APBRmetrics

Mountain · Joined: 13 Mar 2007 Posts: 1527

Here is another large analytic effort with its own blend of components.

http://bucksdiary.blogspot.com/2008/09/2007-08-win-profiles-for-every-nba-team.html

There are other spinoff articles on the blog and more coming.

Drew Gooden ranked 13th best overall PF. Not a big difference-maker at that level but not completely inconsequential either. Realized win value ultimately depends on context.

Mountain · Joined: 13 Mar 2007 Posts: 1527

By position articles, best, worst overall and by offense & defense are all up.

Jamario Moon rates well here (and at basketball-statistics.com) because of the greater weight given defense, measured at the counterpart level (rather than at team level under WP). But he rates very well on offense here too.

Jeff Green the lowest rated player by this method but a rookie who will get a chance to show any improvement.

Bowen sucks here, but the imprecision of counterpart data leaves this debatable.

11 of the best 12 players by this method played for the best 8 playoff teams. That looks good to me.

Pierce with best defensive score in league and Garnett 11th reminds of the importance of help defense that isn't being captured with the reliance on counterpart data.

eyriq · Joined: 04 Jun 2008 Posts: 54 Location: Orlando

I'm interested to see how his methods stack up to this sites scrutiny. I really like his content but mainly because I like any commentary based on objective statistical process. He is getting ambitious with a power ranking system called PVOA, check it out here. I also know that this could be heated as he trends from the WoW school of thought....

Neil Paine · Joined: 13 Oct 2005 Posts: 774 Location: Atlanta, GA

From the looks of things, PVOA isn't really that ambitious -- it's like the SRS we have at Basketball-Reference, except with only 1 iteration instead of hundreds.

Harold Almonte · Joined: 04 Aug 2006 Posts: 616

Yes, it's like SRS but the rankings appear with a more radical weight (Lg. ave). Orlando, #3 DRTG, is showed like a so so defense because the easy SOS. It appears to be designed for bettings, but the variance of teams's additional points every game is too wide to be useful.

Mountain · Joined: 13 Mar 2007 Posts: 1527

Each method has it's advantages. SRS= rigor. His- the useful split of offense / defense.

Harold Almonte · Joined: 04 Aug 2006 Posts: 616

This method has more rigor.... But, there must be an error with Orlando