Breaking down the Boards

Need to talk more about this but for now I will say this:

  • Broke down all the Play by Play for 2009-2010 and my numbers where very close to right (and now it’s even closer).
  • Oh Shiny!

Oh and here’s required reading for all aspiring stat geeks.

Image courtesy of



  1. Some Dude
    April 7

    Did you mean for the last column to be ADJUSTED and not TOTAL? Also, what is being adjusted?

    BTW, have you been paying attention to Bynum lately? His rebounding numbers post all-star break have gone up significantly. Did you notice Pau and Lamar’s have gone down significantly as well, even adjusting for minutes changes.

    Since you have the data, what are the pre all-star break rebound %s and post all-star break prcentages for the 3 players. I think this is a great peek at a rebound’s worth.

  2. Some Dude
    April 7

    Why is Blake Griffin listed as a C? He gets very few minutes there compared to PF.

  3. April 7

    I’m also curious what’s being adjusted, and I noticed the last two columns share the same name.

    Awesome work, Arturo. I’m anxious for some explanation and commentary on it!

    • brgulker,
      The adjusted column assumes the same number of opportunities on Offense and Defense. The last two columns are rank for the total and the adjusted total.

      • Some Dude
        April 7

        ah, okay. i get it now.

        But why is Griffin listed as a center?

  4. Fred Bush
    April 7

    The Celtic bigs should get asterisks since they’re playing a different game than the rest of the league when it comes to offensive rebounding.

  5. April 7

    I’m sure the simple position rating leaves a lot to be desired – Linas Kleiza is not a SG. He’s a SF who even occasionally played PF. Similarly, DeRozan was playing some SF earlier in the year, but ever since the James Johnson trade he’s been almost exclusively playing SG.

    Bayless at #1 for PGs? I can’t say that I would’ve thought that just from watching games (and that’s why we look at the stats).

    I knew you’d have that new xkcd comic up here.

  6. Some Dude
    April 7

    Players adjusted up play for terrible defense or terrible offense. Adjusted down for the opposite. Makes sense since I said as much last post.

    That’s why I question the adjustment as being meaningful. Love’s goes up, but to get equal rebounds he has to play better defense. if he did, does he sacrifice rebounds?

    Anyway, one interesting thing I saw is Kevin Garnett’s disparity in Def and off rebounding. I never noticed it before, but the picture made it stick out like a sore thumb. That is an amazingly large discrepency. I wonder if it’s the most ever.

    It must be a product of the offensive system they run. I’d imagine Doc would be better served by occasionally sending KG to the paint off the ball. Arturo watches the Cs a lot more than me, so I wonder if he thinks it’s the system or something else going on that’s casing such a huge gap. Or any other Cs fan here. 25% to 4%. wowzers.

    • They want Rondo to get the Off Reb and start the break. They also suck at off reb (and it really isn’t the defensive system because Chicago runs the same system and is deadly on the boards).

      • Some Dude
        April 8

        huh? Why would Rondo get an Off rebound and create a break? He’s already on his half!

        What do offensive rebounds and defensive system have to do with each other?

        • SD,
          Was in a hurry and read that wrong. They have no presence in the middle and no ball hawks other than Rondo. KG,Ray and Pierce are typically at least 10 feet from the basket or diving to the bucket. Actually this is why with Shaq they’re better because he parks in the paint and the Celts system is designed to get him easy boards (Baby cannot take advantage of that(.

          • Some Dude
            April 8

            So it’s an offensive scheme issue in that KG is not near the paint on offense, not that he can’t rebound on offense, anymore?

            Cuz KG is killing it on the defensive rebounds.

            • SD,
              Watch em. KG is never close to the hoop on offense and he’s looking to get back on D. They generally clear space in the post. The Bulls are similar but with the marked difference that Noah and Boozer play much closer to to the hoop and are looking to get the rebound. IMHO, this is what will make Chicago a nightmare come playoff time. It’s the same thing that makes the Lakers so evil. Second chance points and fouls in the paint.

              • Some Dude
                April 8

                I do watch em, but not nearly as much as you!

                I don’t get why Doc doesn’t send KG to the glass more often. When the were shooting lights out it was one thing, but not they’re not and he’s not adjusting the offense.

                Perk for Green is looking worse by the day.

  7. Guy
    April 9

    Hey Arturo. I’m checking out your site for the first time in a few months, and these last three posts related to rebounding caught my eye. At first glance, your data seems to show that the spread in team rebounding performance is rather modest. The very best teams are +2% (52%), while the very worst are -2%. Rebounding matters, of course — 2% is worth about 1.6 points, or +/- 4 wins per season — but it appears to be dwarfed by the differences in team shooting efficiency and team defense. Is that now your view as well? (Not trying to reopen old arguments, just curious as to where you stand today.)

    In contrast, the differences in player rebounding percentages you show are far, far greater than what we see for teams. There are quite a few players who grab 3% or 4% more rebounds than their position average (Love is far higher). And yet no single team comes close to being that far above or below average (even the benighted Bargnani can’t drag his team below average). As you might guess, that suggests to me a very large diminishing returns effect is in effect. Otherwise, surely at least a few teams would combine one of these great +4% rebounders with average-rebounding teammates, and end up as a +3% team overall. But assuming you disagree with that interpretation, I wonder if you have a theory as to why team rebounding clusters so closely together, despite players’ ostensibly having such radically different rebounding abilities?

    • Guy,
      Good to hear from you.

      This will probably find it’s way into a future post:
      Standard Dev %
      “Offensive Success Rate
      Pts per 100 Plays Offense” 3.34 74.9%
      “Defensive Success Rate
      Pts per 100 Plays Defense” 2.84 63.5%
      “Rebounding Edge
      Pt. Margin from Incremental Play Generation”
      1.24 27.8%
      Point Margin @ 200 Possesions
      So rebounding does represent a significant impact on wins but it’s only part of the equation. The regression models are assuming that some of the defensive value is colinear with rebounding prowess(and this in general is a very fair assumption). You also need offensive efficiency (at a decent usage level but you can find players to take shots, the trick is to scheme properly and not have too many low usage guys).

      What’s interesting is that you also need low usage, high win producing per minute guys to pair with your high usage high efficiency players.

      Chicago pairs Rose (high usage, decent not great value) with lots of low usage high win producing guys (Noah,Boozer,Deng). Lakers do it with Kobe, Gasol, Bynum and Odom.

      The Heat have too many guys who need possesions to generate value (they really miss Haslem). They need someone to be Scottie/Rodman.

      The Celts I’m not bringing up.

Leave a Reply

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