Finally, Someone Makes A Game About Shooting White PeopleApril 9, 2008
So I woke up and checked my daily subscriptions, one of which linked me to an article in the New York Times about an online ‘test’ hosted by the University of Chicago Psychology department. Here’s a quote from the NYT piece:
To my horror, I turn out to be a racist.
The University of Chicago offers an on-line psychological test in which you encounter a series of 100 black or white men, holding either guns or cellphones. You’re supposed to shoot the gunmen and holster your gun for the others.
I shot armed blacks in an average of 0.679 seconds, while I waited slightly longer — .694 seconds — to shoot armed whites. Conversely, I holstered my gun more quickly when encountering unarmed whites than unarmed blacks.
Being the cunning, ruthless blogger that I am, I decided that there couldn’t be a better way to follow the outrageous success of the Top 13 Hottest Men of Gaming, In Color series than to upstage someone who writes for the Times. So I opened up the test and proceeded to let loose. 60 dead honkeys and negative 740 points later, I realize that the negative 20 points every time I shot an unarmed white man wasn’t a bug – the game wanted me to shoot armed people, not white people. Oops.
So I tried it again and scored a whopping 600 points – I think I accidentally shot 3 unarmed white guys and 1 unarmed black guy. (Sorry to the last guy there. My bad.) Interestingly enough, my reaction times didn’t parallel the NYT writers’; it took me 599ms to react to the armed black guys compared to 609ms to react to the armed white guys, and 672ms to react to the unarmed black guys compared to 682ms to react to the unarmed white guys. So the difference by race was ~10ms, but I seemed to be most suspicious of the unarmed white guys. (“You sure you don’t have a gun? That looks like a knife. Or smallpox – you never know with these guys.”) One would think that, having grown up in a white-dominant world, I’d be quicker to shoot black men and slowest to not-shoot them, but I guess that’s not the case.
As the game attempts to interrogate the internalized prejudices underlying the Amadou Diallo shooting, it’s designed to ‘reward’ shooting over not shooting; the biggest point score is +10 for correctly shooting, as opposed to +5 for correctly not shooting, and the penalty for incorrectly shooting is -20, compared to -40 for incorrectly not shooting. It’s also configured so that the right hand is used to shoot and the left is used to holster, which is significant insofar as most of the population is right-handed, and I’d imagine that would affect our reaction times. I wonder how the results would change if we flipped the controls – which half of our brain is the one that recognizes the shape of a gun?
Still, that ~10ms is pretty significant. I’m going to try switching to a white-skinned player model in Urban Terror and see if these results hold out.
EDIT: I just did a trial run through the game where I held the “/” key down. It reported:
Black Armed: 46.64ms
Black Unarmed: 52.16ms
White Armed: 46.08ms
White Unarmed: 52.28ms
Not really sure why this discrepancy exists between armed/unarmed and black/white if I’m just holding the button down. Huh.