from Amazon
Tuesday, September 16, 2008
Brian Shouse
In close to 700 PA against both lefties and righties, Shouse (LHP): 0.375 wOBA against RHH and 0.275 wOBA against LHH. One SD in 700 PA is 20 points. The typical split is 20 points, so what we have here is 80 points farther than expected, which is 4 SD.
His K/BB ratio (excluding IBB, including HBP) is fairly impressive: 164/30 on one side and 52/58 on the other side. Guess which side is which. We can convert this into an ERA fairly easily: 5.4 minus 12*KBBdifferential/PA. That’s 5.51 against one side and 3.33 against the other side. The split is very real (K, BB splits are more real than other components).
Even his BABIP is fairly substantial: .328 v .278. (Sean shows .319, .258; he seems to count ROE as an out, and removes SF from the denominator; even then, I can’t reproduce his numbers.) That’s based on 500 BIP, meaning 1 SD = 20 points.
And yet somehow, he’s faced almost as many RHH as LHH. This is true of his career, and of the last 3 years. If you have a .275 from one side and .375 on the other side, and the average reliever is around a .330, and if you give him 50% PA for both sides, you get an overall Shouse of .325. Shouldn’t there be a better way to leverage such a pitcher? In Orosco’s last 3 years, he faced 221 LHH and 133 RHH, or 62% LHH. Couldn’t Shouse have been used to that extent somehow? Instead of 100 PA against LHH and 100 PA against RHH, then he gets 80 against LHH and 50 against RHH? If you do that, you get 130 PA of .313 wOBA. As long as you can find someone better than .346 for the other 70 PA, you come out ahead.
Hmmm… that might be tougher than it looks. It seems that even with someone with such an enormous split, it may be pretty difficult to get Shouse to pitch less than he does. The only thing you can possibly do here, to evaluate his usage by his manager, is to go to the gamelogs, and look at his usage, game-by-game.