How new defensive stats change the way we see the Yankees

Let’s talk about the infield again! If there was one weakness on a 103-win Yankees team, it was infield defense. By all the metrics we have, Gio Urshela was merely fine; Gleyber Torres wasn’t, especially at second base; Didi Gregorius took a step back; and Luke Voit was a DH. Only DJ LeMahieu had a really standout year defensively, and that’s one of the reasons the Yankees signed him anyway.

As we look toward 2020, it’s fair to question the infield defense again. Torres seems destined to be the Opening Day shortstop, LeMahieu at second, and Voit at first. One of the big debates this winter has centered around Urshela or Miguel Andujar at the hot corner, with points made on both sides regarding Andujar’s offensive upside and Urshela’s higher defensive floor.

Now we have a new statistic to evaluate infield defense, Infield Outs Above Average from Statcast. Tom Tango has all the math you need, and Mike Petrriello has a great summary of what the metric means, but let’s talk about it specifically in a Yankees context. What I like a lot about the OAA metric is it reflects, roughly, how my gut feels about the defensive prowess of the team:

OAA seems to track fairly closely to the way teams evaluate defense internally: it accounts for a defender’s positioning and the speed of the runner, for example, using data that UZR or DRS doesn’t have access to. With that alone we see why there wasn’t much of an appetite for bringing back Gregorius; his weak bat combined with poor defense, whether Tommy John-related or otherwise, just made him too risky a pick for a team that may well be in a World Series or bust mode in 2020.

The other particular point, brought up in Petriello’s article, is breaking down where Torres is actually playing. He has a 0 OAA rating when he’s standing where shortstops traditionally stand. As shifts have become ubiquitous, especially for teams as data-driven as the Yankees, positions mean less and less, and need to be reflected in the defensive stats we use.

One of the big questions around Torres’ defense has been how the split between second and short has affected his play. He’s been a shortstop his entire life until being called up to the majors, and there’s a good deal of debate about how much time and how many reps a player needs at second, where you cover the bag more frequently, spend more time in short right field, and have different footwork turning double plays than you do at shortstop, even if the latter position needs greater range overall.

When he’s playing where shortstops play, Torres is just fine, exactly average. In context, and I am being very careful here, Derek Jeter was a great hitter and a poor defensive SS, while it looks like Torres is a great hitter and a mediocre defensive SS. However, there’s a flip side to that argument, which is that Torres probably won’t play where shortstops traditionally play, again because of the data the Yankees use.

Out of 207 defensive opportunities at “shortstop”, 160 came when Torres wasn’t standing in what we traditionally consider the shortstop spot. We know that he’s below average the farther he moves from traditional shortstop, and he’s going to have to get better, or else risk being a defensive liability at the most important position.

If there’s one thing I’d like to see from a new version of OAA, it’s how much players depend on their teammates to cover ground for them. Again, using our Torres example, it’s probably easier for Torres to play shortstop when he has LeMahieu at second. The Machine is a bona fide good defender, especially at the keystone, and as of yet there isn’t really an adjustment in the metrics for that. Other players can compensate for the relative weakness of their teammates; think of a great scoop at first base saving a throwing error.

I’m not sure whether Outs Above Average actually settles any of the infield questions for the Yankees. Torres will be given every chance to succeed at short, Andujar was one of the worst infielders in baseball in 2018, and so on. It does give us a new, enlightening tool though, and best reflects the increasing positional flexibility of infielders across baseball,