Sabrheads and what wrong with them!

[gRYFYN1]

An article posted on THT about where Sabrmetrics has run afoul.

Bottom line: the new boss is too much like the old boss. When sabermetrics grew in popularity, the old guard—who felt they were the holders of the Holy Grail—derided the new guys as ignorant since their heads were always stuck in a spreadsheet and they never watched a game. Now the new guys are the old guys who call out the ignorance of those who use observation as part of the basis for reaching conclusions when that conclusion doesn't fit neatly into the foreordained sabermetric framework.


http://www.hardballtimes.com/main/article/i-was-born-a-ramblin-man/

[MeetDaMets]

interesting article.
liked this paragraph :)

My biggest problem with sabermetrics is the same as any ideology that causes some adherents of it to think they have in their possession the Holy Grail of knowledge. The attitude of many who study it is that baseball is a part of sabermetrics rather than sabermetrics being part of baseball, ergo everything that happens within the game can be explained within a sabermetric context and that which cannot be is a fluke of sample size, random variation and “luck.”

[HoustonGM]

Good article

[dickay]

I didn't read the whole article but agree with the premise. It's really not fair however to paint with a broad brush everyone who uses sabre stats as surely there are those who incorporate the games variables and for lack of a better way of putting it, "they believe there is more than one way to get to heaven."

I couldn't tell you much of anything about sabremetrics, and I do find those who support it affirmatively annoying sometimes. But, I do find them a valuable resource and feel they are in most cases more accurate than others including myself because they, in most cases, do the research.

[MeetDaMets]

in the paragraph i quoted , author uses the phrase "some adherents".

i do not perceive that as a broad brush stroke.

[HoustonGM]

in the paragraph i quoted , author uses the phrase "some adherents".

i do not perceive that as a broad brush stroke.
Yeah, I was gonna say the same thing. I don't think he's "stereotyping" at all, but rather speaking about a certain population with the "sabermetric community".

[MeetDaMets]

yeppers :nod:

[dickay]

Yeah, I was gonna say the same thing. I don't think he's "stereotyping" at all, but rather speaking about a certain population with the "sabermetric community".

As I had stated however, I didn't read the whole article, just the premise and the title of the thread which is "Sabrheads and whats wrong with them!"

Thats 'whats wrong with todays media'. They post misleading or broad slanted large font headlines, pushing an agenda while knowing full well the full story/truth won't be relayed until paragraph 7 lol.

Of course, one could say 'thats whats wrong with todays voters'. Not fully informing themselves of an issue, only reading the BS headlines fed to them by a media with an agenda.

Not saying that is what he was doing when making this thread.....just throwing that out there as an observation I made after misinterpreting the title of this thread as an attack on all sabr users.

[MeetDaMets]

Thats 'whats wrong with todays media'. They post misleading or broad slanted large font headlines, pushing an agenda while knowing full well the full story/truth won't be relayed until paragraph 7 lol.

Of course, one could say 'thats whats wrong with todays voters'. Not fully informing themselves of an issue, only reading the BS headlines fed to them by a media with an agenda..

LOL nice, works fer me.:D

[HoustonGM]

Yeah. The title of this thread is misleading.

[gRYFYN1]

Yeah. The title of this thread is misleading.

Not really, the article refers to the people who preach to the Orthodoxy of Sabr as the end-all-be-all of player evaluation. I suppose i could have titled it " Some Sabrheads and some of the things they have sorta wrong" but its just the title I really didn't think it was necessary to over quantify a the title when there was an entire article that follows it pointing out all the quantifiers.

[HoustonGM]

Well, it's just that while you're correct about what the article is about, the title makes it seem as though it applies to all people who use sabermetrics, when the people the article talks about probably make up a minority of "sabermetric savvy" people. No big deal though.

[dickay]

Not really, the article refers to the people who preach to the Orthodoxy of Sabr as the end-all-be-all of player evaluation. I suppose i could have titled it " Some Sabrheads and some of the things they have sorta wrong" but its just the title I really didn't think it was necessary to over quantify a the title when there was an entire article that follows it pointing out all the quantifiers.

Your title was fine. I posted that I didn't read the whole article and cautioned however that everyone who supports sabr not be labeled is all. I was 'sorta' jumped on about that, and then defended my position as to why I said that.

Confusing, I know :p:rolleyes:

[gRYFYN1]

Your title was fine. I posted that I didn't read the whole article and cautioned however that everyone who supports sabr not be labeled is all. I was 'sorta' jumped on about that, and then defended my position as to why I said that.

Confusing, I know :p:rolleyes:

well i did intentionally make the title somewhat polarizing, the point was really the article as a whole, which is actually a great perspective.

[MeetDaMets]

speaking only for myself,

if you felt sorta jumped on, it was not intended.
i was confused.
pointed out why.
is all.
like your reply
and loller :)

peace

[belial]

'I think sabermetrics has fallen into this very trap and it has begun to hinder understanding the game, since the conclusions have been foreordained before research has begun. If the evidence is overwhelming that the conclusion cannot fit within the sabermetric model, it is simply dismissed as a fluke of sample size, luck or random variation rather than considering the possibility that it might be due to human foible or the dreaded “intangible.” ' - I like this guy :)

The point of article is close to what I think. Instead of trying to fit sabr to baseball, it seems sabr is trying to fit baseball into sabr. Won't work because baseball is bigger, and always will be.

[Arctic Blast]

Really good article. He managed to explain what I have tried to, and failed to, many times.

[dickay]

In truth, *much* of the reason why many despise the sabr people is mentioned in this article. The unspoken truth is that a *small* reason why many despise the sabr people is jeolousy, as many know that those truly studying the stats have an understanding of them that makes sense but would take us a ton of time and research to grasp and come up with ourselves.

Feel free not to admit it.....but it's the truth.

[ohms_law]

Yea, I read this last night. Great article.

[SirKodiak]

Here is some of the history that led to that article by John Brattain:
http://www.insidethebook.com/ee/index.php/site/comments/there_are_so_many_problems_with_this_tht_article_from_a_sabermetric_perspec/#comments

[SirKodiak]

Comments on the article at BallHype, including the author:
http://ballhype.com/story/the_hardball_times_i_was_born_a_ramblin_man/#tab=0

[abben]

In truth, *much* of the reason why many despise the sabr people is mentioned in this article. The unspoken truth is that a *small* reason why many despise the sabr people is jeolousy, as many know that those truly studying the stats have an understanding of them that makes sense but would take us a ton of time and research to grasp and come up with ourselves.

Feel free not to admit it.....but it's the truth.

Thank you, thank you, thank you! No one is going to admit this. I can't for the life of me find a thing in that article that isn't subjective and arbitrary.

If I may (from the comments thread that apparently inspired his article),


Granted, it’s a placebo effect but you don’t tell a player to sit on a different spot on the bench during a no-hitter because the numbers show it has no effect on what the pitcher does.
Well, there is no data for something like that. It's possible that there is a psychological effect, which, if real, can be measured. Therefore quantified, therefore turned into those evil, pesky numbers. But because it's difficult to quantify doesn't mean it exists in some magical realm "beyond" numbers. And it certainly doesn't mean that sabermetricians, in the abscense of data, become blithering buffoons so desperate to apply their method that they fall for some cheap mistake like the above.


However, I have yet to see “overwhelming evidence” that the human element within the game is close to irrelevant. Baseball is a game played by individuals with various levels of talent and motivations--however talent has to be motivated, utilized properly and harmoniously for success.

Numbers don’t tell you how to do that.

I don't remember numbers ever claiming they could do that, so as a basis for criticism of sabermetricians, it doesn't fly. What they can do is explain the raw ability which is drawn upon and quantify that. David Eckstein, for example is probably cited as a guy who performs better than someone which his skill set would be expected to- for whatever reasons about personal motivation, work ethic etc. But the fact that he performs reveals itself and can be quantified statistically.

What this really is, is a guy who can't do the math, so he appeals to this "human element" which can't be quantified, looks for any and every way possible to mis-apply knowledge of sabermetrics, and instead of attacking the actual analysis they do (because he can't), he says there is some cult ideology with negative behavior that he doesn't like, then applies the "middle ground" fallacy, that ever re-occuring artifact of the worst in human psychology- "method A isn't perfect, method B isn't perfect, the truth must be in the middle!"

And from the article:


It has become an ideology, an end; when you embrace sabermetrics you have reached enlightenment in all things baseball and all that is left is a crusade to convert the ignorant masses to the new light.
That's just ridiculous and weasely. It's a classic cop-out. Instead of thinking that the other side is using reasons, hey! assume they are just a cult. Mischaracterize your opposition as some more extreme version of itself, then attack that straw man.

You could say the same about any number of well established fields, the "cult" of chemists, perhaps, who are ****-bent on making everyone accept their orthodox belief that hydrogen and oxygen come together to form water, and who tolerate no dissent. They think everything can be explained by the methods in chemistry, but chemistry can't explain the "human element" of our experience with water, so chemistry is incomplete (despite the fact that that is not what chemistry does).... and already we start inventing and injecting pretensions toward completeness into chemistry which it had never aspired after. The method is easy, and he is doing it with sabermetrics.

People know more than him about sabermetrics, so he feels bad. That sounds heartless, but if it's true, it should at least be acknowledged.

(edit: for those that would remind me that he's not talking about sabermetrics in general, I challenge the idea that what he says applies to any sabermatricians, insofar as they are sabermatricians.)

[rogue9]

Thank you, thank you, thank you! No one is going to admit this. I can't for the life of me find a thing in that article that isn't subjective and arbitrary.



add: or necessarily untrue

and you might have started to sound as balanced as his article.

[abben]

(forgive me for editing this quite a bit after I already posted it)


add: or necessarily untrue

and you might have started to sound as balanced as his article.

Right and wrong shouldn't be balanced, nor should my (or anyone's) appearance of obviously being on one side of the issue have anything to do with whether they are right. Which, again, leads to the facts. Sorry if I sound aggressive, though.

I understand. We like to think we have a "knack" for things after watching baseball for a long time, and surely many coaches make brilliant moves based on gut feelings. We like to think there is this personal or subjective thing we possess, an aspect of our individuality that lets us judge things, and we can recoil into this reservoir, speak from the heart, speak from whatever that thing is that's "truly US" and that there is no way this can be replicated or quantified. But every time we use this intuition, its in response to some experience, some fact, some thing we see (the author of the article based his thoughts not on nothing, but on "observation").

To say that intuitive reasoning can't be quantified because it's "beyond numbers" is just a misunderstanding of what is meant by "numbers"- they correspond to instances of repeated experiences and statistics merely uses induction to make inferences, based on which experiences correspond to which.

I can see the objection. A number falls short! How can a #1 truly capture the full meaning of an at-bat! The wrinkles in his shirt, the gestures with the bat! How can one at-bat possibly be lumped together with another in some counting game as though every at bat were exactly the same! BUT, the solution isn't to reject every numeric measurement we could use, but to explore what exactly appears to be insufficient, why, and look for what other influences we think we need to take into account to make our methodology better.

No analysis, no statistics, no intuition, can occur without being grounded in the content of experience. Numbers aren't less real, they correspond to real world events, the same kind that fuel "intuition". There's not really a divide between the two. To use statistics and the like, is just to meet the challenge of speaking clearly about what exactly your intuition consists in.

To reject "numbers", then is at best, to reject one methodology or another for doing a bad job at making predictions. Even if that's true, that doesn't somehow make all of sabermetrics come crashing down. At best, it means our imaginative capacity has fallen short and the methodology doesn't sufficiently capture the complication of the thing it purports to predict. (Throwing the baby out with the bathwater fallacy.)

It hurts on an emotional and not-quite rational level to think that our own personal thought processes can be tracked and analyzed, therefore challenged (e.g. we like to say things like "How DARE you tell ME whether MY opinion, my personal opinion, is right or wrong!"), but even the most personal, subjective analysis is grounded in some fact. In the example in the comments section that apparently helped inspire his article Brian (I think his name is), still had some fact to support his intuition when he was arguing about Blue Jays outfielder Rios, namely, his swing.

If, say, no one sits next to Buchholz, and the reasons for this are "beyond sabermetrics", as though sabermetrics is flawed for not yet incorporating a "how close do you sit to Bucholz algorithm", I have to be blunt and say that's a misunderstanding of what "sabermetrics" is. He wants to say that sabermetrics fails to explain what it purports to explain. But that depends entirely on what it purports to explain. And that's where the author sets up a straw man argument.

Case in point:

because I do not subscribe to the sabermetric orthodoxy. THT is not a hive-minded collective.
and

If the evidence is overwhelming that the conclusion cannot fit within the sabermetric model, it is simply dismissed as a fluke of sample size, luck or random variation rather than considering the possibility that it might be due to human foible or the dreaded “intangible.”
I guarantee he has no concrete example of that actually happening anywhere. Instead of citing some flaw in methodology somewhere, he just accuses sabermatricians of being hive-minded, and invents all of this behavior he thinks he observes. That's an ad hominem, and straw man argument.

[rogue9]

who are you to say that he "thinks" he is observing something. What if I said he i think that Sabrmetricians are just thinking they are observing trends.

You act as though he is arguing that sabermetrics are useless. Hes not, quite the opposite actually. The point of his article though is that many people (you being one seemingly) take it to the extreme. He is arguing that extremes are bad. Which, in many peoples subjective opinion, is a more balanced way of looking at something.

Its two sides to the same coin, two ways to look at it. He is just saying he takes the heads with the tales, not just the tales.

[abben]

(sorry again, I edited this a bit)

who are you to say that he "thinks" he is observing something. What if I said he i think that Sabrmetricians are just thinking they are observing trends.
I'm not sure what you mean by this, though I am positive I never said he just "thinks" he is observing something. Intuition is legit, or often can be. But intuition is supplied in the same way that sabermetrics is, and using numbers and methodology is just a way of figuring out clearly how you use intuition, why it works, and rising to the challenge of clearly saying how your intuition actually works.


Its two sides to the same coin, two ways to look at it. He is just saying he takes the heads with the tales, not just the tales.

Wrong. He isn't just "balancing" the two, he is proactively criticizing sabermetrics, saying it rests upon a flawed assumption. A point which happens to be false and rests on a misunderstanding (perhaps deliberate) of what sabermetrics is. I invite you to re-read my other posts for more, because the point I made stands.

Also, being completely wrong, and actually being correct are two different ways to look at something. It's doesn't follow that the truth is in the middle.

[rogue9]

I guarantee he has no concrete example of that actually happening anywhere. Instead of citing some flaw in methodology somewhere, he just accuses sabermatricians of being hive-minded, and invents all of this behavior he thinks he observes. That's an ad hominem, and straw man argument.
Thats where you said it

I think you miss the point of his article. I don't see where he is saying that sabrmetrics fails as a whole. The way I understood his article was to say that some people use stats and stats only, without any personal obeservation. He is arguing that stats are only stats, pages of numbers, without personal observation. It seems to me that he is saying that to look at baseball solely based off of stats is rather one dimensional.

He also explains how annoying it can be for those of us who enjoy baseball somewhere between the two extremes to have to listen to someone preach down to us that we are enjoying it wrong.

If I want to enjoy baseball for what I perceive to be the intangibles then that is my right. To say that I have to quantify the intangibles to be enlightened as I am entertained is a complete load of BS.

If you can't see that, then I am sorry for you.

It comes down to this. There is an argument going on about how to perceive the entertainment we all love to watch.

How freaking retarded is that.

I am done.

[abben]

Thats where you said it


Ok. I thought you meant with respect to what he observed in baseball players, like Rios.


I think you miss the point of his article. I don't see where he is saying that sabrmetrics fails as a whole. The way I understood his article was to say that some people use stats and stats only, without any personal obeservation.

I've acknowledged that as his central point. And I've responded to it in depth several times. I said 1.) he is characterizing it as a cult instead of pointing out any actual flaws in sabermetrics, which is an ad-hominem argument, and 2.) well...............


I don't see where he is saying that sabrmetrics fails as a whole.

for those that would remind me that he's not talking about sabermetrics in general, I challenge the idea that what he says applies to any sabermatricians, insofar as they are sabermatricians.

I don't see where he is saying that sabrmetrics fails as a whole.

for those that would remind me that he's not talking about sabermetrics in general, I challenge the idea that what he says applies to any sabermatricians, insofar as they are sabermatricians.


I don't see where he is saying that sabrmetrics fails as a whole.

for those that would remind me that he's not talking about sabermetrics in general, I challenge the idea that what he says applies to any sabermatricians, insofar as they are sabermatricians.


I don't see where he is saying that sabrmetrics fails as a whole.

for those that would remind me that he's not talking about sabermetrics in general, I challenge the idea that what he says applies to any sabermatricians, insofar as they are sabermatricians.

Clear?

I'm not interested in trolling so I'll say this:

Either I wasn't clear enough when I described the difference between intuition and sabermetrics, or you aren't reading it very carefully. In any case you haven't actually answered it, and neither has he.

Intuition and sabermetrics both depend on real world events, they aren't different, except that sabermetrics challenges you to state explicitly what your reasoning process is.

[Arctic Blast]

Like in anything else in life, anyone who thinks their way is the ONLY way to perceive or do something is a biased idiot, plain and simple.

[abben]

Like in anything else in life, anyone who thinks their way is the ONLY way to perceive or do something is a biased idiot, plain and simple.

since I'm being called an idiot..............

What's your personal opinion on the answer to 2+2? I met a math teacher who told me it equalled four, but she was just a biased idiot.

Just like my example of a chemist. He thinks mixing hydrogen and oxygen makes water. What a biased idiot.

And, an expert who actually understands how to use sabermetrics? Wait, don't tell me.

Anyway, I invite a thoughtful, civil response to my above posts.

EDIT: Joe Morgan? Maybe you could dish some penetrating insight, please???????

[Arctic Blast]

since I'm being called an idiot..............

What's your personal opinion on the answer to 2+2? I met a math teacher who told me it equalled four, but she was just a biased idiot.

Just like my example of a chemist. He thinks mixing hydrogen and oxygen makes water. What a biased idiot.

And, an expert who actually understands how to use sabermetrics? Wait, don't tell me.

Anyway, I invite a thoughtful, civil response to my above posts.

EDIT: Joe Morgan? Maybe you could dish some penetrating insight, please???????

Where did I say YOU are an idiot? I meant what I said as a NEUTRAL position. Anyone who thinks sabr on its own is the ONLY way of looking at baseball is just as silly as the person who thinks going ONLY by traditional methods is the right way. Just like anyone who thinks that the right wing political view is infallible is a ***** with nothing to offer to discourse, but they're no more guilty than those on the left with the same opinion that only their view is right, and is always right. I wasn't saying that being on one side or another of an argument makes one an idiot, I was saying that seeing your side as the ONLY side as right makes you an idiot. And there is a large difference in discussing the ways of determining the value of a ballplayer and the acceptance of mathematical truths, such as 2+2=4, so I'm really not sure what the point was of making such a ridiculous comparison, as you seem to have felt the need to.

Why you CHOSE to take this as a personal attack, when I said NOTHING to point that way, is beyond me. You also, after all, chose to see the linked-to article in exactly the same way. I didn't see it attacking sabrmetrics as a whole, and it, in fact, seemed to point to a lot of positives from their statistical analyses...but you focused on the criticisms of those who follow it's model as a criticism of anyone who has ever used them for analysis. Do you have some sort of martyr complex, or are you simply so self obsessed that you're unable to accept not being the center of any argument/discussion? Where does this kneejerk assumption of attack against you come from?

I would wait for an insightful response, but thus far, all I've seen from you is baseless accusations and bizarre comparisons that simply do not make any sense. Therefore, I am hardly holding my breath in anticipation.

[OldFatGuy]

Intuition and sabermetrics both depend on real world events, they aren't different, except that sabermetrics challenges you to state explicitly what your reasoning process is.

This is where a lot sabr folks lose us old fashioned folks. Because we look at the stats as a guide only and make decisions based on "non-stat" feelings, we get ridiculed. For example, clutch. I can't tell you how many times I've seen SABR defenders argue there is no such thing as "clutch." And I can't find a single ballplayer that ever played that said "clutch" doesn't exist. The human feelings and emotions that affect us all affect us all differently. And the at-bat in the bottom of the ninth with two outs, bases loaded, and trailing by 1 causes these human feelings and emotions to occur much differently than an at-bat in the 6th inning of a 12-2 game. The sweat glands flow a little more freely, the breathing tightens up, and it affects your peronal performance because we are humans, not machines. Yet when you point out this trueism to many sabr folks, they scream that if that were true it should show up in the stats. But they do you say, just look at such and such. And then they come back say too small a sample size. And around and around we go.

Because sabermetrics and intuition ARE different. One measures the past, the other tries to predict the future. And while sabermetrics, or stats in general, are a good tool in helping to make a decision in the future, they're not the only ones. If I've got two guys on my bench to pinch hit, and one is hitting .500 against this pitcher and the other is 0 for 7, strictly looking at stats would suggest its a no brainer. But if that .500 average was 2-4 with 2 strikeouts and the 2 hits were weak dribblers 25 in front of home plate, it might change that. And if you watched the guy 0-7 hit all 7 of those balls right on the nose only to be right at someone, then you might well choose him as the pinch hitter. I suppose you could invent another stat to keep track of, hard contact per at-bat or something, but that doesn't include other human nature events.

A guy could've gone out late last night and showed up hung over with a severe headache. Another guy could be just starting to get the flu is feeling below average. Another might be in another world because he just caught his wife screwing around on the way to the park. No stats, sabr or otherwise, take the human condition into account as far as predicting the future outcome. This is where good, old fashioned knowledgeable baseball people come in and make intuitive decisions.

[ohms_law]

Yet when you point out this trueism to many sabr folks, they scream that if that were true it should show up in the stats.
a) I never scream.
:)

b) I'm sure they do. The thorniest issue with "clutch" is that there's no definition for it. Precisely what is a "clutch" situation? Is a clutch situation for one player a clutch situation for another? Is there a way to identify what situations are clutch for what players? Without a precise definition... there is no "clutch". All that it is, is some amorphous thing that people somehow "just feel".

The real problem here, on both sides, is the defensiveness. Take a look at all of the last several posts here for examples. Set aside the defensiveness and accept instead, and we'll make progress. Clutch exists, I know it for a fact. It's measurable. Like I said above though, there needs to be something to actually measure. Right now talking about clutch is like asking how many stars are visible in the sky. Well, that's an answerable question as long as you define it some more... can we use a telescope? Is there a set time frame to look at the sky and count? Simply asking how many stars are visible alone isn't a question that can be answered.

This:

A guy could've gone out late last night and showed up hung over with a severe headache. Another guy could be just starting to get the flu is feeling below average. Another might be in another world because he just caught his wife screwing around on the way to the park. No stats, sabr or otherwise, take the human condition into account as far as predicting the future outcome. This is where good, old fashioned knowledgeable baseball people come in and make intuitive decisions.
Simply isn't true though. With enough of the right stats, it's perfectly reasonable to be able to make predictions. How have players who have "just found their wife screwing around" performed in the past? players with the flu? etc...
Simply because the ammount of data may be daunting doesn't mean that the problem is unsolvable. Not having the data does, of course, make the problem basically unsolvable though.

[Reade]

If I want to enjoy baseball for what I perceive to be the intangibles then that is my right. To say that I have to quantify the intangibles to be enlightened as I am entertained is a complete load of BS.


Well said

[ohms_law]

I agree. Sabermetrics and statistics are largely not about enjoyment of the game though. I love just sitting back and watching baseball. I also love sitting down and really analyzing what happened as well.

[belial]

since I'm being called an idiot..............

What's your personal opinion on the answer to 2+2? I met a math teacher who told me it equalled four, but she was just a biased idiot.



It depends on the base system you use. She is biased to the base 10 system.





Just like my example of a chemist. He thinks mixing hydrogen and oxygen makes water. What a biased idiot.



Only in the the right proportions. H2 O is water, while H3 O makes hyrdronium. He would biased towards H2 O.




And, an expert who actually understands how to use sabermetrics? Wait, don't tell me.



Is probably also biased?



Anyway, I invite a thoughtful, civil response to my above posts.


Thank you. I'm sure you'll think I failed. But that's ok.




EDIT: Joe Morgan? Maybe you could dish some penetrating insight, please???????

Please, no.

[belial]

a) I never scream.
:)

b) I'm sure they do. The thorniest issue with "clutch" is that there's no definition for it. Precisely what is a "clutch" situation? Is a clutch situation for one player a clutch situation for another? Is there a way to identify what situations are clutch for what players? Without a precise definition... there is no "clutch". All that it is, is some amorphous thing that people somehow "just feel".

The real problem here, on both sides, is the defensiveness. Take a look at all of the last several posts here for examples. Set aside the defensiveness and accept instead, and we'll make progress. Clutch exists, I know it for a fact. It's measurable. Like I said above though, there needs to be something to actually measure. Right now talking about clutch is like asking how many stars are visible in the sky. Well, that's an answerable question as long as you define it some more... can we use a telescope? Is there a set time frame to look at the sky and count? Simply asking how many stars are visible alone isn't a question that can be answered.

This:

Simply isn't true though. With enough of the right stats, it's perfectly reasonable to be able to make predictions. How have players who have "just found their wife screwing around" performed in the past? players with the flu? etc...
Simply because the ammount of data may be daunting doesn't mean that the problem is unsolvable. Not having the data does, of course, make the problem basically unsolvable though.

You would also never have the data for the individual test subject, which would not be in the control group. The problems are unsolvable, whether you like it or not. Human beings aren't computer programs. You're once again trying to fit something too big into a field too small. Doesn't fit.

[belial]

This is where a lot sabr folks lose us old fashioned folks. Because we look at the stats as a guide only and make decisions based on "non-stat" feelings, we get ridiculed. For example, clutch. I can't tell you how many times I've seen SABR defenders argue there is no such thing as "clutch." And I can't find a single ballplayer that ever played that said "clutch" doesn't exist. The human feelings and emotions that affect us all affect us all differently. And the at-bat in the bottom of the ninth with two outs, bases loaded, and trailing by 1 causes these human feelings and emotions to occur much differently than an at-bat in the 6th inning of a 12-2 game. The sweat glands flow a little more freely, the breathing tightens up, and it affects your peronal performance because we are humans, not machines. Yet when you point out this trueism to many sabr folks, they scream that if that were true it should show up in the stats. But they do you say, just look at such and such. And then they come back say too small a sample size. And around and around we go.

Because sabermetrics and intuition ARE different. One measures the past, the other tries to predict the future. And while sabermetrics, or stats in general, are a good tool in helping to make a decision in the future, they're not the only ones. If I've got two guys on my bench to pinch hit, and one is hitting .500 against this pitcher and the other is 0 for 7, strictly looking at stats would suggest its a no brainer. But if that .500 average was 2-4 with 2 strikeouts and the 2 hits were weak dribblers 25 in front of home plate, it might change that. And if you watched the guy 0-7 hit all 7 of those balls right on the nose only to be right at someone, then you might well choose him as the pinch hitter. I suppose you could invent another stat to keep track of, hard contact per at-bat or something, but that doesn't include other human nature events.

A guy could've gone out late last night and showed up hung over with a severe headache. Another guy could be just starting to get the flu is feeling below average. Another might be in another world because he just caught his wife screwing around on the way to the park. No stats, sabr or otherwise, take the human condition into account as far as predicting the future outcome. This is where good, old fashioned knowledgeable baseball people come in and make intuitive decisions.


Very nice argument. :)

[belial]

Where did I say YOU are an idiot? I meant what I said as a NEUTRAL position. Anyone who thinks sabr on its own is the ONLY way of looking at baseball is just as silly as the person who thinks going ONLY by traditional methods is the right way. Just like anyone who thinks that the right wing political view is infallible is a ***** with nothing to offer to discourse, but they're no more guilty than those on the left with the same opinion that only their view is right, and is always right. I wasn't saying that being on one side or another of an argument makes one an idiot, I was saying that seeing your side as the ONLY side as right makes you an idiot. And there is a large difference in discussing the ways of determining the value of a ballplayer and the acceptance of mathematical truths, such as 2+2=4, so I'm really not sure what the point was of making such a ridiculous comparison, as you seem to have felt the need to.

Why you CHOSE to take this as a personal attack, when I said NOTHING to point that way, is beyond me. You also, after all, chose to see the linked-to article in exactly the same way. I didn't see it attacking sabrmetrics as a whole, and it, in fact, seemed to point to a lot of positives from their statistical analyses...but you focused on the criticisms of those who follow it's model as a criticism of anyone who has ever used them for analysis. Do you have some sort of martyr complex, or are you simply so self obsessed that you're unable to accept not being the center of any argument/discussion? Where does this kneejerk assumption of attack against you come from?

I would wait for an insightful response, but thus far, all I've seen from you is baseless accusations and bizarre comparisons that simply do not make any sense. Therefore, I am hardly holding my breath in anticipation.

Hey man, he isn't worth it.

[abben]

Hey, OldFatGuy (mind if I call you OFG?) Thanks for the thoughtful response. I'm not done with this post, but I'm posting it now for fear of my computer crashing and losing everying.

UPDATE: Not quite done, but this will have to do, feel free to respond to it in this form, though I didn't say everything I wanted to.


This is where a lot sabr folks lose us old fashioned folks. Because we look at the stats as a guide only and make decisions based on "non-stat" feelings, we get ridiculed. For example, clutch. I can't tell you how many times I've seen SABR defenders argue there is no such thing as "clutch." And I can't find a single ballplayer that ever played that said "clutch" doesn't exist. The human feelings and emotions that affect us all affect us all differently. And the at-bat in the bottom of the ninth with two outs, bases loaded, and trailing by 1 causes these human feelings and emotions to occur much differently than an at-bat in the 6th inning of a 12-2 game. The sweat glands flow a little more freely, the breathing tightens up, and it affects your peronal performance because we are humans, not machines. Yet when you point out this trueism to many sabr folks, they scream that if that were true it should show up in the stats. But they do you say, just look at such and such. And then they come back say too small a sample size. And around and around we go.

Yeah, I think you are getting at the overlap, and there is much I agree with here. It's difficult to define, for statistical analysis, what a "clutch" outing is, but we know it, we feel it. What's more, we know for a fact that there isn't some "Clutch Index" that we can just look at that satisfies our understanding of what "clutch" truly is. It can be different things in different cases. It's not always playoff games. Sometimes, if you can come back from 10 runs behind in a regular season game, it feels like a victory. Games against rivals toward the end of the season. But these too can be quantified, (at least loosely) as you rightly pointed out.

What John Brittain is getting at is something like this. We could look at David Ortiz and use some statistic, like "batting average in the ninth inning", and try to use that to measure "clutch" performance. But that's not what clutch is! So if we desperately rely on that stat, we are making a mistake. But wait! Maybe we can do one better: "batting average in the ninth inning in a close scoring game". But yet again, not good enough! What about games where the score is close, but its against a last place team and you already have the division wrapped up?

So we can try even more, "batting average against this team in close scoring games when a playoff spot is in question", or instead of batting average, something like "likelihood to drive in runs". But at each next stop, we find the statistic we are using to be insufficient. So the John Brittains of the world throw up their hands and say "stats just can't do it"! It's our intuition driving this quest for more and more precise stats in the first place, it's our intuition that lets us define the situation we are looking for, but Brittain says "the stat always falls short, and my intuition doesn't". As though there is some magical essence that can never be quantified.

But what's more, a John Brittain would say there is a fundamental failure, fundamental shortcoming in even trying to use stats in the first place. That saber(sabr?)metricians cling, he says, to a stat, and lose all touch with reality. Really? How does he know that? And even if that's true, isn't that the person's fault? What has that to do with sabermetrics itself?

Sure, I can say "batting average is a bad stat, the on-base-percentage is a TRUE measure of how much you get on base". But what did the batting average ever do that was wrong? It's just a statistic.

If for some reason our government (say) decided to dump billions of dollars into research of baseball stats, lending untold resources to the cause that we never had before, it's entirely likely that we would learn all kinds of new things about what makes a "pressure situation" and what makes a player "clutch", and we could understand what makes a guy like David Ortiz so good when he needs to be, and we could know definitively what qualities one has to look for. But we don't have such resources, so, as mortal humans, we do our best to speculate, to use that instinct, baseball intuition or whatever you call it. When we use our instinct we are mulling specific things that are happening, our mind is not a vacuum. For example I saw Tony LaRussa let Chris Duncan pinch hit against a lefty once, and he hit a home run.
So, they aren't different. If a statistician says there is no such thing as "clutch", but there in fact is, then there exists a way to explain what that is, and the statistician made a mistake. Statistics itself didn't make a mistake. The whole problem is, we can lean on one method or one stat too much, get used to it and misinterpret it. "Batting average is insufficient". Why? "It doesn't tell me everything I need to no about batters. Statistics is flawed" But the flaw is NOT with the practice of using stats, it's with US interpreting the stats, trying to use them to explain more than they can. That's what John is claiming Sabermatricians do, but there is no basis for it.

The intangibles that we chase after can be hard to capture. Because we fall short doesn't mean there is a flaw in the stats we use, so long as we understand the meaning of the stats we use. If we understand, we can't blame sabermetrics for our misinterpreting them.


Yet when you point out this trueism to many sabr folks, they scream that if that were true it should show up in the stats. But they do you say, just look at such and such. And then they come back say too small a sample size. And around and around we go.

It is THIS that I contest. That's speculation that can't be substantiated, and yet it's the very heart of the criticism of sabermetrics, an attack on character.

[OldFatGuy]

a) I never scream.
:)


Sorry, didn't mean to imply you. Suffice it say that at the local watering hole of mine this subject has come up. And I have been screamed at. (Although, to be fair, IIRC, I was also screaming. Hard to say who started screaming, other outside influences intervened :D)

And you're right about the defensiveness. I only jumped in when I did because the poster I responded to was obviously being defensive over the original article in question, and expanded it.

You and I may never agree on what stats can and can't tell us, but we do agree they're a useful tool.

[OldFatGuy]

(mind if I call you OFG?)

That would be a decided improvement over the things I'm usually called at the aforementioned watering hole.

[ohms_law]

hehe

ps.: I know you weren't talking to me, but I felt I needed to include that since I was replying immediately after you did.
:)

[SirKodiak]

From the author of the article:

My biggest problem with sabermetrics is the same as any ideology that causes some adherents of it to think they have in their possession the Holy Grail of knowledge.

The most frustrating thing is the supposition by some that I have to tangibly plead fidelity...

To write a certain way to please a certain group so I can be included in their pantheon of writers...

"...the radical fundamentalists of the sabermetric group..."

Not sabermetricians ... just a segment of them.

Many of Bill James later acolytes have gone...

"...it’s guys like them who give hope that sabermetrics will again become more of a journey than a conclusion..."

(again, not criticizng sabermetrics of itself)

I chose my words carefully to make sure it was obvious I wasn't speaking about everyone but just the ideologists among them.

I value the studies of sabermetrics but as you've alluded, any idea--no matter how noble--can become counterproductive when they become an ideology rather than a field of study.

emphasis is from the original

[abben]

I'll bite, belial.


It depends on the base system you use. She is biased to the base 10 system.
No, it doesn't. The values in base 10 can be translated into their equivalent values in any other base, and in each case they describe a single objective value, but they do it in different ways. So, no, subjectivity doesn't play a part in expressing what value it is. The "bias" occurs in a trivial sense, that doesn't change the meaning of the thing communicated. Hence there is an objectively true answer to 2+2, independent of opinion.


Only in the the right proportions. H2 O is water, while H3 O makes hyrdronium. He would biased towards H2 O.
He could expressly describe in sufficient detail, H20 in such a way as to be completely unambiguous. Again, no issue of interpretation.


Is probably also biased?
Whether he's biased or not depends on the facts being contested, which, if sufficiently and clearly described, (just like in the previous cases), terminate in objective truths or falsehoods, independent of opinion. In some cases you can't get sufficiently clear, however that is not always the case.

If someone says "I believe that a single is worth more than a home run", and someone sense "no, that is false, a home run is worth more than a walk", there is an objectively true answer, there is no middle ground.

[belial]

Ahh, but that's the issue. Sabr will always be ambiguous. The problem is they make assumptions, because they have to, and they might not always be right. Hydrogen and oxygen not making water, 2+2 not being 4.

A home run is not implicity worth more than a walk. Clete Thomas drove in the winning run for the Tigers yesterday with a walk. A home run would have led to the same result - a win.

[ohms_law]

uh... a home run is always worth a run. A walk is only worth a run in the specific situation where the bases are loaded. Both have distinct values, which can and have been defined.

...and I just replied to one of your posts. Great.

[abben]

From the author of the article:


emphasis is from the original


for those that would remind me that he's not talking about sabermetrics in general, I challenge the idea that what he says applies to any sabermatricians, insofar as they are sabermatricians.

emphasis is from the original

[belial]

uh... a home run is always worth a run. A walk is only worth a run in the specific situation where the bases are loaded. Both have distinct values, which can and have been defined.

...and I just replied to one of your posts. Great.

You did earlier too.

[ohms_law]

heh

not in one of these threads...
:)

[belial]

A home run is worth more than a walk in almost every situation, but because it isn't always worth more the statement was logically false. Sorry, but true.

[abben]

A home run is not implicity worth more than a walk. Clete Thomas drove in the winning run for the Tigers yesterday with a walk. A home run would have led to the same result - a win.

A home, ceteris paribus, can never be worth less than a walk (if you are describing how they contribute to a team's ability to win a game).


Ahh, but that's the issue. Sabr will always be ambiguous. The problem is they make assumptions, because they have to, and they might not always be right. Hydrogen and oxygen not making water, 2+2 not being 4.

It is precisely this that I contest. A stat is NEVER unambigious. A batting average is clearly defined, as are all stats. How we choose to value a stat, what we try to interpret from a stat can be ambigious. Which is John (the author's) point, but nothing about our subjective assumptions are inherent in a stat itself. He's trying to project ideology into sabermetrics that has nothing to do with what sabermetrics is.

The very act or process of amassing a body of statistics in and of itself carries with it no baggage, no false assumption. To criticize the users of stats, saber/brmetricians for the mere creation of a stat, as though doing so means that they are forced to accept some ideology that comes along with the stat is completely without basis. To say that they subscribe to some faith about the important of a given stat, just because they understand what it actually means, is unsubstantiated ad-hominem speculation.

EDIT: Again, to the point about math, etc. I was pointing out that some things are just true, and there is no middle ground. An opinion that 2+2 = 6 doesn't get to "balance out" an opinion that 2+2 =4, so even if an opinion someone's feelings, we ought to be adults and take sides based on facts and understanding.

[abben]

If I want to enjoy baseball for what I perceive to be the intangibles then that is my right. To say that I have to quantify the intangibles to be enlightened as I am entertained is a complete load of BS.

You can enjoy baseball for any reason you want. You can enjoy the archtecture of the stadiums or the flavor the hot dogs or maybe you like the pair of cleats the home team wears, that's up to you.

But if you proactively, and falsely, assert that something (sabermetrics) has to be "balanced out" because it harbors some inherent flaw that doesn't exist, I'm going to let you know that you are wrong.

[rogue9]

To criticize the users of stats, saber/brmetricians for the mere creation of a stat is completely without basis. To say that they subscribe to some faith about the important of a given stat is unsubstantiated ad-hominem speculation.

I can't recall reading anything in the article, or this thread, criticizing anyone for creating stats.

For the record, I have no problem with the creation of new stats. I think there are some very cool stats out there that alot of people have no idea exist that can give you alot of very good information.

I do not however pretend that it is the only way to look at baseball.

I think the difference is that some people are willing to look at baseball almost completely through stats because it is something that you can wrap your mind around. While others see outside influences that have yet to be quantified (to my knowledge anyway).


for those that would remind me that he's not talking about sabermetrics in general, I challenge the idea that what he says applies to any sabermatricians, insofar as they are sabermatricians.


From the author of the article:
Quote:
Originally Posted by John Brattain
My biggest problem with sabermetrics is the same as any ideology that causes some adherents of it to think they have in their possession the Holy Grail of knowledge.

The most frustrating thing is the supposition by some that I have to tangibly plead fidelity...

To write a certain way to please a certain group so I can be included in their pantheon of writers...

"...the radical fundamentalists of the sabermetric group..."

Not sabermetricians ... just a segment of them.

Many of Bill James later acolytes have gone...

"...it’s guys like them who give hope that sabermetrics will again become more of a journey than a conclusion..."

(again, not criticizng sabermetrics of itself)

I chose my words carefully to make sure it was obvious I wasn't speaking about everyone but just the ideologists among them.

I value the studies of sabermetrics but as you've alluded, any idea--no matter how noble--can become counterproductive when they become an ideology rather than a field of study.
emphasis is from the original

There is the answer to your challenge.

Yah so I wasn't done, but I am now.

[rogue9]

You can enjoy baseball for any reason you want. You can enjoy the archtecture of the stadiums or the flavor the hot dogs or maybe you like the pair of cleats the home team wears, that's up to you.

But if you proactively, and falsely, assert that something (sabermetrics) has to be "balanced out" because it harbors some inherent flaw that doesn't exist, I'm going to let you know that you are wrong.

Oh my word you are dense. No one has asserted that either. The only thing I have seen here or in that article is that some of us think the use of stats should be balanced with "intuition", not that the stats themselves are flawed and need balancing.

You're bending the author's and others words here to shield the fact that you had a knee jerk reaction to the article.

[abben]

Oh my word you are dense. No one has asserted that either. The only thing I have seen here or in that article is that some of us think the use of stats should be balanced with "intuition", not that the stats themselves are flawed and need balancing.

You're bending the author's and others words here to shield the fact that you had a knee jerk reaction to the article.

I'm not going to troll, but

But if you proactively, and falsely, assert that something (sabermetrics) has to be "balanced out" because it harbors some inherent flaw that doesn't exist

No one has asserted that either

The assumption that there is some extreme cult that uses this as the "only way to look at baseball", is just false, and yes, you did claim this. Most recently...


I think the difference is that some people are willing to look at baseball almost completely through stats because it is something that you can wrap your mind around. While others see outside influences that have yet to be quantified (to my knowledge anyway).

The author's point is that there is some essential thing that stats just can't quantify and we (or whatever imaginary minority) are compelled buy into this phoney extreme ideology if we put the effort into understanding sabermetrics.

If you think by using stats, you fail to take note of a more sacred "intuition", you either mischaracterize or misunderstand the people you purport to describe.

[rogue9]

lol, I just looked up the ad-hominem in wiki, your over use of it is rather ironic.

[abben]

Past the point of diminishing returns, but I recommend re-reading what I wrote several posts ago, and trying a bit more sincerely to actually understand what I said, specifically on the fact that intuition and statistics are only "flawed" if you are using them wrong and there is no inherent mistake pervading all of stats that makes them fail for some reason. There is no outside, other magical thing that can't be quantified that somehow "balances out" statistics, and believing this is actually reasonable and logical, it doesn't make anyone part of any cult.

[belial]

A home, ceteris paribus, can never be worth less than a walk (if you are describing how they contribute to a team's ability to win a game).





I didn't say it was worth less. I said sometimes it is equal, thus the statement a home run is more valuable is sometimes false.

[ohms_law]

eh... the walk itself doesn't ever really create a run though. A home run, by itself, always creates at leasty 1 run. A walk may create one run... after 3 other batters get on base somehow. Therefore, in terms of probability, a home run is always more valuable than a walk. Situationally sure, their both worth 1 run after the fact. YOu can predict that a home run will be worth at least 1 run though, but you can't say the same for a walk.

See, you're bringing up a different subject really.

[abben]

I didn't say it was worth less. I said sometimes it is equal, thus the statement a home run is more valuable is sometimes false.

Nor did I characterize you as saying so, though what I said was admittedly not very on topic, so sorry about that. But even in the situation you described, it could be argued that a home run is worth more than a walk. In the same situation, a home run produces more runs than a walk, and a home run would still have produced a win had the team been behind by 1, 2, or 3 runs, producing that same victory as the walk but over a greater variety of situations. However "valuable" isn't clearly defined, so its not a meaningful question.

But that's all beating around the bush. You believe


Sabr will always be ambiguous. The problem is they make assumptions, because they have to, and they might not always be right.

Which is at best unclear, at worst, false. By Sabr do you mean the stats used? Stats operate with specific definitions, what assumptions are involved in the existence of a statistic?

Do you mean analysis of stats? If so, what analysis? Whose analysis?What great secret truth exists that is impossible to quantify? Where is the fatal flaw that is so important that it's worth all of this vague insistence? I've been insulted over and over and over again, and repeatedly mischaracterized (not that my feelings are hurt but it's not much to ask, that for all the bluster, an actual explanation be mixed in there somewhere.) OFG did have a great post, but the things called "intuition" there could theoretically be quantified, and it points to no fatal "underlying assumption" in the use of stats.

I just want to sincerely know what this fatal flaw of Sabr is. It's this mythological temperament and collection of "assumptions" they supposedly make, I've figured out that much. But it amounts to a load of hot air, speculation, and nonsense that can't be backed up by any evidence. The looming and obvious explanation is that these guys are truly experts who know what they are doing, and so everyone who isn't as good with numbers as them gets annoyed. Annoyed at being corrected about a sport they love, annoyed at the idea that their opinions which it feels so good to have, can actually be fact checked, tested against data, proven wrong. But no one wants to admit that because it sounds mean and unfair and anyone who dares suggest it is a big mean jerk.

[belial]

eh... the walk itself doesn't ever really create a run though. A home run, by itself, always creates at leasty 1 run. A walk may create one run... after 3 other batters get on base somehow. Therefore, in terms of probability, a home run is always more valuable than a walk. Situationally sure, their both worth 1 run after the fact. YOu can predict that a home run will be worth at least 1 run though, but you can't say the same for a walk.

See, you're bringing up a different subject really.

I like to be difficult. The homerun is statistically more valuable because it's impossible to always come to the plate with the bases loaded, game tied, in the bottom of the 9th+ inning. I know this. Just let me have my fun. :)

[belial]

You've never read a sabr description of how they came up with the stat and why? Sure, the stat itself is merely observational data, but when so and so says 'pitchers don't control hits given up' and make a stat - the stat is fine to measure something that doesn't mean squat.


The ambiguity is when the sabr person says, 'this stat tries to measure a players speed'. They use triples and sbs and w/e else, when the real answer is only found with a stopwatch.
You're totally missing the point. It's the attitude and sometimes just plain wrongness (it has to be a word if misremembered is) and ingorance of some the stats. True, vorp can't be wrong, because it measures certain things. But it certainly doesn't measure value.

Who insulted you? You said some inaccurate things, and tried to 'expand the stike zone', and were called out on it, but insult?

[dps]

Stats are just an attempt to measure what happens of the field. Sabermetric stats aren't fundmentally different than "traditional" stats like RBIs or innings pitched, they're just newer and hopefully provide better insight to the game. I don't know of any sabermetrician who would claim that the stats that we have to work with measure everything that happens on the field, or that measure with 100% accuracy the things that they do measure. (In this context, I use the term "sabermetrician" to mean people who have actually done publishable research, not someone who has read a couple of Bill James' or Pete Palmer's books and then makes posts about it on the internet.) But (and I've posted basically this same thing before) the fact that we can't measure everything with 100% accuracy doesn't mean that we should ignore what we can measure. On the other hand the opposite is also surely true--there are things that we know exist and happen on the field that we can't measure or can only make very inaccurate measurements of and that doesn't mean that we should pretend that they don't exist--though the fact that some people claim that something exists that can't be measured doesn't mean that it does exist.

[belial]

I don't mind the new metrics, I just mind the people that use them or think they explain more than they do.

Even with the 'old' stats, people would 'use a formula in their head' and weigh performances. Most metrics just build upon the the old 'building blocks' anyway.

[OldFatGuy]

I like to be difficult. The homerun is statistically more valuable because it's impossible to always come to the plate with the bases loaded, game tied, in the bottom of the 9th+ inning. I know this. Just let me have my fun. :)


But your point was valid. I'll give you an example. It's the bottom of the ninth, the bases are loaded and you trail by one with two outs. Your two pinch hit options are 1) a high power, high strikeout, medium on-base guy or 2) a low power, low strikeout, very high on-base guy. The decision here is obvious, you go with 2. Now, bottom of the ninth, bases loaded, two outs, and you trail by 4. Now the situation isn't so obvious, though for me I would then always go with 1.

Thus, statistically speaking, you would pick a different guy depending on a different situation. Now, go back to my first example where the decision was obvious. THEN, to go beyond statistics, you saw guy #1 have the best batting practice he's had in two weeks before the game, and you also saw guy #2 come to the ballpark with his nose and eyes leaking from allergies. Now what???

Stats are never wrong, in that 1+1 is ALWAYS 2. The only way stats are wrong is if you made an error during calculation (the old GIGO effect). But stats can be useless, just as the stat that 100% of the winning teams scored more runs than the losing teams. But how you USE them is where there is so much....defensiveness to use Ohms_Law phrase. It seems, (and yes this is a CHARACTER indictment), it seems that an awful lot of the folks who go on and on about sabr stats use them differently than I would. I know not if that makes me right and them wrong, but I do know that if were still active in baseball I would not be married to any stat, sabr or otherwise.

And I also know that not all of the new stats to grace the baseball world since SABR arrived are all that useful. I'm going to start a whole bunch of **** with this one (I know this because IIRC this was the one that led to the screaming in my local watering hole I referred to), IMHO Dice is way overrated. Way, way overrated. Now I'm sure there are lots of folks that would argue that DICE is not only not overrated, it's one of the best stats to come along since OPS. I'm not one of those folks. And because I'm not, I sometimes feel, (another character indictment) like the author of the article that started all of this mess, that I'm somehow not bowing to the SABR gods in a proper and respectful manner.

That was a very condensending way to make my point (spelt rong im shur), but I've just returned from said watering hole and am not in the best frame of mind to be tactful.

[abben]

You've never read a sabr description of how they came up with the stat and why? Sure, the stat itself is merely observational data, but when so and so says 'pitchers don't control hits given up' and make a stat - the stat is fine to measure something that doesn't mean squat.

Every single time. This is the best "evidence" of some prevailing mentality, and all it is, is a load of nonsensical assumptions and imagined characterics of some vague boogeyman that doesn't actually exist.


The ambiguity is when the sabr person says, 'this stat tries to measure a players speed', . They use triples and sbs and w/e else, when the real answer is only found with a stopwatch.


Yet again. My point is proven over and over. Even if that were true for some specific stat, all you have is an empty, bogus mentality that you think exists in the mind of a sabermetrics guy. Who actually makes such ridiculous claims, that some new measurement of statistical performance IS a measurement of speed? Every example uses a pretend claim.

Let's look at a real description


Secondary Average, or SEC, is a sabermetric tool used to gauge a player’s ability to produce extra bases independent of batting average (The total of a player's extra bases on hits, walks, and stolen bases expressed as a percentage of at bats). The statistic basically covers the three primary factors of an offensive contribution outside of AVG-…power (bases), eye (BB) and speed (SB).


That's reasonable, expansive, and legit way of trying to descibe a players offensive production. So long as you understand what it does. That's the point.

But, to end this debate, I point to (from wiki) the Sabermetrics manifesto, which actually describes what the heart and soul, the true intention behind sabermetrics is (instead of speculating and inventing imaginary claims about what I kinda sorta think it means)


Bill James defined sabermetrics as "the search for objective knowledge about baseball." Thus, sabermetrics attempts to answer objective questions about baseball, such as "which player on the Red Sox contributed the most to the team's offense?" or "How many home runs will Ken Griffey, Jr. hit next year?" It cannot deal with the subjective judgments which are also important to the game, such as "Who is your favorite player?"

The entire point of sabermetrics is that it DOESN'T, repeat, DOESN'T, make vague claims. That is the entire point. No one, says "'this stat tries to measure a players speed" unless they have clearly and meaningfully defined speed, no one says "'pitchers don't control hits given up" unless the notion of "control" is used for convenience to explain some specific phenomena. And there is no reason why they can't do that. Terms like atom and quark were borrowed from philosophy and literature to explain problems in science, with no negative effect on the original understanding of the term.

And why not quote another part from the manifesto:


Since statistics are the best objective record of the game available,
sabermetricians often use them. Of course, a statistic is only useful
if it is properly understood. Thus, a large part of sabermetrics
involves understanding how to use statistics properly, which statistics
are useful for what purposes, and similar things

Game, set, match. Core and central to the philosophy is to have a clear understanding of the meaning and use of numbers. So the idea that it is somehow inherent that sabermetrics leads to pointless and misleading stats, and a misunderstanding of the game, collapses.

They only ask questions that can be clearly defined in the first place, they don't purport to answer abstract questions like measuring "speed". Over and over and over, you miss the point and invent these, false, nonsense claims.

wow. Consider me unsurprised if this is all "biased" and that you, yet again, have somehow discovered the imaginary "true" purpose inside Bill James's head. Because they shouldn't have a right to talk about a field they invented. No, we should listen to those with ideological bents against it.

I asked for a real, coherent example, and you've proven my point- this is vague baseless speculation that you can't back up- you construct some half-hearted straw man and attack the straw man's character and imply all this nonsense about a prevailing mentality among sabermetricians.

If the extent of your claim is "well, yeah, stats can't be wrong (the whole point, but nevermind that)... b-b-b-b-b-but hey, they are imperfect", that's hardly a criticism. With new stats we gain a new galaxy descriptive power that yeilds new insight we never had before, which we never could have gotten from gut intuition alone. And even should some baseless speculation creep in which goes against the core purpose of sabermetrics, that leads us to make bad decisions we wouldn't otherwise make, we could only learn the poor predictive power by further statistical analysis, not by throwing up our hands and offering some unsubstantiated subjective claim about why we personally feel stat X is no good.

The entire point and rise of this whole field of study was to not speculate, but restrict claims about performance to facts, things that aren't vague, and not make the false guesses based on stats that you keep mistakenly attributing to imaginary statisticians who you don't have examples of.

In short, to say Sabrmetrics is vague is to betray a total misunderstanding of its purpose and how it operates.

I invite responses, I'm sure more progress can be had, but I myself now bow out in the interests of civility. Feel free to PM.

[belial]

2.hgm

[ohms_law]

Nope... That is, if I understand what you're saying correctly.

[dickay]

What started out as a good post has turned into the most unreadable thread of all-time.

Or so at least IMO.

[ohms_law]

It's not that bad really. Everyone seems to be keeping their cool... more or less.

[HoustonGM]

2.hgm
Excuse me?

[chuckwillard]

I'm on the fence...

I like certain "sabrmetric" stats. OBP, OPS, WHIP, and I love the X/W-L.

I don't like VORP, I think it's an unnuanced stat. How can someone say David Ortiz's VORP is comparable to, say, Josh Hamilton's? David Ortiz goes down, the Sox plug Manny in at the DH and put Jacoby in LF. Josh Hamilton goes down, God only knows what the Rangers do. True Value over Replacement Player would measure how much better a person is than his likely individual replacement, not some nebulous "replacement" that seems to be suspiciously close to the Mendoza Line.

I don't know if I like these "park averaged" stats...OPS+, ERA+, etc. First and foremost, if I can't see the calculations they're getting these from, I can't trust them. If John McCain/Barack Obama comes out and says "I have a plan to make gasoline $.79 a gallon again and never take another barrel from the Saudis," I'm gonna be damned skeptical until I see the numbers. If Bill James comes out and says, "Jake Peavy is 79% better than the league average" I'm gonna be skeptical until I see the numbers he got there from.

The beautiful thing about all the "traditional" stats, is that when I read that Ted Williams hit .406 is 1941, I know exactly what that means. If I read that Jake Peavy had an ERA+ of 179 (I made that up, I have no idea what his actual stat was), all I have is that stat creators word for it. I can't go plug numbers in and calculate it myself.

My final problem with sabrmetrics (and not SABR itself, I'm a member) is that certain sabrmetricians seem more concerned with proving that "traditional" analysts are wrong than with actually getting to the truth of the matter.

[dps]

I can't see the calculations they're getting these [certain sabermetric stats from, I can't trust them... ...The beautiful thing about all the "traditional" stats, is that when I read that Ted Williams hit .406 is 1941, I know exactly what that means. If I read that Jake Peavy had an ERA+ of 179 (I made that up, I have no idea what his actual stat was), all I have is that stat creators word for it. I can't go plug numbers in and calculate it myself.



One of the things I like about Bill James is that he publishes his formulas, so you can see where he got the numbers from.

[belial]

I'm on the fence...

I like certain "sabrmetric" stats. OBP, OPS, WHIP, and I love the X/W-L.

I don't like VORP, I think it's an unnuanced stat. How can someone say David Ortiz's VORP is comparable to, say, Josh Hamilton's? David Ortiz goes down, the Sox plug Manny in at the DH and put Jacoby in LF. Josh Hamilton goes down, God only knows what the Rangers do. True Value over Replacement Player would measure how much better a person is than his likely individual replacement, not some nebulous "replacement" that seems to be suspiciously close to the Mendoza Line.

I don't know if I like these "park averaged" stats...OPS+, ERA+, etc. First and foremost, if I can't see the calculations they're getting these from, I can't trust them. If John McCain/Barack Obama comes out and says "I have a plan to make gasoline $.79 a gallon again and never take another barrel from the Saudis," I'm gonna be damned skeptical until I see the numbers. If Bill James comes out and says, "Jake Peavy is 79% better than the league average" I'm gonna be skeptical until I see the numbers he got there from.

The beautiful thing about all the "traditional" stats, is that when I read that Ted Williams hit .406 is 1941, I know exactly what that means. If I read that Jake Peavy had an ERA+ of 179 (I made that up, I have no idea what his actual stat was), all I have is that stat creators word for it. I can't go plug numbers in and calculate it myself.

My final problem with sabrmetrics (and not SABR itself, I'm a member) is that certain sabrmetricians seem more concerned with proving that "traditional" analysts are wrong than with actually getting to the truth of the matter.

I agree that certain sabr people are trying to prove traditional stats wrong... but really, all batting is is hits/ab. It tells you one thing. When they say it's horrible to use to rate a player, I agree but so is using wpa, vorp, warp3, ops or any other single stat. Just show me the building blocks - I know a decent performer when I see the basic blocks.

[belial]

Excuse me?

Nothing bad.

[rogue9]

My final problem with sabrmetrics (and not SABR itself, I'm a member) is that certain sabrmetricians seem more concerned with proving that "traditional" analysts are wrong than with actually getting to the truth of the matter.

bingo.

Also, OFG is right on as well.

[HoustonGM]

I don't like VORP, I think it's an unnuanced stat. How can someone say David Ortiz's VORP is comparable to, say, Josh Hamilton's? David Ortiz goes down, the Sox plug Manny in at the DH and put Jacoby in LF. Josh Hamilton goes down, God only knows what the Rangers do. True Value over Replacement Player would measure how much better a person is than his likely individual replacement, not some nebulous "replacement" that seems to be suspiciously close to the Mendoza Line.
Because VORP is meant to be able to compare the value of players, regardless of their team. Obviously, when a team is evaluating the value of players FOR ITSELF, it needs to look at who ITS replacements are. But VORP isn't meant to do that. It's meant to compare players on different teams.


I don't know if I like these "park averaged" stats...OPS+, ERA+, etc. First and foremost, if I can't see the calculations they're getting these from, I can't trust them.
You CAN see those calculations. Park factor calculations are ridiculously simple. Then, they adjust the league OPS or ERA to the park using those park factors, and then compare the player's OPS/ERA to that adjusted OPS. It's really simple, really intuitive, and a very good way to compare players, regardless of era, park, team, etc.


If Bill James comes out and says, "Jake Peavy is 79% better than the league average" I'm gonna be skeptical until I see the numbers he got there from.
It's really simple man. Just compare a player's ERA to the league average...


The beautiful thing about all the "traditional" stats, is that when I read that Ted Williams hit .406 is 1941, I know exactly what that means. If I read that Jake Peavy had an ERA+ of 179 (I made that up, I have no idea what his actual stat was), all I have is that stat creators word for it. I can't go plug numbers in and calculate it myself.
See, I think you're misunderstanding ERA+. This "criticism" is valid for something like VORP, where the exact formula isn't public. But ERA+ is public and easy to quantify on your own. From B-Ref:

- the ratio of the league's ERA (adjusted to the pitcher's ballpark) to that of the pitcher. > 100 is above average and < 100 is below average. lgERA / ERA


My final problem with sabrmetrics (and not SABR itself, I'm a member) is that certain sabrmetricians seem more concerned with proving that "traditional" analysts are wrong than with actually getting to the truth of the matter.
And that is a misconception harbored by many "anti-stat" people. Good sabermetric studies don't start set out to prove something wrong. They start with a question and then examine the facts. Sometimes, it DOES prove the traditional analysts wrong, and that doesn't mean that sabermetrics are more concerned with doing that. It means they ARE more concerned with getting to the actual truth of the matter, regardless what our perception and the media tells us.

I do know that you qualified that with "certain", but this is something that like, abben said, seems like a strawman just to attack sabermetrics as a whole. Who are these mysterious people who only care about proving tradational analysts wrong? Yes, there are people that DO prove traditional analysts wrong, because traditional analysts often say things that ARE wrong and HAVE been proven wrong. But again, that doesn't mean that these people only care about proving them wrong. It means that they care about these analysts actually telling the truth.


Nothing bad.
Well, can you clarify? Because, I'd barely posted in this thread, so I'd like to know why my name was brought up out of nowhere.

[belial]

similiarities in long posts, with multi quoting from several people

[SirKodiak]

Follow-up article by John Brattain (author of the original article):
http://www.hardballtimes.com/main/article/a-conversation-will-bill-baer/

[OldFatGuy]

Follow-up article by John Brattain (author of the original article):
http://www.hardballtimes.com/main/article/a-conversation-will-bill-baer/

That guy is my hero. Does so much better saying what I try to and can't. So much better.

[Arctic Blast]

Well, I sort of feel like it's necessary to 'clear the air' a bit, as it were.

Just to be completely clear about my position...I have no problem with sabermetrics itself. As most people know, while I enjoy baseball, football is my #1 sport of choice. And recently, in football, some interesting new stats have popped up on Football Outsiders (publishers of Pro Football Prospectus) that are fairly similar to sabr stats in baseball...measuring offensive efficiency and offensive line play. For the most part, they're at least interesting, and a new way of really analyzing a team, much like the majority of Sabr stats are in baseball.

I also don't have a problem with 90% of people who would qualify as sabrmetricians. Why would I take issue with someone over a mathematical formula? That's beyond stupid. The issue comes with the other 10%. For some reason, I don't know if it stems from facing irrational criticisms, or feeling like they need to maintain a 'bunker mentality', or whatever, but for some reason, these 10% seem to 'see' criticisms where none exist at all. A fine example here is abben. Nobody else, including many people who are big supports of sabr forumulae, saw any problem with the posted article. Yet he saw it as a one sided assault on what he holds true. I post a neutral assault on 'true believers' on both sides of an argument, and he takes it as a personal attack against only him. I have no patience for people like this. I put people like that in the same category as gaming system fanboyz and political right/left hardcores, and anyone else so utterly biased by their perspective that they can't even admit the possibility that their side is ever the least bit fallible.

Sabrmetrics can be a valuable tool, but it isn't perfection. I think THAT was the point of the original article.

[rogue9]

considering the conversation that has gone on here I especially liked this part of the second article:

The thing is, 2+2=4; it has always been thus and there are no outliers, flukes, random variations in history that caused 2+2 to equal anything other than four. However, in baseball things happen that cause outliers, flukes, random variations which means that a hard and fast approach to run scoring (or anything else) is nowhere near a guaranteed solution. Alas, intangibles are one of those things--a pitcher might live for the big moment while a far more talented hurler might wet the bed. A hitter might focus better and be looser in the big situation than a far more talented batter who might tighten up when the big blow is required.

[ohms_law]

The problem with the criticism of statistics that I see is that the problem usually comes down to "You can't explain this, so your work doesn't mean that much". As has been pointed out several times in this thread already, often the statistic isn't even meant to address the question that the critic is posing.

To directly address the statement rogue9 posted from the article though, "outliers" are always explainable. The narrow scope of statistics often causes outliers to crop up simply because not everything is considered when creating the statistic. With more information, ouliers can always be explained. Even more important, outliers are often acceptable as long as the definition is well formed so that those using the statistic can account for the problems. This is the main reason that, for example, On Base Percentage and Slugging are such valuable statistics that their give right beside Batting Average these days. There are variations in batting average for the simple fact that the possible result set from a plate appearance includes more than either a hit or an out.

Basically, the reason for quite a bit of criticism comes down to either ignorance or a lack of desire to learn. That sort of attitude is frustrating, which I think leads to some "Sabermatricians" acting like ideologues. The reality though, when you read the work that many people have published, couldn't be further from the truth. That being said, there do seem to be some people who are only somewhat knowledgeable and take great pride in regurgitating certain facts and statistics.

[belial]

I think the confusion lies in the percieved value of the sabr stats. I don't place as much weight in wpa as say sirkodiak does, or k/ip as hgm does. This 'ignorance' isn't really ignorance, as I know what the stats measure - I just don't find them to be as important as other things. Pitchers goals are to win the game. Sometimes Leyland tells the pitcher that the runner on 3rd will score, just don't let the batter score. Sometimes hitters through a hittable then 'earn' a walk, which is overemphasized in OBA. (Because a walk = a hit in OBA, when a walk is proven to be less valuable, sometimes even semi-intentionally given to limit a scoring chance by setting up a double play).

[HoustonGM]

similiarities in long posts, with multi quoting from several people
It's called organizing your thoughts and making sure it's obvious what you're responding to.


I think the confusion lies in the percieved value of the sabr stats. I don't place as much weight in wpa as say sirkodiak does, or k/ip as hgm does.
This is probably because you like to decide for yourself how much weight other people put into things, without actually knowing what they think.

[ohms_law]

See though, you've just proven the point that I was making precisely. You're assigning meaning to OBP where none was intended, above. By doing so, you're using OBP in order to attempt to answer a question which it was not intended to answer.

I don't see a pitchers' goal being to win a game, either. It's just not possible for a pitcher to win by himself, especially since he can't score runs (or is extremely limited in doing so). The pitchers goal is, to the best of his ability, to not allow runs, which means that he must earn outs. The situation that you posted about above, where "Jim Leyland tells the pitcher that the runner on 3rd will score, just don't let the batter score." is more about analysis then it is about statistics. The Leyland in your example is predicting that his pitcher can perform better by ignoring the runner on third and concentrating on the guy at the plate. He's also predicting that by allowing the runner on third to score that he can limit the total runs allowed, since the team will be concentrating on earning an out from the guy at the plate. You didn't post any statistics along with that, but there should be (and in real life there are) some statistical evidence to either support or refute the managers decision in that particular situation.

The main point is that statistics measure descreet sets of data. Using a statistic outside of it's intended scope will lead to inappropriate analysis. If you want to answer a question, then you need to use the appropriate statistics to back up your position. If you don't do that, then you may as well just guess.

[belial]

It's called organizing your thoughts and making sure it's obvious what you're responding to.



It's fine, and I know what it is. No one is out to get you.






This is probably because you like to decide for yourself how much weight other people put into things, without actually knowing what they think.

Uh... when I say it's overrated, and you guys say 'no I don't think so', I'm pretty certain you weigh it more than I do. Again, be happy, no one is trying to upset you.

[belial]

See though, you've just proven the point that I was making precisely. You're assigning meaning to OBP where none was intended, above. By doing so, you're using OBP in order to attempt to answer a question which it was not intended to answer.

I don't see a pitchers' goal being to win a game, either. It's just not possible for a pitcher to win by himself, especially since he can't score runs (or is extremely limited in doing so). The pitchers goal is, to the best of his ability, to not allow runs, which means that he must earn outs. The situation that you posted about above, where "Jim Leyland tells the pitcher that the runner on 3rd will score, just don't let the batter score." is more about analysis then it is about statistics. The Leyland in your example is predicting that his pitcher can perform better by ignoring the runner on third and concentrating on the guy at the plate. He's also predicting that by allowing the runner on third to score that he can limit the total runs allowed, since the team will be concentrating on earning an out from the guy at the plate. You didn't post any statistics along with that, but there should be (and in real life there are) some statistical evidence to either support or refute the managers decision in that particular situation.

The main point is that statistics measure descreet sets of data. Using a statistic outside of it's intended scope will lead to inappropriate analysis. If you want to answer a question, then you need to use the appropriate statistics to back up your position. If you don't do that, then you may as well just guess.


The pitchers' goal is to win the game, regardless of how many runs the opposition scores. If you think a pitcher pitches the same with a 5 run lead compared to when the game is tied, then you need some more data.

Don't get mad, but the word you were looking for was, 'discrete'.

[HoustonGM]

Uh... when I say it's overrated, and you guys say 'no I don't think so', I'm pretty certain you weigh it more than I do. Again, be happy, no one is trying to upset you.
You think it's overrated because you like to decide for yourself how other people weight their stats. You like to make up strawmans like that SirKodiak thinks WPA is a good way to judge talent, and then use that to say WPA is overrated, when really, SirKodiak never said such a thing. Like ohms said, you assign meaning to stats where none was intended.


The pitchers' goal is to win the game, regardless of how many runs the opposition scores. If you think a pitcher pitches the same with a 5 run lead compared to when the game is tied, then you need some more data.
Perhaps its you that needs more data, because studies have been done on that, and pitchers show no tendency to pitch differently (statistically) in close games and blowout games. Now, this doesn't mean that they don't pitch differently. All it means is that, when it comes to run prevention, they aren't any different, overall, in those situations.

[ohms_law]

Don't get mad, but the word you were looking for was, 'discrete'.
No spell check at work...
;)


The pitchers' goal is to win the game,
The whole team's goal is to obviously win games.

It's obvious within the correct statistics that some pitchers do pitch differently with a lead or when behind. The difference isn't large though (every example that I've seen is easily within the margin of error), and over the course of an entire season doesn't make any real difference.

The point though is that it's a pitchers job to pitch the ball to the batter. That's it. beyond that, pitchers field their position and (in the National League at least) take their chop at bat. No one can win a game by pitching the ball to the opposition simply because you can't score by doing so.

I think that one of the things that you're referring to is leadership. It's certainly true that many (most?) pitchers assume leadership positions on the teams that they play on. There is no rule that this must be true though, and catchers often assume just as much of a leadership position as pitchers to with other position players occasionally assuming the role of team leader.

One place that I think that the whole "pitchers must win the game" attitude comes from has to do with the fact that the pitcher controls the ball. Baseball is different than pretty much all other sports though in that the defense controls the ball. You can't win any contest with defense alone, since the definition of defense is the prevention of your opponent from achieving it's goal. Defense is vital to winning, but it cannot cause a win.

Purely in terms of baseball, I think that pitchers could legitimately be credited with losses given a strict enough definition. I simply cannot see crediting a pitcher with wins in any meaningful way, though.


and pitchers show no tendency to pitch differently
You can't really say that though because there are some pitchers with consistently different splits... I'll have to find someone on B-R.com now. I remember a really good example that I read about a while back, but I can't recall who it was off the top of my head.
The thing is, that's a highly personal trait to the player. Behavior with or without the lead is certainly not something that it similar in all pitchers.

[HoustonGM]

Well, I meant, pitchers AS A WHOLE. The studies looked at large groups of pitchers.

I've also seen studies regarding specifically two players that are often said to have "pitched to the score" - Catfish Hunter and Jack Morris - and those studies proved that notion wrong.

[ohms_law]

Ah yea, good old Catfish Hunter...
:)

here, take a look at his Game Outcome splits: http://www.baseball-reference.com/pi/psplit.cgi?n1=hunteca01&year=00#outco-outco
you tell me, did he "pitch to the score" at all?
:)

It's the "large groups of pitchers" that becomes a problem, really. The effect that we're looking for gets lost in the noise when you enlarge the group. It's actually somewhat lost in that split as well (it's more noticeable when you look at close vs. blow out games, but I don't see those listed on the B-R.com splits pages.). If you're looking for a specific situational effect, then you need to use stats that represent that situation to see it, you know?

[HoustonGM]

Ah yea, good old Catfish Hunter...
:)

here, take a look at his Game Outcome splits: http://www.baseball-reference.com/pi/psplit.cgi?n1=hunteca01&year=00#outco-outco
you tell me, did he "pitch to the score" at all?
:)
You can't tell by looking at his Game Outcome splits. All pitchers (with a decent enough sample size) have substantially better stats in wins than in losses. Here's an article on "pitching to the score" that looks at it through Pythagoras - http://www.stathead.com/bbeng/spira/pitchtoscore.htm:


The pitchers who get a reputation of "pitching to the score" have one thing in common - they have all generally gotten good run support through most of their careers. It seems apparent to me that pitchers get this reputation because they get better run support than most other pitchers and thus have a W-L record that looks better than their ERA or runs allowed. In general, these pitchers get good run support because they are on good offensive teams for a number of years. Jack Morris is the best example of this; in his entire career, he was on only one below-average offensive team (1989), and he pitched for several great offensive teams. The "pitch to the score" theory seems to be an effort to imbue a pitchers' won-loss record with a value other than luck.

[belial]

Well, I meant, pitchers AS A WHOLE. The studies looked at large groups of pitchers.

I've also seen studies regarding specifically two players that are often said to have "pitched to the score" - Catfish Hunter and Jack Morris - and those studies proved that notion wrong.

They proved results, nothing else.

[HoustonGM]

They proved that the results of Hunter and Morris pitching didn't differ based on the score of the game. Sure, they didn't prove that the pitchers didn't "pitch differently", but that's entirely irrelevent, since its the results that matter, and if the results are the same, it doesn't really matter whether or not they were achieved by a pitcher pitching one way or another.

[ohms_law]

Well... yes and no. It can make a significant difference on the "right" team. As the analysis above points out "they have all generally gotten good run support through most of their careers." If that's not true, things would be very different.

[belial]

It didn't show anything except they got good run support. How do you know all things would be different? Things weren't different, so how can we know for sure?