clock menu more-arrow no yes mobile

Filed under:

Pythagorean Wins and "Luck"

The Astros' Main Man,

I will of course be posting the by-now eagerly anticipated Pundit Count webtacular tomorrow, but as I was checking out Baseball Musings before (finally) hitting the sack this evening, I came across something I just had to share, if only because I am so frequently confused by stuff like this, and had no idea what to make of it.

Seems like the White Sox significantly overperformed for the year when compared with what the Pythagorean formulae were showing.  In fact, the difference between the 99 games they won, and the 92.2 Pythagoras predicted, is the sixth highest differential among World Series participants in history.

And of the "top" seven, only one team--the 1930 Philadelphia Athletics led by Connie Mack, and anchored by Jimmie Foxx and Lefty Grove--managed to actually take their Series.

I'm not a big fan of Pythagorean wins, but the soldiers I enlist in the war to promote the Astro viewpoint don't always get the background check, if you know what I mean.  

These days, if it suggests that the Astros might take the Series, I'm running with it, regardless of the pedigree. Like here with this Pythagorean stuff, or if Eric Karabell, for another good example,  suddenly came aboard and picked the 'Stros in five.

Really, I might be more inclined to suggest that the ability to get more wins out of fewer runs is a positive trait for a ballclub to have, but the hardcore sabermetricians insist that a team's win total deviating very much from its projected wins total is basically ascribable to luck and luck alone.

Using the simplest ra2 + rs2 formula that Pinto uses at the Musings factoid, the Astros finished with a .564 first order pythagorean winning percentage, which works out to 91.4 wins, and which they actually underperformed by 2.4 games.

I guess I'm to take it that this underperformance reflects well for the Astros' chances in the upcoming World Series.

Houston wasn't as good as the White Sox, but they weren't as good more legitimitely.

I think.