Concepts some posters could benefit from learning

[RayClay]

Not for anything but he hasn't exactly been proven wrong in his clutch vs. variance argument.

This is a meaningless phrase, as far as I can see.

Could you explain what it means? I have a textbook right here so i can look it up.

*Variance actually means something in statistics, but it's never vs. anything*

 

[makewayhomer]

I know you don't that is my point. So you think that the effect on the SNS system of a human being is the same for a kick or an at bat in the beginning of a game compared to the same occurrance in the last few seconds or last at bat witht he game on the line, let alone with a championship on the line?

That is "pressure" and it has a PROFOUND effect on the human body and motorskills. "Clutch" athletes are those that produce in that instance and sometimes are able to focus the pressure in a positive way that others are not.

Oh and nice Wiki reference...you do realize that anyone can write that right? It is not even remotely a "fact". Also humor me. You never played a higher level of sports as a starter did you?

I'm just going to repeat what I have said a thousand times, but you haven't listened to. if what you are saying is true, and athletes were PROFOUNDLY affected by pressure, then we would be able to see this PROFOUND affect in their results. but there very little evidence of this. go read through some of the articles I linked, if you don't like wiki.

You never played a higher level of sports as a starter did you?

I played division III football in the 90's. I don't think it matters much though.

 

[makewayhomer]

This is a meaningless phrase, as far as I can see.

Could you explain what it means? I have a textbook right here so i can look it up.

*Variance actually means something in statistics, but it's never vs. anything*

Clutch = the skill to come up with big hits at big moments. if it's truly a skill, then it must be repeatable. if it is not repeatable, then it's...

Variance. Which basically means that players generally perform as expected in big moments. good players perform well, bad players perform poorly. an "unexpected" performance is simply the result of normal performance distribution and can not be attributed to extraordinary skill. if it was skill, he would be able to repeat it over a large sample of clutch situations

 

[emoney_33]

This is a meaningless phrase, as far as I can see.

Could you explain what it means? I have a textbook right here so i can look it up.

*Variance actually means something in statistics, but it's never vs. anything*

This is kind of nit picky on your part. What I meant is he hasn't been proven wrong in his argument that players do not possess an ability to increase their game in pressure situations. He can also not prove for a fact that a player's average output in pressure situations is a standard deviation based on nothing but luck.

 

[Fencer]

His mean batting average is .285. .352 and .369 are well above the mean. That isolated, bur essential (changed RedSox history) is way above the mean.

What are the odds these clutch performances are random given all that data?

Unstoppably clutch is meaningless, nobody hits .1000 over time.

His weekend, aggregate world series and playoff BAs are all well over his mean average.

One (at bat) incidence of anything cannot be statistically significant. No measure so far is less than or even close to his mean batting average.

Another way to measure would be by comparison, say to Ted Williams.

Without running the numbers, I'd guess that Yaz's superiority in postseason to regular season hitting would pass most tests of statistical significance.

EDIT: Only 65 trials? Yikes. I was too quick. I retract the guess. That's only 5 or 6 total hits above his lifetime batting average.

 

[Deus Irae]

Clutch = the skill to come up with big hits at big moments. if it's truly a skill, then it must be repeatable. if it is not repeatable, then it's...

Variance. Which basically means that players generally perform as expected in big moments. good players perform well, bad players perform poorly. an "unexpected" performance is simply the result of normal performance distribution and can not be attributed to extraordinary skill. if it was skill, he would be able to repeat it over a large sample of clutch situations

Given that "clutch" is not strictly baseball related, and is not strictly confined to hitters, your "clutch" definition is clearly erroneous.

 

[makewayhomer]

Given that "clutch" is not strictly baseball related, and is not strictly confined to hitters, your "clutch" definition is clearly erroneous.

we were talking about baseball.

find another nit to pick, or if that is the most intelligent thing you can contribute, then just stop posting.

 

[Fencer]

Yes and it would reveal many ways of testing differently to draw inferences.

For instance you keep mentioning variance. I bet you don't know what it means in statistics. The only thing it means is related to standard deviation, which I've mentioned about three times and you don't seem to be familiar with.

This is Variance



In a normal distribution with 2.61 hits per 8 trials being the mean (for a .326 hitter), perhaps a statistician can work out how many standard deviations 7 hits would be.

RayClay, I've seen many cases where "variance" is used to mean "sample variance." Your claim that it can only be used to mean "population variance" is incorrect.

 

[LiveShot]

whatever, it doesn't matter, I should have been more clear, but it doesn't change the larger point at all

It just proves that variance pertains to truth as you so aptly proved.

 

[RayClay]

Clutch = the skill to come up with big hits at big moments. if it's truly a skill, then it must be repeatable. if it is not repeatable, then it's...

Variance. Which basically means that players generally perform as expected in big moments. good players perform well, bad players perform poorly. an "unexpected" performance is simply the result of normal performance distribution and can not be attributed to extraordinary skill. if it was skill, he would be able to repeat it over a large sample of clutch situations

You don't know what variance means. It does have a meaning in statistics and that's not it.

 

[RayClay]

Without running the numbers, I'd guess that Yaz's superiority in postseason to regular season hitting would pass most tests of statistical significance.

EDIT: Only 65 trials? Yikes. I was too quick. I retract the guess. That's only 5 or 6 total hits above his lifetime batting average.

Unless you're talking yankees, any post season results would have a relatively small sample. Nevertheless, none of the two post season averages or the weekend production are less than the mean for the season or lifetime .285 for lifetime, none are close either.

Given the evidence that is available, It is very difficult to prove it is random.

What is the probability clutchness is correct? Not as much as if he had 300 post season at bats, but clutchness certainly is more probable than randomness.

Statisticians work with small samples and hard to quantify data (rather than batting avg. a stat everyonoe knows) all the time.

 

[RayClay]

RayClay, I've seen many cases where "variance" is used to mean "sample variance." Your claim that it can only be used to mean "population variance" is incorrect.

There's about 8 definitions in the dictionary too. He's using it to mean randomness, which it is not.

 

[shirtsleeve]

Sabremetrics? I tried that once at a health club. Walked bow legged for 3 weeks.

 

[RayClay]

Sabremetrics? I tried that once at a health club. Walked bow legged for 3 weeks.

I'm wondering if this is Bret Saberhagen trying to get some free advertising on the site.

 

[Patriotic Fervor]

PF, there are various successfull baseball teams who are now SABR run organizations, and almost every team has at least 1 SABR type guys on their staff. the Red Sox, Indians, and A's chief among the SABR run team. this stuff IS very prevelant and is being used to make decisions today, whether you know it or not.

Everyone runs numbers. But it's not the defining feature of an organization.

Crunching numbers isn't management, it's a function of management. Before my retirement, I worked in organizations where we did this stuff all the time, but it's only purpose was to aid in making what was deemed the proper decision, at a given time, for the corp. That's all. Didn't rule our lives, didn't make the evening beer taste any better. As BB says, it was what it was.

Invariably, there'd be many variables to consider in any given decision, and there is always an element of uncertainty. The numbers would be useful, yes, but more often than not they weren't the engine that drove the company. They'd be an adjunct to the decision-making process.

In the example you've given below, especially the bolded part, the organization in question seems to realize this. Their terms are "objective" and "subjective", which are good terms, but it's important to remember that each is merely a part of the whole. They'll use both in an effort to minimize the uncertainty (this is not "variance") and maximize productivity/profitability. That's the purpose of management.

I have never said that #'s should be 100% of the decision. they are extremely useful, but the human element must always play a part in the decision making process. this is why the pioneering SABR baseball site Baseball Prospectus has Kevin Goldstein on their staff, who is an old school crusty scout type, not a statistician.

and here is a real world example from just this month from an online chat with the new Pirates GM.

The Pirates upper management has widely ignored OBP (on base percentage) in the past. How important will OBP be in player evaluation under your leadership?
-- Eric S., Pennsboro, W.Va

We are going to utilize several objective measures of player performance to evaluate and develop players. We'll rely on the more traditional objective evaluations: OPS (on base percentage plus slugging percentage) , WHIP (walks and hits per inning pitched), Runs Created, ERC (Component ERA), GB/FB (ground ball to fly ball ratio), K/9 (strikeouts per nine innings), K/BB (strikeouts to walks ratio), BB%, etc., but we'll also look to rely on some of the more recent variations: VORP (value over replacement player), Relative Performance, EqAve (equivalent average), EqOBP (equivalent on base percentage), EqSLG (equivalent slugging percentage), BIP% (balls put into play percentage), wOBA (weighted on base average), Range Factor, PMR (probabilistic model of range) and Zone Rating.

That said, we will continue to stress the importance of our subjective evaluations. Succinctly stated, we believe that a combination of quality objective and subjective analysis will allow us to maximize our probability of success and to make the best possible decisions

Numbers have a place in every organization that adheres to the concepts of scientific management. They have import, and importance, but only to the statistically inclined, and their engineering brethren, do they seem to take on a life of their own.

 

[Fencer]

There's about 8 definitions in the dictionary too. He's using it to mean randomness, which it is not.

I'm not sure he's much further off than you are.

The question is whether the variance is noteworthy or one that was within an a priori likely range under the null hypothesis. Neither you nor he was perfect in your phrasing of the issue.

Anyhow, try http://www.stat.tamu.edu/~west/applets/binomialdemo.html

I used 65 trials, .285 probability of success, 24 actual successes. Probability of having 24 successes or higher was .088.

So we do NOT have statistically significant proof of clutchness.

 

[Deus Irae]

we were talking about baseball.

find another nit to pick, or if that is the most intelligent thing you can contribute, then just stop posting.

Why? You've been exposed as a complete fraud and it's fun watching you act as if you really know what you're talking about.

 

[RayClay]

I'm not sure he's much further off than you are.

The question is whether the variance is noteworthy or one that was within an a priori likely range under the null hypothesis. Neither you nor he was perfect in your phrasing of the issue.

Anyhow, try http://www.stat.tamu.edu/~west/applets/binomialdemo.html

I used 65 trials, .285 probability of success, 24 actual successes. Probability of having 24 successes or higher was .088.

So we do NOT have statistically significant proof of clutchness.

I'll have to trust you on that.

You're not saying his postseason success was random in any of the three cases are you?

You are the only one in this thread that has the slightest idea what your talking about. All I know is the thread starter does not. And I did not start a thread to help people learn some concepts they should know.

I only read these things from a text, but I looked up binomial test and it mentions it with a roll of the dice example. Being that dice only have twelve sides, it seems this test would be perfect for the 7/8 hits against Minnesota example, did you run that?

 

[Fencer]

You're not saying his postseason success was random in any of the three cases are you?

I'm saying that the numbers give no reason to believe it was not random.

Proving non-randomness is straightforward. Proving randomness is much, much trickier.

 

[emoney_33]

I'm not sure he's much further off than you are.

The question is whether the variance is noteworthy or one that was within an a priori likely range under the null hypothesis. Neither you nor he was perfect in your phrasing of the issue.

Anyhow, try http://www.stat.tamu.edu/~west/applets/binomialdemo.html

I used 65 trials, .285 probability of success, 24 actual successes. Probability of having 24 successes or higher was .088.

So we do NOT have statistically significant proof of clutchness.

We can't possibly use this binomial distribution to prove or disprove clutchness right?

This test assumes the probability of the event is .285, but the argument of clutchness is that this probability rises in pressure situations. If we use this on the basis that a player cannot become better in pressure situations, then this only tells us that at least 24 successes out of 65 tries has a probability of .088, correct?


Using the binomial distribution in this discussion seems futile because it is operating under the pretense that "clutchness" does not exist.


To look at proof of clutchness or not, we must study all players' averages in "clutch" situations vs. all other situations. But this will only tell us that most players did not show much improvement over their norm in the clutch. The argument that CERTAIN players posessed this potential human quality can't really be proven or disproven based on their performance numbers. I'd guess it would be more a study of the human brain, affect of extra adrenaline, etc. etc...

Am I wrong here?