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

Mcgrady and Orlando

Expand Messages
  • harlanzo
    I was just looking over the Orlando Magic from the last 3 years. Has anyone noticed that the net effect of adding Mcgrady has not been so impressive? Doc
    Message 1 of 14 , Feb 12, 2002
    • 0 Attachment
      I was just looking over the Orlando Magic from the last 3 years. Has
      anyone noticed that the net effect of adding Mcgrady has not been so
      impressive? Doc Rivers records during the last couple of years is as
      follows:

      year w-l starting lineup
      1999-00 41-41 Armstrong, Wahad/Mercer, Outlaw, Wallace, Amaechi
      2000-01 43-39 Armstrong, Mcgrady, M.Miller, Outlaw, Declerq/Doleac
      2001-02 25-24 Armstrong, Mcgrady, M.Miller, H.Grant, Hunter/Ewing

      Raw win and loss totals seem to indicate that the Mcgrady has not
      been a huge help for Magic. The eastern conference has not really
      improved over these three years. The only major difference in the
      team is Mcgrady for Wallace (and Atkins). But a look at point
      differential kind of shows the orlando may have improved a bit and
      that the won-loss record may start to improve a bit more.

      Pt Differential Per game
      1999-00 100.1-99.4 +0.7
      2000-01 97.5-96.5 +1.0
      2001-02 100.0-97.6 +2.4

      I am not suggesting that Mcgrady is only slightly more valuable than
      Ben Wallace, but the superficial numbers seem interesting. Indeed if
      we look at player evaluation formulas (please take these with the
      usual caveats about tendex and such):
      Tendex Points Per game value
      Wallace .611 +5.1
      Mcgrady .671 +11.2

      Numbers seem to confirm what we all believe, that Mcgrady is a good
      deal better than Wallace. But how do we account for the negligible
      difference in won-loss in ORlando?
    • Michael K. Tamada
      ... [...] ... It s an interesting conundrum; people with more up-close knowledge of the Magic than I might be able to offer an explanation. But quite possibly
      Message 2 of 14 , Feb 12, 2002
      • 0 Attachment
        On Wed, 13 Feb 2002, harlanzo wrote:

        > I was just looking over the Orlando Magic from the last 3 years. Has
        > anyone noticed that the net effect of adding Mcgrady has not been so
        > impressive? Doc Rivers records during the last couple of years is as

        [...]

        > Tendex Points Per game value
        > Wallace .611 +5.1
        > Mcgrady .671 +11.2
        >
        > Numbers seem to confirm what we all believe, that Mcgrady is a good
        > deal better than Wallace. But how do we account for the negligible
        > difference in won-loss in ORlando?


        It's an interesting conundrum; people with more up-close knowledge of the
        Magic than I might be able to offer an explanation. But quite possibly
        the true explanation is the simplest one: plain random error (including
        errors introduced by the limitations of our knowledge and models). Who
        could've predicted that the Mariners would win 116 games last year or that
        the Patriots would beat the Rams? Who would've predicted that the Magic
        would stay about the same (maybe some people did, I don't know).

        Granted, what you've got is 2 1/2 seasons of data, not just one. Although
        wasn't 1999-2000 regarded as an unusual year for Orlando, in which they
        did better than expected?

        Also interesting is: although some fans were predicting good results, I
        think very few people were predicting that Michael Jordan would be having
        as much impact on the Wizards as he has. Although we have to be careful
        not to give him all the credit, similar to the Jason Kidd situation with
        the Nets. But I think with Jordan we literally do have a case of a player
        making his teammates better (or at least to play better). It's not as if
        his stats are those of a player who would singlehandedly add 26 or so
        victories to his team's total. Moreover, my guess is that even in models
        which include superstars' ability to make their teammates better, Jordan's
        results would be beyond what such models could forecast. My guess is that
        much of his contribution has been mental and emotional, rather than in the
        statistical results he gets on the court, which have been very good but
        not ultra-spectacular.


        --MKT
      • HoopStudies
        ... of the ... possibly ... (including ... While a very interesting subject, I m currently not in a position to add to the discussion. Although ... they ...
        Message 3 of 14 , Feb 12, 2002
        • 0 Attachment
          --- In APBR_analysis@y..., "Michael K. Tamada" <tamada@o...> wrote:
          >
          > It's an interesting conundrum; people with more up-close knowledge
          of the
          > Magic than I might be able to offer an explanation. But quite
          possibly
          > the true explanation is the simplest one: plain random error
          (including
          > errors introduced by the limitations of our knowledge and models).

          While a very interesting subject, I'm currently not in a position to
          add to the discussion.

          Although
          > wasn't 1999-2000 regarded as an unusual year for Orlando, in which
          they
          > did better than expected?
          >

          Yes. Rivers won Coach of the Year. Supposedly, the Magic were hurt
          when they lost out on the Duncan Sweepstakes by not being prepared
          for having the hole in the middle. That problem lingers and their
          defense is what has not quite come around.

          > Also interesting is: although some fans were predicting good
          results, I
          > think very few people were predicting that Michael Jordan would be
          having
          > as much impact on the Wizards as he has. Although we have to be
          careful
          > not to give him all the credit, similar to the Jason Kidd situation
          with
          > the Nets. But I think with Jordan we literally do have a case of a
          player
          > making his teammates better (or at least to play better). It's not
          as if
          > his stats are those of a player who would singlehandedly add 26 or
          so
          > victories to his team's total.

          None of us here predicted such success and we took our stabs. I
          think we did make comments that the Wizards defense would get better,
          in part because of Jordan and in part because of Coach Collins, who
          has done that before. My general opinion has been that a lot of
          defensive improvements are often due to a coach's presence (e.g.

          http://www.rawbw.com/~deano/articles/aa091897.htm (end)
          http://www.rawbw.com/~deano/articles/aa020597.htm

          Washington's biggest improvement this year is defensive (though the
          offense is also better). That has to be due, in part, to Collins
          (and Haywood, JWhite).


          Moreover, my guess is that even in models
          > which include superstars' ability to make their teammates better,
          Jordan's
          > results would be beyond what such models could forecast. My guess
          is that
          > much of his contribution has been mental and emotional, rather than
          in the
          > statistical results he gets on the court, which have been very good
          but
          > not ultra-spectacular.

          Two relevant questions:

          If Jordan were replaced by Mitch Richmond, would Washington's defense
          be as good? How different would it be?

          If Collins were replaced by, say, Kevin Loughery, would Washington's
          defense be as good? How different would it be?

          (Ok, 4 questions and, yes, I know that answering questions with
          questions is obnoxious.)

          Dean Oliver
          Journal of Basketball Studies
        • alleyoop2
          I think there s two reasons behind the Magic not improving with McGrady. First, their centers forgot how to play. Doleac and Amaechi were a pretty good combo
          Message 4 of 14 , Feb 13, 2002
          • 0 Attachment
            I think there's two reasons behind the Magic not improving with
            McGrady. First, their centers forgot how to play. Doleac and Amaechi
            were a pretty good combo in 2000; since then I think even their own
            mothers would have to admit that they've sucked.

            Also, the bench detoriorated -- the 2000 team had guys like Atkins,
            Corey Maggette, Derek Strong, etc., and last year's team didn't have
            that.

            I still think this year's Magic team will end up with a much better
            record. They've already played the tough part of their schedule and
            outscored their opponents more than their W/L indicates.


            As for Jordan and the Wizards, there's a couple factors at work here
            beyond Mike being Mike:

            1) They were unusually unlucky in their win total last season; they
            only won 19 games but had the point differential of a 22-23 win team.

            2) They have been unusually lucky in their win total this season:
            They haven't even outscored their opponents yet they're four games
            over .500

            3) They haven't played the tough part of their schedule yet.




            --- In APBR_analysis@y..., "HoopStudies" <deano@r...> wrote:
            > --- In APBR_analysis@y..., "Michael K. Tamada" <tamada@o...> wrote:
            > >
            > > It's an interesting conundrum; people with more up-close
            knowledge
            > of the
            > > Magic than I might be able to offer an explanation. But quite
            > possibly
            > > the true explanation is the simplest one: plain random error
            > (including
            > > errors introduced by the limitations of our knowledge and
            models).
            >
            > While a very interesting subject, I'm currently not in a position
            to
            > add to the discussion.
            >
            > Although
            > > wasn't 1999-2000 regarded as an unusual year for Orlando, in
            which
            > they
            > > did better than expected?
            > >
            >
            > Yes. Rivers won Coach of the Year. Supposedly, the Magic were
            hurt
            > when they lost out on the Duncan Sweepstakes by not being prepared
            > for having the hole in the middle. That problem lingers and their
            > defense is what has not quite come around.
            >
            > > Also interesting is: although some fans were predicting good
            > results, I
            > > think very few people were predicting that Michael Jordan would
            be
            > having
            > > as much impact on the Wizards as he has. Although we have to be
            > careful
            > > not to give him all the credit, similar to the Jason Kidd
            situation
            > with
            > > the Nets. But I think with Jordan we literally do have a case of
            a
            > player
            > > making his teammates better (or at least to play better). It's
            not
            > as if
            > > his stats are those of a player who would singlehandedly add 26
            or
            > so
            > > victories to his team's total.
            >
            > None of us here predicted such success and we took our stabs. I
            > think we did make comments that the Wizards defense would get
            better,
            > in part because of Jordan and in part because of Coach Collins, who
            > has done that before. My general opinion has been that a lot of
            > defensive improvements are often due to a coach's presence (e.g.
            >
            > http://www.rawbw.com/~deano/articles/aa091897.htm (end)
            > http://www.rawbw.com/~deano/articles/aa020597.htm
            >
            > Washington's biggest improvement this year is defensive (though the
            > offense is also better). That has to be due, in part, to Collins
            > (and Haywood, JWhite).
            >
            >
            > Moreover, my guess is that even in models
            > > which include superstars' ability to make their teammates better,
            > Jordan's
            > > results would be beyond what such models could forecast. My
            guess
            > is that
            > > much of his contribution has been mental and emotional, rather
            than
            > in the
            > > statistical results he gets on the court, which have been very
            good
            > but
            > > not ultra-spectacular.
            >
            > Two relevant questions:
            >
            > If Jordan were replaced by Mitch Richmond, would Washington's
            defense
            > be as good? How different would it be?
            >
            > If Collins were replaced by, say, Kevin Loughery, would
            Washington's
            > defense be as good? How different would it be?
            >
            > (Ok, 4 questions and, yes, I know that answering questions with
            > questions is obnoxious.)
            >
            > Dean Oliver
            > Journal of Basketball Studies
          • harlanzo
            ... defense ... Washington s ... I think that MJ is part of the glue that holds Wiz s d together. I have little direct statistical of this, but it looks like
            Message 5 of 14 , Feb 13, 2002
            • 0 Attachment
              --- In APBR_analysis@y..., "HoopStudies" <deano@r...> wrote:
              > Two relevant questions:
              >
              > If Jordan were replaced by Mitch Richmond, would Washington's
              defense
              > be as good? How different would it be?
              >
              > If Collins were replaced by, say, Kevin Loughery, would
              Washington's
              > defense be as good? How different would it be?
              >
              > (Ok, 4 questions and, yes, I know that answering questions with
              > questions is obnoxious.)

              I think that MJ is part of the glue that holds Wiz's d together. I
              have little direct statistical of this, but it looks like MJ plays a
              hard-nosed d. On the other hand, the idea that defense is a majority
              coach-related seems to have merit. The change from Van GUndy to
              Chaney in NY is just the latest example.

              I think this begs the question: to what extent is defense a skill
              that a player has and to what extent is a system/frame of mind
              instilled by a coach? No one can block shots like Mutombo but is it
              really hard to find aggressive perimeter defenders or guys who will
              bang in the post?
            • Michael K. Tamada
              ... We ve certainly seen many examples of teams without much physical talent perform well as teams, and almost always stout team defense is involved -- Golden
              Message 6 of 14 , Feb 13, 2002
              • 0 Attachment
                On Wed, 13 Feb 2002, harlanzo wrote:

                > I think this begs the question: to what extent is defense a skill
                > that a player has and to what extent is a system/frame of mind
                > instilled by a coach? No one can block shots like Mutombo but is it
                > really hard to find aggressive perimeter defenders or guys who will
                > bang in the post?

                We've certainly seen many examples of teams without much physical talent
                perform well as teams, and almost always stout team defense is involved --
                Golden State 1975, many of Bobby Knight's Indiana U teams (when he did win
                championships, he usually did have superstar talent on his team, but he
                often did well with not particularly talented teams).

                Truly superstar defenders do require physical talents that are beyond
                those of most players (Russell, Wilt, Payton). But when we look at the
                good defensive teams vs the bad defensive teams, I'm guessing that talent
                accounts for maybe 1/3 of the difference, and the other 2/3 is the
                "system/frame of mind": how hard and how well the teams play team defense.

                But I think the even deeper and harder question is how much credit to give
                to the coach, vs how much is either random or due to historical accident.
                Collins vs Loughery, yeah there'll be a difference. But it's not
                automatic, I could imagine circumstances where Loughery would help a team
                more in that 2/3 area than Collins would: because Collins (as indeed
                almost all coaches, the great ones just take longer) seems to wear out his
                welcome relatively quickly, the players tune out (at least I'm presuming
                that's what happens), and the team though having the same coach and the
                same physical talent, no longer performs as well in that 2/3 area, and
                plays poorer team defense.

                So eventually the coach must go, even (or especially) if he's a Doug
                Collins.

                So even if most of that 2/3 of team defense is determined by what
                we call coaching, it's not clear that the coach is actually in control of
                that 2/3. Even Lenny Wilkens lost effectiveness first in Seattle and then
                in Cleveland and then in Atlanta, after early coaching success at all
                three places.

                More specifically, if that 2/3 area consists of "system/frame of mind,"
                even though the coach's system may stay the same, the frame of mind
                created by the coach will change. In ways beyond his control. So the
                identity of the coach matters, but that is not the same as saying that he
                controls events.


                --MKT
              • HoopStudies
                ... talent ... defense. ... Uh-oh. Golden Numbers. Don t everyone latch on to these! This is kind of like Baseball is 2/3 pitching. I understand the point
                Message 7 of 14 , Feb 13, 2002
                • 0 Attachment
                  --- In APBR_analysis@y..., "Michael K. Tamada" <tamada@o...> wrote:
                  > On Wed, 13 Feb 2002, harlanzo wrote:
                  > good defensive teams vs the bad defensive teams, I'm guessing that
                  talent
                  > accounts for maybe 1/3 of the difference, and the other 2/3 is the
                  > "system/frame of mind": how hard and how well the teams play team
                  defense.
                  >

                  Uh-oh. Golden Numbers. Don't everyone latch on to these! This is
                  kind of like Baseball is 2/3 pitching.

                  I understand the point and perhaps we should start our own basketball
                  proverbs, but it usually gets me in trouble.

                  Couple of examples where talent is apparently relatively more
                  important.

                  1. George Karl. He has admitted that his current team drives him
                  nuts because they don't have the talent to play defense.
                  2. Detroit Pistons early '90's. After Laimbeer retired, the
                  Piston's defense died. Rodman left afterwards, I think, but the
                  defense was suffering already. Dumars' defense no longer was
                  effective, in part due to age, in part due to the lack of inside
                  presence.

                  > But I think the even deeper and harder question is how much credit
                  to give
                  > to the coach, vs how much is either random or due to historical
                  accident.

                  I instinctively would say that the coach is fairly important and
                  random blips are not large. Even MikeT talks mostly about coaches
                  wearing out their welcomes, which seems to be accurate (esp. wrt
                  Collins).

                  > So even if most of that 2/3 of team defense is determined by what
                  > we call coaching, it's not clear that the coach is actually in
                  control of
                  > that 2/3. Even Lenny Wilkens lost effectiveness first in Seattle
                  and then
                  > in Cleveland and then in Atlanta, after early coaching success at
                  all
                  > three places.
                  >

                  Wilkens would be an interesting one to check out. I think you are
                  seeing a fairly real effect. I am starting to consider studies where
                  effects of players wear off. Maybe it's better and easier to look at
                  the effects of coaches. Control is difficult, of course.

                  This is not an easy question, I will admit. I have beliefs based
                  upon trying to make predictions and I have found that certain coaches
                  do seem to improve defenses (Karl, Riley, Jackson, not enough data to
                  say about Van Gundy)

                  DeanO
                • harlanzo
                  I know we cannot exactly quantify percentages of credit for defense between players and coaches (I know the number isn t always the same in every case even if
                  Message 8 of 14 , Feb 13, 2002
                  • 0 Attachment
                    I know we cannot exactly quantify percentages of credit for defense
                    between players and coaches (I know the number isn't always the same
                    in every case even if we could) but I am inclined to believe that
                    coaches generally are more responsible for this. The reason I
                    believe this is because any individual defender is really only as
                    good as help d. If you have a team of Rodman (circa 1988) playing
                    with Tripucka, Mark Jackson, Brad Lohaus, and Toni Kukoc will reflect
                    poorly on Rodman's D. Yet many of these players have also played on
                    good defensive teams. I realize a team of 5 great defenders would
                    probably limiting scoring even if Don Nelson were the coach, but
                    still I think schemes have at least as much to do with team defensive
                    success. This seems evident by lists of teams imporving and
                    declining in D with changing in coaches. (Hell, even the 90s Knicks
                    couldn't stop anyone when Nelson when was their coach in 1995-96).

                    To resolve this issue it would be great to create a system to
                    quantify defensive stats of individual players. I believe DeanO has
                    already discussed this issue plenty and it has become even more tough
                    with the zone D rule.
                  • HoopStudies
                    ... same ... I am, too, personally. The ramifications are that 1. Michael Jordan (pick your favorite player) is a worse player when a worse coach comes in,
                    Message 9 of 14 , Feb 13, 2002
                    • 0 Attachment
                      --- In APBR_analysis@y..., "harlanzo" <harlanzo@y...> wrote:
                      > I know we cannot exactly quantify percentages of credit for defense
                      > between players and coaches (I know the number isn't always the
                      same
                      > in every case even if we could) but I am inclined to believe that
                      > coaches generally are more responsible for this. The reason I

                      I am, too, personally. The ramifications are that

                      1. Michael Jordan (pick your favorite player) is a worse player when
                      a worse coach comes in, of course.
                      2. Players' values change when they change teams and change
                      coaches. (Those tests put forward before may not be valid?)
                      3. Predicting individual impact means understand how they fit in a
                      coach's system, not just knowing their own stats.

                      gulp.

                      > To resolve this issue it would be great to create a system to
                      > quantify defensive stats of individual players. I believe DeanO
                      has
                      > already discussed this issue plenty and it has become even more
                      tough
                      > with the zone D rule.

                      Let me put out a list of some of the best defenders I had last year:

                      Player Defensive Rating Stops Per Possession
                      Mutombo (A) 94.3 0.692
                      Battie, Tony 96.3 0.636
                      Campbell, Elden 93.2 0.624
                      Magloire, Jamaa 93.9 0.606
                      Traylor, Robert 98.1 0.607
                      Bradley, Shawn 93.1 0.702
                      LaFrentz, Raef 100.0 0.594
                      Williams, Jerom 95.5 0.612
                      Wallace, Ben 91.3 0.719
                      Foyle, Adonal 99.2 0.630
                      Olajuwon, Hakee 95.7 0.667
                      O'Neal, Jermain 94.0 0.660
                      Foster, Jeff 96.3 0.601
                      O'Neal, Shaquil 97.8 0.622
                      Mourning, Alonz 88.9 0.714
                      Jones, Eddie 94.9 0.563
                      Williams, Scott 98.5 0.600
                      Johnson, Ervin 98.6 0.596
                      Garnett, Kevin 97.1 0.616
                      Camby, Marcus 88.0 0.732
                      Thomas, Kurt 93.6 0.592
                      Outlaw, Bo 96.6 0.600
                      Ratliff, Theo 92.1 0.652
                      Mutombo (P) 92.3 0.647
                      Marion, Shawn 91.3 0.644
                      Kidd, Jason 95.5 0.541
                      Sabonis, Arvyda 94.9 0.629
                      Webber, Chris 94.2 0.603
                      Divac, Vlade 94.7 0.590
                      Duncan, Tim 91.3 0.655
                      Robinson, David 89.5 0.701
                      Rose, Malik 92.5 0.625
                      Clark, Keon 94.6 0.661
                      Ostertag, Greg 95.6 0.626
                      Swift, Stromile 98.6 0.622
                      White, Jahidi 100.3 0.613

                      This is a list of selected top defenders from last year. You'll
                      notice that it is mostly populated by big guys. I could have just
                      about the full Miami team on there if I was looking for the best
                      individual defensive ratings. Or I would have no guards on here at
                      all if I were looking for those who stopped the most scores per
                      possession.

                      What I like doing is contrasting players. Jason Kidd vs. Marbury:

                      Kidd, Jason 95.5 0.541
                      Marbury, Stepho 107.2 0.404

                      Kidd is a MUCH better defender, no matter how I look at it. I have
                      taken out context and analyzed them and Kidd looks better.
                      Sometimes, I take out context and the quality of the two guys
                      switches, implying there may be significant influences of
                      coach/system. Doug Christie isn't quite there. He is a pretty good
                      defender, but not great because you put him on a lousy defensive team
                      (like the Warriors) and they get only a little better. A lot of
                      perimeter guys, though, can be decent defenders. Tim Hardaway had a
                      99.0 rating last year with the Heat, but a 0.460 stop/poss number.
                      On that team, he did a decent job funneling things to bad shooting
                      position, basically covering his position. But did he really STOP
                      people? Not too often. My defensive ratings are very subject to
                      context, which I always hated until I sat back and started analyzing
                      defense in greater detail, seeing that system/coach are very
                      important.

                      I could go on and on.

                      DeanO
                    • alleyoop2
                      I m jazzed we re talking about defense (my last column on alleyoop.com was on this very subject), because it s easily the one area where folks like us need to
                      Message 10 of 14 , Feb 14, 2002
                      • 0 Attachment
                        I'm jazzed we're talking about defense (my last column on
                        alleyoop.com was on this very subject), because it's easily the one
                        area where folks like us need to do the most work.

                        I'm doing a study right now analyzing how guys did against the top
                        scorers at their positions, including assists and turnovers.
                        Basically, I've selected a few guys that are the top guns at their
                        positions and tried to select so that cross-matches wouldn't be a big
                        problem (for example, I included Cassell-Allen-Robinson in the study
                        because you'd never put your shooting guard on Robinson or vice
                        versa, so it's easy to tell who guarded who just from the box score.
                        Hope that makes sense), and I'm only selecting games where both
                        player and opponent played at least 25 minutes.

                        My hypothesis is that the quality of a defender may not be readily
                        apparent in his man's points scored, but will show up in assists and
                        turonvers -- the assists reflecting how much double-team help is
                        needed, the turnovers reflecting how much of a pain in the arse the
                        defender was. That's my hypothesis, anyway; I'll see what the results
                        bear out when I'm done with this thing, which may not be for a while.

                        At any rate, my question to the masses is:

                        1) Any other ideas out there on how to study this better?

                        2) Have any other studies similar to this one been done that I'm not
                        aware of? (I'm aware of Doug Steele's defensive difference work, but
                        not much else).

                        John




                        --- In APBR_analysis@y..., "HoopStudies" <deano@r...> wrote:
                        > --- In APBR_analysis@y..., "harlanzo" <harlanzo@y...> wrote:
                        > > I know we cannot exactly quantify percentages of credit for
                        defense
                        > > between players and coaches (I know the number isn't always the
                        > same
                        > > in every case even if we could) but I am inclined to believe that
                        > > coaches generally are more responsible for this. The reason I
                        >
                        > I am, too, personally. The ramifications are that
                        >
                        > 1. Michael Jordan (pick your favorite player) is a worse player
                        when
                        > a worse coach comes in, of course.
                        > 2. Players' values change when they change teams and change
                        > coaches. (Those tests put forward before may not be valid?)
                        > 3. Predicting individual impact means understand how they fit in a
                        > coach's system, not just knowing their own stats.
                        >
                        > gulp.
                        >
                        > > To resolve this issue it would be great to create a system to
                        > > quantify defensive stats of individual players. I believe DeanO
                        > has
                        > > already discussed this issue plenty and it has become even more
                        > tough
                        > > with the zone D rule.
                        >
                        > Let me put out a list of some of the best defenders I had last year:
                        >
                        > Player Defensive Rating Stops Per Possession
                        > Mutombo (A) 94.3 0.692
                        > Battie, Tony 96.3 0.636
                        > Campbell, Elden 93.2 0.624
                        > Magloire, Jamaa 93.9 0.606
                        > Traylor, Robert 98.1 0.607
                        > Bradley, Shawn 93.1 0.702
                        > LaFrentz, Raef 100.0 0.594
                        > Williams, Jerom 95.5 0.612
                        > Wallace, Ben 91.3 0.719
                        > Foyle, Adonal 99.2 0.630
                        > Olajuwon, Hakee 95.7 0.667
                        > O'Neal, Jermain 94.0 0.660
                        > Foster, Jeff 96.3 0.601
                        > O'Neal, Shaquil 97.8 0.622
                        > Mourning, Alonz 88.9 0.714
                        > Jones, Eddie 94.9 0.563
                        > Williams, Scott 98.5 0.600
                        > Johnson, Ervin 98.6 0.596
                        > Garnett, Kevin 97.1 0.616
                        > Camby, Marcus 88.0 0.732
                        > Thomas, Kurt 93.6 0.592
                        > Outlaw, Bo 96.6 0.600
                        > Ratliff, Theo 92.1 0.652
                        > Mutombo (P) 92.3 0.647
                        > Marion, Shawn 91.3 0.644
                        > Kidd, Jason 95.5 0.541
                        > Sabonis, Arvyda 94.9 0.629
                        > Webber, Chris 94.2 0.603
                        > Divac, Vlade 94.7 0.590
                        > Duncan, Tim 91.3 0.655
                        > Robinson, David 89.5 0.701
                        > Rose, Malik 92.5 0.625
                        > Clark, Keon 94.6 0.661
                        > Ostertag, Greg 95.6 0.626
                        > Swift, Stromile 98.6 0.622
                        > White, Jahidi 100.3 0.613
                        >
                        > This is a list of selected top defenders from last year. You'll
                        > notice that it is mostly populated by big guys. I could have just
                        > about the full Miami team on there if I was looking for the best
                        > individual defensive ratings. Or I would have no guards on here at
                        > all if I were looking for those who stopped the most scores per
                        > possession.
                        >
                        > What I like doing is contrasting players. Jason Kidd vs. Marbury:
                        >
                        > Kidd, Jason 95.5 0.541
                        > Marbury, Stepho 107.2 0.404
                        >
                        > Kidd is a MUCH better defender, no matter how I look at it. I have
                        > taken out context and analyzed them and Kidd looks better.
                        > Sometimes, I take out context and the quality of the two guys
                        > switches, implying there may be significant influences of
                        > coach/system. Doug Christie isn't quite there. He is a pretty
                        good
                        > defender, but not great because you put him on a lousy defensive
                        team
                        > (like the Warriors) and they get only a little better. A lot of
                        > perimeter guys, though, can be decent defenders. Tim Hardaway had
                        a
                        > 99.0 rating last year with the Heat, but a 0.460 stop/poss number.
                        > On that team, he did a decent job funneling things to bad shooting
                        > position, basically covering his position. But did he really STOP
                        > people? Not too often. My defensive ratings are very subject to
                        > context, which I always hated until I sat back and started
                        analyzing
                        > defense in greater detail, seeing that system/coach are very
                        > important.
                        >
                        > I could go on and on.
                        >
                        > DeanO
                      • harlanzo
                        I agree with the idea that defensive ability is not necessarily measured by points allowed. But I am not convinced that assists will reflect it any better.
                        Message 11 of 14 , Feb 14, 2002
                        • 0 Attachment
                          I agree with the idea that defensive ability is not necessarily
                          measured by points allowed. But I am not convinced that assists will
                          reflect it any better. For example, say Jordan is being guarded by
                          Starks. Starks can be playing great defense and force MJ into tough
                          fadeaways all day and Mj could hit 65% of them anyway. This does not
                          mean Starks played poorly it merely means that MJ's offense is at a
                          level where any ordinary defender cant stop him. Conversely, you
                          could leave MJ open from the perimeter all night and he might shoot 6-
                          20 anyway. This might also be true of assists. I think to study
                          defense you have to split hairs more than the presents stats allow.
                          Give a rating to the quality of a shot the offensive player takes (ie
                          wide open jumper, fadeaway, buzzer beater etc.) and the kind of
                          assist (dish off after penetration vs. a simple pass around the
                          perimeter) as well as help defense credits. I am not sure the
                          specifics and I am not saying that the study proposed by alleyoop
                          wont yield some interesting information but I think we need more info
                          to better answer defensive questions.


                          --- In APBR_analysis@y..., "alleyoop2" <alleyoop2@y...> wrote:
                          > I'm jazzed we're talking about defense (my last column on
                          > alleyoop.com was on this very subject), because it's easily the one
                          > area where folks like us need to do the most work.
                          >
                          > I'm doing a study right now analyzing how guys did against the top
                          > scorers at their positions, including assists and turnovers.
                          > Basically, I've selected a few guys that are the top guns at their
                          > positions and tried to select so that cross-matches wouldn't be a
                          big
                          > problem (for example, I included Cassell-Allen-Robinson in the
                          study
                          > because you'd never put your shooting guard on Robinson or vice
                          > versa, so it's easy to tell who guarded who just from the box
                          score.
                          > Hope that makes sense), and I'm only selecting games where both
                          > player and opponent played at least 25 minutes.
                          >
                          > My hypothesis is that the quality of a defender may not be readily
                          > apparent in his man's points scored, but will show up in assists
                          and
                          > turonvers -- the assists reflecting how much double-team help is
                          > needed, the turnovers reflecting how much of a pain in the arse the
                          > defender was. That's my hypothesis, anyway; I'll see what the
                          results
                          > bear out when I'm done with this thing, which may not be for a
                          while.
                          >
                          > At any rate, my question to the masses is:
                          >
                          > 1) Any other ideas out there on how to study this better?
                          >
                          > 2) Have any other studies similar to this one been done that I'm
                          not
                          > aware of? (I'm aware of Doug Steele's defensive difference work,
                          but
                          > not much else).
                          >
                          > John
                          >
                          >
                          >
                          >
                          > --- In APBR_analysis@y..., "HoopStudies" <deano@r...> wrote:
                          > > --- In APBR_analysis@y..., "harlanzo" <harlanzo@y...> wrote:
                          > > > I know we cannot exactly quantify percentages of credit for
                          > defense
                          > > > between players and coaches (I know the number isn't always the
                          > > same
                          > > > in every case even if we could) but I am inclined to believe
                          that
                          > > > coaches generally are more responsible for this. The reason I
                          > >
                          > > I am, too, personally. The ramifications are that
                          > >
                          > > 1. Michael Jordan (pick your favorite player) is a worse player
                          > when
                          > > a worse coach comes in, of course.
                          > > 2. Players' values change when they change teams and change
                          > > coaches. (Those tests put forward before may not be valid?)
                          > > 3. Predicting individual impact means understand how they fit in
                          a
                          > > coach's system, not just knowing their own stats.
                          > >
                          > > gulp.
                          > >
                          > > > To resolve this issue it would be great to create a system to
                          > > > quantify defensive stats of individual players. I believe
                          DeanO
                          > > has
                          > > > already discussed this issue plenty and it has become even more
                          > > tough
                          > > > with the zone D rule.
                          > >
                          > > Let me put out a list of some of the best defenders I had last
                          year:
                          > >
                          > > Player Defensive Rating Stops Per Possession
                          > > Mutombo (A) 94.3 0.692
                          > > Battie, Tony 96.3 0.636
                          > > Campbell, Elden 93.2 0.624
                          > > Magloire, Jamaa 93.9 0.606
                          > > Traylor, Robert 98.1 0.607
                          > > Bradley, Shawn 93.1 0.702
                          > > LaFrentz, Raef 100.0 0.594
                          > > Williams, Jerom 95.5 0.612
                          > > Wallace, Ben 91.3 0.719
                          > > Foyle, Adonal 99.2 0.630
                          > > Olajuwon, Hakee 95.7 0.667
                          > > O'Neal, Jermain 94.0 0.660
                          > > Foster, Jeff 96.3 0.601
                          > > O'Neal, Shaquil 97.8 0.622
                          > > Mourning, Alonz 88.9 0.714
                          > > Jones, Eddie 94.9 0.563
                          > > Williams, Scott 98.5 0.600
                          > > Johnson, Ervin 98.6 0.596
                          > > Garnett, Kevin 97.1 0.616
                          > > Camby, Marcus 88.0 0.732
                          > > Thomas, Kurt 93.6 0.592
                          > > Outlaw, Bo 96.6 0.600
                          > > Ratliff, Theo 92.1 0.652
                          > > Mutombo (P) 92.3 0.647
                          > > Marion, Shawn 91.3 0.644
                          > > Kidd, Jason 95.5 0.541
                          > > Sabonis, Arvyda 94.9 0.629
                          > > Webber, Chris 94.2 0.603
                          > > Divac, Vlade 94.7 0.590
                          > > Duncan, Tim 91.3 0.655
                          > > Robinson, David 89.5 0.701
                          > > Rose, Malik 92.5 0.625
                          > > Clark, Keon 94.6 0.661
                          > > Ostertag, Greg 95.6 0.626
                          > > Swift, Stromile 98.6 0.622
                          > > White, Jahidi 100.3 0.613
                          > >
                          > > This is a list of selected top defenders from last year. You'll
                          > > notice that it is mostly populated by big guys. I could have
                          just
                          > > about the full Miami team on there if I was looking for the best
                          > > individual defensive ratings. Or I would have no guards on here
                          at
                          > > all if I were looking for those who stopped the most scores per
                          > > possession.
                          > >
                          > > What I like doing is contrasting players. Jason Kidd vs. Marbury:
                          > >
                          > > Kidd, Jason 95.5 0.541
                          > > Marbury, Stepho 107.2 0.404
                          > >
                          > > Kidd is a MUCH better defender, no matter how I look at it. I
                          have
                          > > taken out context and analyzed them and Kidd looks better.
                          > > Sometimes, I take out context and the quality of the two guys
                          > > switches, implying there may be significant influences of
                          > > coach/system. Doug Christie isn't quite there. He is a pretty
                          > good
                          > > defender, but not great because you put him on a lousy defensive
                          > team
                          > > (like the Warriors) and they get only a little better. A lot of
                          > > perimeter guys, though, can be decent defenders. Tim Hardaway
                          had
                          > a
                          > > 99.0 rating last year with the Heat, but a 0.460 stop/poss
                          number.
                          > > On that team, he did a decent job funneling things to bad
                          shooting
                          > > position, basically covering his position. But did he really
                          STOP
                          > > people? Not too often. My defensive ratings are very subject to
                          > > context, which I always hated until I sat back and started
                          > analyzing
                          > > defense in greater detail, seeing that system/coach are very
                          > > important.
                          > >
                          > > I could go on and on.
                          > >
                          > > DeanO
                        • HoopStudies
                          ... (ie ... info ... Some coaches do this right now, but their information is not public. They typically look only at the quality of shots given up, where it
                          Message 12 of 14 , Feb 14, 2002
                          • 0 Attachment
                            --- In APBR_analysis@y..., "harlanzo" <harlanzo@y...> wrote:
                            > Give a rating to the quality of a shot the offensive player takes
                            (ie
                            > wide open jumper, fadeaway, buzzer beater etc.) and the kind of
                            > assist (dish off after penetration vs. a simple pass around the
                            > perimeter) as well as help defense credits. I am not sure the
                            > specifics and I am not saying that the study proposed by alleyoop
                            > wont yield some interesting information but I think we need more
                            info
                            > to better answer defensive questions.

                            Some coaches do this right now, but their information is not public.
                            They typically look only at the quality of shots given up, where it
                            was and whether there was a hand in the face. I haven't thought
                            about this enough to know whether even that would help (recording
                            information is different from understanding what it means). I have
                            not heard of coaches recording whether guys give up lots of assists.
                            That, frankly, is tough. Whose responsibility is it when a guy gets
                            help defensively, but the guy who gave help left his man open?

                            I have also seen blockout tabulation from coaches. I have never seen
                            it well analyzed, though.

                            We're all, however, saying the same _general_ thing: We need better
                            data.

                            I would add this question: if we get all these different types of
                            data, how are we going to analyze them?

                            JohnH(alleyoop): Doug's analysis is the best I know doing roughly
                            what you want. Otherwise, I am working to put together volunteers to
                            score WNBA games for defensive stats.

                            DeanO
                          • mikel_ind
                            ... Has ... so ... having ... careful ... with ... player ... Not only is it difficult to ascribe a team s sudden turnaround (or lack thereof), to one
                            Message 13 of 14 , Feb 14, 2002
                            • 0 Attachment
                              --- In APBR_analysis@y..., "Michael K. Tamada" <tamada@o...> wrote:
                              >
                              > On Wed, 13 Feb 2002, harlanzo wrote:
                              >
                              > > I was just looking over the Orlando Magic from the last 3 years.
                              Has
                              > > anyone noticed that the net effect of adding Mcgrady has not been
                              so
                              > > impressive?....


                              > [Tamada writes] .... I
                              > think very few people were predicting that Michael Jordan would be
                              having
                              > as much impact on the Wizards as he has. Although we have to be
                              careful
                              > not to give him all the credit, similar to the Jason Kidd situation
                              with
                              > the Nets. But I think with Jordan we literally do have a case of a
                              player
                              > making his teammates better ...> --MKT

                              Not only is it difficult to ascribe a team's sudden turnaround (or
                              lack thereof), to one individual, it's really hard to conjecture that
                              other individuals actually played better (or worse) because of the
                              first guy.

                              One player that always seems overrated to me (statistically speaking)
                              is Jason Kidd. So I have endeavored to debunk the notion that he is
                              All That. (There is even MVP talk out there.)

                              One could argue that Kidd isn't a lot better than Marbury, and simply
                              adding Macculloch, getting Kittles back, Van Horn and Martin staying
                              healthy, have had more to do with the Nets' success.

                              The counterargument is that the "other" Nets are doing better Because
                              of Kidd: the Kidd Effect.

                              The only way I can imagine analyzing this is to look at all the
                              players Kidd has ever played with, who have also played without him,
                              and comparing "adjacent" seasons (chronologically). A sufficiently
                              large sample might show a significant difference.

                              In the 3 cities that have owned Kidd's services, there are some 62
                              players with significant minutes alongside Kidd and without him.
                              Some players are counted more than once: joining Kidd and leaving
                              Kidd count as separate comparisons.

                              I found 62 instances of comparison:

                              Min. Pct. Scor. Reb. Ast. Stl. (.TO.) Blk
                              with Kidd 23.8 .516 13.24 6.30 2.38 1.03 (1.73) .71
                              without K 24.4 .501 12.01 6.24 2.48 1.05 (1.85) .69

                              I believe this is both astounding and significant. I have looked for
                              possible skewings of the results, but I can't find any. Jason Kidd
                              gets 10% more scoring out of his teammates!

                              I had thought perhaps any scoring enhancement would come at the
                              expense of assists (which does drop, but only 4%), or rebounds, since
                              Kidd gets a lot of those. But no.

                              Even the 'anecdotal' evidence indicates that Kidd tends to work
                              wonders for his teammates. Lucius Harris played with Kidd in Dallas,
                              then worked without him for 5 mostly terrible years, and is again
                              flourishing in New Jersey. George McCloud became stellar in Dallas
                              with Kidd, sucked in LA, regained his form in Phoenix, and has mostly
                              stunk in Denver.

                              The only member of the Nets not showing improved numbers this season
                              is Aaron Williams.


                              Mike Goodman
                            • Michael K. Tamada
                              On Thu, 14 Feb 2002, mikel_ind wrote: [...] ... Count me as a Kidd fan, and anti-Marbury, for the standard reasons mentioned by fans and media. Now the
                              Message 14 of 14 , Feb 14, 2002
                              • 0 Attachment
                                On Thu, 14 Feb 2002, mikel_ind wrote:

                                [...]

                                > One player that always seems overrated to me (statistically speaking)
                                > is Jason Kidd. So I have endeavored to debunk the notion that he is
                                > All That. (There is even MVP talk out there.)
                                >
                                > One could argue that Kidd isn't a lot better than Marbury, and simply

                                Count me as a Kidd fan, and anti-Marbury, for the standard reasons
                                mentioned by fans and media. Now the domestic violence stuff against Kidd
                                IS serious, and I can understand why Phoenix decided to get rid of him.
                                But I that they could have and should have gotten more than Marbury for
                                Kidd.

                                [...]

                                > Min. Pct. Scor. Reb. Ast. Stl. (.TO.) Blk
                                > with Kidd 23.8 .516 13.24 6.30 2.38 1.03 (1.73) .71
                                > without K 24.4 .501 12.01 6.24 2.48 1.05 (1.85) .69
                                >
                                > I believe this is both astounding and significant. I have looked for
                                > possible skewings of the results, but I can't find any. Jason Kidd
                                > gets 10% more scoring out of his teammates!

                                I agree that it is significant, not quite as astounded as you because I've
                                believed all along that Kidd makes his teammates better (much more so than
                                Marbury).

                                > Even the 'anecdotal' evidence indicates that Kidd tends to work
                                > wonders for his teammates. Lucius Harris played with Kidd in Dallas,
                                > then worked without him for 5 mostly terrible years, and is again
                                > flourishing in New Jersey. George McCloud became stellar in Dallas
                                > with Kidd, sucked in LA, regained his form in Phoenix, and has mostly
                                > stunk in Denver.

                                I hadn't noticed Harris, but McCloud I sure had. I think he started with
                                Indiana? Did nothing, but became a 3-point bomber in Dallas. But he's
                                clearly a player who needs to be in the right situation in order to be
                                effective. Good teams are usually better at providing the right situation
                                (another example: Bob McAdoo in Buffalo and LA, vs. NY, NJ, Boston, and
                                Detroit).

                                > The only member of the Nets not showing improved numbers this season
                                > is Aaron Williams.

                                Aaron's always been an active, relatively productive per-minute player,
                                even with the Sonics. But Karl never gave him playing time, and (unlike
                                some of Karl's bad playing time decisions) I think for good reason.
                                AWilliams can grab some rebounds, and he'll put up a lot of shots and
                                actually make some of them, but with the Sonics at least he simply didn't
                                know what he was doing out there. Athleticism and hustle can get you so
                                far, and even lead to rebounds and points in the NBA, but to become an NBA
                                starter a certain knowledge of basketball, court sense, team play, self-
                                discipline, or whatever, is needed. In AWilliams case, I don't know what
                                it was that he lacked, but it was definitely something, he simply was not
                                in the team flow and despite his rebounds and points was not contributing
                                in a team way.

                                Possibly the same thing is still happening in NJ. You won't benefit from
                                Kidd's presense if you don't know when to cut to the basket versus when to
                                spot-up for a jumper versus when to post-up inside versus when to set
                                a pick versus when to get position for an offensive rebound.

                                I have not seen the Nets play much recently so I don't know if these
                                statements are still applicable to AWilliams, but that's the way he was in
                                Seattle.



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