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

Problem with Concorcet

Expand Messages
  • Rich Tomasso
    [Moderator note: This is part of a conversation going on elsewhere, but since it s business-related, I ll let this post through. But let s keep the discussion
    Message 1 of 6 , Apr 8, 2003
    • 0 Attachment
      [Moderator note: This is part of a conversation going on elsewhere, but since it's business-related, I'll let this post through. But let's keep the discussion to the Condocet method, okay?]

      --- Elizabeth McKinstry <emckinst@...> wrote:
      > >[Chuck Geshlider wrote:]
      > >8. Add to this intrinsic suspicion, this completely confusing
      > > "CONdorcet" voting method. It reminds me of attempting to understand
      > > the Internal Revenue Code. Every time you ask an official for an
      > > explanation, you get a different story. It's complex, and a
      > > prescription for disaster in this context I bring up in #7. I need not
      > > go into details here. It's self-evident.
      >
      > Really? A different story? That is bad, if true.

      It is happening. In the past few weeks I've been talking with FSP members
      and everyone has a different take on how the voting system works. Even
      people who've read the voting report.

      > I agree that it is not an intuitively understandable voting method.

      That is a major drawback.

      > But after reading several pieces of research on voting methods, it
      became
      > clear that this method is by far the most fair for this kind of choice,
      > and the favored method by voting experts everywhere.

      Is it though? IRV seemed to fit all our needs, and 90% of our audience
      already understands it.

      I think I found another big problem with Condorcet: You can win the vote
      and still not be picked. My quick math seems to indicate that even if we
      hold the vote at 5000 and State A gets 2501 or more first-place votes, it
      can still lose. Esp if State B places very well 2nd or 3rd across the
      board. Concorcet seems to be a method to find the best compromize
      candidate rather than the most preferred candidate, which aren't the same
      thing.
    • jason.sorens@yale.edu
      ... We considered IRV, but everyone informed us that it was inferior to Condorcet. With many candidates, as we have, it can yield real paradoxes when
      Message 2 of 6 , Apr 8, 2003
      • 0 Attachment
        Quoting Rich Tomasso <rich_tomasso@...>:

        > Is it though? IRV seemed to fit all our needs, and 90% of our
        > audience
        > already understands it.

        We considered IRV, but everyone informed us that it was inferior to
        Condorcet. With many candidates, as we have, it can yield real
        paradoxes when Condorcet does not. Also, I'm not sure it's
        intrinsically any easier to understand than Condorcet; it's just more
        frequently used & therefore more familiar to some - many members I
        talked to had real trouble w/ IRV when I proposed it.

        > I think I found another big problem with Condorcet: You can win the
        > vote
        > and still not be picked. My quick math seems to indicate that even if
        > we
        > hold the vote at 5000 and State A gets 2501 or more first-place
        > votes, it
        > can still lose.

        No, that's impossible. If a state wins 2,501 votes against every other
        state in a voting population of 5,000, then the greatest number of
        votes that any state could win against it is 2,499. If a state has
        2,501 first place votes, it will always win, even if it has 2,499 last
        place votes. I guess a new voting explanation is necessary. :-/

        Jason
      • Rich Macy
        ... Notwithstanding Jason s comment that a state with 2,501 first-place votes cannot lose, I do not see how this is a bad thing. I think choosing the best
        Message 3 of 6 , Apr 8, 2003
        • 0 Attachment
          > Concorcet seems to be a method to find the best compromize
          > candidate rather than the most preferred candidate, which aren't the same
          > thing.

          Notwithstanding Jason's comment that a state with 2,501 first-place votes
          cannot lose, I do not see how this is a bad thing. I think choosing the best
          compromise candidate would actually make more people happy than choosing the
          the most preferred.

          =====
          "Where the government fears the people, there is liberty. Where the people fear the government, there is tyranny." --Thomas Jefferson
          Liberty in Our Lifetime - www.freestateproject.org

          __________________________________________________
          Do you Yahoo!?
          Yahoo! Tax Center - File online, calculators, forms, and more
          http://tax.yahoo.com
        • Kelly Setzer
          ... Plurality (popular) voting is exactly what gets you presidents like George Bush or Bill Clinton. I don t point that out as some sort of political
          Message 4 of 6 , Apr 8, 2003
          • 0 Attachment
            On Tue, Apr 08, 2003 at 11:31:32AM -0700, Rich Macy wrote:
            > > Concorcet seems to be a method to find the best compromize
            > > candidate rather than the most preferred candidate, which aren't the same
            > > thing.
            >
            > Notwithstanding Jason's comment that a state with 2,501 first-place votes
            > cannot lose, I do not see how this is a bad thing. I think choosing the best
            > compromise candidate would actually make more people happy than choosing the
            > the most preferred.

            Plurality (popular) voting is exactly what gets you presidents like
            George Bush or Bill Clinton. I don't point that out as some sort of
            political statement, only to illustrate two important points about
            compromise choices:

            1) Neither Bush nor Clinton were elected by even a simple majority of
            the citizens. (I'm ignoring Electoral College gyrations.)

            2) After the Bush and Clinton were elected, the majority of the
            citizens did not want that person to be president. Ergo, a majority
            of the citizens was not satisfied with the selection.

            If you looked at those results in abstract terms (33% of the people vote
            for red jelly beans, etc.), then it seems counterituitive. It also
            illustrates a major issue with the way political offices are filled:
            by the time the citizens get to vote on a candidate, that candidate has
            been the winner of numerous previous compromise choices (i.e. party
            conventions, etc.). If anything, we need more candidates on the
            ballots, rather than just the candidates that manage to be the least
            objectional after repeated compromises. In the end, all we get are
            compromised and corrupt politicians. I guess that's a personal rant.

            As far as the FSP vote, I'll make these two points:
            1) This is not at all like a political election, so my comments above
            are, more or less, off-topic. The FSP vote is more like getting 5,000
            members to agree on what color Corvette we're going to buy.

            2) The voting method topic has been beaten to death. I would really
            reccommend you read through the archive. There has been a tremendous
            amount of technical and social analysis on this matter.

            Kelly
          • Jason P Sorens
            ... There is some virtue to that, I agree. Condorcet is sort of a compromise itself between methods that heavily emphasize second- and third-place votes (like
            Message 5 of 6 , Apr 8, 2003
            • 0 Attachment
              On Tue, 8 Apr 2003, Rich Macy wrote:

              > > Concorcet seems to be a method to find the best compromize
              > > candidate rather than the most preferred candidate, which aren't the same
              > > thing.
              >
              > Notwithstanding Jason's comment that a state with 2,501 first-place votes
              > cannot lose, I do not see how this is a bad thing. I think choosing the best
              > compromise candidate would actually make more people happy than choosing the
              > the most preferred.

              There is some virtue to that, I agree. Condorcet is sort of a compromise
              itself between methods that heavily emphasize second- and third-place
              votes (like Borda) and methods that heavily emphasize first-place votes
              (like IRV). It is likely that no state will win an outright majority of
              first-place ballots, so the winning state will have to have both a high
              number of first-place votes and a high number of 2nd, 3rd, and 4th place
              votes. In other words, Condorcet is most likely to pick the candidate
              that is both a leader in terms of people's first preferences and is
              generally acceptable to a large number of those who do not rank it first.
              Polarizing candidates are unlikely to win unless they have an outright
              majority supporting them, and only a minority dislike them extremely.

              ___________________________________________________________________________

              Jason P Sorens - jason.sorensATyale.edu - <http://pantheon.yale.edu/~jps35>

              <http://www.freestateproject.org> - Do you want liberty in your lifetime?
            • Phyllis
              ... Rich, I was about to reply to Rich Tomaso s statement when I noticed your reply. You have said exactly what I was thinking. With the current air of
              Message 6 of 6 , Apr 8, 2003
              • 0 Attachment
                --- Rich Macy <rich_tiger@...> wrote:
                > > Concorcet seems to be a method to find the best
                > compromize
                > > candidate rather than the most preferred candidate,
                > which aren't the same
                > > thing.
                >
                > Notwithstanding Jason's comment that a state with 2,501
                > first-place votes
                > cannot lose, I do not see how this is a bad thing. I
                > think choosing the best
                > compromise candidate would actually make more people
                > happy than choosing the
                > the most preferred.
                >
                Rich, I was about to reply to Rich Tomaso's statement when
                I noticed your reply. You have said exactly what I was
                thinking. With the current air of division I believe a
                clear win for any one state (as first choice) will only
                defeat the purpose of Freedom in our Lifetime. I believe
                this same spirit is what has kept the Libertarian Party
                from winning elections. IMO, if we don't learn to
                compromise our mission is destined to break up into
                factions.

                Phyl



                =====
                When Hitler came for the Jews... I was not a Jew, therefore, I was not concerned. And when Hitler attacked the Catholics, I was not a Catholic, and therefore, I was not concerned. And when Hitler attacked the unions and the industrialists, I was not a member of the unions and I was not concerned. Then, Hitler attacked me and the
                Protestant church and there was nobody left to be concerned. Pastor Martin Niemoller

                Freedom in our lifetime: http://www.freestateproject.com

                __________________________________________________
                Do you Yahoo!?
                Yahoo! Tax Center - File online, calculators, forms, and more
                http://tax.yahoo.com
              Your message has been successfully submitted and would be delivered to recipients shortly.