At the top of Lenovo [with Yao back it didn't take long]

Explain please.

 

http://www.82games.com/lewin2.htm

Adjusted plus-minus describes the average performance of the team with a given player in the game, taking into account the other nine players on the court.


Rafer's adjusted plus minus would account for the amount of time he spent on court with the starters - and also about how his backup sucks.

 

these stats are biased cause of the recent 50 pt margin of victory, just like the hollinger ranking being influenced by margin of victory. Over the last 25% of the games (in the hollinger ranking) our margin of victory jumps ridiculously with a 50 pt win in there and likewise it will fall alot when that game falls outside of the last 25%. With Yao not playing many minutes this season cause of the injury that 5 man rotation is gonna be benefited greatly by the 50 pt victory.

Stats don't lie, sometimes...

 

screw shooting. Basketball isnt just all about shooting. Luther's passes usually end up at the feet of JVG or of other fans and he makes boneheaded mistakes sometimes. Luther only brings outstanding shooting to the table and decent defence.

 

Stats do lie. In any sport. BJ Simmons... Case closed.

 

You thinkg JVG would be that stupid not to play Head, Tmac, Battier, Hayes and Yao more together if Head fits better than Alston?

http://www.82games.com/0607/0607HOU2.HTM

Alston, Tmac, Battier, Hayes and Yao combination has the best winning percentage, Head, Tmac, Battier, Hayes and Yao didn't even crack top 20.

On the other hand, Rockets are effective with Head replacing Tmac at 2.

Head is an undersized shooting guard, he can't run team as well as Alston, at least for now.

 

Again, how? The link you sent me to says that adjusted +/- adjusts for who is on the court with him and against him, but it says nothing about adjusting for who is on the bench.

I am relatively certain that you are mixing up +/- and +/- differential. I've done the math on these things. Raw +/- tells you nothing about a player's backup. (Although comparing +/- statistics among players who play the same position on the same team can tell you a lot.)

 

Here's an awfully good report authored by some Basketball stat heavy weights:

A Starting Point for Analyzing Basketball Statistics
(under review at the Journal of Quantitative Analysis in Sports)
http://www.uncg.edu/eco/rosenbaum/jqas1.doc

On the whole topic of plus/minus stats...

(1) There are a) +/- stats, b) net +/- stats, c) adjusted +/- stats. The NBA lenovo stats appear to be type a. 82games has type a, type b, and a "Roland Rating" which is now a modified type b.

(2) When looking at type a +/- stats folks have to keep the following in mind: "The biggest disadvantage of most plus/minus statistics is that they confound a player’s performance with the performance of his four teammates while on the floor (and five opponents, as well)." In other words we don't know which is the cause and which is the effect. Did the player actually improve his teams play or was he actually a beneficiary of the other 4 guys on his team playing with him at the time or the other 5 opponents playing against them, or a combination of both? Basic +/- doesn't tell us that.

(3) When looking at type b +/- stats folks need to keep the following in mind: "Net plus/minus statistics account for some of this issue by subtracting off how the team does without the player, effectively assuming that player substitutions are uncorrelated. However, strong correlations between teammates’ minutes on the court and systematic differences in the quality of the opponents different players face are not entirely accounted for using net plus/minus statistics. Starting lineups often play a large number of minutes together, as do deep reserves."

Also net +/- "do not measure the value of a player per se; they measure the value of the player relative to the players that substitute in for him."

So weslinder is correct in his statement about the lenovo +/- stats that they aren't a measure of the value of a player vs. his sub (which is what 82games used to highlight with the original net +/- stat).

However the lenovo stats need to looked at with the type a issues in mind.

 

You've got to be f'ing kidding me.

 

Sorry Tango, I don't want to sound like we shouldn't take Lenovo stats with large grains of salt, but some of the adjusted +/- stats produced by Rosenbaum are no less garbage as well.

Case in point:

http://bbs.clutchfans.net/showthread.php?p=2639595#post2639595

http://bbs.clutchfans.net/showthread.php?p=2639701#post2639701

 

Actually I'm not a real big fan of any +/- stat including adjusted +/- .