Clutch stat

[5-4-3]

Originally Posted by Syd Thrift

Joe Morgan, for all of his color commentary garbage, was one of the smartest on-field players in the history of the game.

And perhaps the stupidest one off of it.

Originally Posted by Syd Thrift

As for the rest of this, you're clearly uninterested in actually reviewing what others have to say. You've decided to align yourself with the orthodoxy and you're just going to stick with this in spite of any issues with what the evidence can or cannot say at this point in time. Which, you do you I guess.

I’m interested in what the data says. And the data says that clutch hitting does not exist. You and others are talking about how you feel, but data doesn’t care about your feelings.

[5-4-3]

Originally Posted by BPS

No "clear, rigorous, and convincing" evidence exists for the absence of clutch hitting. All the studies I've glanced at have serious flaws.

Nonsense.

Originally Posted by BPS

If you want, pick the study you think provides the kill-shot for the idea of clutching hitting and then explain it in your own words, paying particular attention to the statistical techniques used. We can then discuss it. Perhaps you'll convince me.

For all of them: if clutch hitting exists, it would be a repeatable skill. Data has established that it is not.

Do you believe that the earth is flat, too? Or that climate change isn’t real? Because you’re ignoring hard data just like the people who take those stands do.

[Syd Thrift]

Originally Posted by 5-4-3

I’m interested in what the data says. And the data says that clutch hitting does not exist. You and others are talking about how you feel, but data doesn’t care about your feelings.

Bull-crap you care about what the data says, because if you did then you'd have actually read what I'd been saying earlier about how the kind of data I'm interested in doesn't exist yet, at least not in large enough numbers that we can (to my knowledge) make assertions with it.

Otherwise, I think that the earth is round, and yes, I think that climate change is both real and caused by humans (which I see you didn't add in; do *you* dispute anthrogenic climate change? Because the opinion of climatologists is overwhelming in favor of this). I also think that Bigfoot is probably fake, that 9/11 was a conspiracy by 19 people to fly planes into buildings and not a conspiracy by thousands for [insert reason X here], that UFOs aren't real, and that Oswald shot Kennedy. Is there any other skeptic-related question you wanted to ask before we go back into your lack of basic reading comprehension skills?

[BPS]

Originally Posted by NoOne

well, no counting cards is udnerstanding odds as they shift over time... still just playing the odds with updated info. the results? again, confusing results with probability.. playing odds doesn't mean you will win.. merely best probability.. . so a gambler who is more likely to win (if that exists) can go on extended losing treaks. (only in games vs humans can it be possible, even craps favors the house, even though it has the 'best' odds for the player - even black/red on roulette isn't 50/50 because not all resutls are black or red on the wheel)

I'd truly like to respond to your comments, but due to my (apparent) lack of reading skills I'm not certain that the particular claims you are making. So I'm not certain how to respond. Sorry.

[BPS]

Originally Posted by 5-4-3

For all of them: if clutch hitting exists, it would be a repeatable skill. Data has established that it is not.

I've already made a criticism about the study linked to above about the repeatability of clutch hitting. I guess you didn't read my comment.

Oh, well.

[5-4-3]

And you were wrong.

Every rigorous study, every one, with over 100 years of data, says exactly the same thing: there is not one shred of evidence that clutch hitting exists. Not one. Zero evidence. If it existed it would be detectable. It is not detectable.

Bill James runs the numbers and tells you that the data says it doesn't exist.

Dick Cramer runs the numbers and tells you that the data says it doesn't exist.

Pete Palmer runs the numbers and tells you that the data says it doesn't exist.

FanGraphs, Beyond the Box Score, and Baseball Prospectus all run the numbers and tell you that the data says it doesn't exist.

None of you can run the numbers and find a shred of evidence that it exists. You are in opposition to reality. Yet you continue to believe.

William K. Clifford, in “The Ethics of Belief,” wrote that “It is wrong, always, everywhere, and for anyone, to believe anything upon insufficient evidence.” Yet you clutch hitting believers do so.

Christopher Hitchens put it a different way. “That which can be asserted without evidence, can be dismissed without evidence.”

You myth believers have no evidence. None. Therefore I dismiss you.

[Syd Thrift]

And you were wrong.

Every rigorous study, every one, with over 100 years of data, says exactly the same thing: there is not one shred of evidence that clutch hitting exists. Not one. Zero evidence. If it existed it would be detectable. It is not detectable.

Reading comprehension. Learn it.

[Silfir]

I had a look at the Baseball Prospectus study. This is how its conclusion opens:

This analysis, which builds on Pete Palmer’s and Dick Cramer’s research from the 1970s with a much more robust data set, illustrates continued difficulty in identifying clutch hitting as a replicable skill. Again, this is not to say the [sic] clutch hitting does not exist.

The study utilized a specific definition of clutch hitting, as follows (emphasis mine):

A viable statistical metric must be replicable, with results generally consistent over time. Our measure of clutch hitting—the excess performance of a hitter in high win-expectancy plate appearances compared to others—fails to meet this test. We therefore echo Cramer’s conclusion from 41 years ago that while clutch hitting may exist as a feature, it does not exist as a repeatable skill.

Syd Thrift doesn't disagree with any of this:

Originally Posted by Syd Thrift

2. The BP article straight up doesn't address this because they know and I know that there isn't enough data yet. If/when there is, I expect that the evidence for "clutch" in baseball will read something like "most of the time that you see 95+ mph fastballs is when you're down in the 9th inning by 1-3 runs and as such hitters who can get around on fastballs of that speed will have a 10-20 point wOBA boost". I doubt we'll ever find, like, mental evidence because if there was it probably would have shown itself by now. It also may come out that there is no special ability there. We just don't have enough of *that* kind of data.

Again I bolded the relevant passage: Syd Thrift is in perfect agreement with all of the people you just quoted (probably including the non-baseball related ones, given he's written elsewhere he is an atheist himself) that "mental" clutch hitting doesn't exist.

His point is that the pitchers that are usually brought out in clutch situations tend to throw certain kinds of pitches, and that a hitter who is good at hitting those pitches could be said to be a good "clutch hitter", in the same way that there are "pull hitters" or "hitters who are good against lefties" and so on.

What he's done is use a different definition of the word "clutch hitter" that doesn't reference any kind of mental aspect, which I would agree is somewhat confusing since the term "clutch" as its commonly used strongly implies that mental aspect.



On a personal note, I've always regretted being disrespectful to people over the Internet. It's just the worst feeling when you turn out to be wrong, and then you're both wrong and an asshole. I want to be only one of the two at worst.

[Leo_The_Lip]

The biggest questions I have with any clutch study or game implementation are:

"How do you know if a clutch pitcher is facing a clutch hitter?" And how do you determine this?

"How do you know if a choking pitcher is facing a choking dog hitter?" And how do you determine this?

In-game, how does one program this? Are the situations the same, that is, are the chances of a hit exactly the same in these two situations?

My personal bias is that no one is clutch, however, EVERYONE chokes to a greater or lesser degree. My favorite observation comes from basketball where Al Maguire once said, "With the game on the line, I want the C+ student taking the shot because he isn't smart enough to recognize the pressure."

[5-4-3]

What nonsense. It doesn’t exist. It’s just a myth.

You people who believe in it are believing in baseball equivalent of the Tooth Fairy. I can only wonder how many of you believe that the earth is flat as well.

[Syd Thrift]

Originally Posted by 5-4-3

What nonsense. It doesn’t exist. It’s just a myth.

You people who believe in it are believing in baseball equivalent of the Tooth Fairy. I can only wonder how many of you believe that the earth is flat as well.

Are you a troll?

[RonCo]

All good statisticians are careful to point out the they cannot prove a negative. So reputable studies about clutch hitting can only say "we can't find it." For me, that's good enough to say it should not be modeled in a game.

It's much like catcher defense. The Bill James's of the world worked hard to see what kind of impact catchers had on defense for many years, and basically said "we can't find any--or at least not any that really differentiate between catchers." It took advanced work on framing to really find anything. Time moves on. If clutch hitting actually exists, there will be a way to quantify it.

Until then, argue all you want.

[NoOne]

when you can't prove somethng to exist, it's basically saying it doesn't.. .even though technically you cannot say it for certain.

people believe in all sorts of things that are not real. mass hysteria abounds even today.

ffs, there are still "flat-earthers." i gues sthey think the world is like a pac-man screen and you just pop out hte other side? we've circumnavigated the world more than once and it's still a debate for some people, lol.

some refuse to believe the world is only 5000 years old too.

clutch only requires a redefinition for it to be real... nothing can go above and beyond what is possible, and if it is possible you should learn how to do it all the time instead of only during select time periods, lol, which would not technically be 'above and beyond' so it's a catch-22...

why give up all that extra money just to be "clutch" in a short window of time? seem more like a lack of efforct/care in all other situations than "clutch" lol.

[RonCo]

Originally Posted by NoOne

when you can't prove somethng to exist, it's basically saying it doesn't.. .even though technically you cannot say it for certain.

people believe in all sorts of things that are not real. mass hysteria abounds even today.

ffs, there are still "flat-earthers." i gues sthey think the world is like a pac-man screen and you just pop out hte other side? we've circumnavigated the world more than once and it's still a debate for some people, lol.

some refuse to believe the world is only 5000 years old too.

Your examples aren't appropriate for the discussion at hand. Each of those items have data that proves the existence of a different framework, whereas the conversation around "clutch" has only the fact that you can't find it.

For most, at some point, the lack of evidence is enough. But as Carl Sagan said, an absence of evidence is not evidence of absence.

Again, though, we're talking about a computer simulation, so the studies that have been done to date are reason enough to exclude "clutch" from the game, IHMO. They are not, IMHO again, enough to berate people for still thinking (hoping) it might exist. Human beings and human minds operate in weird ways.

[RonCo]

Literally every discovery that humans have ever made were at one time unable to be seen in the existing data. All you can say for certain about Clutch is that no one has been able to find a perspective from which to look at performance data that will prove it exists.

At best today (simplifying a bit) if you want to hold onto the idea of clutch performance all you can say is that it's so small that the sample sizes needed to find it are much larger than the data we've got. But maybe the next ten years will find some way to look at more advanced data and find a different vein.

Who knows?

Another question here, though, is what value does one get out of closing the door on the idea? Why does it matter to you that the existence of "clutch" is settled for once and for all, even though you can't prove it to 100% certainty? In this context ,the only thing that matters to me is that the concept be constrained away from the game (which I honestly don't know is true or not, but for my preferences it would be).

[NoOne]

really, we do know.

genetics, diet, muscles etc.. somethings we can control some we cannot.

you cannot willfully improve your talent relative to contexts with no tangible connection. it's just illogical, that's all. it defies logic and well-understood concepts of biology, physics, and probability.. the things we can control are long-term effort, and can't flip a switch at 7pm on october 25th etc. a playoff game context doesn't make you stronger or improve your hand-eye coordination... simply impossible.

it's small sample for any one player, but all players considered since 1871, and thos contexts provide enough data to know it's random and sample size releated, because it's statistically insignfificant relative to average.

[RonCo]

There are many things about the human body that we know, just as there are many things that we do not. We know it undergoes chemical changes when under stress. We know it has physiological changes. We know that those changes are not voluntary--not able to just be called upon at out command. We know that some bodies do different things than others. We know that periods of intense stress can result in the body achieving things it would not do under other more normal situations.

I agree totally that clutch as it's discussed by most people today is not able to be found. I agree it should not be part of the model unless it's ever able to be found. We can leave it there.

But I see no value in planting a flag that says there will never be a way to find that something involuntary happens inside certain people at certain times that causes a change in how they perform. If that's important to you, then it's important for no real reason I understand. It's all good though.

[Syd Thrift]

Originally Posted by RonCo

All good statisticians are careful to point out the they cannot prove a negative. So reputable studies about clutch hitting can only say "we can't find it." For me, that's good enough to say it should not be modeled in a game.

It's much like catcher defense. The Bill James's of the world worked hard to see what kind of impact catchers had on defense for many years, and basically said "we can't find any--or at least not any that really differentiate between catchers." It took advanced work on framing to really find anything. Time moves on. If clutch hitting actually exists, there will be a way to quantify it.

Until then, argue all you want.

You know that there *is* statistical evidence for pitch framing making a difference, right? Based on evidence that we didn't have available to us before about a decade ago? It's not huge, and since teams realized that it is indeed a thing and not another Baseball Truth it's mostly gone away, but if that's your parallel, I'm all for it.

Again, this isn't about proving a negative. This is about exciting new evidence that will get us closer to truths about the game. And I don't blame you for not reading the first couple pages of this but I for one am not and have not been arguing that clutch is about mental fortitude or whatever. If it exists, it probably exists the same way clutch exists in basketball: certain profiles do better than others in pressure situations. Maybe you'd expect that to show up in the old fashioned stats but maybe it will turn out to be that, say, fastball hitters will otherwise outplay other types of hitters in those situations, all else being equal.

Or it might not. As a skeptic, I'm going to roll with the evidence wherever it lies. I'm not going to sit there and defend the orthodoxy against the unenlightened hordes because that's not really what skepticism is about, and quite frankly that's what led many in the movement to deny climate change. You get into that mindset that it's all about orthodoxy, it's a very small step to going out and finding orthodoxies that fit in with your own opinions rather than accepting the often ephemeral nature of established science.

[RonCo]

Originally Posted by Syd Thrift

You know that there *is* statistical evidence for pitch framing making a difference, right? Based on evidence that we didn't have available to us before about a decade ago?

Yes. My point was that there was a time when sabermetrically inclined people (of which I am one) were going around saying catcher defense didn't matter because it's impact couldn't be detected in the data that Bill James and several others were looking at. Today we know framing is a thing--though less major than originally advertised when it was first looked at properly.

In this light, all the people running around saying differentiation due to catcher defense didn't exist were wrong, despite all the studies around that were unable to see it. It did exist. It just didn't exist in the datasets and methods that were being studied.

In that light, this is about whether you can prove a negative.You can never say "clutch hitting" doesn't exist. All you can say is that there has never been a test devised that has been able to find it.

[RonCo]

Operationally, if you're a baseball GM, then, you look at all the data you can and have to decide how to weigh the idea. If your scout comes in and says "this guy is great in the clutch" you make a decision based on how you assess these things. There are many reasons you might make a decision to go with your scout's opinon regardless of whether "Clutch" exists.

If you're a computer game designer, though, you look at the problem differently. If you consider putting "clutch" into the game and you don't have any data whatsoever to use to design your algorithm, you risk screwing things up if you just pull the algorithm out of your tail end.