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 

2005 VORP
Goto page 1, 2  Next
 
Post new topic   Reply to topic    APBRmetrics Forum Index -> General discussion
View previous topic :: View next topic  
Author Message
ziller



Joined: 30 Jun 2005
Posts: 118
Location: Sac Metro

PostPosted: Wed Jul 20, 2005 12:47 pm    Post subject: 2005 VORP Reply with quote

I did a cursory search and couldn't find anything, so here's a question for those involved in calculating VORP:

I got 0.455 for the replacement level this year using Kevin's formula. Anyone think that's high? It's been that high in the past, but several hundredths lower recently.

That high number has also served to seemingly undervalue the superstars. Only Amare rated over 400 (420 by my calculations), and Garnett came in with 380. Tim Duncan really got hurt by missing games, and finished 19, one spot behind Nash.

Adjusting the replacement level to 0.450 puts Garnett at 404 and Amare at 430. But Nowitzki finishes third with that, seemingly low at 370.

What is everyone else using as the replacement level value?
_________________
SactownRoyalty.com
tziller@gmail.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 679
Location: Seattle

PostPosted: Wed Jul 20, 2005 9:08 pm    Post subject: Reply with quote

I don't think anyone else really does VORP, at least not here.

When I go through my methodology, I get Keith Bogans (efficiency .434) as the 300th-rated player, which is where I'd set replacement level (rounding to .435). That produces the following VORP leaderboard:

Code:
Player       Tm  VORP
---------------------
Stoudemire  PHO   479
Garnett     MIN   452
Nowitzki    DAL   419
James       CLE   418
Marion      PHO   379
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address
jkubatko



Joined: 05 Jan 2005
Posts: 508
Location: Columbus, OH

PostPosted: Wed Jul 20, 2005 10:02 pm    Post subject: Reply with quote

admin wrote:
I don't think anyone else really does VORP, at least not here.

When I go through my methodology, I get Keith Bogans (efficiency .434) as the 300th-rated player, which is where I'd set replacement level (rounding to .435).


I don't keep track of VORP, but it's easy enough to calculate. I also have Keith Bogans as the 300th-rated player, with a "Pelton efficiency" of 0.43387.
_________________
Regards,
Justin Kubatko
Basketball Stats!
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Mike G



Joined: 14 Jan 2005
Posts: 1506
Location: Delphi, Indiana

PostPosted: Wed Jul 20, 2005 10:03 pm    Post subject: Reply with quote

I've got 398 players who appeared last season and contributed net + eWins. Very-near zero on the scale are Geno Carlisle, Pig Miller, Yuta Tabuse, Jahidi White, Damone Brown, Tony Bobbitt, Darrick Martin, Brandin Knight, Kedrick Brown, Torraye Braggs, Maciej Lampe.

Only about 80 came in at less than zero. I guess these could be called replacement-hopefuls who could not even compete effectively against replacements.

398/30 is just about 13 per team. That seems pretty lenient, in a year when there weren't a terrible number of injuries. But the production level equal to no-wins-contributed was determined by the closest correlation to team wins.
Back to top
View user's profile Send private message Send e-mail
Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 679
Location: Seattle

PostPosted: Wed Jul 20, 2005 10:15 pm    Post subject: Reply with quote

Mike G wrote:
Only about 80 came in at less than zero. I guess these could be called replacement-hopefuls who could not even compete effectively against replacements.

You could also say they're guys who are really at replacement level, but didn't show that last season. There are probably many players whose difference from replacement level isn't statistically significant.

Though I still "use" 10, this study seemed to imply a slightly larger number, probably 11, per team. So 13 is certainly reasonable, if on the high side.
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address
ziller



Joined: 30 Jun 2005
Posts: 118
Location: Sac Metro

PostPosted: Thu Jul 21, 2005 2:04 pm    Post subject: Reply with quote

admin wrote:
I don't think anyone else really does VORP, at least not here.

When I go through my methodology, I get Keith Bogans (efficiency .434) as the 300th-rated player, which is where I'd set replacement level (rounding to .435). That produces the following VORP leaderboard:

Code:
Player       Tm  VORP
---------------------
Stoudemire  PHO   479
Garnett     MIN   452
Nowitzki    DAL   419
James       CLE   418
Marion      PHO   379


I went through it rather quickly and just used the available data from DougStats, so Bogans was like 360th on my EFF ranking. I didn't figure for all the multiple-team players, so there were some extra results in there. Do you filter players with miniscule minutes as well?

Also, is there a big reason no one uses VORP? As a baseball stathead, I see the value of it, though it might be interesting to separate the bigs from the guards since a replacement level center is very different from a replacement level point guard (like a RL catcher is different from a RL centerfielder).
_________________
SactownRoyalty.com
tziller@gmail.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
jkubatko



Joined: 05 Jan 2005
Posts: 508
Location: Columbus, OH

PostPosted: Thu Jul 21, 2005 2:10 pm    Post subject: Reply with quote

ziller wrote:
Also, is there a big reason no one uses VORP? As a baseball stathead, I see the value of it, though it might be interesting to separate the bigs from the guards since a replacement level center is very different from a replacement level point guard (like a RL catcher is different from a RL centerfielder).


In baseball VORP is denominated in runs, so it's very easy to grasp. In basketball VORP is denominated in ... I have no idea. Kevin?
_________________
Regards,
Justin Kubatko
Basketball Stats!
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 679
Location: Seattle

PostPosted: Thu Jul 21, 2005 2:14 pm    Post subject: Reply with quote

Ah, that makes sense. Yeah, you have to filter out the multi-team players. No fair counting, say, Antoine Walker as three players.

I think you have two separate questions here, one being "Why isn't this specific version of VORP used?" and the other being "Why isn't replacement level used?"

The first is pretty simple. It's a crappy little formula I put together in my dorm room, and JustinK has shown in the past that it doesn't work as well as a linear-weights system as PER (as far as estimating team performance). It isn't in terms of anything, as he points out, and I usually do PER now when I use linear weights.

As to why replacement level theory is pretty rarely used in the NBA, I'm not sure. If you do a win-percentage based system, you can do a WARP rating, which has a tangible "in terms of" wins. But EdK and I are the only ones who have really done much with that, I think.

[Edit to point out MikeG has also implicitly taken this into account with his more recent stuff, I think.]
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address
ziller



Joined: 30 Jun 2005
Posts: 118
Location: Sac Metro

PostPosted: Thu Jul 21, 2005 2:24 pm    Post subject: Reply with quote

A main reason I ask and a reason I've been looking more into replacement level in the NBA of late is a vast debate in Kings circles about Webber's tenure.

Many say that he actually hurt the team by being on it, citing some stats that show Sacramento's aggregate win percentage since 1999 was better when he wasn't playing than when he was. Some even said that the team would've won a title had Webber disappeared as was replaced by the Kings then back-ups, who were replaced by available replacement-level players.

That doesn't come close to making sense in my head, so I've been looking at a way to really study the phenomenon. Hence, VORP.

I'm going to poke around for more info on WARP. Anything in particular I should read?
_________________
SactownRoyalty.com
tziller@gmail.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 679
Location: Seattle

PostPosted: Thu Jul 21, 2005 2:31 pm    Post subject: Reply with quote

http://www.hoopsworld.com/cgi-bin/news/exec/view.cgi?archive=39&num=8638

http://www.sonicscentral.com/ratingsystem.html
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address
ziller



Joined: 30 Jun 2005
Posts: 118
Location: Sac Metro

PostPosted: Thu Jul 21, 2005 2:40 pm    Post subject: Reply with quote

Incredibly helpful as always, Kevin.
_________________
SactownRoyalty.com
tziller@gmail.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
jkubatko



Joined: 05 Jan 2005
Posts: 508
Location: Columbus, OH

PostPosted: Thu Jul 21, 2005 3:48 pm    Post subject: Reply with quote

Using Dean Oliver's wins and losses (which can be found on my site), I calculated 2005 VORP using the following formula:

Code:

VORP = PW - (0.100)*(PW + PL)

where PW = Player Wins, PL = Player Losses, and 0.100 is the replacement level


In this case VORP is denominated in wins above replacement. Here are the 2005 leaders:

Code:

Kevin Garnett           14.0
Dirk Nowitzki           12.6
LeBron James            12.1
Shawn Marion            11.6
Amare Stoudemire        11.2
Tracy McGrady           10.7
Tim Duncan               9.8
Shaquille O'Neal         9.7
Paul Pierce              9.6
Yao Ming                 9.3


(By the way, I believe this is what Ed Kupfer does, calling the result WAR10.)

The career leaders since 1978 are:

Code:

Karl Malone             197.7
Hakeem Olajuwon         163.0
Michael Jordan          162.1
John Stockton           161.3
David Robinson          149.8
Charles Barkley         146.7
Moses Malone            133.8
Reggie Miller           133.4
Shaquille O'Neal        130.5
Robert Parish           128.5

_________________
Regards,
Justin Kubatko
Basketball Stats!
Back to top
View user's profile Send private message Send e-mail Visit poster's website
jkubatko



Joined: 05 Jan 2005
Posts: 508
Location: Columbus, OH

PostPosted: Thu Jul 21, 2005 3:52 pm    Post subject: Reply with quote

Since you were asking about the Kings, here are the best player seasons since the team moved to Sacramento:

Code:

Peja Stojakovic         2004      10.4
Chris Webber            2000      10.0
Chris Webber            2001       9.8
Brad Miller             2004       8.7
Mitch Richmond          1997       8.2
Otis Thorpe             1987       7.9
Peja Stojakovic         2001       7.9
Peja Stojakovic         2002       7.8
Doug Christie           2003       7.8
Chris Webber            2002       7.8

_________________
Regards,
Justin Kubatko
Basketball Stats!
Back to top
View user's profile Send private message Send e-mail Visit poster's website
ziller



Joined: 30 Jun 2005
Posts: 118
Location: Sac Metro

PostPosted: Thu Jul 21, 2005 4:22 pm    Post subject: Reply with quote

Thanks Justin! That Doug Christie season is surprising, though his DRating was obviously awesome.
_________________
SactownRoyalty.com
tziller@gmail.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
ziller



Joined: 30 Jun 2005
Posts: 118
Location: Sac Metro

PostPosted: Thu Jul 21, 2005 6:25 pm    Post subject: Reply with quote

Thanks to the formula (that would be WARP?) and B-Ref, I took off my training wheels and calculated the Wins Above Replacement Level for the 2005 Kings:

Code:
Name   WARP
Brad Miller                   7.22
Mike Bibby                6.55
Peja Stojakovic   5.9
Darius Songaila   3.96
Chris Webber   3
Cuttino Mobley   2.05
Kenny Thomas   2.03
Brian Skinner   2.01
Maurice Evans   1.76


Miller is amazing. Twenty-five less games than Bibby, and he makes a bigger contribution.

I'm also starting to realize why Songaila is looked at highly among y'all. Nice contribution.
_________________
SactownRoyalty.com
tziller@gmail.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
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  Next
Page 1 of 2

 
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