DeMarre Carroll’s War on Possessions


In today’s basketball analytics paradigm, all of the talk is on efficiency, and for good reason: efficiency is highly correlated with winning. Per possession statistics like Offensive and Defensive Rating, and Synergy’s PPP rule the roost, and when they don’t, per shot statistics like eFG% or TS% step forward. Again, this makes sense: being efficient on both ends of the floor is obviously a good thing.

But when a player dramatically alters the possessions themselves, sometimes we analysts overlook that influence. Enter DeMarre Carroll. Last season, he shot an average-looking 46% from the field (though that was by far the best percentage of his career), and just 28.6% from 3: about average to below average overall for his position. Most look at these numbers and pigeonhole Carroll right there: wing players who can’t shoot tend not to be productive players.

DeMarre, though, is incredibly productive, by uniquely influencing the possession. Let’s look at how he does this:


DeMarre Carroll led the Jazz last season in turnovers per 36 minutes, allowing just 1.1 in that timeframe. To give you a comparison, Gordon Hayward finished with 2.1 TO/36, Alec Burks with 2.3 TO/36, and catch-and-shooters Randy Foye and Marvin Williams ended up with 1.3 and 1.5 TO/36, respectively. Yes, he even beat out notoriously turnover-stingy Al Jefferson, who ended with 1.5 TO/36.

Carroll’s total leaves him 15th in the NBA overall amongst players with over 1000 minutes, with most of the players above him of the extreme catch-and-shoot variety (such as Steve Novak, Shane Battier, Kyle Korver, etc.). Given that estimates for the value of a turnover range from -.9 points to -1.5 points, Carroll may be giving up a point fewer per game than his counterparts in this category alone. But we’re just getting started.


I wish there were a way to make a player famous for a certain facet of their game, but alas, that generally happens through the vagaries of public opinion. DeMarre Carroll’s offensive rebounding at his position is absolutely world-class. Let’s begin with his per 36 numbers again: Carroll averages 2.8 offensive rebounds per 36 minutes, compared to just 0.9 for Gordon Hayward, 1.2 for Alec Burks, 1.1 for Marvin Williams, and 0.3 for Randy Foye. Pretty impressive, no?

But when you compare him to the league overall, Carroll really shines. DeMarre ranks first in the league (again, minimum 1000 minutes) amongst non-PF and non-Cs in offensive rebounds overall: his 2.8 ORB/36 total beating the nearest challenger, Dante Cunningham, by over 10%. Carroll is the only player amongst all PGs, SGs, and SFs to rank in the top 50. He garners 9.1% of the possible offensive rebounds when he’s in the game, another league-leading performance for his position.

This isn’t just a single season fluke either. He led the league for non-big offensive rebounding in 2011-2012.  In his second season, he played only 50 minutes. Even in his rookie season, 2009-2010, he led the league! We have very significant evidence that DeMarre Carroll is very significantly great at offensive rebounding.

The linear weights estimates (link above) for an offensive rebound seem to put the value at about 0.75 points, but I would argue that the value is even greater for Carroll: via Synergy, he averaged 1.31 PPP on his offensive rebounding scoring opportunities last season, good for 11th in the league. That improves significantly on the Jazz’s overall 1.09 points per possession on offensive rebound opportunities. This may just be wish casting, but lets put the value of an Carroll offensive rebound at 0.9 points. Doing the multiplication, he’s getting 1.5 to 2.3 more points per 36 minutes than his wing counterparts based on his offensive rebounding alone.


Carroll also leads the Jazz with 1.9 steals per 36 minutes. This is again higher than his playing time competition: Hayward has 1.0 STL/36, Burks 1.1 STL/36, Foye 1.1 STL/36, and Marvin Williams finished with 0.8 STL/36.

Here, too, DeMarre Carroll compares extremely well with the rest of the league. Carroll ranks second in the league in steals per 36 minutes for a SF, and third in the league for non-guards, only Corey Brewer and Andray Blatche (strangely) finished higher. Again, the Jazz have someone who is world-class at his position at an aspect of the game.

Much like with turnovers, linear weights estimates the value of a steal at about 0.9-1.5 points. Once again, Carroll gains a point on his competition.



So how does it all add up? In these three respects of the game, Carroll is helping the Jazz by somewhere between 2.7 and 4.8 points per 36 minutes. That is a massive improvement: that jump would put the Jazz somewhere between 5th and 10th in the league in scoring margin, up from 15th. In short, the Jazz would have almost certainly made the playoffs, and may have even had home court advantage in the first round.

The brilliant part about Carroll’s season last year, however, was that he was still an effective player even ignoring his war-on-possessions specialties. Carroll’s FG% (46%) was a full 5 percentage points higher than any other season in his career, and his 0.98 PPP on personal offensive possessions was 95th in the league out of 400-500 players. The Jazz offense improved when he was on the floor, even in shooting percentage (albeit by less than 1 percent).

His defense, which I heavily attacked last year, became average this season: metrics on his performance range from somewhat below average (his 0.93 PPP allowed in Synergy) to very good (a 4 point DRTG jump when he was on the floor, via, and his 12.6 PER allowed on Carroll is no longer limited to a specialist’s role.

The end result: a player who has largely eliminated his weaknesses, and improved his league-leading strengths. DeMarre Carroll’s unconventionality shouldn’t stop us from appreciating just how effective he was when on the floor last season, and a free agency offer representative of his talents should be extended.



  1. Misti brunelle

    Fascinating facts that enhances what I’ve said all season, why are his minutes wasted on Marvin Williams? If Corbin can read, send this to him!

  2. Rafael Amarante

    Awesome post, Andy. I had no idea DMC was THAT effective, especially at the offensive boards and stealing the ball.

  3. Dallan Forsyth

    Question: Demarre is notorious for the whole act like he is getting back on D and then running over to the person with the ball and taking it. But a lot of times he gets to it as the same time as the other person and takes it away from them. So if it is a missed shot and someone goes up for the rebound grabs it but before he comes down Demarre takes it from him what is that? Is that a steal? An offensive rebound? Both? What would the stat person mark down for that?

    Because technically it is not a rebound till the other player hits the floor. But if they both grab it in the air and he takes it away the only reason he got the rebound was because he stole it.

  4. Justin Gordon

    Brilliant piece, thanks! I’ve considered Carroll to be one of those “intangibles” type of players, somewhere between an energy guy and a momentum changer. Now I have a lot more to consider. However, yes indeed Carroll should be extended. Heck I didn’t even know this was a question.

  5. Devin Cash

    So what should that offer be? 3 years 10 mil? You don’t want to overpay your role players, but I don’t want to see Run DMC playing for anyone else next year.

  6. DJ

    This article was particularly eye opening. Thanks for posting the facts. I really hope he starts getting some more minutes. As Devin points out above, I would really like to know how much he’s worth and to make sure we don’t overpay him, though based on these stats, he seems a lot more valuable than Marvin Williams to me.

  7. Clint Johnson

    Great information, though I am a little more hesitant to buy into his value being as great as you suggest. DeMarre is predominantly an effort player, and I would argue that Per36 numbers are perhaps more misleading than normal when applied to such players. I think that the Jazz should absolutely retain him at a reasonable rate; I also believe he should be a regular rotation player earning somewhere between 14 to 20 minutes a game, or there about. That being said, I think that he would show diminishing returns if ever given too significant a role or too many minutes: his game would be particularly susceptible to fatigue, complacency, and scouting by an opponent, as well as his peak performance years being shorter than an average player of his value. (Think if Millsap had not developed his skills as he has; I see DeMarre as a similar player type.) He’s a bullpen pitcher, someone you use for 2-3 innings to change the nature of a game, not a starter you count on to win it for you.

    Of course, if DeMarre were to continue to develop his skill set significantly, my opinion on his value would change, of course. When he starts hitting 38% on three shots from range a night, get him in the game 36 minutes a game.

  8. David

    I loved DMC’s game before this article but now I have even more reason to love it! I think he could be a solid bench player for us. Heck, if we lose Big Al AND Millsap he could find himself starting at small forward over Marvin Williams. I’d prefer his hustle over Marvin Williams catch and shoot tendencies any day! I really do hope we keep him.

  9. Jon Midget (Yucca)

    DeMarre Carroll is exactly the kind of guys championship-caliber teams need.

    First, the team needs stars … high-profile players who do a lot of things really, really well. These, of course, are the focal point of the team. And they will represent the vast majority of the team salary.

    But after that, it needs cheap players who don’t need 20 shots but (a) have a set of specific skills that help teams win and (b) don’t do a lot of things that hurt teams’ ability to win.

    And that’s where DeMarre shines. You point out that there’s a bunch of things he does that is very helpful (in contrast to, say, Randy Foye). And at the same time there’s not much that he does that hurts. His defense is fine now. His shooting is fine (3P still poor, but improving). He passes fine.

    No, he’s not a star. But he’s a valuable player.

  10. Post
    Andy Larsen

    Thanks everyone, for the great feedback on this article. DeMarre seems underappreciated by the Jazz world, but it’s clear you guys think he has a bigger role. A few responses:

    Dallan: DeMarre gets credit for the steal if, in the scorer’s opinion, the opposing player had sole possession of the ball. He does not need to land first. If possession was still up for grabs, it’s an offensive rebound.

    Devin: I don’t think 3/$10 is crazy, but honestly I haven’t done the number crunching to stand by a valuation at this point.

    Clint: I think Millsap’s a good example: a guy who teams thought didn’t have enough talent, but showed off enough once he got the playing time. I guess that’s the question: could he keep it up for 36 minutes? I think yes, though largely just because of how he improved the rest of his game this year.

    Jon: I agree. Ideally, since he’s not a lockdown defender or great shooter, you’d have him on a team where he could play alongside one in a starting role, or play off the bench behind a SF in a 6th/7th man role. He’s good enough that you need to make playing time for him, but he may be more hurtful in the playoffs, where teams may lay off of him (see, Tayshaun Prince and Memphis). But, again, great role player.

Leave a Reply

Your email address will not be published. Required fields are marked *