No, Killing Prostitutes in GTA V Won’t Make You More Moral in Real Life


The latest cultural spasm over a video game has arrived. This time, it’s the fact that in the just-released next-generation console versions of GTA V (that is, the versions for Xbox One and Playstation 4) you can play the game from a first-person perspective. Meaning, naturally, that you can have sex with a virtual prostitute from a first-person perspective. Then, if you want, you can kill her and take her money back — charming!

Over at Breitbart London, Milos Yiannopoulos assures us that not only is this a perfectly harmless exercise, but that prostitute solicitation and murder may actually make players better people:

I know what you’re thinking. Isn’t this all a bit sick? Isn’t there something a bit…wrong with men and women who sit at home acting out violent fantasies? As I say, I used to think so too. But the research says there’s absolutely no evidence that violence in games, or depictions of sexy women, make players any more violent, or misogynistic, in real life. 

In fact, at least one university professor, Matthew Grizzard at the University at Buffalo, believes that video games make players more ethical, by increasing their sensitivity to, and awareness of, moral issues. “Violent video-game play may actually lead to increased moral sensitivity,” says Grizzard. “This may, as it does in real life, provoke players to engage in voluntary behavior that benefits others.”

That would be nice. Unfortunately, Yiannopoulos is totally misinterpreting Grizzard’s finding — a fact I confirmed with Grizzard himself. And while I don’t mean to pick on the guy, this is a very useful example of how nuanced social-science research gets a bit butchered when it’s presented for mass consumption. (Possibly unnecessary full disclosure: Though this particular issue has nothing to do with Gamergate, Yiannopoulos is an outspoken proponent of the movement, which I’ve expressed considerable skepticism at.)

So, what was this study, and what did it find? In short, Grizzard and his colleagues took a bunch of university students and put them into two groups. One was asked to write about a memory — either a time they felt guilty or an ordinary day. The other was asked to play a video game either as a terrorist or a U.N. soldier. After the participants completed either the memory exercise or the gaming sessions, the researchers used an established psychological scale to ask participants how guilty they felt (as well as some questions about moral foundations theory that aren’t relevant to the present discussion).

In short, the people who played as terrorists felt more guilty than those who played as U.N. soldiers, and, as expected, the people who thought back on a guilty memory felt more guilty than those who remembered a normal day. It’s also worth pointing out that the guilt effect of memory was stronger than it was for those playing as a terrorist. The researchers concluded, then, that there may be some potential for video games to induce guilt and, potentially, prosocial behavior.

But there are many reasons you can’t just graft this finding onto video-game playing in the real world. For one thing, just because a random person plucked off the street feels guilty about playing a terrorist, that doesn’t mean that an experienced game-player will. It could be the case that the random sample of college students recruited for the study differ in important ways from the sorts of people who play games like GTA V in real life, and that the guilt effects don’t hold for GTA players.

Along those same lines, Grizzard and his colleagues themselves explain that this finding doesn’t necessarily mean that doing bad things in games will make you moral in real life. This is straight from the paper:

Overall, the findings suggest two possibilities. First, repeated play as an immoral character may repeatedly activate guilt and its resultant influence on the increased importance of care and fairness. Under these conditions, we might expect that repeated play as an immoral character would lead gamers to become more sensitive to fairness and more caring overall. Alternatively, guilt resulting from playing as an immoral character may habituate from repeated exposures. Under these conditions, we might expect that repeated play would not lead a gamer to become more sensitive to fairness or become more caring overall, especially if the ability of the game to elicit guilt dissipates with repeated play.

That is, the researchers themselves are saying that maybe playing as a bad guy over and over again dissipates the sort of one-off feelings of guilt they observed in their experiment. In fact, Grizzard said in an email that in an as-yet-unpublished study he conducted, players who repeatedly engaged in violent in-game behavior did, in fact, become habituated to it — that is, the behaviors in question elicited less and less guilt. This is a serious strike against the notion that extended engagement in bad virtual behavior will make you a better person (though it also isn’t evidence, of course, that it’ll make you a worse person).

I know some of this comes across at nitpicky nerdery, but if we’re going to use social-science research to inform real-world debates — and if we’re not, what’s the point of social science? — it’s important for the folks who write about these issues to understand what individual studies do and don’t prove.

I don’t think Yiannopoulos woke up today and said, cackling to himself, “I’m going to misrepresent science to further my agenda!” But I do think he and many others who write about the intersection of video games and behavioral research — Grizzard told me that he’s seen several articles “stretching the results” of his study — should be a bit more cautious. This stuff is complicated.