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 

Estimating Statistics

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



Joined: 18 Feb 2005
Posts: 10
Location: Connecticut

PostPosted: Mon Feb 21, 2005 9:57 pm    Post subject: Estimating Statistics Reply with quote

I was wondering if anyone had methods to estimating certain statistics (such as turnovers and offensive rebounds) for those years before such stats were kept by the NBA?

Thanks,

Dave
Back to top
View user's profile Send private message
Mike G



Joined: 14 Jan 2005
Posts: 1492
Location: Delphi, Indiana

PostPosted: Mon Feb 21, 2005 10:48 pm    Post subject: Reply with quote

Here's my estimated turnover formula:

eTO = .08*Pts + .07*Reb + .16*Ast + .05*Stl + .10*Blk - .005*Min

I don't have real theories behind the correlations; I just tinkered with the factors until listing guys' TO/eTO ratio, I seemed to have all sorts of players (PG,PF, etc) in all parts of the list.

Of course, this formula doesn't help you before 1974; in that era, you further have to estimate steals and blocks. Since these are so variable, I haven't concentrated on anything very precise. But here they are:

eStl = sqrt(Ast) - .4

eBlk = sqrt(Reb) - 1.6

Some guys will come up with negative numbers, so do whatever you think. (SQRT is Excellese for square root.)

In 1974, players were turning it over much more than today. There's a theory that a faster game pace might naturally produce more TO. But I don't know. You still have to take care of the ball for some 24 minutes. Shooting quickly is one way to avoid turnovers.
Back to top
View user's profile Send private message Send e-mail
kjb



Joined: 03 Jan 2005
Posts: 660
Location: Washington, DC

PostPosted: Tue Feb 22, 2005 9:56 am    Post subject: Reply with quote

Check appendix 4 in "Basketball On Paper" for Dean Oliver's method.
Back to top
View user's profile Send private message Visit poster's website AIM Address Yahoo Messenger
Mike G



Joined: 14 Jan 2005
Posts: 1492
Location: Delphi, Indiana

PostPosted: Tue Feb 22, 2005 12:10 pm    Post subject: Reply with quote

This topic reminds me I haven't checked TO/eTO lately. It turns out, of 350 players with at least 215 minutes, there are 15 guys who turn it over at least 50% more than they should (according to the aforementioned formula):

TO/e player pos tm TO
1.97 Araujo,Rafael C Tor 2.4
1.82 Telfair,Sebastia PG Por 3.1
1.76 Livingston,Shaun PG LAC 3.1
1.70 Martin,Kevin SG Sac 1.8
1.69 Norris,Moochie PG NY 2.1
1.66 Rogers,Rodney SF NO 2.7
1.63 Diaw,Boris SF Atl 2.3
1.63 Barbosa,Leandro PG Phe 2.6
1.62 Williams,Aaron PF Tor 1.6
1.59 Brown,Kwame PF Was 2.4
1.56 Gill,Eddie PG Ind 1.9
1.56 Perkins,Kendrick PF Bos 2.4
1.55 Ivey,Royal SG Atl 2.2
1.50 Murray,Ronald SG Sea 2.3
1.50 Collier,Jason C Atl 2.1


Several rookies head the list; only 2 play over 20 mpg (Livingston, 22; and Rogers, 30). The number after the team name is normalized TO/36.

Now here are the best at taking care of the ball, relative to their production:

.28 Hoiberg,Fred SG Min .5
.35 Thomas,Billy SG NJ .4
.37 Varejao,Anderson SF Cle .7
.43 Marshall,Donyell PF Tor .9
.43 Laettner,Christi PF Mia .7
.47 Cook,Brian PF LAL .8
.48 Lafrentz,Raef C Bos 1.1
.50 Bowen,Bruce SF SA .7
.50 Daniels,Antonio PG Sea 1.2
.51 Doleac,Michael C Mia .7
.51 Battier,Shane SF Mem .9
.52 Bonner,Matt PF Tor .8
.54 Griffin,Eddie PF Min 1.1
.54 Marion,Shawn PF Phe 1.4
.54 Booth,Calvin C Dal .7
.55 Jones,Jumaine SF LAL .8
.55 Finley,Michael SG Dal 1.1
.55 Padgett,Scott PF Hou .9
.56 Miller,Brad C Sac 1.5
.57 Duncan,Tim PF SA 2.1

Until Duncan, all are very-low TO players, per-minute. And only 1 player listed as PG (Daniels, who is really just a "guard"). Suggesting that the high-risk Assist category may tend to have built-in turnovers. System turnovers?
Back to top
View user's profile Send private message Send e-mail
HoopStudies



Joined: 30 Dec 2004
Posts: 526
Location: Near Philadelphia, PA

PostPosted: Tue Feb 22, 2005 12:10 pm    Post subject: Reply with quote

WizardsKev wrote:
Check appendix 4 in "Basketball On Paper" for Dean Oliver's method.


Note that this is for teams. I'd probably do teams first then estimate individuals so that the total is roughly equivalent to teams. But I haven't spent much time on this.
_________________
Dean Oliver
Author, Basketball on Paper
http://www.basketballonpaper.com
Back to top
View user's profile Send private message Visit poster's website
replayhoops



Joined: 18 Feb 2005
Posts: 10
Location: Connecticut

PostPosted: Tue Feb 22, 2005 12:24 pm    Post subject: Reply with quote

Thanks for the replies, guys. I just bought Dean's book and will take a look!

Dave
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
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