Can we drop UZr already

[Bobshirleyrules]

I have been on the Yankees board, and for the past several years I have been arguing that the UZR is useless (I did the same thing 11 years ago about the RF/ZR). Sadly, I have watched more journalists trust this stat, and thus report on things that are not happening. I wish one journalist would write a story, please euthanize the UZR stat.

The problem I have with the UZR is that if you plot it among players, and run a regression analysis you will find a minimal correlation to age. Moreover taking some of the best players, the variability within the results from year to year are to put it in layman's terms all over the place.

Last night I made a post on one of the Yankee boards stating how happy I was to see that Jeter's UZR has moved down from costing 50 runs more than the average fielder to 38 runs, I looked this morning and the movement is down to 20 runs. How did one game, specifically the 66th game move Jeter down on an annualized basis 18 runs. Following the logic of this Jeter should have about a dozen ESPN fielding gems. Look I watched the game, Jeter made one great play, but come on already.

Moreover, I was looking at Granderson, according to the UZR he will cost the Yankees about 35 runs more than the average CF. By contrast, Raul Ibanez is actually a plus fielder (he was - 20 last year with Philly)...so tell me if UZR is so accurate why isn't Granderson being pulled rather than Ibanez for better defense. The answer of course is that Girardi isn't an idiot.

I encourage anyone interested in this to pull the best players at respective positions and plot their results, then run a regression analysis. What you are going to find out is that the UZR is meaningless.

[OneTuzSea]

I don't think UZR itself is the problem (although it does have a lot of problems, sure) as much as people misusing it. There is absolutely no reason someone should be using UZR to judge a player for half a season. Minimum before it even matters should be three years. But as soon as it starts populating on FanGraphs people flock to it.

[sexicano31]

But dude, hes got a -100 UZR in like 10 games. SUCKS

[fingerbang]

It fluctuates way too much also. Teix is a good fielder one year and then a **** fielder the next then a good fielder again the year after. Never noticed that when watching. He always seems to be a consistently solid fielder.

[BrianWestKins]

People on this site have taken UZR as gospel for too long now. Rejecting it would be like thinking their whole life is a lie

[sexicano31]

Its better than looking at errors and fielding %

[Bobshirleyrules]

A few points:

@OneTuzSea - You can predict an election of 130 million voters within 3% margin of error surveying 1,000 people....yet 40% of a season is gone and the UZR does not have enough data. The idea that it takes 3 years to measure whether a fielding statistic is accurate is one of the most lame excuses I have heard. I don't fault you for creating it - its been told before.

@sexicano31- At least fielding percentage measures a tangible result. Lousy fielders tend to make a lot of errors. Certainly does not measure range, and in fact some could argue may not correlate well with range (as a player with better range may also commit errors based on his low probability plays; Rey Ordonez of the Mets was a great example of this). Fielding percentage happens to be a very solid stat in that its not skewed and tends to be reliable from year to year (accurate fielders tend to be consistently accurate). Moreover, when you isolate the UZR range stats, you are talking about athletic ability. Thus a player's range should not bounce around so much. Young shortstop should have better range. Is a players bat speed going to diminish game to game. No its physical ability. Look the player might not make solid contact but the force to which he swings is constant. Same with fielding, the players ability to move left, right or catch up to a flyball isn't going to change much from game to game. Yet with UZR its all over the place.

Like I said please Euthanize this statistic.

[flips333]

A few points:

@OneTuzSea - You can predict an election of 130 million voters within 3% margin of error surveying 1,000 people....yet 40% of a season is gone and the UZR does not have enough data. The idea that it takes 3 years to measure whether a fielding statistic is accurate is one of the most lame excuses I have heard. I don't fault you for creating it - its been told before.

@sexicano31- At least fielding percentage measures a tangible result. Lousy fielders tend to make a lot of errors. Certainly does not measure range, and in fact some could argue may not correlate well with range (as a player with better range may also commit errors based on his low probability plays; Rey Ordonez of the Mets was a great example of this). Fielding percentage happens to be a very solid stat in that its not skewed and tends to be reliable from year to year (accurate fielders tend to be consistently accurate). Moreover, when you isolate the UZR range stats, you are talking about athletic ability. Thus a player's range should not bounce around so much. Young shortstop should have better range. Is a players bat speed going to diminish game to game. No its physical ability. Look the player might not make solid contact but the force to which he swings is constant. Same with fielding, the players ability to move left, right or catch up to a flyball isn't going to change much from game to game. Yet with UZR its all over the place.

Like I said please Euthanize this statistic.

Not saying uzr is the way to go but fielding percentage is still not a good stat... And it is skewed, those with better range will tend to get more difficult chances.

[flips333]

A few points:

@OneTuzSea - You can predict an election of 130 million voters within 3% margin of error surveying 1,000 people....yet 40% of a season is gone and the UZR does not have enough data. The idea that it takes 3 years to measure whether a fielding statistic is accurate is one of the most lame excuses I have heard. I don't fault you for creating it - its been told before.

@sexicano31- At least fielding percentage measures a tangible result. Lousy fielders tend to make a lot of errors. Certainly does not measure range, and in fact some could argue may not correlate well with range (as a player with better range may also commit errors based on his low probability plays; Rey Ordonez of the Mets was a great example of this). Fielding percentage happens to be a very solid stat in that its not skewed and tends to be reliable from year to year (accurate fielders tend to be consistently accurate). Moreover, when you isolate the UZR range stats, you are talking about athletic ability. Thus a player's range should not bounce around so much. Young shortstop should have better range. Is a players bat speed going to diminish game to game. No its physical ability. Look the player might not make solid contact but the force to which he swings is constant. Same with fielding, the players ability to move left, right or catch up to a flyball isn't going to change much from game to game. Yet with UZR its all over the place.

Like I said please Euthanize this statistic.

Not saying uzr is the way to go but fielding percentage is still not a good stat... And it is skewed, those with better range will tend to get more difficult chances.

[Superdude]

Kill it with fire!

[Driven]

Hold up. UZR is meaningless? Fine, I don't have TOO much of a problem with you saying that. Of course, there are times when it does indicate something. But then you say how fielding percentage ISN'T meaningless? It's a stat that is measured by what some doofus decides is an error or a hit, which many times can be a silly technicality (can't assume a DP, etc.)

UZR is inconsistant, absolutely. It's a tough stat. I think fielding in general is inconsistant, especially in the outfield. But you can get an indication of who are great fielders, who are bad fielders, and who the jury is till out on.

When you get a large enough sample size (4+ years), and there's consistency in the UZR, you can't tell me that that's meaningless. I shouldn't be generalizing here, but I'm wondering if you're a guy who thinks Jeter is an OK fielder.

When a guy like Jeter's UZR is just downright awful year after year after year, it means he's not a good fielder. When a guy like J.J. Hardy has a strong UZR year after year after year, it means he's a good fielder. I'm not sure how you can label that meaningless.

And yes, I know that one year Jeter had a great defensive UZR. That's not something necessarily wrong with UZR. It could very well just mean that he had a great year, just like many hitters have great years out of nowhere. And pitchers have great years out of nowhere. Production fluctuates, as does UZR.

[Driven]

yet 40% of a season is gone and the UZR does not have enough data.

Welcome to every single baseball stat. Sample size is the key to pretty much everything.

[fingerbang]

Hold up. UZR is meaningless? Fine, I don't have TOO much of a problem with you saying that. Of course, there are times when it does indicate something. But then you say how fielding percentage ISN'T meaningless? It's a stat that is measured by what some doofus decides is an error or a hit, which many times can be a silly technicality (can't assume a DP, etc.)

UZR is inconsistant, absolutely. It's a tough stat. I think fielding in general is inconsistant, especially in the outfield. But you can get an indication of who are great fielders, who are bad fielders, and who the jury is till out on.

When you get a large enough sample size (4+ years), and there's consistency in the UZR, you can't tell me that that's meaningless. I shouldn't be generalizing here, but I'm wondering if you're a guy who thinks Jeter is an OK fielder.

When a guy like Jeter's UZR is just downright awful year after year after year, it means he's not a good fielder. When a guy like J.J. Hardy has a strong UZR year after year after year, it means he's a good fielder. I'm not sure how you can label that meaningless.

And yes, I know that one year Jeter had a great defensive UZR. That's not something necessarily wrong with UZR. It could very well just mean that he had a great year, just like many hitters have great years out of nowhere. And pitchers have great years out of nowhere. Production fluctuates, as does UZR.

Well, Jeter's UZR was donwright awful until 2009. Somehow at the age of 35 he gained some range. Doesn't make any sense to me.

The major problem with UZR is that it gets factored into WAR.

Fielding percentage doesn't tell you everything but it is what is it is, the percentage of screw ups.

[Driven]

Well, Jeter's UZR was donwright awful until 2009. Somehow at the age of 35 he gained some range. Doesn't make any sense to me.

The major problem with UZR is that it gets factored into WAR.

Fielding percentage doesn't tell you everything but it is what is it is, the percentage of screw ups.

It doesn't tell you that when it's up to some nobody who decides whether or not something is an error. When you lose the ball in the sun and it counts as a hit. When you run into a fielder and it's a hit. When you slip and fall and it's a hit.

Not sure if you're agreeing with more or not on that Jeter year, but I don't get why there can't be random years with fielding, when there are random years for hitting and pitching.

[Jeffy25]

What OneTuz said was perfect, people misuse it way too much.

And it also can be really dependent on the guy you are playing defense next to. Ibanez was creating less outs in left field last year because Victorino kept calling him off. So for a lot of those outs, it looked like Ibanez couldn't make the plays. But in reality, he just had an aggressive left fielder. The same goes for Gardner and Granderson in the Yankees outfield. Granderson doesn't call his corner outfielders off that much and just lets them make the plays. Unfortunately, he looks like a light fielder when that happens, and he is playing next to two very plus defensive ball players.

UZR has it's flaws certainly, and it isn't the only defensive metric to suck. But as long as you know what you are looking at and can use the information correctly, it can be a useful statistic. Usually when paired with other defensive metrics it is a lot better.