And it's root, root, root for the vector!

With snow on the ground and icy rain falling outside, it can mean only one thing: Opening Day for the Boston Red Sox, the unofficial MLB team of the MIT Press. Before you ask: No, we don't rigorously police the baseball preferences of our employees - we even have Yankee fans on our staff. Or, we had one, at one point...he left a while ago. And my boss is - get this - a Mets fan. But I digress.

As of this writing, first pitch is just an hour or so away - in Texas, thankfully. So we thought we'd bring you some news that has absolutely nothing to do with the Press but does relate to our beloved home institution. Dimitris Bertsimas, an operations research professor at MIT's Sloan School of Management, recently published a paper called "The Analytics Edge in Baseball." He and a doctoral student crunched the numbers and predicted that the Sox would win - drumroll please - 101 games.

"A player is a vector of numbers and from that, we can forecast overall team statistics," writes Bertsimas,who had the honesty to admit to being a Red Sox fan.

In fact, Bertsimas and his coauthor, Allison O'Hair, ran the numbers for the entire American League East division, predicting that the Tampa Bay Rays would be close on with 100 wins and the Yankees a mere 93 (please, let it be so).

In case you think that all this number crunching is bunk, consider that last year he picked Boston to win 90 games - final total was 89.

So with regression analysis and invariate parametric correlation in mind...play ball!

 

PS - Okay, just one bit of MITP-related info. If you want to know the origins of all this nerdy statistical-analysis-of-baseball stuff, you could check out its founding text: Percentage Baseball by Earnshaw Cook, proudly published by us in 1966 - the year in which the Orioles won the World Series in four straight and the Red Sox won a mere 72 games. Take heart, the Yankees won only 70 and came in last. But again, I digress.