January 19, 2010

Some Thoughts on the NBA’s Best Player

Filed under: Uncategorized — wwinston @ 8:47 am

Based on all games through MLK Day, the three best players by Adjusted +/- ratings are Dirk, LeBron, and DWade in a virtually dead heat. After adjusting for the abilities of their opponents and teammates all three players rate 23 points per game better than an average player. The greatly improved KD is 4th (+18 points.)

  Of course, I will be criticized for not having Kobe at the top of our list. Kobe is a great player (his Adjusted +/- is 12 points per game better than an average player), but due to the fact he has much better teammates than Dirk, LeBron and DWade , I think many people overestimate Kobe’s abilities.   For the algebraically minded reader, here is an analysis of why Kobe (based on play this year)  is probably not  25 points better than an average player.

    With Kobe in the Lakers have played 11 points better than average.  65% of the time Kobe is in with Artest, 67% of the time he is in with Odom and 49% of the time he is in with Gasol. Thus on average Kobe is on the court with 1.8 of these excellent players and 2.2 other (not so good players).

Let’s assume Gasol, Artest and Odom average 5 points better than average (this is consistent with their past abilities and our estimates from this year). Then

(Kobe ability)+ 1.8(average ability of RA, PG and LO) + 2.2(Average ability of other Lakers) = 11

Let’s put in Kobe ability = +25 . Then we get

25 +1.8(5) +2.2(Average ability of other Lakers) = 11

or Average ability of other Lakers is around -10 points. There is no way the other Laker players are this bad.  Farmar, Fisher, Bynum, Brown, Walton and Mbenga may not be superstars, but they are just not this bad.

Well let’s suppose that RA, PG and LO are just average players. Then for Kobe to be 25 points better than average we  must have

25 + 1.8(0) +2.2(Average ability of other Lakers) = 11

or Average ability of other Lakers = -6 points.

Estimating other Lakers to be 6 points below an average player seems plausible. Therefore, if you want to believe Kobe is as good as Dirk, LeBron, or DWade, you should believe that Odom, Gasol and Artest (together) are of average ability. If you believe that, fine. I don’t believe that PG, RA and LO are average players!


  1. I am not sure where you are getting your stats. Basketballvalue shows Dirk clearly behind Wade and Lebron, the 2 leaders. I will use BV’s #s for this post since they are the only ones I have, but if you want to show why the stats there are incorrect then go ahead.

    But as for Kobe’s adjusted number, them being that bad is not as implausible as you think. Fisher is statistically the worst starter in the entire league based on net box score production. The adjusted +/- values of Fisher, Bynum, Farmar, Brown are negative 8, 5, 3.5, and 11. The rest of the bench is much worse.

    If you make the quite plausible assumption that those 2.2 players are about negative 8 to 9, then Kobe should have a value of 19.6 to 21.8. This in fact does match up to his net +/- of 19.1(3rd in the league and barely trailing Wade and Nowitzki among full time starters)

    I am genuinely curious why Kobe seems to be taking such a huge hit to his net +/- number relative to Lebron and Wade.

    Maybe what is happening is that Bynum and Fisher heavily piggyback on the Laker’s good players. This gives them a good net +/- but a bad adjusted figure. However, only the distorted net is used to calculate Kobe’s numbers, which therefore gives an adjusted number that is artificially low. If bad players on the Heat and Cavs do not play as much with the starters as Bynum and Fisher do, then those teams will not have this effect. I am not a statistician, but that is my best guess about what is happening.

    Comment by dave — January 19, 2010 @ 10:05 pm

  2. we do our own adjusted +/-. Dirk is not good in 2 year adjusted on BV.com because they have him por last year. We do not. I think Dirk got blamed in their system last year for Devin George’s bad play . In this years adjusted Basketball value has dirk better than Kobe this year, as do we.I happen to think we do adjusted +/- a better way than other people (but pf course I am biased). Last year basketballvalue.com had Dirk with a negative rating which is unreasonable. We had Dirk at +7.

    Comment by wwinston — January 19, 2010 @ 10:30 pm

RSS feed for comments on this post. TrackBack URL

Leave a comment

Powered by WordPress