from Amazon
Tuesday, November 24, 2009
Pitching to contact and FIP
Kincaid goes through GOOD mathematical gymnastics.
I personally don’t care too much if the difference is under 5 wins for a full-season. FIP gives you a quick view, and that’s all I want from it. And as Kincaid shows, if you try to improve on it to show the pitching to contact scenario, you barely make a dent. If you really wanted to be serious, you’d actually follow the full DIPS calculation to begin with, and use BaseRuns to estimate the runs allowed. That, of course, is a b!tch to do. So, on principle, I’m against using FIP to do anything other than what it is intended for.
Also, FIP tries to remove TWO pieces of information: the impact of the fielders AND the sequence of events. This second part is important to remember. It’s fielding AND sequence independent: SFIP, if you must.
Sort of on the same subject - A couple of weeks ago, I did some simple calculations to see if a 4.70 FIP flyball pitcher (Jarrod Washburn) would be much worse than a 4.70 FIP groundball pitcher. At the very most, Washburn gives up an extra 50 fly balls. An average defense catches 39 of them, a really bad defense (Young, Cuddyer, Span) probably catches 37. Add in the fact that he also wouldn’t be taking advantage of the Twins good infield defense (Morneau, Punto, Hardy, possibly Crede) and you maybe get 5 outs.
And this is a pretty remarkable scenario - an extreme flyball pitcher compared to an extreme groundball pitcher, an extremely good infield defense and an extremely bad outfield defense. Signing a pitcher based on your ballpark might work - signing Washburn to pitch is Safeco is a better idea than signing him to pitch in Coor’s, but the strengths and weaknesses of your defense probably don’t matter much at all.