This won’t be a terribly long post, but I just want some people to understand what UZR is and how much defense actually accounts for a player’s total production.
UZR is ultimate zone rating. UZR/150 is UZR over a 150-game fielding sample size.
UZR/150 is not the definitive defense statistic. Frankly, there is no statistic like that and there probably won’t be until a program called Field F/X begins to chart exactly where and how a fielder caught a flyball or fielded a groundball and how far he ran to get it.
But for now, UZR/150 is the best defense statistic for citation purposes.
How it’s calculated: A man with a computer marks each time a fielder fields a ball, whether it was a flyball or a groundball. He then marks down which zone the ball was fielded or caught, whether there was a throw on the play and if an out was made. As the year rolls on, a complete picture begins to form on just how good the player was defensively: how far he went to catch a flyball, how good he was at converting groundballs into outs, and so on. If I’m understanding Alex Remington’s post on Big League Stew, this is how the field is divided up:
This is what UZR counts, courtesy of SABR library:
- Outfield Arm Runs (ARM) – the amount of runs above average an outfielder saves with their arm by preventing runners to advance.
- Double-Play Runs (DPR) – the amount of runs above average an infielder is in turning double-plays.
- Range Runs (RngR) – is the player an Ozzie Smith or an Adam Dunn? Do they get to more balls than average or not?
- Error Runs (ErrR) – does the player commit less or more errors compared with a league-average player at the position?
So long story short, it accounts for pretty much any possibility that can happen when a ball is hit into play. Your main concern is how much weight we should put into it.
The best way to use UZR is citing three-year sample sizes. Consider that there are three big outcomes that are completely unaffected by fielding: home runs, strike outs and walks. These account for a lot of different outs. Then factor in that a player’s defense will only matter for a fraction of the balls hit into play and you see why a large sample size is necessary.
The other part to consider is that human error comes into play. Though UZR is pretty advanced, a person still has to mark into a notebook where the ball was hit and to which part of the field.
There’s also an important note about first basemen and catchers: UZR does NOT take into account the other jobs a first baseman or a catcher has: catching pitches, catching throws from short stop, and other defensive acts that are specific to the position. So UZR should be used primarily against outfielders, third basemen, shortstops, pitchers and second basemen.
Though UZR has its flaws, it’s still a valuable tool. Citing UZR/150 in three-year sample sizes and you can see who the best fielders in baseball are. Personally, I see UZR/150 as having 20/30 vision. You have a very good idea of what the letters are at the 20/20 level, but it’s still fuzzy. The closer you get (the bigger sample size you receive), the better.
Note that this is completely different from hitting evaluations. A player’s offensive output is much more static and easier to weigh because he gets so many more chances. A player will get 670 plate appearances at the plate in a year, but sometimes only 200 chances to make an out or a play at his position. Sometimes, he may drop a ball that should have been fielded and that’ll skew his UZR.
Now maybe the best thing about UZR/150 is it puts a player’s defensive value into runs ceded or saved. The most fascinating thing I learned over the past few years is how much defense can off-set a player’s offensive value.
Take Adam Dunn, for instance. Dunn is the poster child for underrated offensive players: very high on-base percentage, huge slugging percentage and he takes a ton of walks. His defense, though, is so bad it completely off-set his offensive value in 2009, and that’s with a full 159-game season. UZR/150 and a player’s offensive output measured in Runs Created (wRC) combine to make Wins Against Replacement. His WAR was less than Rafael Furcal’s last year. Basically, Dunn costs his team so many runs in the field that he should only be a DH.
So think about that. Though we know so much about offensive output, we’re only just now starting to get a good grasp on a player’s defensive output. When citing UZR, remember that a total sample is best. And that it can completely off-set any player’s offensive production.
So, with that, here’s how valuable a few Dodgers were in 2009:
Matt Kemp: worth 22.9 runs on offense and saved 2.9 runs on defense. Total WAR: 5.0.
Andre Ethier: worth 25.3 runs on offense and -14.7 runs on defense. Total WAR: 2.7
Manny Ramirez: worth 25.4 runs on offense and -10.7 runs on defense–this despite missing 50 games. Total WAR: 2.5.
Rafael Furcal: worth -5.2 runs on offense and saved 5.7 on defense. Total WAR: 3.0 (note: Furcal’s WAR is weighted differently because of his value as a short stop).
Casey Blake: worth 13.5 runs on offense and saved 11.7 runs on defense. Total WAR: 4.6
Jamey Carroll: worth -2.7 runs on offense and saved 5.2 runs on defense–this despite playing in only 93 games. Total WAR: 1.5.
Nothing on Blake Dewitt yet, but we’ll have a good picture of that at the end of the year. Also, Orlando Hudson was very productive on offensive and average on defense, despite his reputation.
So as you can see, the Dodgers’ defense is considerably more valuable, with the exception of Ethier and Manny, than the replacement team’s. Despite their incredible hitting–and their high on-base percentage from last year, which may have been the key to their success–the defense was key.