The Anomalous Defensive Outcomes of the 2015 Season

On Wednesday, we took a look at the anomalous offensive outcomes of 2015: we had oppo homers from batters who have never hit them before, some of the slowest men in baseball hitting triples, and a guy with below average ISO marks hitting an almost 500-foot home run. It was a great reminder that baseball is really, really weird, and every season there’s at least few events that you will rarely — if ever — see again.

Today, we’re going to revisit that same idea except with defense. We can always watch highlight reels of the best defensive plays of the year, because watching Mike Trout perch on top of the wall in center field to rob a home run is a singular pleasure. It’s simple human nature to enjoy that. What highlight reels lack, however, is the context of the players making the plays: we expect Andrelton Simmons to go in the hole at shortstop to pick a ground ball, leap in the air, and fire a bullet across the diamond to get the runner at first. We do not expect Jhonny Peralta to do the same. That’s why if Peralta made that exact same play (he didn’t, but just suspend your belief for a moment), it wouldn’t necessarily be more impressive on an overall level, but it certainly would be on a personal one. For Simmons, that’s business as usual. For Peralta, it’s a once or twice in a career event. That deserves recognition — and celebration.

The usual caveats apply, given that we’re looking at only one-year samples of defensive data. There’s a lot of noise here — we know that. The aim of this post is to find the joy in that great mixture of noise and talent. There are a spectrum of posts on this illustrious website: toward one end we find incredible batted-ball breakdowns of pitchers and hitters that stretch our understanding of baseball; toward the other, we find the carnivalesque atmosphere of a GIF-addled home run post. This piece will stumble gleefully, Mardi Gras crown askew, in the direction of only one end of that spectrum.

On the technical side, we’re employing a few types of data for this pursuit: UZR/150, DRS, and Inside Edge fielding data, the latter of which ranks each defensive play made on a six-step scale of how often an average fielder at the position would make the play in question — from the categories of “Impossible” to “Almost Certain.” I’ve also used Baseball Reference’s always useful play index, as well as some Baseball Savant. Now onto the findings!

Worst infielder by UZR/150 to make a “remote” (1-10% likely) defensive play: Pablo Sandoval, 4/27/15

Sandoval was actually graded positively by UZR/150 in three out of his first six full seasons in the majors. In 2015, his seventh season, he was the worst defensive player in baseball not named Pedro Alvarez. By DRS, he was tied for sixth-worst at -11 runs saved, and even if you don’t put a ton of stock in defensive metrics, the consensus was that he needs to improve if he wants to keep his job at third. The esteemed Matthew Kory even recently outlined a case for his probable improvement. And here’s a glimmer of hope! Here, in front of our eyes, we catch a glimpse of the elusive flying Panda. This play, following a Kevin Pillar single that was hit sharply past the diving third baseman, should be a lesson not just in camera-friendly fielding, but also in how not to bunt. I, for one, am slightly surprised this defensive play was graded as “remote,” given the fact Sandoval didn’t have to run that far, though it certainly gains that distinction when compared to Sandoval’s season-long contributions to the Red Sox’ infield defense.

Worst outfielder by UZR/150 to make a “remote” defensive play: Matt Kemp, 6/5/15

Matt Kemp! With not one, not two, but three somersaulty-things! UZR/150 didn’t like Kemp this season, giving him the fourth-worst mark among all qualified players at -18.0. This is the point at which someone might scream “regress that rating!” and we should — if not for Kemp’s three-year history of woeful UZR numbers (-35.0 and -25.8 in 2013 and 2014, respectively). DRS also agrees with that assessment, and yet this catch — a fine combination of glove work and balletic grace — would seem to dispute those numbers. It meaningfully does not, of course, though no one — not even an angrily bemused Todd Frazier — can take away this moment of glory.

Hardest-hit batted ball resulting in an out: Hanley Ramirez (7/8/15) & Carlos Gonzalez (4/7/15)



One could certainly argue that these are offensive anomalies, and should have been placed in the Wednesday article. But that assumes (incorrectly) that the author of this post thought of this category before now. By resulting in outs, they in some way have the defense succeeding in a situation in which they very rarely do so: with a batted ball velocity of 117 mph each, these are some of the most infrequent outs recorded. Take a look at a tweet on exit velocity vs. batting average from late July by the well-renowned proprieter of Baseball Savant, Daren Willman:

What’s missing in the tweet is launch angle, which is the reason these balls were turned into outs instead of being taken home by a fan sitting 475 feet from home plate. Still, the outcomes are interesting, and even more surprising when the velocity is viewed from the side angle:

Highest-leverage error: Dee Gordon, 5/5/15

Dee Gordon was both: the second-best second baseman by DRS (13 runs saved) and the owner of the highest-leverage error in 2015 (leverage index of 5.4). These two facts are in no way related, though they would seem to be in line with the shared interests of the readers of this baseball website. As for the error — which put runners at first and second base with no out in the bottom of the ninth during a one-run game — it would prove to have no deleterious effects, as Dan Uggla’s attempted sacrifice during the next at-bat resulted in a force out at third, which was then followed by two consecutive strikeouts to end the game.

Honorable mention: Gregory Polanco, 4/9/15

Though I’d be remiss if I did not include this error — which had a leverage index of 5.39, just .01 points short of the one above — I am not particularly happy about having to do so. Watching it is akin to reliving one’s most vulnerable moment: the fulcrum at which a panoply of life’s talents — upon deciding whether to live or die — fail utterly, and all at once. We are all Gregory Polanco, soothed only by the unrealized gifts of next season’s promise.

Highest Win Probability Added (subtracted) by error: Blake Treinen, 4/14/15

Not to be confused with the previous category — which is the event with the highest potential swing in Win Expectancy — this, a costly two-run throwing error, was the event with the highest actual change. Take a look at the interactive Win Expectancy chart of the game, and you can see quite clearly where it happened:

Source: FanGraphs

There was another 25%+ change in Win Expectancy earlier in the game before the error, but this particular play in the bottom of the seventh inning changed the Red Sox’ chances of winning the game from 39.2% to 73.3% — or, 34.1% in one play. No other error this year even came close, with the next-biggest change clocking in at a measly 24%. That’s not too surprising, as late-inning, game-tying errors in a two-run ballgames away from home tend to be pretty game-changing. Brock Holt drove in the go-ahead run on the very next pitch, and Boston would win by that 8-7 scoreline. Ominously, this mid-April game shared a lot of traits with the types of losses that doomed the Nationals’ 2015 season.

Owen Watson writes for FanGraphs and The Hardball Times. Follow him on Twitter @ohwatson.

Newest Most Voted
Inline Feedbacks
View all comments
Danny Middaughmember
8 years ago

Needs more Elvis Andrus.