March 30, 2004
Once again GDC didn’t let me down. Here are some highlights; pictures will be added over time as they filter in.
First I’ll talk about our presentations, get that out the way, and then talk about the rest of the conference itself.
The 3-day Independent Games Festival, which included a playable demo of Facade, was certainly the high point for me and Michael. We must have had a hundred or more people play it, usually with a crowd of people watching (see pics). (The festival also served as a big user test for us, ferreting out all sorts of bugs.) Noah and William Huber helped monitor the demo from time to time, which was a great help. One my favorite things was seeing Noah watch people play it over and over, and continue to laugh and be surprised as he experienced the range of variation in the dialog.
However, it was sometimes painful to watch people play, since the system is still buggy and untuned in places. I’d guess that at least 50% of the time, the system didn’t react the way we intended it to. But the wonderful thing was, most players were very forgiving. Even when the system clearly screwed up, they kept playing and tried to get stuff to happen. Most people if asked would say the system didn’t seem to be responding well, but that they enjoyed it anyway, or at least found it “interesting”. In a recent comment Sean Barrett said he felt like he was “utterly failing as a participant–flailing badly, not contributing”, blaming himself. On the contrary, I think the system was at fault, not being robust and tuned enough yet to handle your input. Although, it’s true there is a lot of content in the system that will require a pretty active user to discover. One guy who had experience in improv acting (from the CMU ETC group) accomplished a hilarious performance. He seemed to know when to speak to cause maximum response, to accept offers, etc. Perhaps Facade is more of an instrumental text than we realized. … Here’s a list of links so far of others writing about their experience playing Facade (hopefully will add to this over time): Walter Kim, William Huber, Robin Hunicke, Rob Zubek, Calgary’s FFWD, Adventure Gamers article and forum, Notes of Chaos, Celephais, kellee’s blog, auto-translated pages from Czechoslovakia and Japan. (Also, here’s an article I just found from last year’s GDC presentation about Facade, that I gave solo.)
Could it be, Facade crashed?
Sadly we had little chance to play other people’s projects, but I did walk around and look at a few of them. Some cool stuff; Bontago definitely caught my eye.
We didn’t win any awards, which was disappointing, but not the end of the world :-) — see my comment in the previous post. Besides, Gonzalo did a good job channeling our pain.
I’ll briefly describe the programming talk we gave, “Beyond State Machines: Managing Complex, Intermixing Behavior Hierarchies”. I don’t know if it was because people wanted to hear how Facade worked or what, but the room was packed! They were standing in the aisles. (Last year we gave a game design talk, which I did solo because Michael wasn’t able to make it at the last minute; so this year Michael did most of the talk.) The talk gave an introduction to behavior-based programming, which differs in significant ways from traditional imperative languages like C++ and Java, namely that you can easily create parallel behaviors, use context conditions and success tests, behavior conflicts, etc. To help explain the concepts, we showed a series of 6 video clips that started off simple and built up in complexity over time. Several people told us that it was a very effective way of demonstrating the power of this approach. Noah said he heard a pro-state-machine programmer come out of the talk somewhat impressed. (We’ll post our slides and sample videos shortly on interactivestory.net.) Hopefully the talk will spark some interest in some developers wanting to get a hold of ABL (the language implemented for Facade) and/or the related Zoesis technology. Also it’s possible to order the video of our or any other talk in case you missed it and really want to see it, and of course our papers are online. Update: Here’s a bit of reaction to the talk on GameDev.net)
Now on to the rest of the conference. Just like last year, this year’s Experimental Gameplay Workshop was awesome. The first half was a frenetic display of the results of the Indie Game Jam (see Robin Hunicke’s conference pics and pre-conference jam pics), which used Atman Binstock’s very cool 2D physics simulator as its platform. I believe all the games will be downloadable soon, so I won’t go into detail here to describe them, but I will say that I totally love the spirit of this Jam. It’s thrilling to witness a group of talented artist-programmers who really know the medium, improvising some surprisingly powerful, poetic stuff. The almost pure proceduralness of the games was the key; this allowed for some true generativity and emergence. It’s this kind of proceduralness that I and others want to strive for in our own work. (It’s a bit easier to be purely procedural when it comes to physics than character behavior, though…) In extreme contrast to the huge budget and development teams and conservative mindset you normally experience in cut-scene-heavy game development, the Jam games felt very cutting-edge and self-empowering. It inspired several of us think about creating a Jam (or Jams) of our own… more on that in a future post… (update: here it is)
The second half of the EGW was more tame but very interesting. Ken Perlin presented his and Mary Flanagan’s Rapunsel ambitious project on procedural literacy, which has made lots of progress since we first heard about it at last year’s Academic Summit. I think this is one of the most interesting projects currently out there. … A Japanese developer, via translator, demoed a bizarre game in which you roll a sticky ball around an environment (kitchen table, house, a city), snagging every object you touch, turning your ball over time into a ridiculous, gorgeous mess. … Chaim Gingold presented a brief analysis of Wario Ware, a very cool must-see experimental GameBoy game, and Robin Hunicke briefly discussed how several new games have used time in various interesting ways. I liked having two mini-talks mixed in with the EGW (but those less academic-minded in the audience may have felt impatient). … My colleague Peter Weyrauch at Zoesis presented The Demon and the Princess, a new demo I helped create in which the player can engage in physical and conversation interaction with a lifelike intelligent monster character. It seemed to go over very well with the audience, and it didn’t crash, thank God! (We finished the build hours before boarding our plane to GDC, and it still had lingering crash bugs.)
I only had time to see a few talks this year. I barely managed to squeeze into the room to see the Love Story Challenge panel, in which Raph Koster, Will Wright and Warren Spector were each given the challenge to come up with a design for a love story game (go here for a detailed description). Raph Koster’s idea was big and grandiose (not too surprising I suppose, given his bent to create large sprawling worlds), and therefore seemed to me a bit too unfeasible. Will Wright’s idea had a brilliant elegance — he placed his simple love story game inside an already existing war game, Battlefield 1942. And, he left most of the details of the love story to the players themselves. In doing so, his idea seemed very feasible, although he really backed off from the hard problem of creating intelligent NPC’s. Warren Spector didn’t present an idea, because he had to admit he had little idea of how to successfully design such a thing. Instead he talked about the issues and challenges involved, and even briefly mentioned Facade as an example of a project pushing in this direction. (Later we found out he had spent 45 minutes watching others playing it!)
Warren Spector later gave a talk called “What would Aristotle Do?” that we had to miss (it was scheduled opposite the EGW). But Warren jokingly warned Michael away from it because it borrowed from Michael’s neo-Aristotelian theory work ;-)
Game studies in progress
Peter Molyneux gave two talks, both in huge packed rooms, one that I watched and one I walked out of. He’s a good speaker, but tends to talk at a very high level, as if the room is full of newbies. I wish he’d go into more detail. The first talk was about applying AI to game design, but we didn’t learn much other than seeing footage of Lionhead’s upcoming game Fable. It looked cool, but less interesting than the descriptions I read about it previously. Supposedly it has hundreds of thousands of lines of dialog in it though. His second talk was very general, about game design or something, and I wasn’t learning anything so I left.
Michael and I unfortunately had to prepare for our talk when Will Wright was giving his talk, “A Schizophrenic Approach to Game Design“. I heard it was good. Neil Young’s game design keynote on “Lord of the Rings” was too full to get into. And in order to man our IGF booth, we also missed Katherine Isbister’s “10 Tricks from Psychology for Making Better Characters”, “Would the Real Emergent Gameplay Please Stand Up?”, Ernest Adams’ “The Philosophical Roots of Computer Game Design“, the panel “Towards Relevant Research: Collaboration 101” with Robin, Raph and Will, and Gonzalo’s roundtable “Practical Game Theory: Academics Fragging Developers”. Ugh, now I’m depressed…
Gonzalo and Andrew
We did make it to Robert Niedeffer and Celia Pearce’s ArtModJam, but were already familiar with most of the pieces there; I was hoping to see stuff I hadn’t heard of before. I found the video game art presentation at LevelUp had more surprises and a broader survey. But I think the panel would have been pretty valuable to game developers who hadn’t seen these projects before, so I’m glad they did it.
But overall I really loved seeing all the people, finding out what they’re up to and celebrating their latest achievements. Katherine Isbister is starting at RPI this fall.Mike van Lent’s first issue of The Journal of Game Development is out (but I forgot to snag a free copy!) The warm Adam Chapman is still in NY, makin’ aht.Chris Crawford’s working on a new book on Interactive Storytelling, perhaps his best book yet. Noah’s First Person sold out at the GDC bookstore. Patrick Doyle finally finished his PhD, and we surprised him by crashing his celebration party in Palo Alto. Magy Seif El-Nasr is now faculty at Penn State. Ian Wilson is back on the scene with the latest version of his Emotion AI (sorry I didn’t get a chance to see the demo!) I briefly reunited with a good friend I used to work with at PFMagic, Keith Kirby, now a mucky-muck at Mattel. Margaret Wallace’s (another ex-PFM’er) Skunk Studios is going strong. The magnetic Jane Pinckard seemed happy, can’t imagine why. The gang of IDT students were having a fun time. Brenda Harger and her colleagues at CMU’s ETC threw a great party. And Mirjam Eladhari from Sweden managed to make it to GDC only a day late, after her travel agent mistakenly sent her to San Jose, Costa Rica! And I had the pleasure of meeting some in person for the first time: Walter Kim, Dave Thomas, Drew Davidson, Marek Bronstring, Chris Allen, Greg Johnson, Barry Silverman, Michael Sweet, and Gloomy.
Noah describing his text-in-a-cave piece, Screen
The days following GDC I spent in Berkeley and SF with friends Chaim, William, Walter, Gonzalo, Michael and Noah. I used to live there, and so it was fun to return to my favorite spots — Fatapple’s, the Cheese Board, Black Oak, Cody’s, Vik’s (the best!), the Haight, Lulu’s, and Chez Panisse. We also saw the Bang the Machine show at Yerba Buena; it wasn’t terribly substantial, but good to see nonetheless.
Chaim’s humble abode in Berkeley
Feasting at Vik’s Indian Chaat
Reviving those Ms. Pac-Man skillz — like falling off a bike
Phew. That’s it. Can’t wait for next year’s GDC… I hear it’s moving to the Moscone center, in south of Market SF!
Contemplating the future of games on a hilltop in San Francisco