Perimeter defense - by the numbers

Much has been written about the defensive skills of Bradley and Smart while Thomas has been viewed by many as a defensive liability. However, high tech video data paints a very different picture.

   Below data is for the current year and was taken from the below site.

   http://stats.nba.com/league/player/defense/#!/gt15/?sort=PLAYER_LAST_TEAM_ABBREVIATION&dir=-1


        Perimeter shooting - 15 feet plus
 Thomas -  The players that he has been guarding at 15 ft plus have averaged 33.0 fg%
 Smart   - ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 40.4 fg%
 Bradley - ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 42.4 fg%

        Low post - under 10 feet
  Thomas - The players that he has been guarding at under !0 feet have averaged 56.7 fg%
  Smart   - ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 48.8 fg%
  Bradley - ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 64.6 fg%

       #1 - What gives with Bradley
       #2 - Smart dominates down low
       #3 - Thomas is the best out on the perimeter

 Certainly lowering the opponents fg% is a very important part of good defense. But we're only a quarter way through the season so here's hoping that Bradley turns things around. --------------- Are you surprised by the Thomas perimeter numbers ? Obviously quickness plays a huge role in perimeter defense - and Thomas sure as an abundance of it.

  swish

Maybe Bradley has been defending better shooters? Maybe they aren`t getting as many shots off against him or not challenging him unless they are wide open? There are an awful lot of other variables, aren`t there, Swish?

jrleftfoot wrote:Maybe Bradley has been defending better shooters? Maybe they aren`t getting as many shots off against him or not challenging him unless they are wide open? There are an awful lot of other variables, aren`t there, Swish?

Exactly

Thomas guards the worst perimeter scorer on the opposing team. 

Bradley and Smart guard the best. As far as those low post numbers are concerned, how many times have we seen Bradley stuck on someone like Durant this year?

Thomas is not a horrendous liability (if that's what people make him out to be, I personally don't see or hear his D being bashed as much as it used to be). But the fact remains that he is 5'9" and will always need to gamble for steals and use his fouls in order to keep defenders honest. He's also going to have to remain a 22-26ppg scorer and 6 assist to 2 turnover type with 48/36/86 shooting splits to remain a viable starter and 2nd/3rd best player on a contender. Once he loses a step and slips just a bit, he'll be a 16-18ppg guy with a 5:3 ratio, 45% FG and 32% 3 pt FG player who only deserves a bench role. The margin for error is just very small for him b/c of his height and mediocre D.

Great points JR and Ram. That's why stats are dangerous weapons. Bradley always guards the best offensive guard on the other team. Smart is also on a top offensive player, sometimes a player much bigger than him. We've seen Bradley and Smart guarding Durant and Paul George, just to name two, both of whom are superior scorers and much taller than either of them.

I use stats to validate my eyeballs, not to replace them. I also use stats to direct my eyeballs towards aspects of the game that they may have missed (or have not given enough weight or credit to). Basketball is not a laboratory experiment where you can control the inputs, processes and methods. Stats have their place but I don't use them as my lead unless I have nothing else to go on (e.g. I'll use the box score, stats, if I didn't see the game).


bob


.

In this case we're not talking about the number of points scored by the players being defended - but rather the shooting percentages they achieve.

Here's a closer look at the 15 feet plus figures.

Thomas defended against an average of 4.9 shots per game.
Smart,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 5.2 ,,,,,,,,,,,,,,,,,,,
Bradley,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 6.0 ,,,,,,,,,,,,,,,,,,,,

Thomas guarded players who had a league shooting percent average of 37.5 - against Thomas it was 33.0
Smart ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 37.1 - against Smart ,,,,,,,,,, 40.4
Bradley ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 36.2 - against Bradley,,,,,,,,, 42.4

I don't claim to know exactly what it means but I sure in hell find it to be surprising numbers - And based on reputation, not what I expected.

swish

Bradley's numbers in 2015-16

Defended against 5.3 fga per game
Those he guarded averaged 35.9% per game
While those he guarded had a league average of 37.7

Much, much better than his current numbers

swish

swish wrote:In this case we're not talking about the number of points scored by the players being defended - but rather the shooting percentages they achieve.

   Here's a closer look at the 15 feet plus figures.

Thomas defended against an average of 4.9 shots per game.
Smart,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 5.2  ,,,,,,,,,,,,,,,,,,,
Bradley,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 6.0 ,,,,,,,,,,,,,,,,,,,,

Thomas guarded players who had a league shooting percent average of 37.5 - against Thomas it was 33.0
Smart ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 37.1 - against Smart ,,,,,,,,,, 40.4
Bradley ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,    36.2 -  against Bradley,,,,,,,,, 42.4

I don't claim to know exactly what it means but I sure in hell find it to be surprising numbers - And based on reputation, not what I expected.

swish

   

 

swish,

Another relevant stat to this discussion is the ftas by the players guarded by these 3.  If they are fouled in the act of shooting, and the shot does NOT drop, then it does not count as a fga (and, therefore, does not affect their fg%) but they do get the fritos, and that means points.  If you are not playing good defense, and are picking up a lot of fouls because you are out of position, then your defensive fg% will look good but the box score of the player you are guarding will look bad (to us).

Furthermore, I still stand by the position that fg% given up is not a good, revealing stat unless it is put in perspective of who they were guarding and what that player's fg% when guarded by other players.  Marcus Smart is going to give up an awful fg% if he is guarding Kevin Durant, a defensive assignment both he and Bradley have been given.  Does that make them bad defenders or that they were given the hardest assignments?  If IT is assigned to the lesser scorer of their back court does it not make sense that he would have a better defensive fg% than if he was given the better scorer?  IT is never assigned to defend a SF, Bradley and Smart (especially Smart) are all the time despite that not being their natural position.

The danger of using statistics to formulate opinions from scratch is that perspective can be lost and perspective is everything.


bob


.

bobheckler wrote:

swish wrote:In this case we're not talking about the number of points scored by the players being defended - but rather the shooting percentages they achieve.

   Here's a closer look at the 15 feet plus figures.

Thomas defended against an average of 4.9 shots per game.
Smart,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 5.2  ,,,,,,,,,,,,,,,,,,,
Bradley,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 6.0 ,,,,,,,,,,,,,,,,,,,,

Thomas guarded players who had a league shooting percent average of 37.5 - against Thomas it was 33.0
Smart ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 37.1 - against Smart ,,,,,,,,,, 40.4
Bradley ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,    36.2 -  against Bradley,,,,,,,,, 42.4

I don't claim to know exactly what it means but I sure in hell find it to be surprising numbers - And based on reputation, not what I expected.

swish

   

 

swish,

Another relevant stat to this discussion is the ftas by the players guarded by these 3.  If they are fouled in the act of shooting, and the shot does NOT drop, then it does not count as a fga (and, therefore, does not affect their fg%) but they do get the fritos, and that means points.  If you are not playing good defense, and are picking up a lot of fouls because you are out of position, then your defensive fg% will look good but the box score of the player you are guarding will look bad (to us).

Furthermore, I still stand by the position that fg% given up is not a good, revealing stat unless it is put in perspective of who they were guarding and what that player's fg% when guarded by other players.  Marcus Smart is going to give up an awful fg% if he is guarding Kevin Durant, a defensive assignment both he and Bradley have been given.  Does that make them bad defenders or that they were given the hardest assignments?  If IT is assigned to the lesser scorer of their back court does it not make sense that he would have a better defensive fg% than if he was given the better scorer?  IT is never assigned to defend a SF, Bradley and Smart (especially Smart) are all the time despite that not being their natural position.

The danger of using statistics to formulate opinions from scratch is that perspective can be lost and perspective is everything.


bob


.

bob

Quoting you from above "unless it is put in perspective of who they were guarding and what that player's fg% when guarded by other players."

And fortunately this site takes that into consideration. Review my previous post and note guarded players shooting percentage.

swish