Sabermetrics come to the NBA

First of all, you have got to love Mark Cuban.

I mean, who else got out of the Internet bubble at the absolute peak quite like he did? Second, who has spent his money better than Cuban has? If I had $3 billion in liquid assets, then I cannot imagine a better purchase than a sports team. Finally, he really truly thinks about running his franchise. I am life-long L.A. Lakers fan, but watching the ride in Dallas has been a lot of fun. Cuban posts his thoughts on blogmaverick.com as he is running that business. Therefore, when he makes a mistake (like letting Steve Nash leave), you can read Cuban's reasoning in his own words. This is a lot better than trying to reverse engineer the thoughts of your owner and/or GM. Also, Cuban has begun to apply the same types of sabermetric analysis that is transforming some baseball clubs (like the A's, Red Sox and Dodgers). If there is any sport that needs sabermetrics the most, it is the NBA. The manner in which the sport was marketed in the mid-to-late '90s moved the focus away from teams and on to invidual players. This worked to sell tickets, while those players (Jordan, Ewing, Barkley, et al) were active, but it also changed the skills that younger players focused upon. The NBA is badly in need of a movement that helps to de-mystify it as a sport.

Mark Cuban is the first person who can write a big check working on this, so good for him.

More to the point, it is way past time that folks started talking about basketball as a team sport that requires a lot of hard work to master. I am sure there are plenty of dumb NBA players (Ron Artest comes to mind), but in interviews the vast majority of the good ones come of as very smart and self-aware. Anyway, Cuban choses not to spell out his metrics, but there is a great web-site called 82games.com that does. Here are a few things that jump out at me:

• If you read down the list of the Lakers best 5-man rotations, then the substitution pattern Rudy T is using starts to make more sense. Jones is more effective on the floor in the 3-spot than Butler, so Rudy T needs to get him into games ahead of Walton.
• Swapping an expensive Steve Nash for a cheap Jason Terry turns out to be a reasonable move for this season in Dallas. It is still not great to break-up a nucleus of players that works well together, but at least the Mavs found a reasonable statistical replacement.
• In the same vein, the San Antonio Spurs are a case study in the merits of keeping units together. Look at their player pair ratings, when compared with a team like the Houston Rockets who are trying to fit together a lot of new pieces.
• Shaq would've been better if he had stayed with Kobe Bryant. O'Neal and Bryant teamed for +10.5 per game and W% of 81%. Shaq and Dwayne Wade combine for a +6.5 per game and W% of 59%, which is good only comparison to Bryant and Lamar Odom.
• The Baby Bulls may finally be finding their way, but they still need to get rid of Eddy Curry. Look at their player pairs. The kid has an NBA Centers body, but he makes most of his team-mates worse. This is exactly the opposite of what you are looking for in a player. Anyway, here is good article on the kid.

I will have a lot more thoughts on the rise of basketball statistical analysis as I get a handle on it.