View previous topic :: View next topic |
Author |
Message |
Nikos
Joined: 16 Jan 2005 Posts: 346
|
Posted: Mon Jan 17, 2005 1:53 pm Post subject: Anyone have current Bell Curve W-L records? |
|
|
Anyone have a spreadsheet of the current W-L records for each individual NBA player (similiar to the one in Dean Oliver's book)?
Im curious to see some of the %'s for many of the players.
Thanks |
|
Back to top |
|
|
kjb
Joined: 03 Jan 2005 Posts: 864 Location: Washington, DC
|
|
Back to top |
|
|
Ed Küpfer
Joined: 30 Dec 2004 Posts: 785 Location: Toronto
|
Posted: Mon Jan 17, 2005 2:39 pm Post subject: |
|
|
I can do this. Gimme a day or so, and lemme find some webspace.
BTW I don't use the bell curve method to caluclate win/loss -- I use simple Pythagorean, which is close enough. And I don't use the exact methods spelled out in DeanO's book to calculate wins -- I change them very slightly. (Basically, I calculate player possessions, divide by Team Possessions/Game, which gives us Player Games. I multiply player games by player pythagorean win% to get player wins. Player Losses is simply Player Games minus player wins) _________________ ed |
|
Back to top |
|
|
Nikos
Joined: 16 Jan 2005 Posts: 346
|
Posted: Mon Jan 17, 2005 4:30 pm Post subject: |
|
|
Cool, thanks!
I look foward to it Ed. |
|
Back to top |
|
|
Ed Küpfer
Joined: 30 Dec 2004 Posts: 785 Location: Toronto
|
Posted: Mon Jan 17, 2005 6:04 pm Post subject: |
|
|
All right, I've uploaded a csv to here. Hope that's what you're looking for. _________________ ed |
|
Back to top |
|
|
Nikos
Joined: 16 Jan 2005 Posts: 346
|
Posted: Mon Jan 17, 2005 7:44 pm Post subject: |
|
|
Yea thats cool! Only reason I mentioned Dean Oliver's method is because I like to compare it to the player W-L's on Basketball-Reference.com.
Is your W-L drastically different in terms of results from those on B-R.com?
Also Igor or anyone else, how much do you value this type of stat? Is it meaningful IYO, or is it too flawed, given players like Bruce Bowen are not fairly credited for their man defense. Do you feel a great man defender who does not pick up steals to be more valuable than someone who gets a lot of blocks and steals AKA Kirlenko etc.....?
Also, what are STOPS and DPOSS? Who is recording these? Are they valuable or simple calculations of steals and blocks as a rate of how long a player is in a game, and how many potential defensive possesion the player is exposed to?
Is %Team Poss the same as Usage Rate?
Thanks! |
|
Back to top |
|
|
Ed Küpfer
Joined: 30 Dec 2004 Posts: 785 Location: Toronto
|
Posted: Mon Jan 17, 2005 8:26 pm Post subject: |
|
|
Nikos wrote: | Yea thats cool! Only reason I mentioned Dean Oliver's method is because I like to compare it to the player W-L's on Basketball-Reference.com.
Is your W-L drastically different in terms of results from those on B-R.com? |
Identical to a decimal place, among the players I checked.
Nikos wrote: | Also Igor or anyone else, how much do you value this type of stat? Is it meaningful IYO, or is it too flawed, given players like Bruce Bowen are not fairly credited for their man defense. Do you feel a great man defender who does not pick up steals to be more valuable than someone who gets a lot of blocks and steals AKA Kirlenko etc.....? |
I like it a lot. The interpretation is intuitive and straightforward (even if the calculations are complicated). When it comes to distributing value derived from boxscore stats to players who created the offense and defense, I think this is about as far as we can go.
Nikos wrote: | Also, what are STOPS and DPOSS? Who is recording these? Are they valuable or simple calculations of steals and blocks as a rate of how long a player is in a game, and how many potential defensive possesion the player is exposed to? |
DPOSS = defensive possessions. Possessions are estimated as 1/5 of the number of possessions that take place while the player is on the floor. Stops aren't recorded, but estimated. Those calcs for STOPS and other defensive calcuations are listed in one of the appendices of DeanO's book. These represent the iffiest area of all stats analysis in the book.
Nikos wrote: | Is %Team Poss the same as Usage Rate? |
No. It is the percentage of team possessions used up by the player while he's on the floor.
Code: | %Team Poss = PlayerPossessions / PlayerMinutes
--------------------------------
TeamPossessions / TeamMinutes |
With five guys on the floor, the average will be 20%. Typically, the best player on the team gets 25-30%, and only Jordans and Iversons and Kobes get more. _________________ ed |
|
Back to top |
|
|
Nikos
Joined: 16 Jan 2005 Posts: 346
|
Posted: Mon Jan 17, 2005 10:05 pm Post subject: |
|
|
Is there any chance you can post this spreadsheet on your site every week? Or is there a certain worksheet template you can give me that I can make my own based on using DOUG STATS RAW DATA SPREADSHEET?
That would be awesome!
Oh and another thing. Isn't it quite possible that some elite players with poor W-L, make the role players W-L % better because of their presense and the fact that as leaders they see double teams and a drop in efficiency as a result.
Isn't this best to compare only when you are possibly comparing all stars to all stars, role players to other role players etc....? |
|
Back to top |
|
|
Ed Küpfer
Joined: 30 Dec 2004 Posts: 785 Location: Toronto
|
Posted: Tue Jan 18, 2005 1:30 pm Post subject: |
|
|
Nikos wrote: | Is there any chance you can post this spreadsheet on your site every week? Or is there a certain worksheet template you can give me that I can make my own based on using DOUG STATS RAW DATA SPREADSHEET? |
I can post the numbers weekly. The spreadsheet I use to generate the numbers is clunky and huge and has a gazillion user defined functions, so you probably couldn't get it to work. For next season, I'll see if I can come up with a handy one for everybody.
Nikos wrote: | Isn't it quite possible that some elite players with poor W-L, make the role players W-L % better because of their presense and the fact that as leaders they see double teams and a drop in efficiency as a result. |
Absolutely. Although only players with really crappy teammates wouldn't pass out of double teams, so you're more likely to see defenses cause drops in stats totals, not rates -- eg the player defenses focus on may see his points per game fall, but the defense won't cause much of a drop in FG% if the player is smart.
Nikos wrote: | Isn't this best to compare only when you are possibly comparing all stars to all stars, role players to other role players etc....? |
Yes! We know Yao Ming is the focus of the opposition's defense. There is no point in comparing his numbers to Loren Woods, who may be the focus of his mother's love, but that's about it. Comparing Yao to other premier big men is the best way to see how his numbers stack up. _________________ ed |
|
Back to top |
|
|
|