This is Google's cache of viewtopic.php?t=392. It is a snapshot of the page as it appeared on Feb 21, 2011 03:24:05 GMT. The current page could have changed in the meantime. Learn more

Text-only version
These search terms are highlighted: ed küpfer  
APBRmetrics :: View topic - Possession Estimators Revisited
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 

Possession Estimators Revisited

 
Post new topic   Reply to topic    APBRmetrics Forum Index -> General discussion
View previous topic :: View next topic  
Author Message
kjb



Joined: 03 Jan 2005
Posts: 860
Location: Washington, DC

PostPosted: Mon Oct 10, 2005 8:51 am    Post subject: Reply with quote

Ed Küpfer wrote:
I've set up a new four-factors based set of historical stats pages. Nothing too comprensive, but hopefully somethings that you can't find anywhere else. I'm kinda proud of the HTML-based graphs and sparklines.


I like it. Might want to double-check the formulas, though. You have the Wizards with a pts per 100 possession last season of 112.4 scored, and 112.8 allowed -- they were in the 107 range for both last season.
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
Ed Küpfer



Joined: 30 Dec 2004
Posts: 783
Location: Toronto

PostPosted: Mon Oct 10, 2005 10:17 am    Post subject: Reply with quote

WizardsKev wrote:
Might want to double-check the formulas, though. You have the Wizards with a pts per 100 possession last season of 112.4 scored, and 112.8 allowed -- they were in the 107 range for both last season.


The discussion from this thread suggests that we are overestimating (edit, wrote underestimating before) the number of possessions by using the traditional FGA + .44*FTA - OR + TO formula. I used FGM + .4*FTM + OppDR + TO, which gave me a good fit to Roland's pbp possession counts. I don't think it's very important in the context of the website—what matters is the relative differences between offense and defense.
_________________
ed


Last edited by Ed Küpfer on Mon Oct 10, 2005 10:57 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
kjb



Joined: 03 Jan 2005
Posts: 860
Location: Washington, DC

PostPosted: Mon Oct 10, 2005 10:27 am    Post subject: Reply with quote

Ed Küpfer wrote:
WizardsKev wrote:
Might want to double-check the formulas, though. You have the Wizards with a pts per 100 possession last season of 112.4 scored, and 112.8 allowed -- they were in the 107 range for both last season.


The discussion from this thread suggests that we are underestimating the number of possessions by using the traditional FGA + .44*FTA - OR + TO formula. I used FGM + .4*FTM + OppDR + TO, which gave me a good fit to Roland's pbp possession counts. I don't think it's very important in the context of the website—what matters is the relative differences between offense and defense.


I'd thought Justin's numbers were updated to the newer possessions estimator. Justin?
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
jkubatko



Joined: 05 Jan 2005
Posts: 702
Location: Columbus, OH

PostPosted: Mon Oct 10, 2005 12:12 pm    Post subject: Reply with quote

WizardsKev wrote:
I'd thought Justin's numbers were updated to the newer possessions estimator. Justin?


I'm using Dean Oliver's estimator:

Code:
FGA + 0.4*FTA - 1.07*(ORB / (ORB + Opp DRB))*(FGA - FG) + TO


Ed Küpfer wrote:
I used FGM + .4*FTM + OppDR + TO, which gave me a good fit to Roland's pbp possession counts.


For the three years of data I have from 82games, the rmse of the formula above is 2.48 possessions per game. The rmse of Oliver's formula using the same data is 1.49 possessions per game. Ed, if you want to use a formula like the one above, you should probably increase the free throw multiplier to something like 0.52.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Ed Küpfer



Joined: 30 Dec 2004
Posts: 783
Location: Toronto

PostPosted: Mon Oct 10, 2005 4:34 pm    Post subject: Reply with quote

jkubatko wrote:
Ed, if you want to use a formula like the one above, you should probably increase the free throw multiplier to something like 0.52.


I can't in good concience use a coefficient greater than 0.5. It would violate my code of ethics, as well as the Universal Declaration of Free Throw Rights. I will, however, multiply my possession estimates by 0.96, which should bring everything back to the neighbourhood of orthodox estimates.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
jkubatko



Joined: 05 Jan 2005
Posts: 702
Location: Columbus, OH

PostPosted: Mon Oct 10, 2005 4:49 pm    Post subject: Reply with quote

Ed Küpfer wrote:
I can't in good concience use a coefficient greater than 0.5. It would violate my code of ethics, as well as the Universal Declaration of Free Throw Rights. I will, however, multiply my possession estimates by 0.96, which should bring everything back to the neighbourhood of orthodox estimates.


It also changes all of your coefficients, including the free throw coefficient. :-)
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
kjb



Joined: 03 Jan 2005
Posts: 860
Location: Washington, DC

PostPosted: Tue Oct 11, 2005 4:50 am    Post subject: Reply with quote

The formulation I liked from the previous thread -- simple and easy to calculate -- was .96 * FGA + .44 * FTA - ORB + TOV.
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
jkubatko



Joined: 05 Jan 2005
Posts: 702
Location: Columbus, OH

PostPosted: Tue Oct 11, 2005 7:55 am    Post subject: Reply with quote

WizardsKev wrote:
The formulation I liked from the previous thread -- simple and easy to calculate -- was .96 * FGA + .44 * FTA - ORB + TOV.


Be careful. That should be:

Code:
0.96*(FGA + 0.44*FTA - ORB + TOV)


I'm sure that's what you meant, but the way you typed it it looks like the 0.96 multiplier only goes with FGA.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
kjb



Joined: 03 Jan 2005
Posts: 860
Location: Washington, DC

PostPosted: Tue Oct 11, 2005 8:19 am    Post subject: Reply with quote

jkubatko wrote:
WizardsKev wrote:
The formulation I liked from the previous thread -- simple and easy to calculate -- was .96 * FGA + .44 * FTA - ORB + TOV.


Be careful. That should be:

Code:
0.96*(FGA + 0.44*FTA - ORB + TOV)


I'm sure that's what you meant, but the way you typed it it looks like the 0.96 multiplier only goes with FGA.


I was being forced to pay retail for the parentheses. Rolling Eyes
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
Analyze This



Joined: 17 May 2005
Posts: 364

PostPosted: Wed Oct 12, 2005 8:06 am    Post subject: Reply with quote

Ed Küpfer wrote:
WizardsKev wrote:
Might want to double-check the formulas, though. You have the Wizards with a pts per 100 possession last season of 112.4 scored, and 112.8 allowed -- they were in the 107 range for both last season.


The discussion from this thread suggests that we are overestimating (edit, wrote underestimating before) the number of possessions by using the traditional FGA + .44*FTA - OR + TO formula. I used FGM + .4*FTM + OppDR + TO, which gave me a good fit to Roland's pbp possession counts. I don't think it's very important in the context of the website—what matters is the relative differences between offense and defense.


Very nice. Ik hope you will update every season. Smile
_________________
Where There's a WilT There's a Way
Back to top
View user's profile Send private message
Ed Küpfer



Joined: 30 Dec 2004
Posts: 783
Location: Toronto

PostPosted: Wed Oct 12, 2005 11:01 am    Post subject: Reply with quote

The Big Dipper wrote:
Very nice. Ik hope you will update every season. Smile


Actually, I'm kinda hoping someone will steal the idea and do it for me, updating during the season. I did this a few years ago, but I changed ISPs and my new one won't let me publish directly from Excel. I'm not very good with html and web stuff, so I won't be able to figure out a good way of doing it, but maybe someone else wants to do the work. Anyone interested can email or PM me and get technical advice on the stats themselves.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Display posts from previous:   
Post new topic   Reply to topic    APBRmetrics Forum Index -> General discussion All times are GMT - 5 Hours
Page 1 of 1

 
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