In what ways have I been wrong?Originally Posted by jojo
I find it odd that the most popular sabermetric metric to determine value in a given year uses a metric that is not supposed to be used unless you are looking at a minimum of 3 years.
In what ways have I been wrong?
I find it odd that the most popular sabermetric metric to determine value in a given year uses a metric that is not supposed to be used unless you are looking at a minimum of 3 years.
Last edited by edabbs44; 05-24-2011 at 08:29 AM.
WAR correlates exceptionally well with pythag....we're past the stage where cherry picking and posting "WAR....LOL" is compelling.
Choo has been one of the best defensive outfielders in the majors during the early season and that's backed by the eyes. Bruce hasn't been nearly impressive defensively during the young season as he's been so far in totality as a Red. The eyes don't bug at the suggestion that they might rate significantly differently with the glove thus far.
Last edited by jojo; 05-24-2011 at 08:35 AM.
"This isn’t stats vs scouts - this is stats and scouts working together, building an organization that blends the best of both worlds. This is the blueprint for how a baseball organization should be run. And, whether the baseball men of the 20th century like it or not, this is where baseball is going."---Dave Cameron, U.S.S. Mariner
You're conflating estimation of true talent with what actually happened....something that has bitten you each season that Gomes has been a Red...
"This isn’t stats vs scouts - this is stats and scouts working together, building an organization that blends the best of both worlds. This is the blueprint for how a baseball organization should be run. And, whether the baseball men of the 20th century like it or not, this is where baseball is going."---Dave Cameron, U.S.S. Mariner
I haven't watched Choo enough defensively to make any statement about his defense, however I have seen enough fluctuation with UZR to wonder how we can use this to measure someone on a yearly basis, let alone a quarterly basis with any certainty.
So WAR's recent full season valuations of guys like Nyjer Morgan, Brett Gardner and Andres Torres have been spot on?
It is what it is and there are multiple threads on UZR which explain why a single season UZR might diverge from a three season sample. You should search the archives because these conversations have been had.
"This isn’t stats vs scouts - this is stats and scouts working together, building an organization that blends the best of both worlds. This is the blueprint for how a baseball organization should be run. And, whether the baseball men of the 20th century like it or not, this is where baseball is going."---Dave Cameron, U.S.S. Mariner
Are there any threads on why a single season WAR might diverge from a three season sample and, therefore, a player's "true talent level"?
I know why UZR is problematic in small samples...but WAR users seem to ignore that fact sometimes.
The biggest reason is randomness in quality of chances and injury but there are multiple reasons.
"This isn’t stats vs scouts - this is stats and scouts working together, building an organization that blends the best of both worlds. This is the blueprint for how a baseball organization should be run. And, whether the baseball men of the 20th century like it or not, this is where baseball is going."---Dave Cameron, U.S.S. Mariner
Sure, but that doesn't make WAR any more reliable.
I have yet to see a complaint about the UZR component of WAR that isn't answered by the Fangraphs primer.
Regarding Bruce, the vast majority of players go through "hot" stretches and "cold" stretches; it's a part of baseball. Throw a 100-PA window on virtually any hitter's stats and the view will be pretty much the same as Jay Bruce's April.
Reliable for what? That is the important distinction..... As an estimate of true skill or as an estimate of what happened? If looking forward, more data is always better (regardless of offense or defense). If looking back, it's pretty tough to justify throwing the baby out with the bath water IMHO (unless there is a clear reason to do so like the Dunn's last season in the outfield where he was dinged -7 runs in little over 180 inning in RF). That said, I favor taking a survey approach to defensive metrics and it is a flaw of WAR that it only relies upon one (but it is also in advantage in that there is consistency between what a defensive and offensive run is in the system).
If the gold standard metrics diverge greatly, then a player's WAR might be more questionable. But UZR and Dewan's really don't diverge as much as most would intuitively think.
For disclosure (because they do diverge more than usual), here are the outliers mentioned earlier:
Here UZR was consistently higher than Dewan's by about .7 WAR. I wouldn't have a problem with anyone arguing that Morgan was a 4.7 WAR player in 2009. I'd be comfortable with anyone suggesting he was between a 4.7 and 5.2 WAR player that year. The reality is that the conclusion about his season doesn't actually change all that much.Code:Dewan UZR Morgan 22 27 Torres 13 22 Gardner 16 25
Also, if it's possible for a player to have a career year offensively, why do we often act like it's impossible for a player to also have a career year defensively?
Last edited by jojo; 05-24-2011 at 10:49 AM.
"This isn’t stats vs scouts - this is stats and scouts working together, building an organization that blends the best of both worlds. This is the blueprint for how a baseball organization should be run. And, whether the baseball men of the 20th century like it or not, this is where baseball is going."---Dave Cameron, U.S.S. Mariner
Can you point me to the section which explains how to account for UZR randomness in a single season WAR calculation?
For example, is Jay Bruce's 5.3 WAR from last year legit or artifically high due to a randomly high UZR in 2010? Can we confidently use WAR on a single season basis or is this metric just a place holder until something better comes along?
Bruce's defensive value was estimated at 19.7 runs by UZR and 17 runs by Dewan's last year. My eye suggested it was closer to +72 ()so I'm confident that his value wasn't an artifact.
Again it isn't artificial unless one can point to an obvious bias. In other words his value of +20 is what he obtained given the distribution of chances that he had last season. That's not artificial as it is derived from his skill given his opportunities.
"This isn’t stats vs scouts - this is stats and scouts working together, building an organization that blends the best of both worlds. This is the blueprint for how a baseball organization should be run. And, whether the baseball men of the 20th century like it or not, this is where baseball is going."---Dave Cameron, U.S.S. Mariner
Good read, Nate. Thanks for the link.
What I got out of it was that there is much uncertainty in it but it's the best we have (for now).
I have no issue with that. My only issue is that it seems to be weighted to highly in WAR. Using Choo and Bruce as an example - going into last night. Bruce's offensive numbers are significantly better than Choo's, but yet Choo was a 1.6 WAR and Bruce was a 1.2. We already know that Bruce is one of the better RF's in the game. Now Choo is a as well, but could the difference be that significant to manipulate the WAR to that extreme?
Funny, now I checked fangraphs and see that after last night Bruce goes to a 1.7 and Choo 1.5. So I guess, like anything this early, WAR fluctuates greatly day to day.
I don't believe WAR is updated on a daily basis (at least I don't believe the defensive numbers are updated on a daily basis).
Posting Permissions
|
Board Moderators may, at their discretion and judgment, delete and/or edit any messages that violate any of the following guidelines: 1. Explicit references to alleged illegal or unlawful acts. 2. Graphic sexual descriptions. 3. Racial or ethnic slurs. 4. Use of edgy language (including masked profanity). 5. Direct personal attacks, flames, fights, trolling, baiting, name-calling, general nuisance, excessive player criticism or anything along those lines. 6. Posting spam. 7. Each person may have only one user account. It is fine to be critical here - that's what this board is for. But let's not beat a subject or a player to death, please. |
Reply With Quote