Cooperative and Disruptive Behavior in CounterStrike

http://wwwai.wu-wien.ac.at/~hahsler/research/webBasedComm_cs/webBasedComm_cs.pdf

This paper by Austrian business students Michael Hahsler and Stefan Koch was a disappointment to me. I expected a paper entitled "COOPERATION AND DISRUPTIVE BEHAVIOUR – LEARNING FROM A MULTI-PLAYER INTERNET GAMING COMMUNITY" to be mostly about the community and the players. In fact, the paper had almost nothing about in-game interaction or the CS community. An explanation of CS, including the history, gameplay, graphics engine, and server mods takes up about half of the paper. Most of the rest is an explanation of the statistics the authors logged from the CS server they ran, with only one short paragraph about the actual community.

I don't know much about game theory in relation to math, which was a big hindrance to my understanding of the paper. The authors put heavy emphasis on the analysis of what they deemed cooperative behavior, based on relative proximity of players in-game. This very notion is incorrect - CS is a tactical game where sticking together may not always be the best course of action. For example, the Counter-Terrorist team may have to defend two bomb sites simultaneously. In order to do this they may have two people cover one entrance to bomb site A, one cover the other entrance to bomb site A, one cover site B, and a fifth cover an important choke point with a sniper rifle (this is an actual strategy used on one map: de_dust2). According to the authors' model of cooperation maybe two of those players would be considered to be cooperating, and maybe none at all, when in reality this strategy requires quite a bit of cooperation to pull off successfully. In fact, almost every single round of CS requires a lot of cooperation. Their measure of disruptive behavior was also limited to shooting or killing teammates. There are plenty of other ways to disrupt a game, such as sitting in the same spot for long periods of time (especially if the spot is out of the way or hidden), or excessive "flaming" or verbal abuse. Overall, this paper was not very comprehensive, and seemed to focus more on specific (and therefore ultimately useless) numbers rather than on community and interaction between players and the patterns that develop out of those.