Friday, November 13, 2009
A look at arm ratings
Posted by Jeremy GreenhouseUsing 2006-2008 data, I ran a regression using UZR's ARM rating per 150 games as my dependent variable against the three Z-Scored arm ratings on Tangotiger's Fans' Scouting Report-accuracy, strength, and release-and using the number of ballots as my weights. Fans can explain about 10% of the variance in an outfielder's ARM runs per 150 games. It turns out that grading out a 60 on arm strength, one standard deviation above the league average of 50, is worth around half a run over the course of a year. Here are the estimated coefficients of the regressions. An asterisk indicates statistical significance at the 10% level.
| All | Left | Center | Right | |
|---|---|---|---|---|
| Strength | 0.5* | 0.2 | 1.7* | 0.3 |
| Accuracy | 0.7* | 1.4* | -0.5 | 0.4 |
| Release | -0.1 | -0.4 | -0.6 | 0.6 |
Either the fans don't do a good job of assessing the ability of an outfielder to release the ball quickly, or that aspect of an outfielder's game just isn't important in stopping the running game. Arm strength is most important in center field, while accuracy is valued highest in left, confirming our intuition.
Any questions? Feel free to email me.