THE FOREST FOR THE TREES

At lunch today I am reading a little CNNSI. I started reading John Donovan's post about playoff pitching. It was going OK until he got to the White Sox. As an aside, if I have any readers who don't actually know me (doubtful), I am a born and bred Cub fan. As such, I have no "homer" instincts toward the White Sox at all. They are not relevant to me. Thus, this is not a "homer" issue. It is a baseball issue.

Anyway, John Donovan wrote:

From Aug. 8 to Sept. 7, Mark Buehrle, Freddy Garcia, Jon Garland, Javier Vazquez and Jose Contreras have made 30 starts. They threw 191 1/3 innings, won 13 and lost 11 with a 4.28 ERA. From Aug. 6 to Sept. 7 in their championship year, also a 30-game stretch, the White Sox's rotation -- swap out Orlando Hernandez for Vazquez and give Brandon McCarthy a start -- was 14-12 with a 3.17 ERA. The starters threw 207 1/3 innings.

More than a run difference, while throwing 16 more innings.

Yeah, this group could be sucking air.

This is a frustrating side of the statistics revolution in baseball. I generally like the much closer analysis of the game that the sabremetricians apply to the game. For instance, I think that on-base percentage is a fantastic statistic. I think slugging percentage is very useful. I have doubts about OPS, but that's OK. Admittedly, Donovan uses no modern statistics in the above. However, the reliance on statistics, as opposed to Joe Morgan's "understanding" because he "played the game" already puts Donovan's statement on the statistics side of the cultural divide.

Anyway, the one use of statistics I cannot abide is exemplified by Donovan's statement. The reason they play baseball games is not (a) to count the runs given up by pitchers per nine innings, or (b) to accumulate innings. They play baseball games solely to count the number of times your team has scored more runs than the other team when the games end. Thus, the critical statistic that Donovan ignores is that from August 8 to September 7 the White Sox starters won 13 games. Last year in roughly the same time frame they won 14 games. Since they lost one fewer this year this is a standings wash. All of the statistical analysis is meant to help us understand why the standings look the way they do. They are not an end unto themselves. Nobody gets a championship ring solely for leading the league in runs above expected production. They get them for winning baseball games.