templeUsox wrote:
I don't put much weight into UZR.
Why is that?
Also from my understanding Ellsbury also ranks poorly with the other defensive metrics as well.
| Author | Comment | |||
|---|---|---|---|---|
Pantera |
Usefulness of UZR |
Lead | ||
templeUsox wrote: Why is that? Also from my understanding Ellsbury also ranks poorly with the other defensive metrics as well. |
||||
|
|
||||
templeUsox |
#1 | |||
|
Because it's unreliable.
|
||||
|
|
||||
Pantera |
#2 | |||
templeUsox wrote: Care to elaborate? |
||||
|
|
||||
templeUsox |
#3 | |||
|
UZR does not take into account infield positioning, how hard a ball was hit, park adjusting, etc. There are superior defensive ratings which do. That said, all
defensive stats are unreliable at this point. As Aaron says, they are directional at best.
|
||||
|
|
||||
AMarshal2 |
#4 | |||
|
Well, they say they take into account park adjustments, but I'm not sure how effective they are since there seem to be some pretty obvious biases in the
extreme parks (Fenway, Petco, etc.). Eric Van has suggested they don't do it well, though I don't understand his math enough to vouch one way or the
other.
aminahyaquin: Alou is a 37 year old 7 time MVP with a .347 lifetime batting average and a world of hustle. he
can golden glove field in his sleep.
tGG: Alou is actually 42, has never won an MVP, has a .303 lifetime batting average, and is mediocre at best in the field. joshv02: But, other than that, you're wrong. |
||||
|
|
||||
templeUsox |
#5 | |||
|
Balls which hit the Green Monster at Fenway are tallied as catchable balls in UZR. That tells you all you need to know.
|
||||
|
|
||||
Yirmiyahu |
#6 | |||
AMarshal2 wrote:No idea if it's actually true, but I've heard the park adjustments are applied evenly to all fielders. |
||||
|
|
||||
Cem21 |
#7 | |||
|
Ellsbury is a plus defender in CF. Defensive metrics are fine as a point of reference or as a conversation piece, but anything past that is pushing it,
especially when it comes to outfield defense. If you are going to say that Ellsbury has a weak arm and can't throw anyone out, I would agree with that
100%. But, some of the other things floated out there like he doesn't have great range or gets bad jumps on balls are simply not true.
Ellsbury can cover tremendous ground in the outfield and with the exception of the occasional ball hit right at him, he judges balls well off the bat. He goes back on balls hit into the gaps extremely well and hunts balls down. If it weren't for 365 foot flyballs hitting off the Wall, he'd eat up even more balls out in CF. It is true that his elite speed enables him to get to balls that otherwise he shouldn't, but that is a quality you want out of your centerfielder to begin with and that should only bolster his defensive value, not punish it through misconceptions.. Is there room for improvement for Ellsbury in taking the next step towards being one of the elite CF in the game? Yes. But, this notion that he is a subpar CF or that his play in CF is subpar is the result of some metrics thrown together to attempt to quantify an aspect of the game that by all admittance is the most difficult to quantify through metrics or numbers. |
||||
|
|
||||
Hairps |
#8 | |||
templeUsox wrote: Not if I need to know a good recipe for blueberry pancakes.
|
||||
|
|
||||
norbit14 |
#9 | |||
Cem21 wrote:Agree with you 100%.
Here comes Pac-Man Floyd.
|
||||
|
|
||||
JimRiceRocks |
#10 | |||
Hairps wrote:Courtesy of UZR - Ingredients
Directions
"You can never have too much bitching" - left coast bosox fan
"No wonder a lot of opponent's announcers have bullpenis envy - left coast bosox fan |
||||
|
|
||||
Yirmiyahu |
#11 | |||
|
Why would you ever use frozen blueberries?
|
||||
|
|
||||
CT Sox Fan |
#12 | |||
Yirmiyahu wrote: Don't question UZR. |
||||
|
|
||||
JimRiceRocks |
#13 | |||
|
UZR is all about convenience
"You can never have too much bitching" - left coast bosox fan
"No wonder a lot of opponent's announcers have bullpenis envy - left coast bosox fan |
||||
|
|
||||
joshv02 |
#14 | |||
|
UZR uses measured batted ball speed (at least it did in the early iterations) as well as park effects. FWIW, I really liked David Pinto's graphs years ago.
They were fun
|
||||
|
|
||||
Pantera |
#15 | |||
templeUsox wrote:I've written on this in the past but I think it bears repeating. One of the problems is that UZR is widely misunderstood. For instance, even a knowledgeable fan like yourself doesn't know that UZR DOES in fact take into account how hard balls are hit. The more common arguments against the usefullness of UZR are that it can be influenced by factors beyond a players controll and is hence "unreliable". Yet this is also true of offensive stats. OPS, VORP, EQA, and wOBA for instance, can be influenced by several factors that have nothing to do with the players ability. Those who use these metrics do point to these possible influences when using these metrics. But no one uses them as a reason to dismiss the metrics entirely, or even call them "purely directional". UZR isn't completely "reliable" because no one statistic can be completely reliable. UZR is in fact no more or less reliable than any offensive or pitching statistic. Just as EQA won't tell you everything about what a player can do offensively, UZR won't tell you everything about what a player can do defensively. It is certainly tells you more about what a player can do than fielding percentage, zone rating, or range factor. It probably also tells you more than just using your eyes. |
||||
|
|
||||
templeUsox |
#16 | |||
Pantera wrote:Of course it is not. Because it is subjective. And other stats are objective.
Last Edited By: templeUsox 12/16/09 1:58 AM.
Edited 1 time.
|
||||
|
|
||||
templeUsox |
#17 | |||
Pantera wrote: It is certainly tells you more about what a player can do than fielding percentage, zone rating, or range factor. It probably also tells you more than just using your eyes.UZR is just a shined up version of ZR. And even though it has been modified to include speed of batted ball (Josh, you are wrong, it did not incorporate this in the old versions), it is measured on a very basic scale of (soft, medium, hard). Conversely, BIS' stats are measured on a 1-10 scale, with drastically more specific zones, and more highly trained spotters who account for the position of fielders on the field. UZR was a nice start to the whole defensive metric solution in the early oughts. But it has long out-stayed its usefulness. It is not equivalent to EQA or OPS+ or wRC+ in any way, shape, or form. I really can't believe you would stand by that statement. |
||||
|
|
||||
Pantera |
#18 | |||
Cem21 wrote: When defensive metrics point to a conclusion that is different from conventional knowledge one should look at why it is that this metric points to a different conclusion. This is made somewhat harder by not having the play by play data that is used to compile UZR. From my watching of the games I saw that Ellsbury did indeed make great plays when moving laterally. Yet when going back and forth the results seemed to be weaker. For whatever reason, a lot of balls seemed to fall in front of Ellsbury. I still don't understand why Ellsbury wasn't in postion to catch a the looping fly ball from Vladamir Gurerrero that basically ended the Red Sox season. I also don't think that this is just a case of defensive metrics punnishing Ellsbury for not catching an inordinate amount of uncatchable balls. According to Baseball Prospectus's ESPN "Kiss em Goodbye": Jacoby Ellsbury is fast and makes some impressive-looking plays, but he takes awful routes and plays deep because he doesn't go back on the ball well with Vladimir Guerrero's game-winning hit on Sunday a prime example. Additionally in listening to talk radio the last few days, I defintely heard a comment from a beat writer (Maybe Tony Mazz?), indicating that the Red Sox had "considered moving Ellsbury to LF in the past". I wouldn't be certain that the Red Sox completely buy into your comments above. Finally, I am skeptical that the Red Sox would sign a CF with a long term repuation for defensive to a long term deal if they really thought they had a CF on the verge of becomming elite defensively. All and all there seems to be some smoke here. It could very well be that an unusual amount of uncatchable balls at Fenway influenced his poor defensive ratings. But I don't think that taking those away would turn Ellsburry from the lowest ranked CF in the majors to one of the best. |
||||
|
|
||||
Pantera |
#19 | |||
templeUsox wrote: No one is arguing that UZR perfectly measures defense and it maybe that other metrics are more granular. I would only argue that its a good measure of how many balls in play the fielder got to given the type of chances he had. There maybe offensive measures that are more granular than OPS but I've never heard anyone argue that OPS was "useless". Plus as MGL pointed out making data more granular becomes a case of dimminishing returns[url=http://www.baseballthinkf...03-03-21_0/[/url]. I would imagine that UZR likely correlates very well with other complex defesive statistics and that there are very few cases where a player is great by one metric but terrible by another. In the second paragraph you ridicule my claim that UZR is equivlent to well known offensive measures. Riduculing a claim isn't evidence that the claim is false. I stand by that statement because all of the evidence that I have seen shows it to be correct. Its irrelevant if you believe it or not.
Last Edited By: Pantera 12/16/09 3:19 AM.
Edited 1 time.
|
||||
|
|
||||
templeUsox |
#20 | |||
Pantera wrote:That's not a strawman, because you are arguing that UZR perfectly measures defense. How are you doing that? You, in your own words say: UZR is in fact no more or less reliable than any offensive or pitching statistic.OPS is a perfect stat. Why? Because it is 100% perfectly measuring someone's OBP and SL% combined. And those component statistics are perfectly measuring the rate at which people get on base and at which they slug, as defined by conventional statistics. There is no doubt in my mind that someone's slugging percentage is correctly providing the ratio of total bases someone has accumulated over their plate appearances. This is objective information which is being plugged into a simple equation. It is 100% reliable. How does that change with UZR or any defensive statistic? Because there is automatically a level of subjectivity which cloud the numbers to some extent. MLB has taken great strides to remove this subjectivity, with the HitFx system which is on the horizon. But all the antiquated defensive statistics are hugely subjective. That is why they aren't reliable to the baseball fan, who for years has been accustomed to using statistics which are purely objective. |
||||
|
|
||||