Monday, May 26, 2014

How OBP And SLG Contributed To Pitchers' ERA, 1960-79

It is easy to call up this data using the Baseball Reference Play Index. I used all pitchers with 2000+ IP (57 guys). I regressed ERA on OBP and SLG allowed. Here is the equation

ERA = 13.7*OBP + 6.918*SLG - 3.377

The r-squared was 0.867 and the standard error was 0.11.

I calculated each pitcher's predicted ERA and sorted them by how much it differed from their actual ERA (from most below to most above). Click here to see the results. Jim Palmer was the most below his predicted value (.019). His actual ERA was 2.66 while the predicted value was 2.85.

This does not necessarily mean that he was a clutch pitcher, like being good at getting guys out with runners on base. Maybe fielders like Brooks Robinson, Mark Belanger and Bobby Grich turned more than the average number of DPs. There could be other reasons for why some guys were above or below their prediction.

But all but 3 guys were within .20 and 35 were within .10.

I did a similar study on pitchers with 1000+ IP from 1991-2000. Click here to see the results. Here is the regression equation. The coefficients are certainly different than in the above study. Not sure why

ERA = 11.93*OBP + 10.21*SLG – 3.91

I later added in some variables to take into account handedness and strikeouts, as they might affect double plays.

No comments: