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

Re: [APBR_analysis] The Allen Iverson Project

Expand Messages
  • NYFan@aol.com
    This can be seen two ways. A)Lue played more aggressively knowing he was getting help... B) Lue played more aggressive FORCING Al into more double teams. I
    Message 1 of 10 , Oct 16, 2001
    • 0 Attachment
      This can be seen two ways. A)Lue played more aggressively knowing he was getting help... B) Lue played more aggressive FORCING Al into more double teams. I play a lot of 2-2 , 3-3 basketball, and I consider myself a master of defense in those situations. The key to defense in those situations is to force your opponent to dribble into the defense, where your one or two other guys can help you out, while not leaving their man alone. The NBA is very similar because of the zone restrictions. The best teams, force their opponents to come into a double team (watch the Knicks, they force the opponent into the middle of the court, and then from there he is generally double or triple teamed depending on talent, this not only puts him in a tough situation, it allows enough defenders in good position to still make a play on any passes he should make), and I think while Lue may have been more aggressive, that aggressiveness forced Al into more double teams. Kobe tried to play him straight up one on one, but as the results showed, that's not as effective, same for Fisher. Anyway, that's just another point of view.

      ~Ray

      In a message dated 10/16/01 9:43:37 PM Eastern Daylight Time, smckibbi@... writes:


      . It's interesting that the Lakers doubled Al more times when Lue guarded him (25) than they did with Fish (23), even though Fish guarded him for twice as many possessions. I suspect this scheme allowed Lue to play Al more agressively knowing that the help was coming.


    • Dean Oliver
      ... Iverson. ... defensive ... Potentially useful stuff, especially if we start comparing/contrasting players and the defenses that work on them. ... AL shot
      Message 2 of 10 , Oct 17, 2001
      • 0 Attachment
        --- In APBR_analysis@y..., "McKibbin, Stuart" <smckibbi@c...> wrote:
        > Over the summer I've charted the Lakers-Sixers Finals series like an
        > assistant coach would, trying to analyse how the Lakers played Al
        Iverson.
        > Maybe it's a method that might be used to analyze what's a good
        defensive
        > player, maybe it's not, but here are the results:

        Potentially useful stuff, especially if we start
        comparing/contrasting players and the defenses that work on them.

        >
        >
        > Fisher guarded Al on 230 possessions, and allowed Al 174 touches.
        AL shot
        > 35-87 (40.2%) against Fish. Fish fouled him 17 times giving up just
        13 FTA's
        > (a lot of the fouls were "tempo" fouls early in a quarter before
        the bonus
        > kicked in). Fish had 8 steals and 1 block.
        >

        174 touches = 87 shots + 17 fouls (though some may overlap with shots
        on 3-pt plays) + 8 turnovers (?) + 62 passes?

        assists?

        >
        > Lue guarded Al on 108 possessions and allowed Al 82 touches. Al
        shot 9-35
        > (25.7%) against Lue. Lue fouled him 6 times giving up 7 FTA's. Lue
        got 8
        > steals, 1 block.
        >

        82 = 35 + 6 fouls + 8 tov? + 33 passes?

        assists?

        >
        > Kobe guardes Al on 54 possessions and allowed Al 42 touches. Al
        shot 7-18
        > (38.8%) against Kobe. Kobe fouled him 4 times giving up 6FTA's.
        >

        42 = 18 + 4 + ??

        >
        > Al was unguarded 22 times during the 5 games, he went 10-16 (62.5%)
        in that
        > situation. The Lakers improved as the series went on in that matter-
        --Al was
        > unguarded 9 times just in Game 1 (like when he busted that three in
        OT).

        It is unguarded situations that complicate the evaluation of
        individual defense. Whose responsibility was it during those 22
        times? This sometimes happens out of bad rotations. Sometimes off
        fast breaks. I generally do assign these to be TEAM defensive
        lapses, effectively as you have.

        >
        >
        > But scheme is everything. It wasn't a one-on-one battle, there were
        > teammates out there. It's interesting that the Lakers doubled Al
        more times
        > when Lue guarded him (25) than they did with Fish (23), even though
        Fish
        > guarded him for twice as many possessions. I suspect this scheme
        allowed Lue
        > to play Al more agressively knowing that the help was coming.
        Anyway, Al was
        > doubled 58 times, he passed 39 times, was fouled by the doubler 7
        times
        > (good for 8 FTA's), turned it over 4 times, and went 1-8 shooting
        the ball.
        > So doubling is good strategy (duh!). By the way, doubling seemed to
        be the
        > only way to make Al pass the ball, passing it 39 times in 58
        touches is 67%.

        58 = 8 fga + 7 fouls + 4 to + 39 pass OK.

        Another thing that has been proposed for players like Iverson is that
        the doubling and major attention allows easier offensive rebounds or
        easier shots for teammates. Did you track either of those?


        > Helping out after the original defender got beat on the drive was
        markedly
        > less successful, usually because after Al had beat the original
        defender he
        > could use his speed to avoid the help defender. In 61 help
        opportunities
        > Iverson shot 16-36 (44.4%). The help defender(s) blocked six of the
        > attempts, fouled Al another 7 times for 11 FTA's, made him pass it
        18 times
        > (for 3 assists) and got one measley turnover.
        >
        >
        > Alright, so much for the defense. I'm still kind of doubtful of the
        > usefulness of the analysis. It seems to me that even this kind of
        scouting
        > analysis is limited. I mean, we just know Lue can guard this type
        of player
        > provided his team helps him out with doubles 30% of the time that
        his man
        > touches it. I don't know if it means Lue is a "good defender".

        The potential value of this is large. One of the big things I've
        struggled to quantify is how well players do one-on-one. We know
        that team defenses will cover Jerry Stackhouse and Iverson more
        intently even though neither shoots the ball very well. Neither is
        particularly efficient, unlike Kobe or Vince Carter for example. But
        all are "productive". What I think is useful is determining how
        often a player is double-teamed, how big a difference it makes in
        their offensive efficiency (something I can calculate if you fill in
        some of the question marks), and how big a difference it makes in
        their teammates efficiency (the hard part).... It also is
        potentially useful in calculating defensive grades for these 3
        individuals. We qualitatively felt that Lue did OK on Iverson, but
        I'm not sure whether these #'s verify that. Again, filling in
        the ??'s helps.

        Dean Oliver
        Journal of Basketball Studies
      • McKibbin, Stuart
        Dean: Thank you for your interest. Maybe I should explain what I charted. I kept track of his touches, whether they were in half court, advancing the ball, in
        Message 3 of 10 , Oct 17, 2001
        • 0 Attachment
          Dean:
          Thank you for your interest.
          Maybe I should explain what I charted.

          I kept track of his touches, whether they were in half court, advancing the ball, in transition or from an inbounds pass in half court. Sometimes advances and transition touches are the same thing and I made a note of that.

          I kept track of who the defender was, that is who was facing up to Al between him and the basket. Almost always when Al was unguarded it was in transition (like that 3 in OT, technically he was Lue's man but Lue was sprawled out of bounds after missing a godawful layup attempt and no one picked up Al). I kept track of switches, even though it very rarely happened. My e-mail left out certain possessions where Shaw guarded Iverson, or Ho Grant but I can add those up pretty quickly.

          I kept track of how Iverson shot the ball, that is whether it was a driving layup attempt, a drive and a pull up jumper or if he shot the ball immediately after receiving it. If Al got the ball and F'd around dribbling without going anywhere (usually trying to set up his man with a high screen) and then shot it without penetration i lumped it in with shooting a jumper immediately. If he was fouled while shooting and missed I didn't count that as a shot. But I noted what the foul was (hold, in the act, etc.)

          I kept track if he passed and if he drove then passed.

          I kept track of doubles and help. The distinction I made was that if the defender came over before AL attempted to penetrate it was a double, if the defense collapsed on him during penetration or showed out on the pick and roll I called that Help. I kept track of who the doubler or help defender were. And when I could tell I made a note if it was the doubler's man that ended up scoring on the possession.

          I kept track of assists and "assist-like" plays. That is, if Al passed to Ty Hill and Hill was fouled on his layup attempt Al got an "assist-like" credit.

          In some ad hoc notes I kept track if good denial defense was played or if the Sixers were running High pick and roll or if Al was just standing around resting. I wasn't as thorough as i'd should have been about that.

          And finally I kept track of offensive rebounds on Al's shots. That is actually why I started doing this ---because I heard Doug Collins bray about the "Allen Iverson effect", that his misses were just like passes. I thought that that was a bunch of shit so I decided to check. First thing I noticed is that the official offensive rebound definition stinks, sort of like Doug. So I did NOT count recovered block shots as an offensive rebound, nor did I count Al chasing down his own miss as an offensive rebound. I found that of Al's 91 missed shots his teammates corraled 18 of them, 19.7%. Sorry Doug, Al's misses are no more valuable than anybody else's misses. But those 18 off rebs resulted in 32 points (maybe the offensive rebound as I've defined it is a little more valuable than Tendex gives it credit for).

          Anyway, I'll reply a little later to fill in the ??? you had.

          Best Wishes
        • McKibbin, Stuart
          Dean: Thank you for your interest. Maybe I should explain what I charted. I kept track of his touches, whether they were in half court, advancing the ball, in
          Message 4 of 10 , Oct 17, 2001
          • 0 Attachment
            Dean:
            Thank you for your interest.
            Maybe I should explain what I charted.

            I kept track of his touches, whether they were in half court, advancing the
            ball, in transition or from an inbounds pass in half court. Sometimes
            advances and transition touches are the same thing and I made a note of
            that.

            I kept track of who the defender was, that is who was facing up to Al
            between him and the basket. Almost always when Al was unguarded it was in
            transition (like that 3 in OT, technically he was Lue's man but Lue was
            sprawled out of bounds after missing a godawful layup attempt and no one
            picked up Al). I kept track of switches, even though it very rarely
            happened. My e-mail left out certain possessions where Shaw guarded Iverson,
            or Ho Grant but I can add those up pretty quickly.

            I kept track of how Iverson shot the ball, that is whether it was a driving
            layup attempt, a drive and a pull up jumper or if he shot the ball
            immediately after receiving it. If Al got the ball and F'd around dribbling
            without going anywhere (usually trying to set up his man with a high screen)
            and then shot it without penetration i lumped it in with shooting a jumper
            immediately. If he was fouled while shooting and missed I didn't count that
            as a shot. But I noted what the foul was (hold, in the act, etc.)

            I kept track if he passed and if he drove then passed.

            I kept track of doubles and help. The distinction I made was that if the
            defender came over before AL attempted to penetrate it was a double, if the
            defense collapsed on him during penetration or showed out on the pick and
            roll I called that Help. I kept track of who the doubler or help defender
            were. And when I could tell I made a note if it was the doubler's man that
            ended up scoring on the possession.

            I kept track of assists and "assist-like" plays. That is, if Al passed to Ty
            Hill and Hill was fouled on his layup attempt Al got an "assist-like"
            credit.

            In some ad hoc notes I kept track if good denial defense was played or if
            the Sixers were running High pick and roll or if Al was just standing around
            resting. I wasn't as thorough as i'd should have been about that.

            And finally I kept track of offensive rebounds on Al's shots. That is
            actually why I started doing this ---because I heard Doug Collins bray about
            the "Allen Iverson effect", that his misses were just like passes. I thought
            that that was a bunch of shit so I decided to check. First thing I noticed
            is that the official offensive rebound definition stinks, sort of like Doug.
            So I did NOT count recovered block shots as an offensive rebound, nor did I
            count Al chasing down his own miss as an offensive rebound. I found that of
            Al's 91 missed shots his teammates corraled 18 of them, 19.7%. Sorry Doug,
            Al's misses are no more valuable than anybody else's misses. But those 18
            off rebs resulted in 32 points (maybe the offensive rebound as I've defined
            it is a little more valuable than Tendex gives it credit for).

            Anyway, I'll reply a little later to fill in the ??? you had.

            Best Wishes
          • McKibbin, Stuart
            Perhaps a summary of Iverson is in order. 452 possessions. 328 touches. 86 of the touches came from Iverson advancing the ball. 67 of the touches were in
            Message 5 of 10 , Oct 21, 2001
            • 0 Attachment
              Perhaps a summary of Iverson is in order.
               
              452 possessions. 328 touches.
              86 of the touches came from Iverson advancing the ball. 67 of the touches were in transition. Transition touches and advancing touches overlapped 30 times.
               
              Overall, Al shot the 162 times (162/328 = 49.3%). He was 33-85 with jumpshots, including 11 treys for a true shooting percentage of 45.3%. He was 15-33 (45.5%) when he drove and then pulled up for the jumpshot. He was 18-44 (41.0%) when he drove to the basket for a layup.
               
              However, when he was in transition, and many times unguarded, his shooting percentage went way up. In transition situations he was 18-31 (58%).  3-5 with jumpshots, tsp = 70%, 5-9 with the pullup jumper, 10-17 taking it to the rack. So when Allen took it all the way to the basket and wasn't in transition he was a mere 8-27 (29.6%), demonstrating that in the half court at least it's still a big man's game. But in the open court Al's speed is a hell of an advantage.
               
              On his 86 advances Iverson passed the ball 37 times, shot it 33 times, and was fouled 7 times. The other 8 advances consists of various misc. happenings.
               
              Al passed the ball 109 times, (109/328 = 33.2%). 17 assists and 12 "assist-like" plays. .
               
              He had 11 turnovers (11/328 = 3.3%) and was fouled 32 times with the ball (32/328 = 9.7%) and was fouled 7 times away from the ball. 14 other miscellaneous things (like jumpballs, deflections, timeouts, away from the ball fouls, illegal screens, etc.) happened to round out the 328.
               
              There were 170 possessions when Al did not touched the ball. Philly scored 148 points or 0.87 pts/poss.
              There were 119 possessions when Al touched the ball and didn't get a FGA or a FTA. Philly scored 116 points or 0.97 pts/poss.
              There were 172 possesions were Al had either a FGA or FTA. Philly scored 203 points or 1.18 pts/poss. 32 of the 203 points came from Philly hitting the offensive boards (including Al recovering his own shot, so there is a little double counting here).
              I don't count the technicals Al shot as part of any possession.
            • Mike Goodman
              ... touches ... overlapped 30 ... 45.3%. He ... jumpshot. He was ... Stuart, have you broken down the FT attempts resulting from these 3 categories of shot
              Message 6 of 10 , Oct 22, 2001
              • 0 Attachment
                --- In APBR_analysis@y..., "McKibbin, Stuart" <smckibbi@c...> wrote:
                > Perhaps a summary of Iverson is in order.
                >
                > 452 possessions. 328 touches.
                > 86 of the touches came from Iverson advancing the ball. 67 of the
                touches
                > were in transition. Transition touches and advancing touches
                overlapped 30
                > times.
                >
                > Overall, Al shot the 162 times (162/328 = 49.3%). He was 33-85 with
                > jumpshots, including 11 treys for a true shooting percentage of
                45.3%. He
                > was 15-33 (45.5%) when he drove and then pulled up for the
                jumpshot. He was
                > 18-44 (41.0%) when he drove to the basket for a layup.

                Stuart, have you broken down the FT attempts resulting from these 3
                categories of shot attempts? My guess is that adding in those points
                and attempts will yield a much more favorable case for Iverson going
                to the rack.
                And that doesn't even consider damage done to the opponent by picking
                up fouls. Maybe this could be quantified?

                > However, when he was in transition, and many times unguarded, his
                shooting
                > percentage went way up. In transition situations he was 18-31
                (58%). 3-5
                > with jumpshots, tsp = 70%, 5-9 with the pullup jumper, 10-17 taking
                it to
                > the rack. So when Allen took it all the way to the basket and
                wasn't in
                > transition he was a mere 8-27 (29.6%), demonstrating that in the
                half court
                > at least it's still a big man's game. But in the open court Al's
                speed is a
                > hell of an advantage.
                >
                > On his 86 advances Iverson passed the ball 37 times, shot it 33
                times, and
                > was fouled 7 times. The other 8 advances consists of various misc.
                > happenings.
                >
                > Al passed the ball 109 times, (109/328 = 33.2%). 17 assists and 12
                > "assist-like" plays. .
                >
                > He had 11 turnovers (11/328 = 3.3%) and was fouled 32 times with
                the ball
                > (32/328 = 9.7%) and was fouled 7 times away from the ball. 14 other
                > miscellaneous things (like jumpballs, deflections, timeouts, away
                from the
                > ball fouls, illegal screens, etc.) happened to round out the 328.

                So if Iverson was fouled with the ball 32 times, 7 in transition,
                that leaves 25. If 20 of these were while attempting to shoot, and
                15 (just guessing) were on layup attempts, he may have another (24-30
                FT, just a guess) equivalent of 12-15 shooting to add to his 8-27 FG.
                This is 20-42, which is right up there with his long-range TSP.

                > There were 170 possessions when Al did not touched the ball. Philly
                scored
                > 148 points or 0.87 pts/poss.
                > There were 119 possessions when Al touched the ball and didn't get
                a FGA or
                > a FTA. Philly scored 116 points or 0.97 pts/poss.
                > There were 172 possesions were Al had either a FGA or FTA. Philly
                scored 203
                > points or 1.18 pts/poss. 32 of the 203 points came from Philly
                hitting the
                > offensive boards (including Al recovering his own shot, so there is
                a little
                > double counting here).
                > I don't count the technicals Al shot as part of any possession.

                This makes a strong case that Iverson should in fact be taking the
                shots on this team. In fact, he wasn't in the game at all some of
                the time, and if this were more than 1 or 2 minutes it might need to
                be figured separately.
                Is it possible to break down the offensive rebounding resulting from
                the 3 categories (outside jumper, pullup jumper, layup)?
                And how did you classify those times when Iverson beat his man and
                lobbed it up over Shaq? Those seemed to fall at a high percentage.

                Mike Goodman
              • McKibbin, Stuart
                Stuart, have you broken down the FT attempts resulting from these 3 categories of shot attempts? My guess is that adding in those points and attempts will
                Message 7 of 10 , Oct 22, 2001
                • 0 Attachment

                  "Stuart, have you broken down the FT attempts resulting from these 3

                  categories of shot attempts? My guess is that adding in those points

                  and attempts will yield a much more favorable case for Iverson going

                  to the rack. And that doesn't even consider damage done to the opponent by picking

                  up fouls. Maybe this could be quantified? So if Iverson was fouled with the ball 32 times, 7 in transition,

                  that leaves 25. If 20 of these were while attempting to shoot, and

                  15 (just guessing) were on layup attempts, he may have another (24-30

                  FT, just a guess) equivalent of 12-15 shooting to add to his 8-27 FG.

                  This is 20-42, which is right up there with his long-range TSP."

                  I quick run back through my chart breaks down the FT's like this:

                  2 Technicals, 10 bonus fouls for 20 FTs, 8 Driving layups for 15 FTs (that includes one "and-1" FT),  4 pullup jumpers for 7FTs (that includes one "and-1" FT), and 1 jumpshot for 3 Fts. 16 times he was fouled without getting a FT. I'm missing a FT from Game 3, my charts say 12, the boxscore says 13. Perhaps it was a technical foul I didn't have the tape running for.

                  So to continue your train of thought---Iverson took it to the rack with the intention of shooting it 51 times. He made 17, made 1 and was fouled, was fouled on 7 others for 14 FTs, and missed the other 26 attempts. Rather than see how many FTs Al actually made, I think it would be better to just multiply his FTA times his FT%. So the results were (18 + (15/2 x .8))/51 = 47%. (up from 41%) Which is just what you suspected, Mike.  I think what I should do is see which fouls came in transition and subtract those out (I never said 7 fouls came in transition).

                • McKibbin, Stuart
                  Is it possible to break down the offensive rebounding resulting from the 3 categories (outside jumper, pullup jumper, layup)? And how did you classify those
                  Message 8 of 10 , Oct 22, 2001
                  • 0 Attachment

                    "Is it possible to break down the offensive rebounding resulting from the 3 categories (outside jumper, pullup jumper, layup)? And how did you classify those times when Iverson beat his man and lobbed it up over Shaq? Those seemed to fall at a high percentage."

                    The teardrops over Shaq were counted as layups---he wasn't pointing his elbow at the rim.

                    Well here's the total breakdown of offensive rebounding based on shot type:

                    Jumpshot: 52 misses. Teammates recovered 8 offensive rebs plus two team rebounds. The eight off rebs were converted into 18 points, with Al scoring 3 of the 18. The two team rebs were converted into 5 points with Al scoring all 5.------Total 10 second chances, 23 points.

                    Driving Jumpshot: 18 misses. Teammates recovered 6 offensive rebs which were converted into 4 points.

                    Layups: 26 misses. Teammates recovered two that hit the rim and converted them into 5 points. Teammates recovered one blocked shot and converted it into 2 points. 2 team rebounds were converted into 3 points. Al recovered 4 of his own shots and he converted them into 4 points by himself.----Total 9 second chances, 14 points.

                    In my mind what Doug Collins termed the "Iverson effect" should be offensive rebounds by teammates only (recovering blocked shots, team rebounds or Al getting his own misses shouldn't count). So I'd put the Iverson effect as 87 misses, 16 offensive rebounds, 27 points.

                  • Dean Oliver
                    ... offensive ... or Al ... effect as 87 ... I m completely swamped at work these days, so I don t have much time, but I do have one comment on this. It s
                    Message 9 of 10 , Oct 22, 2001
                    • 0 Attachment
                      --- In APBR_analysis@y..., "McKibbin, Stuart" <smckibbi@c...> wrote:
                      > In my mind what Doug Collins termed the "Iverson effect" should be
                      offensive
                      > rebounds by teammates only (recovering blocked shots, team rebounds
                      or Al
                      > getting his own misses shouldn't count). So I'd put the Iverson
                      effect as 87
                      > misses, 16 offensive rebounds, 27 points.

                      I'm completely swamped at work these days, so I don't have much time,
                      but I do have one comment on this.

                      It's interesting.

                      OK, I have more than that. 16 rebounds of 87 misses isn't very
                      good. The hypothesis has always been that teammates should rebound
                      more of his misses proportionately. Apparently not true.

                      However, 27 points on 16 orebs is pretty impressive. Much better
                      than I thought it would be.

                      Sooo... my calculations generally assume that a team performs just
                      about as well off of misses as it does in a normal possession. 27
                      points out of 87 misses is about right because the Sixers rebounded
                      33% of their own missed shots in the postseason and scored 1.03 pts
                      per possession. So my estimate on how many points should have been
                      scored off of Iverson's misses is 1.03*0.327*87 = 29 pts.

                      What I don't estimate very well, however, is that only 16 of his 87
                      misses be rebounded. I estimated 0.327*87 = 28. This makes his
                      offensive efficiency look higher than it was.

                      On the other hand (I argue with myself a lot on these things), I do
                      give credit for other players rebounding

                      In the end, I estimated that Iverson scored on 48% of his 161
                      possessions, "producing" 165 points. I defined possessions slightly
                      differently -- a missed shot of his that is rebounded by a teammate
                      is not considered a possession. I had him "winning" 2 of 5 games vs.
                      the Lakers.

                      OK, that's more time than I can afford. But I will study this stuff
                      more after 11/19 when my life gets a little more normal. Very useful.

                      Dean Oliver
                      Journal of Basketball Studies
                    Your message has been successfully submitted and would be delivered to recipients shortly.