Is there a direct correlation between Height and Rebounding?

[disoblige]

Charles Barkley , Rodman , Dwight , bill rusell , Ben Wallace, Bob petit, etc aren't the tallest players ever in the NBA.

Agree? The biggest factor in rebounding in the nba isn't height?

[Baddy Chuck]

Gerald Wallace is the of the best rebounders in the league at 6'7.

[Ripp]

Why not just make a plot of height versus rebounding percentage for the regular season? You'll almost certainly get something that looks linear

[MSGBallerz]

Why not just make a plot of height versus rebounding percentage for the regular season? You'll almost certainly get something that looks linear

This

[Lakeshow2417]

its the heart and effort if your willing to box out. Dwight WANTS to grab every ball off the boards. which is why he sa top rebounder

[contentode]

Why not just make a plot of height versus rebounding percentage for the regular season? You'll almost certainly get something that looks linear

Other factors might skew the linearity of the height vs rebounding % graph. Ability to block-out, arm length, anticipation (instincts), willingness to scrap and others come to mind.

[Mags FTW]

Bargnani wishes this was the case.

[mysticbb]

Other factors might skew the linearity of the height vs rebounding % graph. Ability to block-out, arm length, anticipation (instincts), willingness to scrap and others come to mind.

This is a part of a really important thing here. Nobody is playing basketball with his head. Standing reach and jumping abilities are a very valuable thing for rebounding. Also Bill Russell or Dwight Howard are significant taller than Rodman or Barkley.

[GJense4181]

standing reach* vs. rebounds per minute would be more accurate.

[Ripp]

Why not just make a plot of height versus rebounding percentage for the regular season? You'll almost certainly get something that looks linear

Other factors might skew the linearity of the height vs rebounding % graph. Ability to block-out, arm length, anticipation (instincts), willingness to scrap and others come to mind.

This is true, but this is also something you can actually test and see. if height alone explains most of the variation, then if you fit a line to the data (thus getting a slope and y-intercept), that line will probably explain most of the variance, indicating that height is the largest factor.

In any case, you can add as many other physical factors as you like. Perhaps a better thing to do would be to grab lots of data from draft express (height, standing reach, speed), and use them to setup a linear regression problem. That way you can get a sense of how important the different factors are.

Actually, now that I think about it, if you perform an experiment like this, you probably want to choose only one of height/standing reach/wingspan as a predictor variable, because they are extremely correlated too. So pick one of those variables, add speed, etc, and then regress against rebounding percentage.

[Ripp]

Height in inches vs. rebounds per minute would be more accurate.

You probably do not want to pick rebounds per minute as the dependent variable in this case...it is affected a lot by pace and how well the other team shoots. Rebounding percentage is a bit more legit, I think.

[GJense4181]

Height in inches vs. rebounds per minute would be more accurate.

You probably do not want to pick rebounds per minute as the dependent variable in this case...it is affected a lot by pace and how well the other team shoots. Rebounding percentage is a bit more legit, I think.

Where would I find such statistics?

[contentode]

Why not just make a plot of height versus rebounding percentage for the regular season? You'll almost certainly get something that looks linear

Other factors might skew the linearity of the height vs rebounding % graph. Ability to block-out, arm length, anticipation (instincts), willingness to scrap and others come to mind.

This is true, but this is also something you can actually test and see. if height alone explains most of the variation, then if you fit a line to the data (thus getting a slope and y-intercept), that line will probably explain most of the variance, indicating that height is the largest factor.

In any case, you can add as many other physical factors as you like. Perhaps a better thing to do would be to grab lots of data from draft express (height, standing reach, speed), and use them to setup a linear regression problem. That way you can get a sense of how important the different factors are.

Actually, now that I think about it, if you perform an experiment like this, you probably want to choose only one of height/standing reach/wingspan as a predictor variable, because they are extremely correlated too. So pick one of those variables, add speed, etc, and then regress against rebounding percentage.

I love statistics Ripp, but frankly, I'm afraid you're way ahead of me when it comes to applying and interpreting them.

[Spicy P]

Yes, I actually did a project on this back in high school for my stats class using stats of 120 starting players from the league and their per 48 minute stats. The correlation coefficient was around 0.80. There were some that were off like Jason Kidd and Eddy Curry, but it's a pretty linear correlation.

[the_warden]

Why not just make a plot of height versus rebounding percentage for the regular season? You'll almost certainly get something that looks linear

Other factors might skew the linearity of the height vs rebounding % graph. Ability to block-out, arm length, anticipation (instincts), willingness to scrap and others come to mind.

Why would that skew the linearity? The fact that other things influence rebounding doesn't mean that, if you were to plot rebounding against height, the result wouldn't be linear.

[Ripp]

I love statistics Ripp, but frankly, I'm afraid you're way ahead of me when it comes to applying and interpreting them.

Not really anything special, just what I study in school. But in this case, the topic isn't that hard...you just want to find a good way of fitting a line onto a plot of data.

Wikipedia has a decent explanation of how this is done in the single variable case (e.g., when we only want to use height to predict rebounding):
http://en.wikipedia.org/wiki/Simple_linear_regression

And here is a tutorial I googled for how to do this in Excel:
http://phoenix.phys.clemson.edu/tutoria ... ssion.html

Might be fun to just grab some data from basketball-reference.com and try it yourself, see how well it fits.

Finally, in the more general case (i.e., multiple predictor variables, for example using both height and vertical leap to predict rebounding), I googled and found a video which explains how to do this in Excel:

http://www.youtube.com/watch#!v=DTjWdmL ... re=related

The audio is a bit bad, but the guy does a decent job of explaining how to do it, I think.

Now that I think about it, maybe start with the video first.

Anyway, hope some of this is helpful.

[Ripp]

Height in inches vs. rebounds per minute would be more accurate.

You probably do not want to pick rebounds per minute as the dependent variable in this case...it is affected a lot by pace and how well the other team shoots. Rebounding percentage is a bit more legit, I think.

Where would I find such statistics?

That is also on basketball-reference.com. For example, for the Raptors, it is listed at the bottom of the following page, in the section called "Advanced":

http://www.basketball-reference.com/teams/TOR/2010.html

The three relevant columns are ORB% (Offensive Rebounding Percentage), DRB%, and TRB%.

If you want to download the data, you can click the "CSV" button. It then puts it into a nice format that you can copy and paste into a text file, and then import into Excel.

Unfortunately, basketball-reference.com doesn't have all the players listed on one page (at least, as far as I know), so you have to manually go through and download for each team. Also doesn't conveniently list height.

[ToffKobe]

Bargnani says hi and laughs at the fact that rebounding has a correlation to height. Well it does, but there's more to just the height.

[disoblige]

Would the co-eff still be excellent if you remove pg ,sg ,sf in the list.

[disoblige]

a post in realgm i read that pf are better rebounder than centers.