After 14 seasons and 325 home runs in the Major Leagues, Jermaine Dye is retiring, according to Ken Rosenthal of FOX Sports (Twitter link). Dye, 37, last played in 2009, though he considered returning to the majors in 2010 and 2011.
Dye, a 17th round selection in 1993, hit 325 homers in his career with a .274/.338/.488 line. He played for the Royals, White Sox and Athletics after breaking in with the Braves as a 22-year-old in 1996. The two-time All-Star and 2006 Silver Slugger recipient was the World Series MVP in 2005 when the White Sox won the World Series.
Dye was traded twice; he went from the Braves to the Royals in 1997 and from the Royals to the A's in a three-team trade in 2001. He earned just shy of $75MM in his career, according to Baseball-Reference.
Meh, he should have just taken a reasonable offer last season and continued playing in some capacity…
You are racist just for saying that!
hes only retiring because no one wants him.
I hope he thought this out…
Solid career. Good luck.
Best of luck to Dye. He had an excellent career and played the game the right way.
an excellent career?
sure he did get a little greedy their at the end but 325 Hr and a world series ring along with 75 million dollars 274 career average , i would call that excellent not HOF excellent and probably should have played another 2 years and end up with 360 plus Hr
i think people get sentimental and lose some perspective when guys retire. you’re listing HR totals, and that’s about all you can mention as far as excellence goes for dye
.274/.338/.488 career line
.353 wOBA, 110 wRC+ (that is, a tick above average)
all while consistently being a significant defensive liability. you’re talking about 18 total WAR over ~11+ seasons
1.6 WAR per year = excellence?
maybe if you said he’d been an excellent hitter in his career. but even then, i dunno. i mean has andruw jones had an excellent offensive career? kevin millar? milton bradley? these players have the same career wOBA
when you factor in defense, the entire other half of the game, there is just no way to call it an excellent overall career imo
Why is everyone so damn serious on here? Everyone always wants to just find as argument. Secondly I really don’t buy into defensive sabermetrics or any sabermetrics for that matter a lot of them are mere estimations.
So just forget it. Dye had a nice career theres no arguing that. Excellent may in the eye of the beholder. Take me for example I think Doug Mientkiewicz is the best player ever just because I love the guy. Is he actually the best player ever? Yeah he is.
yea my problem is that i’m too serious all the time. definitely
and i’m sure your understanding of defensive metrics is bulletproof, but +/-, total zone, and uzr all hate the hell out of jermaine dye’s defense so i have no idea what your argument could be. even if you thought their margin of error was wide, jermaine dye would still rate a defensive liability
And if your team signed Jermaine Dye for his defensive abilities, well…
The problem with sabermetrics and all the fancy numbers is that it takes the human element out of the game. Baseball players spend more time with their teammates then they do with their own families over the course of a calendar year. Clubhouse presence, leadership, etc. are not characteristics that can be quantified. But any team that wins needs guys that fill those roles, and Dye did that for the White Sox.
There was no question that his abilities were slipping, and if he truly wanted to continue playing, then he missed the boat by misjudging his value. But, in my opinion, after watching him play for my team on a daily basis, he was excellent.
the problem with people who don’t understand statistics is that they think their own limited experience watching an extraordinarily small percentage of baseball games played will suffice as a basis for drawing sound conclusions about how the game is played
you are of course entitled to your opinion based on watching him play for your team on a daily basis and you are encouraged to share it, but you should not continue to think that yours is a more meaningful sample than the massive amounts of data accumulated and analyzed by “fancy numbers” and sabermetrics
for all the heat users of advanced stats take for seeming snotty or arrogant, thinking your own private limited experience is more significant than all of the data organized & analyzed objectively is absolutely the height of arrogance
having said that, dye hit quite well 3 out of his 5 seasons in chicago (wOBAs of .361 .417 .376), so it makes sense that you’d remember him fondly
still, for all that he produced with his bat, he gave away almost as much with his defense. had he served as DH it may well have been a different story. but offensive runs produced are not any more valuable than defensive runs cost
His gold glove from 2000 would like to argue with you.
his gold glove from one year would like to argue with me about his career performance? makes sense
and we all know what a pristine measure of defense gold glove awards are. sabermetrics are flawed, see, but what a bunch of guys who definitely didn’t watch every play or even every game of the season (and couldn’t effectively quantify what they’d seen even if they had) say? now that’s reliable
The biggest problem with sabermetrics on defense is that unlike most offensive stats, almost all defensive stats are based on someone perception of a play not an easily verifiable number. While still being useful it doesnt have the same strengths that make offensive numbers so important. So for Jermaine dye, he started his career as a fair to good fielder then declined and steeply declined in the end of his career, but what sabermetrics doesnt measure is his arm did make up for his limited range somewhat. No he wasnt a good fielder overall in his career but at one time he was. BTW his golden glove was a deserved one. Maybe he wasnt the greatest but it wasnt a Jeter situation.
“almost all defensive stats are based on someone perception of a play not an easily verifiable number.”
it’s not nearly as wide open as you suggest. yes, someone has to see what happened and write it down. but there is video available and the margin of error in terms of discerning zones is not very large. you are about to lose this excuse anyway as fieldf/x finally takes over
“hat sabermetrics doesnt measure is his arm did make up for his limited range somewhat”
seeee, i wanted to believe you really did know things about defensive metrics, but then you said this. we absolutely have stats for a player’s arm and how he uses it and they are accounted for in the metrics
“BTW his golden glove was a deserved one.”
love this. are you going to make this case without reference to a stat? or are you going to use the stats that are useless when they don’t favor your preconception? if you’re not going to use a stat, what are you going to tell me?
the problem with defensive stats, see, is that they’re based on faulty human perception. see? but let me tell you, you know, from watching dye play. he deserved that gold glove!
See you are being the arrogant one. You assume that I do not know anything about stats and do not find them useful. I do in fact find them very useful, but they are a tool to evaluate players not a means in themselves. You seem them as the end all be all. You see Jermaine Dye’s name and then you go look up his defensive stats on website, which is what you should do when evaluating a player, but you have only done half your work. You think that you can really only be accurate when you take all human element out of the game. You tell me I wrong because someday you will have a way to tell if they are completely accurate. You tell me I wrong because we have video and people check others work, but how often do they actually check it. “”what sabermetrics doesnt measure is his arm did make up for his limited range somewhat”
seeee, i wanted to believe you really did know things about defensive metrics, but then you said this. we absolutely have stats for a player’s arm and how he uses it and they are accounted for in the metrics”
You used this quote to “show” how I dont understand sabermetrics. Did you ever think that perhaps I was referring to the simple fact that runners, knowing a players arm strength, perhaps you do not try to run on them, dont even attempt, know that they are capable of throwing them out. That is a type of stat that sabermetric defense doesnt understand. the human element. I doubt you even understand the complexities of sabermetrics instead you just throw out numbers that others create and pretend that you know what they mean. I freely understand that the lack of a degree in stats hampers my understand of many elements, thats why I dont pretend to be an expert who fully understands it. Know what a stat measures is different than understanding what it involves and why that is able to tell you what it does. You are just as arrogant was the person who makes claim because they saw it. Saying I know this because I looked it up on the internet and you dont because you didnt is arrogant.
“You assume that I do not know anything about stats and do not find them useful.”
no. exactly wrong. i literally just finished saying that i had assumed you did know about them, and then you said something that suggested otherwise
“You used this quote to “show” how I dont understand sabermetrics”
no, i didn’t. you are being defensive. what you said suggested that you were unaware of what defensive metrics included, which would strongly suggest you are not as familiar with them as you seemed to be
you are now elaborating on what you meant, but what you meant was not clear at all before now. and your elaboration, actually, hasn’t clarified at all that you do know what you’re talking about:
“runners, knowing a players arm strength, perhaps you do not try to run on them, dont even attempt, know that they are capable of throwing them out. That is a type of stat that sabermetric defense doesnt understand”
srsly? i will continue to give you the benefit of the doubt that you understand defensive metrics and have simply misspoken twice in a row. but do you really think the stat runners held does not exist? and if you know it exists, what are you talking about?
I understand that it exist it just doesnt necessarily, in my opinion, translate well into understanding the value it has on defense. I dont think it translates well into a runner taking less bases compared with another player. These are the things that are objective and are hard to quantify. My entire original argument is there are things that you cant really put into stats no matter how hard you try and how accurate you are. Alot of defense is mental and there is no stat for that. Read even when uzr says about fielding, they freely admit there is a ton of leeway compared with offensive stats
yes, there is leeway. i myself have literally already said that in this thread. you are trying to characterize that leeway as mysteriously but enormously significant. i am requiring you to be more precise and justify those characterizations if you want to toss aside or minimize statistics in favor of “we saw him play, he good”. again, i’m not the one making unusual claims
specifically, why do you say the objectively correct number of baserunners that did not advance on a fielder “does not translate well into understanding the value [holding baserunners] has on defense”?
and: “I dont think it translates well into a runner taking less bases compared with another player”
what does that mean?
In that specific instance I mean a player on first taking one or two bases and why he choose not to. Or something else of that nature. You seem to take one specific example out of my entire argument instead of reading it as a whole. My entire point is that defensive stats are much more weighted to a humans opinion of an action rather than an objective fact. i completely argee that people who argue one player is better than other offensively and consistently overlook his useful stats are ignoring reality but in defense it is a different factor. Many of the defensive stats, such are arm strength, are truly debatable. Fan Scouting Report is a great example of this, and in my opinion, the limitations of this arent necessarily the fans. it is the subjective nature. My argument with defensive sabermetric is it is objective analysis that depends largely on subjective data.
how does the reason he chose not to effect the value of the player’s throwing arm? possible outcomes:
a) he does not attempt to take third. a runner is held
b) he does attempt to take third and is thrown out. a runner is not held, but a runner is thrown out and an assist recorded
c) he does attempt to take third and succeeds. a runner is not held and an out is not recorded
with several seasons of data establishing trends which can be contrasted with other players experiences, what important information about the value of the player’s arm is missing here such that it would significantly change the overall value of the player’s defense?
The fact that he probably could have taken third and didnt. The fact that he got caught running stupidly. The fact that it was a close play. etc. Do you see what Im trying to say.
of course, yes. what i don’t see is the connection between that and any claim that these defensive metrics are significantly lacking
especially when the alternative is casual viewer consensus, which is massively less precise
in other words, my position is not and has never been “defensive metrics are perfect”. my position is and has been that they are the best tools available by a mile and absolutely cannot be undercut by claims like “i watched him play and he looked good” or gold glove voting
I just believe that they should be balanced by what an observer sees. Not replaced.
observer experience is massively less reliable in a dozen ways. if they are to be balanced, they have to be balanced at a ratio of like 1000:1 or more
how can you not see that the very same criticisms of the advanced metrics apply to the casual observer’s observation, only a hundred-fold more extremely?
what sense does it make to say we need to balance an marginally imprecise metric with one that is monstrously less precise
Because I see a difference, in certain circumstances, between something being statistically less significant and something actually being less precise. Example A umpire has a 98 percent chance of calling strikes and balls correctly, I see a ball bounce in the dirt and it is called a strike, statistically, I would be fighting the odds to argue against it but in that instance I would be correct to do so. Stats are only as good as the raw data put into it. You feel that data is generally good. I find problems with it. I still use it but I do place a greater than you on what I observe.
response at bottom of thread
“I doubt you even understand the complexities of sabermetrics instead you just throw out numbers that others create and pretend that you know what they mean.”
this is based on nothing but your own private hope that i’m an idiot. there is no evidence anywhere that suggests that i do not “understand the complexities of sabermetrics” (a phrase which itself strongly implies that the speaker considers statistics a mysterious and un-understandable field)
to recap: someone took a position about a player’s performance. i questioned it, and provided statistics that appear to contradict the original position. several people immediately criticized the very field of advanced statistics itself, favoring instead their own tiny samples of first-hand experience. i insisted that the only way to think that is valid is to in fact know very little about statistics in general. and then i was accused of not understanding the field
brilliant!
I would also suggest you read one of the dozen of good articles that explain far better than I am able to difficulties of measuring defense even with advanced stats. Dont take my word for it.
why don’t you recommend some specifically that you think i haven’t read
Because, without knowing what you have and have not read, I really cant. You say sabermetrics defensive stats are solid, many many people, including sabermetics scholars disagree. Read some of the criticism. Many of the point I have brought up are repeated over and over.
you keep saying that and it is useful in that it makes you seem like you’ve read things that i haven’t (who’s arrogant? lol). but i doubt that’s the case. and if it is, i definitely want to learn the things i haven’t learned yet
so please don’t deprive me. you are acting as if you know something i don’t know. just show me an article that explains what you think i don’t know. then i’ll be able to tell you either hey, i didn’t know that. or hey, i did know that – and you forgot to consider this
everyone profits from that. no one profits from what you’re currently doing
I believe that you cannot post links but in baseballnewshare there is a an article, only a place to start, titled something like the limitations of defensive metrics. it is a place to start and it specifically mentions the problems of objective analysis using subjective data. Also due to the fact I was recently banned, long stroy, I dont want to push i and list too many outside sources. This is seriously not a cop out, but a real fear, despite my tone at times I have enjoyed this discussion with you.
haven’t found it through google so far, but i’ll keep looking
the subjective-objective thing is not something new to me, though. you appear to be zooming-out your criticisms (if we can call them that) to a level of generality that simply observed the shortcomings of defensive metrics. this would be enough if i had ever taken the position that they were flawless, but i haven’t
again, the only thing people criticizing the metrics have suggested as a superior point of reference is way, way messier and more subjective than any metric. in other words their suggestions are guilty of the same shortcomings, but to a much greater degree
this is why the objections make so little sense
was the article using jacoby ellsbury as an example? because this article features a criticism of [the use of small samples of the metric] masquerading as a criticism of [the metric in general] =
Again, if you accept the premise that Ellsbury is what he is on defense, you have trouble accepting that his UZR was +3 (3 runs above average) in 2008 and -18.6 (18.6 runs below average) in 2009.
i mean seriously? this is a problem following from the nature of defense in baseball, not the way it’s measured. the equivalent of a year’s worth of offensive data is about 3 seasons of defensive data. that’s not a shortcoming of a stat; that’s a function of the way the game works
That was the article, I think our basic disagreement isnt over the usefulness of sabermetrics but rather you feel that it is good enough not be be overruled for the most part while I feel it is flawed enough to be overruled be reasonable observation. The point of the article i was referencing is the messiness of using objective stats gotten by means of subjective analysis. When I see two players that I regularly see and one player is rated better defensively than the other, not by a significant margin, and I disagree and I have a large enough sample size I will go with what I see. Not because my gut is any better but because stats are a tool to help us not a definitive rating. I think the difference is our outlook, we read the same article and got two different main points from it.
“I feel it is flawed enough to be overruled be reasonable observation”
what is “reasonable observation”? the point is that the most reasonable observation available takes the form of these zone ratings because it is so comprehensive. favoring the casual observer’s limited experience doesn’t make any sense at all because it is flawed in the exact same way as the thing you are favoring it over, only moreso
the article is bothered that a player who is thought to be good defensively can have a two year sample of defensive stats that rates him poorly one year and highly the other. this cannot be written off as subjective messiness, though. that is not what it is
what it is is standard volatility within a small sample. it is exactly the same thing as a player having a two-month slump and then hitting .400 for two months. no one is outraged when this happens about the objectivity of batting average when this happens
sample sufficiency is on a different scale for defense because of how the game works (numbers of opportunities and the physics of playing defense). in other words, two to three months of offense = one year of defense
outlier data samples are not evidence of an imprecise metric. the author of the article is confused. objectively
reposting this comment from below because it is the heart of the issue and you should have room to respond to it if you want:
“I feel it is flawed enough to be overruled be reasonable observation”
what is “reasonable observation”? the point is that the most reasonable observation available takes the form of these zone ratings because it is so comprehensive. favoring the casual observer’s limited experience doesn’t make any sense at all because it is flawed in the exact same way as the thing you are favoring it over, only moreso
the article is bothered that a player who is thought to be good defensively can have a two year sample of defensive stats that rates him poorly one year and highly the other. this cannot be written off as subjective messiness, though. that is not what it is
what it is is standard volatility within a small sample. it is exactly the same thing as a player having a two-month slump and then hitting .400 for two months. no one is outraged when this happens about the objectivity of batting average when this happens
sample sufficiency is on a different scale for defense because of how the game works (numbers of opportunities and the physics of playing defense). in other words, two to three months of offense = one year of defense
outlier data samples are not evidence of an imprecise metric. the author of the article is confused. objectively
I am not arguing the point in the article that you are arguing against. By reasonable observation I dont mean favoring the casual observation over the sabermetric stat in every or even most situation. I mean taking it over when what sabermetric is telling you doesnt match with reality. Example. Alexei Ramirez according to sabermetrics was not a good defensive shortstop on 2009, watching him you could see his style of play was not due to a lack of ability but rather various other problems. I could see his talent, but according to stats he really wasnt there. The next year he was tremendously better, but really did not play much different. That is what I mean. Was he bad one year and then great the next. By style of play yes, but as a player, not really his talent was all there it was execution, but according to sabermetric ifi argued that he was going to become one of the better shortstops I would be incorrect, and I did argue this, was I jsut lucky or did I use observation to overcome a miscalculation.
“Alexei Ramirez according to sabermetrics was not a good defensive shortstop on 2009, watching him you could see his style of play was not due to a lack of ability but rather various other problems. I could see his talent, but according to stats he really wasnt there. ”
oh come on, man. you know the metrics don’t claim to measure talent. they claim to reflect actual performance, which is volatile in the same way offensive is (think slump)
you and the article both write off volatility in the metric to some error in observation. but there is no evidence of an error. there is evidence of standard variance in a small sample. i went over this already
“but according to sabermetric ifi argued that he was going to become one of the better shortstops I would be incorrect”
come ooonnnnn man. anyone literate in defensive metrics would tell you that you couldn’t project jack from a one year sample; we’ve been over this. you are now straw-manning the defensive metric camp and i am not digging it at all
this isn’t an instance of an observation overcoming a calculation at all. it’s a preconception (ramirez is talented) contradicting a small sample. again, the smallness of a sample is not an indictment of the metric. pujols hitting .200 in july isn’t an indictment of batting average. ramirez not getting to as many balls (or not having as many opportunity, or getting particularly unlucky with the type of opportunity he got) for one year is not an indictment of uzr (for example)
Of course it doesnt mean it is trash but it means in this one particular instance that sabermetric did not do one of the things it is absolutely best at, I know it is a small sample size, but this is an example of where one’s observation just worked better, it was not good at predicting future production where it normally cant be beat in this area. You take my criticisms, and they are criticism no matter how you try to deride them, of a system than I general agree with and use to try to make it seem like I am one of those people who just know in their gut who is better. I know that fact that urz didnt predict accurately the future success of a player isnt an indictment of the system, but it is a place where an observation was more successful. Teh fact that sabermetric is limited by small sample sizes is in itself a limitation of the system. Yes Pujols batting .200 doesnt mean he isnt a great hitter, but will you admit that perhap observation might be able, such as someone spotting movement caused by a strain or sprain, as useful information that sabermetrics would not cover. You also miss the point in another area, “you and the article both write off volatility in the metric to some error in observation. but there is no evidence of an error. there is evidence of standard variance in a small sample. i went over this already” See standard variance or standard deviation does not mean a small difference or a large difference. It is just the standard amount that data will deviate from each other. This deviation can be very large and the amount of error can be great. You refuse to admit that faulty data can be introduced into the system or at the very least refuse to admit that enough faulty data can be introduced to cause the data to deviate enough where standard observation can become the superior form of analysis. What do you think causes deviations? It is more than just normal changes in a player. What do you think of Jeter’s defense changing when Cano is introduced? Is that not a variable that is not considered?
I kinda gets down to what is more important, overriding the data at times and using your own judgement or being statistically more likely to be correct.
I know it is a small sample size, but this is an example of where one’s observation just worked better, it was not good at predicting future production where it normally cant be beat in this area.
it didn’t pretend to be! people who do not understand it may have misused it, but that is not a shortcoming of the stat. again, what you are saying is exactly analogous to saying there is a problem with batting average because idiots try to use a two month sample to assess talent or make projections. the creators of the metrics have not and will not claim that small samples will have high predictive value. this is not a shortcoming any more than any other regular, non-saber stat
I know that fact that urz didnt predict accurately the future success of a player isnt an indictment of the system, but it is a place where an observation was more successful
i have been and still am perfectly happy to agree that an inadequate sample of a defensive metric is not more predictive than casual observation
teh fact that sabermetric is limited by small sample sizes is in itself a limitation of the system.
knock it off. this is a limitation of the freakin’ physical world; it isn’t unique to saber stats at all. name one stat to which it doesn’t apply. name one thing in life to which it doesn’t apply
You refuse to admit that faulty data can be introduced into the system or at the very least refuse to admit that enough faulty data can be introduced to cause the data to deviate enough where standard observation can become the superior form of analysis.
if you can show me one place where i’ve refused to admit this, i will shutup forever. it’s been a long thread, but please don’t strawman now
What do you think causes deviations? It is more than just normal changes in a player. What do you think of Jeter’s defense changing when Cano is introduced? Is that not a variable that is not considered?
of course it is. i didn’t say that the only source of variability was a player’s own fluctuations in performance. i actually already said that a player like ramirez could have a varied rating based on no change in his physical ability but instead the type of balls that were hit into his zone within the sample for example
this is exactly why sample size is so critical: the nature of defense in baseball. (as i said already.) there are only so many opportunities a player gets in a game, a series, a season — and [groundball to short] (for example) can take dozens of different forms
in the same way, pujols could be just as on-the-money during his slump months but have faced an extraordinary tough series of pitchers. this wouldn’t be reflected in the data, but we’re not throwing out batting average and we’re not favoring our even less reliable observations over those offensive numbers are we? so why are you proposing it with defense? still makes no sense
one more time: one full season of offensive data is equivalent to about 3 seasons of defensive data. be vigilant about not confusing a sample problem with a metric problem
You just dont understand, I have shown you over and over again ways that observation can be used in place of sabermetric. That was my entire point that ones observation can be used in place of sabermetrics when sabermetric doesnt fit with reality. A sample problem is a metric problem when you are telling me that I cannot decide that sabermetrics is wrong or inaccurate. You say well that doesnt apply. Ok if it doesnt apply then obsevation, in that instance, is superior. You refuse to see any of the limitations of sabermetrics where observation is superior. I am not nor have i ever argued that in general observation is superior to sabermetrics I am just saying that there are times, even with large enough sample sizes that it is.
“A sample problem is a metric problem when you are telling me that I cannot decide that sabermetrics is wrong or inaccurate.”
no. the problem is that you presume the mere existence of a stat in small sample form is making some kind of claim with which you could disagree. it isn’t. it’s a record. you can and should disagree with the idiots who try to make claims supported by that insignificant record on the basis that it is insignificant (too small, not on the on the basis that the record is flawed
“I am not nor have i ever argued that in general observation is superior to sabermetrics I am just saying that there are times, even with large enough sample sizes that it is. ”
you have provided 0 examples of this problem within a sufficient sample and therefore have offered no evidence to support your claim. what am i supposed to do about that? so far we have only discussed ellsbury’s insufficient sample and ramirez’s
You make it impossible to do so. You refuse to accept any analysis that is not backed up by sabermetrics so how could I give you an example you just say no, the sabermetric stats dont agree. The entire argument is over if you can use observation in place of sabermetrics. You claim that I can only prove it using evidence with a large sample size, but you refuse to acknowledge it, the evidence, unless the stats back you up. See the catch 22? I use ramirez to show how you cannot use your data while I can use mine and be correct. I do so not to show that your data is bad or false just that mine can be accurate while your is unable to be.
“You claim that I can only prove it using evidence with a large sample size, but you refuse to acknowledge it unless the stats back you up. See the catch 22?”
not at all. i am asking you to prove a problem exists by citing a significant sample of a stat, because the problems with the small samples are attributable to the smallness, not the stat
i am not requiring you to use anything in particular. just show me a problem that does not amount to a small sample. so far you’ve done ellsbury and ramirez – both small samples
You refuse to accept any evidence that do not fit within your ever changing conception of what is acceptable evidence.
anyone who reads the thread can confirm that that is not true so i am not feeling compelled to argue about it
END SCENE
“this isn’t an instance of an observation overcoming a calculation at all. it’s a preconception (ramirez is talented) contradicting a small sample. again, the smallness of a sample is not an indictment of the metric. ”
this is completely false. It is not a preconception that Ramirez is talented. I did not start the year thinking he was talented. It was from watching him, even though he struggled, that I felt he was talented. This is one of the limitation of sabermetrics. You dont really know why I felt he was talented. You only have your stats to guide you and they are unable to tell the complete story. Watching him, and I know you might not be a white sox fan so never caught him playing enough times, you would have been able to form an opinion of why he was not preforming despite the fact that there clearly was talent. the poor preformance is caught in the stats without any contextual information. Context is everything, that is why RBI are not a good indicator of performance. Can you not admit that it this particular instance, i dont care how small the sample size is, that observation can be used correctly in place of sabermetrics?
there are so many reasons that doesn’t work. the limitations of the human mind and memory are foremost among them. it is simply impossible to rely on one’s own virtually standardless impressions of watching a player perform. even if it could work within one’s own mind, aligning stats with first-hand impressions, it could never function effectively in discussion
you can’t convey that amount of data from memory with any precision at all; you can scarcely process it yourself. this work is sourced to on-paper and in-computer records and computation precisely because we can’t retain it ourselves
so again we have the same problem: even positing imprecision in the stat (which is constructed from the exact same process of observation, only more delicately and with the guidance of an objectively conveyable rubric and other tools), the proposal that we default to a method which is hampered to a much higher degree by the same exact limitations is ludicrous and self-contradictory on its face
i don’t think you’re dumb. i think you’re just misapprehending the nature of the events being measured and the process by which they necessarily must be measured
Ok I am done. I have explained myself enough. I have proved my point to pretty much anyones satisfaction. Careful observation can override sabermetric in certain cases. To go further is to get into a philosophical argument about what it means to be human and the limitations of humans which is just silly to argue here.
fundamental physics and the mechanics of observation very obviously underlie statistical models of physical events like the ones we’ve been discussing at length, so pretending they are somehow tangential is a pretty huge cop-out f you ask me. but i am satisfied anyway
this began with you citing baserunners not running on a player’s arm as an example illustrating some limitation of saber stats, despite the fact that a stat (runners held) measuring that exact thing exists. i have chased you around the playground since then to get you to show me even one example of the shortcomings you repeatedly allude to, only to hear you strawman defensive metrics with sample size confusion
still don’t think you’re dumb, just misunderstanding the nature of statistical models and their construction
You refuse to listen to any example. You just dismiss everything as being already covered even though some things clearly arent. I have given you example after example and you jstu respond with no, that doesnt work even though it does. Does defensvie sabermetrics account for conditions such as sun, wind, rain? Did you know at certain times at the Cell it is almost impossible for the batter to clearly see the ball being pitched? You have no response for these other than sample size which does not cover all these situations. What is, to you, the value of sabermetrics?
i think it’s funny that you just throw out allusions to things like variable weather now for the first time and act as if i refused to respond to them at some imaginary prior point. not funny in a good way tho
“What is, to you, the value of sabermetrics? ”
best tools available for analyzing the way game i (we?) love is played. by a wide margin
like 10 WAR wide
I like how you refuse to respond to variable conditions, the type of things my entire argument is based on. By the value of sabermetric I meant specifically, what are they good at? What are they bad at? Because apparently I dont understand how you understand and use them to justify your points. They arent good in the short term, but I cant use what I observe because that is just my preconceptions. By saying they are not good in the short term I am not trying to knock them Im trying to understand. I understand that they are amazing over a career at comparing players, but if they cannot be used in the short term how do we rate players? Apparently we cant use our judgement, because that is too subjective. OR is that your argument we have to until something better comes along? Because either way my point has been made. There are times where judgement is better than stats.
response at bottom of thread
guy had a major injury and he was no liability in Rf in his early years , was he the best no. was he better than most yes. and he has been more consistent than jones , bradley ,both of them only had like two good years , dye had about 5 , and 325 hr means alot
Dye, thank you for acknowledging what the rest of us already figured out.
Thanks for the memories of 2005 JD!!! We will never forget you for that year!
Thanks for the memories of 2005 JD!!! We will never forget you for that year!
Obviously.
Hope he finds a good post baseball job like Gary Sheffield (that IS Sheffield in the last third of that Rebecca Black video, isn’t it?)
This gets me thinking:
Have either Barry Bonds or Roger Clemens officially retired yet, or are they still holding out hope?
can you play baseball from prison?
WHY WOULD I KNOW THAT?! I AM AN UPSTANDING CITIZEN AND YOUR INSINUATIONS ARE COMPLETELY OUT OF LINE!!!
Probably not. I figure the bats are probably a potential cuase for concern.
Ricky Vaughn played and was scouted from prison. So with that precedence I think you can play baseball from prison.
Califonia Penal League count?
Best of luck to Dye. Trading him to the Royals was Scherholtz’s third worst move with the Braves, only behind the Justice and Teixeira trades.
At least the braves got some sort of value for Justice, albeit not enough. I would rank this as his #2 all time worst move, with the teixeira trade being worst of course. Michael freaking Tucker! What a waste of an at bat this guy was! To think we could have had Dye in the lineup for all of those years. We wouldn’t have had to trade wainwright for drew or picked up sheffield. Granted those guys had excellent years as braves, but the cost to pick them up could have been put elsewhere. Michael Tucker! Ugggh!
At least the braves got some sort of value for Justice, albeit not enough. I would rank this as his #2 all time worst move, with the teixeira trade being worst of course. Michael freaking Tucker! What a waste of an at bat this guy was! To think we could have had Dye in the lineup for all of those years. We wouldn’t have had to trade wainwright for drew or picked up sheffield. Granted those guys had excellent years as braves, but the cost to pick them up could have been put elsewhere. Michael Tucker! Ugggh!
Should say, “Jermaine Dye to remain Unwanted”
Jermaine Dye, we hardly knew ye.
13 years in the league, I’d say that’s quite a long time if you ask me.
Always liked the guy. Too bad he couldn’t have gone out on a better note.
Dye gave me a lot of special memories, especially in 2005. He had some special seasons, and was a good player while healthy. I personally think he was a bit overrated because of how bad his defense was (he was barely above replacement level in some years he was considered to have a good year), but he was still a fun player to watch and a key offensive player in years he was healthy.
It’s his own fault no one wants him now, he passed up many deals last year. Did he really think he was going to get the 1 yr $10mm dollar deal he was asking for? yeah right!
Very good ballplayer that was under-appreciated for much of his career. Wish him nothing but the best in the future.
“Because I see a difference, in certain circumstances, between something being statistically less significant and something actually being less precise. Example A umpire has a 98 percent chance of calling strikes and balls correctly, I see a ball bounce in the dirt and it is called a strike”
how did we establish that he has a 98 percent chance of calling strikes correctly? =P
seriously, though. that’s what’s going on here. in all cases, the zone rating observers are observing the same thing the casual observer is. but unlike the casual observer, they have a useful and precise rubric into which they record their observations
there is nothing precise, reliable, or even measurable really about the casual fan’s observation. even if we posit a 2% margin of imprecision or whatever with the zone observer like the umpire, we should always favor the the most precise metric — with any sample, but especially with large ones
in other words, what you’re omitting from your example there is that you (in the example) are way the f less reliable than the umpire at 98%
so again, if we are to strike a balance (and maybe we should), the balance has to be tipped to an extreme degree. and we can never really take from this that advanced defensive stats are given too much weight over the casual fan’s opinion in a case like dye; that literally makes no sense
They arent good in the short term
they aren’t reliably predictive in small samples, you meant to say. because you were paraphrasing me. and that’s what i’ve been saying.
and this fact — they aren’t reliably predictive in small samples — is not unique to saber stats at all in any way. so it’s way past time to stop pretending it is. a casual observer could count put outs for an inning and that sample wouldn’t be reliably predictive either. the “problem” has nothing whatsoever to do with sabermetrics
but I cant use what I observe because that is just my preconceptions.
back to strawmanning, awesome! the only thing i’ve ever said about casual observation is that it’s a goofy thing to favor over [sabermetrics in sufficent samples] because it is stricken with the exact same issues you object to — to a much much higher degree! it is inferior by your own logic. and your logic here is actually sound
By saying they are not good in the short term I am not trying to knock them Im trying to understand. I understand that they are amazing over a career at comparing players, but if they cannot be used in the short term how do we rate players?
scouting? yes, observation. remember when i already said this earlier? i’ll quote it: “i have been and still am perfectly happy to agree that an inadequate sample of a defensive metric is not more predictive than casual observation”. still holds
Apparently we cant use our judgement, because that is too subjective. OR is that your argument we have to until something better comes along?
bingo!
the problem is that your point at first introduction was not limited to “hey there is at least one scenario where saberstats are not superior to casual observation”. not even close. you introduced the subject in the context of an ongoing discussion of dye’s career performance, which you finally just now acknowledged advanced stats are “awesome” and superior at measuring
you’re right. they are. which is why i used them in the first place to question someone’s casual observation. and that’s all that should have been said!
but you intervened. all for what?
i’ll tell you what for. my bankroll. ’cause even though i enjoyed this exchange more at the beginning when it was fairly good natured, i played much better poker when you were annoying me later on. for this i thank you =D
i’ll check the thread tomorrow if you have more you need to say. but i think we just came full circle
But I was not arguing that Dye’s career defensive performance was excellent or that his stats over his career were not poor for fielding. If that is what you though Im sorry. We were arguing if his golden glove was deserved. I was not nor was I ever arguing about his career. If you go back I was making no such claims anywhere in my arguments. You say it couldnt be because his career performance for fielding is poor or that in that year his sabermetric stats were not good enough for a golden glove. Either case, golden glove is a one year measure of talent, or what writers decide, and as you have put it a single year is not enough evidence to make any decisive conclusions using sabermetrics. This is a quote, edited for length of the comment that spawned our discussions, “So for Jermaine dye, he started his career as a fair to good fielder then declined and steeply declined in the end of his career. (Edited material) No he wasnt a good fielder overall in his career but at one time he was. BTW his golden glove was a deserved one. Maybe he wasnt the greatest but it wasnt a Jeter situation.” See what I said here was nothing about sabermetric measuring his long term defensive ability. I was defending that fact that he had a good year with the glove. I was angry because someone who didnt watch him play alot, small sample size, used a year performance, small sample size, from the internet to determine if they had a good fielding year. Even by you own acknowledgement you cannot use the data in that manner. If you are arguing that it seems unlikely that in one year he would be able to play so much better than his norm well then things do happen like this.
i’m really not gonna go back through the thread, but i’m happy to apologize if i attributed a position to you that you did not actually take. having said that:
“golden glove is a one year measure of talent, or what writers decide, and as you have put it a single year is not enough evidence to make any decisive conclusions using sabermetrics.”
aye, dios mio. a single season of defensive data is insufficient to draw conclusions about a player’s ability. it IS sufficient for conclusions about his performance that year
again, this is perfectly analogous to offense. a two month sample isn’t enough to draw a conclusion about ability, but it absolutely is sufficient for conclusions about those two months
in both cases the data is valid and accurately reflective of past performance. you just need more data to reliably draw sound conclusions about future performance
someone who didnt watch him play alot, small sample size, used a year performance, small sample size, from the internet to determine if they had a good fielding year. Even by you own acknowledgement you cannot use the data in that manner.
you misunderstand. the data tells you what happened that year, literally. (and that’s what the award is based on – what happened that year.) the data is useful to this end, but because of the variability (note: not imprecision) in the quantity and especially the quality of defensive opportunities year to year, a single season isn’t reliable in predicting next year (i.e. determining a player’s ability) the same way two months of offense doesn’t reliably predict the rest of the season
i think we understand each other’s positions now and there isn’t much left to say. you don’t have to accept it, but i’ve made my case as clear as i’m willing to
baseball is back now anyway. it’s all good
each one of your comments have been longer than this article. the dude played. now he doesnt. he could hit. he couldnt field. no need to be all mark twain about it.
Too bad theres not a way to look at a players body of work defensively and give some sort of numerical rating to it.
That was so funny after that marathon thread I mighta peed myself a little.
He was a solid player, one you can depend on. Good luck.
I clicked on this headline just to read the comical comments and not the article, good grief dye your two years too late with this.
haters gonna hate
also dollars to donuts i made more money during this conversation last night than you did all of last week
also also today is opening day part two so my good mood is pretty much invincible right now
I love hearing how successful people are..whether they really are or not. Its so tasteful
dude claimed i was wasting my time here. the exact opposite is true. he has since edited his comment to say something else entirely
“that’s not tasteful” is a sick burn tho
No player besides Dye has ever had a better final season of their career — considering these offensive stats: .250 average, 27 HR, 81 RBI. A play index search of inactive players in their final seasons with .250+/25+/80+ turned up exactly ONE result: Dye.
Barry Bonds’ last season was far better. Less RBI’s but also 160 less AB’s. Half the K’s, 90 more BB’s, OBP +.140. Nobody can touch Bonds, not even close.