Cross: Forecasting pitcher platoon splits

From Jared Cross at The Hardball Times on August 14, 2015:

In The Book, Tom Tango, Mitchel Lichtman and Andrew Dolphin found that pitchers’ platoon splits are considerably more variable than hitters’ splits. In fact, they found that pitchers with true reverse splits, meaning pitchers who are actually less effective against same-handed hitters, are not exceedingly rare. The Book reports that (p. 165):

Among the pitchers with large platoon splits, we find that about three quarters rely largely on a slider or a non-overhand curve. Pitchers with ¾-arm or lower (e.g., Fossum, Bradford and Reed) deliveries also tend to be found on the high-platoon split list. This also suggests that one can estimate pitcher platoon splits even better if one knows the types of pitches being thrown and the arm angle and then divides pitchers accordingly, rather than merely by handedness. Unfortunately, our data do not include pitch types, so we must proceed without exploiting this fact.”

In the 2008 Hardball Times Annual, John Walsh went back further in time and using Retrosheet data and the Neyer/James Guide to Pitchers found that pitchers from 1957 to 2006 with extreme platoon splits were often side-armers and, as a group, threw more sliders than curves. Pitchers with small or reverse platoon splits tended to throw more change-ups and curveballs and fewer sliders. He also noted that low-platoon-split pitchers were often noted as throwing a screwball. Walsh quantified his findings in the following table where a pitcher’s first pitch (according to Neyer/James) was awarded five usage points, his second pitch three points and his third pitch one point.

Read the full article here: http://www.hardballtimes.com/forecasting-pitcher-platoon-splits/



Originally published: August 14, 2015. Last Updated: August 14, 2015.