Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

NUPE wrote:

Bonn1997 wrote:Once you realize the importance of scoring efficiency (and it sounds like you now generally speaking do), when you watch the game you see it entirely differently. You're no longer distracted by flashy, high volume, inefficient offense.

Melo has sufficiently efficient this year. I'd imagine his efficiency will increase slightly when Amar'e returns and he is able to get one or two easy baskets a game as a result, possibly. You must be speaking about yester-year which is irrelevant.

I'm talking about his whole career - it is poor math to look at any small sample of games (including the less than 20 he's played this year).

His sample size under Woody is becoming larger and larger. Watch out!

It's about 50 games. Maybe he's improving maybe not. Players go full seasons with #s that deviate from their averages and then usually return to their career averages. If he goes 100 to 150 games with higher efficiency, I'll become a believer.

Haha.

WS not working? WP it is. WP is not good for the argument at hand it becomes WZ.

15 games too little, how about 25? 30? 50? No good? 100 it is. 100 games and still good? Really my number was 150 - .... hhahaha

Please answer the assist question. And tell me how it relates to his +/- selfishness between the last 2 seasons.

and Kobe proved that a dummy offensive player who takes awful shots can win (if he plays defense)

Yes, he's clearly more selfish than last year on offense. People's interpretations of his performance are just being colored by the team's success.

WRONG AGAIN!

He is clearly passing more, they are just not leading to direct assists.

Oh yeah, he's an MVP level hockey assister. I forgot.

I would love to see a hockey assist stat but I won't claim that even though I have examined the validation data which was quite impressive.

He passes out of the double team far more. Do they (those passes) lead to points? You tell me, doc.

If Melo has improved, my guess is it's just by a couple of tenths of a hockey assist per game.

Is that just a wild guess or is there any shred of evidence behind this assertion?

I GUESS that it improved 769 pct

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:Once you realize the importance of scoring efficiency (and it sounds like you now generally speaking do), when you watch the game you see it entirely differently. You're no longer distracted by flashy, high volume, inefficient offense.

I was always a stat geek. Just not for things that I don't understand like Win Shares.

When people say that Andre Miller was their type of player and he has a WS of .122 for his career, I question it.

When WS fails, we can fall to WP's ... I know.

I have read many dissenting args to WS's and cannot buy into it. You don't get it either - you were just told that its a viable stat. I don't like being told what to believe.

I have been fair in my assessment of what Carmelo is. I also hold out hope that he can be better and believe that he has improved this year. Explain his horrible assists numbers compared to 3.6 last year and tell me which year he was more selfish. PLEASE

HOPE is a good thing. The religion of science can be dangerous.

I haven't closely examined the formula but I do understand the multiple regression analyses used to validate the statistic.

which basically means that you don't have a clue what the phuck it means and can't admit it.

I just did admit that I haven't examined the formula closely. Are you blind?
I've examined the validation data though. The validation data are impressive and I haven't felt like spending the time to examine the details of the formula. I'm more like a medical doctor who reads the abstracts of medical journals than a researcher who reads each issue of each journal cover to cover.

I am the opposite. I tend to put in work and understand something before I mention it again and again as evidence.

Then again. I am an attorney who can't afford to "F" up with my research. You are a dabbling doctor :-)

It depends on what you mean by work. I've put in work to evaluating the validation data.
WS and WP predict a very high percentage of the variance in outcomes (75 to 90% depending on the study), which is all the evidence you need in order to know its importance. The actual formula for these formulas is explained over several hundreds of pages in James' and Berri's work. it's not like it's a 2 page document. If this were my work rather than a hobby, I'd carefully examine it rather than skimming it.

75 - 90 pct?

Ouch try to win a court case with those percentages.

I am sure there is value in them though, just think that it is unproven and often argued. If its right most of the time, its not enough since there is too much room for error.

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:Once you realize the importance of scoring efficiency (and it sounds like you now generally speaking do), when you watch the game you see it entirely differently. You're no longer distracted by flashy, high volume, inefficient offense.

I was always a stat geek. Just not for things that I don't understand like Win Shares.

When people say that Andre Miller was their type of player and he has a WS of .122 for his career, I question it.

When WS fails, we can fall to WP's ... I know.

I have read many dissenting args to WS's and cannot buy into it. You don't get it either - you were just told that its a viable stat. I don't like being told what to believe.

I have been fair in my assessment of what Carmelo is. I also hold out hope that he can be better and believe that he has improved this year. Explain his horrible assists numbers compared to 3.6 last year and tell me which year he was more selfish. PLEASE

HOPE is a good thing. The religion of science can be dangerous.

I haven't closely examined the formula but I do understand the multiple regression analyses used to validate the statistic.

which basically means that you don't have a clue what the phuck it means and can't admit it.

I just did admit that I haven't examined the formula closely. Are you blind?
I've examined the validation data though. The validation data are impressive and I haven't felt like spending the time to examine the details of the formula. I'm more like a medical doctor who reads the abstracts of medical journals than a researcher who reads each issue of each journal cover to cover.

I am the opposite. I tend to put in work and understand something before I mention it again and again as evidence.

Then again. I am an attorney who can't afford to "F" up with my research. You are a dabbling doctor :-)

It depends on what you mean by work. I've put in work to evaluating the validation data.
WS and WP predict a very high percentage of the variance in outcomes (75 to 90% depending on the study), which is all the evidence you need in order to know its importance. The actual formula for these formulas is explained over several hundreds of pages in James' and Berri's work. it's not like it's a 2 page document. If this were my work rather than a hobby, I'd carefully examine it rather than skimming it.

75 - 90 pct?

Ouch try to win a court case with those percentages.

I am sure there is value in them though, just think that it is unproven and often argued. If its right most of the time, its not enough since there is too much room for error.

Actually that's extraordinarily high in the behavioral sciences.
For example, # of cigarettes consumed per day doesn't explain nearly as much variance in lung cancer rates

Here you are trying to assert this science to ONE player. A science with a high a margin of error. Bad analogy.

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:Once you realize the importance of scoring efficiency (and it sounds like you now generally speaking do), when you watch the game you see it entirely differently. You're no longer distracted by flashy, high volume, inefficient offense.

I was always a stat geek. Just not for things that I don't understand like Win Shares.

When people say that Andre Miller was their type of player and he has a WS of .122 for his career, I question it.

When WS fails, we can fall to WP's ... I know.

I have read many dissenting args to WS's and cannot buy into it. You don't get it either - you were just told that its a viable stat. I don't like being told what to believe.

I have been fair in my assessment of what Carmelo is. I also hold out hope that he can be better and believe that he has improved this year. Explain his horrible assists numbers compared to 3.6 last year and tell me which year he was more selfish. PLEASE

HOPE is a good thing. The religion of science can be dangerous.

I haven't closely examined the formula but I do understand the multiple regression analyses used to validate the statistic.

which basically means that you don't have a clue what the phuck it means and can't admit it.

I just did admit that I haven't examined the formula closely. Are you blind?
I've examined the validation data though. The validation data are impressive and I haven't felt like spending the time to examine the details of the formula. I'm more like a medical doctor who reads the abstracts of medical journals than a researcher who reads each issue of each journal cover to cover.

I am the opposite. I tend to put in work and understand something before I mention it again and again as evidence.

Then again. I am an attorney who can't afford to "F" up with my research. You are a dabbling doctor :-)

It depends on what you mean by work. I've put in work to evaluating the validation data.
WS and WP predict a very high percentage of the variance in outcomes (75 to 90% depending on the study), which is all the evidence you need in order to know its importance. The actual formula for these formulas is explained over several hundreds of pages in James' and Berri's work. it's not like it's a 2 page document. If this were my work rather than a hobby, I'd carefully examine it rather than skimming it.

75 - 90 pct?

Ouch try to win a court case with those percentages.

I am sure there is value in them though, just think that it is unproven and often argued. If its right most of the time, its not enough since there is too much room for error.

Actually that's extraordinarily high in the behavioral sciences.
For example, # of cigarettes consumed per day doesn't explain nearly as much variance in lung cancer rates

Here you are trying to assert this science to ONE player. A science with a high a margin of error. Bad analogy.

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:Once you realize the importance of scoring efficiency (and it sounds like you now generally speaking do), when you watch the game you see it entirely differently. You're no longer distracted by flashy, high volume, inefficient offense.

I was always a stat geek. Just not for things that I don't understand like Win Shares.

When people say that Andre Miller was their type of player and he has a WS of .122 for his career, I question it.

When WS fails, we can fall to WP's ... I know.

I have read many dissenting args to WS's and cannot buy into it. You don't get it either - you were just told that its a viable stat. I don't like being told what to believe.

I have been fair in my assessment of what Carmelo is. I also hold out hope that he can be better and believe that he has improved this year. Explain his horrible assists numbers compared to 3.6 last year and tell me which year he was more selfish. PLEASE

HOPE is a good thing. The religion of science can be dangerous.

I haven't closely examined the formula but I do understand the multiple regression analyses used to validate the statistic.

which basically means that you don't have a clue what the phuck it means and can't admit it.

I just did admit that I haven't examined the formula closely. Are you blind?
I've examined the validation data though. The validation data are impressive and I haven't felt like spending the time to examine the details of the formula. I'm more like a medical doctor who reads the abstracts of medical journals than a researcher who reads each issue of each journal cover to cover.

I am the opposite. I tend to put in work and understand something before I mention it again and again as evidence.

Then again. I am an attorney who can't afford to "F" up with my research. You are a dabbling doctor :-)

It depends on what you mean by work. I've put in work to evaluating the validation data.
WS and WP predict a very high percentage of the variance in outcomes (75 to 90% depending on the study), which is all the evidence you need in order to know its importance. The actual formula for these formulas is explained over several hundreds of pages in James' and Berri's work. it's not like it's a 2 page document. If this were my work rather than a hobby, I'd carefully examine it rather than skimming it.

75 - 90 pct?

Ouch try to win a court case with those percentages.

I am sure there is value in them though, just think that it is unproven and often argued. If its right most of the time, its not enough since there is too much room for error.

Actually that's extraordinarily high in the behavioral sciences.
For example, # of cigarettes consumed per day doesn't explain nearly as much variance in lung cancer rates

Here you are trying to assert this science to ONE player. A science with a high a margin of error. Bad analogy.

There is no better alternative though.

Agreed.

Thats why you still need good scouting and the all important eyeball. Mix that with a touch of advance stats that is proven and mainstream and you will make a decent GM.

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:Once you realize the importance of scoring efficiency (and it sounds like you now generally speaking do), when you watch the game you see it entirely differently. You're no longer distracted by flashy, high volume, inefficient offense.

I was always a stat geek. Just not for things that I don't understand like Win Shares.

When people say that Andre Miller was their type of player and he has a WS of .122 for his career, I question it.

When WS fails, we can fall to WP's ... I know.

I have read many dissenting args to WS's and cannot buy into it. You don't get it either - you were just told that its a viable stat. I don't like being told what to believe.

I have been fair in my assessment of what Carmelo is. I also hold out hope that he can be better and believe that he has improved this year. Explain his horrible assists numbers compared to 3.6 last year and tell me which year he was more selfish. PLEASE

HOPE is a good thing. The religion of science can be dangerous.

I haven't closely examined the formula but I do understand the multiple regression analyses used to validate the statistic.

which basically means that you don't have a clue what the phuck it means and can't admit it.

I just did admit that I haven't examined the formula closely. Are you blind?
I've examined the validation data though. The validation data are impressive and I haven't felt like spending the time to examine the details of the formula. I'm more like a medical doctor who reads the abstracts of medical journals than a researcher who reads each issue of each journal cover to cover.

I am the opposite. I tend to put in work and understand something before I mention it again and again as evidence.

Then again. I am an attorney who can't afford to "F" up with my research. You are a dabbling doctor :-)

It depends on what you mean by work. I've put in work to evaluating the validation data.
WS and WP predict a very high percentage of the variance in outcomes (75 to 90% depending on the study), which is all the evidence you need in order to know its importance. The actual formula for these formulas is explained over several hundreds of pages in James' and Berri's work. it's not like it's a 2 page document. If this were my work rather than a hobby, I'd carefully examine it rather than skimming it.

75 - 90 pct?

Ouch try to win a court case with those percentages.

I am sure there is value in them though, just think that it is unproven and often argued. If its right most of the time, its not enough since there is too much room for error.

Actually that's extraordinarily high in the behavioral sciences.
For example, # of cigarettes consumed per day doesn't explain nearly as much variance in lung cancer rates

Here you are trying to assert this science to ONE player. A science with a high a margin of error. Bad analogy.

There is no better alternative though.

Agreed.

Thats why you still need good scouting and the all important eyeball. Mix that with a touch of advance stats that is proven and mainstream and you will make a decent GM.

Bonn1997 wrote:Whenever we look at any research findings and try to apply them to ourselves, we're "asserting the science to one person." When you ask questions like should I eat healthy, exercise, drink and drive etc., you're expressing a belief that applying the science to one person is smarter than ignoring that science.

True. But the possibly of being wrong 25 percent of the time with your analysis? Thats good for a poker game where in the long run you hope to be right 50-75 pct of the time/hands. Or remaining on the smarter side of caution.

Just realize that you can he can be the ONE of 4 where Mr. Berri is wrong.

I need to call it a night. It was certainly enlightening.

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:Whenever we look at any research findings and try to apply them to ourselves, we're "asserting the science to one person." When you ask questions like should I eat healthy, exercise, drink and drive etc., you're expressing a belief that applying the science to one person is smarter than ignoring that science.

True. But the possibly of being wrong 25 percent of the time with your analysis? Thats good for a poker game where in the long run you hope to be right 50-75 pct of the time/hands. Or remaining on the smarter side of caution.

Just realize that you can he can be the ONE of 4 where Mr. Berri is wrong.

The 75 to 90% doesn't mean that his interpretation of specific players will be completely wrong 1 in 4 times. It means that his predictions for team wins will be slightly off. If you add up the WP of the individual players and calculate the team's expected win total, he likely will not be exactly right but will be within 2 to 3 wins of the team's actual performance.
That said, I don't think the advanced stats are perfect. They probably *slightly* devalue shot-creation and slightly devalue Melo's contribution. Most fans dramatically overvalue shot-creation though.

I don't overvalue these shots. I hate them - and hated/hate when kobe takes/took them. With all those horrible/annoying shots, he was still a damn good player.

The question is, does a given player bring enough to the table to overcome these (awful) shots.