Announcement

Collapse
No announcement yet.

Attention Liston!

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Attention Liston!

    Wanna give me a hand here?

    http://thenbehteam.blogspot.com/2011...on-review.html

    See comments.

    Would love to get someone involved in this discussion that knows more about stats than me.

  • #2
    Why do you need help? You made your point.

    Everyone agrees that rebounds contribute to winning, but to know how much any "rebounder" contributed to winning, we need to know more than how many rebounds he got and we need to know more than how many rebounds he got relative to players on other teams.. What we really need to know is how many rebounds his own team got that it would not otherwise have gotten if he weren't there. Do I have a clue how to figure that out? Absolutely not. But, I am convinced that that is the key question and that it has not been answered by any system I know of (although maybe the stats guys on here will say different).

    I remember ages ago when Ben Wallace left Detroit and everyone was bemoaning how it was going to hurt their defence and rebounding. It turned out that the only real effect it had on the team was that they gave up more free throws. In all of the other indicators: efg%, rebounds, blocks, etc. the deltas were negligible. Turned out his best skill was keeping opponents off the free throw line, which obviously helped the team defensively but has nothing to do with rebounds even though Wallace was considered one of, if not the, premiere rebounder in the game.

    Comment


    • #3
      Quirk wrote: View Post
      Wanna give me a hand here?

      http://thenbehteam.blogspot.com/2011...on-review.html

      See comments.

      Would love to get someone involved in this discussion that knows more about stats than me.
      Is this your blog? I like the name of the site. Funny.

      Comment


      • #4
        Apollo, no it's not my blog, I'm just making comments.
        Last edited by Quirk; Sat Apr 30, 2011, 04:41 PM.

        Comment


        • #5
          slaw wrote: View Post
          What we really need to know is how many rebounds his own team got that it would not otherwise have gotten if he weren't there.
          Right, thanks to Mary/ilikeflowers joining the thread, it made it much better. I'm amazed at how smug and defensive the WP community is, including the operator of that Blog.

          Personally, I think WP and PAWS are very interesting indicators, but only for comparing well-rounded, multi-faceted players to each other, they obviously significantly misvalue one-dimensional players. This is plain to even somebody like me who doesn't know very much about stats or basketball. I have explained why I think this is in the thread above, and it's been confirmed by Mary/ilikeflowers who seems, unlike me, to actually know something about the subject.

          The Raptors have two extremes in Bargnani and Evans as examples of players that WP is not likely to value effectively.

          Both of the players have serious flaws in their game, yet WP adherents seem to actually believe that Reggie Evans is a Hall of Fame caliber player, while Bargnani is the worst player to ever play in the NBA. Not only this, but instead of focusing on what WP does well, and cautioning against literal beliefs in such extreme and untenable results, they seem to talk about nothing else but Evans and Bargnani, seeming to have accepted WP as not only a useful indicator that makes best use of imperfect information in the form of boxscore stats, but infallible gospel that *prove* that Evans //really is// a timeless legend and Bargnani //really is// a bust on historic proportions.

          Focusing on players that are plainly outliers in their metric, and therefore, not really good examples of it''s effectiveness does nothing to enhance the credibility of their metric, since they are constantly trumpeting it's less believable conclusions, but they seem to care little about this, and when asked, respond with dismissive evasions.

          But the cold hard truth is that boxscore stats do not provide efficiency stats for rebounds, (and rebounds are highly weighted in WP) and this means that any attribution of wins based on rebounds at the individual level is highly suspect, and should be treated as such.

          I still would like Liston's take. And Ed Kupfer's if he's lurking about somewhere.
          Last edited by Quirk; Mon May 2, 2011, 06:01 AM.

          Comment


          • #6
            Nice post Quirk.

            I think as long as people actually know what these advanced stats are telling them, then they mean something and they can be used to convey legitimate points. But like you said, if you look soley at these numbers, and knew nothing about the NBA, you'd be thinking Reggin is an All-Star and Bargs is lucky to get a minute in Garbage time.

            But if you put these beside the Adjusted +/-, and further advanced stats, it's like adding pages to a book. You don't get the full story until you are given ALL of the information.
            That being said, Evans //really is// a timeless legend. hahaha

            Comment


            • #7
              Quirk wrote: View Post
              Wanna give me a hand here?

              http://thenbehteam.blogspot.com/2011...on-review.html

              See comments.

              Would love to get someone involved in this discussion that knows more about stats than me.
              I've been working on something, but it'll take some time.

              The fatal flaw in their argument is that big rebounders like Evans are often not productive on offense. Their counter is that he's smart not hurting his team by taking less efficient shots. However, that doesn't mean the "big rebounder" doesn't impact their teammates' effectiveness. For example, if Evans gets the ball up high then the defense has a very easy time doubling the Raptors' best player - because its not a good option for Evans to shoot. And WP will encourage him not to.

              This is a classic snapshot I captured - its not uncommon: http://liston.ca/Reggie.jpg

              Ultimately, this means the Raptors are playing 4 on 5. Obviously, Evans has a huge impact on offensive rebounding and this counters some of his challenges.... but, the team's best players will HAVE to shoot the ball in more difficult situations - often double teamed and/or late in the clock (when the ball moves through Evans everyone knows he'll be passing and thus time off the clock before a good shot can be put up).

              In other words, pure rebounders can have a negative influence on their offense and thus a negative impact on their teammates WP scores. I.e. SOMEONE HAS TO SHOOT THE DAMN BALL. So, to maximize your WP score, only rebound and don't take any shots - even late in the clock - make your teammate take difficult shots so you don't hurt your score.

              So teams without many scoring options will see their primary scorers suffer with WP metrics.

              I also have some (smallish) issues with position adjustments as well as team adjustments (defense), but that's for another day.

              WP is a good metric. But the attitude by their proponents is surprising. There are so obvious flaws (as there is with all metrics!!), but they continue to try to insult anyone that points them out. And that works against its popularity. Its too bad, there some good in it.
              Last edited by Liston; Mon May 2, 2011, 03:46 PM.
              http://twitter.com/Liston

              Comment


              • #8
                Liston wrote: View Post
                This is a classic snapshot I captured - its not uncommon: http://liston.ca/Reggie.jpg
                Thanks Tom, yeah I mention the 4/5 thing in the comments on "NBeh?" as well, not for the first time, dismissive evasion as usual.

                They simply say it's common for certain players to play a defensive role and leave the scoring to others, which simply evades the question as to how this affects the given player's productivity.

                It would be funny to argue in response that yes, Evans is not a liability on offense because he takes so few shots, just like Bargnani us not a liability on the boards because he attempts so few rebounds! They'd probably just miss the humour though.

                They can't dismiss the simple fact that there is no efficiency stats for rebounding in the boxscore though, and thus their is simply no way to attribute individual rebound totals to team wins. A major problem since WP weighs rebounds highly in their metric.

                Look forward to that something you're working on!

                Comment


                • #9
                  Quirk wrote: View Post
                  They simply say it's common for certain players to play a defensive role and leave the scoring to others, which simply evades the question as to how this affects the given player's productivity.
                  Par for the course.

                  When you've got a hammer, everything looks like a nail. Berri's hammer is regression analysis, and he goes about hitting everything he can find with it until he finds something that seems vaguely nail-like from a certain angle. Then he proclaims a group of extremely well-paid subject matter experts dumb. When challenged about this, he says things like "regressions are nice, but not always understood by everyone." He calls the protestors dumb.

                  The main problem I have with Berri and his true believers is that, at the very least, the questions Berri attempts to tackle with really complicated regressions are murky things best delivered with a dose of humility. It's obvious with every response to another Berri study that declares someone dumb that different views on the data produce different results. Berri's overarching thesis is that subject matter experts make huge errors because they refuse to look at data from all possible angles. Stuck in their ruts, they robotically bang out decisions like their forefathers. Statistician, heal thyself.

                  Comment


                  • #10
                    Great post Slaw! "Statistician, heal thyself."

                    Comment

                    Working...
                    X