August 6, 2005

Steven D. Levitt Concedes He Was Wrong about His Most Controversial Theory:

No, not his abortion-cuts-crime theory, which generated very little critical comment, but his assertion on his blog that Billy Beane, general manager of the Oakland Athletic's baseball team and hero of Michael Lewis's bestseller Moneyball about how Beane's use of Bill James's statistical tools made the A's a budget powerhouse, was over-rated. Levitt demonstrated that in recent years the A's offensive statistics weren't dramatically better than other good teams.

This proved highly unpopular with the commenters on the Freakonomics blog, who pointed out that Levitt didn't know what he was talking about. The point of Billy Beane's strategy is not to have the best offensive statistics, but to have statistics comparable to the Yankees and Red Sox with much lower payrolls. And, anyway, Levitt was forgetting to adjust for the fact that Oakland doesn't play in a hitter's park. Their field doesn't favor pitchers as much as in the old Reggie Jackson days, but, still, anybody who knows anything about baseball statistics (i.e., not Levitt) knows you have to adjust for the park effect.

All through April and May, Levitt bragged about how badly the A's were doing. On May 28th Levitt began his post entitled "I need some abuse, so here is another baseball post:"

It's probably poor sportsmanship to do another Oakland A's post at a time when the A's have now sunk to a record of 17-31 and are ahead of only one team in the American League.

Well, now the A's are 61-47, having gone 44-16 since Levitt's gloating. Yesterday, Levitt announced, "I concede on the A's," adding:

I don't know anything about baseball, as evidenced by my prior posts... Economists, like everyone else, are much better at explaining the past then predicting the future.

Commenter Jonathan Schwarz replied:

If economists can't predict the future with the same model that they used to explain the past does that pretty much mean that they did not really explain the past?

Good question. The funny thing is that Levitt's abortion theory failed even more catastrophically to predict the past than his baseball theory failed to predict the future (the teen murder rate for the first cohort born after legalization was three times higher than for the last cohort born before legalization), but nobody noticed. It turns out to be less damaging to his reputation to be wrong about the past than about the future ... even though an hour's online research with FBI statistics back in 1999 should have convinced Levitt not to bother foisting his slapdash theory on the world. But he didn't do the due diligence, and when I pointed it out to him in Slate, his gigantic ego was too committed to back down so he's gone on pushing his bad idea on an intellectually defenseless public.

But why is the American intellectual class so defenseless?

What's also funny (and maybe more than a little sad) is how much more brainpower in America is devoted to analyzing baseball statistics than to analyzing crime statistics. Levitt's massive ego encouraged him to talk through his hat about (at least) two subjects he was ignorant about: crime in 1999 and baseball in 2005, and he came up with equally dubious theories about both. Yet, the response to his abortion-cut-crime theory has generally run from trusting to wildly enthusiastic, while his back-of-an-envelope musings about baseball brought down on his head cannonades of detailed analysis of his mistakes.

Now, I've wasted a lot of time looking at baseball statistics, too, but does it strike you that there is something decadent about the disproportion between how much intelligence is devoted to thinking about baseball statistics (a field in which there has been tremendous progress over the last thirty years) and how little effort our people put into analyzing statistics that really matter, like crime rates?

The problem with thinking about social statistics like crime is that they are all intimately tied to racial differences, which aren't supposed to exist, so thinking about them is career-threatening.

My published articles are archived at -- Steve Sailer

No comments: