Want to know when a statistic is useless?

Aramis Ramirez leads the Major Leagues with a .427 batting average in August while teammate Reed Johnson is fifth at .400. Ramirez is tied for tops among any Cubs player in the month of August, dating to 1946. Shawon Dunston batted .427 in August 1991; Bill Madlock hit .425 in August 1975; and Manny Trillo hit .409 in august 1986. Carrie Muskat

You’ll often hear people who appreciate stats say that something is a small sample. This means that the amount of playing time isn’t sufficient to find the statistic as having any value whatsoever. This is something that people learn at very early ages. Did you ever flip a coin as a child that lands 3 times in a row on heads or tails? Did you run to Mommy and tell her that you had a coin that was always going to land on heads or tails? Of course you didn’t. Even at that age, you understood that 3 coin flips was not nearly sufficient to reach that conclusion.

That had .3 and .5 inches. These mean nothing becasue the samples are insufficient and they are nothing other than a subset of a much larger set of statistics.

Anyway, the batting average for the month of August is just that: useless. Know how I know? Because Shawon Dunston is tied for best in Cubs history in a single August month. Because Manny Trillo is 4th best. Dunston, career .269 hitter, and Trillo, career .263 hitter, being at the top of any list of batting average means one or two things: the split is useless and/or the statistic being measured is useless.

In this case, it’s a bit of both. Batting average is mostly useless, but the monthly splits are entirely useless. 


Quantcast