We know that video game violence is certainly correlated with violence — just like smoking is correlated with lung cancer.
Sure. But the hypothesis that violent people gravitate toward violent games is a tad more plausible than the hypothesis that people who are incredibly lung-cancer-prone just happen to gravitate toward inhaling smoke.
We know from many experimental studies that playing violent video games causes less serious forms of aggressive behavior.
Again, true. But there isn’t much we can conclude from this, for two reasons.
First, as Bushman concedes, causing lesser forms of aggressive behavior (say, blowing an air horn to punish another person in a lab experiment) is a far cry from causing murder. And second, even if games do make people feel significantly more aggressive, they also keep young men indoors and usually sober. One study finds that violent-video-game releases are correlated with lower violent crime. It makes no sense to draw conclusions based on the increased aggression without factoring in the other effects.
There isn’t much else to his case. There’s a correlation, and experiments reveal a mild effect, so it’s time to start taking away teenagers’ copies of Grand Theft Auto — and have them run around town, possibly raising hell, instead.