You can find the paper here: http://csjarchive.cogsci.rpi.edu/Proceedings/2011/papers/0138/index.html
This was a paper that I co-authored with Joshua Lewis and David Kirsh that was presented at the Cogsci Conference 2011 in Washington. The paper develops a methodology for analyzing the game Starcraft: Brood War and uses that methodology to study game mechanics and their relation to winning and losing in a much more quantitative sense.
The quantitative nature of the methodology was something that I wished we had time to elaborate on, considering that the nature of studies done using video games often treats the games as a kind of black box through which the scientists push subjects through. There is most definitely room for in-depth analyses of game mechanics when it comes to thinking about the implications that video games have on people during scientific studies. The problem with many studies done using video games is that they treat the game itself as the measurable unit of analysis and not the game’s mechanics (and interactions between player and mechanics as a result of play). This is why I feel that our work with this methodology was important: the game mechanics were primarily responsible for the results we found, and we were able to back this up with quantifiable data.