Let’s Watch Kevin Pillar Play Defense

A good way to get smart people to mock you on the internet is to make a big deal about small sample defensive metrics. There are actually a lot things for which people on the internet will mock you, but that’s definitely one of them. For instance, imagine someone told you that Kevin Pillar is the best defender in baseball because his 2016 UZR/150 is 41.3! That assertion would earn some ridicule because a center fielder with a true talent 41.3 UZR/150 would essentially be some kind of baseball magnet positioned behind second base, and I’m given to understand those violate rules 3.09 and 3.10.

Just pointing to a defender’s metrics, be it UZR or DRS, isn’t enough to win an argument. Pillar has a reputation as a great defender, but no one honestly thinks he’s 40 runs better than the average center fielder. At best, that number suggests that Pillar has made great plays at an unsustainable rate, and at worst, it shows that the metrics are imperfectly designed tools that use imperfect data.

In 2015, Pillar accumulated 14 DRS in center (1,236 innings) and 8 DRS in left (120 innings) to go along with his 14.0 UZR in center and 1.5 UZR in left. Over his most recent full season, he was an elite defender; during the first 66 games of 2016 he’s been even better. Pillar currently owns 11 DRS and a 15.4 UZR (all in center). We all acknowledge those estimates are noisy and the precise run values needn’t be taken at face value, but the underlying message is probably right: Kevin Pillar is playing outstanding defense.

Instead of simply taking the metrics at their word, let’s use a Inside Edge fielding data to identify some of the more interesting plays Pillar has made (or not made) this year. Inside Edge carves up plays into six bins based on the probability the play being made: 0%, 1-10%, 10-40%, 40-60%, 60-90%, and 90-100%.

Below, I’ve presented two spray charts. On the left are all of Pillar’s made plays that fell into the categories at 60% or below. On the right are all of Pillar’s missed plays that fell into the categories at 10% or above.

IE map

As you can see, Pillar has made nine plays rate at the 60% level or below and missed just three at the 10% level or above. Right off the bat, that seems like a good ratio. In fact, there’s almost the hint of a trend in the spray charts. Pillar seems to be good going back and to his right and coming in to his left, but he’s “messed up” on some balls directly over his head.

We’ve heard from the metrics and we’ve dug into Inside Edge to isolate some of the toughest successes and easiest failures, but we can do more if we want to really appreciate Pillar as a defender. Below are the 12 plays found on the Inside Edge graph, in chronological order. I’ve provided the BABIP against on similar batted balls for 2016 based on Statcast exit velocity, launch angle, and horizontal direction. To do this, I grabbed BABIP on balls within three horizontal degrees in either direction, three degrees of vertical launch angle in either direction, and 3 mph of exit velocity in either direction. It’s not a terribly scientific method, but I wound up with no fewer than 19 similar balls for any one play, and far more in most cases.

April 4

Inside Edge: 10-40%

Expected BABIP: .500

What makes this play so impressive visually is that Pillar is starting from essentially straight-away center field and winds up deep in the left-center alley before diving near the wall. This is clearly a difficult play that most center fielders probably wouldn’t make, although if the wall is five feet closer or five feet farther, it probably looks less visually stunning because he would have crashed into it standing up or not at all.

April 17

Inside Edge: 40-60%

Expected BABIP: .263

This is a good play, but it’s not a spectacular play by any means. Pillar gets a good jump and executes a nifty slide at the end, but if he started at normal depth, the ball is caught standing up.

April 17

Inside Edge: 40-60%

Expected BABIP: .742

Pillar gets a fine jump, tracks it well, and times his leap appropriately, but this play doesn’t require that you cover that much ground. He was playing David Ortiz deep enough that this one probably looks a lot easier if the outfield were larger and he could run through the catch rather than make sure he didn’t smash into the wall.

April 18

Inside Edge: 40-60%

Expected BABIP: .500

Alright, now we’re talking. This definitely seems like the kind of play that should earn you some runs saved. The ball is not in the air very long and Pillar covers a ton of ground, dives, and saves extra bases.

April 26

Inside Edge: 40-60%

Expected BABIP: .647

This one looks similar to the previous play, but it’s actually hit significantly harder and at a lower angle. This ball is definitely going for extra bases if Pillar doesn’t make the play, and he takes the right angle to meet the ball before it splits the outfielders and heads up the gap.

May 1

Inside Edge: 40-60%

Expected BABIP: .695

Both Inside Edge and Statcast rate this as a difficult play, but it’s not the kind of catch that’s going to wind up on highlight reels at the same rate as some of the easier dives. Pillar has to track this ball over his shoulder, but he doesn’t crash into the fence or leave his feet. This is clearly a difficult play, but you can understand why it might not stand out.

May 7

Inside Edge: 10-40%

Expected BABIP: .603

This, on the other hand, is the kind of play you will see over and over again the next day. I imagine the expected batting average isn’t off the charts because a lot of similar balls are probably hit with the center fielder shading more toward left field, but from where Pillar starts, that’s a mighty fine grab and finish.

May 9

Inside Edge: 90-100%

Expected BABIP: .144

Anytime someone talented screws up, it’s either painfully uncomfortable or delightfully hilarious. This falls into the latter category. Pillar just gets completely turned around and botches this one big time. This is exactly the kind of play that makes defensive metrics so important relative to fielding percentage. That’s not an error but it’s an easy play that the fielder failed to convert.

May 11

Inside Edge: 10-40%

Expected BABIP: .615

I’ve always maintained that a line drive/fliner hit directly over the center fielder’s head is the hardest defensive play in baseball. This is a good illustration of why. You can see Pillar’s instinct is to track the ball and dive at the last moment, but he probably has a better chance of catching it if he stays on his feet and just sprints the whole way. Try telling your stupid brain not to dive here, though!

May 22

Inside Edge: 10-40%

Expected BABIP: .067

I have no idea why the expected BABIP on this play is so low. It looks like a tough play, Inside Edge thinks it’s a tough play, but guys just don’t get hits when they put the ball in play at this angle, velo, and direction. I went back and checked 2015’s numbers and found a similar effect. Best guess: some combination of wind and spin made this ball travel farther than you would expect. But this is a good example of how complicated defensive stats can be even with quality granular data. Anyone watching this could tell you it was a tough play, but the raw inputs of the batted ball would have you believe it’s easy.

May 31

Inside Edge: 1-10%

Expected BABIP: .441

Ah yes, another famous Pillar diving catch in medium right center.

June 6

Inside Edge: 40-60%

Expected BABIP: .448

This one is pretty weird, mostly because Pillar is playing Nick Castellanos way the heck in right center. He looks really silly getting burned by this one but this certainly appears to be an example of flawed starting position. Pillar doesn’t have much of a chance on this ball, but a center fielder at regular depth probably does.

***

This exercise reinforces Pillar’s greatness, as he makes a number of a very difficult plays and only really screws up one painfully easy one. But the exercise also demonstrates his specific strengths. He appears to come in toward right center well while excelling at covering ground deep into the left center field gap. None of his best plays are deep toward right or in toward left.

Additionally, I think this is a pretty clear demonstration of the difficulties in measuring defense. Pillar makes plenty of great plays, but I’m not entirely sure that I would rank their difficulty in the same order as Inside Edge or Statcast. Defensive metrics are pretty successful at identifying easy plays that the fielder missed without being charged with an error, but getting a complete picture of one’s defense requires a model that separates plays of relatively similar difficulties. I’m much less confident in our ability to distinguish between a 40% play and a 50% play than our ability to find all of the 90-100% plays.

Pretty much any reasonable method will find the really easy plays and the really hard ones, but the classification in the middle is very dependent on the assumptions of the model. That’s not a knock on DRS, UZR, or whatever metrics are coming in the Statcast era, it’s just a reality of trying to measure a process that has a lot of interdependent moving parts.

It’s possible that Pillar has already been worth 10-15 runs more than the average center fielder this year, but having watched these plays I would wager the true value is about half that. Pillar is an excellent center fielder and deserves all of the praise he receives for his efforts, but I wouldn’t bet on him blowing passed the single season UZR record.





Neil Weinberg is the Site Educator at FanGraphs and can be found writing enthusiastically about the Detroit Tigers at New English D. Follow and interact with him on Twitter @NeilWeinberg44.

14 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Damaso
7 years ago
Reply to  Jimmy Turner

yep. this.

grady
7 years ago
Reply to  Jimmy Turner

Yep, came to post the first two. Think this would be an even more interesting exercise with last year’s data as well. Defensive metrics are volatile enough with ~half a season’s worth of data… adding last year’s data to this year’s (the charts especially) would probably allow us to make some different hypotheses or judgements.

jianadaren
7 years ago
Reply to  Jimmy Turner

Agreed. By Neil’s own method, Pillar simply hasn’t had any difficult or tricky plays hit deep right or shallow left this season. They’ve all been either impossible or routine.