Regressions: 2008 MLB WHIP -> Wins

Thursday, February 05, 2009

It's February, so I'm Trying to Find things to Post On. Therefore You're Going to Get Some Simple Regression Tests from the 2008 MLB Season.


The Major High Outlier: Texas with an WHIP of 1.58 an 79 Wins
The Low Outlier: Toronto with an WHIP of 1.24 and 86 Wins

Formula: WHIP = 1.837 + (-0.005525) * Wins
ReFangled Formula: Wins = (WHIP - 1.837) / (-0.005525)

Fun Math Time

Say I want the Yankees to Win a 100 games next year, what would their WHIP likely need to be around? 1.837 + (-0.005525)*100 = 1.28 WHIP. And say I think the Nationals are going to stink up the joint with their horrile pitching and throw up say a 1.56 WHIP. (1.56-1.837) / (-0.005525) = 50 Wins.

Posted by Simon at 3:33 PM   Digg! submit to reddit BallHype: hype it up!

0 comments:

Post a Comment

Advertise Here!