For the third time this season, Derek Jeter is back. It's more of a tentative return than a triumphant one, since the 39-year-old's last two attempts to play led to rapid returns to the disabled list. But maybe this is the one when Jeter's balky lower body will prove capable of keeping him on the field for more than four games at a time. If it is, the .225/.286/.307 line that Yankees shortstops have produced might start to look slightly more respectable. Even an admittedly diminished Jeter, sans muscle strain, can be counted on to deliver a steady supply of opposite-field singles; aside from a low-BABIP blip in 2010, his bat has never been a source of concern. It's his glove that causes all the controversy. "He's a guy that has been very successful at the top of the order and has played very good shortstop for us," said manager Joe Girardi after Jeter's 0-for-3 effort on Monday night in Toronto. No one would take issue with the first half of his statement. Plenty of people would refuse to accept the second.
Jeter has long been baseball's most polarizing defensive player. In the right crowd — a mix of sabermetricians and the regulars at Stan's Sports Bar — it takes just three words ("Derek Jeter's defense") to touch off a debate between people who are equally convinced that the Yankees captain is either one of the best or one of the worst defenders of all time. The "best ever" argument is easy. Jeter has the hardware; only four shortstops can top his total of five Gold Gloves. He's one of the few fielders who have a signature move, the instantly recognizable Jeter jump-throw. He even has a pair of pantheon plays: the 2001 ALDS-saving maneuver commonly referred to as "the Flip," and the header he took into the stands after chasing a popup in 2004.
On the other side are the advanced statistics, which disregard Gold Gloves and treat a flashy-looking jump-throw just like any other assist. According to two historical play-by-play-based systems, Baseball Prospectus's Fielding Runs Above Average and Baseball-Reference's Total Zone, Jeter has cost his team more in the field than any other player in history, with both methods assessing the damage at 230 to 260 runs.1
Granted, that deficit has something to do with Jeter's age, longevity, and durability. Fielding performance peaks early for shortstops, which means that Jeter has spent most of his lengthy career in defensive decline. He's a few healthy weeks away from setting a record for defensive games at shortstop, so he's had more opportunities for misplays (and missed plays) than almost anyone else. But even when looking at smaller slices of his career you can see the story of a flawed fielder unfold. Young Jeter or old Jeter, UZR or DRS, massive sample or small, the stats are unified in their disdain for Jeter's defense.
The last thing the Internet needs is another article about Derek Jeter's defensive statistics. At this point, anyone who's inclined to believe that a number fancier than fielding percentage can tell us something about fielding skill has already been convinced, and anyone who isn't won't be swayed by repetition of the same facts and figures. What might work is an approach that synthesizes stats and the old-fashioned eye test.
Fortunately, that approach already exists. It just needs a new paint job.
In 2006, John Dewan and Baseball Info Solutions published the first edition of The Fielding Bible, which promised to "completely change the entire perception of fielding statistics in Major League Baseball." The book introduced the "plus/minus" system that forms the foundation of Defensive Runs Saved, BIS's proprietary defensive statistic. BIS claimed that by combining computation and observation, plus/minus could provide accurate assessments of fielding performance. The company's video scouts watched every play from every game, recording the direction, type, and velocity of each batted ball. Then the plus/minus model assigned each one an out probability. Armed with that information, BIS believed it could answer the question "How many plays did this player make above or below those an average player at his position would make?" The process has been refined since then, but the basic framework remains the same.
Bill James, then as now a foundational figure of sabermetrics and an adviser to the Red Sox, served as The Fielding Bible's resident skeptic, leading off the book with an essay titled "Jeter vs. Everett." James knew that the numbers said that Jeter was a bad defensive shortstop, and he knew that the Gold Glove voters and sign-waving, marriage-proposing Yankees fans insisted otherwise. He just wasn't sure which to believe.
So Dewan gave him a DVD, which is how people shared data before Dropbox. On the DVD were the best 20 plays by Jeter and the worst 20 plays by Jeter (not including errors), according to plus/minus. Also on the DVD were the best and worst 20 plays by Adam Everett, a good-field, no-hit Astros shortstop whom the system rated as baseball's best. James watched all 80 plays, noted the stylistic distinctions between Jeter and Everett, and came away converted.
As James acknowledged in the essay, comparing a relatively small sample of a couple fielders' best and worst plays would seem, on the surface, "to be an ineffective way to see the difference between the two of them." But as he went on to explain, this exercise isn't "designed to reveal the differences between them; this is designed to make them look the same."
Under those circumstances, the differences that do stand out can be instructive. One fielder's lows will be lower, or his highs not as high; maybe he'll make most of the same plays, but he won't make them as smoothly or with as much time to spare. As James put it, "Watching Derek Jeter make 40 plays and then watching Adam Everett make 40 plays at the same position is sort of like watching video of Barbara Bush dancing at the White House, and then watching Demi Moore dancing in Striptease."
We can't add Jeter vs. Everett '06 to our Netflix instant queue to see what James saw. (Sadly, Striptease isn't available either.) And we can't watch Everett in action, since he retired in 2012. But we can re-create Dewan's DVD with GIFs and a new costar, substituting a current player for Everett to see whether the visual contrast between fielders is still as striking today.
I asked Ben Jedlovec, vice-president of Baseball Info Solutions, to send me what BIS deemed to be Jeter's 20 best and 20 worst plays from 2011-12. Playing the part of Everett is Mariners shortstop Brendan Ryan, Everett's spiritual successor as a weak hitter with a phenomenal glove. Jedlovec included Ryan's best and worst plays from 2011-12, too. Jeter, at minus-33 runs, was BIS's worst-rated shortstop over that two-season span; Ryan, at plus-45, was the best.
While we're comparing the two players, we'll also watch out for any reasons to doubt the data we're using. The validity of advanced defensive stats is more widely accepted than it was when James wrote his essay, but metrics like DRS and UZR still attract considerable skepticism, some of it warranted. In 2009, Jeter dismissed his poor defensive ratings, saying, "Everybody doesn't play the same position, everybody doesn't get hit the same ground ball, everyone doesn't have the same runner. So you can't figure out a mathematical equation on it. If Ichiro hits a ball in the same spot that a slower runner does, how can you compute that in a computer? You can't do it."
Raul Ibanez, another player to whom defensive stats haven't been kind, described advanced fielding metrics as "as flawed as the Gold Gloves"2 because they don't consider "defensive positioning or alignment for certain hitters."
BIS has addressed some of these complaints by timing every batted ball, but positioning remains a partial unknown. And as we're about to see, positioning can be pretty important.
The Best Plays
I plotted the 20 best plays by Jeter and Ryan on the diagram below, making the best estimate I could of the point where each ball was fielded. If you're plotting along at home, you can find video of the top 20 plays for Ryan here, and the top 20 plays for Jeter here.
The first thing I noticed was that plotting plays from TV feeds, as BIS video scouts do, isn't easy. Sometimes the broadcast cooperates and you get a panorama at the moment of the play:
It's simple enough to determine that ball's location, since both bases are visible. But in many cases, the camera is so zoomed in that by the time the ball gets to the glove the bases are hidden. This makes the fielder the only frame of reference. How do you pinpoint the ball's location when this is all you can see?
You can back up until you get to a frame that gives you a better sense of where the ball is, but some amount of guesswork is involved. This uncertainty introduces a potential bias into the plotting process: If the fielder's location is all you have to go by, then you might be tempted to judge the difficulty of the play by his proximity to the ball. That would tend to favor players with poor range, whose distance from the ball might cause certain opportunities to be deemed more difficult than they are. If anything, though, that means that DRS could be understating the difference between Jeter and Ryan.
And that difference is noticeable, even when comparing each player's best plays. On the plot above, the points representing Ryan's plays are slightly more spread out. Unless my eyes deceived me, Ryan was responsible for the most far-ranging assist in every direction.3
The stylistic difference is easiest to pick up on balls in the third baseman–shortstop hole. BIS has Jeter at minus-16 plays on balls to his right in 2011-12; Ryan rates at plus-26 plays. When Ryan goes to his right, he often takes the time to plant his feet before throwing, as he does on this grounder hit by Torii Hunter:
Because Ryan isn't running when he lets that ball go, he gets his full force behind it. The throw is a little off line, but Ryan has no trouble getting it to first on the fly, even from a few feet into the outfield. He does this repeatedly, retiring Mark Reynolds, Kurt Suzuki, and, most impressively, Brent Morel:
If a speedier player is running and Ryan doesn't have time to plant and unload off his back foot, he'll throw without breaking stride, whipping the ball to first after an incredibly quick transfer. Here he uses that technique to beat Sam Fuld by half a step:
Jeter, too, sets himself sometimes, as on this grounder hit by Paul Konerko:
It's worth noting that on the Hunter ground ball above, Ryan gets rid of the ball in less than a second, while Jeter takes roughly a second and a half on this Konerko hit. Often that can mean the difference between an out and an infield hit, although we don't know whether Jeter took that long because he couldn't move more quickly or because he knew Konerko was running. (Plus/minus doesn't account for the speed of the batter.)
But with Jeter, balls like that still sometimes lead to jump-throws like this:
We're looking only at Jeter's best plays right now, so this is one that worked. (Even below-average fielders make above-average plays, just as weak hitters like Ryan occasionally have four-hit games.) But jumping takes time. In this case, it works by the slimmest of margins with a slow player running, and the ball bounces well before it gets to first.4
When asked last season whether he had better range to his right than he did up the middle, Ryan responded:
I think I probably do. The play on the backhand — the one where Jeter likes to go with the jump throw. That's what he does best, so it's good for him, and I'll do it from time to time. But for me, it's more the backhand and throwing on the run. I'm able to get a lot on that throw, so it's easier for me to do it in that fashion.
It's at the extremes that Ryan's edge is most evident. This is as far as Jeter gets to the right of second base on any of his top plays:
Ryan gets to this Endy Chavez bouncer …
… and this one by Hank Conger …
… both of which require him to range slightly farther to the right. He also snags an Andre Ethier humpback liner to the right of second that hits the ground at the lip of the outfield grass:
Finally, Ryan flashes a barehand grab, which we don't see Jeter do:
The Worst Plays
The first thing that stands out about Brendan Ryan's worst plays: Most of them aren't that bad. (You can find them here; Jeter's bottom 20 are here.) There are a few flat-out flubs, balls that he juggled or deflected but that were hit hard enough or far enough away from his starting position that he wasn't charged with an error:
However, there's only one ball that gets by him and rolls into the outfield (oddly enough, it was hit by Jeter, whose lack of range may have rubbed off on Ryan for one play). Ryan's range is such that he gets to anything hit in his vicinity, even if he can't quite convert.
Jeter is just the opposite. His bottom 20 plays include four balls that bypass him:
But he doesn't have any that look like Ryan's bobbles. That's one of the reasons (along with his offensive production and overall visibility) why Jeter has won over Gold Glove voters; the fielders who win the award tend to be sure-handed, rather than rangy. Jeter gets outs on an above-average percentage of the balls he gets to, which helps obscure the fact that he gets to so few. It's telling that errors are recorded only when a fielder has mishandled the ball, even though not even getting close to it might be the greater crime. We're more likely to remember a fielder's sins of commission than we are his sins of omission.
Comparing worst plays to best plays is a good way to spot the importance of positioning. This ball hit by Eric Thames would have been an easy play for Ryan if he were positioned in his typical spot:
But since he seems to be holding the runner at second prior to the pitch, he's out of position and can't make it over in time. This qualifies as one of Ryan's worst plays not because of his range, but because of where he's standing before the ball is hit. Conversely, this Alex Gonzalez grounder shows up among Ryan's best efforts:
It's hit harder and deeper in the hole than the ball from Thames, which makes it a more difficult play, in the abstract. But this time, Ryan has the hitter played perfectly, so he hardly has to move. You can see the same effect with this top-20 effort by Jeter on a Martin Prado grounder up the middle:
With Jeter at short, that ball sneaks into center more often than not. In this case, though, there's a runner on first with no outs, so Jeter is playing at double-play depth, pinched in toward second. Thanks to that positioning, he needs to take only a few steps to gobble up what otherwise might have been a base hit.
In a sense, that Ibanez quote about flawed metrics and flawed Gold Gloves is right. BIS tracks pronounced shifts whenever it can, but it doesn't record more subtle shifts in positioning. Those same shifts generally aren't detectable on TV. As a result, the stats might overrate the range of a perfectly positioned fielder, or suggest that a poorly positioned fielder moves much worse than he does.
What we don't know is how much this matters. If a well-positioned fielder is responsible for where he's standing, then his smart positioning is a skill, and he deserves to be praised. If he's being positioned by a manager or coach, he's getting credit for someone else's work. In a retrospective sense, the distinction is irrelevant, but it might matter when projecting future performance. If you plan to trade for a highly rated defensive player, for instance, you'd want to know whether his ratings are the result of his own ability or attributable to a coach. The positioning/range quandary could have implications for aging as well.
A player whose defensive performance depends heavily on positioning might be expected to experience a more gradual defensive decline, since while range and reaction time decrease over one's career, positioning should only improve with age and experience. Without information from a precise motion-tracking system like FIELDf/x, there are still a lot of unknowns.
Here's the really interesting thing about our 2013 reboot of Dewan's DVD: Jeter doesn't look good, but he doesn't seem to be struggling in the same ways that James detailed last decade. James notes that Jeter "played much, much more shallow than Everett," and that "many or most of the good plays made by Jeter were plays made in the infield grass, slow rollers that could easily have died in the infield." That's not the case in Jeter's updated top 20, which doesn't include a single slow roller. In fact, three of Jeter's worst plays came on the kind of opportunity at which James says he used to excel:
Jeter is now playing so far back that he can't charge in quickly enough to convert. James also observes that "Jeter, in 40 plays, had maybe three plays in which he threw with his feet set" and "there was literally not one play in the collection of his 20 best plays in which Jeter planted his feet in the outfield grass and threw." While there's still the odd jump-throw or rushed delivery in this updated selection, Jeter throws with his feet set in almost half of his top 20, and we've already seen him throw out Hunter with both feet firmly planted outside the infield.
So what is Jeter doing differently, and how can he still rate so poorly in BIS's eyes if he's no longer making the same mistakes?
In 2007, his age-33 season, Jeter's DRS fell to minus-24, and Yankees GM Brian Cashman reportedly noticed that he "hadn't lost one step, but two." After the season, Cashman took Jeter to dinner and found a tactful way to tell him that his defense was hurting the team. This came as news to Jeter, who'd never heard that message from anyone else in the organization. To his credit, he took the tip to heart and spent the winter working on his lateral movement with a new fitness trainer.
In 2008, Jeter's DRS improved to minus-10. And in 2009, after working with Yankees infield coach Mick Kelleher on "aggressive defensive positioning," Jeter posted a plus-3 DRS, the first time he'd ever cracked positive territory. Given how well Jeter hit that season — .334/.406/.465 — the thought of him wielding a good glove to boot is almost as scary as a mask of his face. (No wonder the Yankees went all the way.) But that was as good as it got: The next year he fell back to minus-9, and then minus-15, and then minus-18.
So this is the tragedy of Derek Jeter's defense: Just when he finally found out how to play shortstop, he began to get old. Jeter no longer makes as many fundamental mistakes as he did in his early thirties, the last time someone studied his best and worst plays. But at an age at which only a handful of players have managed to spend a full season at shortstop, he lacks the speed to take advantage of his improved positioning. All of which makes Jeter's defensive evolution one of baseball's best might-have-beens. If Cashman had taken Jeter to that dinner about defense a decade earlier, how valuable could he have been? And how many World Series would the Yankees have won?