December 10, 2012
Resident Fantasy Genius
A Manager's Impact on Steals
Major League Baseball clubs often pride themselves on being aggressive on the basepaths. While the scope of this article isn’t to examine whether this strategy helps win ballgames, we can say with absolute certainty that being aggressive on the bases is the favored approach among fantasy owners. Stolen bases can be difficult to come by, so finding unexpected sources of steals can make a big difference for a team looking to win a championship.
One way to come across such bargains is to look at players who are changing teams, going from one with a passive organizational philosophy to one with an aggressive philosophy. Lots of teams claim to be aggressive, but as with most things in this game, you can’t always take people at their word or reputation.
The most your fantasy competition likely does is read articles like these or look at the teams that attempt the most steals, but that can be very misleading. After all, raw stolen-base totals can be influenced by the personnel the manager has at hand. The most passive manager in the world will still lead the league in steals if he has Michael Bourn, Mike Trout, and Ben Revere patrolling his outfield and Jose Reyes and Jose Altuve manning the middle of his infield. No, if we’re going to be making decisions based on manager philosophies, we need to be more scientific than that. To that end, today I’m going to be examining which teams and managers are actually aggressive and which tend to be more prudent with their baserunners.
My study uses a similar format as the With or Without You methodology Tom Tango made famous. I looked at all hitters who played under a particular manager in one season but under a different manager in either the previous or the following season. I then compared that player’s stolen-base attempt rate between the two years (while accounting for league average) to see how it changed when he was playing under a different manager. I then summed up all managers by year, with each hitter’s contribution to the manager’s bucket weighted by the lesser of his two stolen-base opportunities (with the manager and without the manager).
I ran some correlations to find the proper amount of regression and regressed each manager’s five-year total to come up with an initial estimation of his true talent level. After that, I started the process over again, but this time I used these initial estimates to adjust the “without manager” year. That is, a manager will get less credit for raising a player’s attempt rate if he played for a manager who didn’t like to run the year before. I ran several iterations of this process to come up with my final figures.