I’m not a big baseball fan, but I am an analytics fan, so I’m curious to see how the concept is portrayed in the movie Moneyball, set to hit theatres on Sept. 23. The movie is based on the book of the same name by Michael Lewis. It tells the story of Billy Beane (played by Brad Pitt), the general manager of the Oakland A’s in the late 1990s, who needed to field a winning team on a meagre payroll.
You might already know about this analytics success story. Beane and his assistant general manager Paul DePodesta had an idea: There had to be champion players out there whom everyone else had overlooked; if no one was looking for them, they would be affordable. Their prospecting tool would be statistical analysis.
For example, their analysis showed that new stats such as on-base percentage and slugging percentage were better predictors of offensive success than the tried-and-true qualities valued by traditional buyers of talent. The analytical approach paid off, and it was soon copied by other major league baseball teams.
I don’t know if the movie will have much to do with analytics, but the trailer (click here to watch on YouTube), does seem to portray a geeky practitioner trying sell crusty, tradition-bound decision makers on the value of his craft. I like that story.
Better than that, though, is a line spoken by the fictional character Peter Brand (based on DePodesta and played by Jonah Hill): “Your goal shouldn’t be to buy players,” Brand says to Beane. “Your goal should be to buy wins. In order to buy wins, you need to buy runs.”
It’s a statement about success with statistical modeling: You have to start by properly framing the question. The rest is just technique.
The analytical approach doesn’t pay off for professional sports like it used to, because now everyone does it. (The Oakland A’s are not doing very well these days.) But analytics will always pay off in the nonprofit world because in the hunt for potential donors we still mostly compete against ourselves — we either identify our friends and supporters effectively, or we don’t.