part of play. (To repeat, the engine's extent and detail aren't the point; I could be talking about a notecard of brief "stay in character" requirements or a 300-page set of probability charts.) By the way, moving the GM into a position of authority over the rules/system is a derived state of the rules' authority; I'll discuss that later.  
Many Simulationist systems also emphasize modularity - you've got the baseline engine for what happens, so for specialty phenomena, whatever new rules go on top must not violate or devalue that baseline. When a system is very strong in this regard, it's what most people call "universal" or "generic," by which they mean customizable through addition. 
My final point is that this mode requires clear player-character/real-person boundaries, in terms of in-character knowledge and metagame knowledge. There's no single set of boundaries that applies to all ways to play Simulationist, but whatever they are in a given instance, they must be clear and abided by.  
How-to-play text 
A lot of game texts in this tradition reach for a fascinating ideal: that reading the book is actually the start of play, moving seamlessly into group play via character creation. Features of some texts like the NPC-to-PC explanatory style and GM-only sections are consistent with this ideal, as well as the otherwise-puzzling statement that character generation is a form of Director stance.

Character gen as director stance basically means: while during normal play you mostly only control your PC, during gen you might invent or alter things only a god could irl--like decide you eye color or where you were born, etc. You are setting more of the scene. (Also true during leveling-up sometimes--like you get to pick some new outside-you thing.)

It supports the central point of this essay, that the value of Simulationist play is prioritizing the group imaginative experience, to an extent that expands the very notion of "play" into acts that from Narrativist or Gamist perspectives are not play at all. 
This ideal poses two problems: one for the GM in particular, and one for the group as a whole.  
The GM problem, only partly solved by GM-only sections, is that it makes it very hard to write a coherent how-to explanation for scenario preparation and implementation. Putting this sort of information right out "in front of God and everybody" is counter-intuitive for some Simulationist-design authors, because it's getting behind the curtain at the metagame level. The experience of play, according to the basic goal, is supposed to minimize metagame, but preparation for play, from the GM's perspective, is necessarily metagame-heavy, and if reading the book is assumed to be actually beginning to play ... well, then a certain conflict of interest sets into the process.  
The usual textual solution is to assume that the GM is already on the same page and to address him or her as a co-conspirator. In many games, however, such information is outright punted, such that a GM must bring a particular set of experiences and values to the text in the first place in order to play the game. 
I've never seen it but ok. I think part of what Ron's detecting may be the GM sections' assumption that the GM has bought and read the book--unlike the players--and so can be assumed to be "on the same page" more than the player.

The whole-group problem is that individually-conducted character creation often produces differing conclusions about the point of play from player to player, which is to say, the characters are fully plausible and created by the rules, but are also manifestly incapable of interacting in terms of any one person's desired genre/setting. The classic example in fantasy-adventure play is the party including a paladin and an assassin; the one in superhero play is the super-team that includes both a Spider-Man clone and a Wolverine clone. 

The usual textual solution is to urge that all character creation be subject to the approval of the GM, which in practice poses some problems. For instance, it assumes that the Social Contract of the game group permits such authority and presents no procedure to follow if that happens not to be the case.

I can't imagine adults trying to build "functional play" when the players don't even have a "social contract" where they trust their GM to enforce subgenre rules. It'd be like trying to play without dice or a pencil or any other necessary material. All of GNS makes a lot more 'sense' if you assume a parallel dimension where people willingly play with GMs but somehow don't ever like them.

And it's not like Ron is blind to this--iirc he's the guy who said "don't roll with someone you wouldn't see a movie with"--so I don't get why he's laying this lack of a fundamental essential on the doorstep of the game design not "the social contract".

Also, I have never seen any text explaining what a GM is supposed to do or to say to the player regarding how to re-write the character or to design a new one

I can't imagine an adult thinking they could roll with a GM for whom this was in any way a challenge. If you can't even go "No you can't play a beholder" how do you do any basic GMing task like say "hey you looked for secret doors and rolled well but there aren't any!"?

And as for kids: clear rules never stopped them from arguing about anything--ever play basketball. I suspect the lack of Connect Four bloggers suggests they found a way to fun in RPGs while arguing they way they did with everything else.

 ...every example, and there are many, seems to assume that the GM "just knows" how to communicate the je ne sais qua to the player. 
I suggest that genuinely helpful, teaching-oriented text that does not fall into synecdoche ("real role-players," etc) would be a tremendous benefit to presenting straightforwardly Simulationist games. Such text would include methods for GMs to prepare scenarios from a fully-metagame perspective - which is to say, the ideal of the book "being play" would have to be lost temporarily - as well as methods for the GM's work during character creation itself. Furthermore, this text would have to be practical and compelling to players in a way that "All character creation is subject to the approval of the GM" is not - for instance, it would inspire players to avoid the paladin-assassin problem on their own, during the creation of the first characters rather than the second ones.  
Historically, such text has been rare. Well, actually, it's rare for any mode of play, but I submit that Simulationist-oriented games have tended to have special trouble with it due to the widely-held ideal of treating the text experience as play. 
Internal Cause is King  
Consider Character, Setting, and Situation - and now consider what happens to them, over time. In Simulationist play, cause is the key, the imagined cosmos in action. The way these elements tie together, as well as how they're Colored, are intended to produce "genre" in the general sense of the term, especially since the meaning or point is supposed to emerge without extra attention. It's a tall order: the relationship is supposed to turn out a certain way or set of ways, since what goes on "ought" to go on, based on internal logic instead of intrusive agenda. Since real people decide when to roll, as well as any number of other contextual details, they can take this spec a certain distance. However, the right sort of meaning or point then is expected to emerge from System outcomes, in application. 

Not sure what "meaning" is supposed to mean here. Like in Westworld, I don't know what activities Ron would consider do or do not have "meaning" or how it would be difficult to achieve.

Clearly, System is a major design element here, as the causal anchor among the other elements. As I outlined in the previous essay, System is mainly composed of character creation, resolution, and reward mechanics.  
During character generation, layering and overt currency are frequently employed to engage the player in Simulationist play during the process.  
Layering may be employed to establish and identify the character's plausibility in terms of the game-world itself. For a look at the historical differences among games, compare the methods for establishing player-character skill competence in early RuneQuest (Simulationist) with those of Hero Wars (Narrativist). In Hero Wars, the system limits how many of the thirty or so starting abilities are assigned high values (two really good ones and one great one), but not which ones. Whereas in RuneQuest, every skill has a starting-character value based on its commonality and difficulty to learn, and every skill is rated in money regarding learning higher values of competence, based both on difficulty to learn and who teaches the skill. Hero Wars character creation, which is minimally layered, isn't concerned with the implausibility of having a mastery-level in Greatsword be just as "likely" as having it in Farming; RuneQuest character creation, which is maximally layered, emphatically is. 
Edwards is doing a weird thing here--I know he thinks these games are Sim-friendly and I know he thinks they have certain characteristics in common but I am not sure if these characteristics they have in common are the things that made him declare them Sim-friendly to begin with or whether he is noting them after the categorization as discoveries of correlations. He acts like someone told him these were the Sim games and his job after is to impartially catalogue what was in them.

He does this a lot.

To repeat, the above point is historical. Whether the distinction I've drawn holds for any and all Simulationist play potential, I don't know.  
A related issue is prerequisite attributes and abilities for a given ability, which represent a further step of layering. Prerequisites are common in historical Simulationist and Gamist design, and in the former, I think they are present specifically to reinforce the same plausibility/likelihood issue.  
For currency, consider Champions or many of the games based on its principles. From a Simulationist perspective on play, if a given feature costs more than another, or if it can be traded off with some other feature, or if it plus another feature mathematically yield a third, then it's all built to focus attention and assign cause from "is" to "does" in the imagined game-context. That cause must be (a) engaging (as for any RPG) and (b) causally continuous through the layers, providing for many equally-functional, equally-plausible, and potentially equally-enjoyable options. 
I think this combined approach and perceived purpose of layering and currency is why attribute + skill systems have remained entrenched - a strong sub-set of the Simulationist perspective demands that the in-world ontogeny of a character's ability be integrated into the process of establishing it on the character sheet. 
No, attribute+skill is common because someone who is great at Astronomy (skill) but really stupid (attribute) is implausible and trips any kind of players suspension of disbelief alarm, so linking attributes to skills ("layering") somewhat avoids the issue. Also: sketchily described characters ("Pick 2 things you're good at and a job") can take a while to invite investment and don't force you to reconcile the variety of stats in the interesting way that most character gen does. 7 Wis, 17 Int, hmmmm...

Resolution mechanics, in Simulationist design, boil down to asking about the cause of what, which is to say, what performances are important during play. These vary widely, including internal states, interactions and expressions, physical motions (most games), and even decisions. Two games may be equally Simulationist even if one concerns coping with childhood trauma and the other concerns blasting villains with lightning bolts. What makes them Simulationist is the strict adherence to in-game (i.e. pre-established) cause for the outcomes that occur during play. 
Again: Westworld. He means outcomes aren't shaped by the need for challenges or for interesting plot.

Before talking about dice or other specific resolution mechanics, I'll discuss two elements of Resolution which are rarely recognized: the treatment of in-game time and space. These are a big deal in Simulationist play as universal and consistent constraints, which must apply equally to any part of the imagined universe, at any point during play. 
To talk about this, let's break the issue down a little: 

  • In-game time occurs regarding the actually-played imaginary moments and events. It's best expressed by combat mechanics, which in Simulationist play are often extremely well-defined in terms of seconds and actions, but also by movement rates at various scales, starship travel times, and similar things.
  • Metagame time is rarely discussed openly, but it's the crucial one. It refers to time-lapse among really-played scenes: can someone get to the castle before someone else kills the king; can someone fly across Detroit before someone else detonates the Mind Bomb. Metagame time isn't "played," but its management is a central issue for scene-framing and the outcome of the session as a whole.
  • Real time is, of course, the real time of play as experienced by the people at the table. I think comparing between its flow and that of the in-game time is a crucial issue as well - when is a huge hunk of real time necessary to establish a teeny bit of in-game time, and vice versa?
The following text is also from the first edition of the Dungeon Master's Guide (TSR, 1979); the author is Gary Gygax. 
One of the things stated in the original game of D&D was the importance of recording game time with respect to each and every player-character in a campaign. In AD&D it is emphasized even more: YOU CAN NOT HAVE A MEANINGFUL CAMPAIGN UNLESS EXTENSIVE RECORDS ARE KEPT.[provides an example, then:]

You may ask why time is so important if it causes such difficulties with record-keeping, dictates who can or can not go adventuring during a game session, and disperses player characters to the four winds by its strictures. Well, as initially pointed out, it is a necessary penalty imposed on characters for certain activities [making magic items - RE]. Beyond that, it also gives players yet another interesting set of choices and consequences. The latter tends to bring more true-to-life quality to the game, as some characters will use precious time to the utmost advantage, some will treat it lightly, and some will be constantly wasting it to their complete detriment. Time is yet another facet which helps to separate the superior players from the lesser ones.
That latter point bears close, close examination. Gygax is not talking about winning, I think, but about a quality. This is his value judgment about how to play this game. His "true to life quality," I think, is synonymous with his earlier reference to creativity and imagination, or Simulationism (prioritizing Exploration) as defined by me. 
Game time is of the utmost importance. Failure to keep careful track of time expenditure by player characters will result in many anomalies in the game. ...

Edwards provides no explanation of why he thinks game-time isn't important in challenge-based play. Or why Gygax wouldn't be thinking about winning with all that free time. I've played games without strict game-time around "race the clock" plots and they are open to lots of exploits that both kill challenge ("The bomb won't even threaten to go off until we're in the last scene, so let's just interrogate everyone in the building") and trip the suspension-of-dis alarm.

Gygax's text perfectly states the Simulationist view of in-game time. It is a causal constraint on the other sorts. One can even find, in many early game texts, rules that enforce how in-game time acts on real time, and vice versa. However, most importantly, it constrains metagame time. It works in-to-out. In-game time at the fine-grained level (rounds, seconds, actions, movement rates) sets incontrovertible, foundation material for making judgments about hours, days, cross-town movment, and who gets where in what order. I recommend anyone who's interested to the text of DC Heroes for some of the most explicit text available on this issue throughout the book.  
So much for time; now let's talk space. Rules for characters' movement in the imagined space of the situation go all the way back to wargaming, in the (to us oldies) familiar forms of grids and hex-maps, counters, and even rules or tape-measures. The original context was pretty large-scale: the movement of troops, heavy vehicles, squadrons, and so on. For role-playing in the "new" sense, the scale got bumped down to the individual level, and so came to emphasize facing, movement rate, turn rate, number of personal actions, and similar.  
The interesting thing is that most of these specific details have been lost in most, although not all, Simulationist rules design over the decades, with nary a whimper. Why? Because second-to-second kinetics ceased to be (or rarely were) the issue of Exploration at hand, particularly in genre-heavy play (see later). The Situation of interest typically isn't "facing" when we want Character, Setting, System, Situation, and Color to fire on shared cylinders with full internal-consistency and agreed-upon thematic outcomes.
It's significant, I think, that movement-specific mechanics do remain in many Gamist RPG design as an element of tactical challenge. 
That last statement seems something like the result of a No True Scotsman fallacy--"Sim doesn't emphasize spatial relationships, Oh that game does? Then it's Gamist" (look back at the last essay for the thin thin lines that cause one game to be declared Sim and another Gamist).

Now for the more nitty-gritty resolution mechanics, or DFK (Drama, Karma, Fortune). Historically speaking, the System has been based on task resolution, not conflict resolution, regardless of scale. Don't mistake "conflict" for "large-scale task." This point is independent of the system's complexity; it applies to rock-paper-scissors and GM-fiat as well as to dice and tables.  
The causal sequence of task resolution in Simulationist play must be linear in time. He swings: on target or not? The other guy dodges or parries: well or badly? The weapon contacts the unit of armor + body: how hard? The armor stops some of it: how much? The remaining impact hits tissue: how deeply? With what psychological (stunning, pain) effects? With what continuing effects? All of this is settled in order, on this guy's "go," and the next guy's "go" is simply waiting its turn, in time. 
Again: all this is just avoiding different folks' individual suspension-of-disbelief failure alarms. Not specific to someone whose goal is to enjoy simulation.

The few exceptions have always been accompanied by explanatory text, sometimes apologetic and sometimes blase. A good example is classic hit location, in which the characters first roll to-hit and to-parry, then hit location for anywhere on the body (RuneQuest, GURPS). Cognitively, to the Simulationist player, this requires a replay of the character's intent and action that is nearly intolerable. 

Or, alternately Ron, regardless of creative agenda or goal, it's a little distracting if the robot horse has to be wound up before you can ride it.
It often breaks down in play, either switching entirely to called shots and abandoning the location roll, or waiting on the parry roll until the hit location is known. Another good example is rolling for initiative, which has generated hours of painful argument about what in the world it represents in-game, at the moment of the roll relative to in-game time. 
The most common Simulationist resolution is handled through Fortune, specifically Fortune-at-the-End. This term refers to a dice roll (or cards, or whatever) which is consulted after all possible pre-resolution description of the actions in question has been delivered. Its alternative, Fortune-in-the-Middle, is not historically observed in Simulationist game design. (See glossary for definitions and links.) 
(These confusing indie-RPG "fortune" terms are like "I shoot him" (roll) (description of hit or miss) vs "We fight" (roll) "Ok, looking at these dice what must've happened is...". Fortune-in-the-middle is a real storygamey technique about regularly interpreting the die roll after it hits the table beyond whatever description is necessary to communicate "it worked, you crush his skull" or "it didn't, the arrow flies wide and strikes a bee". Some details about the flow this creates here. I'm guessing Edwards would say Apoc World is fortune-in-the-middle. You might say a random encounter table is fortune-at-the-beginning.)


A useful way to look at Fortune in much Simulationist play is to think of anything that isn't rolled as being a 100% outcome on an implied roll. The extreme view (see the Purist for System category below) is to interpret the whole shootin' universe as tacitly operating according to the d100 or the 3d6 or whatever that's used to handle character task resolution. 
An entire discussion awaits concerning the shape of dice curves, modifiers' effects, separate vs. incorporated effects, and more. I look forward to this on the forums. Also, more details about resolution in Simulationist games are presented below, when I break down the sub-types in detail. 
Finally, reward mechanics remain a topic of vast debate and design potential in Simulationist games. I think the following historical categories barely scratch the surface.  
BRP style: character improvement is literally a function of play just as any other action, via practice and study. This is the famous "if you succeed with a skill during play, roll over your skill percent between sessions in order to improve." The pitfall is graininess, such that one can then start debating about whether one should learn more or less across ten "hits" against one opponent vs. one hit each for ten opponents, why one does or doesn't learn from a failed attempt, and how study is to be rated and applied (much less how it's to be played) relative to the "experience" methods. 
Hero style: the player gains points simply for being there (despite attempts at parsing it, that's what it amounts to), and the point-allocation based cost of character creation continues to be applied. The character is added to in terms of the points that were originally used to assemble him, and arguably as an expression of the same in-game developmental processes involved. In this case, the point-gains are metagame, but the spending is supposed to use in-game logic, sometimes reinforced by "corralling" sections of the character off from one another. The pitfall is reaching degrees of improvement which themselves violate the genre-level standards of that particular play, which some games overcome by making the intersession correspond to substantial in-game time. 
The thing about improvement making characters so good they get out of genre has a flipside Edwards misses: in D&D for sure on purpose and in other games maybe accidentally--it makes the campaign genre-shift over time. When people who play a different indie game every month seem confused how someone can play the same campaign for years, this is one thing they're missing.

That Edwards doesn't make this observation is trivial--except as an opportunity to reflect that nobody in the narrative sphere has picked it up. Characters changing over time is a staple of Burning Wheel and DiTV, but D&D-style genre shifts over time are anathema and perhaps somewhat aesthetic alien to the scene as they require a serial-fiction mentality rather than a drama or film one, and because a genre shift changes a lot of thematic assumptions ("premises" as Ron might say).

A principle in art is that you never claim any territory you're not going to completely use, and the indie narrativist hyper-focused emphasis on specific kinds of thematic stories and dilemmas--one Big Theme per game--means that dealing with the genre shift that comes from real character improvement may just be like taking out a loan to get a summer home after spending 10 years building the perfect patio. It's just inviting in all these new complications.

As a general emotional principle, GNS tends to want to avoid unexpected complications--including weird rules interactions, weird player interactions, weird genre interactions, weird thematic shifts etc--and DIY play seems to embrace them. This, again, goes back to GNS being about a remedy for bad things seen and DIY being a celebration of possibilities--a familiar split in art in any era ("whoa, chaos, let's not let this new thing called oil paint/abstraction/hiphop/roleplaying hurt people" vs "whoa have you seen what you can do with oil paint/abstraction/hiphop/roleplaying ") . Faced with the radical text of D&D and its immediate successors ("anything that you can do in life has an analogue here in the game--plus magic") GNS folks want to experiment but only in order to devise a more reliable game, DIY seems happy to see experiment as part of the fun.

In either case, the key issue is that character change potentially disrupts the current relationship among the components of the character. Options to fix the problem are generally unsatisfactory: (1) slow it down, and (2) permit only tiny changes. One option, rarely seen, is to include kind of a secondary, add-on game with its own set of components, as with Rune status in RuneQuest. (I realize that not everyone knows all of the games I'm referencing, and I certainly don't have all historical RPGs memorized. This topic definitely calls for more discussion in the forums, where we have room to describe all the various examples in detail.)  
The diversity of Simulationist game design 
Here's a quick overview of existing diversity in Simulationist play. I'm focusing on fun, functional, coherent play - none of the following is a criticism or indictment. Also, I've tried to represent as many creator-owned titles as possible, but I'll refer to others as needed. 
My overall point is that, although Simulationist play is defined as prioritizing Exploration of the five elements, its diversity is not a five-headed, one-element-per-submode hydra. All five elements are always involved. In defining the subtypes of this mode of play, here are the issues: (1) whether Exploring System is primary, and (2) which of the other elements are necessary "support" or "chassis" and which ones are diminished in emphasis. 
Purists for System  
What games are these? EABA, JAGS, SOL, Pocket Universe, and Fudge are deliberately "generalist" regarding setting. The big commercial models are GURPS, BRP (in its "unstripped" form), DC Heroes (now Blood of Heroes), Rolemaster, D6 (derived and considerably Simulationized from Star Wars), and the Hero System (as such, mainly derived from Danger International and Fantasy Hero rather than early Champions). Whether D20 should be included in this category is a matter for some debate.  
These games' five-element structure is consistent: System + Color thereof, Setting, then Character + Situation. I'm trying to think of one which switches the role of character before setting, which might include some some superhero games. It might seem odd that Color is placed so high in priority, but consider the engineering-text model for the game text of GURPS - this is, actually, Color for System.  
A lot of people have trouble with the notion of "Exploring System." They argue that playing a game like Fudge is necessarily Setting-first. I disagree, but this debate properly belongs in the forums. 
In these games, the System is all about Fortune and all about Currency. 
Regarding Fortune, probabilities are the key to achieving the basic Simulationist internal-cause priority. Consider both comparative probabilities among characters at a given moment as well as probabilities in transition within a character over time - in action (actually resolving tasks), these are what drive the game. For these games, a unified probability mechanic to handle any game-modelled instance is the ideal, usually resulting in a single tables-based concept such as the Universal Table in DC Heroes. 
Purist-for-System designs tend to model the same things: differences among scales, situational modifiers, kinetics of all kinds, and so forth. The usual issues surrounding incorporated vs. unincorporated effects, opposed vs. target number mechanics, the interaction of switches and dials, and probability-curvature shape are therefore the main things to distinguish these systems from one another. Compared to other designs, high search and handling times, as well as many points-of-contact, are acceptable features. (Please see the Glossary for the definition of points-of-contact). 
The idea of "purist-for-system" is just raw error. This is just more of Edwards not getting why any number of gamers might want the horses in Westworld to work kinda like (meaning in terms of the abstraction of RPGs, analogously to-) real horses. 

DC Heroes is instructive. DC systems deal with the perennial scaling problem of how do we get Batman and Superman in the same system, with folks like Wonder Woman and Martian Manhunter in-between and then magic on top, etc. Without a system that genuinely scales, playing Batman trying to outwit Superman is tactically boring because it's too easy.
Necessity is the mother of invention and without a scaling system Batman doesn't have to invent the sonics to hurt Superman's super-hearing, Batman doesn't have to invent the suit that feeds off the coldbringer missile's detonaion, Batman doesn't have to get Green Arrow to shoot that kryptonite arrow, or synthesize kryptonite, etc etc

A narrativist player would be happy to say that its Batman's comic and he's been accumulating Plot Points for the last 4 issues so he has better than even odds of winning (or, AW style, winning at a cost) and then add all these tactical complexities as color on top, but a challenge-oriented player wants the numbers against them (the Edwardsian narrativist emphatically does not want the odds against getting the story to be there) and wants to try to figure out how to overcome them--and have the color be the result of that.

That is: to face something more analagous to Batman's problem than to Frank Miller's.

Edwards continues in this vein for a few more paragraphs about Purist-For-System which I'll skip, before saying:

As I see it, Purist for System design is a tall, tall order. It's arguably the hardest design spec in all of role-playing. 

Because rather than seeing it as a technical design to allow for certain genres to exist (or, in the case of GURPS and its ilk, coexist) as a challenge or even as a natural plot-engine (if I have to invent kryptonite, Alfred may well ask me what I'm doing), he assumes this detail is basically an all-devouring fetish for the perfect system.

Significantly, GNS acknowledges that players can drift from goal to goal (challenge-desire, to plot-desire, etc) but not the reasons why they might drift--especially the role of the game-as-played and its causal engine as a collaborator in suggesting goals. Edwards imagines the player having goals (perhaps influenced by the game's text and marketing) and then hoping/demanding the game satisfies them. He does not see how a sufficiently robust cause-and-effect engine generates unintended consequences out of player actions and therefore offers new tactical and thematic opportunities/problems which might get a player to change what they're interested in.

For Edwards, Westworld (and a game system) is like a restaurant: you show up and try to get what you want out of the options it provides. For many other people it's more like an ever-changing smorgasbord--options abound and if you go at the kipper platter long enough, the consequence is that its replaced with a bowl of nachos, and then you decide whether you want nachos. To a GNS fan, a causal engine which goes around spitting out themes, plots and enemies willy-nilly (the bank robber you shot was the son of the mayor of the next town over, the old mine is full of dynamite, the father owns the mine, the gun jammed so you had to steal a new one, the one you stole was the one the old rancher was depending on to defend himself from the Glantons...) may seem chaotic, even....incoherent?...but this chaos is not just endured to get to the challenge or character moment or satisfying scenes scooped from it, it is often welcomed for its ability to continuously surprise players as a full collaborator with their own creativity--another player at the table.

Edwards is right in not seeing starkly classical moral-thematic plot easily emerging here (at least not unmolested by a handful of other interests), he is wrong in not seeing this causality as a force in its own right with a constant appeal to G and N impulses beyond a desire to watch the clockwork tick.

I would suggest that players harnessing and satisfyingly "narrativizing" an emergent situation is a lot more common than Edwards suggests but its hard for him to recognize it without the Classic Plot attached. I'd also suggest that it's especially hard for GNS to recognize (with its obsession with assigning optimal modes to systems) because a common part of the process is allowing casual contact and less-creatively-invested incident to build a connection between player and character until the character begins to take on a thematic resonance that then can be more consciously manipulated--much like Miller took Batman (a character who began life as the uncomplicated star of several serial, disconnected, tactical, wholly challenge-driven adventures) and invested in themes that emerged (often causally) around the character until a Big Story (easily RPGable from both a tactical and premise-driven POV) resulted.

Now Edwards moves on to...
High Concept  
In cinema, "High Concept" refers to any film idea that can be pitched in a very limited amount of time; the usual method uses references to other films. Sometimes, although not necessarily, it's presented as a combination: "Jaws meets Good Will Hunting," or that sort of thing. I'm adopting it to role-playing without much modification, although emphasizing that the source references can come from any medium and also that the two-title combo isn't always employed.  
The key word is "genre," which in this case refers to a certain combination of the five elements as well as an unstated Theme. How do they get to this goal? All rely heavily on inspiration or kewlness as the big motivator, to get the content processed via art, prose style, and more. "Story," in this context, refers to the sequence of events that provide a payoff in terms of recognizing and enjoying the genre during play. 
High-concept seems a bit strange because while, yes, Edwards is using the phrase the same way its used in the rest of art discourse (for once), it isn't clear why this is somehow a category that can't overlap in any number of ways with others. If DC Heroes had exactly the same "purist-for-system" system but the default setting was Watchmen would it be High Concept?

He seems to be categorizing these games based on why he thinks other people like them rather than on how they actually work.
This sort of game design will be familiar to almost anyone, represented by Arrowflight (Setting), Pax Draconis (Setting), Godlike (Setting), Sun & Storm (Setting + Situation), Dreamwalker (Situation), The Godsend Agenda (Character-Setting tug-of-war), The Collectors (applied Fudge, Situation + Character), Heartquest (applied Fudge; Character), Children of the Sun (Setting), Fvlminata (Setting), and Dread (Situation + Character), Fading Suns (Setting), Earthdawn (Setting), Space: 1889 (Setting), Mutant Chronicles (Setting), Mage first edition (Character), Mage second edition (Setting), Ironclaw (Setting), and Continuum (Setting with a touch of System). Many Fantasy Heartbreakers fall into this category, almost all Setting-based. 
Some of the best-known games of this type include Tekumel, Jorune, Traveller (specifically in its mid-80s through mid-90s form), Call of Cthulhu, Pendragon, Nephilim, Feng Shui, the various secondary settings for AD&D2 like Al-Qadim, and quite a few D20 or WEG games which rely on licensing. I am coming to think of D20 as a kind of High Concept chassis, a very new and interesting development in RPG design.
 Also, most incoherent game designs are partly or even primarily High Concept Simulationist as well, with AD&D2 and Vampire (first edition) as the best-known examples.  
It seems to just be code for "a lot of setting stuff included in the game", but then...

At first glance, these games might look like additions to or specifications of the Purist for System design, mainly through plugging in a fixed Setting. However, I think that impression isn't accurate, and that the five elements are very differently related. The formula starts with one of Character, Situation, or Setting, with lots of Color, then the other two (Character, Situation, or Setting, whichever weren't in first place), with System being last in priority.
Translation of that last sentence:

"starts with" = "the pitch is"

"Character, situation, or setting" = "You are a vampire, The world is Dark Sun, or You are about to meet Cthulhu or whatever"

" then the other two" = "and then if you read it you find out about"

"last in priority"= "not as much part of the appeal, as far as I can tell, from reading the ads in Dragon and sitting at the far end of the table guessing why people like these games"

This might seem complicated or different than what I said above but its not: "character, situation and setting" pretty much cover the range of things covered in a Hollywood high-concept pitch.

 Character creation is far more delimited as well, relying heavily on Setting and Situation. In this case, the "points" are pure metagame for purposes of making characters; they don't reflect or underly the universe in action as in the Purist for System games.
I also recommend examining Theme carefully. In this game, it's present and accounted for already, before play. The process of prep-play-enjoy works by putting "what you want" in, then having "what you want" come out, with the hope that the System's application doesn't change anything along the way. 
DC Heroes (allegedly "purist-for-system" and not "high concept") has built-in-themes and the "points" don't represent anything. Unless he means that in DCH everything is measured in "AP"s (sound, damage, money, etc) which: really? You think that fundamentally changes how people play?

Starting characters tend to be very colorful

DC Heroes.

...and described by many terms and numbers, 

DC Heroes.

but relatively static: waiting for their hook, so to speak. 

DC Heroes.

Hooks are often built-in; unlike the Purist for System methods, the player-to-character relationship usually includes my second "role level" in addition to the third and fourth.  
[From the glossary:
Roles, "role levels"
(1) The player's social role in terms of his character - the mom, the jokester, the organizer, the placator, etc. (2) The character's thematic or operational role relative to the others - the leader, the brick, the betrayer, the ingenue, etc. (3) The character's in-game occupation or social role - the pilot, the mercenary, the alien wanderer, etc. (4) The character's specific Effectiveness values - armor rating, weapon attributes, specific skills and their values, available funds, etc. See The class issue and all internal links.]
DC Heroes.

Quantitatively, the more common character creation methods (which are not unique to Simulationist design) include less layering but more nesting (i.e. options within options, as well as the one-from-column-A, one-from-column-B approach established by Vampire), and almost always the relatively clumsy "GM approval" proviso. The specific method is usually based on points, but sometimes with Fortune methods to render a character role/type less likely to occur (making them more expensive with points also aims at this function). Notably, in-game money isn't modeled by the point-system during play. 
The System is not all about Fortune, either, and these games can be very uneasy in this regard. Dice-based resolutions sometimes represent much noise and effort about not much effect, i.e., random factors tend not to deviate from expected results very much.

....say: 1-in-20? Or: about often enough for it to be a surprise and an event? If only there were, say, 40-50 years of testing to see if about 1 out of every 20 times is rare enough to make an extreme result seem viscerally exciting when it coincides with a critically important roll.

 Some games display a small range of possible Effect (i.e. damage rarely harms an opponent very much at a time), slight metagame adjustments to minimize extreme results, or a lot of offered strategies for the GM to soften or redirect the effects that occur. 
These last are just railroading and illusionism. Yes: they suck. But they are of special interest to Edwards because at this point he's trying to find systemic solutions to wrestle "story" away from the spectre of the Evil Controlling GM.

Points-of-contact are far more directional; things which aren't relevant to the Explorative focus are often summarized and not "System'ed" with great rigor. When done well, such that the remaining, emphasized elements clearly provide a sort of "what to do" feel, this creates an extremely playable, accessible game text. Dread offers the perfect example for the lower points-of-contact end; Arrowflight and Godlike are similar but more reassuringly nail-even-the-irrelevant-down at the higher points-of-contact end. The truly outstanding games illustrating this latter approach are Call of Cthulhu, Unknown Armies, and Pendragon.  
However, when it's done badly, resolutions are rife with breakpoints and GM-fiat punts, and a great deal of effort during character creation yields little sense of what the character is is about to do. 

Why would anyone need that? Especially the supposed simulationist.

High Concept play can be divided neatly into those which are greatly concerned with "the big story" and those which are not. Historically, the latter used to be the most common: Call of Cthulhu, Jorune, or more recently Dread and Godlike, in which "the story" only refers to a record of short-term events and set-pieces. However, following the spearhead for this type of game text, Ars Magica, now the long-term story-type is more common. A lot of internet blood has been spilled regarding how this phenomenon is or is not related to Narrativist play, but I think it's an easy issue. The key for these games is GM authority over the story's content and integrity at all points, including managing the input by players.

A mundane point in 2018, but one the Pop Usage Example didn't grasp and a lot of people don't: narrativism isn't about story, there being a story "promoting story" etc, it's about plot and specifically players directly influencing the plot.

Even system results are judged appropriate or not by the GM; "fudging" Fortune outcomes is overtly granted as a GM right.  
The Golden Rule of White Wolf games is a covert way to say the same thing: ignore any rule that interferes with fun. 

Edwards ties together three phenomena:

-GMs controlling the plot


-Ignoring rules

The first two are obviously related. A GM usually fudges to get a given thing in-game to occur (control plot). Ignoring rules, however, can and often is utterly unrelated and is by no means a thing done only in response to the GM's desires. Rules can be ignored because they don't suit anyone's purpose and in ideal practice this can't be used to control plot because the players have to ratify rules changes. In a properly-customized ruleset, nobody at the table when the rule changes has anyone to blame but themself.

Edwards--well on the way to the kind of Always Follow The Rules We Handcrafted Them From Our Norwegian Vegan Wood They're Perfect ideologies that characterize hippie narrativist games--assumes a totally adversarial power-struggle will ensue in the absence of clear procedures, with obvious consequences:
No one, I presume, thinks that any player may invoke the Golden Rule at any time; what it's really saying is that the GM may ignore any rule (or any player who invokes it) that ruins his or her idea of what should happen.  
The functional version of such play is properly called Illusionism, which has undergone a good deal of debate and clarification at the Forge (see glossary).
"Illusionism" is giving the players the illusion their choice mattered to the plot. Ron spends some time going over some various railroady texts and railroady advice which anyone familiar with mainstream RPGs knows well.

He makes this distinction:

Players are enjoined to immerse, by which they mean "keep your metagame [plot demands -z.] agenda out of it," at the aesthetic level. It's best understood as Illusionism by full consent, which is what keeps it from being railroading,

I agree that consent (and awareness) are what keep a thing from being railroading and, even at low levels of "railroadiness" essentially it's player consent to a restrictive structure that marks the precise border of "railroad". Semi-exhaustive discussion of that here.

I’d also note that there’s a later term not used here where the players essentially consent to a situation where they don’t look too hard at their choices to see if they’re being nullified called “participationism” and I’ve heard the GNS expert mentioned at the top of this essay say that he’s into that now because he can’t imagine how he’d handle the PCs trying to flood the dungeon.

Participationism is also commercially important for two reasons: it’s super common in “celebrity” streaming play (though we didn’t do it on I Hit It With My Axe) because of that kind of play’s dual nature as spectator-sport and actual play—and fixed-plot adventures are easier for mainstream RPG companies to create and market.

Probably super-ethical RPG theory for 21st century would demand that mainstream companies urge DM’s to at least tell players that sometimes their choices will be nullified or quantum-ogre’d (the path forks, there's a choice, both lead to the same ogre) in the name of the “story” if they want to use the modules. But a lot of GMs like illusionism and still disagree with this approach (see the comments).

The whole problem can be avoided by buying sandbox stuff but a lot of people aren’t cool and smart enough for that hardcore lyfe so whatevs.

After a couple closing remarks, Edwards moves on to a topic which apparently has been controversial in his own circles up til that point.

In what follows, he’s basically trying to separate the nascent baby of narrativism from the bathwater of other experimental and arty RPG practices that also aren’t mostly about trying to kill monsters. He’s trying (with who knows how much success) to explain that, from a GNS POV—that is from the point of view of “metagame” (plot demands of gamists and narrativists) these games are technically in the same categories as GNS whipping boys (the "incoherent mainstream") AD&D2 and Vampire—while not offending fellow arty hippie role-players.

Without that social context, a lot of this doesn’t make sense, but here goes:

Rules-lite Story or Character priorities  

This section is likely to get me into trouble, so I'll tread carefully. I suggest that many self-described "rules-lite" or "story-oriented" role-playing games represent a derived version of the High Concept model, slanted heavily toward Situation - especially Situation which is under complete GM control, overt or covert.

I believe he's referring to things like "Ok, now we're all gonna see what it's like to be a political prisoner"-style games where the high concept is arty and the goal is in some sense actorly. "Being a political prisoner" is the "Situation"--situation being a thing that takes place in a larger setting (sometimes implied, like "the real world") and, int his case, usually strongly implying a Thematic Premise, though one that might be baked in an predetermined in a way Edwards' wouldn't call "Narrativist". Like narrativism, the Premise may be provocative, but unlike narrativism, the premise is baked in: it doesn't ask you to consider an ideological question but basically makes provocative ideological assumptions and subjects you to them. Not Hamlet but rather Survival In Auschwitz.

This also clarifies something I noted in passing in the last essay: Edward separates Color from Situation. This is unnecessary for a sandbox roleplayer of any kind (any element of Color might be seized on and used to manipulate the situation and thus become part of it or even have been secretly part of the Situation all along in the case of clues hidden in plain sight in GM description), but for a sadbox roleplayer (Now we will play ze game vhere you are subjected to ostracism!) the "situation" is specifically the ideologically provocative bit.

Players get to contribute tons of Color, even content, but never outcomes or final-resolutions, and playing the character as conceived is the first priority, sometimes taken to extremes of Actor Stance (e.g. Turku play, see the Glossary). Character and Situation are prioritized with Color, with Setting next, and lastly the formal System, which is slanted strongly toward Drama-mechanics. This mode of play may be strongly linked with LARP crossovers. 
Here's my point: in application, a covert System is heavily, heavily entrenched, regardless of whatever to-hit modifiers or dice rolls have been peeled away. This system is based on Social Contract (what we all agree is "good" or "fun") and Social Context (i.e. the subculture that players belong to), and it is sternly reinforced through these means. I think it's significant that literal referees - on-the-spot judges of what can and cannot happen - are a necessary feature as soon as groups get beyond a certain size.  
It's not just High Concept though. It looks like it - the heavy emphasis on story/genre, with overt eschewing of System, but it's also (a) actually pretty heavy on Drama-driven or Karma-driven System and (b) emphasizes customizable Settings as in Purist for System play. So I think it's worth its own category.  
From the introduction to Theatrix (1993, Backstage Press, authors are David Berkman, Travis Eneix, and Brett Hackett): 
 Making a story come to life can be a difficult task. Previous generations of game systems have been rules bound, trapped within their own structure and rigidity. We wanted to produce a game that would help you in every way, not hinder you. So we developed a system of rules that is written to evolve along with your style of storytelling and roleplaying. These rules can be used to guide every facet of the game's progress, without becoming intrusive. You can use all the rules, or easily peel them away in layers, until you're running free-form games. The rules heavily encourage adopting this style of play, making themselves unnecessary.
In other words, the system helps create story by fading away, much like the Dictatorship of the Proletariat. I think that this whole design effort arises from a desire for "big story" in the face of Purist-for-System design and mainly Fortune-driven High Concept design. In the effort to get out of that sort of Simulationist play, the thought is to get rid of the System that supports it, with any explicit System being perceived as that sort of system. I consider this a problematic design goal but it's widespread enough to merit a category. What makes it difficult to discuss is that its explict story-creation goals are similar to those of Narrativist play, but the operational process is stripped-down High Concept Simulationism. (See the GNS stuff below for further discussion.) 
Basically Edwards is criticizing is a game which starts off like you're feeling what it's like to be sad in a box but then moves on to allow the player to go beyond being sad in a box and maybe morally and thematically question boxness. He is skeptical you can't do this without clear rules about who controls what. He is skeptical you can do anything without clear rules about who controls what.

He finishes this thought, names a few examples then...

Setting-creation and universe-play mechanisms  
Another derivation of the Purist for System approach brings the Setting creation process directly into play itself. The System-driven elements of the Setting are as "active" as any particular character might be, during play as well as during preparation. Basically, the setting is played, even created, as a part of regular play.  
Boink! I just realized that the original Traveller, or at least one way to play it, represents an example of this approach. Star system and planet creation are written right into the process of play, such that adventures and missions become not only a means of enjoying and improving characters, but also a means of enjoying and basically mapping the game-space.
This is basically the exploration/hexcrawl/Red Marches situation (we start with a vague map and clarify as the group chooses to explore it), which was not big on Edwards' radar at this time.

This is very distinct from later versions of Traveller, which were emphatically High Concept with a Setting emphasis. (Oh, and just for credit where it's due, I should also mention that Traveller pioneered the mechanics of overt character-creation-as-play.)
This mode of play is not merely creating more setting through preparation as play progresses. It relies on doing so in a system-driven fashion much like character creation, carried out as an overt or near-overt part of actual play.  
This, 4e trolls, is why you can die in some versions of Trav character  creation--it's a solo game of trade-offs and risks you take to try to get a cool character.
It's a pretty rare form of play and design, probably because the economics of splat-book publishing overwhelmed the hobby, and Traveller itself, from the mid-1980s onwards. The more recent examples include Aria, Multiverser to some extent, and the currently-in-development The Million Worlds. The design spec is to achieve the Color/kewl power of High Concept with the uncompromising power and consistency of the Purists for System, via inserting the explicit metagame world-creating ability. I think this approach is interesting for the level of Director stance potentially involved and I look forward to more role-playing evolution along these lines. 
Historical note: BRP 
Pound for pound, Basic Role-Playing from The Chaosium is perhaps the most important system, publishing tradition, and intellectual engine in the hobby - yes, even more than D&D. It represents the first and arguably the most lasting, influential form of uncompromising Simulationist design. 
If you take out the word "simulationist" and put in any number of words meaning "rpg but not narrativist" ("trad" is better than nothing) then I totally agree.

And Basic Role Playing is: D&D stats + % skill system (only sometimes derived from those stats) + bolt-on mechanics for the specific setting you're playing (Sanity for Cthulhu, personality metrics for Pendragon, etc). It's the AK-47 of RPGs--extremely flexible and robust because it's not clever or focused. If you wanted to invent a game tomorrow for a  random genre and do a one-shot, just plug the skills from that genre into the BRP skill list and you've got a (trad) game that's good enough for government work.

"Basic Role Playing" is a good name for it: it's within a stone's throw of any more-focused design you might wanna make after. Greg Stafford was Egon Schiele to Gygax and Arneson's Gustav Klimt.

It's kind of hard to discuss just how it was influential, as its very first appearance as a pamphlet accompanying a boardgame wasn't widely distributed. The influence operated primarily through the popularity of both RuneQuest and Call of Cthulhu. Looking across the early versions of these games as well as Superworld, Questworld, and more, I think BRP is identifiable as a Purist for System design and publishing. It's really probably the precursor for the later GURPS mode of publishing.  
However, that vision, plan, or phenomenon, whatever, swiftly evolved into High Concept, both in RuneQuest (Setting) and Call of Cthulhu (Situation) as they hit their early-mid-80s forms, which is what most people are familiar with, I think. Call of Cthulhu remains High Concept to the present day, whereas RuneQuest, upon being licensed to and redesigned to the specifications of Avalon Hill, essentially evolved into a new Purist for System game, with the setting, Glorantha, relegated to the background at most. Moving into the late 80s and early 90s, the new BRP games (Pendragon, Nephilim) represented fairly radical Drifting of Cthulhu-style BRP into their respective High Concepts. 
GNS crossover issues 
As usual for GNS-heavy text, I'll speak of games themselves in the GNS terms, but with the proviso that I'm really speaking about the play itself that is typical of or best supported by the rules of those games. 
First, the FAQ 
Q: Can Simulationist design be Abashed? 
A: Sure. "Abashed" refers to design that must be Drifted in order to play because incompatible priorities are present among different parts of the rules. It's different from Incoherent design in that such Drift is easy and minor. Technically, an Abashed game is already at least two modes (or sub-modes); e.g. I've said that Little Fears represents Abashed Narrativist design, but it should really be called Abashed Narrativism/Simulationism. 
Q: So "Abashed" means combined? 
A: No. Combined GNS modes which work well together would be "Hybrid." There's a whole section on that below. Abashed games must be Drifted (i.e. their rules must be operationally changed, or some sections ignored) in order to play. 
Q: Can Simulationist play be Vanilla? 
A: Well, we don't say Vanilla and Pervy any more (too rude for some, apparently). Now we talk about Points-of-Contact being low or high for given portions of rules. But to lapse back into the old terminology, yes, it can. Dread is a veritable poster child for Vanilla Sim, which I would generalize to mean a High Concept Simulationist design with low Points-of-Contact and a high emphasis on Situation. Pervy Sim basically just ups the Points-of-Contact as well as the emphasis on Exploring anything regardless of topic, which pretty much describes any member of the Purist-for-System category. 
This refers to some old online-RPG jargon that, sadly, has fallen by the wayside. 

I mean "sadly" not just because I'm a porn actor but because we lost a gradation of meaning here:

The phrase "points of contact" means how much you look stuff up. High p-o-contact and low p-o-contact used to be referred to as "pervy-vs-vanilla". The metaphor that equates using crunchy rules with a sexual fetish was (and continues to be) extremely important in Edwards' mentality toward "simlationist" games. Some people said the liked looking stuff up and Edwards assumed that was the fun. The whole concept of "purist for system" assumes that these rules are there purely because people like them as an end in themselves and that this desire is a "black box" with no further utilitarian meaning.

This also scans a flashlight across some things, RPG-historically:

1. A lot of the energy Edwards and his predecessors put in here was about simply disentangling the "crunchy, complicated" quality of some RPG from being specific to any given philosophy, approach, or goal of RPGing, including "realism". I don't know how stupid things were before they got there, but experience suggests: very. Which means we owe them a debt for at least that.

2. The decision to move from the Vanilla/Pervy language to "points of contact" may reflect the sexual and overall aesthetic conservatism that was going to come to dominate the post-GNS indie scene (not so much because the members were all ok with it, more that the ones that were managed to seize the moral and social high ground and everybody else just grinned and let it slide for social and eventually commercial reasons, as Ron does--"too rude for some, apparently"). I'd love to see whatever discussion that decision was made in.

3. Making handling RPG books kinda fun (not just easy, but fun in itself) is a goal (though not a primary goal) of some DIY RPG design and these are (rare and unusual) examples of enjoying system for its own sake. The idea is 2-5 times a session you pull out a terrible chart, like the d1000 mutation table, with full pervy performance like "let's see...Cloud of Flies, Cloven Hooves, Cowardice...ah, Crest--you grow a crest on your head!". Play Vaults of Vyzor with Jeff Rients for full perv action when he pulls out his Arduin Crit Table. This is an example of acknowledging that if you have these kinds of esoteric mechanics at the table they take up space, and once you take up space you need to make that fun.

Character generation  
Character generation text and methods are extremely diverse within each GNS mode, which is one of the reasons I favor group communication during this phase of pre-play. For instance, some Gamist-ish games utilize point-allocation systems, which looks similar to the widespread method in Simulationist-ish games. However, for Gamist purposes, this method is all about strategizing tradeoffs, rather than establishing a fixed internal-cause to "justify" the character. Similarly, Gamist character creation utilizing Fortune methods isn't the same as the few Simulationist randomized methods - in the former, it's a lot like gambling, whereas in the latter, it's about a character maturing through Fortune's vagaries represented by in-game effects like culture, weather, disease, and so forth (e.g. Harnmaster).  
Culture, weather and disease are all weaponizable, from a challenge Pov, and all actionable from a narrative pov.
Narrativist character creation in some games requires a fair amount of back-story, just as some Simulationist play does, but in the former, it's about establishing a chassis for conflict, metagame [plot demands], and reward, and in the latter, it's about Coloring the character and providing opportunities for GM-created hooks.
Again the GM creating a hook and then a player narrativizing the hell out of it doesn't seem to be an important thing to Ron although, deciding to play a focused narrativist RPG like Dogs In the Vineyard is essentially responding to another person's hook--just a game designer's rather than a GM's. And the difference in consenting to one rather than the other should be minor.

 I rank the conflict between these concepts, during play, among the highest-risk situations for the survival of a gaming group. Strategies to resolve this conflict, whether social or design-oriented, are currently not well-developed in the hobby. 


He then goes on to talk about "metagame" a lot. I've replaced the word with "demanding-stuff-from-the plot" to clarify:

Demanding stuff from the plot mechanics  
The term "Demanding stuff from the plot" is problematic throughout this essay for Simulationist play and rules design. Demanding stuff from the plot mechanics, by definition, entail the interjection of real-people priorities into the system-operation. Now, it is foolish to speak of Simulationist play as lacking Demanding stuff from the plot; that would only apply if the people at the table were themselves rules-constructs as well as the rules, and that's silly. But compared to Gamist and Narrativist play, Simulationist play may be spoken of as lacking Demanding stuff from the plot [i]interpersonal agenda[/i], like "winning" or "doing well" in Gamism, or addressing a Premise in Narrativism. Its Demanding stuff from the plot, although fully social, is self-referential, to stay in-game. I recognize that it's a problematic issue and I look forward to some discussion about it.  
To clarify for purposes of the essay, compare the following: (1) an in-game essence or metaphysical effect called "Karma," which represents the character's moral status in that game-universe according to (e.g.) a god or principle in that game-world; (2) a score on the sheet which has literally nothing to do with the character's in-game identity, also called "Karma," recognized and applied by the real people with no in-game entity used to justify it. In both systems, Karma is a point-score which goes up and down, and which can be brought into play as, say, a bonus to one's dice roll. But I'd say that #1 is not Demanding stuff from the plot at all, and #2 is wholly Demanding stuff from the plot.
Mechanically, how do they differ? One thing to consider is how the score goes up and down - by player-use, or by in-game effects? Another is whether the score is integrated with the reward/improvement system - does spending a Karma reduce one's bank of improvement points? In fact, is Karma a spent resource at all? Still another issue is whether in-game effects must be in place, or inserted into place, to justify its use. No one of these indicators is hard-and-fast, however; one must consider them all at once, and how they relate to Simulationism (and non-Simulationism) is a fascinating issue. At this point I tend to think that the main issue, basically, is who is considered to "spend" them - character or player. 
I suggest that Trouble in Orkworld, Hero Points in Hero Wars, and Spiritual Attributes in The Riddle of Steel are Resource-based Demanding stuff from the plot mechanics, whereas Power in RuneQuest, Sanity in Call of Cthulhu, and these mechanics' many derivatives in other games, are straightforward, non-Demanding stuff from the plot Resources. Similarly, I suggest that the role-playing bonuses based on out-of-game neatness in Sorcerer are Demanding stuff from the plot, whereas the Stunt rules based on difficulty or unlikelihood in Feng Shui are not. 
It's a tough discussion, though. One confounding factor is that Demanding stuff from the plot mechanics are often present as "fixes" of otherwise-Simulationist systems that proved to be mildly broken in play. The trouble with such a thing is that it can lead to serious Drift of the sort that breaks Social Contracts or renders systems incoherent. 
He may be referring to like fate points in Warhammer (which he calls incoherent, iirc). Someone somewhere pointed out the get-out-of-death-free point of Fate pts in Warhammer: they show you (more clearly than a Saving Throw) that, mechanically, you should've died in this extremely brutal system and you only got out of it because of fate points, so up your game next time. 

As far as I can tell, Simulationist game design runs into a lot of potential trouble when it includes secondary hybridization with the other modes of play. Gamist or Narrativist features as supportive elements introduce the thin end of the metagame-agenda wedge. The usual result is to defend against the "creeping Gamism" with rules-bloat, or to encourage negatively-extreme deception or authority in the GM in order to preserve an intended set of plot events, which is to say, railroading. In other words, a baseline Simulationist focus is easily subverted, leading to incoherence.  
Whether this issue can be resolved by future designs and Social Contracts is unknown. Speaking historically, though, AD&D2, Vampire, and Legend of the Five Rings are especially good examples of incoherent design that ends up screwing the Simulationist. You have Gamist character creation, with Narrativist rhetoric (especially in Vampire). You have High Concept Simulationist resolution, which is to say, easily subverted by Gamism because universal consistency is de-emphasized.
He's confusing Gamism (which he equates with challenge) with a specific kind of self-imposed challenge: munchkinism--or "character op"--that makes a challenge of gaming the system and taking advantage of system-specific tactics. System-agnostic challenge doesn't do this.

 And finally, you have sternly-worded "story" play-context, which in practice becomes game-author-to-GM co-conspiracy. The net result is a fairly committed Simulationist GM presiding over a bunch of players tending toward more agenda-based play of different kinds. 
What happens? All the wedges widen, and the unfortunate thing is that the more everyone likes the basic, fun interest of the topic ("genre") at hand, the worse the rift becomes. 

  • The aggravated Narrativist leaves the play situation after butting heads with the GM over the "story." Arguably, the early White Wolf games in general are responsible for what amounted to a mass exodus of Narrativist-oriented role-players from the hobby in the mid-1990s.
...and GNS.

  • The Gamist runs rampant, moving from sportsmanlike challenge/competition (as would be found in a coherent Gamist design) to "break the system" vs.-game, vs.-GM challenge/competition. The group typically either dissolves or evicts the Gamist player; evictees find one another and enjoy themselves with gusto, Drifting the rules significantly and focusing on player-vs.-player challenge/competition. They tend to be quite public and large-group oriented, via on-line and LARP play. [AEG was clever enough to recognize this phenomenon and incorporate it into the L5R market strategy.]
  • The Simulationist, whether GM or player, fights a losing battle against the Gamist, often feeling betrayed and desperate. Simulationist groups which survive this conflict tend to be very insular, clique-ish, and GM-centered, with the GM seen as the conduit or channeller to "the game" as published. Such a GM is usually given carte blanche authority over the social, system, and plot-oriented content of the game, and the players become fairly subordinated to the content of play. The group often Drifts the rules significantly to reflect and reinforce the immediate Social Contract; simultaneously, they become defensive and protective regarding the game title as a subcultural item.
This last one is just "it's hard for me to talk to mainstream gamers, they like their GMs, hack their game and don't want to play my game, I don't get it". And if you wanna see "defensive and protective regarding the game title as a subcultural item" try co-writing a Vampire: TM video game with your transgender author with a bunch of GNS refugees staring over your shoulder waiting for you to Do It Wrong.

Edwards gives some examples of hybrid games, then (my interjections in not-italics)...

Shit! I'm playing Narrativist  
In Simulationist play, morality cannot be imposed by the player or, except as the representative of the imagined world (Westworld), by the GM. Theme is already part of the cosmos; it's not produced by metagame (Plot-demand) decisions. Morality, when it's involved, is "how it is" in the game-world (how things are in Westworld), and even its shifts occur along defined, engine-driven parameters. The GM and players buy into this framework in order to play at all.  
The point is that one can care about and enjoy complex issues, changing protagonists, and themes in both sorts of play, Narrativism and Simulationism. The difference lies in the point and contributions of literal instances of play; its operation and social feedback.  
I'll provide two examples, a simple one and a complex one.  
The simple one: Consider the behavioral parameters of a samurai player-character in Sorcerer and in GURPS. On paper the sheets look pretty similar: bushido all over the place, honorable, blah blah. But what does this mean in terms of player decisions and events during play? I suggest that in Sorcerer (Narrativist), the expectation is that the character will encounter functional limits of his or her behavioral profile, and eventually, will necessarily break one or more of the formal tenets as an expression of who he or she "is," or suffer for failing to do so. No one knows how, or which one, or in relation to which other characters; that's what play is for. I suggest that in GURPS (Simulationist), the expectation is that the behavioral profile sets the parameters within which the character reliably acts, especially in the crunch - in other words, it formalizes the role the character will play in the upcoming events. Breaking that role in a Sorcerer-esque fashion would, in this case, constitute something very like a breach of contract. 
He then gives a more complicated example comparing Pendragon (Sim) to Riddle of Steel (Nar), pointing out the Pendragon guy basically doesn't have a lot of choices. Then:

I may be a little biased about this issue, but it seems to me that a character in Narrativist play is by definition a thematic time-bomb, whereas, for a character in Simulationist play, the bomb is either absent (the GURPS samurai), present in a state of near-constant detonation (the Pendragon knight, using Passions), or its detonation is integrated into the in-game behavioral resolution system in a "tracked" fashion (the Pendragon knight, using the dichotomous traits). Therefore, when you-as-player get proactive about an emotional thematic issue, poof, you're out of Sim. Whereas enjoying the in-game system activity of a thematic issue is perfectly do-able in Sim, without that proactivity being necessary. 

Before anyone flips out, stop for one more point, which is that my perceived time-scale of play for all the above points is quite high. I'm talking about whole sessions and sets of sessions, not moment-to-moment combat decisons or dialogue. So the "poof" is a pretty prolonged thing (and I better not develop this metaphor any further either). 

Many people mistake low time-scale techniques like Director stance, shared narration, etc, for Narrativism, although they are not defining elements for any GNS mode. Misunderstanding this key issue has led to many people falsely identifying themselves as playing Simulationist with a strong Character emphasis, when they were instead playing quite straightforward Narrativist without funky techniques. 
There are probably ways to change pop misconceptions about RPG terms--using this much jargon's not one of them. Thus: people still saying Narrativist to mean "its got story stuff".

He complains about L5R and then...

El Dorado and Drift
El Dorado is a term coined by Paul Czege based on some ideas proposed by Joachim Buchert (see glossary for links). As originally proposed, it was essentially Narrativist play with a strong Simulationist supportive element - a functional hybrid. When I surprised this debate by shrugging and stating that hybrids, with one mode dominant, are viable, possible, and functional, and when The Riddle of Steel demonstrated an exceptionally fine example, the term changed a bit. Over time, it has come to mean as well an experientially smooth and perhaps even unnoticeable shift from Simulationist play-assumptions to Narrativist ones. 

Some technical stuff and then...

I'll discuss this issue in much more detail in the Narrativism essay, but I'll pose the most serious problem facing the seekers of El Dorado: idealizing story creation but refusing to do it. Oh, am I going to catch it for this section ... well, people are just going to have to disagree about whether stories can "create themselves." 
Personally, I don't think they do, and we won't get anywhere by pushing and pulling. In practical terms, lots of hassles and possibilities arise when expecting story to "emerge" from metagame-absent play. Here are the two extremes which arise. 

  • The bad one: A frustrated Narrativist-ish player takes over as GM and relies on railroading. He or she insists that everyone care about the story, but also insists upon everything going as he or she desires. I consider this approach to rank among the least functional role-playing in existence.
  • The good one: Everyone agrees that story is a wonderful and desirable emergent property, but commits to no metagame meddling or prioritizing by anyone. In theory, this is quite functional, but the tricky part is that everyone also has to accept that story might not happen at all, and to be all right with that.
As I said in the first essay, this second thing sounds perfectly fun to a lot of people--there are a lot of things to do in a campaign and people like variety. Sometimes you end up in a dungeon and sometimes you end up in a wilderness, it's all good. Relax. 

More examples, then:

Pitfalls of design  
The first and most serious problem in Simulationist design is to rely on habit and imitation for some mechanics features of the game and then to try to tack on one's own ideas. I'm not talking about simple influence, which is part and parcel of any RPG design, but the porting of whole assumption-sets out of their integrated contexts with all aspects of the parent game. 

Also the problem with anyone making an RPG theory after another person's RPG theory. Including Ron.

This is very common in Fantasy Heartbreakers and usually results in a lot of broken math. Obviously this problem is not unique to Simulationism, but when it occurs in that context, it's really painful.  
Another serious problem is the ideal of "transparency," especially as applied to the High Concept approach. I cannot help but be blunt: System is experientially inescapable. One cannot make Character, Setting, Situation, and Color "go" without it. Drama-driven systems are just as System as any other, for instance. (See the Transparency entry in the Glossary.)  
Really to remove System requires that anything and everything that happens during play be mediated solely through the Social Contract, without any formalized method even to do that. I think that such play would be awfully difficult, requiring so much negotiation regarding how to play per unit of play as to be hopeless. (Again, I am not discussing well-organized systems based mainly on Drama, which are perfectly wonderful and not subject to these criticisms.)  

I don't know what Ron would make of the many different rulesets for comedy-improv games or actor's-studio exercises. And how they transition to the "systemless" improv of professional comedy and dramatic performance we see on Tv, stage and in movies--do they stop being games?  Were they games to begin with? Does the presence of a system (rules) make them games? Is co-hosting a stand-up show a game or just a format?

Edwards makes a series of technical recommendations for what not to do when making a Sim game which are a combination of regular practical stuff and misconceptions about the purpose of crunch we've already seen then...

For play really to be Simulationist, it can't lose the daydream quality: the pleasure in imagination as such, without agenda. For game design to promote this goal, it must be openly valued and its virtues articulated, not assumed (as it often is) to be "good role-playing" by anyone's standards and hence left unstated. Design should be inspiring and elegant in its own right, promoting the desire to see this Setting or Character unfold, or to see this System do its stuff.  
I now offer a couple of points that are probably going to draw some objections.
It's a hard realization: devoted Simulationist play is a fringe interest. It is not the baseline or core of role-playing, which is Exploration. (Here is where my interpretation of the Scarlet Jester's Exploration differs the most from his original presentation.)
Quite a bit of role-playing theory and design has taken a training-wheels approach, especially using Purist for System games like GURPS, in the assumption that role-playing at the Simulationist "level" or "type" is the necessary skill to develop or grow to any other type. 
I think he's mostly saying that people who wanna kill orcs or tell a story about their guilt about kiling orcs don't need to learn to act first so stop giving them shit about it. Fair.

Another good question is whether the claim is valid that role-playing has been "Sim-dominated" through its history, whether in play or in design. Regarding play, I think all the evidence points to all the GNS modes, and much diversity within those modes, being present since the beginning of the hobby. Regarding design and publishing, I think that we need to distinguish between Simulationist elements vs. coherent design - the former have certainly been widespread, but mainly in incoherent games, with AD&D and Vampire as the chief examples. 
Clarifying here near the end: Role-playing specifically by definition requires simulation but the goals of the people doing it have varied wildly (which Ron appreciates) as have the ways systems can enable them to pursue them (which Ron does not--or not enough).

At this point, the end for sure tells us more about Ron's headspace and ambitions than about games...

The Hard Question 
Well, here it is. Before getting bent out of shape, remember that each mode is gonna get one of these. 
Role-playing is a hobby, leisure activity. The real question is, what for, in the long term? For Simulationist play, the answer "This was fun, so let's do it again," is sufficient.  
However, for how long is it sufficient? Which seems to me to vary greatly from person to person. Is the focus on Exploration to be kept as is, permanently, as characters and settings change through play? Some say "sure" and wonder what the hell I'm talking about, or perhaps feel slightly insulted. Or, is Drift ultimately desirable? Is play all about getting "it" to work prior to permitting overt metagame agendas into the picture? Some might answer "of course" and wonder why anyone could see it otherwise.  
So! Is there an expected, future metagame payoff, or is the journey really its own reward? Is Simulationist play what you want, or is it what you think you must do in order, one day, to get what you want?  
I judge nothing with these questions. I think that they're important to consider and that answers are going to vary widely, that's all. 
So I hope this clarifies a few things I didn't already get to in the analysis of the first GNS essay:

What "maintenance simulation" is.

What "metagame" means in GNS.

Why GNS is obsessed with it.

Some things that obsession makes GNS obscure or ignore.

In reading the last two GNS essays--on gamism and narrativism--I'll do my best to focus only on stuff we haven't gone over yet, while still making sure what they intend to mean and how they fit in here is as clear as possible. If we're lucky, we'll have laid enough groundwork for this to go way faster.