Concepts some posters could benefit from learning

[fair catch fryar]

14 pages of math? I hate math. Maybe I'm in the minority here but I could give a crap what advanced algorithims are going to tell me about clutch vs non clutch. I know that I love having #12 as my QB and I loved having #4 as our kicker.

I've never had a bad feeling with them on the field (except watching Adam on the field with the Colts which makes me want to throw up) and they both brought a lot of happiness to this city and it's fans. Is it really necessary to try and negate all of those feelings because statistically they can't be clutch?

I'm starting to go there too!! These guys can't seem to understand that no matter how hard they "spin" their argument, I just don't believe there is a mathematical explanation for "clutchiness." I believe it is innate, some people are born with it and some are not, while others can obtain it through experience, practice, and reflection.

 

[makewayhomer]

So why haven't the Yankees been able to perform in "clutch" situations with all of their "statistically superior" talent in recent years??

if you're asking why the Yankees haven't won a WS in recent years, the answer is similar to the one I gave in my OP about the Patriots winning all 3 of their SB's, with a twist.

in baseball, even a bad team will beat a good team a healthy % of the time in a short series (5 or 7 games). if the Yankees are 60% or even to 70% to win a series, that makes them a big favorite. that means that a full 30% to 40% of the time a clearly inferior team will beat them - that's a lot!

in addition, teams need to win 3 straight series to win the WS these days.

so if you assume the best team is a 70% favorite to win each series, that means that win all 3 (.7 * .7 * .7) or 34% of the time. basically 1/3 of the time - that means even the best team can only expect to win every 3 years.

the Yankees ran really well when they won a bunch of WS in a row, and ran really poor more recently.

 

[Joker]

What makes me happy is winning. if advanced analysis can help my team win, then I'm all for it.

My school of thought is this.."the race doesn't always go to the swiftest runner nor the fight to the strongest guy...but that's the way to bet"

 

[Tommysgirl]

Tommy,

What makes me happy is winning. if advanced analysis can help my team win, then I'm all for it.

this isn't about taking away credit, it's about using information to make the best decisions to help the team win

I do understand about advanced analysis and stats and such I just don't think clutch is one to debate. When Adam came out on the field to kick the 29 yarder to win the game nobody thought about statistical analysis I would think 99% of the people watching thought that was money in the bank. That has nothing to do with stats and everything to do with watching this man over the last 7 years.

 

[makewayhomer]

I'm starting to go there too!! These guys can't seem to understand that no matter how hard they "spin" their argument, I just don't believe there is a mathematical explanation for "clutchiness." I believe it is innate, some people are born with it and some are not, while others can obtain it through experience, practice, and reflection.

fair, nobody is trying to explain it.

however, if "some people are born with it and some are not, while others can obtain it through experience, practice, and reflection" then the statistics would support that, and they don't. baseball is great in that it's allowed us to study these big sample sizes

 

[makewayhomer]

What makes me happy is winning. if advanced analysis can help my team win, then I'm all for it.

My school of thought is this.."the race doesn't always go to the swiftest runner nor the fight to the strongest guy...but that's the way to bet"

hey we agree

 

[Fencer]

Really, so your telling me that you can take any Joe Scmoe off the street and calculate whether or not his mental fortitude and experience can be measured. Take away the fact that natural athletic ability is a pre-requisite, but everyone has a heart and a mind, and these can not be measured by statistics. It's OK to disagree with me and not call it "BS", but what I'm saying is that some people are able to focus and perform while others "freeze" and choke in certain situations. Two athletes in similar situations are not going to have similar statistics over the same period of time in "clutch" moments.

Please get a grip on the difference between the present and future tense of verbs. Then maybe we can try to have a reasonable discussion on this subject.

Or we can just agree to wildly disagree.

 

[Fencer]

I hate math.

Do you also hate reading? One is pretty much as basic as the other.

 

[Tommysgirl]

Do you also hate reading? One is pretty much as basic as the other.

No, I love to read. I don't think it's a matter of basics I'm not in 4th grade anymore. Some people such a yourself, love to spend a lot of time and get really in depth with mathematical analysis others don't. I don't think that's a bad thing. People are different.

I'm also grown up to admit what are my strong suits and what aren't. Math simply isn't one of them. Fortunately, for me it's not a great downfall in my life.

 

[emoney_33]

I hate this stuff, but. You can't really "prove" anything. What you try to do is estimate the probability something happened by chance. If you try to prove the null hypothesis, that means you proved a high probability it could happen by chance.

You wouldn't test batting against a coin flip, nless you know lots of .500 hitters.

I can't crunch the numbers, but random chance says Yaz had a 3/10 chance of getting a hit each time because hw was a .326 hitter. He performed almost almost 9/10. (7/8 or .875).

You put that stuff on a curve and it's way off, It could be random, but it seems significant for a small sample.

Things regress to the mean, in this case, his world series and playoff performances regress closer to his mean average and they do.

His mean average .285 career. His 1967 .326 was his best.

That means his career WS .352 and playoff .369 are still way over his mean .285.

Looks statistically significant to me.

Are you suggesting Yaz never had a streak of 7/8 or close to it during the regular season?

You can't really judge performance in 8 at bats against his career average of thousands of at bats.

I can flip a coin and get heads 9 out of 10 times. The odds of getting heads never changed, it remained 50% but sometimes I'll get it at a 90% rate over a TINY sample size. I flip that coin a million times and there's no way I'm getting that much variance.

 

[fair catch fryar]

Please get a grip on the difference between the present and future tense of verbs. Then maybe we can try to have a reasonable discussion on this subject.

Or we can just agree to wildly disagree.

Hey professor..put this in past tense...G.F.Y.! You go from telling someone they can't have an opinion on a subjective topic to telling them how to use verbs on a fan's messageboard. Maybe someone 'ought to kick your soapbox out from underneath you and give you a little humble pie to eat.

 

[aluminum seats]

Wow, lot of responses here.

By the way, I love phrases like this......"now, I realize this is a foreign concept to most of you...." Really makes you sound smart.

Anyway, the issue of "clutchness" of been pretty much beaten to death, in that by definition truly huge moments won't come up often enough to really generate any kind of decent sample size. So, in football, you have to largely go by what you see. AV was clutch, hit clutch kicks. Beyond that, it's wasted words......it is what it is.

 

[RussFrancis]

Random chance is going to factor into everything in life. Its not a new concept. But determining how some rise to the occasion or are more 'clutch' than others under pressure boils down to a few vital elements: preparation, skill, poise, focus and execution. And courage. Some do more of each or have more of each than others.

Success isnt without risk. The attempt may result in failure or pain. It means knowing what youre trying to accomplish may come at a cost. It may mean youll get hammered or have your head ripped off while executing a play. But youre willing to suffer the pain. The Kurt Warner TD pass to Isaac Bruce in the 2000 Superbowl that won the game is a good example. A perfect example of a 'clutch' play. The man stood in the pocket with heat directly in his face bearing down, got completely lit up the second he released that ball, but had the courage to stand in there, make that throw and deal with the pain. Most QBs simply arent willing to make that sacrifice and take that punishment. Theyd see a 290lb guy running at them and fold like a blanket. Not the special guys. The guys with the medals, the rings, and the most respect. The guys with the courage needed to maintain focus and execute knowing the costs are 'clutch'. They dont always succeed, but theyre willing to take the chance. The Brady pass against Denver in the playoffs that got picked off by Champ Bailey and returned 100yds the other way comes to mind. Brady saw the best corner in the game covering his most clutch receiver, Troy Brown, and had the balls to make the throw. And a great player simply made a greater play and won the moment. But the fact that Brady had the guts to make the throw despite the risk are what you take away from the play. He's not always gonna win. But it wasnt 'chance' that decided that fate. Sometimes, the other guy or the other team is just better.

So far this year, NE has just been better. Theyve been more clutch. It hasnt been by chance. Its been a result of skill and hard work. And courage. The hospital pass Moss caught in the Colts game to jump start an important drive, rising several feet in the air to make a brilliant one-handed snag in the middle of the field surrounded by defenders comes to mind. The man knew the risk of what he was doing. He could have let that one go with good reason. But he stood up, rose up, and was clutch.

Courage and skill. I'll take my 'chances' with it every time.

 

[scout]

You bring up very rare occurences and the human factor to combat statements. You are correct in the sense that these blanket statements are not true 100% of time (what is true 100% of the time though?). However sabermetrics is based around the idea that the sample size has to be large enough. In general these statements are true, which you could see better if you compiled the numbers over decades, hundreds of players, thousands of at bats, situations etc...

Which reminds me a little of hold em odds. Professional poker plays make money in the long run because they stick with the odds in their favor. Of course bluffing, reading opponents etc.. come into play more at that level. In the very basic sense playing holdem, if you always make "good" plays you will make a profit in the long run. Calling an all-in when you have a 67% chance to win the hand is always the right call, even though 33% of the time you WILL lose.

Also I think the "human" element is overrated a ton in sports today by the fans, media, analysts etc.. I can't PROVE it obviously, although I believe I remember seeing a study where there was only one major league baseball player ever that had better "numbers" in "clutch" situations than his average in all situations.

Whether an insane adrenaline pump in a tight situation can increase professional ability, well that is not something I am least bit qualified to discuss lol.


Football is a different game, different factors and a far smaller sample size, so it's much harder to study these things for football. The effect coaching and teamwork has on an NFL game is a ton more than the effect teamwork and coaching has on an MLB game.

The theory that Homer brings into the discussion is flawed. It does not bring into account the human element. Back in the Tom Landry days of the Dallas Cowboys, they used a drafting technique which was interpreted the same way as this theory. It did not take into account characteristics of the person. After missing out on some great athletes, the last one I recall was Mike Singletary at Baylor, the strategy was scraped. I think they said he was too small or something.
In the baseball example, the author mentions a walk as being equal to a home run in the ninth inning. In other words, a walk to Big Pappi would have the same effect during a playoff game as a home run? Ever hear the fans at Fenway rejoice after a "clutch" home run. Then, the pitcher has to deal with those frenzied fans and the next batter (Manny). Seriously, his next pitch has no variance between the two options. Another example is a walk is equal to a single. Yes, taken at its face value. But the sport is more complicated then that. The runner is going hard to first, and the infielder hurries the throw causing a (human) error. Everyone of the authors reasons can be disputed because they lack the human element. How about Adam Vinartieri? My description of clutch is used when there are no more kicks possible. Time is out and the game is on the line. Does the sound from the fans from an indoor stadium play into the factor? Does failure play into the equation (Donnie Moore). Opposing coaches call time outs before/after kicks, is this a waste of time? It is well document, that in any sport a player will rise to the occasion. Reggie Jackson (Mr. October), Robert Horry, Ken Dryden, Larry Bird, USA-USA-USA Olympic hockey team, James Braddock, and the list is endless of athletes who rose to the occasion. You cannot put any variable on whatever drove them to that point. To say that football is different or that it will all come down to a theory is pure BULL. Tell it to the players who overcame all odds to win everything.

 

[Rainy-Day Patriot]

The first post prompts a fundamental confusion. (The poster may or may not be fundamentally confused--I don't know.)

To translate the poster's comment into classical test theory:
For any so-called clutch performer we have merely a sample of performances from an infinite domain of possible clutch performance situations. If the performer got to replay all the relevant clutch situations an infinite number of times (without fatigue or learning), sometimes he would succeed, sometime he would fail, but his success rate would be a really nice thing to know--this would be his "true" score. Of course we never get an infinite number of trials. We are always stuck with a sample. Because there is variance, we can never be sure about our inferences from the sample of performance to the domain of performance. This is why good test reporting uses error bands. The poster's main point, which is a good one, I believe, is that before we get all fired up about the performances that we observe from Vinatieri, Brady, Ortiz etc., we should consider the error bands that would inform our inference from the sample to the domain. Welcome to the world of quantifying measurement error.

Here is my concern:
Error bands go in both directions. The post might lead a reader to believe that error bands for outliers (e.g., superstar clutch performers) only stretch toward the average. While Vinatieri, Brady, or Ortiz might be the "beneficiaries" of variance, they may equally be the "victims" of variance.

Here is the tricky tricky part:
My point is only true for individuals cases! When we look at groups, we do want to pull in the outliers to account for measurement error. Measurement error artificially spreads out groups of performers when we compare them. However, when it comes to assessing the performance of a single individual, which I think we are trying to do, observed performances are just as likely to be above or below the average performance over infinite trials. I.e., variance 'helps" just as much as "hurts." Imagine if Ortiz, as kick-ass as he is, is nevertheless the victim of BAD luck!

BTW, great post!

Sean

 

[PatsWickedPissah]

One interesting football related case study might be to look at QBs ratings in the regular season vs in the playoffs. Yes, the sample size for any significant stats based conclusion is alas, too small. You'd have massive under achievers like Rthbrgr with something like a 22 rating in his SB win vs Tom Brady whom I'll conjecture had higher QB ratings in the playoffs leading to his SB wins than he did in the preceding regular seasons. That's excellence under pressure. And even more amazing given that the quality of the opponent in post season is markedly better than the average opponent in regular season.

However it's most unlikely that he'll have as high playoff ratings in Jan 2008 as his off the charts 2 to 3 sigma extrordinary regular 2007 season. Too much variance to overcome.

 

[makewayhomer]

One interesting football related case study might be to look at QBs ratings in the regular season vs in the playoffs. Yes, the sample size for any significant stats based conclusion is alas, too small. You'd have massive under achievers like Rthbrgr with something like a 22 rating in his SB win vs Tom Brady whom I'll conjecture had higher QB ratings in the playoffs leading to his SB wins than he did in the preceding regular seasons. That's excellence under pressure.

I think his playoff QB rating is pretty similar to his regular season rating, I'll have to look

 

[makewayhomer]

The first post prompts a fundamental confusion. (The poster may or may not be fundamentally confused--I don't know.)

To translate the poster's comment into classical test theory:
For any so-called clutch performer we have merely a sample of performances from an infinite domain of possible clutch performance situations. If the performer got to replay all the relevant clutch situations an infinite number of times (without fatigue or learning), sometimes he would succeed, sometime he would fail, but his success rate would be a really nice thing to know--this would be his "true" score. Of course we never get an infinite number of trials. We are always stuck with a sample. Because there is variance, we can never be sure about our inferences from the sample of performance to the domain of performance. This is why good test reporting uses error bands. The poster's main point, which is a good one, I believe, is that before we get all fired up about the performances that we observe from Vinatieri, Brady, Ortiz etc., we should consider the error bands that would inform our inference from the sample to the domain. Welcome to the world of quantifying measurement error.

Here is my concern:
Error bands go in both directions. The post might lead a reader to believe that error bands for outliers (e.g., superstar clutch performers) only stretch toward the average. While Vinatieri, Brady, or Ortiz might be the "beneficiaries" of variance, they may equally be the "victims" of variance.

Here is the tricky tricky part:
My point is only true for individuals cases! When we look at groups, we do want to pull in the outliers to account for measurement error. Measurement error artificially spreads out groups of performers when we compare them. However, when it comes to assessing the performance of a single individual, which I think we are trying to do, observed performances are just as likely to be above or below the average performance over infinite trials. I.e., variance 'helps" just as much as "hurts." Imagine if Ortiz, as kick-ass as he is, is nevertheless the victim of BAD luck!

BTW, great post!

Sean

agree with all, great post

 

[makewayhomer]

Wow, lot of responses here.

By the way, I love phrases like this......"now, I realize this is a foreign concept to most of you...." Really makes you sound smart.

.

in hindsight, I probably could have chosen better words. apologize if I insulted anyone.

I'm not trying to say that I'm smarter than anyone; however I am saying that the conventional wisdom often spouted by the media and even players/coaches is often wrong, and since we hear that all the time, we take it as gospel truth.

the reality is much of conventional wisdom is wrong, but many people don't realize it, b/c hey when you hear something your whole life you take it as truth. that is what I was trying to point out in my OP.

if I insulted anyone, apologies, but sometimes saying some outrageous is the best way to get people thinking

 

[aluminum seats]

in hindsight, I probably could have chosen better words. apologize if I insulted anyone.

I'm not trying to say that I'm smarter than anyone; however I am saying that the conventional wisdom often spouted by the media and even players/coaches is often wrong, and since we hear that all the time, we take it as gospel truth.

the reality is much of conventional wisdom is wrong, but many people don't realize it, b/c hey when you hear something your whole life you take it as truth. that is what I was trying to point out in my OP.

if I insulted anyone, apologies, but sometimes saying some outrageous is the best way to get people thinking

No problem, and props to the thread, which is a nice shift from "look who wrote something mean about the Pats!"