A Cub-fan friend of mine recently said to me that Carlos Marmol is two entirely different pitchers. One day, he comes out of the bullpen and is unhittable, but the next day he is wild and ineffective. He assured me he could predict how well Carlos Marmol would fare in the game based on his immediate performance when he entered the game. I found this fascinating. Could you really predict whether good Marmol or bad Marmol is going to pitch that day after, say, ten pitches? If you could, that would be a tremendous advantage to have.

The thing about Carlos Marmol is that he is basically unhittable. The batting averages against him since 2007 are incredibly paltry, .169 in 2007, .135 in ‘08, .170 in 2009 and .148 this year. The only way to beat him is to hope that he is wild. Let’s take a look at every game in which Marmol has appeared since 2007 and we’ll break down his first ten pitches into the total number of balls and strikes. What we’ll try to do is see if Marmol’s performance can be predicted based how wild he is. We can hypothesize that the more balls he throws in the first 10 pitches, the less effective he will be. Our measure of effectiveness will be opposing OPS for at-bats that begin after the 10th pitch in each of Marmol’s appearances from 2007 through last night. Here is the opponent OPS based on how many balls Marmol throws in his first ten pitches.

Our hypothesis was incorrect:  Marmol's first ten pitches aren't indicative of things to come. The good news for Cubs fans is that regardless of the number of balls thrown in the first ten pitches, Marmol is remarkably effective and consistent.