Dexter Fowler and UZR
by jseiner on Nov 7, 2009 4:08 PM EST
Email
PrintSo I was killing some time on FanGraphs looking over some UZR numbers when I stumbled upon Dexter Fowler's stat line. According to FanGraphs, his UZR for his career is an abysmal -14.5 (-20.0 UZR/150). That's awful. Like disturbingly, he should be playing a corner OF spot, awful.
Granted, I know zone ratings and other defensive metrics should be taken with a grain of salt.
I guess I'm just wondering if there are any Rockies' fans out there who can comment on what Fowler looked like in center. I'm from Pittsburgh, and didn't get to see much of Fowler at all this summer. But being from Pittsburgh, I saw a lot of Nate McLouth in center, and when I saw his lousy zone ratings, I wasn't at all surprised because it was very clear that, despite his athleticism, he wasn't getting to a lot of balls that a good center fielder should get to. His jumps/routes were consistently lousy. You didn't need to be an MLB scout to see it.
Scouts more or less raved about Fowler's defense in the minors.
I'm just wondering if there's any chance they were wrong, or if this is just a fluky outlier from an imperfect rating system?
35 comments
|
1 recs |
Do you like this story?
Comments
Ellsbury
I thought the same thing about Jacoby Ellsbury, who had a -18.3 UZR/150 last season, but a plus value at all three OF positions in 2008. I saw some plays where Ellsbury possibly could have made more plays if he had a better break on the ball, but I also saw him use his speed to get to balls that many others wouldn’t. Regardless, from what I’ve seen, I wouldn’t rank Ellsbury as the worst fielding full time CF.
And from what I’ve heard, I wouldn’t have guessed Fowler would be the third worst. Maybe the two are being rated so poorly for similar reasons. I don’t know too much about UZR to say.
I think we did this argument in another thread a while back.
UZR takes a combination of several factors into account and spits out a number that reflects the overall runs contribution (or lack thereof) of a player in a single season. It could mean that soandso is a bad defender. It could mean that soandso didn’t have a lot of opportunities to showcase their range and / or made too many mistakes on the ones they had. It could mean that soandso had a butterfingers year with the errors. It could mean that they weren’t good at starting or turning double plays (mostly middle infield only).
The best thing to do with UZR is to take it on a larger sample size. Matthew Carruth over at Fangraphs likened a single season worth of defense to be only worth about 1/3rd of a season’s worth of hitting stats with regard to trying to determine skill sets of a player. Basically, you wouldn’t use 230 plate appearances to determine a hitter’s skills. You would use all ~700 or so plate appearances. And even then that might not be enough, if there’s enough noise in the data.
Defensive data is the same way. A single season of defensive data is not enough. It is best to look at, at least, three seasons worth of UZR data to get a decent grasp on whether or not a player is a good, bad, or middling defender.
In the names that get thrown around a lot, Ellsbury had a decent 2008 but a miserable 2009. Does this mean he’s good (2008) or bad (2009)? Both? Neither? We don’t know yet. If he plays a full season in 2010, we’ll have a much better idea of where Ellsbury is going to be going forward. If he comes back and has a decent but not great 2010, we’ll say that he’s probably averagish in center field. If it’s bad, we’ll have to accept that he’s not that great. If it’s good, we’ll have to accept that 2009 might be an aberration.
Fowler is in the same boat.
Fans are typically idiots.
by The Typical Idiot Fan on Nov 7, 2009 5:35 PM EST reply actions
and it could be the case
actually definitely is the case, that every batted ball is different and just because so many are in Player X’s zone and the average player made more outs out of all balls hit in that zone does not mean the average player would have made more outs out of the balls actually hit to Player X.
You're using the "sure hands" philosophy.
Tango studied this and found that the perception of defensive strength is directly proportional to how “sure handed” they are. In other words, and I hate using him as this example but he’s just the easiest to pick on, Derek Jeter. Jeter is often regarded as being a good defender because every ball he gets his hands on turns into an out. He makes few errors, rarely makes the wrong play, and looks like he’s working hard. So everybody assumes he’s a good defender. They ignore his range because they think that making the outs you should is more important than making outs that you normally shouldn’t.
The question then becomes one of personal bias. Do you favor someone who makes the outs he should or do you favor someone who makes fewer of those outs but makes more outs that the other person wouldn’t have. Range factors, which are heavily weighed into most of the modern metrics, rely upon the latter more than the former.
Fans are typically idiots.
by The Typical Idiot Fan on Nov 8, 2009 8:50 PM EST up reply actions
not really
I am saying that if you take thousands of balls hit in a zone, and say throughout the majors x% were turned into outs, and player Y turned fewer than x% into outs, it is information, sure, but it doesn’t really tell you that the average player would have turned more of those particular balls hit to player Y into outs.
But i understand the concept that if 250 balls are hit to a player and he makes no errors and converts 220 on average, whereas another player makes 10 errors but converts 240, you take the guy who converts more balls into outs.
And I agree with your other post that they will eventually calculate the park effects better, and that in general the more seasons of UZR data you get the more reliable it is. I get a problem with the way fangraphs says, “oh, Jay Bay is worth x dolars, Franklin Gutierrez is worth $26.8 million this year” blithely tossed around. Maybe they are worth that, but i think it is a little more nuanced than that.
I think you're confusing worth with value.
Franklin Gutierrez, based on what teams pay for 1 WAR, was as valuable as a player that would have made 26.8 million. It doesn’t mean he’s worth that, it just means he produced like someone who would be.
In doing so, they can also show how undervalued certain players are by the league and / or show how underpaid players bring more value to their team than the overpriced and underperforming veteran.
Fans are typically idiots.
by The Typical Idiot Fan on Nov 9, 2009 4:02 PM EST up reply actions
Jeter
BTW, his UZR has improved last couple of years. It was +6.6 this year.
Did anyone see the Curb Your Enthusiam episode last week where Larry gets in the fight with the stonemason about Jeter? The stonemason says he is overrated, terrible fielder, etc. You know Larry doesn’t know from UZR, BABIP, clutch-hitting, etc.
Jeter had a good year.
As for the CYE moment, I found it to be Larry David exposing the general polarizing nature that has permeated baseball fandom in the last 10 years or so. The stats and online blog fans are seeing one thing while the traditionalists are stuck in their ways. I don’t look at it as Larry David making an indication that one way or the other is better (even if Larry’s character is the traditionalist), I just look at it as Larry David bringing attention to the other side of the coin.
Fans are typically idiots.
by The Typical Idiot Fan on Nov 9, 2009 4:04 PM EST up reply actions
"Larry"
There’s larry in real life, and i don’t know what he knows about sabermetrics etc. but i doubt it is all that much, and “Larry” the character, which i am sure is mostly him but not all. So you may be right. He likely is aware of the divide or else he wouldn’t have featured it, although there isn’t much to explain why the stonemason beluieves Jeter is an awful fielder. He may just be going on visuals or a Mets or Sox fan, unless i am recalling the scene incorrectly.
No matter. it was pretty funny.
Coors Field
Could Coors’ spacious outfield have something to do with it. Went back over the past couple seasons and only like three Rockies’ center fielders have had positive UZR’s over the last 3-4 years, and all have done so over small sample sizes.
I suspect that has a lot to do with it
If someone that knows how the exact calculations work can contradict this then please do, but I believe that every stadium’s “in play” area is sectioned off into the same number of “zones”. Since Coors is so huge, the OF zones are much larger and players there are penalized by extra batted balls dropping in as hits that they really have no chance of getting to.
The above is true, even if you ask MGL who created the measure
Coors is just too spacious. Also it doesn’t take into account initial positioning, so if Fowler was shading over to help Hawpe, his score would take a hit with Hawpe getting a bonus.
Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.
by Jeff Zimmerman on Nov 7, 2009 7:36 PM EST up reply actions
Hawpe and his -21.3 UZR thank you for giving him your bonus.
Positioning is something to consider. However, I question how much positioning would really impact a player’s ability to play defense.
I also question the “COORS IS HEUG” argument. There are other spacious ballparks out there and we don’t see the same problems.
Fans are typically idiots.
by The Typical Idiot Fan on Nov 7, 2009 10:08 PM EST up reply actions
I know MGL has voiced consern about UZR at Coors
Here is the best link I could find for now (pretty week) — I am will keep looking one of his better ones:
Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.
by Jeff Zimmerman on Nov 8, 2009 12:10 AM EST up reply actions
Based on that comment...
…he recognizes that Coors is unusual without knowing exactly how we should weigh it. Which is pretty much the problem here. Some people are rushing to the defense of Fowler and using Coors as an excuse. That’s great, but in another thread I went and looked at the various players who’ve played center field for Coors. If there is an effect, it’s inconsistent. Guys like Juan Pierre and Preston Wilson didn’t show any noticeable Coors effect, while Willy Tavaras did. The rest is a pack of small sample size that is worthless as data.
The short answer is nobody knows what Fowler’s talent level is and nobody knows what the effect of Coors is. However, Hawpe is a bad defender. Just really really bad. If Hawpe is somehow getting a bonus here, it’s getting lost in how much he sucks.
Fans are typically idiots.
by The Typical Idiot Fan on Nov 8, 2009 2:07 AM EST up reply actions
Coors dimensions
You’re right TTIF, I had always just assumed that Coors had a larger in-play area, but it’s actually smaller than Arizona and about the exact same size as a few other parks.
Where are you getting this info?
the lf power alley and CF are very deep in Coors, in part to deal with the altiude. I guess dimensions of other parts of the OF could even it out, so I’m wondering where you’re getting the info.
Coors tends to have a pretty high PF for BABIP. What causes that, I dunno, but it would seem to make it difficult on the fielders. MGL does make park adjustments, but I suppose it’s possible park effects affect CF more than LF or RF.
Anyone have a sub to Bill James online? I thought plus/minus had home/road breakdowns. That info might be helpful.
Leave Dexter alone! You're lucky he even performs for you!
I can't read, apparently
http://www.hardballtimes.com/main/article/speed-and-defense/
Coors is actually the largest according to the article I was looking at.
Fowler shouldn't be shading to "help Hawpe." He should be shading to where a player is most likely to hit the ball!
If the Rockies coaching really is having him compensate instead of playing the batter, that’s a really, really bad job of coaching.
by thejd44 on Nov 8, 2009 4:13 AM EST up reply actions
not quite
The outfield as a whole should shift based on where the batter is likely to hit the ball – but within that setup, Fowler should be shading towards the weaker fielder to spread out coverage better and maximize the chances of getting an out.
We have seen this before.
Ichiro was amongst the worst “ball hogs” in baseball in 2007 while playing center field. He made 97 plays out of his zone, far more than the next highest player. In his defense, to his right was Raul Ibanez and to his left was Jose Guillen. Both are not exactly carpet burners in the outfield. Ichiro’s UZR that year was 6.7. Good for a center fielder, but his range rating was -1.7. His range rating fluctuates in right field from 2.3 (lowest) to 12.2 (highest) and his other meaningful stretch in center was a positive range rating (2008, with Raul still in left but the three headed dog of Wilkerson / Balentien / Reed playing most of right before Ichiro took over there later, superior defenders to Guillen). What we saw was that Ichiro going out of his zone to make plays more often was actually hurting his range rating, which was probably due to him shading more towards left or right than he usually would. He had shorter distances to go, so his range was not as obvious.
If that is the case with Fowler, we should be able to see the same thing. We still have a small sample size to work with, but it’s worth keeping an eye on.
Thanks to the The Hardball Times, we can see that Fowler made 62 plays on balls out of his zone. For 127 games played, that seems like a high number to me. Looks like he was making one out of zone play every 2 games (roughly). If that was a constant rate, we would see 78 over a full 162 game season. I don’t know what the major league average is, but that’s only 19 away from Ichiro’s 97 back in 2007 and while Ichiro did make another 83 of them this season, he has only been above 50 in a full season twice (the two already mentioned seasons). Twice in 9 years implies that making that many is uncommon.
So is there something to this? Maybe. Maybe Fowler was shading over to cover Hawpe more and got busted on his range. There’s simply not enough information to really judge. But like I said, it’s worth keeping an eye on.
Fans are typically idiots.
by The Typical Idiot Fan on Nov 8, 2009 9:03 PM EST up reply actions
I was right.
Looking at this season’s OOZ play numbers, the theoretical 78 number I came up with for Fowler would have put him 5th, tied with Aaron Rowand in the NL and 12th, tied with Rowand, DeJesus, and Choo in both leagues. However, most of the top 20 OOZ playmakers are center fielders. Not sure what that means, if anything.
I wonder if MGL and Tango have already done a plot of OOZ plays and range rating to see if there’s any correlation.
Fans are typically idiots.
by The Typical Idiot Fan on Nov 8, 2009 9:14 PM EST up reply actions
A while back, MGL paid a bunch of people to try to mark starting spot, time of ball travel and final spot for a few years worth of data
I am wondering if they were looking into OOZ plays
Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.
by Jeff Zimmerman on Nov 8, 2009 10:45 PM EST up reply actions
Fowler had a bit of a problem with all-around execution this year. At times, he would exhibit great range in getting to a ball on a given play, only to not execute the actual catch. He also had some problems with gap routes. A lot of the cause is simply inexperience. Keep in mind he skipped AAA, and therefore never had to play a CF as spacious as Coors Field. My take is when he settles in he’s going to improve and be one of the better defensive center fielders in the game.
Good stuff
This is what I was looking for. Thanks.
Defensive stats
Have to be taken with a grain of salt. A year’s worth of offensive representation isn’t equaled by defensive reps until 2-3 years.
I like steak.
SSS or no SSS, it's pretty likely at this point that Carlos Gonzalez is the superior defender
and should be starting in center.
Linda's in the cold ground, won't see her anymore
Somewhere out on the highway tonight, the drunken engines roar
It's just one of those things, one of those things
-- Al Stewart, "Accident on 3rd St."
In memory of Nick Adenhart and all victims of drunk driving
In times like these, go with the scouting reports.
If the scouting report says Fowler is a slick center fielder, go with that. If the scouting report says he sucks, and UZR says he sucks, maybe he sucks.
Fans are typically idiots.
by The Typical Idiot Fan on Nov 8, 2009 2:08 AM EST up reply actions
Gonzalez
Gonzalez was the better defender this year in center. His ability to get to balls and make plays when he gets there is pretty amazing.
StickRat did a pretty good job summing up Fowler’s season. He was raw and made rookie mistakes but he also displayed, at times, every defensive ability that you could ask for in an elite defensive CFer.
Way too small a sample to make that conclusion.
Gonzalez spent a whole 309.3 innings in center field to Fowler’s 977.3. If Fowler’s sample is too small to judge, Gonzalez’s is near worthless.
Fans are typically idiots.
by The Typical Idiot Fan on Nov 8, 2009 9:06 PM EST up reply actions
Observation not statistical analysis
My statement had nothing to do with any sort of statistical comparison.
Gonzalez also played good defense in center in ~530 innings in 2008 in Oakland.
According to the fan scouting report, CarGon is the better CF...
Fowler is probably -5 or so now with +5 upside with his speed and would be pretty great for a corner OF.
I wonder how much of this has to do with their strides?
Gonzalez is relatively short, Fowler is tall. Gonzalez looks like a fast 5’9" guy—quick, short strides. Fowler is the classic smooth glider who elicts comments about being nonchalant or lazy. He seems plenty fast, but doesn’t look like he’s running hard.
Fowler seemed to have tremendous range to me, equal to or better than Gonzalez’s. Most of the Rockies games I watched were on tv rather than at the park, which makes it a bit harder to judge routes, but I don’t remember bad routes on gap hits that Stickrat notes above. I guess I wish I had seen more of CarGo at the park to see if he tended to play deeper than Fowler. After noticing Fowler’s poor showing in UZR, I began to watch for missed balls and he would miss some to the deep gaps. On tv, it didn’t seem like he or any CF would have had a chance at it, but I could definitely be wrong there.
Leave Dexter alone! You're lucky he even performs for you!
Doesn't seem to effect BJ Upton who is considered a fantastic defender
despite the fact that he almost never appears to be running hard.
Tools Whore
three years
is the approximate defensive sample size needed in uzr to be the equivalent of one season’s worth of offensive statistics.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
