BASEBALL: PECOTA PROJECTIONS ARE A FUZZY CRYSTAL BALL FOR 2006
The New York Times column, "Keeping Score" highlighted Pecota on Nov 13.
Pecota was developed four summers ago by Nate Silver, a 24-year-old Chicago financial consultant. "In some ways it was boredom," Silver said. "If I had a spreadsheet on my computer, it looked like I was busy." (Its official name, Player Empirical Comparison and Optimization Test Algorithm, joyfully boils down to the last name of Bill Pecota, a former major league utilityman.) The projections are anticipated enough among some executives that Silver has received e-mail messages saying, "Hey, are the Pecotas done yet?"
Though the system naturally cannot predict all player fluctuations, it succeeds more than most. Last winter, it identified Jonny Gomes, a relatively unknown Tampa Bay Devil Rays outfielder, as having an excellent chance at significant improvement. Gomes hit .282 with 21 home runs in just 101 games. It also foresaw even greater production from established players like Andruw Jones, Derrek Lee and Dontrelle Willis.
Pecota helps quantify the danger of long-term pitching investments, and points out which types of arms tend to project best. The strongest indicators of future performance are rates of strikeouts, walks and home runs, but also ground ball-to-fly ball ratio and even body type.
As for this year's free-agent pitching market, Silver sees a fine future for A. J. Burnett, an enigmatic 28-year-old right-hander. Burnett projects to have a 3.65 earned run average next season, but also has a 15 percent chance of blossoming into a Cy Young award candidate for several seasons. Of the two most intriguing relievers available, the veteran Billy Wagner and the less-known B. J. Ryan, Pecota chooses Ryan as the better long-term investment thanks to his age next season (30 to Wagner's 34) and several other peripheral factors.
NYTimes: Predicting Futures in Baseball, and the Downside of Damon by Alan Schwarz
Pecota site: Baseball Prospectus