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

Re: Shane Battier - a great player?

Expand Messages
  • dan_t_rosenbaum
    Battier always shows up very high in my ratings. Right now, he is in my top 50 and my statistical ratings (derived from the plus/minus data) suggests that
    Message 1 of 22 , Jan 26, 2005
    • 0 Attachment
      Battier always shows up very high in my ratings. Right now, he is in
      my top 50 and my statistical ratings (derived from the plus/minus
      data) suggests that replacing an average NBA player with 40 minutes
      of Shane Battier would improve the team 3.5 points per game.
      Factoring that back down to 25-30 minutes per game, your estimate of
      an 8-10 game improvement by adding Battier is quite reasonable.

      In my stuff, I find that turnovers, steals, and blocks are very
      important. That tends to elevate players like Battier.

      My strong belief is that players who do lots of little things, but
      are not good at delivering the memorable shot, the memorable rebound,
      or memorable assist are often overlooked by non-stats folks and maybe
      even moreso by not-so-good stats folks. I think that much of the
      value of what we do in terms of player evaluation comes in finding
      these guys who do the little things.

      Best wishes,
      Dan





      --- In APBR_analysis@yahoogroups.com, bchaikin@a... wrote:
      >
      > not just this season but even last season whenever i run a
      simulation with
      > shane battier on a team that team always seems to improve alot (8-
      10 games per
      > average 82 game season). kind of hard to understand for a player
      that
      > doesn't even score 9-10 pts/g playing 25-30 min/g, and has
      touches/min of just 0.6,
      > very low for an SG/SF. i have him rated as an excellent defender,
      but even
      > if i had him rated as just an average defender he still improves
      teams by
      > upwards of 6-8 games per average 82 game season...
      >
      > looking for some proof/verification i see on _www.82games.com_
      > (http://www.82games.com) that battier does indeed have the
      highest +/- difference of all
      > grizzlie players at +11.3 points. i wanted to see if that was the
      highest +/-
      > differential of any player on a similar average team (memphis is
      close to a
      > .500 team, 23-19 as of last sunday), so looking at LAL, DET, ORL,
      HOU, MIN,
      > IND, CHI, PHI, BOS, and LAC, who are all near .500 W-L records, i
      came up with
      > these players:
      >
      > battier +11.3
      > r.hamilton +10.2
      > t.mcgrady + 7.8
      > k.garnett +11.5
      > j.tinsley +12.2
      > k.hinrich +11.1
      > l.deng +11.0
      > k.korver +10.5
      > e.brand +11.8
      > b.simmons +10.7
      >
      > i only picked players who had played at least half of their team's
      total
      > minutes, and of those that did the above group had the best +/-
      differentials.
      > battier has by far the lowest scoring average of the above group,
      and also the
      > lowest touches/min, but as you can see has just as high a +/-
      differential as
      > the best above...
      >
      > what sticks out about battier is he gets many more steals than
      turnovers
      > committed (he rarely turns the ball over), is an excellent
      rebounder (especially
      > offensive) for a SG/SF, when he does shoot can hit it (career 48%
      on 2s and
      > 38% on 3s), blocks a good amount of shots, and is a great defender
      (from what
      > i can tell). but he is not a scorer and does not dominate in any
      way on
      > offense. evidently tho his skills do translate into a good number
      of wins...
      >
      > i was curious if others have noticed this about battier, and was
      wondering
      > what others may think is the reason he doesn't play more minutes,
      especially
      > from those that may have seen him play alot...
      >
      > bob chaikin
      > _bchaikin@b..._ (mailto:bchaikin@b...)
    • John Hollinger
      ... I agree on the turnovers part, but steals and blocks? Those two stats are shown to have very little correlation with overall defensive effectiveness, so it
      Message 2 of 22 , Jan 26, 2005
      • 0 Attachment
        >
        > In my stuff, I find that turnovers, steals, and blocks are very
        > important. That tends to elevate players like Battier.

        I agree on the turnovers part, but steals and blocks? Those two stats are shown to have
        very little correlation with overall defensive effectiveness, so it puzzles me to hear you say
        that. For every Battier and Kirilenko there's a guy like Stromile Swift or Jason Williams who
        puts up decent defensive stats but doesn't play good defense.
      • Mike G
        ... average NBA player with 40 minutes ... .. ... rebound, ... maybe ... I guess I m one of these not-so-good with stats. Battier looks like a very average
        Message 3 of 22 , Jan 26, 2005
        • 0 Attachment
          --- In APBR_analysis@yahoogroups.com, "dan_t_rosenbaum"
          <rosenbaum@u...> wrote:
          >
          > Battier always shows up very high in my ratings. ...replacing an
          average NBA player with 40 minutes
          > of Shane Battier would improve the team 3.5 points per game.
          ..
          >
          > My strong belief is that players who do lots of little things, but
          > are not good at delivering the memorable shot, the memorable
          rebound,
          > or memorable assist are often overlooked by non-stats folks and
          maybe
          > even moreso by not-so-good stats folks. ..

          I guess I'm one of these not-so-good with stats. Battier looks like
          a very average player to me. No glaring weaknesses, and no
          particular strengths.

          Some teams desperately need such a player. He's like Rodney McCray
          lite. A steadying influence.

          But such players, when they lose a step, may quickly go from
          eminently useful to imminently useless. A shooter may keep on
          shooting well after his other attributes are gone. A rebounder
          keeps on banging after his shooting touch has left him.

          But a guy who sets picks and plays defense may find himself a
          journeyman in search of a niche. If he's got one now, both he and
          his team might consider themselves fortunate.
        • bchaikin@aol.com
          ... I agree on the turnovers part, but steals and blocks? Those two stats are shown to have very little correlation with overall defensive effectiveness, so
          Message 4 of 22 , Jan 26, 2005
          • 0 Attachment
            >
            > In my stuff, I find that turnovers, steals, and blocks
            are very
            > important.  That tends to elevate players like
            Battier.

            I agree on the turnovers part, but steals and blocks? Those two stats are shown to have
            very little correlation with overall defensive effectiveness, so it puzzles me to hear you say
            that... 

            fwiw i have found through simulation that steals and blocks have quite a large bearing on a team's overall defensive effectiveness. you can draw all the correlations you want showing how some teams can have good defenses with few steals and blocks and other teams can have bad defenses with alot of steals and blocks, but that is irrelevant because:
             
            (1) a team's overall defensive effectiveness encompasses far more than just steals and blocks - last season (2003-04) there was on average about 8 steals per team per 48 min game, and 5 blocked shots per team per 48 min game, but the average team had about 90 possessions per game. so at a maximum that's just 13 out of a possible 90 team possessions for a ST or BS to occur, and on average another 77 possessions that a team has to play defense on without a steal or block occurring, and....
             
            (2) a steal or block are by definition excellent defensive plays. a steal on the vast majority of occasions prevents the opposition from scoring on one team possession (on occasion can occur after one free of two free throws is made, the team gets an off reb, etc), and since the average team possession can be shown to have the value of approximately one point, that steal is the vast majority of the time preventing on average one point from being scored. a block always prevents a shot from going in, and since the average shot goes in about 44% of the time, that block reduces that to 0%. again you can make all the assumptions you want about how going for a steal or a block and not getting it puts you in a worse defensive position than if you hadn't gone for it, but that does not negate the fact that when a ST or BS do in fact occur, in and of themselves they are excellent defensive plays, because they are....
             
            For every Battier and Kirilenko there's a guy like Stromile Swift or Jason Williams who
            puts up decent defensive stats but doesn't play good defense....
            now here i wholeheartedly agree with this statement. but in the simulation i also rate players for how much they increase or decrease the FG% of the opponents they are guarding (on a scale of -5.0% being an excellent defender and a +5.0% being a very poor defender) and can thus reflect this by battier and kirilenko having low individual defensive FG%s (like a -4.0% or -5.0%) and swift and williams having very high individual defensive FG%s (like a +4.0% or +5.0%)... 
             
            for example, as of last sunday the average FG% of the league, subtracting off all BS, was:
             
            FGM/(FGA - BS) = 42664/(96023 - 5871) = 42664/90152 = .473
             
            the chi bulls were the team with the best best def FG% (minus the BS) with:
             
            1236/(3012 - 213) = 1236/2799 = .442
             
            and the charlotte bobcats the team with the worst def FG% (minus the BS) with:
             
            1354/(2915 - 202) = 1354/2713 = .499
             
            so the average player on the bulls has an individual defensive FG% of:
             
            .442 - .473 = -.031 or -3.1%
             
            and the average player on the bobcats has an individual defensive FG% of:
             
            .499 - .473 = .026 or +2.6%
             
            so as you can see a team that has many players with individual defensive FG%s in the minus region (-5.0% being the best) but that does not get alot of steals and/or blocks will assuredly be a better defensive team than a team with individual defensive FG%s in the plus region (+5.0% being the worst) but that does get alot of steals and/or blocks, because the vast majority of team possessions (something like at least 77 out of 90) do not see a ST or BS occur....
             
            (also a team that forces more turnovers than another team will on average be the better defensive team, but unfortunately the league does not keep stats for this)...
             
            one player who is a prime example of this similar to battier this season is shawn marion of the suns. right now he is turning the ball over on just 4% of his touches (0.8 - 0.9 touches/min). that is very low for a PF. simulation shows me that right now he is the 3rd to 4th best PF in the league behind tim duncan and andrei kirilenko, about equal to kevin garnett, in terms of generating wins for their respective teams, because he averages 19.5 pts/g on a scoring FG% of 56% and a very low 1.4 turnovers/game playing 40 min/g, while garnett is getting 23.0 pts/g on a scoring FG% of 55% but with 2.8 turnovers/game, twice as many as marion (but still quite low, turnovers on just 5% of his touches but 1.6 touches/min)....
             
            looking at this another way, as of this past sunday garnett had 64 steals and 114 turnovers (for a difference of -50) and marion had 88 steals and just 57 turnovers (for a difference 31). that's a difference of 81 team possessions in just 41 games, or 2 per game, between the two players. since each team possession can be shown to be worth on average about one point, this shows marion is about 2 pts/g better than garnett - just from the ST and TO. now every rating system i know of will show garnett is the vastly superior player to marion, but simulation shows marion's penchant for not turning the ball over and also getting a good number of steals is quite valuable in terms of generating wins for his team...
             
            this is kind of analogous to an argument i remember bill james once used for saying craig biggio was as valuable a player as was jim rice, if not more (this was when jim rice was hitting home runs and getting a ton of rbi's), when most would say such a statement was preposterous. but his argument went something like this - even though both players had about the same number of ABs and hits, biggio got on base by error something like 40 times in a season, while rice got on base by error something like just 5 times. rice also grounded into something like 35-40 DPs, and biggio less than 5. so james' argument was that right there biggio was on base 60-70 times more than rice in one season where both got about the same number of ABs and hits....
             
            bob chaikin
             
             
             
             
          • dan_t_rosenbaum
            First of all, I hope that at the very least you guys are reading the posts at APBRmetrics. There are lots of fascinating discussions going on over there. 368
            Message 5 of 22 , Jan 26, 2005
            • 0 Attachment
              First of all, I hope that at the very least you guys are reading the
              posts at APBRmetrics. There are lots of fascinating discussions
              going on over there. 368 posts over the last month. And I think
              the quality of the posts is every bit as high as over here.

              My work shows that when teams put guys with lots of steals and
              blocks on the floor, they tend play a lot better. Of course, this
              is an average effect, so there are exceptions. But the results are
              so striking that it suggests that steals and blocks are worth quite
              a bit more than theory would suggest.

              It appears that players with lots of steals and/or blocks tend to do
              other things, perhaps such as disrupt the offensive flow by
              pressuring the ball or altering whether a player shoots inside, that
              results in those guys, on average, being worth a lot more than a PER
              or TENDEX analysis of steals and/or blocks would suggest.

              Remember my analysis is at the individual-level. Blocks and steals
              are likely to appear less valuable in a team analysis, because their
              value will be subsumed by other statistics, such as defensive field
              goal percentage. Moreover, if your defensive effectiveness is
              largely just opponenets' PER, steals and blocks may not do a great
              job of explaining defensive effectiveness if a large fraction of
              steals and blocks are not on the guy assigned to a particular player.

              And Mike G., I was not picking on folks like you. If you are a not-
              so-good stats guy, then I am a not-so-good stats guy as well. I was
              really referring to folks who don't get much past looking at points,
              rebounds, and assists. Your disagreement about the value of Battier
              is a substantive disagreement and not simply a by-product of an
              overly simplistic method of assessment.

              Best wishes,
              Dan

              --- In APBR_analysis@yahoogroups.com, "John Hollinger"
              <alleyoop2@y...> wrote:
              >
              >
              > >
              > > In my stuff, I find that turnovers, steals, and blocks are very
              > > important. That tends to elevate players like Battier.
              >
              > I agree on the turnovers part, but steals and blocks? Those two
              stats are shown to have
              > very little correlation with overall defensive effectiveness, so
              it puzzles me to hear you say
              > that. For every Battier and Kirilenko there's a guy like Stromile
              Swift or Jason Williams who
              > puts up decent defensive stats but doesn't play good defense.
            • tajallie@hotmail.com
              Bob, Intuitively the FG% adjustment seems logical, but have you looked at stats that support the range of adjustment between +/- 5% . . is it possible if you
              Message 6 of 22 , Jan 27, 2005
              • 0 Attachment
                Bob,

                Intuitively the FG% adjustment seems logical, but have you looked at
                stats that support the range of adjustment between +/- 5% . . is it
                possible if you take away a single good defensive player all other
                players defensive rating would be effected . For example looking at
                the Bulls data from 82games Griffith appears to have a 5.5% effect on
                the entire teams FG% allowed (and 8% effect from the average). Do you
                already adjust on team basis?

                PS Although just looking the FG% allowed results in some strange
                conclusions (like Griffin being more impactful than Wallace and
                Garnett on FG%)


                --- In APBR_analysis@yahoogroups.com, bchaikin@a... wrote:
                > >
                > > In my stuff, I find that turnovers, steals, and blocks are very
                > > important. That tends to elevate players like Battier.
                >
                > I agree on the turnovers part, but steals and blocks? Those two
                stats are
                > shown to have
                > very little correlation with overall defensive effectiveness, so
                it puzzles
                > me to hear you say
                > that...
                >
                > fwiw i have found through simulation that steals and blocks have
                quite a
                > large bearing on a team's overall defensive effectiveness. you can
                draw all the
                > correlations you want showing how some teams can have good
                defenses with few
                > steals and blocks and other teams can have bad defenses with alot
                of steals
                > and blocks, but that is irrelevant because:
                >
                > (1) a team's overall defensive effectiveness encompasses far more
                than just
                > steals and blocks - last season (2003-04) there was on average
                about 8 steals
                > per team per 48 min game, and 5 blocked shots per team per 48 min
                game, but
                > the average team had about 90 possessions per game. so at a maximum
                that's
                > just 13 out of a possible 90 team possessions for a ST or BS to
                occur, and on
                > average another 77 possessions that a team has to play defense on
                without a
                > steal or block occurring, and....
                >
                > (2) a steal or block are by definition excellent defensive plays. a
                steal on
                > the vast majority of occasions prevents the opposition from scoring
                on one
                > team possession (on occasion can occur after one free of two free
                throws is
                > made, the team gets an off reb, etc), and since the average team
                possession can
                > be shown to have the value of approximately one point, that steal
                is the vast
                > majority of the time preventing on average one point from being
                scored. a
                > block always prevents a shot from going in, and since the average
                shot goes in
                > about 44% of the time, that block reduces that to 0%. again you
                can make all
                > the assumptions you want about how going for a steal or a block
                and not
                > getting it puts you in a worse defensive position than if you
                hadn't gone for it,
                > but that does not negate the fact that when a ST or BS do in fact
                occur, in
                > and of themselves they are excellent defensive plays, because
                they are....
                >
                > For every Battier and Kirilenko there's a guy like Stromile Swift
                or Jason
                > Williams who
                > puts up decent defensive stats but doesn't play good defense....
                >
                > now here i wholeheartedly agree with this statement. but in the
                simulation i
                > also rate players for how much they increase or decrease the FG% of
                the
                > opponents they are guarding (on a scale of -5.0% being an excellent
                defender and
                > a +5.0% being a very poor defender) and can thus reflect this by
                battier and
                > kirilenko having low individual defensive FG%s (like a -4.0% or -
                5.0%) and
                > swift and williams having very high individual defensive FG%s (like
                a +4.0% or
                > +5.0%)...
                >
                > for example, as of last sunday the average FG% of the league,
                subtracting
                > off all BS, was:
                >
                > FGM/(FGA - BS) = 42664/(96023 - 5871) = 42664/90152 = .473
                >
                > the chi bulls were the team with the best best def FG% (minus the
                BS) with:
                >
                > 1236/(3012 - 213) = 1236/2799 = .442
                >
                > and the charlotte bobcats the team with the worst def FG% (minus
                the BS)
                > with:
                >
                > 1354/(2915 - 202) = 1354/2713 = .499
                >
                > so the average player on the bulls has an individual defensive FG%
                of:
                >
                > .442 - .473 = -.031 or -3.1%
                >
                > and the average player on the bobcats has an individual defensive
                FG% of:
                >
                > .499 - .473 = .026 or +2.6%
                >
                > so as you can see a team that has many players with individual
                defensive
                > FG%s in the minus region (-5.0% being the best) but that does not
                get alot of
                > steals and/or blocks will assuredly be a better defensive team than
                a team with
                > individual defensive FG%s in the plus region (+5.0% being the
                worst) but
                > that does get alot of steals and/or blocks, because the vast
                majority of team
                > possessions (something like at least 77 out of 90) do not see a ST
                or BS
                > occur....
                >
                >
                > (also a team that forces more turnovers than another team will on
                average be
                > the better defensive team, but unfortunately the league does not
                keep stats
                > for this)...
                >
                >
                > one player who is a prime example of this similar to battier this
                season is
                > shawn marion of the suns. right now he is turning the ball over on
                just 4% of
                > his touches (0.8 - 0.9 touches/min). that is very low for a PF.
                simulation
                > shows me that right now he is the 3rd to 4th best PF in the league
                behind tim
                > duncan and andrei kirilenko, about equal to kevin garnett, in terms
                of
                > generating wins for their respective teams, because he averages
                19.5 pts/g on a
                > scoring FG% of 56% and a very low 1.4 turnovers/game playing 40
                min/g, while
                > garnett is getting 23.0 pts/g on a scoring FG% of 55% but with 2.8
                > turnovers/game, twice as many as marion (but still quite low,
                turnovers on just 5% of his
                > touches but 1.6 touches/min)....
                >
                > looking at this another way, as of this past sunday garnett had 64
                steals
                > and 114 turnovers (for a difference of -50) and marion had 88
                steals and just 57
                > turnovers (for a difference 31). that's a difference of 81 team
                possessions
                > in just 41 games, or 2 per game, between the two players. since
                each team
                > possession can be shown to be worth on average about one point,
                this shows
                > marion is about 2 pts/g better than garnett - just from the ST and
                TO. now every
                > rating system i know of will show garnett is the vastly superior
                player to
                > marion, but simulation shows marion's penchant for not turning the
                ball over
                > and also getting a good number of steals is quite valuable in
                terms of
                > generating wins for his team...
                >
                > this is kind of analogous to an argument i remember bill james once
                used for
                > saying craig biggio was as valuable a player as was jim rice, if
                not more
                > (this was when jim rice was hitting home runs and getting a ton of
                rbi's), when
                > most would say such a statement was preposterous. but his argument
                went
                > something like this - even though both players had about the same
                number of ABs
                > and hits, biggio got on base by error something like 40 times in a
                season,
                > while rice got on base by error something like just 5 times. rice
                also grounded
                > into something like 35-40 DPs, and biggio less than 5. so james'
                argument was
                > that right there biggio was on base 60-70 times more than rice in
                one season
                > where both got about the same number of ABs and hits....
                >
                > bob chaikin
                > _bchaikin@b..._ (mailto:bchaikin@b...)
              • bchaikin@aol.com
                Bob, Intuitively the FG% adjustment seems logical, but have you looked at stats that support the range of adjustment between +/- 5% . . is it possible if you
                Message 7 of 22 , Jan 27, 2005
                • 0 Attachment
                  Bob,

                  Intuitively the FG% adjustment seems logical, but have you looked at stats that support the range of adjustment between +/- 5% . . is it possible if you take away a single good defensive player all other players defensive rating would be effected . For example looking at
                  the Bulls data from 82games Griffith appears to have a 5.5% effect on the entire teams FG% allowed (and 8% effect from the average). Do you already adjust on team basis?

                  PS Although just looking the FG% allowed results in some strange conclusions (like Griffin being more impactful than Wallace and Garnett on FG%)
                   
                  back in the late 1980s early 1990s when i first developed the simulation software i felt i had to come up with a way to increase and/or decrease opponent FG%s for the defense to work properly - on an individual player basis, not on a team basis (the simulation does not have any "team" ratings, just player ratings, because players can be traded from team to team). at that time i looked at the previous decade's worth of data and found that in any one season the largest difference between the best and worst shooting teams was about 9%, and the largest difference between the best and worst shooting teams over the span of a few seasons was about 11%. so i picked a happy medium of about 10%, figuring no one actual team would ever have all -5.0% players on defense at any one time nor all +5.0% players. but since using the computer game one could draft up players to create teams of drafted players, this could be done so the results had to be realistic...
                   
                  if you run a simulation of say 8200 games for a team, but first switch all of the players on a team to individual defensive ratings of -5.0%s, then do the same thing but switch all of that team's players to +5.0%, you'll see a swing of upwards of 20-25 games over 82 games (like making a 30-52 team a 55-27 team, or vice versa). keep in mind this means that no other player parameters were changed, i.e. all player rebounding ratings were still the same, as were their steals and blocks ratings, etc - just the individual defensive FG% ratings. so as you can see the individual defensive FG%s do indeed have a great deal to do with the results, which they should (i.e. as the 04-05 chi bulls and charlotte bobcats data in a previous posting showed)...
                   
                  now what would be even better is if we had data showing that certain players actually increase or decrease the amount of FGAs by the players they guard, not just increase or decrease their FG%s, and that may be coming soon, but i haven't had this kind of data over the past two decades (forced turnover data would be equally welcome to)...
                   
                  and keep in mind this individual defensive FG% rating is competely separate from a player's defensive ability to block shots (or get steals). thus a player who is a -5.0% on defense, but who also blocks alot of shots, has a much greater impact on decreasing an opponent team's FG% than a player with just a -5.0% individual defensive player rating who does not block alot of shots or a +5.0% defender who does block alot of shots. a defender's individual defensive FG% rating affects only the shooting of the opponent's player he guards directly (including those he picks up on defensive switches) but his shot blocking rating affects all opponent's FGAs, not just the player he is directly guarding..
                   
                  having said that keep in mind that the bulls griffin data you see at www.82games.com and quoted above is not an individual defensive mark, but what the team did when that person was in the game (and not in the game). valuable information, but not quite the same..
                   
                  also in developing the simulation shot blocking had to be of import, i.e. whenever a shot was attempted there had to be an opportunity for it to be blocked by the defense (i.e. the vast majority of the time the average FGA is not blocked). if it was not blocked it has to have a chance to go in (based on the shooting player's FG% and the defender's ability to increase or decrease that FG%), but if blocked have no chance of going in. so in the simulation you can determine the "value" of shot blocking (along with the value of other player abilities). so for example the year mark eaton blocked 456 shots for the jazz in the early/mid 1980s, i can run the sim with him blocking all those shots, then set his shot block rating to 0.0%, and when i do this playing him 36 min/g both ways, you'll see a swing of about 8 games - i.e. his 456 blocks, and those blocks alone, represent about 8 more wins for his team...
                   
                  bob chaikin
                   
                   
                   
                   
                   

                • Coach McCormick
                  Regarding the blocked shots, I can t see them being worth 8 games a season. Do you have stats on how many of those blocks the Jazz recovered vs how many the
                  Message 8 of 22 , Jan 27, 2005
                  • 0 Attachment
                    Regarding the blocked shots, I can't see them being worth 8 games a season. Do you have stats on how many of those blocks the Jazz recovered vs how many the offense recovered vs how many went out of bounds?
                    bchaikin@... wrote:
                    Bob,

                    Intuitively the FG% adjustment seems logical, but have you looked at stats that support the range of adjustment between +/- 5% . . is it possible if you take away a single good defensive player all other players defensive rating would be effected . For example looking at
                    the Bulls data from 82games Griffith appears to have a 5.5% effect on the entire teams FG% allowed (and 8% effect from the average). Do you already adjust on team basis?

                    PS Although just looking the FG% allowed results in some strange conclusions (like Griffin being more impactful than Wallace and Garnett on FG%)
                     
                    back in the late 1980s early 1990s when i first developed the simulation software i felt i had to come up with a way to increase and/or decrease opponent FG%s for the defense to work properly - on an individual player basis, not on a team basis (the simulation does not have any "team" ratings, just player ratings, because players can be traded from team to team). at that time i looked at the previous decade's worth of data and found that in any one season the largest difference between the best and worst shooting teams was about 9%, and the largest difference between the best and worst shooting teams over the span of a few seasons was about 11%. so i picked a happy medium of about 10%, figuring no one actual team would ever have all -5.0% players on defense at any one time nor all +5.0% players. but since using the computer game one could draft up players to create teams of drafted players, this could be done so the results had to be realistic...
                     
                    if you run a simulation of say 8200 games for a team, but first switch all of the players on a team to individual defensive ratings of -5.0%s, then do the same thing but switch all of that team's players to +5.0%, you'll see a swing of upwards of 20-25 games over 82 games (like making a 30-52 team a 55-27 team, or vice versa). keep in mind this means that no other player parameters were changed, i.e. all player rebounding ratings were still the same, as were their steals and blocks ratings, etc - just the individual defensive FG% ratings. so as you can see the individual defensive FG%s do indeed have a great deal to do with the results, which they should (i.e. as the 04-05 chi bulls and charlotte bobcats data in a previous posting showed)...
                     
                    now what would be even better is if we had data showing that certain players actually increase or decrease the amount of FGAs by the players they guard, not just increase or decrease their FG%s, and that may be coming soon, but i haven't had this kind of data over the past two decades (forced turnover data would be equally welcome to)...
                     
                    and keep in mind this individual defensive FG% rating is competely separate from a player's defensive ability to block shots (or get steals). thus a player who is a -5.0% on defense, but who also blocks alot of shots, has a much greater impact on decreasing an opponent team's FG% than a player with just a -5.0% individual defensive player rating who does not block alot of shots or a +5.0% defender who does block alot of shots. a defender's individual defensive FG% rating affects only the shooting of the opponent's player he guards directly (including those he picks up on defensive switches) but his shot blocking rating affects all opponent's FGAs, not just the player he is directly guarding..
                     
                    having said that keep in mind that the bulls griffin data you see at www.82games.com and quoted above is not an individual defensive mark, but what the team did when that person was in the game (and not in the game). valuable information, but not quite the same..
                     
                    also in developing the simulation shot blocking had to be of import, i.e. whenever a shot was attempted there had to be an opportunity for it to be blocked by the defense (i.e. the vast majority of the time the average FGA is not blocked). if it was not blocked it has to have a chance to go in (based on the shooting player's FG% and the defender's ability to increase or decrease that FG%), but if blocked have no chance of going in. so in the simulation you can determine the "value" of shot blocking (along with the value of other player abilities). so for example the year mark eaton blocked 456 shots for the jazz in the early/mid 1980s, i can run the sim with him blocking all those shots, then set his shot block rating to 0.0%, and when i do this playing him 36 min/g both ways, you'll see a swing of about 8 games - i.e. his 456 blocks, and those blocks alone, represent about 8 more wins for his team...
                     
                    bob chaikin
                     
                     
                     
                     
                     


                    Do you Yahoo!?
                    Yahoo! Search presents - Jib Jab's 'Second Term'
                  • bchaikin@aol.com
                    Regarding the blocked shots, I can t see them being worth 8 games a season. Do you have stats on how many of those blocks the Jazz recovered vs how many the
                    Message 9 of 22 , Jan 27, 2005
                    • 0 Attachment
                      Regarding the blocked shots, I can't see them being worth 8 games a season. Do you have stats on how many of those blocks the Jazz recovered vs how many the offense recovered vs how many went out of bounds?
                       
                      456 blocks in a season is about 5.5 per game. if you assume half of those shots would have gone in if they weren't blocked (50%, or even 45% or 40%), and for the sake of argument assume they were just 2pters, that means eaton alone with his shot blocking ability is saving a potential 4.4 pts/g - 5.5 pts/g. that's just with his blocks, not his ability to prevent shots going in by just playing defense other than with blocks...
                       
                      you don't think preventing 4-5 pts/g is worth at least 8 wins over the span of an 82 game season?...
                       
                      bob chaikin
                       
                       
                       
                       
                       

                    • tajallie@hotmail.com
                      My question is that the simulation assigns team variance to individual player averages. Perhaps some players have much greater effects than +/- 5% even though
                      Message 10 of 22 , Jan 27, 2005
                      • 0 Attachment
                        My question is that the simulation assigns team variance to
                        individual player averages. Perhaps some players have much greater
                        effects than +/- 5% even though the overall effect of five players
                        rarely deviates by more than that. Intuitively it seems logical that
                        better defensive players will cover for poorer ones, thus smoothing
                        the effects.


                        --- In APBR_analysis@yahoogroups.com, bchaikin@a... wrote:
                        > Bob,
                        >
                        > Intuitively the FG% adjustment seems logical, but have you looked
                        at stats
                        > that support the range of adjustment between +/- 5% . . is it
                        possible if you
                        > take away a single good defensive player all other players
                        defensive rating
                        > would be effected . For example looking at
                        > the Bulls data from 82games Griffith appears to have a 5.5% effect
                        on the
                        > entire teams FG% allowed (and 8% effect from the average). Do you
                        already
                        > adjust on team basis?
                        >
                        > PS Although just looking the FG% allowed results in some strange
                        conclusions
                        > (like Griffin being more impactful than Wallace and Garnett on FG%)
                        >
                        > back in the late 1980s early 1990s when i first developed the
                        simulation
                        > software i felt i had to come up with a way to increase and/or
                        decrease opponent
                        > FG%s for the defense to work properly - on an individual player
                        basis, not on
                        > a team basis (the simulation does not have any "team" ratings,
                        just player
                        > ratings, because players can be traded from team to team). at that
                        time i
                        > looked at the previous decade's worth of data and found that in
                        any one season
                        > the largest difference between the best and worst shooting teams
                        was about 9%,
                        > and the largest difference between the best and worst shooting
                        teams over the
                        > span of a few seasons was about 11%. so i picked a happy medium of
                        about
                        > 10%, figuring no one actual team would ever have all -5.0% players
                        on defense at
                        > any one time nor all +5.0% players. but since using the computer
                        game one
                        > could draft up players to create teams of drafted players, this
                        could be done
                        > so the results had to be realistic...
                        >
                        > if you run a simulation of say 8200 games for a team, but first
                        switch all
                        > of the players on a team to individual defensive ratings of -5.0%s,
                        then do the
                        > same thing but switch all of that team's players to +5.0%, you'll
                        see a
                        > swing of upwards of 20-25 games over 82 games (like making a 30-52
                        team a 55-27
                        > team, or vice versa). keep in mind this means that no other player
                        parameters
                        > were changed, i.e. all player rebounding ratings were still the
                        same, as were
                        > their steals and blocks ratings, etc - just the individual
                        defensive FG%
                        > ratings. so as you can see the individual defensive FG%s do indeed
                        have a great
                        > deal to do with the results, which they should (i.e. as the 04-05
                        chi bulls
                        > and charlotte bobcats data in a previous posting showed)...
                        >
                        > now what would be even better is if we had data showing that
                        certain players
                        > actually increase or decrease the amount of FGAs by the players
                        they guard,
                        > not just increase or decrease their FG%s, and that may be coming
                        soon, but i
                        > haven't had this kind of data over the past two decades (forced
                        turnover data
                        > would be equally welcome to)...
                        >
                        > and keep in mind this individual defensive FG% rating is competely
                        separate
                        > from a player's defensive ability to block shots (or get steals).
                        thus a
                        > player who is a -5.0% on defense, but who also blocks alot of
                        shots, has a much
                        > greater impact on decreasing an opponent team's FG% than a player
                        with just a
                        > -5.0% individual defensive player rating who does not block alot
                        of shots or
                        > a +5.0% defender who does block alot of shots. a defender's
                        individual
                        > defensive FG% rating affects only the shooting of the opponent's
                        player he guards
                        > directly (including those he picks up on defensive switches) but
                        his shot
                        > blocking rating affects all opponent's FGAs, not just the player
                        he is directly
                        > guarding..
                        >
                        > having said that keep in mind that the bulls griffin data you see
                        at
                        > _www.82games.com_ (http://www.82games.com) and quoted above is not
                        an individual
                        > defensive mark, but what the team did when that person was in the
                        game (and not
                        > in the game). valuable information, but not quite the same..
                        >
                        > also in developing the simulation shot blocking had to be of
                        import, i.e.
                        > whenever a shot was attempted there had to be an opportunity for it
                        to be
                        > blocked by the defense (i.e. the vast majority of the time the
                        average FGA is not
                        > blocked). if it was not blocked it has to have a chance to go in
                        (based on
                        > the shooting player's FG% and the defender's ability to increase or
                        decrease
                        > that FG%), but if blocked have no chance of going in. so in the
                        simulation you
                        > can determine the "value" of shot blocking (along with the value of
                        other
                        > player abilities). so for example the year mark eaton blocked 456
                        shots for the
                        > jazz in the early/mid 1980s, i can run the sim with him blocking
                        all those
                        > shots, then set his shot block rating to 0.0%, and when i do this
                        playing him 36
                        > min/g both ways, you'll see a swing of about 8 games - i.e. his
                        456 blocks,
                        > and those blocks alone, represent about 8 more wins for his team...
                        >
                        > bob chaikin
                        > _bchaikin@b..._ (mailto:bchaikin@b...)
                      • Coach McCormick
                        the flip side is that probably 3 of those were recovered by the offensive team; so, they still could score on that possession. and, while there is the
                        Message 11 of 22 , Jan 27, 2005
                        • 0 Attachment
                          the flip side is that probably 3 of those were recovered by the offensive team; so, they still could score on that possession. and, while there is the acknowledgement that intimidation plays a factor, how many fouls were committed in the pursuit of blocked shots?
                           


                          bchaikin@... wrote:
                          Regarding the blocked shots, I can't see them being worth 8 games a season. Do you have stats on how many of those blocks the Jazz recovered vs how many the offense recovered vs how many went out of bounds?
                           
                          456 blocks in a season is about 5.5 per game. if you assume half of those shots would have gone in if they weren't blocked (50%, or even 45% or 40%), and for the sake of argument assume they were just 2pters, that means eaton alone with his shot blocking ability is saving a potential 4.4 pts/g - 5.5 pts/g. that's just with his blocks, not his ability to prevent shots going in by just playing defense other than with blocks...
                           
                          you don't think preventing 4-5 pts/g is worth at least 8 wins over the span of an 82 game season?...
                           
                          bob chaikin
                           
                           
                           
                           
                           


                          Do you Yahoo!?
                          Yahoo! Search presents - Jib Jab's 'Second Term'

                        • John Hollinger
                          The thing with both shot blocks and steals is that a missed attempt isn t recorded. If Raef LaFrentz goes for a block and misses, and as a result his man gets
                          Message 12 of 22 , Jan 28, 2005
                          • 0 Attachment
                            The thing with both shot blocks and steals is that a missed attempt
                            isn't recorded.

                            If Raef LaFrentz goes for a block and misses, and as a result his man
                            gets an easy follow dunk on the rebound, that isn't recorded. If
                            Allen Iverson goes for a steal and misses, which happens about five
                            times a game, and as a result his man gets a wide-open shot, that
                            isn't recorded either.

                            I actually think Eaton's blocks might have been as valuable as you
                            say, or at least somewhere close to it. Utah was first or second in
                            the league in FG% defense in most of his best seasons. But the thing
                            about Eaton is that he could get those blocks without jumping half
                            the time, and almost never took himself out of the play. Tim Duncan
                            is like that, too. But most of the top guys in blocks and steals
                            aren't.


                            --- In APBR_analysis@yahoogroups.com, Coach McCormick
                            <highfivehoopschool@y...> wrote:
                            > the flip side is that probably 3 of those were recovered by the
                            offensive team; so, they still could score on that possession. and,
                            while there is the acknowledgement that intimidation plays a factor,
                            how many fouls were committed in the pursuit of blocked shots?
                            >
                            >
                            >
                            > bchaikin@a... wrote:
                            > Regarding the blocked shots, I can't see them being worth 8 games a
                            season. Do you have stats on how many of those blocks the Jazz
                            recovered vs how many the offense recovered vs how many went out of
                            bounds?
                            >
                            > 456 blocks in a season is about 5.5 per game. if you assume half of
                            those shots would have gone in if they weren't blocked (50%, or even
                            45% or 40%), and for the sake of argument assume they were just
                            2pters, that means eaton alone with his shot blocking ability is
                            saving a potential 4.4 pts/g - 5.5 pts/g. that's just with his
                            blocks, not his ability to prevent shots going in by just playing
                            defense other than with blocks...
                            >
                            > you don't think preventing 4-5 pts/g is worth at least 8 wins over
                            the span of an 82 game season?...
                            >
                            > bob chaikin
                            > bchaikin@b...
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            > ---------------------------------
                            > Yahoo! Groups Links
                            >
                            > To visit your group on the web, go to:
                            > http://groups.yahoo.com/group/APBR_analysis/
                            >
                            > To unsubscribe from this group, send an email to:
                            > APBR_analysis-unsubscribe@yahoogroups.com
                            >
                            > Your use of Yahoo! Groups is subject to the Yahoo! Terms of
                            Service.
                            >
                            >
                            >
                            > ---------------------------------
                            > Do you Yahoo!?
                            > Yahoo! Search presents - Jib Jab's 'Second Term'
                          • bchaikin@aol.com
                            The thing with both shot blocks and steals is that a missed attempt isn t recorded. agreed - but, that still does not negate the fact that both a steal and a
                            Message 13 of 22 , Jan 28, 2005
                            • 0 Attachment
                              The thing with both shot blocks and steals is that a missed attempt isn't recorded.
                               
                              agreed -
                               
                              but, that still does not negate the fact that both a steal and a block - when they do indeed occur - are positive (and excellent) defensive plays. you can assume all you want that by going for a steal and block and by not getting it may put you in a worse defensive position, but are you trying to intimate that it is in some way bad to try for steals and blocks? if so i have yet to see any data supporting this (you may be quite right, i just haven't seen any data to support this - just coaching lore)...
                               
                              on the contrary, i can easily show you data (such that harvey pollack used to keep for the 76ers in his nba statistical yearbook) that for example very good shot blockers, even when they don't block a shot, many times alter the flight path of the shot such that the shot doesn'y go in. he called these "intimidations", and when he kept track of these for 76ers players like shawn bradley and manute bol, his data revealed that they stopped just as many if not more shots from going in without physically touching the shot than they did with their actual blocks....

                              If Raef LaFrentz goes for a block and misses, and as a result his man gets an easy follow dunk on the rebound, that isn't recorded.
                               
                              i'm sure this occurs, and often. but on the other hand what about all those times lafrentz goes for a block, misses the block, but forces the shooter to alter the shot just enough that it still does not go in? that is still an excellent play because then on average the defense will get about 2/3 of the available rebounds of that missed shot. again anything preventing a shot from going in in the 1st place is good defense...
                               
                              If Allen Iverson goes for a steal and misses, which happens about five times a game, and as a result his man gets a wide-open shot, that isn't recorded either.
                               
                              correct - it is not recorded. but iverson has throughout his career averaged 2.5 steals per game. that's 2.5 opponent team possessions where they do not score. if he misses another 5 times, are you saying you know the opponent will score more than 2.5 of those times? those steal attempts that are missed may result in a 5 on 4 situation, but i don't know that the opponent will score more often or not than if iverson hadn't gone for a steal. the assumption may assuredly be true, but i have yet to see any data supporting it (i don't mean from you, i mean anywhere, pollack's books, roland's website, etc)...
                               
                              if you don't get a steal but play great defense and force a shot to be missed, your opponent still has on average about a 1/3 chance to get an offensive rebound. but when you do get a steal you end their possession and stop them from scoring...

                              I actually think Eaton's blocks might have been as valuable as you say, or at least somewhere close to it. Utah was first or second in the league in FG% defense in most of his best seasons. But the thing about Eaton is that he could get those blocks without jumping half the time, and almost never took himself out of the play. Tim Duncan is like that, too. But most of the top guys in blocks and steals aren't.
                              i certainly don't believe that there is an nba coach out there that is telling their players not to go for blocks or steals because that would put their team in a worse defensive position. they always say don't commit fouls, but i doubt they say don't go for good defensive plays like ST and BS....
                               
                              also players who do get alot of steals and blocks do not commit significantly more fouls than players who do not. in a previous posting (APBR posting 19440) i showed the foul rates for players who got 200+ BS or ST in a season, 150+ BS or ST, and 100+ BS or ST, versus those that didn't, and it wasn't even 1-2% more fouls per 100 minutes on average...
                               
                              bob chaikin
                               
                               
                               
                               
                               
                               
                               
                            • Coach McCormick
                              Of course, there is no guarantee that the shot would have gone in had it not been blocked. So your argument works both ways. If you miss a steal/block and are
                              Message 14 of 22 , Jan 28, 2005
                              • 0 Attachment
                                Of course, there is no guarantee that the shot would have gone in had it not been blocked. So your argument works both ways. If you miss a steal/block and are not sure whether the miss is punished by the made basket, you cannot say the steal/block prevented a made basket. It may have prevented a missed shot.

                                bchaikin@... wrote:
                                The thing with both shot blocks and steals is that a missed attempt isn't recorded.
                                 
                                agreed -
                                 
                                but, that still does not negate the fact that both a steal and a block - when they do indeed occur - are positive (and excellent) defensive plays. you can assume all you want that by going for a steal and block and by not getting it may put you in a worse defensive position, but are you trying to intimate that it is in some way bad to try for steals and blocks? if so i have yet to see any data supporting this (you may be quite right, i just haven't seen any data to support this - just coaching lore)...
                                 
                                on the contrary, i can easily show you data (such that harvey pollack used to keep for the 76ers in his nba statistical yearbook) that for example very good shot blockers, even when they don't block a shot, many times alter the flight path of the shot such that the shot doesn'y go in. he called these "intimidations", and when he kept track of these for 76ers players like shawn bradley and manute bol, his data revealed that they stopped just as many if not more shots from going in without physically touching the shot than they did with their actual blocks....

                                If Raef LaFrentz goes for a block and misses, and as a result his man gets an easy follow dunk on the rebound, that isn't recorded.
                                 
                                i'm sure this occurs, and often. but on the other hand what about all those times lafrentz goes for a block, misses the block, but forces the shooter to alter the shot just enough that it still does not go in? that is still an excellent play because then on average the defense will get about 2/3 of the available rebounds of that missed shot. again anything preventing a shot from going in in the 1st place is good defense...
                                 
                                If Allen Iverson goes for a steal and misses, which happens about five times a game, and as a result his man gets a wide-open shot, that isn't recorded either.
                                 
                                correct - it is not recorded. but iverson has throughout his career averaged 2.5 steals per game. that's 2.5 opponent team possessions where they do not score. if he misses another 5 times, are you saying you know the opponent will score more than 2.5 of those times? those steal attempts that are missed may result in a 5 on 4 situation, but i don't know that the opponent will score more often or not than if iverson hadn't gone for a steal. the assumption may assuredly be true, but i have yet to see any data supporting it (i don't mean from you, i mean anywhere, pollack's books, roland's website, etc)...
                                 
                                if you don't get a steal but play great defense and force a shot to be missed, your opponent still has on average about a 1/3 chance to get an offensive rebound. but when you do get a steal you end their possession and stop them from scoring...

                                I actually think Eaton's blocks might have been as valuable as you say, or at least somewhere close to it. Utah was first or second in the league in FG% defense in most of his best seasons. But the thing about Eaton is that he could get those blocks without jumping half the time, and almost never took himself out of the play. Tim Duncan is like that, too. But most of the top guys in blocks and steals aren't.
                                i certainly don't believe that there is an nba coach out there that is telling their players not to go for blocks or steals because that would put their team in a worse defensive position. they always say don't commit fouls, but i doubt they say don't go for good defensive plays like ST and BS....
                                 
                                also players who do get alot of steals and blocks do not commit significantly more fouls than players who do not. in a previous posting (APBR posting 19440) i showed the foul rates for players who got 200+ BS or ST in a season, 150+ BS or ST, and 100+ BS or ST, versus those that didn't, and it wasn't even 1-2% more fouls per 100 minutes on average...
                                 
                                bob chaikin
                                 
                                 
                                 
                                 
                                 
                                 
                                 


                                Do you Yahoo!?
                                Yahoo! Mail - You care about security. So do we.
                              • Stephen Greenwell
                                If Allen Iverson goes for a steal and misses, which happens about five times a game, and as a result his man gets a wide-open shot, that isn t recorded either.
                                Message 15 of 22 , Jan 28, 2005
                                • 0 Attachment
                                  If Allen Iverson goes for a steal and misses, which happens about five
                                  times a game, and as a result his man gets a wide-open shot, that
                                  isn't recorded either.

                                  Out of curiosity, has anyone ever looked at what NBA players shoot when they're WIDE OPEN?  I'm talking about practice, game time, etc.  From anecdotal experience, I generally shoot about the same (or at least I think I do!) assuming I have a clean look and don't have to alter a shot - The distance of the guy/girl from me doesn't matter.

                                  Even if there *is* a difference, does the increase in shooting percentage negate the value of an occasional block or steal?  For example, does Iverson's man improve from a 40 to 75 percent shooter if he's left unguarded?  Can anyone shoot a jumper for this high of a percentage even if they're left wide open?  Watching players in practice, I've seen some people go on streaks of 3 to 10 threes in a row, but I've also seen this in a game where the shot sample size is much smaller.  At what point does improved shooting percentage from an out-of-position defender make up for steals and blocks generated?

                                  I doubt there are answers to all (or any) of my questions, but I'm interested to know if anyone has any thoughts.

                                  Stephen Greenwell
                                • Philip Maymin
                                  I tracked practice shots for a variety of players in shootaround before the media access times (the real shootaround as opposed to the one for the fans) and
                                  Message 16 of 22 , Jan 29, 2005
                                  • 0 Attachment
                                    I tracked practice shots for a variety of players in shootaround
                                    before the media access times (the "real" shootaround as opposed to
                                    the one for the fans) and though it seemed a little higher, especially
                                    when someone initially unexpected like Richard Jefferson started
                                    nailing consecutive very long threes halfway to halfcourt, when I
                                    compared the percentages they weren't all that different from his game
                                    performances. There's a confirmatory bias at work if you just watch
                                    someone shoot: you dismiss a couple clanks as "near hits", a bunch of
                                    clanks as "warming up," and when they hit a couple in a row you think
                                    they are 100%. Turns out your experience is about right in that the
                                    defender tends not to matter too much, though I don't have an
                                    extremely large sample size. Certainly free throw shooting didn't seem
                                    to be any different in practice than in games.


                                    On Sat, 29 Jan 2005 02:03:21 -0500, Stephen Greenwell
                                    <sgre6768@...> wrote:
                                    > If Allen Iverson goes for a steal and misses, which happens about five
                                    > times a game, and as a result his man gets a wide-open shot, that
                                    > isn't recorded either.
                                    > Out of curiosity, has anyone ever looked at what NBA players shoot when
                                    > they're WIDE OPEN? I'm talking about practice, game time, etc. From
                                    > anecdotal experience, I generally shoot about the same (or at least I think
                                    > I do!) assuming I have a clean look and don't have to alter a shot - The
                                    > distance of the guy/girl from me doesn't matter.
                                    >
                                    > Even if there *is* a difference, does the increase in shooting percentage
                                    > negate the value of an occasional block or steal? For example, does
                                    > Iverson's man improve from a 40 to 75 percent shooter if he's left
                                    > unguarded? Can anyone shoot a jumper for this high of a percentage even if
                                    > they're left wide open? Watching players in practice, I've seen some people
                                    > go on streaks of 3 to 10 threes in a row, but I've also seen this in a game
                                    > where the shot sample size is much smaller. At what point does improved
                                    > shooting percentage from an out-of-position defender make up for steals and
                                    > blocks generated?
                                    >
                                    > I doubt there are answers to all (or any) of my questions, but I'm
                                    > interested to know if anyone has any thoughts.
                                    >
                                    > Stephen Greenwell
                                    > ________________________________
                                    > Yahoo! Groups Links
                                    > To visit your group on the web, go to:
                                    > http://groups.yahoo.com/group/APBR_analysis/
                                    >
                                    > To unsubscribe from this group, send an email to:
                                    > APBR_analysis-unsubscribe@yahoogroups.com
                                    >
                                    > Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
                                  • Roland Beech
                                    We ve done some charting of contested versus uncontested shots, and yes there is quite a difference: http://www.82games.com/saccon.htm ... From: Stephen
                                    Message 17 of 22 , Jan 29, 2005
                                    • 0 Attachment
                                      We've done some charting of contested versus uncontested shots, and yes there is quite a difference:
                                       
                                      ----- Original Message -----
                                      Sent: Friday, January 28, 2005 11:03 PM
                                      Subject: Re: [APBR_analysis] Re: Shane Battier - a great player?

                                      If Allen Iverson goes for a steal and misses, which happens about five
                                      times a game, and as a result his man gets a wide-open shot, that
                                      isn't recorded either.

                                      Out of curiosity, has anyone ever looked at what NBA players shoot when they're WIDE OPEN?  I'm talking about practice, game time, etc.  From anecdotal experience, I generally shoot about the same (or at least I think I do!) assuming I have a clean look and don't have to alter a shot - The distance of the guy/girl from me doesn't matter.

                                      Even if there *is* a difference, does the increase in shooting percentage negate the value of an occasional block or steal?  For example, does Iverson's man improve from a 40 to 75 percent shooter if he's left unguarded?  Can anyone shoot a jumper for this high of a percentage even if they're left wide open?  Watching players in practice, I've seen some people go on streaks of 3 to 10 threes in a row, but I've also seen this in a game where the shot sample size is much smaller.  At what point does improved shooting percentage from an out-of-position defender make up for steals and blocks generated?

                                      I doubt there are answers to all (or any) of my questions, but I'm interested to know if anyone has any thoughts.

                                      Stephen Greenwell
                                    • Coach McCormick
                                      I ve seen Peja Stojakovic make about 25-30 threes in a row in practice. Heck, I saw Mateen Cleaves make 10 in a row... Philip Maymin
                                      Message 18 of 22 , Jan 29, 2005
                                      • 0 Attachment
                                        I've seen Peja Stojakovic make about 25-30 threes in a row in practice. Heck, I saw Mateen Cleaves make 10 in a row...

                                        Philip Maymin <pmaymin@...> wrote:
                                        I tracked practice shots for a variety of players in shootaround
                                        before the media access times (the "real" shootaround as opposed to
                                        the one for the fans) and though it seemed a little higher, especially
                                        when someone initially unexpected like Richard Jefferson started
                                        nailing consecutive very long threes halfway to halfcourt, when I
                                        compared the percentages they weren't all that different from his game
                                        performances. There's a confirmatory bias at work if you just watch
                                        someone shoot: you dismiss a couple clanks as "near hits", a bunch of
                                        clanks as "warming up," and when they hit a couple in a row you think
                                        they are 100%. Turns out your experience is about right in that the
                                        defender tends not to matter too much, though I don't have an
                                        extremely large sample size. Certainly free throw shooting didn't seem
                                        to be any different in practice than in games.


                                        On Sat, 29 Jan 2005 02:03:21 -0500, Stephen Greenwell
                                        <sgre6768@...> wrote:
                                        > If Allen Iverson goes for a steal and misses, which happens about five
                                        > times a game, and as a result his man gets a wide-open shot, that
                                        > isn't recorded either.
                                        > Out of curiosity, has anyone ever looked at what NBA players shoot when
                                        > they're WIDE OPEN?  I'm talking about practice, game time, etc.  From
                                        > anecdotal experience, I generally shoot about the same (or at least I think
                                        > I do!) assuming I have a clean look and don't have to alter a shot - The
                                        > distance of the guy/girl from me doesn't matter.
                                        >
                                        > Even if there *is* a difference, does the increase in shooting percentage
                                        > negate the value of an occasional block or steal?  For example, does
                                        > Iverson's man improve from a 40 to 75 percent shooter if he's left
                                        > unguarded?  Can anyone shoot a jumper for this high of a percentage even if
                                        > they're left wide open?  Watching players in practice, I've seen some people
                                        > go on streaks of 3 to 10 threes in a row, but I've also seen this in a game
                                        > where the shot sample size is much smaller.  At what point does improved
                                        > shooting percentage from an out-of-position defender make up for steals and
                                        > blocks generated?
                                        >
                                        > I doubt there are answers to all (or any) of my questions, but I'm
                                        > interested to know if anyone has any thoughts.
                                        >
                                        > Stephen Greenwell
                                        > ________________________________
                                        > Yahoo! Groups Links
                                        > To visit your group on the web, go to:
                                        > http://groups.yahoo.com/group/APBR_analysis/
                                        >  
                                        > To unsubscribe from this group, send an email to:
                                        > APBR_analysis-unsubscribe@yahoogroups.com
                                        >  
                                        > Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.


                                        Do you Yahoo!?
                                        The all-new My Yahoo! � Get yours free!
                                      • John Hollinger
                                        I m with you on that -- I ve seen NBA players (and not even good ones) make an amazing number of consecutive 3s in warm ups. ... practice. Heck, I saw Mateen
                                        Message 19 of 22 , Jan 30, 2005
                                        • 0 Attachment
                                          I'm with you on that -- I've seen NBA players (and not even good
                                          ones) make an amazing number of consecutive 3s in warm ups.


                                          --- In APBR_analysis@yahoogroups.com, Coach McCormick
                                          <highfivehoopschool@y...> wrote:
                                          > I've seen Peja Stojakovic make about 25-30 threes in a row in
                                          practice. Heck, I saw Mateen Cleaves make 10 in a row...
                                          >
                                          > Philip Maymin <pmaymin@g...> wrote:I tracked practice shots for a
                                          variety of players in shootaround
                                          > before the media access times (the "real" shootaround as opposed to
                                          > the one for the fans) and though it seemed a little higher,
                                          especially
                                          > when someone initially unexpected like Richard Jefferson started
                                          > nailing consecutive very long threes halfway to halfcourt, when I
                                          > compared the percentages they weren't all that different from his
                                          game
                                          > performances. There's a confirmatory bias at work if you just watch
                                          > someone shoot: you dismiss a couple clanks as "near hits", a bunch
                                          of
                                          > clanks as "warming up," and when they hit a couple in a row you
                                          think
                                          > they are 100%. Turns out your experience is about right in that the
                                          > defender tends not to matter too much, though I don't have an
                                          > extremely large sample size. Certainly free throw shooting didn't
                                          seem
                                          > to be any different in practice than in games.
                                          >
                                          >
                                          > On Sat, 29 Jan 2005 02:03:21 -0500, Stephen Greenwell
                                          > <sgre6768@p...> wrote:
                                          > > If Allen Iverson goes for a steal and misses, which happens about
                                          five
                                          > > times a game, and as a result his man gets a wide-open shot, that
                                          > > isn't recorded either.
                                          > > Out of curiosity, has anyone ever looked at what NBA players
                                          shoot when
                                          > > they're WIDE OPEN? I'm talking about practice, game time, etc.
                                          From
                                          > > anecdotal experience, I generally shoot about the same (or at
                                          least I think
                                          > > I do!) assuming I have a clean look and don't have to alter a
                                          shot - The
                                          > > distance of the guy/girl from me doesn't matter.
                                          > >
                                          > > Even if there *is* a difference, does the increase in shooting
                                          percentage
                                          > > negate the value of an occasional block or steal? For example,
                                          does
                                          > > Iverson's man improve from a 40 to 75 percent shooter if he's left
                                          > > unguarded? Can anyone shoot a jumper for this high of a
                                          percentage even if
                                          > > they're left wide open? Watching players in practice, I've seen
                                          some people
                                          > > go on streaks of 3 to 10 threes in a row, but I've also seen this
                                          in a game
                                          > > where the shot sample size is much smaller. At what point does
                                          improved
                                          > > shooting percentage from an out-of-position defender make up for
                                          steals and
                                          > > blocks generated?
                                          > >
                                          > > I doubt there are answers to all (or any) of my questions, but I'm
                                          > > interested to know if anyone has any thoughts.
                                          > >
                                          > > Stephen Greenwell
                                          > > ________________________________
                                          > > Yahoo! Groups Links
                                          > > To visit your group on the web, go to:
                                          > > http://groups.yahoo.com/group/APBR_analysis/
                                          > >
                                          > > To unsubscribe from this group, send an email to:
                                          > > APBR_analysis-unsubscribe@yahoogroups.com
                                          > >
                                          > > Your use of Yahoo! Groups is subject to the Yahoo! Terms of
                                          Service.
                                          >
                                          >
                                          > ---------------------------------
                                          > Yahoo! Groups Links
                                          >
                                          > To visit your group on the web, go to:
                                          > http://groups.yahoo.com/group/APBR_analysis/
                                          >
                                          > To unsubscribe from this group, send an email to:
                                          > APBR_analysis-unsubscribe@yahoogroups.com
                                          >
                                          > Your use of Yahoo! Groups is subject to the Yahoo! Terms of
                                          Service.
                                          >
                                          >
                                          >
                                          >
                                          > ---------------------------------
                                          > Do you Yahoo!?
                                          > The all-new My Yahoo! – Get yours free!
                                        • Dean Oliver
                                          Even worse, I ve seen me make 8 in a row. A hand in the face matters. How you do it matters. When you do it matters. Defense is what I did... because I
                                          Message 20 of 22 , Jan 30, 2005
                                          • 0 Attachment
                                            Even worse, I've seen me make 8 in a row.

                                            A hand in the face matters. How you do it matters. When you do it
                                            matters. Defense is what I did... because I certainly couldn't make 8
                                            in a row in a game.

                                            DeanO

                                            Dean Oliver
                                            Consultant to the Seattle Supersonics
                                            Author, Basketball on Paper
                                            http://www.basketballonpaper.com
                                            "Oliver's book provides an insightful framework for basketball. His
                                            approach highlights and simplifies the basic goals of team offenses
                                            and defenses, with an interesting description of how teamwork among
                                            players with different roles can be evaluated. This book is a unique
                                            and surprisingly practical addition to a coach's library." Dean
                                            Smith, Hall of Fame Basketball Coach, University of North Carolina


                                            --- In APBR_analysis@yahoogroups.com, Coach McCormick
                                            <highfivehoopschool@y...> wrote:
                                            > I've seen Peja Stojakovic make about 25-30 threes in a row in
                                            practice. Heck, I saw Mateen Cleaves make 10 in a row...
                                            >
                                            > Philip Maymin <pmaymin@g...> wrote:I tracked practice shots for a
                                            variety of players in shootaround
                                            > before the media access times (the "real" shootaround as opposed to
                                            > the one for the fans) and though it seemed a little higher, especially
                                            > when someone initially unexpected like Richard Jefferson started
                                            > nailing consecutive very long threes halfway to halfcourt, when I
                                            > compared the percentages they weren't all that different from his game
                                            > performances. There's a confirmatory bias at work if you just watch
                                            > someone shoot: you dismiss a couple clanks as "near hits", a bunch of
                                            > clanks as "warming up," and when they hit a couple in a row you think
                                            > they are 100%. Turns out your experience is about right in that the
                                            > defender tends not to matter too much, though I don't have an
                                            > extremely large sample size. Certainly free throw shooting didn't seem
                                            > to be any different in practice than in games.
                                            >
                                            >
                                            > On Sat, 29 Jan 2005 02:03:21 -0500, Stephen Greenwell
                                            > <sgre6768@p...> wrote:
                                            > > If Allen Iverson goes for a steal and misses, which happens about
                                            five
                                            > > times a game, and as a result his man gets a wide-open shot, that
                                            > > isn't recorded either.
                                            > > Out of curiosity, has anyone ever looked at what NBA players shoot
                                            when
                                            > > they're WIDE OPEN? I'm talking about practice, game time, etc. From
                                            > > anecdotal experience, I generally shoot about the same (or at
                                            least I think
                                            > > I do!) assuming I have a clean look and don't have to alter a shot
                                            - The
                                            > > distance of the guy/girl from me doesn't matter.
                                            > >
                                            > > Even if there *is* a difference, does the increase in shooting
                                            percentage
                                            > > negate the value of an occasional block or steal? For example, does
                                            > > Iverson's man improve from a 40 to 75 percent shooter if he's left
                                            > > unguarded? Can anyone shoot a jumper for this high of a
                                            percentage even if
                                            > > they're left wide open? Watching players in practice, I've seen
                                            some people
                                            > > go on streaks of 3 to 10 threes in a row, but I've also seen this
                                            in a game
                                            > > where the shot sample size is much smaller. At what point does
                                            improved
                                            > > shooting percentage from an out-of-position defender make up for
                                            steals and
                                            > > blocks generated?
                                            > >
                                            > > I doubt there are answers to all (or any) of my questions, but I'm
                                            > > interested to know if anyone has any thoughts.
                                            > >
                                            > > Stephen Greenwell
                                            > > ________________________________
                                            > > Yahoo! Groups Links
                                            > > To visit your group on the web, go to:
                                            > > http://groups.yahoo.com/group/APBR_analysis/
                                            > >
                                            > > To unsubscribe from this group, send an email to:
                                            > > APBR_analysis-unsubscribe@yahoogroups.com
                                            > >
                                            > > Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
                                            >
                                            >
                                            > ---------------------------------
                                            > Yahoo! Groups Links
                                            >
                                            > To visit your group on the web, go to:
                                            > http://groups.yahoo.com/group/APBR_analysis/
                                            >
                                            > To unsubscribe from this group, send an email to:
                                            > APBR_analysis-unsubscribe@yahoogroups.com
                                            >
                                            > Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
                                            >
                                            >
                                            >
                                            >
                                            > ---------------------------------
                                            > Do you Yahoo!?
                                            > The all-new My Yahoo! – Get yours free!
                                          Your message has been successfully submitted and would be delivered to recipients shortly.