If half of being a faculty is teaching, the other half is performing research.
Game Studies research is pretty straightforward in terms of how to approach it; it's similar to other cultural and anthropological studies, where you conduct interviews and observe player reactions to games and analyze demographic tables and such, and then you find the correlations.
Game Design isn't like that. Yes, some researchers conduct studies that try to quantify "fun", in the same way that economists might try to quantify "utility" or psychologists quantifying the exact definition of various disorders.
But look at cinema. Is anyone writing research papers on the crafting of an audience reaction, or the correlation of fun to car chase scenes, or creating emotional tension in the viewer? These kinds of things are artistic, creative and highly subjective; they don't lend themselves well to number-crunching quantitative research.
Instead, the academic side of film creation concentrates on, well, creation. Students make their own experimental films and showcase them.
Should game design research be the same way? Shouldn't we be making experimental games, using mechanics or themes that haven't been used before? If we must, we can always write an academic-style paper about the game we made, that talks about the design intent and observed player reactions.
And yet, I haven't seen anything like that coming from academic researchers. Grad students, yes, but not faculty. Am I barking up the wrong tree here, or is this a direction that game research should be heading in? Or has it already, and I'm missing out on a lot of experimental games out there?
Monday, August 21, 2006
Subscribe to: Post Comments (Atom)
I may be biased, but quantitative papers like this make me very happy, and are incredibly useful from a design perspective.
What about building small games that record what the players do, so you can analyze stuff about player behavior? I'm not talking those projects where they measure people's heart rates while they play a game or whatever, I'm just talking about collecting statistics on gameplay behavior over a wide cross-section of the player population, much like in the paper above.
Research that goes in this direction might be interesting: http://lostgarden.com/2006/01/creating-system-of-game-play-notation.html
Personally, as a grad student in communication, I don't conduct much quantitative research; I prefer qualitative methods like interviews and participant observation with a focus on use and culture. I once tried to take a game design course, but I ended up having to drop it because the 3/4 of the class about theory had already been covered in more depth by readings I had done for other work, and the final bit of class involving game design required background knowledge of programming that I don't have.
Lots of great comments here.
Darius: you're right, I could definitely see things like player behavior being useful. But it begs the question, where is the similar research in the film industry? Wouldn't it be equally useful for a script writer to know what kinds of scenes evoke what kinds of audience responses? Where's the research that "proves" it?
Duncan: The thing I love about that particular blog post on Lost Garden is the intense discussion in the comments that follow. Lots of people arguing over whether such a thing is even possible. Might be interesting to try, to see if gameplay notation can work at all...
Jason: A game DESIGN course consisted of 3/4 theory and 1/4 programming? For shame! Chalk it up to one more example of the game industry's terminology clashing with the terms used in academia. Would you be willing to give the school and course number, or a link to the syllabus or something, so I can see it for myself?
Ian, the film industry's content has doesn't consist of audience response, and they do do test screenings to gauge that. Games are interactive, therefore player behavior dynamics are an order of magnitude more meaningful to us than passive audience response is for film.
I agree we should make games, it seems obvious when you think about it.
Theory should be testable in an interaction model and rapidly iterated on.
Post a Comment