This is Google's cache of viewtopic.php?p=33187&sid=2aaf1e0e1900ad2635600cf081f360af. It is a snapshot of the page as it appeared on Mar 25, 2011 23:58:10 GMT. The current page could have changed in the meantime. Learn more

Text-only version
These search terms are highlighted: mtamada  
APBRmetrics :: View topic - With or Without You in Basketball
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 

With or Without You in Basketball

 
Post new topic   Reply to topic    APBRmetrics Forum Index -> General discussion
View previous topic :: View next topic  
Author Message
DSMok1



Joined: 05 Aug 2009
Posts: 608
Location: Where the wind comes sweeping down the plains

PostPosted: Fri Jan 07, 2011 2:22 pm    Post subject: With or Without You in Basketball Reply with quote

On my blog, I started to delve into the basic topic of game-level With-or-Without-You. I couldn't find any threads on here discussing this topic. Has any WOWY work been done?

Here's what I said in the intro to my post:
Quote:
The concept of With-or-Without-You is very basic. If you are playing, is our team better or worse? If the team is worse with you available, then that’s a really bad sign! It’s the core concept behind such basketball metrics as +/-, Statistical Plus/Minus and Advanced Plus/Minus. In baseball, Tom Tango and MGL[url=site:insidethebook.com wowy with without you] work with it a lot.[/url]


After pondering this further, I think I may have stumbled upon a way to validate/adjust VORP for basketball. Isn't the very definition of VORP equivalent to With-or-Without-You? If we can compile WOWY data on a large scale, won't that basically generate what value over replacement player should come up with?
_________________
GodismyJudgeOK.com/DStats
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Serhat Ugur (hoopseng)



Joined: 13 Oct 2006
Posts: 208
Location: Basketball Research

PostPosted: Fri Jan 07, 2011 6:10 pm    Post subject: Reply with quote

I recall an argument at B-R blog. Not sure that it is about what you suggest here

http://www.basketball-reference.com/blog/?p=8446#comment-37425
_________________
http://www.nbastuffer.com
Back to top
View user's profile Send private message Visit poster's website
greyberger



Joined: 27 Sep 2010
Posts: 49

PostPosted: Sat Jan 08, 2011 12:20 am    Post subject: Reply with quote

With or wiiiitthout you aaaaiiiiiieeee

I can't live

with or withooooouuuut you
Back to top
View user's profile Send private message Send e-mail
mtamada



Joined: 28 Jan 2005
Posts: 376

PostPosted: Sat Jan 08, 2011 3:17 am    Post subject: Re: With or Without You in Basketball Reply with quote

DSMok1 wrote:
After pondering this further, I think I may have stumbled upon a way to validate/adjust VORP for basketball. Isn't the very definition of VORP equivalent to With-or-Without-You? If we can compile WOWY data on a large scale, won't that basically generate what value over replacement player should come up with?


They're not quite the same: VORP compares a player against a replacement level player. WOWY compares a player to ... basically whoever his backup happens to be, which varies widely from team to team. I.e. if you're the mid-1980s Trailblazers, losing Jim Paxson wouldn't cause you to lose much, because you've got the young Clyde Drexler backing him up (and a key reason why the Blazers didn't draft yet another SG, even though his name was Michael Jordan). If you're the early-1990s Timberwolves, losing Tony Campbell might hurt you more, because of lack of scorers to replace him. But Paxson was almost certainly a better player than Campbell.
Back to top
View user's profile Send private message
DSMok1



Joined: 05 Aug 2009
Posts: 608
Location: Where the wind comes sweeping down the plains

PostPosted: Sat Jan 08, 2011 11:35 am    Post subject: Re: With or Without You in Basketball Reply with quote

mtamada wrote:
DSMok1 wrote:
After pondering this further, I think I may have stumbled upon a way to validate/adjust VORP for basketball. Isn't the very definition of VORP equivalent to With-or-Without-You? If we can compile WOWY data on a large scale, won't that basically generate what value over replacement player should come up with?


They're not quite the same: VORP compares a player against a replacement level player. WOWY compares a player to ... basically whoever his backup happens to be, which varies widely from team to team. I.e. if you're the mid-1980s Trailblazers, losing Jim Paxson wouldn't cause you to lose much, because you've got the young Clyde Drexler backing him up (and a key reason why the Blazers didn't draft yet another SG, even though his name was Michael Jordan). If you're the early-1990s Timberwolves, losing Tony Campbell might hurt you more, because of lack of scorers to replace him. But Paxson was almost certainly a better player than Campbell.


Yes, I realize they are not the same, but shouldn't they scale linearly to one another? So that the rankings and relative value in VORP would be the same as in WOWY (on average, for the average team).

If a player is at replacement level, theoretically when he is lost/isn't available, it shouldn't hurt the team.

Of course, there's a ton of noise involved in a metric with such small sample sizes.
_________________
GodismyJudgeOK.com/DStats
Back to top
View user's profile Send private message Send e-mail Visit poster's website
mtamada



Joined: 28 Jan 2005
Posts: 376

PostPosted: Wed Jan 12, 2011 8:48 pm    Post subject: Re: With or Without You in Basketball Reply with quote

DSMok1 wrote:
Yes, I realize they are not the same, but shouldn't they scale linearly to one another? So that the rankings and relative value in VORP would be the same as in WOWY (on average, for the average team).


VORP in theory is context independent. I.e. each player's VORP is getting measured against the same, constant standard, namely the Replacement Level Player.

WOWY cannot be context independent, because the numbers that pop up for say Jim Paxson change depending who's backing him up on the bench: is it Clyde Drexler or Doug West?

So WOWY could rate Paxson as just-above-benchwarmer status, or as an all-star. I.e. his relative ranking is subject to huge fluctuations.

If all players played for average teams, then yes WOWY would be basically VOAP (Value Over Average Player) and would simply scale up from VORP. But players don't all play for average teams.
Back to top
View user's profile Send private message
DSMok1



Joined: 05 Aug 2009
Posts: 608
Location: Where the wind comes sweeping down the plains

PostPosted: Thu Jan 13, 2011 9:01 am    Post subject: Re: With or Without You in Basketball Reply with quote

mtamada wrote:
DSMok1 wrote:
Yes, I realize they are not the same, but shouldn't they scale linearly to one another? So that the rankings and relative value in VORP would be the same as in WOWY (on average, for the average team).


VORP in theory is context independent. I.e. each player's VORP is getting measured against the same, constant standard, namely the Replacement Level Player.

WOWY cannot be context independent, because the numbers that pop up for say Jim Paxson change depending who's backing him up on the bench: is it Clyde Drexler or Doug West?

So WOWY could rate Paxson as just-above-benchwarmer status, or as an all-star. I.e. his relative ranking is subject to huge fluctuations.

If all players played for average teams, then yes WOWY would be basically VOAP (Value Over Average Player) and would simply scale up from VORP. But players don't all play for average teams.


You're right, WOWY would be subject to huge fluctuations. However, I think you might be over-simplifying what it says. You're forgetting the effect of chaining. In other words, look at the following case:

Team PF minutes and +/-:
Player A: 33 MPG +1.00
Player B: 12 MPG -1.00
Player C: 3 MPG -2.50
Player D: 0 MPG -3.50

If Player A goes down, what do the minutes do? Player B doesn't take all of them. In fact, it would probably look more like this:

Player B: 33 MPG
Player C: 12 MPG
Player D: 3 MPG

What would be the effect on the team?

Before: +0.28
After: -1.53

So the what is the difference? What is the effective "replacement level"? The effective replacement level is (-1.53-0.2Cool*48/33=-2.63 for this team.

Does that make sense? The effect of chaining is to cause WOWY to sort of mimic a VORP valuation.

On paper, at least.
_________________
GodismyJudgeOK.com/DStats
Back to top
View user's profile Send private message Send e-mail Visit poster's website
mtamada



Joined: 28 Jan 2005
Posts: 376

PostPosted: Sat Jan 15, 2011 2:25 am    Post subject: Reply with quote

I see chaining reducing but not eliminating the impact of the player's substitutes in his WOWY estimate. I.e. his valuation will depend on the quality as well as the minutes given to players B, C, and D. That's better than depending on just the value of his substitute, because with more players and more variables, those variables will tend to drift more to the overall average.

But it's still going to be the case that the quality and minutes of players B-D will be different for say the Celtics vs the Timberwolves. If players changed teams every year, we could argue that those B-D values would average out over the course of a career, but most players play for just a small number of teams during their career, and some will get have lower value B-D teammates over their career than others.
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