« NP: Where ignorant armies clash by night | Main | @FullComment.com: A book recommendation... »

A major event in sabermetrics

I’ll compress this as tightly as possible for the benefit of the casual sports fan. “Sabermetrics” is just a term for the organized study of baseball using the traditional means of science and scholarship; it most commonly refers to the study, in particular, of baseball statistics.

There is a predictable, empirically established relationship between (a) the number of runs a major-league baseball team scores and allows in a season and (b) its won-loss record. This relationship is usually called the “Pythagorean expectation” because there’s a sum of squares in the equation. Bill James, who originally discovered the Pythagorean formula, announced a finding in the early ‘80s that teams which do better in the standings than their run totals would suggest will revert to the Pythagorean norm the next year; on the basis of a limited sample he could find no tendency for teams to outperform Pythagoras from year-to-year.

He has now searched again for such a tendency, using 100 years of data, and found it. Beating Pythagoras appears to be a repeatable skill. If this result holds up under peer review, it will overturn (or at least modify, as Einstein did Newton) one of the most widely accepted statistical laws of baseball.


TrackBack URL for this entry:

Comments (3)

The 1960s Dodgers treated the night before Sandy Koufax pitched as their night-on-the-town, because they figured that they could play hung over because Sandy only needed one or two runs to win (e.g., the 1965 game deep in the pennant race against the Cubs in which they got one-hitted and still won because Koufax threw a perfect game).

So, I could see a team with an especially disparate quality pitching staff (some outstanding pitchers and some really bad ones) winning more games than their runs totals would suggest because they routinely win low-scoring close ones with their best pitchers and lose big blowouts with their bad pitchers. That could be reasonably stable from year to year. But it probably wouldn't be because bad pitchers tend to get replaced by mediocre pitchers.

Another possibility is that a veteran team might steal some close wins through savviness. But veteran teams tend to get worse if they maintain the roster.

Apparently the game in question took all of 103 minutes to finish. No wonder Banks used to say "Let's play two."

Thanks. The one hit of the game was Sweet Lou Johnson's "double to first" -- he blooped one over Ernie Banks's head and it dribbled off into foul territory before the first baseman ran it down. What an offensive classic!


This page contains a single entry from the blog posted on October 15, 2007 8:45 PM.

The previous post in this blog was NP: Where ignorant armies clash by night.

The next post in this blog is @FullComment.com: A book recommendation....

Many more can be found on the main index page or by looking through the archives.

Powered by
Movable Type 3.35