Friday, October 12, 2007

ALCS 2007: The Usual Suspects

Over the last 15 seasons (1993-2007) the top 4 teams in the majors in aggregate regular season winning percentage are:
Yankees .598
Braves .594
Red Sox .549
Indians .543

The Mets are 14th in the majors with a .501 regular season winning percentage overall over the last 15 years: 1183 win and 1179 losses, the closest of any team to exactly .500 over this period. The bottom 4 teams in winning percentage over the last 15 years include, unsurprisingly, Tampa, KC and Pittsburgh, all of whom were in last place again this season, plus Detroit, which has come out of a long swoon the last two years.

Over this 15 year period, the Mets "Pythagorean" expectation, the number of wins they would normally have expected to win based on the number of runs they scored and gave up, is about 22 wins higher than they actually won, or about a game and half fewer wins per year on average than they would be normally have been expected to win. To put it another way, the Mets, based on their runs scored and runs surrendered, would have been expected to have a .510 winning percentage over the past 15 seasons rather than a .501 percentage. Not much difference there, and that is the largest underperformance by any team in the majors over that time period. The largest overperformance (more actual wins than expected based on runs scored and runs surrendered) was by the Yankees, who won 42 more games over the past 15 years than their "Pythagorean" expectation predicted, almost three a season on average. This was the largest overperformance of any team over the last 15 years, by a large margin.

technical note: there are slight variations in the Pythagorean expectation formula among different sabermetricians -- here I'm using the formula from, which is the simplest in use, i. e., the original one developed by Bill James: expected wins equals runs scored squared, divided by the sum of runs scored squared plus runs allowed squared.

No comments: