THE BOOK--Playing The Percentages In Baseball

Tuesday, April 06, 2010

Pitcher batting 8th

In my Markov tests in The Book, putting the pitcher 9th or 7th was breakeven, with 8th being the optimal spot, but on the order of 2-3 runs or so.  (When MGL does his studies, he find the pitcher batting 9th to be ever-so-slightly preferable.) So, when Dave reports this:

The average pitcher-bats-9th team was 0.23 runs below optimal, while Pittsburgh and the AL teams averaged 0.07 runs below optimal. This would suggest flipping the pitcher and 8th hitter on the other NL teams would result in an improvement of about 0.16 runs. Over 162 games that is 25 runs or 2.5 wins, a surprisingly high number to me.

It should be surprising because it is impossible.  In my tests, moving the pitcher from 8th to 4th caused a 0.10 runs per game difference.  That’s the absolute limit for moving one player to the worst spot, while keeping the other 8 guys in the same order.  I presume Pinto’s model used SLG, OBP and regression, which I think he based it on Cy’s work (all this was done a few years ago)?  If that’s the case, then there is a limit to its usefulness.