By now you’ve probably heard the story that yesterday ESPN’s Buster Olney tweeted about Trevor Cahill and his chances for the AL Cy Young and Keith Law’s replies about Cahill not being worthy based on advanced stats. Today Buster, who is generally a pretty smart guy, made some interesting strawman arguments comparing Cahill to Roger Maris and some other things that really don’t mean anything.
The obvious question is: since FIP and xFIP are mainly predictive, what role do they play in awards vs. stats that are more results based?
Personally, I think they all have semi-equal roles to play. Wins and ERA are thrown out because they both have a huge amount to do with either luck or how good the team is. As Phil Mackey pointed out in a tweet tonight, Matt Capps’s ERA gets penalized because Delmon Young kicked a flyball in Tampa, but his xFIP remains unblemished. At the same time, FIP, and more to the point xFIP is more of a “this should be what happened” stat. Francisco Liriano is leading the league in least home runs given up at .18/9 innings. xFIP normalizes his home run rate to the league average, making his xFIP 3.00. While Liriano’s great home run rate is not sustainable (or at the very least not very sustainable; I think it’s probable that pitchers have at least some control over their HR rate), the fact remains that Liriano has not given up many homers this year. He shouldn’t be penalized for the fact that he “should have” given up more homers than he has.
To put this a better way, let’s examine an analogy. Let’s say I am in a class of 50 people and I am taking a 100 question multiple choice test, which is amazingly difficult. I know roughly 75 of the questions, but have to guess on the final 25. By some sort of magic, let’s say I guess 24 of the 25 right and get a 99% on the test, whereas other people, some who might know 70 of the questions, some who might know as many as 80 or 85, guess horribly. I end up with the best grade in the class and should absolutely get credit for that, but at the same time there is no way I am smarter than the people who knew the answers to 80-85 of the questions, and in no way am I the smartest kid in the class.
Cahill is a very good pitcher. In time he could be one of the best in baseball, but for right now he is more like the kid who gets 75 questions right and has to guess on the final 25 and for whatever reason he is guessing correctly this year. It’s not going to happen again, but at the same time he should still get credit for doing so.
Last year I thought Franklin Gutierrez should have gotten some MVP votes because his WPA was 3.79, one of the best in baseball. Obviously clutch hitting is not sustainable (he’s at -1.28 this year), but it does happen, and people should get credit for that.
Olney is wrong thinking that Cahill’s ERA and WHIP make him a viable Cy Young candidate, but people who say only FIP and xFIP should determine who wins the Cy Young are equally incorrect.