Assessing Defensive Impact

Moderator: Doctor MJ

User avatar
yosemiteben
Forum Mod - Hornets
Forum Mod - Hornets
Posts: 20,282
And1: 13,634
Joined: Mar 20, 2013
   

Assessing Defensive Impact 

Post#1 » by yosemiteben » Mon Mar 2, 2015 9:34 pm

I'd enjoy reading a discussion of the best metrics to use to assess the impact a player has on the defensive end of the floor. On the Hornets you have two players, MKG and Bismack Biyombo, who are most valuable on that end of the floor, but quantifying that value can be difficult.

I typically start by looking at +/- numbers like those on 82games, then look at defensive tracking and opponent shooting data on stats.nba.com. Unless I have a lot of experience watching players, that's normally where I leave off.

Now let's say there is a debate going on about how much Kyrie Irving's defense has improved this year. What metrics do you think are most appropriate to use to assess the extent of the improvement in his defense? I find myself often feeling like I'm not capturing enough of the picture with the approach I mentioned above, but I'm not sure where else to turn to provide meaningful data.
blabla
Sophomore
Posts: 156
And1: 76
Joined: May 23, 2012

Re: Assessing Defensive Impact 

Post#2 » by blabla » Mon Mar 2, 2015 10:07 pm

Best defensive metric is probably ESPN's Real Plus Minus http://espn.go.com/nba/statistics/rpm/_/sort/DRPM
User avatar
bondom34
Retired Mod
Retired Mod
Posts: 66,587
And1: 50,209
Joined: Mar 01, 2013

Re: Assessing Defensive Impact 

Post#3 » by bondom34 » Mon Mar 2, 2015 10:20 pm

Yeah, DRPM is probably easiest/most accessible. For bigs, I'd add Nylon Calculus's Rim Protection numbers:
http://nyloncalculus.com/stats/rim-protection/
MyUniBroDavis wrote: he was like YALL PEOPLE WHO DOUBT ME WILL SEE YALLS STATS ARE WRONG I HAVE THE BIG BRAIN PLAYS MUCHO NASTY BIG BRAIN BIG CHUNGUS BRAIN YOU BOYS ON UR BBALL REFERENCE NO UNDERSTANDO
User avatar
yosemiteben
Forum Mod - Hornets
Forum Mod - Hornets
Posts: 20,282
And1: 13,634
Joined: Mar 20, 2013
   

Re: Assessing Defensive Impact 

Post#4 » by yosemiteben » Tue Mar 3, 2015 5:03 pm

So here is my question - let's say you are having a debate about who is the better defender, Michael Kidd-Gilchrist or Cody Zeller. I think the overwhelming majority of Charlotte fans, even those who are more sophisticated, would say MKG, yet Cody's DRPM is 11th in the league whereas MKG is 19th. MKG's defensive net impact is -11, whereas Cody's is just -3. On that Nylon Calculus rim protection chart, Cody appears just average at rim protection - he's ranked 35th in rim FG% allowed but he has a pretty low contest %.

So how would you assess the question of who is the better defender between Cody and MKG?
User avatar
bondom34
Retired Mod
Retired Mod
Posts: 66,587
And1: 50,209
Joined: Mar 01, 2013

Re: Assessing Defensive Impact 

Post#5 » by bondom34 » Tue Mar 3, 2015 11:06 pm

yosemiteben wrote:So here is my question - let's say you are having a debate about who is the better defender, Michael Kidd-Gilchrist or Cody Zeller. I think the overwhelming majority of Charlotte fans, even those who are more sophisticated, would say MKG, yet Cody's DRPM is 11th in the league whereas MKG is 19th. MKG's defensive net impact is -11, whereas Cody's is just -3. On that Nylon Calculus rim protection chart, Cody appears just average at rim protection - he's ranked 35th in rim FG% allowed but he has a pretty low contest %.

So how would you assess the question of who is the better defender between Cody and MKG?

Well, RPM tends to rate bigs a little higher in general, so there's that. Most bigs just have an inherent greater influence on defense. Zeller likely isn't a great rim protector and I haven't seen enough of him to say about his man defense or really how he works. So, I went to a final source that maybe will help. I'm typing this as I ran the query so I don't bias myself, but using nbawowy.com, running different combos:

These are opponent numbers (so lower is better):

Zeller and MKG both off court:
875 minutes, 1.103 points per possession, 50 percent eFG, 54.1 TS
http://nbawowy.com/query/4j6tb333jzv3g14i

Both on:
1148 minutes, 0.971 PPP, 46 percent eFG, 48.5 TS
http://nbawowy.com/query/e19ml1fyvkbakyb9

MKG on, Zeller off:
1010 minutes, .975 PPP, 48.1 eFG, 51.8 TS
http://nbawowy.com/query/wdyjym5hu4dt2o6r

Zeller on, MKG off:
1482 minutes, 1.058 PPP, 48.9 eFG, 53.6 TS
http://nbawowy.com/query/wj2sw3vs34cdte29

So looking at these, I'd say MKG seems to be putting up a slightly better profile. I'd say check NBA.com's player tracking for defense by play type too.
MyUniBroDavis wrote: he was like YALL PEOPLE WHO DOUBT ME WILL SEE YALLS STATS ARE WRONG I HAVE THE BIG BRAIN PLAYS MUCHO NASTY BIG BRAIN BIG CHUNGUS BRAIN YOU BOYS ON UR BBALL REFERENCE NO UNDERSTANDO
DanishLakerFan
Veteran
Posts: 2,957
And1: 569
Joined: Jan 02, 2015

Re: Assessing Defensive Impact 

Post#6 » by DanishLakerFan » Thu Jun 11, 2015 11:22 am

yosemiteben wrote:So here is my question - let's say you are having a debate about who is the better defender, Michael Kidd-Gilchrist or Cody Zeller. I think the overwhelming majority of Charlotte fans, even those who are more sophisticated, would say MKG, yet Cody's DRPM is 11th in the league whereas MKG is 19th. MKG's defensive net impact is -11, whereas Cody's is just -3. On that Nylon Calculus rim protection chart, Cody appears just average at rim protection - he's ranked 35th in rim FG% allowed but he has a pretty low contest %.

So how would you assess the question of who is the better defender between Cody and MKG?


Defense is tough to gauge since there are so many ways to prevent opposing team from scoring, but in genereal i think you have to have the same approach as you described - +/- stats, DRPM, opponent shooting data etc. I like the rim protection stat on Nyloncalculus as well and you can use the nbawowy.com as well to look at different scenarios too. However, at some point you have to try to take all that data and put it into some kind of context and watch what actually happens on the floor. For instance, Steph Curry have improved quite a bit and looking at the stats would suggest that he's among the best defenders at the PG position. But he's contantly being hidden on the opposing teams' worst offensive player, which really skews those stats.

Just based on a quick review of these different stats, i'd say that MKG is the better defender. He always gets the toughest defensive assignment, he can guard multiple defenders and from 82games.com you can see that the Hornets are 8.8 points better defensively when MKG is on the floor compared to Zeller's 3.1. From Nyloncalculus you can see Zeller contest a relatively low number of shots for his position and size (32%), so he's not exactly a rim protector. At the same time, the stats does suggest that Zeller probably is quite underrated on that end. Whether that's because he's used extremely effectively or some other reason, i think you'd have to look at some game film and see what's happening.

Return to Statistical Analysis