I had bought in to the claims of improved defense from Jhonny Peralta, and anecdotally it seems as if he is much improved, but UZR certainly pours a lot of cold water on that theory.

If Andy Marte and Asdrubal Cabrera both continue their current career paths I would not at all be surprised to see 2008 open with Jhonny Peralta as the Indians starting third baseman.

Albert Pujols is a gold-glover.

That's just boggling.

Eh? Pujols has been getting acclaim for his fielding ever since 2005.

Anyway, J.D. Drew: -10. Whoa. Where did that come from?

Eh? Pujols has been getting acclaim for his fielding ever since 2005.


I know; he just doesn't look it; I mean, yeah, he makes good plays, but he doesn't really give off that vibe of being a great fielder.

Add to the fact that you expect most career .330 hitters to not be as good with the glove, and..

Renteria and Chipper Jones are missing from the list?
All the Braves fielders have been above average otherwise, which kind of agrees with visual observation

Are the three year averages available? That might be more interesting right?

Albert Pujols is a gold-glover.

He maybe... but this information is very limited in determining it, as UZR doesn't deal with the most important aspect of 1b defense. Receiving.

I know; he just doesn't look it; I mean, yeah, he makes good plays, but he doesn't really give off that vibe of being a great fielder.


Really? He looked like crap in left field, but he's looked outstanding at first base. He pounces on anything close to the bag. As for receiving, he's looked as good as you can look to me. Receiving may be the most important thing, but I don't think there's a huge range of talent levels for that among first basemen who shouldn't obviously be DHs.

He maybe... but this information is very limited in determining it, as UZR doesn't deal with the most important aspect of 1b defense. Receiving.

Pujols excels at receiving. You get pretty good at taking bad throws when David Eckstein is the one "rifling" balls from shortstop.

Dewan credited Pujols with an obscenely good number of scoops.

When I go to the link above, there's data for the years 2003-2007.

Good to see that Manny still sucks, even under a park adjusted system like this.

Darren, here's Link the to the 2007 numbers.

When I go to the link above, there's data for the years 2003-2007.


where is that?

I must have been blind, found the Chipper and Renteria #s

So, I feel a little better about Tony Pena Jr.'s 265/296/354. Not much, but a little bit.

So, I feel a little better about Tony Pena Jr.'s 265/296/354. Not much, but a little bit.

that hitting is a plus! When they traded for TP Jr, they were trading for those +9 runs on defence

Jeff Kent's numbers are the ones that are boggling -- -15 in only 34 games? It is hard to be that bad even on purpose.

As if Figgins offensive numbers weren't depressing enough...

Srul - I'm pretty sure you cant be -15 in 34 games thats .5 plays less than average per game.

Srul - I'm pretty sure you cant be -15 in 34 games thats .5 plays less than average per game.

Did I misread the spreadsheet? Or are you saying the UZR just does not make sense empirically?

I agree with you that it sounds extreme.

Even in this small sample size the leaders tend to be the guys who UZR has liked from 2003 through 2006...Pedro Feliz, Everett, Utley, O-Huds, Mark Ellis, Rolen, Sizemore.

With the glaring exception of Aramis Ramirez, what the heck is he doing with 6 runs saved already?

I've always felt that UZR had a very high standard deviation. And I'm also pretty sure MGL has discussed this plenty in the past...

When you see a number like Kent has, it is probably that some of the gound balls that the database "sees" as easy were not so easy or he was in a shift which was not accounted for (I don't assume any "shifts" for certain batters) or any one of many reasons why sample data has sample error in it and why we don't just prorate someone's sample numbers especially for small samples. We don't look at A-Rod's 15 homers in the first 30 games and say, "Gee I am surprised at his 90 homer per 150 game number," do we?

Maybe Carlos Zambrano has been giving up a lot of groundballs which were very easy to catch but the database thinks were hard to catch.

mgl,

Do you have the /150 games posted somewhere?

This is simply brilliant.

Just on the 03-07 combined file. Trust me, for 07, you don't want to be influenced by the "per 150 games" numbers and they serve little purpose - no more than if someone hits 2 HR on the first game of the season andf you want to see his "per 162" HR numbers.

I think they just provide a frame of reference. For offense, I can look at something like 9 HR in 40 games and know that that's a good pace, but it's not as readily apparent when it's 4 UZR above average in XX games. In other words, I'm not assuming the player is actually that good, I'm just using it to understand the actual rate that he's posted thus far.

pretty sure that should be 75 HR/150, which is a reasonable pace actually...

Kent's pace is probably unsustainable, but if you had to pick anyone for the shitty fielder of the year award, a near 40 year old 2B with no speed who's always been big for the position is a pretty good candidate.

I've added up the team data here:
http://spreadsheets.google.com/ccc?key=pdta7ixUc9m5D9He3hJXyFg

and wrote about it here:
http://mvn.com/mlb-stats/2007/06/01/uzr-data-for-the-last-4-13-years/

Chone - are you blogging on that site now, rather than your old blog spot site?

Couple of weird ones. Wily Mo Pena, Matt Stairs and Chris Duncan near 0. Lyle Overbay and Tori Hunter in the negatives. Pairing this info with what you have observed allows you to assess pretty well which ones apply and which ones don't.

Pairing this info with what you have observed allows you to assess pretty well which ones apply and which ones don't.

I sound like a broken record but you have to be careful with that kind of thinking! One of the ways that small samples can have error is when a players looks good but they really are not. That happens all the time on offense but somehow its not supposed to happen on defense. Well it does. I imagine that if I had lots of video and time on my hands, I could construct video clips of the worst fielders in baseball making Sportscenter highlight catches and the best fielders in baseball looking like the reincarnation of Skates Lonnie Smith. What makes you think that it is not the same thing when you watch a fielder for 5, 10 or 20 games?

Now, if you are an intelligent and somewhat objective and keen observer of defensive talent, you can certainly make inferences about players' true defensive talent by their speed, jumps, etc. But even those can be deceiving in the short run and sometimes even the long run. I have spoken many times about how one of the ways that defenders get overrated is that they simply "look good" in the field even to the trained eye. How the heck do you think that Jeter gets so overrated in the field by lots of players, coaches, managers, GM's, scouts, and ex-player broadcasters? How is that possible? Because some players who don't get to many balls give the illusion that they do and vice versa. Not a lot, but some. And the shorter you observe a player for, the more likely that they will fall into one of those categories.

Look at the above comment from a very smart and knowledgeable guy about Kent. His near zero UZR for 2007 is suprising? He was -7 in 06, 0 in 05, +10 in 04, and +9 in 03. (OK I cherry picked the 03 cutoff because he was -9 in 02, but then again in 01, he was +10!) This guy has been a very good, nearly gold glove defender at 2B for almost a decade, yet he gets no credit for his defense. I am not sure why. He is certainly capable of putting up zero UZR numbers now even at 40 years old. Capable - what am I saying? My projection for him in 07, which is based on his past and his age, is +1! Maybe AROM meant that the 0 was surprisingly LOW?

One other thing that really surprised me about the data: Biggio at +2. Even with the small sample size, that very much contrasts with what just about every writer is saying about his defense -- that he is immobile, has no range, etc. Well, at least compared to the average, and for the period of time he has been in, he has not hurt his club in the field.

At bat . . . well, another 20+ hits, and they can relegate him to only occasional duty. But what the heck are they doing, batting him lead-off?

Chone - are you blogging on that site now, rather than your old blog spot site?

Trying, poorly, to leep up with both.

Maybe AROM meant that the 0 was surprisingly LOW?

I'm confused here. We are talking about Jeff Kent, who's listed at -15 for this year (at least in the file I downloaded)? In 2006 he was -11 (-14 per 150), average in 2005, very good before that. Seems like age has caught up to him.

That Jeff Kent was ever a good second baseman always surprised me, just seemed too big and too slow for the spot, but that's a comment about the limited utility of observation more than of defensive stats.

well, another 20+ hits, and they can relegate him to only occasional duty. But what the heck are they doing, batting him lead-off?

More atbats? The quicker he gets 3000, the quicker they can replace him.