The Limits of Formalism, my talk from BIRS

I spent last week up in the mountains around Banff, Canada, with a spectacular group of people, as we talked about “computational modeling of games.” This was a workshop held at the Banff International Research Station, or BIRS, and organized by Andy Nealen of NYU and Michael Mateas of UC Santa Cruz. As you may be able to tell from the title, it was moderately mathy, though I was assured by several of those there that it was dramatically less so than an actual mathematician would expect, and certainly less so than the other workshops in the series!

I was asked to give a “seed talk” on the question of “the limits of formalism.” The questions Michael and Andy asked me to answer were

What leverage does a formalist approach to game design give you? What might it leave out? What are the broad contours of the landscape of formalist game design theories? What intellectual commitments are formalist game designers making? What are the biggest holes in our current understanding of formalist approaches to game design?

A page for the talk can be found here, and it includes the slides, a video, and a downloadable PDF of the slides.

The big point I was trying to get across can be boiled down to the word “consilience.” To me, the limit of formalism is obvious, and it’s the same damn thing we have been fighting over since lomg before I wrote “Two cultures and games.” As we approach the increasingly larger complex knot that is “games” we must become at minimum more tolerant of the varying disciplines that go into them, and ideally become more cross-disciplinary ourselves. Fissures have appeared in what was once a small community; these stress fractures are perhaps inevitable as the scale of everything grows, but are nonetheless worrying to me. Gaps in communication between AAA and indie, computational modelers and game grammarians, formalists and game studies folk, game critics and gamers, and many more, speak to an increasing Balkanization of our understanding of game.

Now, the group there in Banff was not only extremely open to alternate viewpoints, but openly enthusiastic about them, and already exploring lots of interesting cross-disciplinary directions themselves. So to an extent I was building a strawman. But a common experience is to find oneself touching on a field about which one knows little; how does one ramp up on that, when there’s not a layman’s introduction or simplified version easily available — or even when there might be, but nobody from outside knows about it? I felt this keenly as they tossed around math and modeling terms with which I was not familiar, and I felt it for them as they explored modeling wildly diverse areas of human experience.

Basically, all sorts of disciplines have their formalisms, and to really cover “games” you kind of need to learn all of them; but if you try, you may never get to go deep on anything, or you may simply never finish.

Even the way in which we selected problems to address in workshops suffered from this; people stood up and put extremely varied problem topics on the board, ranging from how to build game rule sets that encouraged players to tell stories about the game after play, to generating meditation gardens. We then signed up to varying groups based on what sounded like a good workshop. But of course, we then tended to select the topics for which we had some affinity, which unless you held to rigid discipline, could very well work against your joining a group that could really have needed your insight. I know I avoided groups which had problems that I felt I had spent a lot of time on, in favor of groups that had something fresh and different; to what degree does that leave that given group without the benefit of past work? Similarly, it would have been nice to have more narrative folks in the workshops I was in that dealt with more mathy topics.

All in all, the workshop was a tremendous success; I don’t want to portray this larger problem as having had a negative effect. In fact, quite the opposite, as by the end of the week it was clear that there was quite a lot of very interesting cross-pollination going on. Among the developments that were most interesting to me:

• Machinations and derivations thereof seem to have had strong traction in this particular crowd, which is interesting because my sense is that more commercially oriented parts of the design community haven’t picked it up as a tool that much. Instead, I see a lot of “napkin languages” and use one myself. Further, the napkin languages seem to be very verb-centric and also more amenable to visualizing higher-order semantics, tactics, and heuristics — in other words, using simple game grammar diagrams it’s much easier to show things like “attack” and “defend” whereas the more formal models are down at granular actions like “take card.” Machinations is far superior at modeling economies; grammar diagrams are better at conveying play in the way a player thinks about a game. There are big things these two approaches could learn from one another. We did one workshop specifically on fruitful communication between designers and academics, and video of our brief five minute summary can be seen here.
• It strikes me that many in the academic world are very very interested in procedural generation of all sorts of things, to a degree that both the AAA and the narrative indie community are not, given their emphasis on control of the player’s experience. There is a lot for the two groups to learn from each other there as well. We had several interesting workshops on generating things that fall on the experiential side of things. Among them were the aforementioned meditation gardens and another on generating environmental storytelling spaces, a procedurally generated karaoke game, and Vi Hart’s boobscape.io (output of a workshop on PCG and feminism)…

I need to mull over some of the other things more before being able to blog coherently about them. But if any of this is vaguely interesting, I definitely urge you to check out the complete set of videos, photos, and notes available at the website BIRS provided for the workshop. And also track down the work done by all of the individual people who were there! I know I will be doing so.