In science, correlation doesn't imply causation. Changes in brain activity, for instance, don't necessarily lead to changes in behavior. Even an obvious relationship between virtual aggression and real-life aggression, like acting out the specific behaviors portrayed in "Grand Theft Auto," isn't necessarily one of cause and effect. It may be that real-life violent psychopaths enjoy being virtual violent psychopaths, and they choose games based on that preference [source: Lillebuen].
In fact, there are considerable deficiencies in studies connecting violent game play with violent (or, more accurately, aggressive) behavior, including a failure to contextualize and a general inconsistency among results [source: Ferguson].
For instance, a 2010 study published in the "Review of General Psychology," reviewed past studies that reported ties between violent games and violent responses. The researchers found the subjects most deeply affected by violent game play were those who exhibited personality traits indicative of psychoticism, which include lack of empathy, nonconformity and impulsiveness. People who fit this model are predisposed to see violence as an appropriate response to social conflict [sources: Markey, Harris].
National trends, which rely on much larger sample sizes than do lab studies, also go against a causative relationship between video games and behavior: While sales of violent games are increasing (and games themselves are getting more violent), violent crime rates in the United States are going down [source: LiveScience]. In 2013, crime rates in the 10 biggest U.S. cities were the lowest in four decades [source: Zadronsy].
And one 2013 study, published in the "Journal of Youth and Adolescence," found the behaviors of "at risk" kids to be unaffected by playing violent video games. "At risk," in this case, meant they exhibited symptoms of attention-deficit disorder or depression, conditions widely believed to increase vulnerability to the potentially negative effects of video game violence. The researchers looked at the behaviors of 377 such U.S. children, examining both their game-playing habits (violence levels, time spent playing) and their negative social habits (getting into fights, bullying, cutting class), and found no link between the two. In fact, in a small number of children with attention-deficit symptoms, playing violent video games actually seemed to correlate with slight reductions in bullying [source: Ferguson].
In an open letter to the American Psychological Association (APA) published in 2013, a group of more than 200 psychology experts petitioned the APA to re-examine its resolution that "all violence be reduced in video games and interactive media marketed to children and youth" [source: APA]. As people of science, they explained, they simply could not support an official position based on what they considered dramatically insufficient evidence. As of early 2014, an APA task force was still looking into the matter.
Scientific proof aside, you won't find many people in favor of children (or adults, for that matter) playing pretend murder for hours a day. Most retailers refuse to sell violent "rated M" (mature) games to kids under 17, and every console on the market has built-in parental controls [sources: ESRB, Jackson]. It's likely we'll never know whether virtual violence begets actual violence. Imagine the ethical concerns in designing a study that could definitively prove that. In the meantime, video game sales increase: "Grand Theft Auto 5" broke a Guinness World Record in 2013 for "highest revenue generated by an entertainment product in 24 hours" [source: Lynch].