Obligatory Moneyball Reference

Yes, many folks in the analytics space flock to “The Moneyball Reference”.  It’s a great example of how data analysis seeped into the mainstream using a powerful vehicle known as Brad Pitt.  The usual reference points out that Billy Beane’s analytical approach to players and statistics was counter to the decades-long logic of the established way of thinking.  Furthermore, that logic led to improved success while spending fewer dollars.  Call it the anti-Yankee approach.  As an analyst and a lifelong baseball fan, though, there’s a more nuanced takeaway from Moneyball that I like to reference.  That point is that the Moneyball approach is predicated on knowing the rules of the game.  In baseball, the team with the most runs wins.  Period.  Here’s a (crude) video clip of the scene from the movie that refers to this concept.  For the purists, here’s the passage from page 127 of the book:

“Paul DePodesta had been hired by Billy Beane before the 1999 season, but well before that he had studied the question of why teams win. Not long after he’d graduated from Harvard, in the mid-nineties, he’d plugged the statistics of every baseball team from the twentieth century into an equation and tested which of them correlated most closely with winning percentage. He’d found only two, both offensive statistics, inextricably linked to baseball success: on-base percentage and slugging percentage. Everything else was far less important.”

Moneyball: The Art of Winning an Unfair Game by Michael Lewis, 2003

So why am I pointing out something relatively obvious…that he who scores the most runs in baseball wins the game?  It’s because in other industries, the rules aren’t always so clear and succinct.  I’ve done a fair bit of work in higher education, and the rules are absolutely not clear when it comes to colleges and universities.  What is “winning”?  Is it graduating?  If so, graduating in how many years?  Is it getting a good job after college?  If so, what are the parameters of “good”?  Is it learning?  If so, pull up a chair, pour yourself a drink, and let’s have a lengthy discussion about measuring learning.

The point here is that since the rules aren’t always clear, one can’t apply a rote statistic like OPS to definitively determine student success.  Now that doesn’t mean analytics don’t work in education.  On the contrary, a good analytical approach can give the institution wonderful insight.  It’s just that the institution needs to take a step-by-step approach – ask a question, answer it, implement the results, and then ask the next logical question.  This incremental approach to effective analytics is a more reasonable way to think about things in education as opposed to a binary winning/losing viewpoint.  Now excuse me while I go calculate Miguel Cabrera’s WAR just for fun.