BABIP Regression

Prior to the Cubs signing Carlos Marmol to a 3-year extension this past offseason, there was a lot of talk about how Marmol was one of those guys who could have a very low BABIP. His BABIP as a reliever since 2007 had been .241, which is well below the average and even below the average for an extreme fly ball pitcher. It’s well over .300 this season and since the beginning of 2010 his BABIP is .312. That’s higher than we’d expect even for an average pitcher, but quite a bit higher for someone like Marmol. 

I’m not highlighting Marmol’s BABIP regression as an I told you so. I thought the contract they signed him to was a good one at the time and projections said he’d be worth it. I still feel confident he’ll be worth it, but I wasn’t so sure the Cubs should sign a reliever to that contract who still had 2 years of club control. One of the reasons people pointed to in support of signing him was that he could maintain a low BABIP, but there wasn’t really any reason to believe that he could. 

I didn’t expect him to have a .331 BABIP this season, but it’s not any more surprising than a .250 BABIP would be. I’m only posting this as an example of how we have to regress statistics even if our eyes tell us he’s just very hard to hit. Marmol has still maintained a low HR/FB rate, which has been the difference between being a decent closer and being a terrible one. His HR/FB rate is still more than double what it was the previous two seasons combined. Here’s to hoping his HR/FB rate can remain quite low. 

The increased BABIP results in an increased batting average allowed and thanks to Marmol’s poor control, he’s allowed batters to have an OBP of .357, which is quite a bit worse than league average. 


Quantcast