It's been a while since I did any planning for the conference. I've spent more time in the last week staring at maps of the coast than I have thinking about the actual purpose of the trip. So I'm getting back down to it. Last year was the first time for the poster sessions at the Game Developers Conference. It sounds like a good way to pack more content in to the time available. They must have no shortage of ready speakers and poster sessions are a good way to accommodate them. I know I could come up with something to present.
The way the posters work is the presenter puts together the information on their subject and might have a talk or a paper ready to go. They don't get a room to talk in or overhead slides though. They get an area I'd guess about 4'x9' on which they can lay out the diagrams and information that go along with the presentation. They also have a one hour time slot to give their talk that goes along with the ever present poster.
The great thing about a poster is that it's always out there. People are going between sessions (the non-posters, that is) and might have a few minutes to read or they could be wandering in or out of the expo and they get drawn in. So the poster is up there with the basics of the presentation and the viewers know that they can come back at a certain time (I think it's posted) and talk to you in person. So there is a period of time when the poster presenters have to be on the spot and get to give their talk. I suppose people might be more likely to wander away than they would if they had seats, but that goes both ways: people are also likely to wander over to see what the crowd is looking at.
There are eight sessions listed
in total now. Most are on the Game Design or Programming tracks. The two outliers are both being delivered by Byung-ho Park, who worked as a game columnist in Korea for over 20 years. He's lectured at GDC
before and this year his two posters are Online Gaming Business in the Asian Market: Understanding the New Business Model
and You Deny, But You're Lovin’ It: Using Psychophysiology for Product Evaluation
His poster on psychophysiological responses to games promises to be one of those topics that makes you queasy but you can't help agreeing with. Monitoring people instead of relying on their reports sounds like a lab test, but it removes a lot of human factors that cloud marketing statistics and studies. The other presentation he'll give, on the Asian game market sounds obvious given his writing background. Although this presentation is targeted at companies entering that market, it sounds like something anyone in the industry could benefit stopping by to have a short read.
Poster Sessions on Game Design
Christian Baekkelund will present what sounds like an overview of the different kinds of AI
and how they can be applied in games. I'd like to have a look at this one. I know that the path planning concepts I dug up when looking into Graphplan
are one of many approaches. Then there are numerous other machine learning and AI
topics that span many fields including game design. I used to always think neural networks were the core of AI
because that's what we covered in university. Neural networks are just one huge field of many when it comes to AI
. I'm sure Christian will give us a good survey of the buzzwords and the methods that touch on games today.
This one covers a military application of games and game based training. It sounds like a definite Serious Games application to me. Curtis Conkey has covered related topics
at the Serious Games Summit
last fall. It sounds like figuring out what the player learned during a simulation session is hard to determine. I can relate - web based training sessions given to employees in the private sector often seem ineffectual because they rely on an exit quiz that can be duped. I hope the Navy has better solutions to this problem for their training. Check out the poster and find out.
Carlos Martin will be disaggregating and recontextualizing in this poster presentation. In public no less. For a whole hour. Just goes to show that wild "anything goes" attitude at the GDC
. Seriously, the summary says he'll be trying to separate the components that usually just get lumped under the heading fun
for interactive games. It's can be hard to explain that making a fun game that might feel like a mindless diversion can take a whole lot of thought and planning. This poster session sounds like an introduction to that kind of planning.
I definitely have my own point of view (actually several) on this subject, and I'm sure a lot of other developers do to. I like the idea that Dorothy Warner (the presenter) includes game developers in the research, it will be informative to see what the general feelings are. I just hope it doesn't turn into a brawl. This is a pretty emotionally charged issue, but I think that developers are generally a rational bunch. Software people have to be objective and can usually discuss a problem without getting drawn in too deeply. Of course a poster presentation on violence in games can really go anywhere, depending heavily on definitions chosen for violence, methodology of the study, and any preconceived notions that the researchers might have.
Poster Presentations on Programming
These last two poster sessions are on the Programming track. I've left programming for last because that's where a lot of my background and interests lie. If I did the programming ones first I'd wear myself out and never get to any others.
Alan Kimball will present a poster on working with the potentially visible set (PVS
) using the dedicated processing power of the graphics processor (GPU
) instead of the general purpose hardware on a computer's CPU. The idea of using the GPU
for heavy-duty number crunching that doesn't result in colouring dots on the screen has been around at least since the advent of programmable shaders and gets more worthwhile as GPU
pipelines get deeper. The interesting part about applying this to the PVS
is that the PVS
is a precomputed set of data for the game player - it's not their GPU
he's talking about. GRAPE works at build time and utilizes the GPU
to speed up building PVS
data as part of the overall software build process. Really cool concept. I wonder how portable GRAPE is - if it can be extended beyond PVS
data and into other types of computation.
I haven't done any Direct3D or HLSL
in a couple years now, so this one sounds very advanced to me. Nonetheless, water effects are pretty. There are concepts in 3D that keep getting repeated and refined. It sounds like there's some very detailed refinement that Francesco Carucci will describe in this poster presentation. Making water realistic is one thing, but expressing emotion is a mastery of that art.
So it looks like there's a good crop of poster presentations to see this year at the Game Developers Conference. Take your time to stroll around and talk with the presenters about their work. The setting of the posters really makes the speakers more accessible to the audience than in the more traditional rooms.
While posters are accessible to all comers, the traditional sessions are more interesting with a little prep work. Jeff
and I have done some reading to get ready and shared what we've gathered in the blogs over the last few weeks. Have a look through our unofficial preview of GDC 2006 sessions
to get a feel for what to expect, along with pointers into the official site