I came to baseball later in life. As a kid growing up in Greece, I played soccer, which is very popular in Europe. I then got interested in American football, and basketball, and I also followed track and field. Even after living years in Boston – a rabid Red Sox town – I never gave baseball much of a thought.
Then, a while back, a friend gave me a copy of the book Money Ball: The Art of Winning an Unfair Game, because he knew I was interested in analytics. (You see, I have a passion for sports, but my greater passion is analytics.) The book, by Michael Lewis, is the story of how Billy Beane – a former player who later became general manager of the Oakland A’s – used an analytical, sabermetric approach to create a winning team. The A’s were the first team to heavily depend on quantitative methods, and at the time – we’re talking 2002 here – many considered Beane’s approach very radical, even foolish. But he made believers out of his skeptics. Even with the third smallest payroll in major league baseball, the A’s were able to use quantitative methods to gain an edge.
Now I am a big baseball fan. It’s not so much the sport that captivates me; it’s the math. More than any other sport, baseball lends itself perfectly to quantitative analysis. There are 162 regular season games in baseball, compared to 82 in basketball and 16 in football. This amounts to a larger sample size for statistical analysis. Plus, the ratio of the seven largest payrolls to the seven smallest in major league baseball is 4 to 1, compared to 1.75 to 1 in the NBA and 1.5 to 1 in the NFL. This gives more incentive for general managers of less wealthy teams to get great players at lower prices. All teams have the same goal: make the playoffs. The wish to make the playoffs is due to the extra revenue gained from more ticket sales, an increased fan base, and more media attention. However, for the less rich teams to even think about gaining more revenue, they have to get to the playoffs on a much smaller budget. And that’s not easy to do.
The use of quantitative methods in baseball also provides a great learning opportunity for my MBA students. My new paper* with Allison O’ Hair includes a case study where I ask them to imagine they are assistants to Theo Epstein, the 37-year-old general manager of the Red Sox who already has two World Series rings on his hand. Epstein, who has spent countless hours evaluating all of the Sox players, must decide which players to renew or instead offer a contract to one of the players that is on the market from another team. There are 25 players on the roster, and 7 of them will become free agents this year, meaning that they can accept a contract from any team. But after that, he has an even bigger decision to make: how much will he offer them? He wants to be competitive with the other offers they will get, but he also has to keep in mind that he has to remain within his set budget.
The assignment asks students to propose a plan about which players to offer contracts to, and how much to offer them. I want to see their decision-making based on statistical reasoning.
My goal is to get students to think about teambuilding in a different way. Many of them will graduate to become upper-level executives at companies, and at various points in their career, they will be tasked with building a team. Most often they look at qualitative attributes, but I’m trying to get them to look at quantitative facts, and to make trade-offs. I don’t want them to ignore qualitative factors. Team chemistry matters, but how do you model that? You can try but has more randomness to it.
* The Analytics Edge in Baseball; Sloan School and Operations Research Center, Massachusetts Institute of Technology; Dimitris Bertsimas and Allison O’Hair
A winning season: Professor Bertsimas uses quantitative analytics to predict the Red Sox will win 101 games this season to the Yankees’ 93: http://www.reuters.com/article/2011/04/01/us-baseball-redsox-mit-idUSTRE7301OR20110401