It does seem like in chess, each move is a new game in a way. It changes the whole thing, for one thing.

RK: Right. So I say that all games are iterative. I also say that all games are turn-based. And finally, I say that all games have more than one turn. They're always iterative. There's always a starting state, and then you do shit, and then there's an ending state. And sometimes, the ending state becomes a new starting state. So when you move the pawn in chess, now you've got a new chess layout that you have to think about. But then you have the choice of, "Well, which piece am I going to move?" And each piece has its own topological space around it, because they all move in completely different ways.

The fact that they move in different ways -- that they have different rules -- is that content, or is it mechanics? Are landscapes content? I tend to think that landscapes are content, and therefore, arrangements of chess pieces are content. But then when you go forward, they are verbs, and that means they are actions that you take. It gets weird. That's probably the kind of thing that will make people reading this interview go, "What the fuck? Who cares! This cannot possibly be useful!" (laughs)

It seems like an area in which the game academia should be looking seriously. That's the kind of problems that those people's brains are designed to solve and write long-winded things about.

RK: Yeah. But you know what? When music went through its big formalization period, it was actually musicians who did it. The Well-Tempered Clavier was done by a musician, not by an academic. We usually have tended to see that, but a lot of that stuff actually gets done by practitioners. I've listed off a bunch of practitioners who are doing stuff with it.

But they don't have time to do it all the time, or get together and figure out the one true way.

RK: But we do get together! It's the hidden secret of the industry. We have these cool secret retreats where we go meet.

And talk about grammar?

RK: Sometimes, yeah!

Cool.

RK: I'm not making this up! Project Horseshoe, that got some ink. That's what that was -- that was a game designer retreat. There's several of them, and we have forums where we talk about stuff like this. Game grammar was actually born on a private game designer forum. That's where I put together the first outline of game grammar.

When I was looking at all the stuff you put down in the way it was there, just laid-out and dissected like that, it didn't look like building from those pieces... it was difficult to discern how, by using these pieces, you would create fun.

RK: The pieces did not give me a recipe for fun. They gave me a really good way to see when something wouldn't be fun, which is kind of interesting, right? I'll go back to music notation -- you look at a piece of music notation, and you're not necessarily going to know if it's going to be a great piece of music. You can often tell if it will be a crappy piece of music! I think that's kind of an interesting thing -- a lot of notation systems do that. They show you the absence of stuff more than the presence.

I found it when I was doing it, and then other guys, like when Andrew McLennan did it -- he's the guy at Slam Games, working at ITI Techmedia and Metaforic -- they did the GDC talk last year, where they used a game grammar approach to quantify the difficulty of going through levels of an FPS. The kinds of things they found, was they were able to find shelf events -- places where people would just quit. That kind of thing.

When you put together even a simple diagram -- even one using the kind of level that I have or [Lost Garden blogger and Gamasutra columnist] Dan Cook has [in his 'Chemistry Of Game Design' article], which is pretty high-level -- it's not near as complex as the Stéphane Bura one that I showed. Pretty straightforward. You can see places where we ask the player to solve a problem that is trivial, like "Push a button." Then we ask them to solve another problem that's trivial, then another problem that's trivial, then another problem that's trivial... and pretty soon you have crafting in an MMO.

You can see in the diagram where there's no systems -- it's "UI Action, UI Action, UI Action, UI Action, UI Action." That sticks out as unfun. (laughs) Right? You can see places where nothing branched, for like, forever. You can spot the unfun, but it doesn't tell you if it's fun.

I ended up with a list of criteria that's in A Theory of Fun -- it's in the book, actually -- which is nine questions you can ask. It's stuff like, "Did the state change from last time, and does that matter?" If it didn't change, then what you're making is probably less fun than it ought to be. Chess would suck if after you made a move, the board stayed the same! "Did you have to make a decision, and was skill required in the decision?" If you could roll dice and do as well, it's not as fun. There were like nine things. If you're missing some of the nine, then your game is probably in trouble.