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

Re: predicting bust-out years (and bust years)

Expand Messages
  • Dean Oliver
    Not sure what inspired what, but the NBA GM s had their picks for bust out years, too. Online at http://www.nba.com/preview2002/General_Manager_Survey.html
    Message 1 of 20 , Oct 10, 2002
    • 0 Attachment
      Not sure what inspired what, but the NBA GM's had their picks for
      bust out years, too. Online at

      http://www.nba.com/preview2002/General_Manager_Survey.html

      Other than Darius Miles and Richard Jefferson, it's pretty random.
      Both those guys would bust out because, using BobC terminology,
      they'll be getting more touches.

      Darius Miles - 15.4%

      Richard Jefferson - 11.5%

      Rashard Lewis - 7.7%

      Jason Richardson - 7.7%

      Courtney Alexander - 3.8%

      Vin Baker - 3.8%

      Jonathan Bender - 3.8%

      Mike Bibby - 3.8%

      Pau Gasol - 3.8%

      Eddie Griffin - 3.8%

      Al Harrington - 3.8%

      Andrei Kirilenko - 3.8%

      Corey Magette - 3.8%

      Stephon Marbury - 3.8%

      Kenyon Martin - 3.8%

      Andre Miller - 3.8%

      Yao Ming - 3.8%

      Jermaine O'Neal - 3.8%

      Jason Terry - 3.8%



      --- In APBR_analysis@y..., "Mike G" <msg_53@h...> wrote:
      > Okay, back with an improved list of who's hot and who's not.
      >
      > Now I have a list of 240 guys who have played the last 4 years, and
      > another 60 who have played 3 of the last 4 years.
      >
      > I do not figure in career averages this time around. Rather, I
      take
      > the ratio of last year to previous 3 years (weighted 3:2:1). Then
      I
      > take the ratio of the last 2 years to the 2 previous years. Then I
      > take the average.
      >
      > Pretty crude, I know. The idea is to be comprehensive.
      >
      > Once again, here are players on the rise:
      >
      > 1.53 Jermaine O'Neal 33.9
      > 1.53 Nazr Mohammed 24.2
      > 1.43 Rashard Lewis 29.7
      > 1.42 Al Harrington 25.3
      > 1.40 Kevin Ollie 19.2
      > 1.37 Steve Nash 31.6
      > 1.33 Dirk Nowitzki 37.7
      > 1.32 Popeye Jones 23.8
      > 1.31 Corliss Williamson 30.7
      > 1.30 Calbert Cheaney 18.1
      >
      > 1.30 Rick Brunson 17.2
      > 1.30 Chauncey Billups 28.4
      > 1.29 Jeff McInnis 25.0
      > 1.29 Lucious Harris 23.5
      > 1.28 Predrag Stojakovic 31.0
      > 1.27 Alvin Williams 25.2
      > 1.27 Tyronn Lue 22.3
      > 1.27 Radoslav Nesterovich 23.4
      > 1.25 Bruce Bowen 15.6
      > 1.24 Vladimir Stepania 21.4
      >
      > 1.24 Jason Williams 28.9
      > 1.24 Keon Clark 29.2
      > 1.24 John Crotty 22.6
      > 1.23 Aaron McKie 26.7
      > 1.22 Michael Stewart 15.5
      > 1.22 Ben Wallace 30.2
      > 1.21 Tracy McGrady 40.3
      > 1.21 Derek Fisher 22.3
      > 1.21 Troy Hudson 23.9
      > 1.21 Ricky Davis 24.7
      >
      > Here are some 3-year players (figured a little differently).
      >
      > 1.39 Rafer Alston 20.2
      > 1.37 Devean George 20.9
      > 1.28 Scott Padgett 23.3
      > 1.27 Jonathan Bender 18.4
      > 1.26 Kenny Thomas 26.4
      > 1.26 Jumaine Jones 22.8
      > 1.23 Baron Davis 33.5
      > 1.21 Richard Hamilton 28.6
      > 1.21 Todd MacCulloch 28.2
      > 1.20 Dion Glover 21.3
      >
      > 1.18 Rusty LaRue 18.8
      > 1.18 Jason Terry 31.2
      > 1.17 Anthony Johnson 15.4
      > 1.15 Adrian Griffin 20.1
      > 1.13 Ron Artest 27.2
      >
      >
      > Now, the bottom of the list, 4-year players whose stocks are
      dropping.
      >
      > .44 Dee Brown 6.8
      > .58 Eldridge Recasner 7.6
      > .64 Glen Rice 14.7
      > .66 Felton Spencer 8.9
      > .66 John Starks 13.8
      > .67 Brent Price 10.9
      > .67 Mookie Blaylock 15.3
      > .68 Jelani McCoy 13.5
      > .68 Shawn Kemp 21.5
      > .70 Johnny Newman 11.7
      >
      > .70 Mitch Richmond 16.6
      > .71 Tom Gugliotta 19.1
      > .71 Nick Anderson 14.3
      > .73 Bimbo Coles 14.4
      > .73 Shandon Anderson 14.9
      > .76 Lindsey Hunter 16.0
      > .76 Patrick Ewing 22.9
      > .77 Kendall Gill 16.7
      > .77 Sam Mitchell 14.4
      > .78 Grant Long 15.0
      >
      > .78 Tim Hardaway .78
      > .78 Vitaly Potapenko 17.2
      > .78 George Lynch 16.4
      > .78 Ervin Johnson 16.0
      > .79 Shawn Bradley 20.1
      > .79 Theo Ratliff 21.0
      > .79 Jaren Jackson 13.2
      > .79 Robert Pack 17.1
      > .80 Howard Eisley 16.3
      > .80 Travis Knight 12.6
      >
      >
      > And the dimmest of the 3-year bunch:
      >
      > .66 Sam Mack 10.8
      > .77 John Amaechi 13.8
      > .81 Carlos Rogers 17.2
      > .82 Anthony Mason 22.9
      > .82 Mikki Moore 18.7
      > .83 Andrew DeClerq 15.7
      > .84 Obinna Ekezie 14.1
      > .86 Michael Ruffin 13.4
      > .87 Emanuel Davis 14.5
      > .87 Evan Eschmeyer 15.5
    • John Hollinger
      By weighting your list toward last season and not including an age factor, I think your first list unwittingly was more like guys who played over their head
      Message 2 of 20 , Oct 10, 2002
      • 0 Attachment
        By weighting your list toward last season and not including an age
        factor, I think your first list unwittingly was more like "guys who
        played over their head last season". A lot of them - Williamson and
        Bobby Jackson, for instance -- showed up on my "guys who have no
        chance of repeating last season" list in the study I did for the book.



        --- In APBR_analysis@y..., "Michael K. Tamada" <tamada@o...> wrote:
        >
        >
        > On Thu, 10 Oct 2002, Mike G wrote:
        >
        > [...]
        >
        > > Once again, here are players on the rise:
        > >
        > > 1.53 Jermaine O'Neal 33.9
        >
        > [...]
        >
        > > 1.22 Ben Wallace 30.2
        > > 1.21 Tracy McGrady 40.3
        >
        > Interesting stuff, it'll be interesting to see how much predictive
        value
        > they have. Some of these stats clearly are describing young
        players who
        > are improving, e.g. O'Neal. Other may be describing players who
        aren't so
        > much improving as finding their niche on the court, or their teams
        are
        > figuring out how to utilize them.
        >
        > But in both cases, especially the latter, can we detect or predict
        when a
        > player is approaching his ceiling, and getting as good as he'll
        ever get?
        > I think that Ben Wallace might be such a player, it's no surprise
        to see
        > him on a most-improved list, but can he keep on improving or has he
        > reached the limit of his potential? At a higher level, the same
        question
        > might be asked of McGrady -- how much better could we expect him to
        get,
        > compared to where he is already?
        >
        > Every once in a very long while, players who might seem to be
        approaching
        > their ceiling keep on busting through, Jordan being the most
        prominent
        > example. He might've reached a statistical peak fairly early on,
        but in
        > non-statistical ways kept on displaying a will to win and amazing
        > consistency.
        >
        >
        > --MKT
      • Michael K. Tamada
        Yes, I had that in the back of my mind and should ve added it as third interpretation of the players on this most improving list: 1. Young players who are
        Message 3 of 20 , Oct 11, 2002
        • 0 Attachment
          Yes, I had that in the back of my mind and should've added it as third
          interpretation of the players on this "most improving" list:

          1. Young players who are getting better (Jermaine O'Neal, Rashard Lewis,
          etc.).

          2. Not-so-young players who might be getting better, i.e. late bloomers
          (Ben Wallace), or who might be players who are just finding a better role
          or niche.

          3. Players whose recent performance was better than their previous
          performance, but more due to luck, having a career year (ie luck), or
          having a really good role/niche which is unlikely to be repeated.


          Players in category 1 can be usually be expected at worst to repeat their
          performance and many have a good chance of continuing to improve.
          Category 2 players are less likely to improve this year, but they might
          stay steady (I don't see anyone predicting a downfall for Ben Wallace
          e.g., even though many people are down on the Pistons as a team).
          Category 3 players are more likely to decline rather than improve.


          And so the question is, can we use other variables such as age as JohnH
          suggested, combined with these improving-players lists, to predict this
          year's improving/declining players? Shouldn't be hard in some cases.
          Only injuries e.g. I think could derail the improvement that I expect to
          see from the West Coast generation of athletic, not-so-good-
          shooting-but-improving SGs: the Richardsons (Jason and Q) and Desmond
          Mason.

          In addition to age, the other variable that I was trying to invoke is the
          notion of a performance ceiling; after a certain level, improvement
          becomes unlikely because a player has become so darned good that it's not
          reasonable to expect continued improvement. Maybe McGrady is at that
          ceiling, maybe not. Shawn Kemp reached it in 1996; he improved every
          single year until then, but once you reach the level of (IMO) best player
          in the finals (including Jordan), it's an awful stretch to predict
          continued improvement for the next year. (Jordan however arguably did
          continue to improve or at least add additional ways of helping his team
          win even after achieving his initial MVP, championship level). But how
          many players continue to improve after reaching all-pro status? A very
          few, the Birds, Magics, etc. of the world. A few more can stay steady at
          an all-pro level for a few years. Most or at least many fall back to a
          lesser level after a brief stay at the top.


          --MKT


          On Fri, 11 Oct 2002, John Hollinger wrote:

          > By weighting your list toward last season and not including an age
          > factor, I think your first list unwittingly was more like "guys who
          > played over their head last season". A lot of them - Williamson and
          > Bobby Jackson, for instance -- showed up on my "guys who have no
          > chance of repeating last season" list in the study I did for the book.
          >
          >
          >
          > --- In APBR_analysis@y..., "Michael K. Tamada" <tamada@o...> wrote:
          > >
          > >
          > > On Thu, 10 Oct 2002, Mike G wrote:
          > >
          > > [...]
          > >
          > > > Once again, here are players on the rise:
          > > >
          > > > 1.53 Jermaine O'Neal 33.9
          > >
          > > [...]
          > >
          > > > 1.22 Ben Wallace 30.2
          > > > 1.21 Tracy McGrady 40.3
          > >
          > > Interesting stuff, it'll be interesting to see how much predictive
          > value
          > > they have. Some of these stats clearly are describing young
          > players who
          > > are improving, e.g. O'Neal. Other may be describing players who
          > aren't so
          > > much improving as finding their niche on the court, or their teams
          > are
          > > figuring out how to utilize them.
          > >
          > > But in both cases, especially the latter, can we detect or predict
          > when a
          > > player is approaching his ceiling, and getting as good as he'll
          > ever get?
          > > I think that Ben Wallace might be such a player, it's no surprise
          > to see
          > > him on a most-improved list, but can he keep on improving or has he
          > > reached the limit of his potential? At a higher level, the same
          > question
          > > might be asked of McGrady -- how much better could we expect him to
          > get,
          > > compared to where he is already?
          > >
          > > Every once in a very long while, players who might seem to be
          > approaching
          > > their ceiling keep on busting through, Jordan being the most
          > prominent
          > > example. He might've reached a statistical peak fairly early on,
          > but in
          > > non-statistical ways kept on displaying a will to win and amazing
          > > consistency.
          > >
          > >
          > > --MKT
          >
          >
          >
          > To unsubscribe from this group, send an email to:
          > APBR_analysis-unsubscribe@yahoogroups.com
          >
          >
          >
          > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
          >
          >
        • Mike G
          ... Well, time will tell if guys are playing over their heads . Sometimes players just go on to the next level. ... book. I have to assume no chance is
          Message 4 of 20 , Oct 11, 2002
          • 0 Attachment
            --- In APBR_analysis@y..., "John Hollinger" <alleyoop2@y...> wrote:
            > By weighting your list toward last season and not including an age
            > factor, I think your first list unwittingly was more like "guys who
            > played over their head last season".

            Well, time will tell if guys are playing 'over their heads'.
            Sometimes players just go on to the next level.

            > A lot of them - Williamson and
            > Bobby Jackson, for instance -- showed up on my "guys who have no
            > chance of repeating last season" list in the study I did for the
            book.

            I have to assume "no chance" is sports lingo for 'less than likely',
            rather than literally 0% chance.

            Bobby Jackson has played 5 years. He was a starter his rookie year,
            and backed up Brandon and now Bibby as point guard. Last year he
            discovered his shot. Are you saying his shot was on loan, and now he
            has to give it back?

            Corliss Williamson has been an underachiever until now. He has
            played 7 years.

            A quick look shows there are quite a few players whose peak years
            were after their 7th: Blaylock, Bogues, Elden Campbell, Cassell,
            Dell Curry, Antonio Davis, Gilliam, Avery Johnson, Karl Malone,
            Mashburn, Reggie Miller, Mourning, Shaq, Payton, Perkins, Wesley
            Person, Rice, Richmond, Schrempf, Thorpe, Van Exel, Webber, Willis,
            and others.

            It seems that 6-8 years in the league is the normal peak. But to
            declare that a player has certainly reached his zenith at year 7, is
            going out on a limb.

            Much of what leads to improvement is just hard work. Dedication and
            concentration are essential both to achieving excellence and
            maintaining it. Karl Malone has to work very hard just to remain so
            good. I have no doubt Corliss Williamson worked very hard on his
            game, and luck had very little to do with his improvement.
          • Dean Oliver
            ... is ... What you re arguing here, of course, is that age has little predictive value. Actually, I think this is fairly true. I did a rather extensive
            Message 5 of 20 , Oct 11, 2002
            • 0 Attachment
              --- In APBR_analysis@y..., "Mike G" <msg_53@h...> wrote:
              > A quick look shows there are quite a few players whose peak years
              > were after their 7th: Blaylock, Bogues, Elden Campbell, Cassell,
              > Dell Curry, Antonio Davis, Gilliam, Avery Johnson, Karl Malone,
              > Mashburn, Reggie Miller, Mourning, Shaq, Payton, Perkins, Wesley
              > Person, Rice, Richmond, Schrempf, Thorpe, Van Exel, Webber, Willis,
              > and others.
              >
              > It seems that 6-8 years in the league is the normal peak. But to
              > declare that a player has certainly reached his zenith at year 7,
              is
              > going out on a limb.
              >

              What you're arguing here, of course, is that age has little
              predictive value. Actually, I think this is fairly true. I did a
              rather extensive study of how players performed as a function of age
              and saw little relationship. There is a little more strength in
              performance vs. experience (which is different due to so many kids
              coming out), but not much. Tying age and experience together was
              little help, though I thought it would, with experience defining how
              quickly a player reached a peak and age defining how long they
              maintained it. Not so much.

              Great players are simultaneously more predictable and less
              predictable than others. They are more predictable in that they are
              going to get time, they are going to have offenses built around them
              so they never have to adjust to different roles. They usually have
              nicely shaped peaks on their offensive ratings, increasing early in
              their careers (as Nowitzki has), and slowly declining. The
              unpredictable part is exactly how high they will get when looking at
              their early years, which, of course, can be key. Jordan's 2nd
              season, with injury didn't show how good he'd get (though the famous
              63 pt playoff game might have).

              Jordan

              Season G Sc.Pos. Poss. Fl % ORtg. PP/G %TmPoss Age
              1985 82 1106 1921 0.58 118 27.6 30% 21
              1986 18 179 336 0.53 109 20.4 36% 22
              1987 82 1336 2355 0.57 117 33.6 36% 23
              1988 82 1292 2143 0.60 123 32.2 33% 24
              1989 81 1250 2104 0.59 123 31.9 32% 25
              1990 82 1226 2092 0.59 123 31.4 32% 26
              1991 82 1152 1892 0.61 126 29.0 31% 27
              1992 80 1088 1841 0.59 121 27.9 30% 28
              1993 78 1106 1938 0.57 119 29.7 33% 29
              1995 17 208 399 0.52 109 25.7 31% 31
              1996 82 1064 1833 0.58 124 27.6 31% 32
              1997 82 1020 1791 0.57 121 26.5 31% 33
              1998 82 1033 1857 0.56 114 25.8 31% 34
              2002 60 647 1334 0.49 99 22.0 35% 38

              Less than great players often change roles, get traded, play with
              different teammates, and just fundamentally are more affected by
              other players around them. That makes age less a predictor and more
              a constraint. Being age 33 almost always means that they won't be
              getting any better. But being age 24 vs 28 may not mean as much.

              > Much of what leads to improvement is just hard work. Dedication
              and
              > concentration are essential both to achieving excellence and
              > maintaining it. Karl Malone has to work very hard just to remain
              so
              > good. I have no doubt Corliss Williamson worked very hard on his
              > game, and luck had very little to do with his improvement.

              Hard to say. His efficiency wasn't much higher, but he basically had
              more looks.

              Season Tm G ScPoss Poss Fl % ORtg. PtsPr/G %TmP
              1996 SAC 53 140 316 0.44 87 5.2 27%
              1997 SAC 79 434 847 0.51 103 11.0 22%
              1998 SAC 79 670 1281 0.52 103 16.7 24%
              1999 SAC 50 305 583 0.52 104 12.1 21%
              2000 SAC 76 367 700 0.52 106 9.8 20%
              2001 TOR 42 180 366 0.49 97 8.5 22%
              2001 DET 27 182 336 0.54 107 13.4 21%
              2002 DET 78 474 907 0.52 107 12.4 28%

              He was 28 last year, presumably about a peak year. Efficiency-wise,
              he was about where he's been, but he took a much greater load of the
              offense. Can he do the same this coming year? Probably not, though
              probably he can get close. I wouldn't be surprised at a year with a
              rating of 105 and using 25-28% of possessions. If he's a starter,
              probably on the lower end because he'll get tired. That's a good
              year if the defense maintains its prowess.

              We seem to have 2 operational hypotheses out there on how players
              react to more responsibility. Mine generally is that players won't
              get more efficient because of it, but can get less efficient. The
              other is that players generally maintain their efficiency but their
              touches change (correct me if I'm wrong). I'd say that it's
              difficult to "prove" either one of them, but it matters in cases like
              Williamson, where he basically got more touches and maintained his
              efficiency. I would say (and JohnH would seemingly say) that
              Williamson is unlikely to maintain that, but a touch perspective
              would suggest (I think) that he should stay about the same if his
              touches stay high. With the big change at the top of the Detroit
              personnel, that could make a lot of predictions flaky. I guess we'll
              see.

              DeanO
            • John Hollinger
              What I found was that over the past five years (paraphrasing myself since I don t have the book in front of me), there were I believe 23 players who were 28
              Message 6 of 20 , Oct 11, 2002
              • 0 Attachment
                What I found was that over the past five years (paraphrasing myself
                since I don't have the book in front of me), there were I believe 23
                players who were 28 and older and improved their PER (the rating I
                use in my book) by 2.00 or more in one season. Every single one of
                them was worse the next year. Every one. On average, the sunk right
                back to the level of the year previous, which is consistent with
                the 'over their head' hypothesis.




                --- In APBR_analysis@y..., "Dean Oliver" <deano@r...> wrote:
                > --- In APBR_analysis@y..., "Mike G" <msg_53@h...> wrote:
                > > A quick look shows there are quite a few players whose peak
                years
                > > were after their 7th: Blaylock, Bogues, Elden Campbell, Cassell,
                > > Dell Curry, Antonio Davis, Gilliam, Avery Johnson, Karl Malone,
                > > Mashburn, Reggie Miller, Mourning, Shaq, Payton, Perkins, Wesley
                > > Person, Rice, Richmond, Schrempf, Thorpe, Van Exel, Webber,
                Willis,
                > > and others.
                > >
                > > It seems that 6-8 years in the league is the normal peak. But to
                > > declare that a player has certainly reached his zenith at year 7,
                > is
                > > going out on a limb.
                > >
                >
                > What you're arguing here, of course, is that age has little
                > predictive value. Actually, I think this is fairly true. I did a
                > rather extensive study of how players performed as a function of
                age
                > and saw little relationship. There is a little more strength in
                > performance vs. experience (which is different due to so many kids
                > coming out), but not much. Tying age and experience together was
                > little help, though I thought it would, with experience defining
                how
                > quickly a player reached a peak and age defining how long they
                > maintained it. Not so much.
                >
                > Great players are simultaneously more predictable and less
                > predictable than others. They are more predictable in that they
                are
                > going to get time, they are going to have offenses built around
                them
                > so they never have to adjust to different roles. They usually have
                > nicely shaped peaks on their offensive ratings, increasing early in
                > their careers (as Nowitzki has), and slowly declining. The
                > unpredictable part is exactly how high they will get when looking
                at
                > their early years, which, of course, can be key. Jordan's 2nd
                > season, with injury didn't show how good he'd get (though the
                famous
                > 63 pt playoff game might have).
                >
                > Jordan
                >
                > Season G Sc.Pos. Poss. Fl % ORtg. PP/G %
                TmPoss Age
                > 1985 82 1106 1921 0.58 118 27.6 30% 21
                > 1986 18 179 336 0.53 109 20.4 36% 22
                > 1987 82 1336 2355 0.57 117 33.6 36% 23
                > 1988 82 1292 2143 0.60 123 32.2 33% 24
                > 1989 81 1250 2104 0.59 123 31.9 32% 25
                > 1990 82 1226 2092 0.59 123 31.4 32% 26
                > 1991 82 1152 1892 0.61 126 29.0 31% 27
                > 1992 80 1088 1841 0.59 121 27.9 30% 28
                > 1993 78 1106 1938 0.57 119 29.7 33% 29
                > 1995 17 208 399 0.52 109 25.7 31% 31
                > 1996 82 1064 1833 0.58 124 27.6 31% 32
                > 1997 82 1020 1791 0.57 121 26.5 31% 33
                > 1998 82 1033 1857 0.56 114 25.8 31% 34
                > 2002 60 647 1334 0.49 99 22.0 35% 38
                >
                > Less than great players often change roles, get traded, play with
                > different teammates, and just fundamentally are more affected by
                > other players around them. That makes age less a predictor and
                more
                > a constraint. Being age 33 almost always means that they won't be
                > getting any better. But being age 24 vs 28 may not mean as much.
                >
                > > Much of what leads to improvement is just hard work. Dedication
                > and
                > > concentration are essential both to achieving excellence and
                > > maintaining it. Karl Malone has to work very hard just to remain
                > so
                > > good. I have no doubt Corliss Williamson worked very hard on his
                > > game, and luck had very little to do with his improvement.
                >
                > Hard to say. His efficiency wasn't much higher, but he basically
                had
                > more looks.
                >
                > Season Tm G ScPoss Poss Fl % ORtg.
                PtsPr/G %TmP
                > 1996 SAC 53 140 316 0.44 87 5.2 27%
                > 1997 SAC 79 434 847 0.51 103 11.0 22%
                > 1998 SAC 79 670 1281 0.52 103 16.7 24%
                > 1999 SAC 50 305 583 0.52 104 12.1 21%
                > 2000 SAC 76 367 700 0.52 106 9.8 20%
                > 2001 TOR 42 180 366 0.49 97 8.5 22%
                > 2001 DET 27 182 336 0.54 107 13.4 21%
                > 2002 DET 78 474 907 0.52 107 12.4 28%
                >
                > He was 28 last year, presumably about a peak year. Efficiency-
                wise,
                > he was about where he's been, but he took a much greater load of
                the
                > offense. Can he do the same this coming year? Probably not,
                though
                > probably he can get close. I wouldn't be surprised at a year with
                a
                > rating of 105 and using 25-28% of possessions. If he's a starter,
                > probably on the lower end because he'll get tired. That's a good
                > year if the defense maintains its prowess.
                >
                > We seem to have 2 operational hypotheses out there on how players
                > react to more responsibility. Mine generally is that players won't
                > get more efficient because of it, but can get less efficient. The
                > other is that players generally maintain their efficiency but their
                > touches change (correct me if I'm wrong). I'd say that it's
                > difficult to "prove" either one of them, but it matters in cases
                like
                > Williamson, where he basically got more touches and maintained his
                > efficiency. I would say (and JohnH would seemingly say) that
                > Williamson is unlikely to maintain that, but a touch perspective
                > would suggest (I think) that he should stay about the same if his
                > touches stay high. With the big change at the top of the Detroit
                > personnel, that could make a lot of predictions flaky. I guess
                we'll
                > see.
                >
                > DeanO
              • Mike G
                ... We aren t looking at exactly the same numbers, partly because I don t have age in my data. However, I have looked at a couple hundred players whose
                Message 7 of 20 , Oct 13, 2002
                • 0 Attachment
                  --- In APBR_analysis@y..., "John Hollinger" <alleyoop2@y...> wrote:
                  > ... over the past five years .... 23
                  > players who were 28 and older and improved their PER (the rating I
                  > use in my book) by 2.00 or more in one season. Every single one of
                  > them was worse the next year. Every one. On average, the sunk right
                  > back to the level of the year previous, which is consistent with
                  > the 'over their head' hypothesis.

                  We aren't looking at exactly the same numbers, partly because I don't
                  have 'age' in my data. However, I have looked at a couple hundred
                  players whose careers went at least 10 years.

                  In the old days, I guess age 28 would be about one's 6th-7th year in
                  the pros, on average. 6 years is long enough to judge a player, so I
                  compared these guys' first 6 years with the next 6 (or 4 or 5, if
                  they didn't play 12 years).

                  Players are divided by 'statistical eras': careers completed before
                  1974; those spanning '73-'74; and those since 1974.

                  Old-Timers (1952-73)
                  Sample size: 40

                  subgroup- min. pct. sco. reb ast PF - total
                  1st 6 yrs 32.5 .476 15.3 7.2 3.0 3.6 - 25.6
                  2nd 6 yrs 29.3 .487 14.9 6.3 3.2 3.8 - 24.5
                  avg ratio 0.88 1.02 .96 .85 1.02 1.04 - .95
                  # improve- 12 - 28 - 14 - 7 - 22

                  This confirms one old suspicion -- that younger players rebound
                  better. At least in the old days. Meanwhile, shot selection
                  improves with age, and assists go up, while scoring load goes down.


                  Group 2 (from before and after 1973-74)
                  Sample size: 95

                  subgroup- min. pct. sco. reb ast PF - total
                  1st 6 yrs 32.6 .507 15.7 6.7 3.5 3.4 - 26.0
                  2nd 6 yrs 30.3 .512 15.1 6.0 3.9 3.3 - 25.0
                  avg ratio 0.92 1.01 .95 .88 1.07 .95 - .95
                  #improved- 30 - 57 - 34 - 16- 58

                  Looks like this '70s-dominated group lost fewer minutes and rebounds,
                  and continued the tradition of better passing. The small increase in
                  shooting pct may just reflect a universal trend.


                  Group 3, after 1974 (now we have steals and blocks)
                  Sample size: 182

                  subgroup- min. pct. sco. reb ast PF- Stl Blk - total
                  1st 6 yrs 28.5 .534 15.4 6.8 3.4 3.5 1.3 0.9 - 27.8
                  2nd 6 yrs 26.4 .528 14.1 6.2 3.4 3.3 1.2 0.7 - 25.5
                  avg ratio 0.90 0.99 0.90 .90 .96 .94 .87 .75 - .91
                  #improved- 61 - 68 - 56 - 32- 85 121- 44- 33

                  This latter group (which does not include active players) bucks the
                  trend of better passing, as well as the fact that rebounds and points
                  are equally scarce.

                  The overall reduced minutes in this group suggests that more 'scrubs'
                  are playing beyond a few years. The 'avg ratio' line is geometric
                  means. (that is, the 182nd root of all 182 ratios multiplied
                  together, for a given category.) Thus, Joe Kleine was the 'most
                  improved' passer in his 2nd 6 years, though he didn't collect that
                  many more total assists.

                  In any case, I see 68 of 182 players improving their 'scoring
                  efficiency' after their first 6 years; and 56 who improved their
                  scoring output; and etc.



                  Mike G

                  > > I would say (and JohnH would seemingly say) that
                  > > Williamson is unlikely to maintain that, but a touch perspective
                  > > would suggest (I think) that he should stay about the same if his
                  > > touches stay high. With the big change at the top of the Detroit
                  > > personnel, that could make a lot of predictions flaky. I guess
                  > we'll
                  > > see.
                  > >
                  > > DeanO

                  Sorry, I butchered your response, Dean. Predicting one year is
                  conceivable; beyond that, forget it.

                  I think of NBA players similar to racing cars. They may run fine for
                  some years, but there is always intense maintenance. Then at some
                  point, they are totalled. Either mothballed, scrapped, or put to use
                  as a parade piece.
                • Dean Oliver
                  ... for ... use ... Daniel Kahneman just won a Nobel Prize (in economics) for some of his work in looking at how people perceive the world. He worked with
                  Message 8 of 20 , Oct 13, 2002
                  • 0 Attachment
                    --- In APBR_analysis@y..., "Mike G" <msg_53@h...> wrote:
                    > ... Much useful info on how players change through time...
                    >

                    > Sorry, I butchered your response, Dean. Predicting one year is
                    > conceivable; beyond that, forget it.
                    >
                    > I think of NBA players similar to racing cars. They may run fine
                    for
                    > some years, but there is always intense maintenance. Then at some
                    > point, they are totalled. Either mothballed, scrapped, or put to
                    use
                    > as a parade piece.

                    Daniel Kahneman just won a Nobel Prize (in economics) for some of his
                    work in looking at how people perceive the world. He worked with
                    Amos Tversky, who did all the hot hand studies on basketball. Among
                    the other things they were interested in is how people make and
                    evaluate predictions (which is why I'm mentioning it here). Most
                    people make a lot of predictions, considering a lot of different
                    things, but then are wrong. When they go back and look at their
                    predictions, most people say, "I should have known," or "I thought of
                    that." They just didn't give enough weight to certain factors or
                    didn't link all their thoughts. I'm looking for some of the papers
                    on this to see how they lend insight, since all I'm saying is common
                    sense.

                    Interestingly, most people say that Tversky would have shared this
                    Nobel Prize with Kahneman had he not died a few years ago (no
                    posthumous awards, I guess). That would mean that he would be the
                    first person to have won a Nobel Prize for basketball!

                    DeanO
                  • Mike G
                    ... 23 ... Constraining yourself to the last 5 years may not be representative of either the past or the future. Even so, it looks like Kurt Thomas would
                    Message 9 of 20 , Oct 14, 2002
                    • 0 Attachment
                      --- In APBR_analysis@y..., "John Hollinger" <alleyoop2@y...> wrote:
                      > What I found was that over the past five years (paraphrasing myself
                      > since I don't have the book in front of me), there were I believe
                      23
                      > players who were 28 and older and improved their PER (the rating I
                      > use in my book) by 2.00 or more in one season. Every single one of
                      > them was worse the next year. Every one. On average, the sunk right
                      > back to the level of the year previous, which is consistent with
                      > the 'over their head' hypothesis.

                      Constraining yourself to the last 5 years may not be representative
                      of either the past or the future. Even so, it looks like Kurt Thomas
                      would definitely qualify under your criteria. Elden Campbell and
                      Johnny Newman might, as well. (These 3 for the last 3 seasons).

                      I don't know what your PER measures, or what the magnitude of 2.00
                      is, relative to my own system. But speaking of which, here are the
                      aforementioned players as I had them for the last 3 years:

                      Kurt Thomas
                      yr aj min pct sco. reb ast - total
                      00 27 25 .535 12.3 `9.3 1.5 - 23.8
                      01 28 28 .556 16.0 `9.2 1.2 - 27.9
                      02 29 34 .534 16.0 10.4 1.2 - 28.5

                      'aj' is age, sized for the data column
                      rates are per-36-min, adjusted
                      pct is combined
                      total includes steals, blocks, and turnovers

                      All Thomas' rates are steady or up, and minutes are way up, so he
                      seems like a sure exception to your statement, John.


                      Elden Campbell
                      yr aj min pct sco. reb ast - total
                      00 31 33 .490 13.6 8.6 1.9 - 27.0
                      01 32 30 .493 16.1 9.6 1.7 - 29.9
                      02 33 28 .545 19.3 8.9 1.8 - 32.6

                      Campbell's minutes are down, but his scoring and efficiency are way
                      up.

                      Johnny Newman
                      yr aj min pct sco. reb ast - total
                      99 35 19 .517 11.2 2.9 1.6 - 16.4
                      00 36 22 .556 16.2 2.9 1.2 - 20.3
                      01 37 25 .531 15.3 3.1 2.0 - 20.6


                      In all, I have found about 40 players who had a year of significant
                      improvement after age 27, and followed with an even stronger year.

                      In approximate chronological order:

                      player, 3-yr-period (age)

                      Gene Shue 59-61 (27-29)
                      Johnny Kerr 61-63 (28-30)
                      Sam Jones 64-66 (30-32)
                      Connie Dierking 67-69 (30-32)
                      Elgin Baylor 66-68 (31-33)
                      Johnny Green 69-71 (35-37)
                      Darrall Imhoff 66-68 (27-29)
                      Lenny Wilkens 71-73 (33-35)
                      Bob Boozer 67-69 (29-31)
                      Tom Meschery 67-69 (28-30)
                      John Havlicek 69-71 (28-30)
                      Chet Walker 69-71 (28-30)
                      Dave DeBusschere 72-74 (31-33)
                      Kevin Loughery 68-70 (27-29)
                      Zelmo Beaty 67-72 (28-32)

                      Zelmo made his big jump between ages 28 and 29, and improved steadily
                      to age 32.

                      Keith Erickson 72-74 (27-29)
                      Jerry Sloan 73-75 (30-32)
                      Dick Snyder 73-75 (28-30)
                      Steve Mix 78-81 (30-33), see note on Zelmo
                      Artis Gilmore 80-82 (31-33)
                      Tom Owens 77-79 (27-29)
                      Julius Erving 79-81 (28-30)
                      Brian Taylor 79-81 (27-29)
                      Larry Bird 84-88 (27-31), like Zelmo
                      Tom Chambers 88-90 (28-30)
                      Sedale Threatt 91-93 (29-31)
                      Fred Roberts 89-91 (27-29)
                      Derek Harper 94-96 (32-34)
                      Kevin Willis 96-98 (33-35)
                      John Stockton 93-95 (30-32)
                      A.C. Green 91-93 (27-29)
                      Terry Porter 97-99 (33-35)
                      Karl Malone 94-97 (31-33)
                      Wayman Tisdale 92-94 (27-29)
                      Ron Harper 97-99 (32-34)
                      Johnny Newman
                      Mark Jackson 95-98 (29-32)
                      Kevin Johnson 95-97 (28-30)
                      Cliff Robinson 98-00 (31-33)
                      Dee Brown 97-99 (28-30)
                      Elden Campbell
                      Bimbo Coles 97-99 (28-30)
                      Kurt Thomas


                      John H., do any of these players qualify in your specifications?

                      I've listed about 40 players from among my 'top 527' that have made
                      a 'significant jump' after age 27, and followed with an even better
                      year. I don't know how many of the 527 had such a 'jump year' after
                      age 27, but 40 of them had at least one year of improvement
                      immediately after that.


                      Mike G
                    • John Hollinger
                      At first glance I d say most of the guys made jumps that weren t as big as the ones in my study. Elden Campbell didn t qualify in 01 but did for 02 - meaning
                      Message 10 of 20 , Oct 14, 2002
                      • 0 Attachment
                        At first glance I'd say most of the guys made jumps that weren't as
                        big as the ones in my study. Elden Campbell didn't qualify in '01 but
                        did for '02 - meaning the prediction is that he'll decline in '03.
                        Thomas didn't qualify either year, and I don't believe Cliffy did
                        either, although I'd have to doublecheck.

                        I was looking for guys who blasted off at middle age -- such as
                        Clarence Weatherspoon with Cleveland in '01, who didn't make it on
                        your list. I wasn't looking for slow, steady improvement a la Kurt
                        Thomas -- I was hunting for years that could be considered flukes,
                        and by and large, they all turned out to be exactly that.




                        --- In APBR_analysis@y..., "Mike G" <msg_53@h...> wrote:
                        > --- In APBR_analysis@y..., "John Hollinger" <alleyoop2@y...> wrote:
                        > > What I found was that over the past five years (paraphrasing
                        myself
                        > > since I don't have the book in front of me), there were I believe
                        > 23
                        > > players who were 28 and older and improved their PER (the rating
                        I
                        > > use in my book) by 2.00 or more in one season. Every single one
                        of
                        > > them was worse the next year. Every one. On average, the sunk
                        right
                        > > back to the level of the year previous, which is consistent with
                        > > the 'over their head' hypothesis.
                        >
                        > Constraining yourself to the last 5 years may not be representative
                        > of either the past or the future. Even so, it looks like Kurt
                        Thomas
                        > would definitely qualify under your criteria. Elden Campbell and
                        > Johnny Newman might, as well. (These 3 for the last 3 seasons).
                        >
                        > I don't know what your PER measures, or what the magnitude of 2.00
                        > is, relative to my own system. But speaking of which, here are the
                        > aforementioned players as I had them for the last 3 years:
                        >
                        > Kurt Thomas
                        > yr aj min pct sco. reb ast - total
                        > 00 27 25 .535 12.3 `9.3 1.5 - 23.8
                        > 01 28 28 .556 16.0 `9.2 1.2 - 27.9
                        > 02 29 34 .534 16.0 10.4 1.2 - 28.5
                        >
                        > 'aj' is age, sized for the data column
                        > rates are per-36-min, adjusted
                        > pct is combined
                        > total includes steals, blocks, and turnovers
                        >
                        > All Thomas' rates are steady or up, and minutes are way up, so he
                        > seems like a sure exception to your statement, John.
                        >
                        >
                        > Elden Campbell
                        > yr aj min pct sco. reb ast - total
                        > 00 31 33 .490 13.6 8.6 1.9 - 27.0
                        > 01 32 30 .493 16.1 9.6 1.7 - 29.9
                        > 02 33 28 .545 19.3 8.9 1.8 - 32.6
                        >
                        > Campbell's minutes are down, but his scoring and efficiency are way
                        > up.
                        >
                        > Johnny Newman
                        > yr aj min pct sco. reb ast - total
                        > 99 35 19 .517 11.2 2.9 1.6 - 16.4
                        > 00 36 22 .556 16.2 2.9 1.2 - 20.3
                        > 01 37 25 .531 15.3 3.1 2.0 - 20.6
                        >
                        >
                        > In all, I have found about 40 players who had a year of significant
                        > improvement after age 27, and followed with an even stronger year.
                        >
                        > In approximate chronological order:
                        >
                        > player, 3-yr-period (age)
                        >
                        > Gene Shue 59-61 (27-29)
                        > Johnny Kerr 61-63 (28-30)
                        > Sam Jones 64-66 (30-32)
                        > Connie Dierking 67-69 (30-32)
                        > Elgin Baylor 66-68 (31-33)
                        > Johnny Green 69-71 (35-37)
                        > Darrall Imhoff 66-68 (27-29)
                        > Lenny Wilkens 71-73 (33-35)
                        > Bob Boozer 67-69 (29-31)
                        > Tom Meschery 67-69 (28-30)
                        > John Havlicek 69-71 (28-30)
                        > Chet Walker 69-71 (28-30)
                        > Dave DeBusschere 72-74 (31-33)
                        > Kevin Loughery 68-70 (27-29)
                        > Zelmo Beaty 67-72 (28-32)
                        >
                        > Zelmo made his big jump between ages 28 and 29, and improved
                        steadily
                        > to age 32.
                        >
                        > Keith Erickson 72-74 (27-29)
                        > Jerry Sloan 73-75 (30-32)
                        > Dick Snyder 73-75 (28-30)
                        > Steve Mix 78-81 (30-33), see note on Zelmo
                        > Artis Gilmore 80-82 (31-33)
                        > Tom Owens 77-79 (27-29)
                        > Julius Erving 79-81 (28-30)
                        > Brian Taylor 79-81 (27-29)
                        > Larry Bird 84-88 (27-31), like Zelmo
                        > Tom Chambers 88-90 (28-30)
                        > Sedale Threatt 91-93 (29-31)
                        > Fred Roberts 89-91 (27-29)
                        > Derek Harper 94-96 (32-34)
                        > Kevin Willis 96-98 (33-35)
                        > John Stockton 93-95 (30-32)
                        > A.C. Green 91-93 (27-29)
                        > Terry Porter 97-99 (33-35)
                        > Karl Malone 94-97 (31-33)
                        > Wayman Tisdale 92-94 (27-29)
                        > Ron Harper 97-99 (32-34)
                        > Johnny Newman
                        > Mark Jackson 95-98 (29-32)
                        > Kevin Johnson 95-97 (28-30)
                        > Cliff Robinson 98-00 (31-33)
                        > Dee Brown 97-99 (28-30)
                        > Elden Campbell
                        > Bimbo Coles 97-99 (28-30)
                        > Kurt Thomas
                        >
                        >
                        > John H., do any of these players qualify in your specifications?
                        >
                        > I've listed about 40 players from among my 'top 527' that have made
                        > a 'significant jump' after age 27, and followed with an even better
                        > year. I don't know how many of the 527 had such a 'jump year'
                        after
                        > age 27, but 40 of them had at least one year of improvement
                        > immediately after that.
                        >
                        >
                        > Mike G
                      • Dean Oliver
                        ... Thomas ... Thomas offensive rating reached 105 in 2000 using 17% of his team s possessions when on the floor, then 111 and 19% in 2001, then 105 and 20%.
                        Message 11 of 20 , Oct 14, 2002
                        • 0 Attachment
                          --- In APBR_analysis@y..., "Mike G" <msg_53@h...> wrote:
                          > Constraining yourself to the last 5 years may not be representative
                          > of either the past or the future. Even so, it looks like Kurt
                          Thomas
                          > would definitely qualify under your criteria. Elden Campbell and
                          > Johnny Newman might, as well. (These 3 for the last 3 seasons).
                          >
                          > I don't know what your PER measures, or what the magnitude of 2.00
                          > is, relative to my own system. But speaking of which, here are the
                          > aforementioned players as I had them for the last 3 years:
                          >
                          > Kurt Thomas
                          > yr aj min pct sco. reb ast - total
                          > 00 27 25 .535 12.3 `9.3 1.5 - 23.8
                          > 01 28 28 .556 16.0 `9.2 1.2 - 27.9
                          > 02 29 34 .534 16.0 10.4 1.2 - 28.5
                          >
                          > 'aj' is age, sized for the data column
                          > rates are per-36-min, adjusted
                          > pct is combined
                          > total includes steals, blocks, and turnovers
                          >
                          > All Thomas' rates are steady or up, and minutes are way up, so he
                          > seems like a sure exception to your statement, John.
                          >

                          Thomas' offensive rating reached 105 in 2000 using 17% of his team's
                          possessions when on the floor, then 111 and 19% in 2001, then 105 and
                          20%. He's been touching the ball more, but his rating declined last
                          year, implying he is unlikely to get any better. With McDyess down
                          and Sprewell down, he's likely to try to do more, but he won't be a
                          very efficient player if he keeps using more possessions.

                          >
                          > Elden Campbell
                          > yr aj min pct sco. reb ast - total
                          > 00 31 33 .490 13.6 8.6 1.9 - 27.0
                          > 01 32 30 .493 16.1 9.6 1.7 - 29.9
                          > 02 33 28 .545 19.3 8.9 1.8 - 32.6
                          >
                          > Campbell's minutes are down, but his scoring and efficiency are way
                          > up.

                          His efficiency went back up to where it was between 96 and 99.
                          Probably not sustainable. Though there may be a story here. I'd
                          need to look to see why he bounced back so much last year. He was
                          definitely down for 2000 and 2001 in terms of efficiency (around
                          101), but was 106-109 last year and in '96-99.

                          >
                          > Johnny Newman
                          > yr aj min pct sco. reb ast - total
                          > 99 35 19 .517 11.2 2.9 1.6 - 16.4
                          > 00 36 22 .556 16.2 2.9 1.2 - 20.3
                          > 01 37 25 .531 15.3 3.1 2.0 - 20.6
                          >
                          >

                          Johnny Newman? Maybe I have inaccurate numbers for him. I do indeed
                          have his efficiency staying about where it has been throughout his
                          career -- a pretty good 109 -- but the rate at which he used
                          possessions declined immensely, from around 18% to 11%. 11% usually
                          signals end of career. He's always been a pretty miserable defender,
                          so I'd be surprised if that kind of offensive rating is going to get
                          him a slot.

                          > In all, I have found about 40 players who had a year of significant
                          > improvement after age 27, and followed with an even stronger year.
                          >
                          > In approximate chronological order:
                          >
                          > player, 3-yr-period (age)
                          >
                          > Gene Shue 59-61 (27-29)
                          > Johnny Kerr 61-63 (28-30)
                          > Sam Jones 64-66 (30-32)
                          > Connie Dierking 67-69 (30-32)
                          > Elgin Baylor 66-68 (31-33)
                          > Johnny Green 69-71 (35-37)
                          > Darrall Imhoff 66-68 (27-29)
                          > Lenny Wilkens 71-73 (33-35)
                          > Bob Boozer 67-69 (29-31)
                          > Tom Meschery 67-69 (28-30)
                          > John Havlicek 69-71 (28-30)
                          > Chet Walker 69-71 (28-30)
                          > Dave DeBusschere 72-74 (31-33)
                          > Kevin Loughery 68-70 (27-29)
                          > Zelmo Beaty 67-72 (28-32)
                          >
                          > Zelmo made his big jump between ages 28 and 29, and improved
                          steadily
                          > to age 32.'

                          Can't look at these guys due to weakness of stats.


                          >
                          > Keith Erickson 72-74 (27-29)
                          > Jerry Sloan 73-75 (30-32)
                          > Dick Snyder 73-75 (28-30)
                          > Steve Mix 78-81 (30-33), see note on Zelmo
                          > Artis Gilmore 80-82 (31-33)
                          > Tom Owens 77-79 (27-29)
                          > Julius Erving 79-81 (28-30)
                          > Brian Taylor 79-81 (27-29)

                          Advent of 3pt shot.

                          > Larry Bird 84-88 (27-31), like Zelmo

                          I think this relates to the experience vs. age thing. Most guys do
                          peak after 4-8 yrs in the league, which is close to where Bird was
                          here.

                          > Tom Chambers 88-90 (28-30)

                          His 1990 season was his best, but it's not all that different
                          efficiency-wise and possession-rate wise than what he did from '84-
                          88. Ratings all around 106-110, using 24-26% of team possessions.

                          > Sedale Threatt 91-93 (29-31)

                          With some of these guys, I looked to eyeball peak periods before
                          reading yours. With Threatt, I saw a sustained period from '87
                          to '95. He got more regular minutes in '91 to '93 and his rating did
                          go up each year, from 111.6 to 112.4 to 113.5, but his % of the
                          offense went down from 21% to 20% and 20%.

                          > Fred Roberts 89-91 (27-29)

                          His first 3 years out of Boston. His rating went 105, 108, 111 and
                          he used 15%, 17%, and 18% of team possessions. Still a role player,
                          but a better one each year. In 1992, he tried to use 20% of the
                          possessions and his rating dropped back down to 106.

                          > Derek Harper 94-96 (32-34)

                          Efficiency went down in '96 pretty bad.

                          > Kevin Willis 96-98 (33-35)

                          He was supposedly history in '96 after a horrid year. He basically
                          returned to about his previous days (a little below in '98) then
                          plummeted after '98.

                          > John Stockton 93-95 (30-32)

                          He had a down year in '93 as defenses were allowed to pretty much
                          slam anyone who came through the middle, which John does a lot.

                          > A.C. Green 91-93 (27-29)
                          > Terry Porter 97-99 (33-35)

                          Classic case of a guy whose efficiency went up when he took a lighter
                          load in the offense. He dropped below 20% tm poss usage and his
                          rating repeaked. Guys change roles in the NBA a fair amount.

                          I'm tired. And hungry. Italian food tonight, some chicken and
                          pancetta thing. Sounds better than looking at more stats.

                          DeanO

                          > Karl Malone 94-97 (31-33)
                          > Wayman Tisdale 92-94 (27-29)
                          > Ron Harper 97-99 (32-34)
                          > Johnny Newman
                          > Mark Jackson 95-98 (29-32)
                          > Kevin Johnson 95-97 (28-30)
                          > Cliff Robinson 98-00 (31-33)
                          > Dee Brown 97-99 (28-30)
                          > Elden Campbell
                          > Bimbo Coles 97-99 (28-30)
                          > Kurt Thomas
                          >
                          >
                          > John H., do any of these players qualify in your specifications?
                          >
                          > I've listed about 40 players from among my 'top 527' that have made
                          > a 'significant jump' after age 27, and followed with an even better
                          > year. I don't know how many of the 527 had such a 'jump year'
                          after
                          > age 27, but 40 of them had at least one year of improvement
                          > immediately after that.
                          >
                          >
                          > Mike G
                        • Michael K. Tamada
                          ... [...] ... Yes, in econ it s always Kahneman and Tversky who we cite. Both were/are actually more psychologists than economists (I share an office with a
                          Message 12 of 20 , Oct 15, 2002
                          • 0 Attachment
                            On Sun, 13 Oct 2002, Dean Oliver wrote:
                            > Daniel Kahneman just won a Nobel Prize (in economics) for some of his

                            [...]

                            > Interestingly, most people say that Tversky would have shared this
                            > Nobel Prize with Kahneman had he not died a few years ago (no

                            Yes, in econ it's always "Kahneman and Tversky" who we cite. Both
                            were/are actually more psychologists than economists (I share an office
                            with a psych instructor who has a list from some psych textbook or
                            organization: "100 notable pyschologists". Kahneman is on the list, I
                            think Tversky also, and also Herbert Simon, another Econ Nobel winner,
                            who's rather hard to classify: he did some econ, some cognitive science,
                            some psych. For that matter, economists like to claim John Nash, of _A
                            Beautiful Mind_ fame, but he's really more of a mathematician than an
                            economist.)

                            > posthumous awards, I guess). That would mean that he would be the

                            Correct, longevity helps.

                            > first person to have won a Nobel Prize for basketball!

                            The "Kahneman and Tversky" duo were so tightly linked in my head that I'd
                            always assumed that they had both done hot hand studies. But I guess it
                            was Tversky by himself.


                            --MKT
                          • Mike G
                            ... but ... I don t see how Campbell could qualify for 01-02, if Thomas didn t qualify for 00-01. Thomas improved his per-minute production more than
                            Message 13 of 20 , Oct 15, 2002
                            • 0 Attachment
                              --- In APBR_analysis@y..., "John Hollinger" <alleyoop2@y...> wrote:
                              > At first glance I'd say most of the guys made jumps that weren't as
                              > big as the ones in my study. Elden Campbell didn't qualify in '01
                              but
                              > did for '02 - meaning the prediction is that he'll decline in '03.
                              > Thomas didn't qualify either year, and I don't believe Cliffy did
                              > either, although I'd have to doublecheck.

                              I don't see how Campbell could qualify for '01-02, if Thomas didn't
                              qualify for '00-01. Thomas improved his per-minute production more
                              than Campbell, AND his minutes went up:

                              Kurt Thomas
                              > > yr aj min pct sco. reb ast - total
                              > > 00 27 25 .535 12.3 `9.3 1.5 - 23.8
                              > > 01 28 28 .556 16.0 `9.2 1.2 - 27.9
                              > > 02 29 34 .534 16.0 10.4 1.2 - 28.5

                              > > Elden Campbell
                              > > yr aj min pct sco. reb ast - total
                              > > 00 31 33 .490 13.6 8.6 1.9 - 27.0
                              > > 01 32 30 .493 16.1 9.6 1.7 - 29.9
                              > > 02 33 28 .545 19.3 8.9 1.8 - 32.6

                              Thomas' '00-01 improvement is +4.1 in my system, while Campbell's '01-
                              02 is +2.7. Since Thomas' minutes went up, while Campbell's went
                              down, the per-game numbers would be even more pronounced.


                              > I was looking for guys who blasted off at middle age -- such as
                              > Clarence Weatherspoon with Cleveland in '01, who didn't make it on
                              > your list. I wasn't looking for slow, steady improvement a la Kurt
                              > Thomas -- I was hunting for years that could be considered flukes,
                              > and by and large, they all turned out to be exactly that.

                              Weatherspoon from '00-'01 had basically the same scoring numbers, in
                              more minutes. Rebounding was up, a bit:

                              yr min pct sco. reb ast - total
                              00 21 .553 13.0 `9.4 2.0 - 25.6
                              01 34 .552 13.2 11.1 1.4 - 27.9
                              02 31 .470 `9.5 10.0 1.3 - 22.6

                              Again, not a huge 'jump' to begin with, so I am wondering if Spoon
                              qualifies in your survey only because he -- after the fact -- proved
                              to have a fluke year.

                              Cliff Robinson had twice that much jump in '97-98 (age 30-31), and
                              improved the next 2 years as well.


                              Mike Goodman
                            • Mike G
                              ... way ... Having a very good backup (Magloire) probably helps. Elden s foul rate has gone up, while minutes have gone down. Blocks up a bit -- unusual for
                              Message 14 of 20 , Oct 15, 2002
                              • 0 Attachment
                                --- In APBR_analysis@y..., "Dean Oliver" <deano@r...> wrote:
                                > >
                                > > Campbell's minutes are down, but his scoring and efficiency are
                                way
                                > > up.
                                >
                                > His efficiency went back up to where it was between 96 and 99.
                                > Probably not sustainable. Though there may be a story here. I'd
                                > need to look to see why he bounced back so much last year. He was
                                > definitely down for 2000 and 2001 in terms of efficiency (around
                                > 101), but was 106-109 last year and in '96-99.

                                Having a very good backup (Magloire) probably helps. Elden's foul
                                rate has gone up, while minutes have gone down. Blocks up a bit --
                                unusual for an older player. Shooting pct is at an alltime high
                                (.545 last year, career .508).

                                > >
                                > > Johnny Newman
                                > > yr aj min pct sco. reb ast - total
                                > > 99 35 19 .517 11.2 2.9 1.6 - 16.4
                                > > 00 36 22 .556 16.2 2.9 1.2 - 20.3
                                > > 01 37 25 .531 15.3 3.1 2.0 - 20.6
                                > >
                                >
                                > Johnny Newman? Maybe I have inaccurate numbers for him. I do
                                indeed
                                > have his efficiency staying about where it has been throughout his
                                > career -- a pretty good 109 -- but the rate at which he used
                                > possessions declined immensely, from around 18% to 11%. 11%
                                usually
                                > signals end of career.

                                The above data block refers to '99-01, and I guess you are looking
                                at '02.


                                > > Brian Taylor 79-81 (27-29)

                                > Advent of 3pt shot.

                                Taylor and fellow ABA-er Ron Boone started for the Kings in 1977.
                                Then Taylor had 2 really sinking years. Indeed, the 3-pointer helped
                                revive his career.

                                >
                                > > Larry Bird 84-88 (27-31), like Zelmo
                                >
                                > I think this relates to the experience vs. age thing. Most guys do
                                > peak after 4-8 yrs in the league, which is close to where Bird was
                                > here.

                                Bird had 2 years as a mere star, 3 as a mere superstar, then 4 as the
                                best of the best.

                                > > Tom Chambers 88-90 (28-30)
                                >
                                > His 1990 season was his best, but it's not all that different
                                > efficiency-wise and possession-rate wise than what he did from '84-
                                > 88. Ratings all around 106-110, using 24-26% of team possessions.

                                Enough of both adds up. His 1st year in Phoenix ('89) was also his
                                career-best as a rebounder, and his 2nd year there was the
                                astronomical scoring year.

                                >
                                > > Sedale Threatt 91-93 (29-31)
                                >
                                > With some of these guys, I looked to eyeball peak periods before
                                > reading yours. With Threatt, I saw a sustained period from '87
                                > to '95. He got more regular minutes in '91 to '93 and his rating
                                did
                                > go up each year, from 111.6 to 112.4 to 113.5, but his % of the
                                > offense went down from 21% to 20% and 20%.

                                Yep. With the Lakers, he pretty much maintained his scoring pace but
                                added the assists, which boosted him a lot (from 4.3 to 7.2, per-
                                36). Rebounds also rose.


                                > > Fred Roberts 89-91 (27-29)
                                >
                                > His first 3 years out of Boston. His rating went 105, 108, 111 and
                                > he used 15%, 17%, and 18% of team possessions. Still a role
                                player,
                                > but a better one each year. In 1992, he tried to use 20% of the
                                > possessions and his rating dropped back down to 106.

                                Roberts was a steady 5 reb/2 ast guy, while his scoring went from 10
                                to 15 to 10.

                                >
                                > > Derek Harper 94-96 (32-34)
                                >
                                > Efficiency went down in '96 pretty bad.

                                NOT!

                                Derek Harper
                                yr age pct sco. reb ast - total
                                95 32 .482 11.8 2.3 5.4 - 21.3
                                96 33 .539 13.3 2.8 6.4 - 23.8
                                97 34 .551 16.2 2.8 4.7 - 24.7

                                You might be thinking of some other Harper?


                                > > Kevin Willis 96-98 (33-35)
                                >
                                > He was supposedly history in '96 after a horrid year. He basically
                                > returned to about his previous days (a little below in '98) then
                                > plummeted after '98.

                                Yes, in '96 he was bad in Miami and worse in Golden State. In
                                Houston, he carried the team while Hakeem was hurt. Pretty good for
                                a 35-year-old.


                                > > John Stockton 93-95 (30-32)
                                >
                                > He had a down year in '93 as defenses were allowed to pretty much
                                > slam anyone who came through the middle, which John does a lot.

                                1993 was known for this?

                                Outside of '93, Stockton has had the smoothest career trajectory of
                                any player I have seen. And for the longest period.

                                > > A.C. Green 91-93 (27-29)
                                > > Terry Porter 97-99 (33-35)
                                >
                                > Classic case of a guy whose efficiency went up when he took a
                                lighter
                                > load in the offense. He dropped below 20% tm poss usage and his
                                > rating repeaked. Guys change roles in the NBA a fair amount.

                                After 3 benchwarming years in Portland and Minnesota, Porter
                                rediscovered his shot and gave up his passing duties. This is rather
                                counter to the trend of older players trying to find a niche, but it
                                worked for him.


                                > I'm tired. And hungry. Italian food tonight, some chicken and
                                > pancetta thing. Sounds better than looking at more stats.

                                Not to mention, a hungry statistician is an ugly thing. But outside
                                of the Newman and Harper misses, you shot about 80%.
                              • Dean Oliver
                                ... his ... outside ... Hey, Newman I ll grant you (though his efficiency went down in 2001 a little and, like Willis, it just points to a surprisingly bad
                                Message 15 of 20 , Oct 15, 2002
                                • 0 Attachment
                                  --- In APBR_analysis@y..., "Mike G" <msg_53@h...> wrote:
                                  > > >
                                  > > > Johnny Newman
                                  > > > yr aj min pct sco. reb ast - total
                                  > > > 99 35 19 .517 11.2 2.9 1.6 - 16.4
                                  > > > 00 36 22 .556 16.2 2.9 1.2 - 20.3
                                  > > > 01 37 25 .531 15.3 3.1 2.0 - 20.6
                                  > > >
                                  > >
                                  > > Johnny Newman? Maybe I have inaccurate numbers for him. I do
                                  > indeed
                                  > > have his efficiency staying about where it has been throughout
                                  his
                                  > > career -- a pretty good 109 -- but the rate at which he used
                                  > > possessions declined immensely, from around 18% to 11%. 11%
                                  > usually
                                  > > signals end of career.
                                  >
                                  > The above data block refers to '99-01, and I guess you are looking
                                  > at '02.
                                  > >
                                  > > > Derek Harper 94-96 (32-34)
                                  > >
                                  > > Efficiency went down in '96 pretty bad.
                                  >
                                  > NOT!
                                  >
                                  > Derek Harper
                                  > yr age pct sco. reb ast - total
                                  > 95 32 .482 11.8 2.3 5.4 - 21.3
                                  > 96 33 .539 13.3 2.8 6.4 - 23.8
                                  > 97 34 .551 16.2 2.8 4.7 - 24.7
                                  >
                                  > You might be thinking of some other Harper?
                                  >
                                  > > I'm tired. And hungry. Italian food tonight, some chicken and
                                  > > pancetta thing. Sounds better than looking at more stats.
                                  >
                                  > Not to mention, a hungry statistician is an ugly thing. But
                                  outside
                                  > of the Newman and Harper misses, you shot about 80%.

                                  Hey, Newman I'll grant you (though his efficiency went down in 2001 a
                                  little and, like Willis, it just points to a surprisingly bad year in
                                  that bad year for several, 1999).

                                  But Harper is different. He had more turnovers and fewer assists and
                                  didn't go to the line as much (relative to FGA) as the previous
                                  year. You don't seem to track turnovers much as part of your
                                  preferred stats (at least you don't post them, though I know they go
                                  in your overall power rating).


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