Evan Turner’s Defensive Prowess, Part 2: Defensive Percentiles

[Blaz06Draft]

Evan Turner is a standout defender according to NBA.com’s Play Type defensive statistics. This summarizes defensive percentiles. A percentile is what percent of NBA players you rank better than in a given stat. So a 90% percentile means that you are better than 90% of NBA players.

I also compare him to All-NBA Defensive Team wings.

Spot Up: 20% of NBA possessions
Turner 85.0%
Avery Bradley 60.3%
Jimmy Butler 32.3%
Paul George 78.7%
Kawhi Leonard 75.1%
Comment: Turner is best of the group in the most common NBA possession.

Ball Handler: 17% of NBA possessions
Turner 92.3%
Bradley 58.7%
Butler 37.8%
George 35.7%
Leonard 89.7%
Comment: Turner is best of the group on the 2nd most common NBA possession.

Transition: 13% of NBA possessions
No individual defensive data provided.

Cut: 8% of NBA possessions
No individual defensive data provided.

Isolation: 7% of NBA possessions
Turner 73.7%
Bradley 81.5%
Butler 76.9%
George 59.6%
Leonard 83.0%
Comment: Turner is 4th, but still better than ¾ of NBA players.

Post Up: 7% of NBA possessions
Turner 44.9%
Bradley 25.0%
Butler 25.8%
George 34.0%
Leonard 70.9%
Comment: Turner is 2nd, well above all but Leonard.

Roll Man: 7% of NBA possessions
Turner 34.2%
Bradley 91.7%
Butler 27.7%
George 27.0%
Leonard 97.8%
Comment: Leonard and Bradley are elite, while Turner is best of the rest.

Off Screen: 5% of NBA possessions
Turner 34.2%
Bradley 91.7%
Butler 27.7%
George 27.0%
Leonard 97.8%
Comment: Leonard and Bradley are elite, while Turner is best of the rest.

Hand Off: 4% of NBA possessions
Turner 60.4%
Bradley 19.6%
Butler 17.0%
George 33.7%
Leonard 77.5%
Comment: Turner is 2nd to Leonard, and well above the others.

Putback: 6% of NBA possessions. Data available. Didn’t investigate.

Misc: 6% of NBA possessions. Data available. Didn’t investigate.

Weighted Average: If you take a weighted average of these players’ percentiles (weighted by number of NBA possessions), you get the following rating (in order):
Leonard 82.4%
Turner 72.8%
Bradley 56.7%
George 51.1%
Butler 37.2%
Comment: While not in the same class as Defensive Player of the Year Kawhi Leonard, Evan Turner is elite at defending the two most common NBA possessions, and better than the other All-NBA Defensive Team wings in most other categories. As a result, he ranks 2nd on this list made up on Turner and the All-NBA Defensive Team wings.

 

[BlazerCaravan]

@SlyPokerDog, can you get this into the main Blazers forum?

 

[SlyPokerDog]

@SlyPokerDog, can you get this into the main Blazers forum?

Anything for you.

 

[Blaz06Draft]

@SlyPokerDog, can you get this into the main Blazers forum?

What is "the main Blazers forum?"

 

[blue9]

Pretty cool!

I find it quite interesting that - according to this metric - Jimmy Butler doesn't really seem to be all that great of a defender.

 

[SlyPokerDog]

What is "the main Blazers forum?"

You originally posted this in the off topic section. I moved it to the main/Blazer forum. You didn't do anything wrong.

 

[SlyPokerDog]

Evan Turner is a standout defender according to NBA.com’s Play Type defensive statistics. This summarizes defensive percentiles. A percentile is what percent of NBA players you rank better than in a given stat. So a 90% percentile means that you are better than 90% of NBA players.

I also compare him to All-NBA Defensive Team wings.

Spot Up: 20% of NBA possessions
Turner 85.0%
Avery Bradley 60.3%
Jimmy Butler 32.3%
Paul George 78.7%
Kawhi Leonard 75.1%
Comment: Turner is best of the group in the most common NBA possession.

Ball Handler: 17% of NBA possessions
Turner 92.3%
Bradley 58.7%
Butler 37.8%
George 35.7%
Leonard 89.7%
Comment: Turner is best of the group on the 2nd most common NBA possession.

Transition: 13% of NBA possessions
No individual defensive data provided.

Cut: 8% of NBA possessions
No individual defensive data provided.

Isolation: 7% of NBA possessions
Turner 73.7%
Bradley 81.5%
Butler 76.9%
George 59.6%
Leonard 83.0%
Comment: Turner is 4th, but still better than ¾ of NBA players.

Post Up: 7% of NBA possessions
Turner 44.9%
Bradley 25.0%
Butler 25.8%
George 34.0%
Leonard 70.9%
Comment: Turner is 2nd, well above all but Leonard.

Roll Man: 7% of NBA possessions
Turner 34.2%
Bradley 91.7%
Butler 27.7%
George 27.0%
Leonard 97.8%
Comment: Leonard and Bradley are elite, while Turner is best of the rest.

Off Screen: 5% of NBA possessions
Turner 34.2%
Bradley 91.7%
Butler 27.7%
George 27.0%
Leonard 97.8%
Comment: Leonard and Bradley are elite, while Turner is best of the rest.

Hand Off: 4% of NBA possessions
Turner 60.4%
Bradley 19.6%
Butler 17.0%
George 33.7%
Leonard 77.5%
Comment: Turner is 2nd to Leonard, and well above the others.

Putback: 6% of NBA possessions. Data available. Didn’t investigate.

Misc: 6% of NBA possessions. Data available. Didn’t investigate.

Weighted Average: If you take a weighted average of these players’ percentiles (weighted by number of NBA possessions), you get the following rating (in order):
Leonard 82.4%
Turner 72.8%
Bradley 56.7%
George 51.1%
Butler 37.2%
Comment: While not in the same class as Defensive Player of the Year Kawhi Leonard, Evan Turner is elite at defending the two most common NBA possessions, and better than the other All-NBA Defensive Team wings in most other categories. As a result, he ranks 2nd on this list made up on Turner and the All-NBA Defensive Team wings.

I'm really enjoying your posts, thanks for taking the time and research to post these.

 

[Rastapopoulos]

I have to say, while I appreciate the OP's work, I am always suspicious of defensive stats. Here's one problem: suppose your team employs a ton of double/triple teams. Won't that make your stats look better? Ironically, if you're thought of as a weak defender, your coach will send help all the time, so your guy will actually look pretty bad (and maybe the guy being doubled-off will look good, thereby making the doubler look defensively bad because he had to help you out). Or, suppose you're Bruce Bowen and your coach won't double, but puts you on the best scorer on the other team. This stacks the stats against you because you're always defending someone awesome who is bound to score more than the crappy guy on the other team (who your coach is hiding his worst defender on). How can any stat take things like that into account, i.e., coaching decisions and style of play?

 

[Rastapopoulos]

I think, for example, that Khris Middleton may benefit from the Bucks' weird style of defense (or at least, he did the first year they implemented it, before other teams knew what to do with it).

 

[Boob-No-More]

I have to say, while I appreciate the OP's work, I am always suspicious of defensive stats. Here's one problem: suppose your team employs a ton of double/triple teams. Won't that make your stats look better? Ironically, if you're thought of as a weak defender, your coach will send help all the time, so your guy will actually look pretty bad (and maybe the guy being doubled-off will look good, thereby making the doubler look defensively bad because he had to help you out). Or, suppose you're Bruce Bowen and your coach won't double, but puts you on the best scorer on the other team. This stacks the stats against you because you're always defending someone awesome who is bound to score more than the crappy guy on the other team (who your coach is hiding his worst defender on). How can any stat take things like that into account, i.e., coaching decisions and style of play?

Advanced defensive stats are still in their infancy. As I commented in the other Turner thread yesterday, DRPM is a severely flawed stat that doesn't come close to passing the eye test.

In this case, I noticed that the Defensive Play Type stats are heavily dependent on sample size and position/role. I do think they have some value, as long as you are comparing players with comparable sample sizes and roles - which is what the OP has done here.

Just for grins, I looked at the first two categories (most significant sample sizes) to see how Turner compared to the former player many want to compare him to, Nicolas Batum. I also tossed in Dame and C.J., since the common accepted wisdom (aka: eye test) says they are poor defenders. Here's what I found:

Spot Up: 20% of NBA possessions
Turner 85.0%
Batum 64.4%
Dame 11.3%
C.J. 24.6%
Comment: Edge to Turner. Dame and C.J. suck at defense.

Ball Handler: 17% of NBA possessions
Turner 92.3%
Batum 83.5%
Dame 26.0%
C.J. 34.2%
Comment: Sight edge to Turner. Dame and C.J. suck at defense.

Not sure if the data is sufficient to declare Turner a better defender than Batum, but since he will be making $55 million less, I'm happy we have Turner. The eye test is met with Dame and C.J. The data passes the eye test filter in this case.

BNM

 

[Rastapopoulos]

Dame and CJ must be world-beaters on offense, then, to be "the second-best backcourt in the NBA" ---BNM.

 

[Rastapopoulos]

I will concede that Batum may be overrated on D and Turner underrated. I will certainly look forward to watching Turner (a signing I was very happy about - it struck me that I was in the minority). One complicating factor is that Turner probably wasn't guarding their better players (they have All-NBA defender Avery Bradley for the guards, also Marcus Smart, and Jae Crowder for the SFs) whereas Batum probably was (after MKG went down). And also Batum is much more versatile. He's guarded 1-4 and has a much better wingspan and can do those nice chasedown blocks.

Certainly happy to see stats saying Turner is good. Let's see if he helps the Blazers get better.

 

[Blaz06Draft]

I have to say, while I appreciate the OP's work, I am always suspicious of defensive stats. Here's one problem: suppose your team employs a ton of double/triple teams. Won't that make your stats look better? Ironically, if you're thought of as a weak defender, your coach will send help all the time, so your guy will actually look pretty bad (and maybe the guy being doubled-off will look good, thereby making the doubler look defensively bad because he had to help you out). Or, suppose you're Bruce Bowen and your coach won't double, but puts you on the best scorer on the other team. This stacks the stats against you because you're always defending someone awesome who is bound to score more than the crappy guy on the other team (who your coach is hiding his worst defender on). How can any stat take things like that into account, i.e., coaching decisions and style of play?

AGREED that defensive stats can be misleading, and are in part system dependent. However, they are all we have right now. And all players are evaluated the same way.

You could also say that Crabbe's and Meyer's good 3 point shooting is inflated by having Dame and CJ as high level threats taking the majority of the defensive focus. And that they benefit from Stotts' offensive system wizardry. So some of the same type of comments can be made about offensive stats. It's just that the ball going in the hoop is pretty definitive, while defensive acts (other than blocks and steals), are not.

 

[Rastapopoulos]

AGREED that defensive stats can be misleading, and are in part system dependent. However, they are all we have right now. And all players are evaluated the same way.

You could also say that Crabbe's and Meyer's good 3 point shooting is inflated by having Dame and CJ as high level threats taking the majority of the defensive focus. And that they benefit from Stotts' offensive system wizardry. So some of the same type of comments can be made about offensive stats. It's just that the ball going in the hoop is pretty definitive, while defensive acts (other than blocks and steals), are not.

Yes, I agree. All stats are useless out of context. And they pretty much always will be.

Eventually they'll probably program an AI to do a better job than any human scout, but it'll be from watching reams and reams of video, not from looking at stats.

 

[Boob-No-More]

Dame and CJ must be world-beaters on offense, then, to be "the second-best backcourt in the NBA" ---BNM.

They are. And, I'm not the only one who thinks so. A lot of people smarter than me kept calling them that all through the playoffs.

BNM

 

[BonesJones]

Can someone compare Turner to Aminu?

 

[Rastapopoulos]

Can someone compare Turner to Aminu?

Also: Andre Roberson.

 

[Rastapopoulos]

Looking at the three threads you've started @Blaz06Draft... are you Evan Turner's agent?

 

[Boise Blazer]

Looking at the three threads you've started @Blaz06Draft... are you Evan Turner's agent?

I love stats and I appreciate these as well but I couldn't help but laugh at your comment because my first thought was, "Is this Neil Olshey showing us why and how he chose Turner?"

 

[Boob-No-More]

Also: Andre Roberson.

I'm not going to do them all, just the two most statistically relevant play types (like my previous post):

Spot Up: 20% of NBA possessions
Turner 85.0%
Aminu 58.8%
Roberson 32.5%
Comment: This exercise is left to the reader.

Ball Handler: 17% of NBA possessions
Turner 92.3%
Aminu 41.9%
Roberson 77.3%
Comment: This space intentionally left blank.

BNM

 

[Blaz06Draft]

I love stats and I appreciate these as well but I couldn't help but laugh at your comment because my first thought was, "Is this Neil Olshey showing us why and how he chose Turner?"

No, just a fan trying to understand why NO chose Turner, what Turner's strengths are as a player, and how he might fit into the Blazers.

 

[rasheedfan2005]

cool thread, fun to see him stomp on paul george in every stat

 

[Blaz06Draft]

Evan Turner is a standout defender according to NBA.com’s Play Type defensive statistics. This summarizes defensive percentiles. A percentile is what percent of NBA players you rank better than in a given stat. So a 90% percentile means that you are better than 90% of NBA players.

Weighted Average: If you take a weighted average of these players’ percentiles (weighted by number of NBA possessions), you get the following rating (in order):
Leonard 82.4%
Turner 72.8%
Bradley 56.7%
George 51.1%
Butler 37.2%
Comment: While not in the same class as Defensive Player of the Year Kawhi Leonard, Evan Turner is elite at defending the two most common NBA possessions, and better than the other All-NBA Defensive Team wings in most other categories. As a result, he ranks 2nd on this list made up on Turner and the All-NBA Defensive Team wings.

Here are DRtg for the list, plus the Blazer wings from last year, plus Nic and Wes from their Blazer years: basketball-reference:

Turner 104
Bradley 106
Butler 106
George 101
Leonard 96
Aminu 107
Harkless 108
Crabbe 110
Batum 107 (Blazer career)
Matthews 110 (Blazer career)

Here, Turner compares favorably to all but Leonard and George, who rates higher here than the detailed analysis I did yesterday. He is also better than Nic and Wes.

His DRtg was around 4 points per 100 possessions better than last year's Blazer wings, which is significant

 

[e_blazer]

Here's another question for you stats guys: how does Turner compare to Parsons? Obviously, Parsons is the better offensive player, but as Olshey's first free agent target, how does Parsons stack up against Turner? Was Olshey mainly looking to bolster the defensive prowess of the Blazers, or was that just a secondary thing that Turner happened to bring to the table?

 

[Harry's Raincoat]

It is now obvious that Olshey was trying to make the roster a better one defensively, without hurting the offense.

 

[Trackjack]

Parson does not compare to Turner on defense. What Parson is better is shooting the 3 ball. I like Turner better due to he more rounded the Parson on his overall game. I see Turner increase his points this year to 15-17 points a game. His stats will be 15/6/6.

 

[Boob-No-More]

Here's another question for you stats guys: how does Turner compare to Parsons? Obviously, Parsons is the better offensive player, but as Olshey's first free agent target, how does Parsons stack up against Turner? Was Olshey mainly looking to bolster the defensive prowess of the Blazers, or was that just a secondary thing that Turner happened to bring to the table?

Once again, just the top two play types by sample size:

Spot Up: 20% of NBA possessions
Turner 85.0%
Parsons 64%

Ball Handler: 17% of NBA possessions
Turner 92.3%
Parsons 76.7

Parsons is known as a good 2-way payer, which is exactly what this team needs. He's not a lockdown defender, but passable to the point of probably being better than average. At, least he doesn't suck on defense and would have given us a reliable 3rd option on offense to take some defensive pressure off Dame and C.J.

If Parsons was Neil's Plan A and Turner his Plan B, I'm cool with that. Parsons is a better shooter/scorer, but Turner is a better distributor/defender. 3 and D guys are all the rage these days, and many of them vastly overpaid. Turner lacks the 3-point shooting to fit the latest buzzword, he's more of a create and D guy, which given we have so many other 3-point shooters may end up being a better fit for this team than a guy like Parsons.

BNM

 

[Boob-No-More]

cool thread, fun to see him stomp on paul george in every stat

I wouldn't get too excited about that. Paul George is still a better defender and a MUCH better offensive player. He's also a max player and wasn't available.

As I stated in a previous post, these play type stats suffer from small sample sizes. In some play types, even a full season's worth of data produces a very small sample size. Because of this, they are also dependent on position, role, defensive assignment and quality of opponent.

For example, Paul George is a bigger player. Basketball-reference still lists him as a SF, but he played 39% of his minutes last season at PF (in the playoffs, he played 67% of his minutes at PF). So, that means, for example, when IND plays CLE, Paul George is tasked with guarding LeBron James. Not so for Evan Turner, he ends up guarding a smaller, perimeter player, like J.R. Smith (in this case). Obviously, LeBron James is a MUCH better offensive player than J.R. Smith, a one-dimensional, streaky shooter.

So, Paul George could actually do a good job on LeBron and hold him below his season scoring average, but still get "lit up" for 25 points. Because Smith runs hot and cold, he could easily end up going 2-9 from 3-point range for 6 points. Was that Turner's defense, or Smith's streaky shooting? Hard to say, but at the end of the day, it makes Turner look like a defensive stud, whether he deserves the credit or not. And, because the sample sizes are so small, one exceptionally cold shooting night from a volume shooter like Smith can have a statistically significant impact on Turner's play type defensive numbers.

Point being, don't read TOO much into these stats. As I said earlier, they are somewhat useful, and certainly interesting, for players who play similar minutes and similar roles. But, you still need to look at other advanced stats, defensive assignment, quality of opponent, use the eye test, and most of all, common sense. I'm thrilled we got Evan Turner, he's a good defender and a good distributor - TWO things this team needed. Still, I'd take Paul George in a heartbeat, for his defense alone, over Evan Turner. And, I think pretty much every GM in the league would too, if given that option.

Edit: Here's a few more. more established (but not necessarily more "accurate" defensive metrics comparing George and Turner):

DRtg:
Paul George:
2015-16 = 101
Career = 99
Evan Turner:
2015-16 = 104
Career = 105

DWS:
Paul George:
2015-16 = 4.8
Career = 22.8
Evan Turner:
2015-16 = 3.0
Career = 14.8

DBPM:
Paul George:
2015-16 = 1.0
Career = 1.8
Evan Turner:
2015-16 = 1.3
Career = 1.5

STL/G:
Paul George:
2015-16 = 1.9
Career = 1.7
Evan Turner:
2015-16 = 1.0
Career = 0.8

STL%
Paul George:
2015-16 = 2.7
Career = 2.7
Evan Turner:
2015-16 = 1.7
Career = 1.5

Opponent PER (from 82games.com):
Paul George:
2015-16 = 14.2
2014-15 = 6.1 (small sample size, only played 6 games due to injury)
2013-14 = 12.3
2012-13 = 11.2
Evan Turner:
2015-16 = 12.2
2014-15 = 14.4
2013-14 = 15.8 IND, 15.0 PHI
2012-13 = 15.4

George has been 1st or 2nd team All Defense three times, Turner zero. George is regularly among the top 10 in steals, DRtg, and Defensive Win Shares. Turner has never finished in the top 20 in any of these, or any other, defensive category.

The good news is Turner's defense has been steadily improving. George's defense hasn't quite returned to the elite level he was at prior to his injury, but he is still a top 20 defender in the league.

BNM

 

[Natebishop3]

It is now obvious that Olshey was trying to make the roster a better one defensively, without hurting the offense.

Seems obvious to me that Neil sees the writing on the wall and is building a highly versatile, long, athletic, defensive team to stop (or at least slow down) the Warriors.

Turner
Crabbe
Harkless
Aminu

All wings that can play multiple positions and defend.

 

[Boob-No-More]

Seems obvious to me that Neil sees the writing on the wall and is building a highly versatile, long, athletic, defensive team to stop (or at least slow down) the Warriors.

Turner
Crabbe
Harkless
Aminu

All wings that can play multiple positions and defend.

I don't think it's just the Warriors, I think it's a good plan, in general. Traditional positions are being blurred a replaced by roles and skill sets. Such is the world of analytics. Having versatile players with complementary skill sets gives the coach flexibility for different opponents and different situations.

BNM