This is Google's cache of viewtopic.php?t=2741&sid=21e6f9bd1814d356189bf30fd8a5c03c. It is a snapshot of the page as it appeared on Apr 11, 2011 00:34:26 GMT. The current page could have changed in the meantime. Learn more

Text-only version
These search terms are highlighted: ilardi  
APBRmetrics :: View topic - Adjusted rebounding
APBRmetrics Forum Index APBRmetrics
The statistical revolution will not be televised.
 
 FAQFAQ   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   RegisterRegister 
 ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

Adjusted rebounding
Goto page 1, 2, 3, 4, 5  Next
 
Post new topic   Reply to topic    APBRmetrics Forum Index -> General discussion
View previous topic :: View next topic  
Author Message
back2newbelf



Joined: 21 Jun 2005
Posts: 276

PostPosted: Thu Feb 03, 2011 5:11 pm    Post subject: Adjusted rebounding Reply with quote

Using the same technique as in viewtopic.php?t=2669 I computed adjusted rebounding for the last 1 1/2 years.
Home teams grab more available offensive rebounds (30.4% vs 29.7%).
Lambdas were 1500 for offensive rebounding and 3000 for defensive rebounding.
Offensive and Defensive numbers are per 100 "rebound opportunities"

If someone wants the raw data in bbv matchup format I can upload the file.

http://www.stats-for-the-nba.appspot.com/rebounding
_________________
http://stats-for-the-nba.appspot.com/
Back to top
View user's profile Send private message
Crow



Joined: 20 Jan 2009
Posts: 825

PostPosted: Fri Feb 04, 2011 2:08 am    Post subject: Reply with quote

Thanks.

10 guys estimated at +2 better on total rebounds impact per 100 possessions at each end of the court. Less than 2%.

Only about 15% of the 538 players listed are estimated at +1 rebound or better on this and about 15% are -1 or worse. So it is estimated, by this regularized model, that about 70% of players are between +1 and -1. FWIW, it appears there were even fewer values beyond equivalent thresholds in Joe Sill's earlier estimates.

Food for thought about the relative importance of individual rebounding and maybe the degree of regularization.
Back to top
View user's profile Send private message
Statman



Joined: 20 Feb 2005
Posts: 242
Location: Arlington, Texas

PostPosted: Fri Feb 04, 2011 5:51 am    Post subject: Reply with quote

Crow wrote:
Thanks.

10 guys estimated at +2 better on total rebounds impact per 100 possessions at each end of the court. Less than 2%.

Only about 15% of the 538 players listed are estimated at +1 rebound or better on this and about 15% are -1 or worse. So it is estimated, by this regularized model, that about 70% of players are between +1 and -1. FWIW, it appears there were even fewer values beyond equivalent thresholds in Joe Sill's earlier estimates.

Food for thought about the relative importance of individual rebounding and maybe the degree of regularization.


Which is why Berri's stuff often looks so absurd to us (well, at least me) - valuing player rebounds at such a high rate.
_________________
Dan

My current national college player rankings (and other stuff):
http://www.pointguardu.com/f136/statmans-ratings-56243/index6.html#post355594
Back to top
View user's profile Send private message Send e-mail
back2newbelf



Joined: 21 Jun 2005
Posts: 276

PostPosted: Fri Feb 04, 2011 7:45 am    Post subject: Reply with quote

Crow wrote:
10 guys estimated at +2 better on total rebounds impact per 100 possessions at each end of the court.

I should have been clearer. It's not per possession. It's per "rebound opportunity". A "rebound opportunity" is every time a player of your team misses a shot that can be rebounded by your team (basically every shot except technical- and "first of two"-freethrows).
Compared to APM with points a missed shot (that can be rebounded) by Team A is equivalent to a possession by Team A. Team A getting the subsequent offensive rebound would be equivalent to Team A scoring one point.

I've also thought about doing this without all the available rebounds from free throws(when the last free throw was missed). I would imagine those kind of rebounds get very rarely rebounded by the offensive team and the guards of the offensive team have almost no influence on those. Thus, guards from teams that shoot (and miss) a lot of free throws get unfairly punished
_________________
http://stats-for-the-nba.appspot.com/
Back to top
View user's profile Send private message
EvanZ



Joined: 22 Nov 2010
Posts: 307

PostPosted: Fri Feb 04, 2011 8:29 am    Post subject: Reply with quote

back2newbelf wrote:

I've also thought about doing this without all the available rebounds from free throws(when the last free throw was missed). I would imagine those kind of rebounds get very rarely rebounded by the offensive team and the guards of the offensive team have almost no influence on those. Thus, guards from teams that shoot (and miss) a lot of free throws get unfairly punished


FWIW, I don't count give credit for dreb off free throws for similar reasons. I do give credit for oreb, though.
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
Mike G



Joined: 14 Jan 2005
Posts: 3623
Location: Hendersonville, NC

PostPosted: Fri Feb 04, 2011 10:24 am    Post subject: Re: Adjusted rebounding Reply with quote

back2newbelf wrote:
...
Home teams grab more available offensive rebounds (30.4% vs 29.7%).

Hmm, how is it that league avg ORb% is .263 ?
Quote:
I've also thought about doing this without all the available rebounds from free throws(when the last free throw was missed). I would imagine those kind of rebounds get very rarely rebounded by the offensive team

Even then, off missed FG the league ORb% is around .280.
What is it I'm missing?
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
Ilardi



Joined: 15 May 2008
Posts: 265
Location: Lawrence, KS

PostPosted: Fri Feb 04, 2011 10:38 am    Post subject: Reply with quote

back2newbelf -

Nice work (as usual). Have you taken a look to see how your regularized adjusted rebound rates compare with simple ORR% and DRR%? I would guess the correlations must be pretty high, but would love to see the actual r's if you or anyone else have time to check.
Back to top
View user's profile Send private message
back2newbelf



Joined: 21 Jun 2005
Posts: 276

PostPosted: Fri Feb 04, 2011 10:50 am    Post subject: Re: Adjusted rebounding Reply with quote

Mike G wrote:
Quote:
I've also thought about doing this without all the available rebounds from free throws(when the last free throw was missed). I would imagine those kind of rebounds get very rarely rebounded by the offensive team

Even then, off missed FG the league ORb% is around .280.
What is it I'm missing?

I didn't do this yet
Mike G wrote:

back2newbelf wrote:
...
Home teams grab more available offensive rebounds (30.4% vs 29.7%).
Hmm, how is it that league avg ORb% is .263 ?


Is it .263 of *all rebounds* or just of those that can actually be rebounded by the attacking team? I don't count it as rebounding opportunity if the first of two free throws is missed, but it's always listed as a team rebound for the defending team
_________________
http://stats-for-the-nba.appspot.com/
Back to top
View user's profile Send private message
mystic



Joined: 18 Dec 2006
Posts: 21

PostPosted: Fri Feb 04, 2011 11:05 am    Post subject: Reply with quote

Ilardi wrote:
back2newbelf -

Nice work (as usual). Have you taken a look to see how your regularized adjusted rebound rates compare with simple ORR% and DRR%? I would guess the correlations must be pretty high, but would love to see the actual r's if you or anyone else have time to check.


I just ran a linear regression with ORB% obtained by boxscores and those values for offensive rebounds:


Code:
      Model Summary
Model   R   R Square   Adjusted R Square   Std. Error of the Estimate
1   ,363a   ,132   ,130   ,7704233
a. Predictors: (Constant), ORB%




Same for DRB%


Code:
      Model Summary
Model   R   R Square   Adjusted R Square   Std. Error of the Estimate
1   ,318a   ,101   ,100   ,4245693
a. Predictors: (Constant), DRB%




And for the combination:

Code:
      Model Summary
Model   R   R Square   Adjusted R Square   Std. Error of the Estimate
1   ,366a   ,134   ,132   ,8530314
a. Predictors: (Constant), TRB%




Doesn't look like a high correlation to me. Which even gives more fuel to the "why Berri overrates rebounds" conversation.
_________________
http://bbmetrics.wordpress.com
Back to top
View user's profile Send private message Visit poster's website
Ilardi



Joined: 15 May 2008
Posts: 265
Location: Lawrence, KS

PostPosted: Fri Feb 04, 2011 11:42 am    Post subject: Reply with quote

Thanks, mystic. Wow, r's down in the .32-.36 range. I would have guessed much higher.

So, if we were to fully trust these adjusted rebounding numbers, they would imply that simple ORR% and DRR% figures are misleading as often than not, i.e., that a player's rebounding performance is so heavily contextual - dependent on teammates and opponents - that simple unadjusted rebound rate is nearly uninterpretable on its own.

The implications are not trivial . . .
Back to top
View user's profile Send private message
EvanZ



Joined: 22 Nov 2010
Posts: 307

PostPosted: Fri Feb 04, 2011 11:53 am    Post subject: Reply with quote

I agree!

(Mystic, nice to see you here.)
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
Ilardi



Joined: 15 May 2008
Posts: 265
Location: Lawrence, KS

PostPosted: Fri Feb 04, 2011 11:55 am    Post subject: Reply with quote

A few of the adjusted rebounding values are so surprising that I'm having trouble understanding how they could be valid. For example, Dwight Howard doesn't even place among the top 50, while he ranks #3 among NBA starters in overall rebounding percentage. Any thoughts?
Back to top
View user's profile Send private message
DSMok1



Joined: 05 Aug 2009
Posts: 611
Location: Where the wind comes sweeping down the plains

PostPosted: Fri Feb 04, 2011 11:57 am    Post subject: Reply with quote

Wow, those are low r-values!

I wonder if that's why, when I ran my ASPM regressions, rebounding was weighted so lightly?
_________________
GodismyJudgeOK.com/DStats
Twitter.com/DSMok1
Back to top
View user's profile Send private message Send e-mail Visit poster's website
EvanZ



Joined: 22 Nov 2010
Posts: 307

PostPosted: Fri Feb 04, 2011 12:04 pm    Post subject: Reply with quote

This is probably germaine.

When I take the REB100 from ezPM and correlate it with ORR (which is the offensive rebounding rate for a player relative to his counterpart), the R^2 is 0.43. When I do the same thing for DRR, the R^2 is 0.0001.

It's sort of blowing my mind. Either I'm giving not enough credit for defensive rebounds, or this has something to do with diminishing returns. I guess. Question
_________________
http://www.thecity2.com
http://www.ibb.gatech.edu/evan-zamir
Back to top
View user's profile Send private message
back2newbelf



Joined: 21 Jun 2005
Posts: 276

PostPosted: Fri Feb 04, 2011 12:06 pm    Post subject: Reply with quote

Everyone, please don't jump to conclusions. There might be an error in my code as there seem to be discrepancies with this and average OReb%, it might still be a not-so-good approximation or maybe the methodology isn't perfect for that kind of problem.

All things considered, we can only know if this is useful data if we compare it to standard OReb/DReb% (averaged over all on-court-players) as an predictor for out of sample data, which isn't hard to do but I don't have much time
_________________
http://stats-for-the-nba.appspot.com/
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic    APBRmetrics Forum Index -> General discussion All times are GMT - 5 Hours
Goto page 1, 2, 3, 4, 5  Next
Page 1 of 5

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


Powered by phpBB © 2001, 2005 phpBB Group