Wednesday, May 02, 2007

Basketball and race.

This is sort of interesting; the NYT discusses a study (not yet published) that purports to show a slight bias in NBA officiating on the basis of race. That is, white refs call fouls on black players more often than white players; and black refs call fouls on white players more often than black players. (At least, I think that's the conclusion; the article isn't totally clear on the claim that was tested.) The NBA has supposedly done a study that reaches the opposite conclusion, but because they won't release the data for independent confirmation, this is little more than PR.

That said, here's a quick reply from Nicholas Beaudrot (here):
  • It's possible ... and I say this having not read the paper ... that fouls per 48 minutes is not the proper rate state to examine. It would be interesting to examine only fouls during the act of shooting, to see if black players are called for more fouls in the act of shooting fouls against players in the act of shooting than white players.
  • It's also possible, though unlikely, that ... how to say this delicately ... African-American players and white players play the game differently, which accounts for the slight difference in foul rates.
  • An increase in fouls of 2.5-4.5% per 48 minutes for African-American players is incredibly low. If state troopers only stopped black drivers 5% more often than white drivers we would be crowing about the incredible reduction of racial profiling.
I'm not sure about the first point; it's sufficiently technical to be beyond my expertise. As for the second, we'd have to see the paper to really tell if they controlled for this or not. The third, though, is the kicker: this is a tiny bias. What it says, if the result can be repeated (which, if -- big if -- the NBA study holds up, it can't) is that officiating is generally fairer than other areas of society where racial bias is a problem. So, really, the question should be: what are NBA refs doing right?


Dan said...

Considering that EVERYONE has some level of hidden prejudice, that's actually a pretty good rate.

ADHR said...

Exactly. So, one wonders what they're doing right.