Wednesday, May 2, 2007

Ok, I can't take much more of this

Unlike NBA referees, race is not an issue for these guys.

The New York Times is reporting today that when you put a bunch of white referees together, they call more fouls against black players than an all-black crew (and, conversetly, that all-black crews call more fouls on white guys). Are we really doing this repressed racial bias jive again? What's next? WNBA players make less than NBA players because they are women?

Fortunately, a deeper look at the information reveals that the study is a bit flawed.

I think I should be clear that the people that came up with this study are bright guys (Wharton Professor and Cornell Grad Student) and the third parties who evaluated it are also bright (Harvard, Yale and Ca-State Graduate studies Profs). So I'm not calling these people stupid in the same way I would Pete Prisco both fat and stupid. I am just saying that their study doesn't reveal anything.

The study looked at NBA referee calls from 1991 to 2004 and came to the conclusion:
"that players who were similar in all ways except skin color drew foul calls at a rate difference of up to 4 ½ percent depending on the racial composition of an N.B.A. game’s three-person referee crew."
They then went on to conclude that the difference is large enough that:
"the probability of a team winning is noticeably affected by the racial composition of the refereeing crew assigned to the game.”
Really? 4.5% has a "noticeable" effect? Let's delve into this claim a little more deeply....

First, on the noticeable effect side, let's concede that white refs call 4.5% more fouls on black players than black refs. Now, an average NBA game has around 40-45 (depending on the teams playing) fouls called during the game. I'll use the higher number of 45. An average NBA game is decided by around 9pts. Then, while black players constitute 83% of the NBA population, let's even assume that for this argument all the players on team A are black and all the players on team B are white. If the increase is 4.5% with a white crew and black players, the MAXIMUM effect on the game would be 3 extra fouls called per game called against the black team (and that's a LIBERAL statistical reading. It's much closer to 2 extra fouls a game). Unless all those fouls are shooting fouls on 3 pointers, the effect on the game is that the game is closer but ends the same way. So the effect on the game argument doesn't fly with me. Two extra fouls per game doesn't make a significant difference, even if the study is true.

Lastly, aren't we getting tired of these race related statistical arguments. Here's a statistic for you: The NBA is 83% African-American. However, teams with African-American GMs feature rosters in which 90% are African-American. Does that mean that African-American GMs are racially biased? And what about Joe Dumars? Detroit only had three white players and he TRADED two of them (Darko and Arroyo) for a black player (Kelvin Cato) and a draft pick... and they will probably take a black player with a the draft pick they got! Isn't that clearly racial bias? If that argument is absurd, why isn't the referee bias argument? It's just all so stupid.

Why do we use benign statistics to prove the unprovable? I will never understand it.

No comments: