RealGM

http://www.basketballgeek.com/2009/10/3 ... play-data/

I have to admit being biased to any system that rates KG as the best defender in the league, but this looks like a decent method to evaluate defensive talent.

Camby is considered a poor man defender. (great help defender)
Is Odom really on that list?

its kinda loose, but Odom did have a good season though, plus that metric says it gives credit to defensive rebounding which is going to boost him, plus he plays 6th man role often last year so he's gonna get more minutes against backups. Also I'd assume big men are gonna get a slight advantage in the "Defenders that commit fouls that lead to made free throws are assigned full credit for allowing the opponent to score these points." category since I'd assume big men are going to usually foul other big men and vice versa for guards, and since big men are usually way worse free throw shooters, wouldn't that skew in favor of big men rankings?

Its cool though, I always like people that put in the time and effort to try and find a formula that gives you something

Good thread. Good OP, and good comments.

I spent a couple years looking for the quixotic measurement of defense, but interaction effects always seemed to thwart every system. This one is better than nothing, but I'm not sure one can be created.

I know he's getting old but I have a hard time believing any metric that rates TD behind Sideshow Flopper.

shrink wrote:I spent a couple years looking for the quixotic measurement of defense, but interaction effects always seemed to thwart every system. This one is better than nothing, but I'm not sure one can be created.

I think to create a "PER"-like measurement for defense, you need more and better information than what's in a conventional box score. This measure is better, IMO, than plus-minus, but still limited by what is traditionally tracked (blocks, rebounds, etc.) I've read implications that some NBA teams (the Rockets are one example) are tracking their own unique stats that are not in conventional box scores, but I've never seen the specific methodology.

As far as the Wolves go, I think adding a player like Wallace would instantly make the Wolves a .500 team. Not elite, but at least able to play with anyone in the league. Everyone would like the next Shaq or Duncan, but I think adding a two-way player like Wallace on the wing is more realistic.