Thursday, March 6, 2014

MIT Study: MLB Skippers Flunk Pitcher Choices

Where Was MIT When Really Needed? (NY Daily News)
A new study by computer scientists at the Massachusetts Institute Of Technology suggests that MLB managers usually make poor decisions when yanking or staying with pitchers. The computer model designed by MIT scientist John Guttag (who is, by the way, a Yankee fan) shows that managers made the wrong decisions about 48% of the time—mostly by leaving pitchers in. Of course, the opposite situation transpired when Tiger manager Jim Leyland yanked Max Scherzer in Game Two of the 2013 ALCS at Fenway. The Detroit bullpen blew the lead and the series went back to Detroit tied at one game each. Guttag's model would have told Leyland to stick with Scherzer. The scientific paper "A Data-driven Method for In-game Decision Making in MLB" is a geeky sabermatrician's dream, which uses Pitcher’s Total Bases (PTB) as the key measure of hurler effectiveness. The database of 21,538 innings resulted in definite patterns emerging—which conceivably could guide future managers. As Guttag put it, "Clearly the most important decision a manager makes is changing pitchers. I think there’s definitely room for improvement." You think?