Those of you who follow the comments may have seen the debates about whether a particular player's bad defense cancels out his good hitting (or whether good defense compensates for poor hitting). One of the most popular ways to measure the "value" of a player's defense these day is a stat called Ultimate Zone Rating, or UZR. UZR is a counting stat, like RBI: over a season, UZR is an estimate of the number of runs a player has saved or cost the team with his defense, compared to a league-average player at his position. For example, Ryan Zimmerman had an UZR of 13.1 in 2009, highest in the NL--his defense was worth 13.1 extra runs for the Nationals. UZR/150 is the accompanying rate stat: it's how many runs above or below average a player's defense is over 150 games (150 games being considered roughly a full season). Ryan Zimmerman's UZR/150 was 13.5 in 2009, also highest in the NL.
Some people who follow baseball stats have concerns with UZR. The math is complicated, and there are some subjective judgments that go into the calculations (for instance, was a ball "hit hard" or not). Also, UZR tends to bounce around a lot--conventional wisdom is that offensive stats like OBP and SLG take something like half a season to settle down on their "true" values; UZR takes something like 2 seasons, depending on what position you play. It's also hard enough to figure out that the first batch for 2010 has only gone up today, and it only gets updated weekly (as opposed to OPB/SLG, which gets updated immediately on the stadium scoreboard after a player's plate appearance). That out of the way, let's take a look at some meaningless (but fun!) early-season stats! Whose glove is gold, and whose is granite? Answers after the jump! Thanks, as usual, to fangraphs for the data.
Early-season silliness
Remember when I said just now that UZR bounces around a lot? Well, if we limit ourselves to players with 100 or more innings at a position, here's what we get:
- Josh Willingham (LF): +2.3 (+25.4 UZR/150)
-
Ian Desmond (SS): -0.4 (-1.7 UZR/150)
- Adam Dunn (1B): -0.5 (-6.9 UZR/150)
- Nyjer Morgan (CF): -3.2 (-36.9 UZR/150)
Now, I don't think anyone seriously considers Adam Dunn to be worth 3 wins more than Nyjer Morgan on defense--this is that "settling down" problem I talked about earlier. I expect Nyjer's UZR to climb on the season (his career UZR/150 in the OF is +24). Those of us watching Dunn play at first haven't seen him obviously embarrass himself with the glove, and so far the numbers bear that out--his fielding is slightly below average, which I think we can all agree would be a huge win for the Nats (assuming that he keeps it up and he starts hitting--career, he's -21 UZR/150 at 1B). Josh Willingham has also made a few nice plays in LF, as my ballpark neighbor CookieLover constantly reminds me. So far, his UZR is top-notch, but I think we can expect that to look a little more like league average with time (career he's -3.2 UZR/150 in LF). As for Desi, so far he's a league-average SS. That's pretty awesome if he keeps it up (it's comparable to Guz' career SS UZR/150 with the Nats, which is about -2).
Early-season sillierness
Now let's take a look at players with 50ish innings or so of playing time.
- Willie Harris (RF): +0.9 (+25.5 UZR/150)
- Cristian Guzman (2B): +0.0 (+0.5 UZR/150)
- Adam Kennedy (2B): -0.1 (-7.5 UZR/150)
- Ryan Zimmerman (3B): -0.2 (-8.4 UZR/150)
Oh noes! Zimmy's glove is broken! No, not exactly--career, he's +12 UZR/150 at the the hot corner, so expect to see that number come up as the season progresses. To date, Guzman seems to be showing league average chops at 2B, despite his "discomfort." Let's hope his arm holds up. Kennedy wouldn't be a terrible replacement there by the look of things so far. TAWH seems to be adapting to RF quite nicely (career, he's +8.7 UZR/150 in the OF, and +17.6 in LF--we'd expect him to be at least +10 in RF, despite his limited playing time there before this season).
Early-season silliestness
Now let's have some fun with the real outliers. These numbers are the fielding equivalent of a relief pitcher with an OPS of 2000 because he got a seeing-eye single in his only AB. For your amusement, a selection of players with 20 or less innings at a position:
- Justin Maxwell is king of the RF platoon with an UZR/150 of +296.3! That's right, he'll win us 27 more games than even Willie Harris! He's going to be making plays on balls hit to 3B! He'll be charging sac bunts from RF for DPs! Okay, that's based on only 10 chances, but how can it be wrong?
- Cristian Guzman is a disaster in RF, with an UZR/150 of -76.7! Sure, he's only had 3 chances in 6 innings out there, but haven't we seen enough? Guz in RF = 100 losses, but J-Max in RF = 100 wins!
- Willie Harris shows that his true love is LF, with an UZR/150 of +102.9 in the other corner--see what one brilliant catch vs the Mets will do when you only have 10 innings in a position?
- Alberto Gonzalez and Adam Kennedy each make the case why they should replace the Z-man, racking up UZR/150s of +85.0 and +79.0, respectively--so what if they only have 5 games there between the two of them.
- Adam Dunn should probably stay at 1B for late-inning defense, as Adam "Dunnside Assistance" Kennedy is tracking a dismal -38.2 UZR/150 on first.... in 17 innings.
What have we learned?
Well, we threw out everything that looked odd (it's noise! low sample size!) and kept everything that matched our expectations (that looks about right!). We've learned how to use advanced metrics to confirm what we thought we knew already, as a certain Nationals manager described in the pre-Opening Day Q&A. In other words, we can't say much of anything about players' true fielding ability from UZR until we're a lot farther along in the season. However, I strongly encourage people to argue about Adam Dunn in the comments, anyway.