the exposure to doom study is equivocal at best. anyone who follows the heuristics and biases program in social psychology knows that framing effects are well established (in this case exposure to violence), but the trick is to show that the effect is anything more than temporary (studies on framing and racism i've seen produce only a temporary effect), and further, to show whether it is actually a significant /causes/ violence. (the "10 times as likely" in the gentile study could be misleading if the likelihood of getting into a fight without exposure to games is itself low, and also needs comparison to other effects which might be much more significant -- socioeconomic conditions for example -- ill have to have a look at it)

you would think that the genitile study and the doom study proves this in conjunction, but it doesn't unless it controls for all the other relevant variables (a serious problem with most of these studies). and the stories we hear about simple imitation of violence in video games are pretty silly judging by the statistics comparing exposure to violent video games versus actual incidence of the serious violent behavior, the kind we often see in video games.