FREE hit counter and Internet traffic statistics from freestats.com

Sunday, October 10, 2004

Design the Residue of Luck?

Branch Rickey is often quoted as saying that "Luck is the residue of design". Jay Bennett (the co-author of Curve Ball) and Aryn Martin turn this phrase on his head in this chapter entitled "The Numbers Game: What Fans Should Know About the Stats They Love". The chapter is included in the book Baseball and Philosophy: Thinking Outside the Batter's Box.

In particular I think its instructive to consider their analysis of the variation in batting average among the 146 players who qualified for the batting title. They conclude that the variation resulted from two sources: ability and chance, and thatfully half of the variation (the averages ranged from .215 for Jeremy Burnitz to .370 by Barry Bonds) is the result of chance, the left - the residue - the result of ability. For most fans that just doesn't sound right since we think of .300 hitters, for example, as having the ability to hit .300 when in reality their actual ability is best thought of as a range instead of as a point. This is similar to Stephen Jay Gould's often repeated statement that variation, not type, is the central aspect of biology.

More interestingly for sabermetricians Bennett and Martin then go on perform a similar analysis with on-base percentage, slugging percentage, and OPS. They found that ablity played a much larger role in determining these values, as much as 3 to 4 times that of chance. So not only does OPS correlate better with run scoring, it also is a better measure of a hitter's true ability. While general managers are just now figuring out the former, I haven't heard much discussion of the latter.

No comments: