[drk3351Participant]

There is the basic steals and blocks, but everyone knows that’s not the proper way to measure a defender’s impact. Defensive win shares and defensive box plus minus is availible. But I find + – in particular to be very iffy, Damian Lillard is constantly talked about as a poor defender, though his + – is above average. Same for Kelly Olynyk. Opinions? Stat I should know about?

[Sewok15Participant]

That is pretty much all of them outside of Block Steal and DReb %’s. I think the eye test is the only true way to determine a players defensive worth. Team defense and individual defense are two way different animals which are both difficult to track statistically. Also great offense can beat great defense so there really isn’t a stat for shots that were well defended that went in.

[Sewok15Participant]

That is pretty much all of them outside of Block Steal and DReb %’s. I think the eye test is the only true way to determine a players defensive worth. Team defense and individual defense are two way different animals which are both difficult to track statistically. Also great offense can beat great defense so there really isn’t a stat for shots that were well defended that went in.

[arambone2]

 Drawn offensive fouls/charges is another useful one. Rookie Marcus Smart is "only" averaging 1.4 steals per game, but I discovered that stat, offensive fouls drawn, and Smart is tied for 5th in the entire NBA as a rookie. Add that 47 to his steals per game, and he’s well over 2 steals/drawn fouls per game, and basically elite already as a rookie. 

Will he make 3rd team All Defensive Team? He probably should, but probably won’t.

And Smart isn’t just flopping and drawing charges. Only 9 of his 47 drawn offensive fouls were charges.

http://www.nbaminer.com/player-foul-details/

 

 

 

 

 

[arambone2]

 Drawn offensive fouls/charges is another useful one. Rookie Marcus Smart is "only" averaging 1.4 steals per game, but I discovered that stat, offensive fouls drawn, and Smart is tied for 5th in the entire NBA as a rookie. Add that 47 to his steals per game, and he’s well over 2 steals/drawn fouls per game, and basically elite already as a rookie. 

Will he make 3rd team All Defensive Team? He probably should, but probably won’t.

And Smart isn’t just flopping and drawing charges. Only 9 of his 47 drawn offensive fouls were charges.

http://www.nbaminer.com/player-foul-details/

 

 

 

 

 

[TheMOSTHATEDoneParticipant]

There is defensive rating that measures many defensive statistics, some are team based.

The other is Opposite players FG% against the player, that gives you the idea of individual defensive abilities.

 

 

 

 

[TheMOSTHATEDoneParticipant]

There is defensive rating that measures many defensive statistics, some are team based.

The other is Opposite players FG% against the player, that gives you the idea of individual defensive abilities.

 

 

 

 

[treytalkssports.comParticipant]

 I follow opponents field goal percentage at the rim in order to determine true rim protection. You can find that here

Nylon calculus also has a rim protection stat that combines the above mentioned stat with the percentage of contests a defender actual has. In other words, they try to combine effectiveness at the rim with effort in protecting the rim. 

You can also go into each players profile on nba.com and view how players shoot against a specific defender and how they shoot against everyone else. In other words, fof Tony Allen, Players FG% drops from 45% to 37% when he is guarding them. They also break it down into different kinds of shots (3s, long 2s, midrange, etc.). Per your example, Lillard is decent at contesting threes, but players shoot 15% (less than 6 feet) and 12% (6-10 feet) higher against him. In other words, those players can easily get to the rim when they want to against him. 

The thing I do not understand is the "frequency" column on that page. Does that mean compared to normal shot frequency, or what? 

Those are some of the ones I use.

 

 

 

 

 

 

[treytalkssports.comParticipant]

 I follow opponents field goal percentage at the rim in order to determine true rim protection. You can find that here

Nylon calculus also has a rim protection stat that combines the above mentioned stat with the percentage of contests a defender actual has. In other words, they try to combine effectiveness at the rim with effort in protecting the rim. 

You can also go into each players profile on nba.com and view how players shoot against a specific defender and how they shoot against everyone else. In other words, fof Tony Allen, Players FG% drops from 45% to 37% when he is guarding them. They also break it down into different kinds of shots (3s, long 2s, midrange, etc.). Per your example, Lillard is decent at contesting threes, but players shoot 15% (less than 6 feet) and 12% (6-10 feet) higher against him. In other words, those players can easily get to the rim when they want to against him. 

The thing I do not understand is the "frequency" column on that page. Does that mean compared to normal shot frequency, or what? 

Those are some of the ones I use.

 

 

 

 

 

 

[sweaterflexParticipant]

 Frequency is just the percentage of the time that the shot scenario occurs out of the total distribution.  36% of the time the person Tony Allen is guarding shoots a three, 64% of the time he shoots a two, that kind of thing.

 

Lillard’s RPM is definitely altered because he plays most of the time with Robin Lopez, a very good defender. The one stat nba.com is missing that the SportVU tracking should have is drives allowed/36 minutes or something that indicates how many times someone blows by Lillard.  That kind of stat would be particularly helpful with players like Westbrook who gamble too much for steals.

[sweaterflexParticipant]

 Frequency is just the percentage of the time that the shot scenario occurs out of the total distribution.  36% of the time the person Tony Allen is guarding shoots a three, 64% of the time he shoots a two, that kind of thing.

 

Lillard’s RPM is definitely altered because he plays most of the time with Robin Lopez, a very good defender. The one stat nba.com is missing that the SportVU tracking should have is drives allowed/36 minutes or something that indicates how many times someone blows by Lillard.  That kind of stat would be particularly helpful with players like Westbrook who gamble too much for steals.

[treytalkssports.comParticipant]

That’s interesting – They indicate that the number does not have to equal 100%. Why is this? 

[sweaterflexParticipant]

You have to look at it like a Venn Diagram with overlapping points.  Shots that are within 6 ft are also within 10 ft.  It’s kind of strange they chose to map it that way instead of exclusive sets like 1-5 ft., 6-10, 11-15, etc.  It’s also strange that shot player charts are just now becoming publicly available when you can see them as far back as NBA Live 96. 

[sweaterflexParticipant]

You have to look at it like a Venn Diagram with overlapping points.  Shots that are within 6 ft are also within 10 ft.  It’s kind of strange they chose to map it that way instead of exclusive sets like 1-5 ft., 6-10, 11-15, etc.  It’s also strange that shot player charts are just now becoming publicly available when you can see them as far back as NBA Live 96. 

[treytalkssports.comParticipant]

That’s interesting – They indicate that the number does not have to equal 100%. Why is this? 

[Author]

Posts