This past week has served as a mid-season review of sorts, recapping the activities-both surprising and expected-in the performance realms of teams, hitters, starting pitchers, and relievers. We conclude this series with a look at how the fielding has shaken out so far. Unlike the previous reviews, in which my colleagues were able to employ comparisons between projections and actual results, the area of fielding is generally immune to such strategies. In fact, fielding stats are really more along the lines of performance snapshots at a specific point in time rather than irrefutable truths about talent levels. The reasoning deals with the methodologies behind the systems in place, so before moving onto the leader boards, let’s briefly review what these metrics are currently evaluating.

For fielding, Ultimate Zone Rating (UZR) and John Dewan’s Plus-Minus system are generally considered cutting edge, and both work in somewhat similar fashions. Developed by Mitchel Lichtman, UZR essentially breaks up the fielding grid into a wide array of different zones, measures the overall number of outs converted by each position in each zone, and compares the individuals at the positions to the average. The various components include range, errors, double plays, and/or throwing arms with the end result translating plays above or below average into runs, with the type of batted ball and park taken into account, providing a tangible quantification of how a player performed at a position relative to the average of himself and his positional peers. Dewan’s system measures performance in plays in relation to the average without the run conversion. Due to the relative nature of this system, an influx of slick fielders can lead to an improved league, paving the way for situations in which a player with identical skills from one year to the next does not measure up as well.

This is one of the reasons that defensive metrics have not necessarily been embraced as much as some would like; an EqA drop from .335 to .313 isn’t ambiguous at all, while a fielding mark drop from +7 to +4 does not necessarily mean the player worsened in any way. The other caveat to toss out there before showing the numbers is that fielding data should not automatically be extrapolated. Jack Wilson‘s +10.5 at shortstop right now does not indicate that he is on pace for a +21 season; it merely shows that, as of this moment, he has been 10.5 runs better than the average shortstop. For all we know, he could finish the season at +9 or some other mark not necessarily suggested by his current performance. Having said all of that, below you will find the top five fielders at each position through July 8 relative to UZR, with Dewan’s results accompanying for the basis of comparison.

First Base              UZR (Runs) Dewan (Plays)
Paul Konerko              4.6         5.0
Miguel Cabrera            3.9         0.0
Derrek Lee                3.8         4.0
Adrian Gonzalez           2.6         5.0
Kendry Morales            2.5         3.0

Keep in mind when viewing these tables that the data from Dewan’s system will be measured in plays, not runs; the 4.0 from Derrek Lee means he has performed four plays above an otherwise average fielding first baseman at this juncture. Interestingly, Cabrera rates as being the definition of average in one area yet ranks second in another, which would be noteworthy for the discrepancy if it weren’t for the fact that he is on the leader board, period.

Second Base               UZR        Dewan
Brandon Phillips          6.1         3.0
Placido Polanco           5.8         0.0
Ian Kinsler               5.4        18.0
Dustin Pedroia            5.0         7.0
Aaron Hill                3.9        19.0

The keystone is certainly producing some divergent results. Brandon Phillips has been one of the best fielders over the last several seasons, even winning the Fielding Bible Award for his efforts last season. UZR currently recognizes him as having the best glove so far, while Plus-Minus does not acknowledge him on its leaderboard. Conversely, both Aaron Hill and Ian Kinsler are having solid defensive seasons, but Plus-Minus makes them out to be singularly great twin towers dwarfing the field.

Shortstop                 UZR       Dewan
Jack Wilson              10.5        19.0
J.J. Hardy                8.0         7.0
Marco Scutaro             6.2        21.0
Elvis Andrus              5.6         9.0
Rafael Furcal             4.7         2.0

Wilson has been fantastic in the field so far, confirmed by both systems. He makes the most money on the Pirates-$7.25 mil this season with an $8.4 mil option for next year-but his fielding prowess to date has overpowered his below-average bat. The Pirates have been trying to move him for quite some time, but we may finally see the longest-tenured Bucco don a new uniform as potential contenders try to sure up infield depth. And to answer the question everyone is dying to have answered-Adam Everett ranks sixth in UZR, right behind Furcal.

Third Base                UZR       Dewan
Joe Crede                11.4        16.0
Adrian Beltre             9.3        15.0
Brandon Inge              8.3        10.0
Evan Longoria             8.3        19.0
Ryan Zimmerman            7.6        17.0

No surprises here. If one were to poll analysts or scouts, curious to learn of the best fielding third basemen in the game, the majority would likely include some combination of these five names and Pedro Feliz, who would find himself under Zimmerman were the table to be extended an additional player. In fact, this happens to be the most convergent position amongst the two systems, with high ratings all around.

Left Field                 UZR       Dewan
Nyjer Morgan              14.1         3.0
David DeJesus             12.6         4.0
Juan Rivera                9.4        15.0
Carl Crawford              6.5        12.0
Juan Pierre                6.2        10.0

Left field is a pretty interesting position, given that it generally houses two types of fielders with the rare occasion for a third type. The first type of players is comprised of the slower, lead-footed guys like Adam Dunn and Pat Burrell. The second type is equivalent to what analysts generally term as tweeners, players who tear the cover off of the ball in Triple-A but not so well in the majors; they can play center field, but do not particularly well. The third group is made up of those relegated to a corner outfield spot even though they could ably man the middle. Consider the Mariners‘ outfield of Endy Chavez, Franklin Gutierrez, and Ichiro from this season; all three could play center field, but only one name can be penciled in at the spot.

Center Field               UZR       Dewan
Franklin Gutierrez        12.0        14.0
Matt Kemp                 10.1        -1.0
B.J. Upton                 5.4       -12.0
Mike Cameron               5.1        -3.0
Curtis Granderson          2.1         6.0

The most interesting results this season can be found in center, where players characterized as solid with the glove by UZR not only fail to make the leaders in the Plus-Minus system, some even rate as below average. In cases like this it best serves interested parties to investigate the granular components of the systems. UZR breaks data down into the aforementioned categories of runs saved via the arm, double play “turnability,” error prevention, and range. Plus-Minus also factors arm quality into the equation and breaks range up into different segments-to left, straight on, and to right for infielders, for instance. Carving your way through this data may help to explain the vast discrepancies. Unfortunately, I don’t personally see much in the numbers capable of offering any sort of definitive explanation. It could just be small-sample-size antics at work.

Right Field               UZR       Dewan
Jay Bruce                10.2         2.0
Justin Upton              9.8         8.0
Nelson Cruz               9.4        10.0
Ichiro Suzuki             6.9        17.0
J.D. Drew                 5.0        10.0

Right field has the reputation of encompassing the criteria of its corner outfield colleague but filtering for those with tremendous arms. Whether or not this is true can be debated, but comically, the first four guys listed go in reverse when switching systems. As we have seen throughout these comparisons, slight alterations in methodology can lead to wildly divergent results, and if a few other systems were to be incorporated, more discrepancies would likely surface.

Given the widely divergent interpretations of defensive performance, what is there to do? First, remember that we are dealing with a very small sample of performance. A half-season is not nearly enough fielding time with which to discern true talent levels. Factor in that the average can change each year and that the individual numbers are relative, not absolute, and deciphering these talent levels becomes even tougher. Perhaps the best way to figure out how good a player is at a certain position would be to run through the same methodology, but over a three-year span. Find the number of all balls fielded in each zone for a specific position over this time span, and then compare the three-year absolute numbers of the individuals to the rolling average. Another solution would be to use several different metrics in addition to scouting reports from both professionals. If a good number of funs note that Fielder X struggles mightily to his left, then the data has some context to which it can be compared.

Also remember that not all positions are created equally. A +10.5 UZR at shortstop is worth much more than a +10.5 at first base. Tom Tango came up with positional adjustments to be applied in order to level the playing field, so to speak, prorated over the number of games played. The adjustment for first basemen is -12.5 runs per 162 games, whereas shortstops receive 7.5 runs per 162 games for their efforts. Therefore, if two players-one a first baseman, and one a shortstop-with an entire season under their belts surface, who each have +10 marks, the shortstop is actually 20 runs better in the field, since he put up equivalent numbers at a much tougher spot.

Much can change from now until the end of the season, and it might actually be a worthwhile study to take note of the fielding marks at this point in time in order to see where the bottom lines of these players reside. Defensive metrics will continue to improve as more data, like HITf/x, becomes available. However, the variability across systems, based primarily on the inner workings of said systems and not the players themselves, lessens the credibility of defensive marks in general for some. Also, our eyes can only tell us so much, causing problems when the numbers fail to match up neatly with some observer’s personal opinions. The example I like to give involves Carlos Beltran and Shane Victorino; Beltran can glide to the same balls that Victorino scurries for. Both are solid fielders-though the numbers this year would disagree-but many fans will see Victorino’s fleet feet and assume he worked harder to get to the same ball. In actuality, Beltran just has more range and his making it look easier should not count against him.

These factors, as well as the fluctuation of numbers in certain instances, make defensive data tougher to grasp and to use as the basis of definitive claims, but if we are careful, a lot can be parsed from defensive information.

A version of this story originally appeared on ESPN Insider Insider.

You need to be logged in to comment. Login or Subscribe
This comment: "A +10.5 UZR at shortstop is worth much more than a +10.5 at first base." Is a little misleading. Both of them save the same number of runs defensively. I think your point is that if the two players hit at the same level and each have a 10.5 UZR, the shortstop is more valuable because there are fewer shortstops who can do that, but as written, it sounds like you are saying that a +10.5 UZR SS saves more runs defensively than the 1B.
Agree. Another dimension to the question would be: Is it harder to achieve a +10.5 UZR at SS than it is at 1B? The answer to that question may be "yes", in the sense that most major league regulars at more demanding positions could probably, given time to acclimatize to a new postion, become a GG-calibre defender at 1B (although it wouldn't be the 'highest and best use' of their skill set), but not vice versa.
First basemen also participate more often in plays than shortstops do, so they establish their level of performance a bit better where shortstops might fluctuate more since they are involved in fewer plays than first basemen.
What it meant is that even though they saved the same amount of runs, when we normalize for difficulty, the shortstop is a MUCH better fielder.
I don't know if that's a true statement because different positions require different skill sets. The way you write that suggests that a +10.5 SS would be even better at 1B, like a +15.0 1B, but that's not necessarily true. It could be that his speed is nearly useless at 1B, and an inability to catch wild throws shows up to ruin his defensive performance.
The positional adjustments are based off of what would happen if you were to plug in, say, Bloomquist or Bruntlett at any position, using data from players who have historically played multiple positions. We can argue semantics but a +10 1B is not equivalent to a +10 SS. The actual UZR and + - is relative to positional peers but we need to use some form of positional adjustment to level the playing field.
To elaborate on the Bloomquist/Bruntlett front, here are the positional adjustments per 162 games: C: 12.5 runs 1B: -12.5 runs 2B/3B/CF: 2.5 runs LF/RF: -7.5 runs SS: 7.5 runs This is based on what would happen if you put Bloomquist in at a position. The average SS would be 7.5 runs better per 162 games than Willie. However, the average LF/RF would be 7.5 runs worse than Willie. So when I say that a +10 1B is actually 20 runs worse as an overall fielder than a +10 SS in 162 game spans, it is based on historical data of fielding at several positions and the difficulty in playing certain positions. The adjustments are essentially a way to level the playing field, literally, a baseline for comparisons.
Interesting article. I would be interested in knowing how well BP's measures (Rate, RAR, EqA, etc.) correlate with these other two measures. And why the two systems mentioned in the article are generally considered cutting edge.
At least for the CFs who were wildly divergent between the two systems, BP's measures agree with UZR on Kemp and Cameron and split the difference between the two on Upton. BP and UZR both generally use play-by-play data though, so it's natural that they would agree more than the Plus/Minus system.
I have to ask the question... where does Jeter fall on the list?
UZR = 0.2, good for 12th in MLB, behind Rollins and ahead of H. Ramirez.
The Cards took some fan heat last year for the unsexy group of Kennedy/Izturis/Glaus in the INF, with arguments generally being that it would be better to hand those chances to guys playing at the major league minimum (esp. for 2B/SS). One year later, it would be interesting to see how how the Cardinals polyglot of Barden/Thurston/Ryan/Greene/Greene/Hoffpauir/Etc compares to the unsexy veterans. The Kids of 09 have certainly not separated themselves from the 08 Unsexy Vets with anything terrific at the plate...but have they given up anything in the field?
What "Kids of 09"? You're listing a bunch of late 20s players...
Rasmus not playing enough CF to count on the list? I was sure he'd make the top 5 CF with his UZR of 11...
Must not qualify.
Just a silly question, but what about pitchers' defense? Do they play so few innings, relatively, that it isn't measured?
I find it interesting that the right side of the Detroit infield is looked upon favorably by UZR but not so much by Plus/Minus.
Why didn't you use Dewan's Runs Saved? Then you could do a straight-up comparison. Also, it's important to note the difference between the two systems. Dewan's system is more focused on 'value', but Lichtman tries to measure 'skill'.