Rubbing Mud: The Rise of the Rover and the Fall of Dozier

The Angels and Twins played a four-game series over the weekend. It was a matchup of two teams with plenty in common—low preseason expectations, a good positional corps somewhat wasted by too-thin pitching, and yet surprising (if modest) early-season success. In one small and strange way, however, it was also a meeting of two teams at opposite ends of a philosophical spectrum.

With left-handed batters at the plate, the Angels’ second basemen play deeper (on average) than those of all but one other team (the Mariners). The Twins’ second basemen (we’re talking mostly about Brian Dozier here) play the shallowest against lefties.

Average Starting Distance for Second Basemen, Against Left-Handed Batters, 2017

This data is as wide-lens as it gets. I’m not going to try to take shift frequency or shift type out of the equation, nor to balance everything for runners on base or outs in the inning. Some of that would be possible, but not all of it would be desirable. A shift, for instance, is made up on individual changes in defensive positioning. Even if it’s as part of a shift, I want to know if a given fielder is playing deep against a certain hitter or not.

Nor should you assume that this data is shaped solely by shift differences. For instance, the Angels played their second basemen well onto the outfield grass against Joe Mauer and Max Kepler during the series, even though the left side of their infield largely stayed put, and even though Kepler is fairly fleet of foot. Dozier, by contrast, played even Kole Calhoun a few feet shy of the grass on some plays.

Of course, we’ve never really had this data before, and we don’t have an easy way to contextualize it. That difference—a 15-foot average disparity in depth—surely makes some impact. But how much? To find out, I ran a search for all balls hit by left-handed batters at a launch angle of seven degrees or less, and with a spray angle filter that essentially covered the typical range of a second baseman’s responsibility.

Then I sorted the 30 teams by the difference between the expected batting average (xBA, a metric devised by Tom Tango and the rest of the public-facing Statcast team at MLB Advanced Media, which uses exit velocity and launch angle to estimate the expectation on a given batted ball) of all the balls they allowed that fit those criteria, and the actual batting average they allowed on those balls. A positive number means teams prevented hits, relative to xBA. A negative one signifies that they’re allowing more hits than expected, given the quality of contact.

The Mariners rank third on that leaderboard, with an xBA-BA of .130 (.280-.150). The Angels rank ninth, at .100 (.273-.173). The Twins rank 29th, at -.001 (.289-.290). It sure seems like allowing Dozier to play so shallow a second base is costing Minnesota outs.

A caveat or two should be noted here. First of all, xBA is not perfect for this purpose. It’s not direction-adjusted, which means that whatever the xBA on a given sharp grounder to second base might be, it’s probably much higher if the ball is hit right up the middle than if it’s hit right to the second baseman’s typical position. This search doesn’t capture that. (The Twins still rank second-worst in raw batting average allowed on the hits in question, and the Mariners and Angels still rate above average.) Nor does it tell us who made each play; a shift might have put the shortstop or third baseman in front of some balls, making this a deeply imperfect measurement of the second baseman’s ability to cut them off.

In the big picture, though, there’s still a clear Thing to see here. Both the Mariners and the Angels are positioning their second basemen deeper against lefties than any team did in either 2015 or 2016. They each seem to have realized that the physical symmetry of the infield is a lie, and that if you have any reason to expect a hard ground ball to second base, you’re better off having a guy playing deeper there. The throw from even shallow right field to first base is shorter than the throw from deep short to first. The progression is natural, once one embraces the idea that shifting itself first introduced to the game: the notion that defensive positions need not be as they have always been.

It’s easy now, in fact, to imagine a world in which the second baseman (as we have known him thus far in baseball history) ceases to exist. When I played coach-pitch baseball in the summer of 1998, we had a second baseman, and we had three outfielders, but we also had a 10th guy, and he was neither outfielder nor infielder. We called the position rover, and his job was to get to balls that were hit too hard for the inexpert reflexes and underdeveloped athleticism of 9-year-old infielders to handle. Most of the time, it was too late to turn such balls into outs, but he could at least hold the batter to a single. (He also served as a much-needed cutoff man on throws in from the outfield, which is why our team had maybe the strongest thrower at rover most of the time.) Perhaps this is the future of the second baseman, only he won’t be backing up anyone.

He’ll be a second shortstop, playing up the middle against right-handed pull hitters, drifting well out into shallow right field against lefties, and being ready to go out and get shallow flies (balls the outfielders, playing ever deeper in response to the ever-stronger and more power-hungry hitters, can't get) no matter where he starts. Slightly insufficient athletes have been hidden at second base for a long time, just as they were hidden in the corner outfield spots for a long time. The future of run prevention is such that there will be no hiding bad athletes, and no shortage of positions at which to try out good ones.

Instead of priding themselves on playing shallow and taking away cheap hits, second basemen will become focused on taking away hard hits—and, eventually, experts at charging softer grounders, the way a third baseman charges a bunt. When it comes to defensive alignment, teams are still unearthing new ways to turn convention on its head and get outs where there once were hits.