Wednesday, December 01, 2004
Proving Moneyball
Posted by Dave StudemanJC, who runs the Sabernomics blog, recently attended an economics conference that included, among other things, presentations of baseball economics. One paper in particular, An Economic Evaluation of the Moneyball Hypothesis, is fascinating for two reasons:
First, the authors set out to prove that major league teams truly did undervalue OBP before the publication of Moneyball -- and they make their case very well. But to take it a step further, they also show that the baseball "market" subsequently valued OBP correctly after Moneyball's publication -- proving that General Managers really did read and learn from the book (despite the widespread disdain shown toward it).
Second, they used Win Expectancy/PGP analyses of the 1999 and 2000 seasons to prove a number of things that have been discussed before. For instance, they show that OBP really should be valued twice as highly as SLG. And they developed their own set of Linear Win Weights for each event on the baseball field (closely following the approach already used by Tangotiger).
This is geeky stuff, but the article is very readable, and I recommend it to my fellow baseball geeks.
Dave was called a "national treasure" by Rob Neyer. Seriously. Comments about this article can be sent to him through the miracle of e-mail.