Regressions: 2008 WHIP -> ERA

Monday, February 09, 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 1.51 and an ERA of 4.73
The Low Outlier: San Francisco with an WHIP of 1.29 and an ERA of 3.98

Formula: WHIP = 0.574305 + (0.189032) * ERA
ReFangled Formula: Wins = (WHIP - 1.837) / (-0.005525)

Fun Math Time

If you look at the chart you'll notice just how tightly coupled WHIP and ERA are. One is a very solid predictor of the other. So if you see your team has a dot below the line than you probably got a little unlucky and vice versa if the dot is above the line. So say I want the Yanks to have an ERA of 3.50 this year, what will their WHIP need to be? 1.235. On the opposite end of the spectrum, what if I think the Rangers are going to have a WHIP of 1.50, what will their ERA likely be? 4.90.

Posted by Simon at 2:30 PM   Digg! submit to reddit BallHype: hype it up!


Post a Comment

Advertise Here!