Department of Defense (with a special guest appearance by GV!)

[JimiCliff]

http://grantland.com/features/department-of-defense/

Absolutely fantastic stuff here.

Can't wait to see how the Raptors stack up.

[Scraptor]

Some other interesting notes from the source article:

DeMar received the most on-ball defensive attention of any player (ahead of KD).

DeMar was third-best among guard/forwards at hurting shot efficiency on the perimeter, after Carlos Lee and DWade.
TRoss was second-worst among guard/forwards at preventing/discouraging shot frequency in the restricted area.

Toronto had the 5th highest "defensive entropy" of any team. This is a measure of defensive activity in terms of switches/double teams.

.....

Super-interesting, but also highlights just how tricky it is to draw conclusions from this.

Is DeMar receiving the most attention simply because he cannot effectively create space and frequently drives into double teams and traffic? We've already highlighted in other threads that he takes one of the highest proportion of contested shots relative to his peers.

Does DeMar disrupt perimeter D at the expense of getting blown by, because he can't recover? On isolation PPP per synergy, he is at the 30th percentile. Greivis, who's also highlighted in the Grantland article, is at the 48th percentile. If our guys are getting broken down on the perimeter it seems like the model simply assigns defensive blame to whoever's closest: "Our model circumvents the challenges associated with identifying the most appropriate counterpart for a player, since we directly infer who is guarding whom at every instant of a possession."

Hopefully I read and understood everything correctly, but it seems super tricky. But a great first step, and I'm sure the models will evolve as time goes on.

[Axel]

Some other interesting notes from the source article:

DeMar received the most on-ball defensive attention of any player (ahead of KD).

DeMar was third-best among guard/forwards at hurting shot efficiency on the perimeter, after Carlos Lee and DWade.
TRoss was second-worst among guard/forwards at preventing/discouraging shot frequency in the restricted area.

Toronto had the 5th highest "defensive entropy" of any team. This is a measure of defensive activity in terms of switches/double teams.

.....

Super-interesting, but also highlights just how tricky it is to draw conclusions from this.

Is DeMar receiving the most attention simply because he cannot effectively create space and frequently drives into double teams and traffic? We've already highlighted in other threads that he takes one of the highest proportion of contested shots relative to his peers.

Does DeMar disrupt perimeter D at the expense of getting blown by, because he can't recover? On isolation PPP per synergy, he is at the 30th percentile. Greivis, who's also highlighted in the Grantland article, is at the 48th percentile. If our guys are getting broken down on the perimeter it seems like the model simply assigns defensive blame to whoever's closest: "Our model circumvents the challenges associated with identifying the most appropriate counterpart for a player, since we directly infer who is guarding whom at every instant of a possession."

Hopefully I read and understood everything correctly, but it seems super tricky. But a great first step, and I'm sure the models will evolve as time goes on.

Your interpretation is the same as mine. No way GV is really a good defender. The model gets skewed by passing the responsibility to the help defender, which would enhance the stats of blown by defenders. Really seems to emphasis contesting shots more than anything.

[MACK11]

GV a good defender? Hahahahah