from Amazon
Friday, January 12, 2007
Solving Barry Zito, or Why Does DIPS Not Work With Him?
When you get data into the hands of smart, resourceful people, you get answers. And boy, did we get answers:
First, here are three excellent articles by Ken Arneson. The third is the best on to read, but the second one is the insightful one that concerns us at the moment:
http://catfishstew.baseballtoaster.com/archives/575440.html
http://catfishstew.baseballtoaster.com/archives/575894.html
http://catfishstew.baseballtoaster.com/archives/576977.html
Many argue that Zito is a fly-ball pitcher who has been greatly helped by the large foul territory and the damp air of the Oakland Coliseum. And yet there’s this: his career ERA is better on the road (3.44) than in Oakland (3.66).
I simply treated Zito as a FB pitcher, and since FB have lower BABIP, that, I figured, was enough for me. The rest of his statement, regarding the foul territory, I was never too concerned with, but we know it has some effect.
No, the interesting part is when he says:
In fact, I’d guess that Zito has such a low BABIP because he makes batters hit easy-to-catch fly balls. He keeps his BABIP low by inducing batters to hit weak fly balls. Zito is consistently among the MLB leaders in popup percentage.
Which was further supported by Adam Morris as he looks at popup percentages by handedness of batter. Fascinating stuff. It then led me to Baseball Reference and I looked for Zito’s BABIP against LH and RH. Here it is:
RH: .260
LH: .292
A 32-point split. Is that alot? Sean hasn’t (yet?) implemented splits on a team or league level. So, we move on to Retrosheet, where the data is there, and we just have to figure the BABIP ourselves. Left handed pitchers had a BABIP of this against batters:
RH: .302
LH: .294
That’s right, while Zito’s career BABIP against LH is virtually exactly league average, he has a BABIP of 42 points less against RH! Zito has 3200 BIP against RH, meaning one standard deviation would be 8 points. His performance is FIVE standard deviations from the mean. That is about as significant as significant gets.
Thank you Ken and Adam for solving a huge mystery for us.
***
As an aside, for the sake of completeness, here is the 2006 BABIP for righty pitchers against:
RH: .298
LH: .306
As you can see, there is about an 8-point platoon advantage in BABIP. (You need to do a better job though, because not each sample is equally represented. The Book handled the issue better, but from a wOBA perspective, but, I should have looked at BABIP as well.) Anyway, the key here is that DIPS, in its pure form, should not be held so closely. There are alot of groups of data that shows that there is a definite difference, and platoon is yet another.