Areae president Raph Koster is perhaps best known as a designer of Ultima Online and the previous CCO of Sony Online Entertainment, but his new venture capital-funded project, Metaplace, hooks snugly into the 'mainstream online world' angle that is a world away from the hardcore MMOs he formerly developed.

Koster is one of the most notable cheerleaders for a paradigm shift in the game industry away from alleged hardcore-centric insularity, In this extensive interview, conducted at last month's Austin GDC, Gamasutra discusses his ideas for a working "game grammar" for developers and much, much more.

I wanted to talk about your work on game grammar, as discussed in your AGDC lecture and in previous and upcoming books. Have other people done that before? It struck me as very much like a structuralist, or post-structuralist kind of theory. I don't know if you agree.

Raph Koster: It's certainly not developing out of thin air. I think there's a bunch of predecessors. In the late '90s, we had Doug Church talking about "formal mechanics," or something. We had Doug Church back then, and we had what I think of as the "Looking Glass Mafia" of people -- the MDA stuff, which was Marc LeBlanc, and Robin Hunicke, and somebody else's name that I'm blanking on. So it was that kind of thing. There was all of the work that Chris Crawford did -- an awful lot of what I described in the model is premised on his description of how interactivity works. He wrote the accessible book for game designers, and the very dense and really rich book on interactivity.

Didn't [game veteran and Marble Madness creator Mark] Cerny do something, too?

RK: Mark Cerny's thing isn't quite the same. There's lots of stuff. The grammar stuff is just taking it a little bit further. My goal, originally -- it's kind of funny that the grammar stuff has actually gotten interest and attention this year. It was my lowest-rated lecture ever when I gave the original lecture at GDC three years ago. It was actually lowest-rated because it was split. A bunch of people thought it was awesome, and then a bunch of people said, "Whoa, total freaking waste of time!"

I think now they're paying attention to it because you're doing some weird new company thing, and they're like, "Well, we have to listen to him now!"

RK: That certainly helps! Pretty much every other creative field has ways of notating what they do. That was the original question -- can we notate this? That's actually what I wanted -- a notation system. Nobody has been able to come up with a good notation system yet, but the thing about trying to come up with a notation system -- we want it, because god damn do design documents suck as a means of communicating game design.

The thing I always say is that building a game off of a game design document is like trying to film a movie off of the director's commentary. One is not the same as the other at all. What I wanted was a way to notate what was going on, so that we could communicate it effectively. In trying to do that, what you end up at is, "Well, what are the things that we want to write down? What is pitch? What is key? What is tempo?"

The way that you broke it down, I felt like I needed to look at it and analyze it to make sure that it wasn't missing anything.

RK: Oh, I skipped over lots. It can't be that simple.

You said it so succinctly that it was easy to go along with, and so I just wanted to make sure that I didn't really go along. Have you written a book about it yet or anything?

RK: No. It's actually for sale on Amazon now, but it doesn't exist! Isn't that a neat trick?

Hey, that's cool. You're making good money!

RK: (laughs) No, not really.

Selling virtual items, that's what you're doing!

RK: (laughs) Yeah! The publisher's been wanting this book for a while. Part of the reason why I put this talk together in the way I did was because it prodded me to actually organize some of it. I've been writing bits and pieces of game grammar stuff on the blog for a while too. I've had a few knock-down, drag-out fights about whether the word "grammar" was even right. Frankly I don't even care very much. It's there in part just for the alliteration, I'll admit it. The process is kind of hard, actually. It's easy to look at something like Space Invaders and do that breakdown... It's really hard to do it for poker, as an example.

Anything that has an element of luck or risk, really.

RK: Well, it isn't so much luck. This is something I gloss over entirely in the lecture, but it isn't so much luck. Luck is easy. Luck is just a black box that spits out a number.

We do suck at randomizing things, really.

RK: We do. Computers suck at randomizing in one way, and we suck at odds assessment really badly. We're awful at it. It seems to be a human brain issue. It's not something we do well -- just like computers just don't do some things really well, our brain does that really badly.

The thing that's complicated in grammar is the question of assets. Are the pieces in chess -- you look at the array of moves -- are the pieces in chess verbs, or are they assets you manipulate? In a deck of cards, when you have a hand in poker, are you manipulating the cards as tokens? It's weird questions like that that are kind of picky. They lead to questions like, "Are chess problems content, or is each chess problem actually a new game?" It leads to weird nomenclature questions like that. That's actually a thing that was really weird and tricky, especially when you try and diagram it and end up going, "Well, do I have to actually notate every single one of the 52 cards in poker, or what?"