Kicking around the idea of regression today brought me back to an excellent read from one of the best statistical writers out there, Russell Carlton. The idea of regressing baseball data is far from new, but when someone does this much legwork you just have to put it to use, right? With that in mind I wanted to take regress and show how the Rays batters would be affected by regression over the past season. We could entire careers to get a better idea of the player’s true talent, but let’s just look at this past season.
Anything worth doing is worth doing right so I created a calculator that you can play around with that regresses Russell’s key statistics to the 2013 American League average for all non-pitchers. It’s probably easiest just to download your own copy under the files tab, but if you don’t want to do that just right-click the “Calculator” tab and click on “Duplicate”. From here just enter a players name in the yellow cell and the calculator will do everything else.
Using this I was able to compile the actual and regressed statistics for each player on the 2013 Tampa Bay Rays:
There’s a lot going on here, but I think it does a good job showing how affected the batters with few PA/AB/BIP/FB are by regression and how even guy closing in on 700 PA are affected somewhat. Keep in mind these are not projections this is merely a look at what these guys seasons would have looked like if properly regressed. Probably not all that useful, but I think this makes a nice reference to look at and the calculator could/should be a solid addition.