Concepts some posters could benefit from learning

[CrazyDave]

Wow, just...wow.

 

[pats-blue]

you don't know what you're talking about. you are saying that Palmer created a stat that said X, and someone else used parts of Palmers research to say Y. that is perfectly reasonable

No you don't know wht you are talking about. I am talking about PALMER created a stat that he considered to measure offensive performance of a player that corrolated directly to WINS. Your guy used PALMERS stats for his paper and made a STATEMENT that said the opposite. That is not reasonable at all unless you are saying that there is not really a fact based conclusion only complete BS. (still have not seen one single fact or stat to back up his statement, unlike plenty of stats I provided BTW...where is it?) I will also point out you completely avoided that FACT that this guy was talking about the Sox in a pennent race in 69 and 70 when the Sox finished 21 and 22 games out yet referenced how Yaz didn't hit well during the games that mattered to te pennent? Please address this as to me that really calls into question ANYTHING that comes out of this guys mouth because he is obviously clueless.

Also just FYI...if someone says it is "highly overrated" and another says it is "generally underestimated" that is NOT considered the same in fact it is refuting your argument in your original post that was going to explain to us this "foriegn concept to us" how we could "benefit from learning" from you. Unfortunately the link you provided actually disputes your "theory". IN fact the entrie article actually thrashes pretty good Sabremetrics as it relates to "clutch" and is the entire point of the article. So there you have it PROVEN that you can make stats say anything you like since you linked two articles that come to exactly OPPOSITE conclusions. Boy this was fun!

 

[Fencer]

That's the funny point, statisticians will never say so. The data suggests, research indicates....

And rightly so.

 

[PatsWickedPissah]

The nuclear option is hereby invoked...

 

[Fencer]

I'm in graduate school and it's the end of the semester. Actually statistics is about as far from my areas of interest as anything could be, even if I had the time right now.

I'm sure there are some mathematically inclined fans who will read those articles and explain them to you if you ask them nicely.

I think he's pointing out that the article, in a very simple way, blows theories like yours all to hell.

It uses a different metric of clutchness than you do -- namely, important situations in regular-season games. The benefit of that choice, obviously, is that every player on every team can be measured every season.

And it provides strong evidence against the theory that that kind of clutchness goes beyond random chance.

 

[Fencer]

7/8=35/50? 7*5=35 8*5=40

What's the success rate of 7/10 then?

If you're so curious, why don't you go to the link I gave earlier and type in a few numbers yourself?????????

 

[Fencer]

I don't know how 7/10 could come into it when we're analyzing 7/8 at bats.

Dude, YOU'RE the one who brought 7/10 up. Nobody else did.

As for the analogy between 7 trials of 8 at .326 probability vs. 35 trials out of 50 at .5 probability -- that's been explained multiple times already!!!!!!!!!!!!

 

[Fencer]

Is it because there are more "stats" to pick from

More precisely, it's because different atbats in baseball are more similar than different plays in football. There are no differences in field position, and you're not relying on your teammates. The main differences are opposition, weather, and whatever is going on with tiredness/clutchness/stress/whatever.

What's more, because baseball is a game of precision, one might hypothesize there was a greater difference due to adrenaline than in some other sports. Adam Vinateri may need to fight off the physical shakes to preserve his aim, just as a pitcher or batter does. But that's less of an issue for Brady, and less of one yet for clutch passcatchers David Givens and Mike Vrabel.

 

[Fencer]

I think he is doing his best to ignore your evidence because it goes against the 'study' he found that supports his hypothesis that clutch doesn't exist, or is so statistically irrelevant as to be meaningless. Clearly the study of one special individual, Yaz, violates his overarching hypothesis.

That study SPECIFICALLY mentioned Yaz.

 

[Fencer]

I'm sorry, but it's ridiculous to imply that all the players in the NFL perform at the same exact level in relation to each other when under extreme pressure. Which is what being clutch relates to. I know if I have to go up and kick a SB winning FG, I'd buckle, which is why AV gets paid millions and I get paid jack ****. You seem to totally misunderstand the concept of sports with the whole mathematical formula thing.

But you're not in the NFL!

We're only comparing different members of the very small group that has successfully earned NFL jobs.

 

[RayClay]

I think he's pointing out that the article, in a very simple way, blows theories like yours all to hell.

It uses a different metric of clutchness than you do -- namely, important situations in regular-season games. The benefit of that choice, obviously, is that every player on every team can be measured every season.

And it provides strong evidence against the theory that that kind of clutchness goes beyond random chance.

I agree that I have no idea how Yaz performs in seasons where the team is more than 20 games out.

Never said i did.

 

[RayClay]

If you're so curious, why don't you go to the link I gave earlier and type in a few numbers yourself?????????

Told you I don't know how to work the math in these things. Trouble is, as I think has been shown in this thread but I'm not sure, it's easy to come up with the wrong numbers or put them in the wrong place.

I'm no mathematician.

 

[RayClay]

Dude, YOU'RE the one who brought 7/10 up. Nobody else did.

As for the analogy between 7 trials of 8 at .326 probability vs. 35 trials out of 50 at .5 probability -- that's been explained multiple times already!!!!!!!!!!!!

No, emoney did. There was never any 7/10 situation.

 

[pats-blue]

That study SPECIFICALLY mentioned Yaz.

Yeah and that "study" also provided NO stats what so ever to back up that statement. Also the statement on Yaz in the "Study" mentioned that Yaz didn't come through in the 69 and 70 pennent race...when the SOX WERE NEVER IN ONE those years. Considering he was basing his paper on stats and formulas that Palmer came up with as he referenced in his footnotes....and accorrding to PALMERS stats for hose years Yaz was #1 in MLB in 1970 in games his offense contributed directly to wins and #10 in 1969. Now I still think most of the stats are still just that numbers trying to define a contest between humans not machines...yet the stats he used to "prove" his point that I found said exactly the opposite that NO ONE in MLB was better than Yaz in 1970...and he WAS NOT IN a pennent race so his statment is worse than BS it is an ill informed statment that calls into question all his work if he can make such a statement that is worng on so many levels.

 

[Fencer]

No, emoney did.

No, he did NOT!!!!!!!!!!!!

Do you know anything about probability at all? 7 successes out of 10 trials and 35 successes out of 50 trials are NOT the same thing.

 

[mb6592]

there is not formula that can perfectly predict the results of a football game that has yet to be played. is that your point?

however, you can make reasonable guesses. for various reasons, I think the Patriots win tomorrows game about 90% of the time. ie, they have a 90% chance to win the game.

Reasonable guesses...good grief. And WE DON'T NEED STATISTICS TO PREDICT THE PATRIOTS HAVE AT LEAST A 90% CHANCE OF WINNING TOMORROW. IN MOST CASES, STATS DON'T BRING ENOUGH TO THE TABLE.

 

[mb6592]

Reasonable guesses...good grief. And WE DON'T NEED STATISTICS TO PREDICT THE PATRIOTS HAVE AT LEAST A 90% CHANCE OF WINNING TOMORROW. IN MOST CASES, STATS DON'T BRING ENOUGH TO THE TABLE.

IN MOST CASES....holy crap.....you see what's happening

 

[RayClay]

No, he did NOT!!!!!!!!!!!!

Do you know anything about probability at all? 7 successes out of 10 trials and 35 successes out of 50 trials are NOT the same thing.

Not much. I have only the coincidence of a required course on research for a library course. We do zero math.

At least i know enough not to count the difference between 5th and 6th inning hits in seasons where the team is behind by 21, 22 games as being similar to the difference between postseason or season deciding games and regular season performance.

 

[PatriotsReign]

Nice post and I tend to agree. I certainly support the concept of using metrics to staff your team. BUT, sports is not the weather. When human beings compete, there is more than physics at work. These intangibles are what makes a player clutch. It would be easy for a team to call it a year down three games to the Yankees in the ALCS. It took something special for Schilling to get out to the mound with his bloody sock. And it isn't variance that raises TB's performance when the game is on the line with time running out.

I have no patience for sports intellectuals because they don't know their a** from their elbow most times. Football is the least intellectual of all the major sports and the coach has the most influence.

Football is also by far the most emotional of all the major sports. The reason why the Patriots have consistently beaten the odds is due to Belichick's ability to fuel the team's collective emotion and focus toward one goal...to win the next game.

IMO, Saber-geeks have no place in football discussions. You can give me all your statistics and I would still say I'd beat 9 out of 10 of them on a game to game prediction because the most important elements determining who will win can't be found in the statistics.

Saber-geeks can help in the draft and free-agency process, but as for good football discussion, this is not the place for them

 

[PatriotsReign]

No, he did NOT!!!!!!!!!!!!

Do you know anything about probability at all? 7 successes out of 10 trials and 35 successes out of 50 trials are NOT the same thing.

Keep your probability in the classroom. This board will never become a haven for stats men because football fans prefer NOT to hear what they have to say.

I'm an analyst by profession. My job is to predict sales based upon several causal factors and combine that with historical probabilty...but you know what?

NO ONE GIVES A CRAP!! That is why I never look at football that way. Real football fans are just like the players on the field. We are motivated & driven by EMOTION, not intellectual BS. Very few of us have the patience to listen to statistical probability and most people who profess the value of it here post with an arrogance that is not found in the average fan.

I think Stat-geeks really believe they know more about football than the average fan which probably makes up 95% of all our members. So most of us laugh when someone posts some lengthy statistical/probability/success/failure/ratio crap. They typically forget about the truly important factors that drive the game of football and probably think what works for baseball should apply in football. If they stay & debate long enough, they realize they don't possess the knowledge required for a good fact based football discussion.

Personally, I deal with statistics & probability everyday in my job. when I come here, it's the last thing I want to discuss or read. Another factor that makes it difficult to welcome the Saber-geek is that he will often post something negative or give a bleek prediction on our team. And the truth is, we just don't want to hear it. Fans want to always believe their team is going to win. So if we see some lengthy post on the probability/likelihood of the Patriots losing, we just want to beat that person silly!