Happy Labor Day! Regularly Scheduled Articles Will Resume on Tuesday, September 2.
January 9, 2014
Rereading Nate Silver: 16. Batting Average and Stablizing Stats
In which batting average gets seeeeeerrved.*
Randomness: Catch The Fever!
Abstract: Nate shows that players with extremely high batting averages one season tend to see their batting averages decline, and that even relative to other offensive rate stats (isolated power, walk rate) the regression is severe, because batting average is so random. This is so obvious that he acknowledges that it is so obvious, anticipates criticism that it is so obvious, but “really that's the point. Regression to the mean isn't some lofty concept or novel idea--it's the cold hard fact.”
Key quote (as part of an unrelated tacked-on section about Mike Piazza giving up catching): “Catching is hazardous to a player's health. If you quit smoking at age 34, how much of the damage is reversible?”
Russell: Last year, Russell Carleton updated his own work on when different metrics stabilize, considerably advancing the relatively modest point Nate was making. Batting average takes longer to stabilize than anything (well, other than 2B+3B rates), Russell found. Batting average takes more than 15 times longer to stabilize than strikeout rate. Swing rates stabilize even more quickly. In an essay for this year’s BP Annual (pre-order now!) Carleton gives a pretty good demonstration of how to use this fact and why to use this fact; it’s actually exceptionally relevant to the future of sabermetrics that Russell predicts in his essay (and which I won’t spoil for you here).
On the Nate Silver Must-Read Scale: 1