Okay, this is a call to point to old, relevant threads (if any) and for new discussion.

Before I get started, a disclaimer: I seem to have the super-power of accidentally pissing people off through my casual mode of speech. I'd like to encourage a "no harm, no foul" mode of thought in this thread: I'm not attacking anyone or dismissing anything anyone has said, I'm just trying to work things out in my head. I apologize in advance for any offense I'm giving, particularly to Ron, as I quote him a lot here.

Basically, in light of several of my assumptions being turned on my head by the Simulationist essay, I've re-read Ron's original GNS essay, with a very practical eye: What good is GNS theory? How is it useful in, say, designing my own game, Unsung? The latter because, frankly, I'm a selfish bastard and I'm mainly concerned about how GNS can benefit me and my own play. ;-D

Now, several people have taken pains to point out that GNS is about actual play, not design, per se. However, Ron himself highlights, in his essay, two uses for GNS theory. One is to identify GNS conficts in a group, with en eye to preventing disfunctional play. I think GNS works admirably in this sense, and has allowed me to improve the play of my current group, even if we've never discussed GNS in detail.

However, in the original GNS essay, Ron compares GNS, when it comes to game design, to building a bridge. The assertion is one can use GNS to build a better game, like one uses physics to build a better bridge. This assertion is echoed in the "System Does Matter" essay, where it says one can use GNS to determine a good or bad design.

But if the only way to determine if a design succeeds or fails through actual play, what difference does GNS make in this process? To extend the bridge metaphor: while the ultimate test is walking accross the bridge to see if it falls, knowledge of the laws of physics gives me a sense, before we even build the bridge, of whether it will fall or not. How is this true with GNS?

I'm not asking for an equation I can plug a game into to see if it's Narrativist or not, but certainly there must be some sort of criterion one is using to judge if a game is likely to support, say, Narrativist goals, to determine if a game is "incoherent" or "abashed", without having to walk accross the bridge. Otherwise, what good is it? Why not just throw something together, playtest, and tweak?

I was an English major as well as a CS major, and the reasion I didn't go into graduate studies of English was because I didn't feel that knowledge of literary theory enhanced my ability to, say, write poems -- which was my primary interest. In fact, it seemed to do the opposite -- it got in the way. Now, I'm not saying GNS is like that, but I'm starting to get that vibe. This is what I mean by "crisis of faith".

To attempt to illustrate my point further: Take it as given that John Wick had no interest in GNS when writing Orkworld, but instead he "utilized and developed principles of Narrativism, metagame mechanics, and focused Premise on Character and Situation, precisely as outlined in the theory, but did not articulate them overtly." Given this, how does GNS theory aid me in not having to do as John Wick did? That is, how am I benefitted by knowing GNS in such a way that I don't have to have his background to acquire his level of coherence? How do I recognize an incoherent design?

When I wrote Rise Again, I realized that I was leaning in a strong Simulationist direction, but I wanted a more Narrativist tint to it. At least, that's what I thought at the time. Through modification, I ended up with Unsung, by stripping away the Setting and Color and concentrating on what I thought was a more Narrativist priority. I was aiming at a functional Narrativist/Simulatinist hybid that one could cause to lean more or less in either direction through the use of optional rules. But if I have to walk accross the bridge to test it, how is this different from the pre-GNS days of RPGs? Sure, I understand that some things will need to be thrashed out through playtest, but how does one know that, even in its rough form, Mongrel is a Simulationist design? And what good does knowing that do me, as a game designer?

Kirt,

Here's my take on it.

Every game is designed to do something. Let's call that ineffable something X. Now, when people play this game, it should - under optimal circumstances - enable them to do X, and perhaps even push them towards doing X.

GNS theory isn't necessary to design your game to do X. If you have a clear goal of what X is, and craft your game to do that, you should reach your goal. However, GNS theory is a damn useful tool to help you determine what X is and how to get there.

You brought up Orkworld, which is a good example. It did its X very well, where X = "play noble misunderstood savages in a world that's very harsh towards them, and determine what noble really means." It did this for many reasons, in my opinion:
a) John Wick knew exactly what he wanted when writing it.
b) Thomas Denmark provided him with artwork that was on the spot for focusing that.
c) John Wick is one of the best damn game designers, period. (This is where my opinion runs strong. The more I look back at my time playing 7th Sea, the more I agree with it.)

Now, if you don't know exactly what that X you're aiming for is, or if you don't have Wick's talent, it might help to have some tools, just like a study of symbolism and literary criticism isn't necessary to write a good novel, but can't hurt. GNS theory is one of those tools, proven to be a big help.

Did that help you at all? To distill my point, "GNS theory isn't necessary to write a focused game, but it sure can help."

Practical example of design application:

Note this is going to be stilted because I'm just trying to get the point across.

Let's say I'm designing a game, and my goal is for it to be all about having players explore what it's like to colonize the moon, and what the mindset of such explorers would be, and their interactions. So, I get done with chargen which seems to address making characters appropriate to the setting. I decide that "balancing" characters is unimportant, as the system will make appropriate sorts of characters, and that's what matters in this game.

Thinking I'm most of the way home, I sit down to think about a resolution and reward system. Looking at previous designs I decide that it's best to have a resolution system that focuses on using character effectiveness to overcome specific tasks (pretty standard), and for a reward system I award experience points for foes "defeated" and conflcits overcome, and allow these poits to be spent on making the character more effective.

Well, what do I have? The chargen is Sim, and the rest is Gamist, and the combination is going to be horribly incoherent. The players seeing that the resolution and reward systems are all about success and advancement will probably take advantage of the "unbalanced" chargen and make silly powerful characters. Or, if it prevents that, the players with the less powerful characters are likely to feel like they are being shafted. If the players are very Sim in preference, they will wonder at the unreality of how the reward system changes their characters, and wonder why the system doesn't really do more to support their exploration of the situation.

If you get both sorts of players all hell breaks loose.

OK, as I said, obvious example. But avoiding incoherent designs is the first and most simple use of GNS. Now, I know that you're a better designer than to allow that sort of crud into your game, even if you didn't know about GNS. But GNS allows you to look for more subtle differences, and to consider things like creating functional hybrid designs, etc.

Mike

Clinton R. Nixon wrote: Did that help you at all? To distill my point, "GNS theory isn't necessary to write a focused game, but it sure can help."

See, here's where I think the three focused essays are going to come in really handy.

Let's say, to take the already-present example, that you want a game about exploring the moon, and your focus is (to make a long story short) Sim. Now you look at Ron's big Sim essay, and you think about the different kinds of systems there are, and what their pitfalls have been. And you think, "Hey, I'm going to do a system just like that one there, but I know to watch out for that other thing there."

So you go and read the games in question, and you think about how their systems work. And you build your own system, numbers and whatnot, based on a really similar structure. But when you get to that big dangerous thing Ron's warned you about, you don't do it the way your models did --- you fix it.

Everybody cribs off everybody. Along with Ron, I'd like to express the hope that everybody will start being much more explicit about this, i.e. giving references (e.g. "Thanks to XXX game by YYY, who presented this mechanic in a really helpful way for me."). But when you crib, it really helps to have some distance for thinking about what you're getting. If you want to design a Sim game, you'll get more bang for your cribbing buck if you work from other Sim games. To put it differently, you'll be wary of saying, "Hey, here's this fantastic mechanic I'm going to crib" if you understand that the game it comes from is a Nar game. Furthermore, you'll have an easier time noticing and avoiding pitfalls if you know how the models have tended to slide into incoherence.

Does that help?

Okay, I cross-posted with Mike, so to reply to him and Clehrich:

Yeah, I think I'm starting to get it. I think my problem is that I'm worried that I've produced an "abashed" design, but I think that's just a function of the fact I was working on a hybrid. I guess, in the end, I'll have to walk over the bridge some more. I guess started to feel like there was something I was missing here...

Now, I know that you're a better designer than to allow that sort of crud into your game, even if you didn't know about GNS

Remember, nothing wrong with either Hybrid or Abashed. If you think you've got something that works then just look at how you can leverage the place you've landed at. For example, if you've landed at Abashed, consider making optional rules that empower players to Transition to the desired mode more easily and effectively. IOW, predict what work will need to be done, and have it already done.

BTW, what particularly makes you think that your game is Abashed or Hybrid?

Mike

xiombarg wrote: But if I have to walk accross the bridge to test it, how is this different from the pre-GNS days of RPGs? Sure, I understand that some things will need to be thrashed out through playtest, but how does one know that, even in its rough form, Mongrel is a Simulationist design? And what good does knowing that do me, as a game designer?

Mike Holmes wrote: BTW, what particularly makes you think that your game is Abashed or Hybrid?

Cassidy wrote: Firstly GNS has helped me recognise the priorities and goals that the individual players in our group have.

Hey,

Kirt, first, these are an excellent series of threads and I really appreciate your efforts in them.

My comment for this one is that even physics, the most hard-core applied and multiply-tested engineering physics, still has to include the "walk across the bridge" phase of the development process. That hasn't left the picture, and I'm comfortable with the idea that it won't, no matter what.

Given that, that we aren't going to lose or eliminate the need for that phase, maybe all GNS can do for designers is to help them to avoid (a) trying to please everyone at once (and in fact concentrating on pleasing themselves in aesthetic terms), and (b) thinking that they're creating a One True Way to play. Even if it's nothing else, those two things would sure be nice to see in greater proportion.

By the way, are we using "GNS" to indicate just that one single level of my multi-box model? Or as short-hand for the Exploration box, containing the GNS box, and then all the System box, and then within that the play and Stance box?

Best,
Ron

Kirt,

I don't think that GNS is quite as Zen as your posts make it seem. :)

GNS is practical. If you want to know what it's good for and how it can be used, just look around you. Example: I designed Cornerstone using GNS theory. It's not hidden. The guts of the process are right there in two threads for everyone to read.

In spite of Ron always saying that GNS is only about decisions and instances of play, that's not *really* what it's about. Decisions and instances of play are just the starting place. The title of the first essay - "system does matter" is a clue. GNS is really about interfacing. How does the system affect play? How will the players use the system?

Systems are not mystical spoons. They're concrete; they exist as entities; they do things. They're like computer programs. You can analyse a system in GNS terms by looking at it and imagining how it would work. You can design from GNS by visualizing how you want the game to work.

GNS helps you to identify and classify game elements; it helps you construct elements that support your goals, and it helps you prune off elements that *don't* support your goals.

I read the entire thread; as I was drawing to the bottom, I was thinking that no one took the answer in the direction I would have gone--and then Paganini got close, maybe hit it. But since I'm not sure, I'm going to post anyway.

GNS is about the kinds of decisions players make during play. That's the obvious part, and the part we always try to bring to the fore. However, in system design, GNS is about the mechanics that encourage or discourage particular kinds of decision-making during play.

Understanding GNS is very much about recognizing what kinds of decisions people make. Gamist decisions are generally about improving characters, overcoming obstacles, and winning the game. Narrativist decisions are generally about exploring moral, ethical, and personal themes and ideas, and building to a story climax and denoument. Simulationist decisions are often hardest to recognize, as they can mimic the others to some degree (for example, in a certain kind of simulationist play you can have characters who are striving to become the best they can be, and in another kind of simulationist play you can be running characters who are wrestling with personal, ethical, or moral issues). However, decisions are generally based on what is right for the character, setting, and/or situation, and not on what improves the chance of winning or makes a better story. In the ultimate example, if you gave the player the power to determine whether his character would die of a particularly brutal wound, the gamist decision would usually be that the character survives, the narrativist might decide either way based on whether it furthers the story toward its climax, and the simulationist would evaluate whether the wound seemed serious enough to be fatal.

Hopefully you've already got that part; consider that focus.

What matters in game design is, what will tend to encourage the players to make the kinds of decisions I want them to make, and what will tend to encourage them to make decisions contrary to what I want?

I have elsewhere observed that the D&D reward system encourages gamism in two ways, and that most attempts to fix it fail to see this. That is, the character gathers experience points for killing monsters and gathering loot; that's half of it. The other half is that when he has killed enough monsters and gathered enough loot, he becomes better at killing monsters and gathering loot--the second way in which the reward system encourages gamist play. If that's not clear, think of the standard fix. We will change the system so that you get points for good roleplaying that advances the story. Now whenever you roleplay well, you are encouraged by a reward of being able to kill more monsters and gather more loot.

Once you see this, you approach your mechanics, individually and collectively, and ask yourself, what kind of decision making does this encourage? For a simple example, a game in which characters are usually more powerful than opponents will tend to encourage courageous confrontational play, as the players will believe that they can take on the world and usually be proved correct. A different mechanic in which combats are going to be bloody and injurious, with a good chance the characters will not survive and a better chance that they will suffer aftereffects for some time to come will encourage reticent cowardly play, as the players try to keep their characters away from danger. A game (like Feng Shui?) which rewards creative over-the-top descriptions of action with greater success will encourage characters to become more and more flamboyant in such descriptions. A game which reduces all attacks to "roll the dice to see if you hit" will at least require effort to produce similar results, and one which says, "I'm sorry, that flying leap spin flip kick to the face is very difficult and will cost you a 20% penalty on your chance to hit" is going to very quickly let players know that "I hit him" is the correct statement (unless it is balanced by significantly better damage on success).

Thus as you look at each mechanic, you ask yourself what it encourages and what it discourages in terms of player decisions. If it discourages them from making the kinds of decisions you want them to make (e.g., if it signals that a gamist power-building priority is important here when you wanted to tell realistic or compelling stories of coming to maturity), then you try to find a different mechanic that will encourage the kind of play you want.

I hope this helps.

--M. J. Young

Okay, let's take this one person at a time. ;-D


Ron, I certainly agree that you always have to walk the bridge. My concern was GNS's ability, like physics, to allow you to throw out bad bridges from the start. When I started the thread, I was skeptical that GNS let you do that. After everyone's replies, I'm less skeptical, tho I think GNS is less useful for game design than physics is for bridge design, but that's the danger of hewing too close to a metaphore.

Given that, that we aren't going to lose or eliminate the need for that phase, maybe all GNS can do for designers is to help them to avoid (a) trying to please everyone at once (and in fact concentrating on pleasing themselves in aesthetic terms), and (b) thinking that they're creating a One True Way to play. Even if it's nothing else, those two things would sure be nice to see in greater proportion.

Hi Kirt,

I think your question is definitely valid, and if I understand it correctly, it boils down to "How do I take theory and apply it?"

I think the primary point of GNS is using it as a diagnostic tool, that is, you can use it to identify or classify things, and assist in understanding what you have. The primary things that I find useful about it is that you can look at a game that you know fits into one of the categories, decide what it is that makes it that kind of game, and apply the same concept or similar principles to what it is you're trying to create.

For example, if I take a look at making a gamist game, I can pull elements from Magic, Chess, Poker, D&D, or whatever tickles my fancy. Likewise with Sim and Nar type games. I can analyze and decide what it is about a particular game that appeals to me(strategy, resource management, bluffing/social politicking, etc.) and apply it to what I 'm trying to do.

In the case of Unsung, I would be doing a few things to make sure that I'm getting what I want: Digging deep into movies/books/etc. that give me the kind of "feel" for the type of play I want, and picking out what elements really make it happen, and second, I would be playing other games of similar vein to figure out what does and doesn't work and why. It may be a worthwhile thing to take the backstory from a movie or book, run an "Unsung" game as a one shot in a few diffferent systems and see what happens. Also check out pacing and focus from your source material and find out what is crucial to making that sort of thing happen.

Chris

M.J. nailed it, with one addition. You can also use the process in reverse.

GNS attempts to classify *all* potential role-playing decisions. Think of it as a description of a set containing every possible play decision. (This is a lofty generalization, I know, but stick with it.) Once you recognize and understand every type of decision that can possibly be made, you can choose a subset of decisions and design mechanics around them. What mechanic would best encourage *this* particular decision type that I've identified? This is pretty much how Cornerstone worked. Rather than starting with a game, we started with GNS goals and figure out how best to serve them.

I want to ramble about this for a bit if you don't mind. The three big GNS axes (G, N, and S, respectively ;) are really the *end* of the thought process, not the beginig. In order to understand what narrativism is, you have to understand all of sub-compnents that define it.

You don't just say "I want to design a narrativist game!" You look at the subcomponents. You find *definite* goals to choose frome. What premise? How can the mechanics best encourage players to address it? To encourage players, you have to understand about reward systems. To understand reward systems you have to understand currency.

Stance, exploration, currency, and so on are not just interesting ideas loosely associated with GNS. They're the ingredients that GNS is built out of.

Edit: I forgot to mention, Vincent's thread "Egri and Me" and the subthreads thereof is a great example of this kind of thing in action. It's not "how can one make a narrativist game?" It's *very* specific. It's about narrativism, but it's *focused* on addressing premise. Not just on addressing premise, but on a specific way of addressing premise - via a particular kind of character. Mechanics are designed to fit specific goals, not generalizations.

Kirt asked: wrote:
(Quick digression: Similiar issues drove the Gift system in Unsung, but I'm ashamed to admit I'd never looked at Cornerstone until now. Quick question about Cornerstone: Can you activate a trait that's in the attribute being used for the roll?)

I have a few comments here and like Kirt its all No HArm no Foul.

First Kirt, I personally focused on your question of (to paraphrase) "How does it help me know what bridge designs to throw out before hand?" When it comes to Game Design, GNS does not. From a player group perspective I think it works perfectly at getting the bugs of play out before hand. However, in design I think it comes up short.

Why? Quite simply it only focuses on on theoretical play. Till someone, as you and Ron both mentioned, someone crosses the bridge, then you will never know if the game works. What is the ultimate goal of good game design? To create a sound and coherent game and to create a game PEOPLE WILL PLAY. Its not about marketing, getting people to fork over money, but it is about Player Currency ie Time, Money, and Thought.

If you build the ultimate two lane bridge, but it gets built in a country where the people are strange and only like one way bridges, you have failed utterly to satisfy your audience. Consequently, if you build the Ultimate Narrativist game (in GNS definition) and most role players are Sim or Gamist, then you have also failed.

I think GNS can create a coherent game but it cannot accurately predict audience reaction. You could answer every question posted here, follow all guidelines and create a beautiful game no one wants to play.

Perhaps GNS was never meant to answer the question of "What does the RPG audience want to play and how will they React to my game" and it would be unfair to judge it that way if its never meant to do so. I just thionk thats it hard to assume if you make a NArrativist game, NArrativist gamers will play it because, well its a perfect (if such a thing exists) Narrativist game.

My second point is that, with reference to the Forge and the RPG community in general, GNS is pretty much the only Industry Tool available. I have mad more then one critic of the Forum and GNS etc admit in the same breath that most of the serious THINKING is happening right here. I do not think its arrogant to say that nor am I trying to be a sychophant, but one of the many things that drew me here is all the serious thought that goes on. (also for the most part we are a damn friendly crowd, hard to find on the net)

However, other then some other less well thoguht out fragments, as far as I know GNS is the only serious model of gaming and game design available. Like Relativity. DDoes it solve every riddle? No it really spawns more questions then it answers but right now, it works pretty well for what we use it for.

Sean
ADGBoss

GNS is not at all about predicting what an audience will like. It's about creating mechanics that reach design goals. If your goal is to create a Narrativist game, then GNS theory will help you achieve that. And to the extent that nobody likes an incoherent game, GNS will help you avoid that.

Who knows if people want more Narrativist games? GNS certainly won't tell you.

OTOH, given that I and others here (this is Ron's theory from his recent social series of threads) feel that it's more important to make games that function as we want them to function than to try and figure out what people want, this is a good thing. I and others believe that you can't sell games by predicting the RPG market, and the successful game will be the one that reaches interesting goals in a well designed manner.

That is, design a good Narrativist Game, and people will become Narrativists.

Mike

Ok I can see that as a possibility but I think that more then GNS concerns go into game design. Certianly it has to be something I want to play as a gamer or else I will not even begin the design process. I think it also has to have Interest Elements I will call them lay outside strict GNS explanation.

For example: A Narrative game: God of Pineapples
"What is it worth, to you, to grow the greatest pineapples"
No super Powers, No dark conspiracies, no aliens. Just Pineapple growers.

IS this going to lure people into Narrativist play? I don't think it will BUT I could be wrong about that. Afterall as I recently mentioned I am always intrigued Mike, by your Paraplegic Photographers idea, which is certianly not main stream.

So perhaps it is worth the theory that "Buil it and they will Come" can work.

Certianly something to think about.

Sean
ADGBoss

ADGBoss wrote: I think it also has to have Interest Elements I will call them lay outside strict GNS explanation.