Jump to content


Yo! You're not logged in. Why am I seeing this ad?

Photo

State of Defensive Statistical Systems


This topic has been archived. This means that you cannot reply to this topic.
77 replies to this topic

#51 Resonance Wright


  • It's a put-on


  • 1910 posts

Posted 02 July 2006 - 10:07 AM

It's easy to lose track of it in a relatively enlightened forum like this, but the average fan really disdains the notion that stats can tell the complete story on a baseball player. Fans like to believe that a 'real' ballplayer transcends stats -- that a player can have 'heart' and hence value that isn't reflected in their physical performance. That's the PR side of MLB footing a significant cost so teams can derive improved fielding data at no overall value to MLB.

Plus, who would really benefit from the availability of improved statistical data? The answer is 1) teams savvy enough to derive financial gain from improved statistics and 2) fantasy league guys who are similarly smart. And neither of those groups would benefit from that knowledge being widely available. In fact, they'd probably take a hit.

So basically what you're looking at is a proposal that would hurt the people smart enough to add value from it and would have to be funded by a money-hungry group that wouldn't benefit from it but might take a PR hit. The Babe will rise from the grave, forgive the Sox, and come back to hit cleanup before this will happen.

Edited by Resonance Wright, 02 July 2006 - 10:09 AM.


#52 Resonance Wright


  • It's a put-on


  • 1910 posts

Posted 02 July 2006 - 10:14 AM

The only way you'd ever get a network in place to exhaustively catalog fielding events and determine from such data a fielder's skill at a position would be if a private entity -- say, the Boston Red Sox -- could make, or save, more money in doing it than it would cost. Well, to be more accurate, you'd only get that if a private entity BELIEVED it would derive that sort of financial benefit.

Which is possible, I suppose, but it'd be proprietary.

#53 OCD SS


  • SoSH Member


  • 6827 posts

Posted 02 July 2006 - 10:51 AM

No, these stats are NOT "the expressions of a probability."  They are perfect measures of what has already happened.  Big difference. If a batter gets two hits out of five plate appearances, he has batted successfully 40% of the time, and hence, has a batting average of .400.  That's a measure of past performance and not an expression of a future probability.

<{POST_SNAPBACK}>


I phrased that badly, but if you are going to use what happened (and was perfectly measured) in the past as an expectation of what will happen in the future, you are talking about a probability. The reason your example is not an expression of future probablility is because the margins for error outstrip your sample size.

#54 Arock78

  • 1953 posts

Posted 02 July 2006 - 11:10 AM

Plus, who would really benefit from the availability of improved statistical data? The answer is 1) teams savvy enough to derive financial gain from improved statistics and 2) fantasy league guys who are similarly smart. And neither of those groups would benefit from that knowledge being widely available. In fact, they'd probably take a hit.

So basically what you're looking at is a proposal that would hurt the people smart enough to add value from it and would have to be funded by a money-hungry group that wouldn't benefit from it but might take a PR hit.

<{POST_SNAPBACK}>


Winning is the best PR out there. If the team's winning, whether the players are players a fan would love or not, the team would do very well--Bonds, I rest my case.

Besides, the Sox have a lot of players they determined would help them win through a combination of statistical analysis and scouting, and if you've watched over the past few years, they've aquired fewer headache-type players and more intelligent character guys in the process. Just this past offseason, pretty much everyone they picked is a player fans can easily get behind from person perspective (the only exception being the eminently insane Tavarez).

OBP and OPS are stats that give a team an advantage over BA and RBI. Still, not every team has come around on that. They indeed help the teams that are interested in and understand how to use them. And they don't help the other teams. History disagrees with you about your idea that a team like the Sox would take a hit from the information being widely available.

edit: i dont spel goood

Edited by Arock78, 02 July 2006 - 11:24 AM.


#55 redsox1918

  • 4253 posts

Posted 02 July 2006 - 03:32 PM

FWIW,...yes, I basically stay away from the sabermetric world mainly because of the probability factor, but I don't want to keep going there. Pumpsie beat me to it with his example. I'll politely agree to disagree.

I applaud all the posts here! Good job! Only on SoSH... :P

#56 Vermonter At Large


  • SoxFan


  • 3060 posts

Posted 03 July 2006 - 04:05 PM

The only way you'd ever get a network in place to exhaustively catalog fielding events and determine from such data a fielder's skill at a position would be if a private entity -- say, the Boston Red Sox -- could make, or save, more money in doing it than it would cost. Well, to be more accurate, you'd only get that if a private entity BELIEVED it would derive that sort of financial benefit.

Which is possible, I suppose, but it'd be proprietary.

<{POST_SNAPBACK}>


Although I have no knowledge of it, I would really be surprised if some teams, including the Red Sox, weren't already doing something like this.

#57 OCD SS


  • SoSH Member


  • 6827 posts

Posted 03 July 2006 - 04:47 PM

I think it's fairly common knowledge that the Sox and some other teams have their own proprietary fielding metrics; Oakland has one, and the Cardinals simply bought UZR and the person who developed it.

On another board I've been engaging in a debate over what the Sox Proprietary Fielding Metric might look like. IIRC, someone here talked to Bill James last year and manged to get something along the lines of "We have something, and we're very happy with it, end of comment."

In addition, we know that Mr. James is on the Red Sox payroll, and is good friends with John Dewan who has his own feilding system (and Bill James also happened to write two chapters for the Fielding Bible). Dewan is associated with Baseball Info Solutions (who supplied the PBP data for the Fielding Bible) and the Red Sox are one of BIS's biggest clients. I think it's safe to say that the SPFM is similar to Dewan't +/- system.

#58 Bucknahs Bum Ankle


  • SoSH Member


  • 8489 posts

Posted 07 July 2006 - 08:29 AM

Is defense the new "Moneyball" skill that the Red Sox are targeting?  We all know it was OPS a few years ago that was undervalued and teams like the Red Sox and A's were able to exploit that for a while.  Now that the rest of the league seems to have caught on, have the Red Sox brass turned their eyes toward defense in terms of getting the best value out of new acquisitions?  Taking a look at the off-season position player acquisitions:

Gonzalez - Certainly fits the mold, but we don't know if he's in the team's long-term plans. 

Lowell  - Comes to mind as a guy they were willing to take on despite his high salary in part because of his gold glove-caliber defense.  Obviously the main reason they took him was to get Beckett, but I suspect that quality of his defense might have been enough to get the Red Sox to do a deal they otherwise might not have considering his struggles at the plate last season.

Loretta -  Considered somewhat above average defender at 2B.  Probably picked him up more for his bat than D, but he's certainly no Todd Walker out there.

Crisp - I think Damon's future ability to continue playing CF effectively was a significant factor in the team not re-signing him.  While Crisp hasn't often been cited as one of the best defensive CF in the game, we have seen some spectacular plays out of him, and he makes all the routine ones, even if his arm is a little weak.  I think the team got Crisp with an eye toward moving him a corner OF position next season, but that may be changing given how well he has played CF.

Also, the willingness of the team to shift Youkilis to 1B as likely a plus defender shows a leaning toward defense considering he isn't your typical 1B-caliber hitter. 

Personally I think it's all about finding the right balance.  The team has significantly improved the defense over last year, yet they remain a very good hitting club.  Obvious they have been playing over their heads recently both offensively and defensively, and are sure to come back down earth in both regards, but I think going forward we will see a better balance of defense to offense.  Both being very good overall now, rather than having a great offense at the expense of defense as we had seen the past few years.

<{POST_SNAPBACK}>


Theo was on Mike and Mike this morning and basically confirmed what I wrote above. They can no longer find good value in high OBP guys but can get good value in great defenders. Mentioned AGon in particular and the value of his 1-year contract. Said they focused in the offseason on improving run prevention since they really couldn't hope to improve run scoring compared to past seasons. Also mentioned trying to find the right balance of offense to defense rather than relying on guys who can mash the ball out of the park on a regular basis but can't play D.

#59 Hairps

  • 1725 posts

Posted 07 July 2006 - 12:40 PM

A little while ago, I became fleetingly obsessed with defense. As a result, I scoured around for any and all publicly-released 2005 UZRs. Here's the compilation of what I dug up:

2005 UZR (Based on results publicly-released by MGL)
NAME       POS     R/150  Runs         NAME       POS       R/150    Runs
Closser    C                 -7        Beltran    CF*                   12
Matheny    C                  8        Damon      CF*                   -6
Molina     C                  8        Edmonds    CF*                   13
Posada     C                  4        Finley     CF           -19
Pudge      C           14              Griffey    CF*                  -44
Tek        C           -4              Hunter     CF             2
Giambi     1B           0              Jones      CF*                   -1
Laroche    1B         -14              Kotsay     CF           -22
Lee        1B           6              Logan      CF            30
Morneau    1B          15              Rowand     CF            24
Sexson     1B         -18              Taveras    CF*                   22
Teixeira   1B          16              Wells      CF             4
Ward       1B                11        Williams   CF*                  -12
Cano       2B                -1        Cabrera    LF*                  -12
Castillo   2B                 5        Crisp      LF            38
Counsell   2B                26        Holliday   LF*                   14
Ellis      2B          15              Manny      LF           -47
Hudson     2B          12              Matsui     LF*                   -2
Loretta    2B                -9        Abreu      RF*                   -4
Soriano    2B         -20              Guillen    RF            27
Cabrera    SS          14              Hawpe      RF*                  -16
Everett    SS                21        Jones      RF            13
Jeter      SS         -14    -8        Lane       RF***                -20
Reyes      SS               -11        Nixon      RF           -13
Vizquel    SS                 5        Sheffield  RF*          -12     -20
Young      SS         -21              Ichiro!    RF             2
Bell       3B                16        Swisher    RF            37
Chavez     3B           5
Glaus      3B               -25
Inge       3B          13
Lowell     3B                -3
Mora       3B          14
Rodriguez  3B                 1
Teahan     3B         -28

* Does not take into account "arm runs"
** Might not be park-adjusted enough

Edited by Hairps, 07 July 2006 - 01:26 PM.


#60 Frisbetarian


  • ♫ ♫ ♫ ♫ ♫ ♫


  • 4631 posts

Posted 10 July 2006 - 07:38 PM

A little while ago, I became fleetingly obsessed with defense. As a result, I scoured around for any and all publicly-released 2005 UZRs. Here's the compilation of what I dug up:

2005 UZR (Based on results publicly-released by MGL)
NAME       POS     R/150  Runs         NAME       POS       R/150    Runs
Closser    C                 -7        Beltran    CF*                   12
Matheny    C                  8        Damon      CF*                   -6
Molina     C                  8        Edmonds    CF*                   13
Posada     C                  4        Finley     CF           -19
Pudge      C           14              Griffey    CF*                  -44
Tek        C           -4              Hunter     CF             2
Giambi     1B           0              Jones      CF*                   -1
Laroche    1B         -14              Kotsay     CF           -22
Lee        1B           6              Logan      CF            30
Morneau    1B          15              Rowand     CF            24
Sexson     1B         -18              Taveras    CF*                   22
Teixeira   1B          16              Wells      CF             4
Ward       1B                11        Williams   CF*                  -12
Cano       2B                -1        Cabrera    LF*                  -12
Castillo   2B                 5        Crisp      LF            38
Counsell   2B                26        Holliday   LF*                   14
Ellis      2B          15              Manny      LF           -47
Hudson     2B          12              Matsui     LF*                   -2
Loretta    2B                -9        Abreu      RF*                   -4
Soriano    2B         -20              Guillen    RF            27
Cabrera    SS          14              Hawpe      RF*                  -16
Everett    SS                21        Jones      RF            13
Jeter      SS         -14    -8        Lane       RF***                -20
Reyes      SS               -11        Nixon      RF           -13
Vizquel    SS                 5        Sheffield  RF*          -12     -20
Young      SS         -21              Ichiro!    RF             2
Bell       3B                16        Swisher    RF            37
Chavez     3B           5
Glaus      3B               -25
Inge       3B          13
Lowell     3B                -3
Mora       3B          14
Rodriguez  3B                 1
Teahan     3B         -28

* Does not take into account "arm runs"
** Might not be park-adjusted enough

<{POST_SNAPBACK}>


I actually have a few more if you'd like them, the most interesting being Tejada at +13.

I've been meaning to post here for a while to attempt to constrain the use of UZR as a definitive defensive stat on SoSH. All of the numbers you posted are from MGL's "old" UZR calculations. During the offseason, he devised a new, more accurate system that, as I understand it, looked at each play individually for run value, and considered speed of the batted ball, as well as arm runs. Unfortunately, not many of the "new UZR" numbers were made public, and those that were shared showed drastic differences from the old system. Trot Nixon, for example, went from -13 to +4, Manny from -47 to -32, Sheffield from -12 to somewhere around 0. These huge changes were never explained publicly, but there was a rumor of some kind of a calculation error that was never, to the best of my knowledge, confirmed or denied. The system is proprietary. If there was not an error in the new UZR calculations, however, it would certainly call into question all of the old numbers that are out there. A change from the worst RF'der in the AL to an above average player, a 17 run swing (or just over 20 plays), is enough to cause a re-examination ofthe entire metric, especially when it is repeated with multiple players (Jeter had a large change too, if I recall correctly).

I've noticed that posters are using UZR numbers in their arguments in other topics, and I guess I'm trying to say that UZR, with the confusion regarding the change in metric, the limited availabilty, and the often single year number, may not be a good number to use for defensive evaluation without some kind of disclaimer or explanation. I know people like to quote the 3 year numbers of Tango's site, but those numbers are both up to 6 years old, and from a calculation that has undergone multiple tweaks and adjustments as the data and knowledge got better.

Edited by Frisbetarian, 10 July 2006 - 07:40 PM.


#61 Tangotiger

  • 445 posts

Posted 11 July 2006 - 10:10 AM

MGL's UZR has changed into a more Pinto-like model. It has nothing to do with adding extra variables (which in fact were part of the old UZR). The old UZR's methodology was completely explained and publicly released a few years ago.

The big difference is that the new UZR uses more granular zones.

Here are the most-recent, and very limited, UZR:

http://www.insidethe...nd_outfielders/

I have the whole list for all the years, and if/when MGL wants to release them, I will do so.

I will also say that the Fans, AS A GROUP, do a very good job at evaluating fielding, as noted in the subsequent comments int he link above.

#62 Pumpsie


  • The Kilimanjaro of bullshit


  • 10577 posts

Posted 11 July 2006 - 11:14 AM

So, my question after viewing the stats and the comments below them is "What's happened to Coco Crisp's first step this season?"

According to the experts, Coco's first step was one of the very best in baseball between 2000- 2005 for ALL outfielders. This season, OTOH, he's appeared unsure of himself on a number of occasions, and has seemed to have had trouble reading the ball, leading to bad, or delayed, first steps, and having to use his speed to make up for his indecisions. What gives?

Edited by Pumpsie, 11 July 2006 - 11:15 AM.


#63 Tangotiger

  • 445 posts

Posted 11 July 2006 - 12:03 PM

According to the experts, Coco's first step was one of the very best in baseball between 2000- 2005 for ALL outfielders.  This season, OTOH, he's appeared unsure of himself on a number of occasions, and has seemed to have had trouble reading the ball, leading to bad, or delayed, first steps, and having to use his speed to make up for his indecisions.  What gives?

<{POST_SNAPBACK}>


Who are these "experts"? The Indians baseball fans I surveyed, for three years in a row, gave him around an average mark for "instincts", and very good marks for "first step / acceleration", and even better for his speed.

What you call "first steps" may be what the Fans I surveyed call "instincts". When I roll out the scouting report in a month, it will be interesting to see if the Sox fans see Crisp differently than Indians fans did.

#64 Rudy Pemberton


  • Just a string of characters


  • 27883 posts

Posted 11 July 2006 - 12:07 PM

Tango- I question how stastically valid the sample of fans in your survey is, the error bars must be huge. This isn't a random sample of fans of a team, is it? Aren't these samples somewhat biased? If they are found on websites like SOSH, I would think they are very prone to groupthink (the Millar example and how the opinion of him changed over the years makes me wonder), as well.

Can we accurately compare the Indians fans rankings vs. the Sox fans rankings? Wouldn't we need to know more about how the fans differ (do they normally have high rankings? low? etc etc.) Just wondering.

#65 Frisbetarian


  • ♫ ♫ ♫ ♫ ♫ ♫


  • 4631 posts

Posted 11 July 2006 - 12:48 PM

MGL's UZR has changed into a more Pinto-like model.  It has nothing to do with adding extra variables (which in fact were part of the old UZR).  The old UZR's methodology was completely explained and publicly released a few years ago.

The big difference is that the new UZR uses more granular zones.

Here are the most-recent, and very limited, UZR:

http://www.insidethe...nd_outfielders/

I have the whole list for all the years, and if/when MGL wants to release them, I will do so.

I will also say that the Fans, AS A GROUP, do a very good job at evaluating fielding, as noted in the subsequent comments int he link above.

<{POST_SNAPBACK}>


Thanks for sharing those numbers, Tango. I have a few questions about them that I hope you will have the time (and inclination) to answer. First, are these six year numbers all calculated using the new method, or are they a combination of new and old UZR numbers? If they are a combination, has MGL ever explained the drastic differences in the new and old UZR numbers of players such as Nixon, and Ramirez, and would a combination of the systems be accurate? Also, I believe I read somewhere that MGL had changed a bit more than the zones, specifically that he was assigning an actual run value for each ball missed as opposed to an average run value. Still, it does not seem as though this, along with the granular zone adjustment, could account for a change of 17 runs for a player like Nixon, taking him from the worst rightfielder in the AL to an above average one. Do you have any insight into this?

Don't get me wrong, I'm not knocking the system, and I certainly appreciate all the work both you and MGL have done in the area of defensive evaluation, but I am understandably confused right now. I understand that UZR is a proprietary system, and that MGL may not wish to share the details of the changes, but the limited info that has been made available is difficult for me to see as reliable at this point.

One final word to any of you reading this thread - chances are very good that you will love The Book by Tango, Litchtman, and Dolphin. Outstanding stuff, imo, that I've been reading at the beach for the past few weeks, much to the chagrin of my son who is calling me a geek. (My wife took a great picture of me reading it sitting on my board in the sand and waiting for 5 PM so we could surf - the look on my son's face behind me is classic.)

Yo la tengo,
Fris

Edited by Frisbetarian, 11 July 2006 - 12:55 PM.


#66 wade boggs chicken dinner


  • SoSH Member


  • 6602 posts

Posted 11 July 2006 - 01:09 PM

I've been meaning to post here for a while to attempt to constrain the use of UZR as a definitive defensive stat on SoSH. All of the numbers you posted are from MGL's "old" UZR calculations. During the offseason, he devised a new, more accurate system that, as I understand it, looked at each play individually for run value, and considered speed of the batted ball, as well as arm runs.

<{POST_SNAPBACK}>

i think you have it basically right, here's mgl's own description:

Anyway, I am also working on an "ultimate, ultimate zone rating (UUZR)" which, rather than using distinct zones or vectors and the probabilities of catching a certain type ball within them, uses a smooth function such that we can basically plug in the x, y coordinates of a batted ball (along with the usual characteristics - speed, type, etc.) and come up with the probability of that ball being caught, regardless of whether we already have an historical "baseline" for that particular type of ball at those coordinates. I am also going to incorporate into the UUZR methodology subjective ratings on all plays made (which STATS routinely provides) to improve the integrity of the data.

As well, I am working on better ways of "park adjusting" player stats in order to do better context-neutral projections as well as to determine the future value of a player in a specific park, especially when that player changes home teams. I am continually working on improving my projection models, as these are really at the heart of what a sabermetrician can do for a team. Tom might disagree with this as he tends to think that one projection system is basically as good as another.



#67 Tangotiger

  • 445 posts

Posted 11 July 2006 - 01:11 PM

Tango- I question how stastically valid the sample of fans in your survey is, the error bars must be huge.


Nope, pretty small.

This isn't a random sample of fans of a team, is it? Aren't these samples somewhat biased?


Yes, biased towards hardcore fans.

If they are found on websites like SOSH, I would think they are very prone to groupthink (the Millar example and how the opinion of him changed over the years makes me wonder), as well.


Can you cite the numbers, if you are making the claim?

Can we accurately compare the Indians fans rankings vs. the Sox fans rankings? Wouldn't we need to know more about how the fans differ (do they normally have high rankings? low? etc etc.) Just wondering.

<{POST_SNAPBACK}>


We are about to find out. From what I've seen in the past, the results are fairly consistent, about as consistent as two groups of fans from the same team.

#68 Tangotiger

  • 445 posts

Posted 11 July 2006 - 01:17 PM

First, are these six year numbers all calculated using the new method, or are they a combination of new and old UZR numbers?


All-new.

Also, I believe I read somewhere that MGL had changed a bit more than the zones, specifically that he was assigning an actual run value for each ball missed as opposed to an average run value.


I don't think he did, because we talked about it a couple of months ago, and he thought it was a good idea to do. He may have implemented it shortly thereafter though.

Still, it does not seem as though this, along with the granular zone adjustment, could account for a change of 17 runs for a player like Nixon, taking him from the worst rightfielder in the AL to an above average one. Do you have any insight into this?


Anything's possible, especially with Fenway. I don't have any further insights.

...but the limited info that has been made available is difficult for me to see as reliable at this point.


The y-to-y is around .50 for CF with 80 games and LF/RF with 110 games. A bit worse for IF. It's internally consistent.

Your point is why I like to use the Fans' Scouting Report as well. You need to look at both, as Epstein's famous "looking at players through both lenses" comment.

One final word to any of you reading this thread - chances are very good that you will love The Book by Tango, Litchtman, and Dolphin. Outstanding stuff, imo, that I've been reading at the beach for the past few weeks, much to the chagrin of my son who is calling me a geek. (My wife took a great picture of me reading it sitting on my board in the sand and waiting for 5 PM so we could surf - the look on my son's face behind me is classic.)


Thanks! And can you send me the pic to [email protected] ?

#69 Vermonter At Large


  • SoxFan


  • 3060 posts

Posted 11 July 2006 - 01:27 PM

So, my question after viewing the stats and the comments below them is "What's happened to Coco Crisp's first step this season?"

According to the experts, Coco's first step was one of the very best in baseball between 2000- 2005 for ALL outfielders.  This season, OTOH, he's appeared unsure of himself on a number of occasions, and has seemed to have had trouble reading the ball, leading to bad, or delayed, first steps, and having to use his speed to make up for his indecisions.  What gives?

<{POST_SNAPBACK}>

Pumpsie,

I think that part of this is the difference between LF and CF in terms of getting early reads.

It also occurs to me (and I have no proof of this, of course) that the backdrop behind home plate may have a lot to do with how well an outfielder reads a ball coming off the bat. Since the wall behind home plate is so low, could outfielders sometimes have a harder time getting reads off the bat in Fenway than in other parks? I've never heard of this, but it seems possible.

#70 Vermonter At Large


  • SoxFan


  • 3060 posts

Posted 11 July 2006 - 01:35 PM

Tango,

Since you're here, we've been having a discussion in the Varitek thread about the significance of game-calling/pitcher-handling by catchers in terms of run scored. Obviously its not currently quantifiable, but I'd be curious to know if you have ever worked on that and how large you estimate such an effect might be.

#71 Tangotiger

  • 445 posts

Posted 11 July 2006 - 01:43 PM

Tango,

        Since you're here, we've been having a discussion in the Varitek thread about the significance of game-calling/pitcher-handling by catchers in terms of run scored.  Obviously its not currently quantifiable, but I'd be curious to know if you have ever worked on that and how large you estimate such an effect might be.

<{POST_SNAPBACK}>


The only work I've done is this:
http://www.tangotige...t/catchers.html

And your "obviously" comment doesn't apply. Using an age and park adjustment, you can follow the exact same process I did to get to a catcher's game-calling ability. I'm just surprised that no one has done it yet.

#72 Rudy Pemberton


  • Just a string of characters


  • 27883 posts

Posted 11 July 2006 - 01:51 PM

That's interesting, but I don't feel it's geting at what we want to know. Things like SB, WP, etc. show up in a stat sheet- the "intangibles" that Tek supposedly provide evidently don't (or at least aren't obvious stats), or would show up in the pitchers performance (perhaps in terms of BB/K, HR, R, K Rate, etc, etc.)

(Thanks PBK, that was what I am referring to. Good to see consistency across years, I think. I do wonder, though, how often does one's opinion of a defender change? I doubt it changes too much year to year, whereas the stats might? I don't know).

Edited by Rudy Pemberton, 11 July 2006 - 02:02 PM.


#73 PedroKsBambino


  • SoSH Member


  • 14161 posts

Posted 11 July 2006 - 01:54 PM

Tango, Rudy was referring earlier to a post I made yesterday about Kevin Millar here. My suggestion was that perhaps in Kevin Millar's case the fan perceptions of his defense was affected by the precipitous drop in his public image (and bat) in 2005, and I used the fan survey to show that SOMETHING definitely changed, be it real or perception

I want to say, though, that the numbers I cited for all the players suggests that on the whole, the fans were exceptionally consistent in rankings year to year, which was something I noted then as well. There's essentially no significant change in the rankings for Manny, Nixon, Damon, Mueller, Varitek, and Bellhorn between 2004 and 2005 in spite of an intervening world championship and, almost assuredly, a somewhat different set of fans responding in the two years (eyeballing it, there's something like 50% more responses in 2005, so we know there's many 'new' participants). Thus, on the whole, that look I took yesterday made me feel quite confident in the overall accuracy of the fan survey.

The Millar thing is interesting becasue I think it's the exception (or a possible exception), not the rule from Tango's surveys. And as I noted there, one could also suggest that Millar did crater in the field in 2005 justifying the changed perception (though I personally didn't feel this way). Tango likely has UZR numbers available to comment on whether Millar's measured defense changed between 2004 and 2005, so he'll have some info on whether it's my eyes or those of the collective voters that aligns more closely with that system :lol:

Edited by PedroKsBambino, 11 July 2006 - 02:01 PM.


#74 Tangotiger

  • 445 posts

Posted 11 July 2006 - 02:07 PM

Things like SB, WP, etc. show up in a stat sheet- the "intangibles" that Tek supposedly provide evidently don't (or at least aren't obvious stats), or would show up in the pitchers performance (perhaps in terms of BB/K, HR, R, K Rate, etc, etc.)


Right, I'd look at K and BB per BFP, etc, and even BABIP. If he has an impact, it'll show up in the stats.

#75 Tangotiger

  • 445 posts

Posted 11 July 2006 - 02:19 PM

PKB: great job! I am not surprised by the consistency, since I show an r of .9x when I look at 20 fans from the Redsox and compare their ballots to 20 other fans from the Redsox. I believe that the survey is representative of the true fan. Whether the true fan knows what he's talking about is another story.

***

I do agree that fan perception exists, and that can be seen with the Mets / Matsui. Mets fans despise him, and as poor a fielder as he is, he's nowhere near one of the worst fielders in the league across any position. It will be interesting to see how non-Mets fans rate him.

***

As for Millar, the numbers really won't show you a trend, even if they do show you a trend. The sampling error overwhelms the trend-data. That said, Millar as a 1B in 03-05 was pretty consistent, at -2 to -7 per 150 games. Just giving you a straight sum across all his positions (not something I'd recommend), here's his UZR:

year Runs Games
2000 -1 48
2001 -6 88
2002 -5 103
2003 -5 145
2004 -13 139
2005 -6 140

Among the 901 players in the 2000-2005 database, and doing a positional adjustment, he ranks at the 10th percentile (i.e., 90% of all fielders, excluding catchers, out there are better than he is).

#76 PedroKsBambino


  • SoSH Member


  • 14161 posts

Posted 11 July 2006 - 02:30 PM

Out of curiousity, and not doubting for an instant that Kevin Millar is a very poor fielder measured against all position players, does including all non-catchers and using a positional adjustment tend to relegate 1B (and most LF) to the bottom half?

I suppose it just depends how aggressive the position adjustment is, really.

#77 Tangotiger

  • 445 posts

Posted 11 July 2006 - 02:37 PM

Out of curiousity, and not doubting for an instant that Kevin Millar is a very poor fielder measured against all position players, does including all non-catchers and using a positional adjustment tend to relegate 1B (and most LF) to the bottom half?

I suppose it just depends how aggressive the position adjustment is, really.

<{POST_SNAPBACK}>


Yes, the adjustment adds several runs to the SS/CF, a few to the 2B, chops off a few for LF/RF and several for 1B. The gap between the average 1B and average SS/CF is around 15 runs. So, if you have an average fielding 1B and an average fielding SS (or top and top), they are 15 runs apart. If your 1B produces 20 runs more with the bat than the SS, then your average fielding, good hitting 1B is worth a bit more than your average fielding, average hitting SS.

#78 Hairps

  • 1725 posts

Posted 12 July 2006 - 09:08 AM

Chris Dial: Individual [Defensive] Leaders at the All-Star Break - AL

Chris Dial: Individual [Defensive] Leaders at the All-Star Break - NL

PKB has posted the Sox-specific stuff in the Sox Defense Thread.