Another Way to Tell a Hitter’s Having a Good Year
Near the end of the game between the Blue Jays and Rays yesterday, the camera panned to center field. Evan Longoria was at the plate, and the Jays broadcast team was talking about the third baseman’s power. “He’s got some power to right field, too, now, and I think that’s why you’ll see the outfielders, especially the center fielder and right fielder playing a couple steps back,” said Dan Shulman. “Look at how deep Kevin Pillar is in center field. That’s only a couple of steps, it seems like, for Pillar, from the warning track!” he continued. “We have not seen Kevin Pillar play that deep,” concurred Buck Martinez.
It was impressive. That little dot in center is Pillar. Looks like a wallflower at a middle-school dance.
He was 361 feet from the plate at that moment. It makes sense, given Longoria’s spray chart this year. You’ll notice that Pillar is shaded a little bit to right, which is where Longoria hits many of his deep outs.
Source: FanGraphs
But the Blue Jays were pushing the envelope a bit. Call it situational defense, maybe, because Pillar was playing more than 30 feet further back than the average center fielder against Longoria this year. Given that there were two outs in the eighth inning of a tie game and Brad Miller and Nick Franklin were scheduled to hit behind Evan Longoria, there’s a certain amount of making sure to stop the big hit doesn’t sink the team. In a league where it probably pays to play deep, this was playing just a bit deeper on a guy who hits them deep.
The Rays’ third baseman is actually in the top tier of the league when it comes to how deep the center fielder plays, even if this was further than usual. Here are the top-15 average depths among hitters that have seen 1000 pitches or more.
Name | Count | Average CF Distance |
Nolan Arenado | 2411 | 329.0 |
Carlos Gonzalez | 2041 | 327.9 |
Edwin Encarnacion | 2544 | 327.0 |
Mark Reynolds | 1733 | 326.8 |
Trevor Story | 1753 | 326.4 |
Jose Bautista | 1797 | 326.2 |
Josh Donaldson | 2620 | 326.0 |
Giancarlo Stanton | 1752 | 325.5 |
Evan Longoria | 2199 | 325.4 |
Steven Souza Jr. | 1674 | 325.1 |
Steve Pearce | 1172 | 324.9 |
Chris Davis | 2381 | 324.7 |
Troy Tulowitzki | 1796 | 324.6 |
Miguel Cabrera | 2199 | 324.6 |
Ryan Raburn | 1009 | 324.6 |
Maybe that’s the best takeaway from this iconic moment: Longoria is being played deeper, because, you know, he’s hitting the ball deeper. Like most of the guys on this list, he’s slugging. He’s outperformed his projected slugging percentage by almost 90 points.
We’ve used stats like these to reverse-engineer an understanding of the type of data that teams have. We know they have better data, and they move quicker when it comes to adjustment, so we’ve looked at things like fastball distance from the center of the zone to identify slugging breakouts. Could this one also tell us which power breakouts teams deem sustainable?
Here’s a new way of looking at the center-field depth stat. I’ve taken the hitters with the biggest difference between their actual slugging percentage and their projected one, and converted the average distance into a rank stat. Look how quickly the “believable” slugging outbreaks come into focus.
Name | Count | Average CF Distance | Distance Rank | Proj SLG | Actual SLG | Diff |
Sandy Leon | 1016 | 316.6 | 146 | 0.318 | 0.553 | 0.235 |
Tyler Naquin | 1337 | 313.1 | 216 | 0.374 | 0.548 | 0.174 |
Ryan Schimpf | 1231 | 317.2 | 135 | 0.398 | 0.567 | 0.169 |
Brian Dozier | 2531 | 316.4 | 151 | 0.405 | 0.574 | 0.169 |
Daniel Murphy | 1997 | 315.2 | 182 | 0.433 | 0.598 | 0.165 |
Trevor Story | 1753 | 326.4 | 5 | 0.422 | 0.567 | 0.145 |
Aledmys Diaz | 1556 | 317.2 | 136 | 0.382 | 0.517 | 0.135 |
David Ortiz | 2146 | 324.2 | 16 | 0.498 | 0.629 | 0.131 |
Jake Lamb | 2129 | 320.8 | 53 | 0.413 | 0.538 | 0.125 |
Matt Joyce | 1062 | 316.2 | 158 | 0.375 | 0.497 | 0.122 |
Corey Seager | 2228 | 316.5 | 150 | 0.421 | 0.536 | 0.115 |
Jung Ho Kang | 1226 | 318.4 | 105 | 0.420 | 0.534 | 0.114 |
Jose Altuve | 2190 | 317.9 | 117 | 0.431 | 0.543 | 0.112 |
Yoenis Cespedes | 1842 | 320.1 | 69 | 0.456 | 0.566 | 0.110 |
Hernan Perez | 1332 | 311.3 | 243 | 0.347 | 0.453 | 0.106 |
DJ LeMahieu | 2402 | 321.3 | 48 | 0.384 | 0.490 | 0.106 |
Charlie Blackmon | 2194 | 321.7 | 41 | 0.432 | 0.537 | 0.105 |
Cameron Rupp | 1371 | 318.1 | 113 | 0.356 | 0.460 | 0.104 |
Yasmany Tomas | 1716 | 321.5 | 44 | 0.409 | 0.510 | 0.101 |
Matt Carpenter | 2060 | 317.6 | 122 | 0.425 | 0.525 | 0.100 |
It’s easy to be glib, point to Leon’s age and declining exit velocity, and call that a win for our center-field defense distance stat. Look at the players that have been hitting for power all year and have explanations behind them, like Trevor Story, Jake Lamb, and Yasmany Tomas, and how they’ve been defended, and you’ll really like this stat.
But what about Daniel Murphy? Teams must know he’s standing closer to the plate and attacking fastballs and that his power breakout is sustainable on some level. Maybe they do, and this stat isn’t that awesome at capturing breakouts. Especially if the hitter is a pull hitter. Because look at Murphy’s spray chart. His deepest hits are all down the line.
Source: FanGraphs
That pull bias might also explain Tyler Naquin’s center-field depth, since his furthest hit balls are either pushed or pulled. So maybe Ryan Schimpf is the most interesting case here: teams are lukewarm on the old rookie’s power to center. But he has hit homers and long outs in that direction, and has power to all fields. Are teams missing on him when it comes to where they play their center fielder? Is it a function of the parks in which he’s played? Or will he come to earth quickly, as his projections suggest?
Source: FanGraphs
Perhaps we could wrestle our way backwards to a predictive stat here. If we included the average depth of all the outfielders instead of just the center fielder, we would overcome this pull/push bias. We could correct for home parks, since all the Coors’ sluggers see deep center fielders. If we did these things, we could maybe use the teams’ defenses to help us understand which breakouts they believe in. But, even then, it would probably be better to start with the player’s batted-ball profile itself, because that’s what informs the defense. It’s what comes first. And we have a lot more information there.
Still, it’s fun to look up, see a center fielder playing deeper than you’ve ever seen before late in a close game, and wonder what that means.
With a phone full of pictures of pitchers' fingers, strange beers, and his two toddler sons, Eno Sarris can be found at the ballpark or a brewery most days. Read him here, writing about the A's or Giants at The Athletic, or about beer at October. Follow him on Twitter @enosarris if you can handle the sandwiches and inanity.
The first thing I noticed about the leaderboard was the amount of Rockies on there. Do you think that has something to do with their park being so big? In other words, do you think it’d be beneficial to do the calculation as how close they are, on average, relative to the centerfield dimension at their home park? Nolan Arenado, for example, would have the average centerfielder playing 79.3% (329/415) of the way between home plate and center
it probably does have something to do with the distance. it’s a huge outfield out there.
I had the exact same thought. Arenado is at least a believable #1, but Ryan Raburn at #15? Add in the fact that they have 4 of the top 5, plus Blackmon and Lemahieu hanging out in the 40’s, and there is definitely something else going on here.
It’s not just the large outfield either – the thin air probably scares defenders back a couple more feet, too. I thinks this article is really cool and goes a LONG way for a “Proof of Concept,” but there would have to be some control along the lines of general Park Factors before it could be implemented into some real leaderboard or projection system.
I was going to mention something similar, but with the Jays bats (EE, JB, JD and Tulo).
This is absolutely the first thing I noticed. I think this needs to be a Dist+ kind of stat. It would take a bit of work, but I envision something like (PlayerCFDist/ParkCFDist)*100, weighted to the number of PA in each park. I would think this would not only normalize the effect of the park, but reduce the effect of the CF (if a particular CF likes to play deeper or shallower).