Discussion Board

http://www.economist.com/blogs/gametheory/2015/03/baseball-statistics

In the article above, D.R. explains of the widely-held believe that Spring Training Baseball statistics mean absolutely nothing.  However, he challenges this using lesser known statistics and the fact that there are some ways, albeit somewhat misunderstood, to predict growing trends in players as well as in the sport itself.  He uses data that is seldom seen to show that there is some correlation between spring training numbers and "real game" numbers.  The fact that strikeouts per at-bat increased each year from 2006 to the present can easily be seen during the season, but the fact that the same trend has occurred during spring training should have been noticed by at least somebody, right?  The continually belief that spring training can't "predict" the season is a form of ignorance on the part of many in the game.  Some very important stats that people don't often think about, such as a hitter's strikeout or walk rate, tell a story much different than the commonly held beliefs of the "great baseball minds" of today.  The correlation between many stats are too similar to simply be coincidental, right?

A baseball stats forecasting system called ZiPS has attempted to utilize statistics in order to project and predict future statistics for ballplayers and ball clubs alike.  The fact that adding in spring training statistics into an accurate system such as ZiPS allowed the ZiPS system to become even more accurate has to mean something, right?  Well, looking at the numbers on his slides, the spring training stats only count slightly, but make a pretty large difference when it comes to the forecasting of statistics for the season ahead.  Many of the more accurate prediction are for younger players with little experience in the major leagues.  This happens because there isn't much minor league data to use for the model, but at least with spring training, there is abundant statistical data that can be used.  This isn't to say that spring training stats are an all-telling statistic that should be looked at for all decision making, but it does show that with the right utilization of knowledge, predictions can become more accurate.

Still, however, people disregard spring training stats because there haven't been enough success stories to actually have baseball managers and general managers take risks on young players having good springs or veterans "finding the groove" again.  One of the most important things to keep in mind when forecasting is something Donald Rumsfield has said: "Absence of evidence is not evidence of absence."  This statement is important in all forms of predication and forecasting, whether it be a sport such as baseball, or a social science such as economics.

E-mail me when people leave their comments –

You need to be a member of UNH Economics Collective to add comments!

Join UNH Economics Collective