Miguel Sano, Defensive Superstar

I’m going to be up front with you: that headline is seriously misleading. Based on what we can tell, Miguel Sano is probably not a good defensive player. Last year, though, to try and get the club’s offense going, the Twins decided to try Miguel Sano in right field. Regret kicked in pretty quickly.

He played right field most days for the first couple of months of the season, until a strained hamstring put him on the disabled list for the entire month of June. During the month that the team didn’t have to watch Sano chase balls around the outfield, they decided that they didn’t really want to see that ever again. When he returned to the team at the beginning of July, he was promptly moved back to third base. He split his time between there and DH over the rest of the year.

As a pretty large human being, Sano certainly doesn’t look like an outfielder, and while we only saw him out there for a little over 300 innings, the early returns weren’t particularly positive.

Those regular Minnesota viewers who filled out the Fans Scouting Report were harsh in their assessments of his defensive abilities, giving him a 33 overall ranking (out of 100), with arm strength representing his only above-average tool. UZR and DRS both hung negative ratings on his outfield work, as well, with DRS grading his small-sample performance like he was one of the worst outfielders in baseball, giving him a -8 rating despite only a quarter of a season of playing time there.

Of course, one of the problems with metrics like UZR and DRS is that there are only a small number of plays by which to judge a player’s quality, so larger samples are needed to start separating the signal from the noise. This is one of the primary reasons everyone is so excited about the potential value of Statcast data: the ability to accurately measure starting position and distance traveled while the ball is in flight should allow us to better understand which defenders really were making difficult plays that saved runs for their team.

And this year, MLB is making it’s Catch Probability metric public, using hang time and distance traveled to estimate the difficulty of every ball hit to an outfielder. Over the weekend, Baseball Savant rolled out a Catch Probability Leaderboard, so we can look at the different rates of catches by difficulty for players in comparison to their peers.

In general, the results align with what we’d expect based on what we think we know about defensive outfield quality right now. The three outfielders who made the most “five star” catches — plays with a catch probability below 25% — in 2016 were Ender Inciarte, Adam Eaton, and Billy Hamilton. Among regular outfielders, no player turned a higher percentage of difficult plays into outs than Hamilton, with Kevin Kiermaier not too far behind. On the other end of the spectrum, Mark Trumbo had the lowest percentage of “one star” catches, converting only 58% of the balls hit into the area where the metric thinks the catch probability is between 91-95%.

Hamilton good, Trumbo bad. So far, nothing shocking here.

But if you click around on the sortable leaderboard enough, you’ll find a few surprises. For instance, if you look at the highest rate of five-star catches without any kind of minimum number of opportunities, Hamilton actually only ranks third. The guy right in front of him, Mac Williamson, was graded as an above-average defender by the fans and rated really well by both UZR and DRS, so while scouting reports haven’t called him an elite defender, at least the small-sample data we did have suggested that he made some nice plays last year.

The guy in front of Williamson? Yeah, the misleading headline gives it away. Miguel Sano had three balls hit to him in right field that were considered “five star” opportunities, and he caught two of them, good for a 67% success rate, the best of any outfielder in baseball in 2016.

So, I was intrigued. This is exactly the kind of play that shouldn’t be all that easy to make just by randomness. Since the catch probability variables are simply hang time and distance traveled — spray direction will be added in the not too distant future, to make it a three-variable model — there shouldn’t be many plays where a guy who can’t cover any ground gets credit for a spectacular play simply because he was positioned well, which was always one of the primary concerns with metrics like UZR and DRS. Going by time and distance, you’d think that this may be a measure of just something a guy can or can’t do.

And mostly, the leaderboards support that idea. The guys we are pretty sure are lousy defensive outfielders just didn’t make any plays like this last year. Matt Kemp had 37 five-star opportunities and failed to turn any of them into outs. Melky Cabrera was 0 for 34, Jayson Werth 0 for 33, Carlos Gonzalez 0 for 30, and Yasmany Tomas 0 for 24. J.D. Martinez, who recorded the worst UZR of any corner outfielder last year, went 0 for 22. Trumbo was 0 for 14 when the Orioles were willing to let him play the outfield.

But Sano, who was deemed bad enough defensively to have the plug pulled on the experiment after just two months, made two plays that Statcast deemed to be the highest level of difficulty possible, in just three attempts. With my curiosity piqued, I asked Mike Petriello if he could point me to those two plays, and he was kind enough to do just that. So let’s watch Miguel Sano’s two spectacular catches together.

I wonder what Sano’s Fans Scouting Report rating would have been like if it were filled out only by Tigers fans? A couple of weeks apart, they saw him rob Miguel Cabrera and then Victor Martinez on balls that looked like sure hits off the bat.

Now, we should note that these plays, and especially the second one, aren’t necessarily what people have in mind when they think of incredible defensive plays by an outfielder. As you can see from Sano’s Statcast fielding chart, he didn’t have to run more than 40 feet to make either play, and both were close to the upper range of what constitutes a “five star” catch.

But perhaps it should be somewhat encouraging to the Twins that Sano made two plays in the outfield that required a good read of the ball off the bat, a pretty quick first step, and a coordinated dive to get a ball that is just about to hit the ground. While the Twins decided they’d seen enough of Sano in the outfield to call that experiment a failure, these two plays show an interesting level of athleticism in the types of skills that you would think may translate well to third base.

Range matters everywhere, but conventional wisdom suggests that, at third base, it’s more about footwork, reaction time, and hands than pure speed or ability to cover ground. And while the fans were down on Sano in each of those areas in the FSR, these two plays show some level of ability that the other OFs-who-should-be-DHs just didn’t display at any point last year. And, for whatever it’s worth, in 450 career innings at third base, UZR has rated the range portion of his fielding ability at +8, though he’s given back most of that by making 15 errors in 51 games. Which is, you know, atrocious.

As the Statcast guys have noted, we’re in the first or second inning of Statcast’s existence. At this point, there’s still way more we don’t know than that we do know. In reality, we don’t really know what Sano making those two plays tells us about his defensive talent level, or if it tells us anything at all. In a few years, we may look back at this post as a great example of how we were fumbling around the dark. I certainly wouldn’t say that, because Sano made two diving catches in the outfield, on balls hit fairly close to him, that he’s a better defender than he’s generally been given credit for.

But this is the kind of thing that maybe, just maybe, Statcast can do really well. Most guys who just can’t field didn’t make any plays like this last year. That Miguel Sano made two of them, in just three opportunities, might be a clue that there’s a bit more athleticism in there than we might expect from a guy who certainly looks like a DH.

Sano is probably always going to be a bat-first player, but his profile looks a lot better if he’s a reasonable defensive third baseman. He’ll have to fix the error problems, but perhaps the Twins pulling the plug on the outfield experiment was more because his abilities could be best utilized at third base, and not because they’re running out of places to hide him.