« Rundown Updates | Main | Colby Does Not Love PECOTA »

January 31, 2009

PECOTA Does Not Love Texas

Baseball Prospectus just released its first run of PECOTA ratings for 2009. I’m not going to print their subscription-only data, but I have performed a quick analysis to answer the question of what PECOTA thinks of the Rangers this season.

Answer: not much.

Before I divulge the results, some background. PECOTA estimates plate appearances and innings pitched for all players, but it doesn’t attempt to justify them on a team level. Many of their predictions read as “what ifs? (for example, what if Martin Perez pitched 99 innings in Arlington?). Thus, the team-wide sums are preposterous (in Texas’s case, 8,700 PA and 2,600 IP). To correct this “problem? for my purposes, I made my own estimates of plate appearances and innings but still used PECOTA’s rate estimates verbatim. I also chopped Texas’s unearned runs to 75 (a little worse than the league average) from last year’s 107.

Offensively, PECOTA understandably thinks poorly of Elvis Andrus and Omar Vizquel. It also predicts Taylor Teagarden will bat .209 (albeit with ample walks and power). The more pressing issue is the regression toward the mean for most of the established hitters. Only Josh Hamilton and Ian Kinsler merit OBPs in excess of .350. The less said about Michael Young, the better.

PECOTA, as always, hates Texas pitching. The only hurlers with sub-5.00 ERAs are relievers Frank Francisco, C.J. Wilson, Josh Rupe and Warner Madrigal. PECOTA really thinks ill of Matt Harrison, and folks like Derek Holland and Neftali Feliz aren’t projected as saviors yet.

My team-level use of PECOTA predicts Texas will score about 790 runs and allow 920. That’s worth 69 wins. Ugh.

Big caveat: One potential problem with my analysis is scaling. That is, I could do this with every team and discover that the MLB as a whole has a net deficit in runs scored versus allowed, which of course is impossible. If, for example, I discovered that the “average? team had a net deficit of 30 runs, I’d have to add about three wins to every team’s total.

You might recreate this exercise with different results, but I think the general theme is clear. PECOTA forecasts a significant decline in Texas’s offense with only modest improvement in pitching. I look forward to BP’s own team-wide PECOTA-based predictions.

Posted by Lucas at January 31, 2009 11:52 AM