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

Value of the triple double

Expand Messages
  • rose7654
    I m not so sure that you can dismiss the triple double as a fun but meaningless stat. I just did some quick research of the first 28 triple doubles of
    Message 1 of 8 , Dec 8, 2003
      I'm not so sure that you can dismiss the triple double as a fun but
      meaningless stat. I just did some quick research of the first 28
      triple doubles of 2002-2003 season, from Nov. 1 to Feb. 26. (see
      http://cbs.sportsline.com/nba/story/6113668). The team that had a
      player that made a triple double won 25 of those games and lost only
      3 (I did not count as a loss the games where a player from both teams
      had a triple double which, incidentally, happened twice).


      More interesting tidbits:

      1) of those three losses, two of them were by Sacramento when Chris
      Webber had a triple-double. And Sacramento won 59 games last year!

      2) Conversely, the Bulls only won 30 games last year but won both
      games in which they had a player with a triple double--and both of
      those games were against good teams (NJ and Pho).

      3) Finally, perhaps the strangest little tidbit of them all, of the
      28 triple doubles over that 4 month period, 15 of those came on a
      night when somebody else had another NBA triple double while only 13
      came on nights where nobody else had one.
    • Stephen Greenwell
      3) Finally, perhaps the strangest little tidbit of them all, of the 28 triple doubles over that 4 month period, 15 of those came on a night when somebody else
      Message 2 of 8 , Dec 8, 2003
        3) Finally, perhaps the strangest little tidbit of them all, of the
        28 triple doubles over that 4 month period, 15 of those came on a
        night when somebody else had another NBA triple double while only 13
        came on nights where nobody else had one.


        To me, this sounds like a case of one of those "happy scorers" who gives out too many assists and rebounds for a respective player.  As for the correlation between wins and triple doubles, this doesn't surprise me a lot.  While you'd have to include more years to make a definitive conclusion, the players who most often record triple doubles are multitalented and play for decent teams.  You need to have a decent team to be able to record ten assists usually (they need to make it), unless you're an Andre Miller type who focuses solely on passing a la his Cavs years.

        Steve Greenwell
      • Matthew Lloyd
        More evidence of not dismissing a triple-double as a pop stat : According to STATS, Inc. there have been 436 triple-doubles recorded since 1991-92. Teams
        Message 3 of 8 , Dec 9, 2003
          More evidence of not dismissing a triple-double as a
          "pop stat":

          According to STATS, Inc. there have been 436
          triple-doubles recorded since 1991-92. Teams with
          players recording a triple-double are 333-103 (.764).

          --- rose7654 <rose7654@...> wrote:
          > I'm not so sure that you can dismiss the triple
          > double as a fun but
          > meaningless stat. I just did some quick research of
          > the first 28
          > triple doubles of 2002-2003 season, from Nov. 1 to
          > Feb. 26. (see
          > http://cbs.sportsline.com/nba/story/6113668). The
          > team that had a
          > player that made a triple double won 25 of those
          > games and lost only
          > 3 (I did not count as a loss the games where a
          > player from both teams
          > had a triple double which, incidentally, happened
          > twice).
          >
          >
          > More interesting tidbits:
          >
          > 1) of those three losses, two of them were by
          > Sacramento when Chris
          > Webber had a triple-double. And Sacramento won 59
          > games last year!
          >
          > 2) Conversely, the Bulls only won 30 games last year
          > but won both
          > games in which they had a player with a triple
          > double--and both of
          > those games were against good teams (NJ and Pho).
          >
          > 3) Finally, perhaps the strangest little tidbit of
          > them all, of the
          > 28 triple doubles over that 4 month period, 15 of
          > those came on a
          > night when somebody else had another NBA triple
          > double while only 13
          > came on nights where nobody else had one.
          >
          >
          >
          >


          __________________________________
          Do you Yahoo!?
          New Yahoo! Photos - easier uploading and sharing.
          http://photos.yahoo.com/
        • Dean Oliver
          I call it a pop stat because it doesn t really say much to help analyze a team or a player. The problem with relating triple-doubles to win-loss is that it
          Message 4 of 8 , Dec 9, 2003
            I call it a pop stat because it doesn't really say much to help
            analyze a team or a player. The problem with relating triple-doubles
            to win-loss is that it doesn't really say anything. "Someone get a
            triple double and we'll win! Now go out there and get a triple
            double!" No, that doesn't work. Generally it is good players who get
            triple doubles. But I don't think triple doubles define how good they
            are. Other things do and triple doubles are the occasional
            consequence of that, not the defining cause.

            DeanO

            --- In APBR_analysis@yahoogroups.com, Matthew Lloyd <mlloyd333@y...>
            wrote:
            > More evidence of not dismissing a triple-double as a
            > "pop stat":
            >
            > According to STATS, Inc. there have been 436
            > triple-doubles recorded since 1991-92. Teams with
            > players recording a triple-double are 333-103 (.764).
            >
            > --- rose7654 <rose7654@y...> wrote:
            > > I'm not so sure that you can dismiss the triple
            > > double as a fun but
            > > meaningless stat. I just did some quick research of
            > > the first 28
            > > triple doubles of 2002-2003 season, from Nov. 1 to
            > > Feb. 26. (see
            > > http://cbs.sportsline.com/nba/story/6113668). The
            > > team that had a
            > > player that made a triple double won 25 of those
            > > games and lost only
            > > 3 (I did not count as a loss the games where a
            > > player from both teams
            > > had a triple double which, incidentally, happened
            > > twice).
            > >
            > >
            > > More interesting tidbits:
            > >
            > > 1) of those three losses, two of them were by
            > > Sacramento when Chris
            > > Webber had a triple-double. And Sacramento won 59
            > > games last year!
            > >
            > > 2) Conversely, the Bulls only won 30 games last year
            > > but won both
            > > games in which they had a player with a triple
            > > double--and both of
            > > those games were against good teams (NJ and Pho).
            > >
            > > 3) Finally, perhaps the strangest little tidbit of
            > > them all, of the
            > > 28 triple doubles over that 4 month period, 15 of
            > > those came on a
            > > night when somebody else had another NBA triple
            > > double while only 13
            > > came on nights where nobody else had one.
            > >
            > >
            > >
            > >
            >
            >
            > __________________________________
            > Do you Yahoo!?
            > New Yahoo! Photos - easier uploading and sharing.
            > http://photos.yahoo.com/
          • Mikey Stewart
            Agreed. I was probably too harsh in my criticism of the triple double, and I don t think that I could have explained my real position better than Dean just
            Message 5 of 8 , Dec 9, 2003
              Agreed. I was probably too harsh in my criticism of the triple double,
              and I don't think that I could have explained my real position better than
              Dean just did. Triple doubles correlate highly with winning, and that
              makes sense because, ceteris paribus, more points, rebounds, and assists
              indicate a better performance. If you looked at things like turnovers and
              shooting percentages, I'm sure that you'd find that the triple doubles
              with few turnovers and high shooting percentages correlate even more
              highly with winning. Then we get into a whole discussion of linear
              weights as we try to throw everything into the regression.

              So yeah, triple doubles do correlate highly with winning, but there's the
              confounding variable of the player just having a solid game, period.


              On Tue, 9 Dec 2003, Dean Oliver wrote:

              > I call it a pop stat because it doesn't really say much to help
              > analyze a team or a player.  The problem with relating triple-doubles
              > to win-loss is that it doesn't really say anything.  "Someone get a
              > triple double and we'll win!  Now go out there and get a triple
              > double!"  No, that doesn't work.  Generally it is good players who get
              > triple doubles.  But I don't think triple doubles define how good they
              > are.  Other things do and triple doubles are the occasional
              > consequence of that, not the defining cause.
              >
              > DeanO
              >
              > --- In APBR_analysis@yahoogroups.com, Matthew Lloyd <mlloyd333@y...>
              > wrote:
              > > More evidence of not dismissing a triple-double as a
              > > "pop stat":
              > >
              > > According to STATS, Inc. there have been 436
              > > triple-doubles recorded since 1991-92.  Teams with
              > > players recording a triple-double are 333-103 (.764).
              > >
              > > --- rose7654 <rose7654@y...> wrote:
              > > > I'm not so sure that you can dismiss the triple
              > > > double as a fun but
              > > > meaningless stat.  I just did some quick research of
              > > > the first 28
              > > > triple doubles of 2002-2003 season, from Nov. 1 to
              > > > Feb. 26.  (see
              > > > http://cbs.sportsline.com/nba/story/6113668).  The
              > > > team that had a
              > > > player that made a triple double won 25 of those
              > > > games and lost only
              > > > 3 (I did not count as a loss the games where a
              > > > player from both teams
              > > > had a triple double which, incidentally, happened
              > > > twice).
              > > >
              > > >
              > > > More interesting tidbits:
              > > >
              > > > 1) of those three losses, two of them were by
              > > > Sacramento when Chris
              > > > Webber had a triple-double.  And Sacramento won 59
              > > > games last year! 
              > > >
              > > > 2) Conversely, the Bulls only won 30 games last year
              > > > but won both
              > > > games in which they had a player with a triple
              > > > double--and both of
              > > > those games were against good teams (NJ and Pho).
              > > >
              > > > 3) Finally, perhaps the strangest little tidbit of
              > > > them all, of the
              > > > 28 triple doubles over that 4 month period, 15 of
              > > > those came on a
              > > > night when somebody else had another NBA triple
              > > > double while only 13
              > > > came on nights where nobody else had one. 
              > > >
              > > >
              > > >
              > > >
              > >
              > >
              > > __________________________________
              > > Do you Yahoo!?
              > > New Yahoo! Photos - easier uploading and sharing.
              > > http://photos.yahoo.com/
              >
              >
              > Yahoo! Groups Sponsor
              > ADVERTISEMENT
              > click here
              >
              > 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.
              >
            • igor eduardo küpfer
              In the 2002-03 season, I count 42 triple doubles (3D), with the 3D player s team winning 29 of those
              Message 6 of 8 , Dec 9, 2003
                <snip discussion of triple double correlation with wins>

                In the 2002-03 season, I count 42 triple doubles (3D), with the 3D player's
                team winning 29 of those games -- 69%. In the same season, I count 233 teams
                having a combined 3D -- that is, at least one player having double-digit
                points, rebounds, and assists (or double-digit points, rebounds, and blocks,
                or double-digit points, assists, and blocks) although not necessarily the
                same player. Those teams won 64% of those games. A chi-squared test shows
                that the difference is not statistically significant. This would seem to
                indicate that an individual player triple-double has little, if any,
                predictive value.

                ed
              • Dean Oliver
                This seems like a good way of looking at it. Not significant at what level? ... player s ... 233 teams ... digit ... and blocks, ... necessarily the ... shows
                Message 7 of 8 , Dec 9, 2003
                  This seems like a good way of looking at it. Not significant at what
                  level?

                  --- In APBR_analysis@yahoogroups.com, igor eduardo küpfer
                  <igorkupfer@r...> wrote:
                  > <snip discussion of triple double correlation with wins>
                  >
                  > In the 2002-03 season, I count 42 triple doubles (3D), with the 3D
                  player's
                  > team winning 29 of those games -- 69%. In the same season, I count
                  233 teams
                  > having a combined 3D -- that is, at least one player having double-
                  digit
                  > points, rebounds, and assists (or double-digit points, rebounds,
                  and blocks,
                  > or double-digit points, assists, and blocks) although not
                  necessarily the
                  > same player. Those teams won 64% of those games. A chi-squared test
                  shows
                  > that the difference is not statistically significant. This would
                  seem to
                  > indicate that an individual player triple-double has little, if any,
                  > predictive value.
                  >
                  > ed
                • igor eduardo küpfer
                  ... From: Dean Oliver To: Sent: Tuesday, December 09, 2003 1:30 PM Subject: [APBR_analysis] Re: Value of
                  Message 8 of 8 , Dec 9, 2003
                    ----- Original Message -----
                    From: "Dean Oliver" <deano@...>
                    To: <APBR_analysis@yahoogroups.com>
                    Sent: Tuesday, December 09, 2003 1:30 PM
                    Subject: [APBR_analysis] Re: Value of the triple double


                    > This seems like a good way of looking at it. Not significant at what
                    > level?

                    Chi-squared = 0.264, df = 1, two-tailed p value = 0.6075

                    I should add that I didn't double count the triple doubles, ie group 1 was
                    individual 3Ds and group 2 was combined team 3Ds minus individual 3Ds.

                    ed

                    --- In APBR_analysis@yahoogroups.com, igor eduardo küpfer
                    <igorkupfer@r...> wrote:
                    >> <snip discussion of triple double correlation with wins>
                    >>
                    >> In the 2002-03 season, I count 42 triple doubles (3D), with the 3D
                    >player's
                    >> team winning 29 of those games -- 69%. In the same season, I count
                    >233 teams
                    >> having a combined 3D -- that is, at least one player having double-
                    >digit
                    >> points, rebounds, and assists (or double-digit points, rebounds,
                    >and blocks,
                    >> or double-digit points, assists, and blocks) although not
                    >necessarily the
                    >> same player. Those teams won 64% of those games. A chi-squared test
                    >shows
                    >> that the difference is not statistically significant. This would
                    >seem to
                    >> indicate that an individual player triple-double has little, if any,
                    >> predictive value.
                    >>
                    >> ed
                  Your message has been successfully submitted and would be delivered to recipients shortly.