from Amazon
Saturday, September 20, 2008
WPA/LI
I posted this at Primer:
I’m not discussing whether WPA/LI is appropriate or not for Cy Young (I would lean toward not). I’m asking a more general question about what Balrick thinks that WPA/LI is trying to do.
***
Thank to fret for bringing up that game.
In that particular inning, the WPA totaled -.049 wins. Indeed, every scoreless top of the half will give the defense -.049 wins in WPA, regardless whether you had 3 or 6 batters. People who like just runs like that. Those who care about number of runners allowed don’t.
WPA/LI for the game in question for the top of the 1st was +.019. (Plus is bad for the defense.) What WPA/LI does is treat each PA in isolation, unleverages WPA so that it scales the performance so that each PA is weighted at exactly 1.0, and weights each event (walk, homer, strikeout, etc) with respect to the game state (e.g., K are more impactful with runner on 3b and less than 2 outs than otherwise).
Is it necessarily a bad thing that a team that allows 2 runners out of 4 batter to reach base to show that they are below average (even though they got out of the inning scoreless)? I don’t think it’s necessarily a bad thing. I don’t know that it’s necessarily a good thing.
But to dismiss WPA/LI would be preferable with reasoning behind it.
WPA/LI is exactly perfect for a hitter, if you believe that it matters when a hitter strikes out with runner on 3B and less than 2 outs, and you don’t think that it’s just another out. That is, you believe that the batter and pitcher realize that that situation requires a fairly strong change in approach. Only WPA/LI will give you what you need, under this belief premise. Linear Weights by the 24 base/out states is a close second.
WPA/LI is not as good for a pitcher, since a pitcher is his own team, and each PA should not necessarily be treated as if they are as impactful as any other PA. WPA/LI forces each PA to be worth exactly 1.0.