Predictive Value of Spring Training

Are Spring Training stats an accurate predictor of regular season performance?

As one of the new kids on the block (and if you’re thinking of that music group, save yourself and stop it), I’d like to say hello and hope you enjoy what you are about to read from me in the coming weeks and months. And now, on to the post…..

A few days ago I ran across this little post on Beyond the Boxscore, and it has me thinking about what predictive value spring training may have. I know, I know, the traditional logic says that there’s no way that you can predict anything from spring training. And I haven’t explored the data (in part because I lack it at the moment to do the proper analyses) to debunk or affirm that notion. However, I can’t help but think that there has to be something to these numbers of the spring.

Again, I haven’t done any statistical analysis for this, but think about what kind of decisions are usually decided based on spring training. How valuable or accurate would it be to determine who the Nationals’ primary first baseman should be for this year based on Nick Johnson’s and Dmitri Young’s performance in camp? Or if a kid like Boone Logan should have even been kept on the 2006 White Sox roster despite a great performance in the spring? Or if a fantasy manager should adjust a draft strategy based on performance in camp if it could be an indicator of a better or worse than expected season?

My point is that these are decisions that the baseball people make that the members of the sabermetric community who don’t work in major league offices have yet to broach (as far as I know, and some days that isn’t much). Some of it may simply be data availability. Some of it may be a sample size issue that easily invalidates any results of such an inquiry. Some of it may be the issue of Major League Equivalences being impossible or extremely difficult to calculate. Heck, someone may have even explored the subject already and I just haven’t read their work yet.

If you have any pointers on articles to read, other questions that could be answered, links to some other research, or just think this idea is a dead end analysis waiting to happen, post a comment. I’m going to see what I can dig up, and will hopefully come back with a better analysis once I have some data to work with.