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 

Hollinger's latest
Goto page 1, 2, 3, 4  Next
 
Post new topic   Reply to topic    APBRmetrics Forum Index -> General discussion
View previous topic :: View next topic  
Author Message
jkubatko



Joined: 05 Jan 2005
Posts: 505
Location: Columbus, OH

PostPosted: Tue Jun 26, 2007 3:49 pm    Post subject: Hollinger's latest Reply with quote

John Hollinger has a new system for evaluating the pro potential of a player using his college statistics.
_________________
Regards,
Justin Kubatko
Basketball Stats!
Back to top
View user's profile Send private message Send e-mail Visit poster's website
TexasEx



Joined: 12 May 2006
Posts: 17
Location: Houston, TX

PostPosted: Tue Jun 26, 2007 4:32 pm    Post subject: Re: Hollinger's latest Reply with quote

Interesting article.
Back to top
View user's profile Send private message
Statman



Joined: 20 Feb 2005
Posts: 79

PostPosted: Tue Jun 26, 2007 4:53 pm    Post subject: Reply with quote

Combine this and the Hoops Analyst articles - one could get a good idea who the busts (and suprises) may be statistically.

The only problem I have the analysis is that it doesn't seem to adjust for team pace in many circumstances, and no adjustment for SoS. Both articles like Fazekas - but come on now - his SoS isn't near that of many of the others. Same with Stuckey.

But - the stuff is still quite cool - and I think informative.
_________________
www.goodstats.net
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Don Babbitt



Joined: 01 Feb 2007
Posts: 9
Location: California

PostPosted: Tue Jun 26, 2007 5:34 pm    Post subject: Reply with quote

The Football Outsiders have a system for ranking NCAA quarterbacks that are anticipated to be picked in the first two rounds.

It's interesting to me watch the development in stat. analysis for the NBA and the NFL.
Back to top
View user's profile Send private message
jeffpotts77



Joined: 18 Feb 2005
Posts: 142
Location: Cambridge, MA

PostPosted: Tue Jun 26, 2007 9:04 pm    Post subject: Reply with quote

His results seem pretty good to me. He seems to have come to a lot of the same conclusions as Weiland about which stats are most relevant when evaluating college prospects. Excellent article.
Back to top
View user's profile Send private message Yahoo Messenger
tmansback



Joined: 12 Aug 2005
Posts: 120

PostPosted: Tue Jun 26, 2007 9:14 pm    Post subject: Reply with quote

Best one I've seen. Really only good when taking other things into account. Mainly do they have the physical tools to get it to translate to the pros. Shelden Williams had no shot at being a great shotblocker and post player in the pros.

Hawes numbers are a concern. His illness during the year might be effecting his numbers though.

The best players in this draft are the freshman. Whoever get Young and Wright are going to be the steals of the draft I think.
Back to top
View user's profile Send private message
Mark



Joined: 20 Aug 2005
Posts: 807

PostPosted: Wed Jun 27, 2007 1:15 am    Post subject: Reply with quote

John, your system does well particularly in finding the top guy but my initial rough impression is that elevates properly in roughly the same proportion of cases where there is a significant difference between rank and actual draft # as it elevates guys who havent validated that- at least yet. Are you willing to discuss the weights here? I agree with age being a variable of some importance but wondered if you gave it a linear function or a more variable one as you moved from the age center and if it was the same for all positions or different to reflect league experience/interpretation of age/success by position?

I may be jumping the gun for a planned post here but I noticed a post on this topic and plenty more draft stuff at http://hoopinion.blogspot.com/.

Also wages of wins has a relevant draft piece with a useful split of performance ranking by tournament opponent / not.
Back to top
View user's profile Send private message
gabefarkas



Joined: 31 Dec 2004
Posts: 879
Location: Durham, NC

PostPosted: Wed Jun 27, 2007 5:38 am    Post subject: Reply with quote

Hindsight is 20/20. I'm curious how much hunting and pecking there was. In other words, how much of it was a case of John knowing roughly what he wanted the results to look like, and then adding/removing factors and weights until it came out nicely?
Back to top
View user's profile Send private message Send e-mail AIM Address
Chronz1



Joined: 22 May 2006
Posts: 108

PostPosted: Wed Jun 27, 2007 7:51 am    Post subject: Reply with quote

gabefarkas wrote:
Hindsight is 20/20. I'm curious how much hunting and pecking there was. In other words, how much of it was a case of John knowing roughly what he wanted the results to look like, and then adding/removing factors and weights until it came out nicely?


Is that really such a bad thing? You cant succeed in the future unless you learn from the mistakes of the past.

Im curious to see if any player whos rated poorly ended up having a great career.
Back to top
View user's profile Send private message
jeffpotts77



Joined: 18 Feb 2005
Posts: 142
Location: Cambridge, MA

PostPosted: Wed Jun 27, 2007 8:22 am    Post subject: Reply with quote

Chronz1 wrote:
gabefarkas wrote:
Hindsight is 20/20. I'm curious how much hunting and pecking there was. In other words, how much of it was a case of John knowing roughly what he wanted the results to look like, and then adding/removing factors and weights until it came out nicely?


Is that really such a bad thing? You cant succeed in the future unless you learn from the mistakes of the past.

Im curious to see if any player whos rated poorly ended up having a great career.


I gotta agree with Chronz1 here. Adjusting the formula until the results look reasonable is exactly the method that should be used in this case.
Back to top
View user's profile Send private message Yahoo Messenger
Harold Almonte



Joined: 04 Aug 2006
Posts: 224

PostPosted: Wed Jun 27, 2007 9:22 am    Post subject: Reply with quote

Wouldn't be that the same as a regression?
Back to top
View user's profile Send private message
gabefarkas



Joined: 31 Dec 2004
Posts: 879
Location: Durham, NC

PostPosted: Wed Jun 27, 2007 9:59 am    Post subject: Reply with quote

Chronz1 wrote:
gabefarkas wrote:
Hindsight is 20/20. I'm curious how much hunting and pecking there was. In other words, how much of it was a case of John knowing roughly what he wanted the results to look like, and then adding/removing factors and weights until it came out nicely?


Is that really such a bad thing? You cant succeed in the future unless you learn from the mistakes of the past.


Well, he doesn't explain how or why he chose the factors that he did. Was it a regression analysis? Was there any formality in the methods? Or did he make a list apriori and then try to see which factors need to be included to match his list?

Also, I would suspect that the evolution of the game would make his results less valid the farther back we go. With that in mind, perhaps the factors that were important in 2006 may not be nearly as important in 2011 (when these guys are peaking).
Back to top
View user's profile Send private message Send e-mail AIM Address
davis21wylie2121



Joined: 13 Oct 2005
Posts: 359
Location: Atlanta, GA

PostPosted: Wed Jun 27, 2007 10:10 am    Post subject: Reply with quote

Harold Almonte wrote:
Wouldn't be that the same as a regression?


That's what I assumed he did -- just take all the pace-adjusted (and competition-adjusted) college stats from drafted players over the past however many years, throw in a few dummy variables for 7-footers and the like, and then regress them on whatever stat you think indicates "NBA Success" (since this is Hollinger, I'm guessing PER was used here). I'm not sure whether more points indicates a better career value, or peak value, or what, but that's the way I'd do it. He says he'll post the formula on his blog at some point anyway.

The main problem I have with Hollinger's system is that it doesn't seem to improve on the current method of drafting (ostensibly pure scouting) all that much... Instead of eliminating the mistakes, it simply makes different ones: Curtis Borchardt 4th in 2002? Mike Sweetney over Wade and Bosh in 2003? Deron Williams below Chris Taft? Paul Davis over LaMarcus Aldridge? Nick Fazekas and Josh McRoberts as top-10 picks this year? I know the method is nowhere near a finished product, but I don't think it's really usable at all for actual decision-making in its current incarnation.
Back to top
View user's profile Send private message
hoopinion



Joined: 26 Jun 2007
Posts: 5

PostPosted: Wed Jun 27, 2007 10:10 am    Post subject: Reply with quote

Mark wrote:


I may be jumping the gun for a planned post here but I noticed a post on this topic and plenty more draft stuff at http://hoopinion.blogspot.com/.



The specific post where I took a superficial, first impression look at Hollinger's study is located here: http://hoopinion.blogspot.com/2007/06/hollingers-collegiate-evaluation.html. I did a quick run-through of Wilson Chandler v. Julian Wright as my impression, both analytically and observationally, is that Wright is a far better prospect. (Admittedly, my collegiate sympathies could be a source of bias.)

As one would expect and is probably most sensible, in Hollinger's list the players I consider to be the better defensive prospects in the draft (Noah, Brewer, Julian Wright) rank behind players (Dudley, Fazekas, Thaddeus Young) who figure to concentrate their value in offensive production.
Back to top
View user's profile Send private message Visit poster's website
basketballvalue



Joined: 07 Mar 2006
Posts: 51

PostPosted: Wed Jun 27, 2007 11:26 am    Post subject: Reply with quote

gabefarkas wrote:
Hindsight is 20/20. I'm curious how much hunting and pecking there was. In other words, how much of it was a case of John knowing roughly what he wanted the results to look like, and then adding/removing factors and weights until it came out nicely?


Yes, I think it's interesting work but also a good question you raise, Gabe. I'd be curious to know if the formula was built using data from the 90s how well it would predict this decade.

John, I'd love to see you look back on your predictions for this year about this time annually. It would be fascinating reading given the strength of this class.

Thanks,
Aaron
www.basketballvalue.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  Next
Page 1 of 4

 
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