Thursday, September 10, 2015

The math is in the Astros favor (not the Rangers).

The Pythagorean Theorem of Baseball created by the Godfather of sabermetrics, Bill James, is a simple yet useful statistic for baseball teams. If you want the full breakdown of the statistic click hereIf you want the TL;DR version (and you trust me) it's taking runs scored and runs allowed and creating an estimated winning percentage. Essentially, it's what a team's record should be based on runs scored and runs allowed.

The Astros are +99 runs prior to Wednesday's night game against he A's (+102 as of writing); the Rangers are -26 runs prior to Wednesday's night game against the Mariners (-31 as of writing). The +99 runs works out to an 81-58 Pythagorean W-L record for the Astros and the -26 runs works out to a 66-71 Pythagorean W-L record for the Rangers. Current standings have the Astros at 75-64 and the Rangers at 73-64. Based on the numbers above the Astros have under performed and the Rangers have over-performed so far on the season, which is why I'm not all that worried about the Astros being overtaken by the Rangers in the American League West standings.

The Athletics, with +5 runs, are closer to the Astros in run differential than the Rangers, which should probably be a warning for the Pythagorean W-L record. It's not all that it seems. The Athletics should be better this year and probably right behind the Astros in the standings, but they're not. They're last in the standings. That's the caveat to Pythagorean record: records don't regress towards the Pythagorean record like other advanced statistics do. Teams will over and under perform in seasons. In 2013, for example, the Astros under performed their Pythagorean record of 57-105 by six games. Yes, that's right, the Astros should have been six wins better than their 51-111 record they ended up with based on runs scored and runs allowed.

Back to 2015 and why I'm not worried about the Astros, and more importantly why the math is against the Rangers. Going back to at least 2000, no team has made the playoffs with a negative run differential.  Sure the Rangers are only one game back of the division (likely two, based on scores as of this writing), but they've allowed more runs than they've scored. The closest team to the Rangers current run differential to make the playoffs is the 2012 Orioles, who made the playoffs with a +7 runs scored. To look at it from another angle, the Rangers need to post a +33 run differential the rest of the season to match the Orioles run differential in 2012 (the Rangers have not posted a +33 run differential in any month this season). At this point, the Angels with a -10 run differential have a much better chance of making of the playoffs than the Rangers do, according to the Pythagorean records.

Pythagorean W-L record is hardly a fool proof stat, however, if the Rangers do make the playoffs between their current run differential and the +32 run differential needed to match the 2012 Orioles, they'll be making history.