Loading ...
Sorry, an error occurred while loading the content.

[APBR_analysis] And while I'm at it...

Expand Messages
  • igorkupfer@rogers.com
    [A veritable posting explosion from your usually lurking author...] Speaking of goofy ratings... Whenever I look at a boxscore, I ve always subtracted misses
    Message 1 of 5 , Feb 4, 2003
    • 0 Attachment
      [A veritable posting explosion from your usually lurking author...]

      Speaking of goofy ratings...

      Whenever I look at a boxscore, I've always subtracted misses from points. It just seems
      like an easy way to mentally calculate scoring efficiency. Recently, I decided to
      formalise the calculations, and give it a little stronger theoretical basis. This is
      what I came up with:

      Efficiency Points = pts + 1*(2pointers made) - 1*(2pointers missed)
      + 2*(3pointers made) - 1*(3pointers missed)
      + 0.5*(free throws made) - 0.5*(FT missed)
      - 1*(turnovers) + 1*(offensive rebounds)

      The assumption is this: teams generally average pretty close to 1 point per possession.
      That means that whenever a player ends a possession without scoring, he "costs" his
      team 1 point -- the point his team could normally expect. A missed 3 pointer ends a
      possession with zero points -- as does a missed 2 -- so that player has 1 EffPT
      subtracted from his total. Similarly, a turnover ends a possession with zero points,
      and the player will be docked 1 EffPT for each one.

      On the other hand, a made 2 pointer gives a team 1 point more the
      expected-one-point-per-possession. That extra point gets added to that players total. A
      trey is 2 points more than the expected.

      Offensive rebounds take a zero-point possession, and turn it into a normal
      expected-one-point-possession. One EffPT goes to that player.

      The free throws get a 0.5 coefficient, for purely arbitrary reasons. I can't figure out
      how much they are actually worth, or even conceive of a way to calculate it.

      Assists are not included because there's no easy way to deduct the EffPTS from the
      assistee. This, I think, limits the usefulness of the rating.

      That's all there is to it. As an example, here's the box of tonight's Rap-Bucks game.
      EffPts are in parentheses.

      Bucks 95, Raptors 98
      2/4/2003 Bradley Center, Milwaukee, WI

      Raptors

      PLAYER POS MIN FGM-A 3GM-A FTM-A OFF AST TO PTS EffPTS
      VINCE CARTER G 34 11-24 1-3 2-2 2 4 0 25 [27]
      ALVIN WILLIAMS G 34 5-12 1-2 0-2 0 2 2 11 [7]
      JEROME WILLIAMS F 39 5-7 0-0 3-6 4 1 1 13 [19]
      MORRIS PETERSON F 27 6-9 1-2 1-3 2 1 1 14 [19]
      ANTONIO DAVIS C 44 7-13 0-0 2-2 7 3 3 16 [22]
      Voshon Lenard 35 3-12 1-3 6-7 2 2 2 13 [11]
      Rafer Alston 14 2-5 0-2 0-0 0 3 2 4 [1]
      Michael Bradley 13 1-1 0-0 0-0 1 1 2 2 [2]

      TOTAL 240 40-83 4-12 14-22 18 17 13 98 [107]


      Bucks

      PLAYER POS MIN FGM-A 3GM-A FTM-A OFF AST TO PTS
      SAM CASSELL G 39 8-14 1-1 3-4 2 7 2 20 [24]
      RAY ALLEN G 35 9-17 3-7 0-0 1 3 2 21 [24]
      ANTHONY MASON F 36 5-13 0-0 3-4 1 7 2 13 [10]
      TIM THOMAS F 21 1-6 0-1 0-0 0 0 3 2 [-5]
      ERVIN JOHNSON C 25 0-2 0-0 0-0 6 0 2 0 [2]
      Michael Redd 30 10-14 2-5 2-2 1 0 1 24 [33]
      Toni Kukoc 20 2-5 1-3 2-2 0 2 2 7 [6]
      Kevin Ollie 18 0-4 0-0 0-0 0 0 0 0 [-4]
      Jason Caffey 16 4-6 0-0 0-0 2 1 1 8 [11]

      TOTAL 240 39-81 7-17 10-12 13 20 15 95 [101]

      Note that EffPTS don't add up exactly to actual points. I've run this test on about a
      dozen games so far, and while they never seem to add up exactly, EffPTS is almost
      always within 10 points of the actual total, usually overestimating. Considering the
      shaky basis of my assumptions, I think that's pretty good -- although I'd like to get
      it closer. The one rule I gave myself: I'm not allowed to change any of the weights,
      except for free throws. I still have no idea how much they are worth with regards to
      possessions. Anybody have any ideas?


      ed
    • Dean Oliver <deano@rawbw.com>
      Ed -- Take a look back at Message 14. That message discusses some of the other methods for weighting stats. JohnH of this group wrote a book with different
      Message 2 of 5 , Feb 5, 2003
      • 0 Attachment
        Ed --

        Take a look back at Message 14. That message discusses some of the
        other methods for weighting stats. JohnH of this group wrote a book
        with different weights. There is an economist named Dave Berri who
        also has something effectively similar. There are a few more on the
        web, too. I've developed a matrix showing the relative weights that
        the developers put on the different stats. I can't put that matrix
        out here since it's in the book I've got coming out.

        Yours is slightly different, of course, because you don't include any
        defensive stats, making it purely an offensive tool. That makes it
        somewhat unique.

        DeanO


        --- In APBR_analysis@yahoogroups.com, <igorkupfer@r...> wrote:
        > [A veritable posting explosion from your usually lurking author...]
        >
        > Speaking of goofy ratings...
        >
        > Whenever I look at a boxscore, I've always subtracted misses from
        points. It just seems
        > like an easy way to mentally calculate scoring efficiency.
        Recently, I decided to
        > formalise the calculations, and give it a little stronger
        theoretical basis. This is
        > what I came up with:
        >
        > Efficiency Points = pts + 1*(2pointers made) - 1*(2pointers missed)
        > + 2*(3pointers made) - 1*(3pointers missed)
        > + 0.5*(free throws made) - 0.5*(FT missed)
        > - 1*(turnovers) + 1*(offensive rebounds)
        >
        > The assumption is this: teams generally average pretty close to 1
        point per possession.
        > That means that whenever a player ends a possession without
        scoring, he "costs" his
        > team 1 point -- the point his team could normally expect. A missed
        3 pointer ends a
        > possession with zero points -- as does a missed 2 -- so that player
        has 1 EffPT
        > subtracted from his total. Similarly, a turnover ends a possession
        with zero points,
        > and the player will be docked 1 EffPT for each one.
        >
        > On the other hand, a made 2 pointer gives a team 1 point more the
        > expected-one-point-per-possession. That extra point gets added to
        that players total. A
        > trey is 2 points more than the expected.
        >
        > Offensive rebounds take a zero-point possession, and turn it into a
        normal
        > expected-one-point-possession. One EffPT goes to that player.
        >
        > The free throws get a 0.5 coefficient, for purely arbitrary
        reasons. I can't figure out
        > how much they are actually worth, or even conceive of a way to
        calculate it.
        >
        > Assists are not included because there's no easy way to deduct the
        EffPTS from the
        > assistee. This, I think, limits the usefulness of the rating.
        >
        > That's all there is to it. As an example, here's the box of
        tonight's Rap-Bucks game.
        > EffPts are in parentheses.
        >
        > Bucks 95, Raptors 98
        > 2/4/2003 Bradley Center, Milwaukee, WI
        >
        > Raptors
        >
        > PLAYER POS MIN FGM-A 3GM-A FTM-A OFF AST TO PTS
        EffPTS
        > VINCE CARTER G 34 11-24 1-3 2-2 2 4 0
        25 [27]
        > ALVIN WILLIAMS G 34 5-12 1-2 0-2 0 2 2
        11 [7]
        > JEROME WILLIAMS F 39 5-7 0-0 3-6 4 1 1
        13 [19]
        > MORRIS PETERSON F 27 6-9 1-2 1-3 2 1 1
        14 [19]
        > ANTONIO DAVIS C 44 7-13 0-0 2-2 7 3 3
        16 [22]
        > Voshon Lenard 35 3-12 1-3 6-7 2 2 2
        13 [11]
        > Rafer Alston 14 2-5 0-2 0-0 0 3 2
        4 [1]
        > Michael Bradley 13 1-1 0-0 0-0 1 1 2
        2 [2]
        >
        > TOTAL 240 40-83 4-12 14-22 18 17 13
        98 [107]
        >
        >
        > Bucks
        >
        > PLAYER POS MIN FGM-A 3GM-A FTM-A OFF AST TO PTS
        > SAM CASSELL G 39 8-14 1-1 3-4 2 7 2
        20 [24]
        > RAY ALLEN G 35 9-17 3-7 0-0 1 3 2
        21 [24]
        > ANTHONY MASON F 36 5-13 0-0 3-4 1 7 2
        13 [10]
        > TIM THOMAS F 21 1-6 0-1 0-0 0 0 3
        2 [-5]
        > ERVIN JOHNSON C 25 0-2 0-0 0-0 6 0 2
        0 [2]
        > Michael Redd 30 10-14 2-5 2-2 1 0 1
        24 [33]
        > Toni Kukoc 20 2-5 1-3 2-2 0 2 2
        7 [6]
        > Kevin Ollie 18 0-4 0-0 0-0 0 0 0
        0 [-4]
        > Jason Caffey 16 4-6 0-0 0-0 2 1 1
        8 [11]
        >
        > TOTAL 240 39-81 7-17 10-12 13 20 15
        95 [101]
        >
        > Note that EffPTS don't add up exactly to actual points. I've run
        this test on about a
        > dozen games so far, and while they never seem to add up exactly,
        EffPTS is almost
        > always within 10 points of the actual total, usually
        overestimating. Considering the
        > shaky basis of my assumptions, I think that's pretty good --
        although I'd like to get
        > it closer. The one rule I gave myself: I'm not allowed to change
        any of the weights,
        > except for free throws. I still have no idea how much they are
        worth with regards to
        > possessions. Anybody have any ideas?
        >
        >
        > ed
      • igorkupfer@rogers.com
        ... From: To: Sent: Wednesday, February 05, 2003 10:29 AM Subject: [APBR_analysis] Re: And while I m at it...
        Message 3 of 5 , Feb 5, 2003
        • 0 Attachment
          ----- Original Message -----
          From: <deano@...>
          To: <APBR_analysis@yahoogroups.com>
          Sent: Wednesday, February 05, 2003 10:29 AM
          Subject: [APBR_analysis] Re: And while I'm at it...


          > Ed --
          >
          > Take a look back at Message 14. That message discusses some of the
          > other methods for weighting stats. JohnH of this group wrote a book
          > with different weights. There is an economist named Dave Berri who
          > also has something effectively similar. There are a few more on the
          > web, too. I've developed a matrix showing the relative weights that
          > the developers put on the different stats. I can't put that matrix
          > out here since it's in the book I've got coming out.
          >

          Hmm, mostly 0.5 - 1.0, which is what I guessed.

          > Yours is slightly different, of course, because you don't include any
          > defensive stats, making it purely an offensive tool. That makes it
          > somewhat unique.
          >

          I focus only on offense (and only part of the offense, for that matter) because I'm
          incapable of developing an all-encompassing system. I prefer to let you guys slug out a
          more comprehensive model -- I've seen how difficult it is. I'll focus on the small
          stuff.

          I think what makes it different than the others is it's theoretical basis. Most of the
          methods you describe are linear weights-type systems, the
          add-the-good-subtract-the-bad-to-arrive-at-a-rating models. Mine is a little different.
          (First of all, I wanted it to look like a player's points total for a game, so you
          would have an easy way of grasping what the rating means.) I assume those other ratings
          used some kind of regression to arrive at weights that would produce results that jibe
          with a qualitative assessment of player ability. I don't do that (I share your distaste
          with linear weights). I only add and subtract stuff with regards to a teams expected
          points per possession: Teams can expect 1 point per possession. If a player misses a
          shot, he cost the team 1 point. If he turns the ball over, he cost the team 1 point.

          Like I said, it's a pretty shaky basis on which to construct a rating system -- but it
          seems to produce half decent results. I'd like it to be more accurate, but I can't seem
          to do that without changing weights. I'm thinking that there may be a conceptual
          problem with my approach that I'm overlooking. (Here's one: since missed shots usually
          result in an offensive board 30% of the time, they can't possibly be worth -1 with
          regards to a possession. But why does -1 give good results? I don't know.)
        • dlirag <dlirag@hotmail.com>
          ... rating system -- but it ... accurate, but I can t seem ... a conceptual ... missed shots usually ... be worth -1 with ... know.) Would -0.7 per missed FGA
          Message 4 of 5 , Feb 15, 2003
          • 0 Attachment
            --- In APBR_analysis@yahoogroups.com, <igorkupfer@r...> wrote:

            > Like I said, it's a pretty shaky basis on which to construct a
            rating system -- but it
            > seems to produce half decent results. I'd like it to be more
            accurate, but I can't seem
            > to do that without changing weights. I'm thinking that there may be
            a conceptual
            > problem with my approach that I'm overlooking. (Here's one: since
            missed shots usually
            > result in an offensive board 30% of the time, they can't possibly
            be worth -1 with
            > regards to a possession. But why does -1 give good results? I don't
            know.)

            Would -0.7 per missed FGA give better results?
          • Dean Oliver <deano@rawbw.com>
            ... conceptual ... don t ... There s not a lot of solid guidance on what the weights should be. If all you want are better results -- better is a personal
            Message 5 of 5 , Feb 18, 2003
            • 0 Attachment
              --- In APBR_analysis@yahoogroups.com, "dlirag <dlirag@h...>" > a
              conceptual
              > > problem with my approach that I'm overlooking. (Here's one: since
              > missed shots usually
              > > result in an offensive board 30% of the time, they can't possibly
              > be worth -1 with
              > > regards to a possession. But why does -1 give good results? I
              don't
              > know.)
              >
              > Would -0.7 per missed FGA give better results?

              There's not a lot of solid guidance on what the weights should be.
              If all you want are better results -- "better" is a personal
              decision, so you can try it and see whether the results
              are "better". Whether you also account for some teams rebounding
              only 20% of their misses and other teams rebounding close to 40% is a
              question. And if you modify the weight on missed FGA, should you
              also modify the weight on offensive rebounds?

              DeanO
            Your message has been successfully submitted and would be delivered to recipients shortly.