<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix">Forest,<br>
      <br>
      Sorry you're right, I somehow miscounted.<br>
      <br>
      However the method certainly fails "Pairwise Plurality", and I
      can't see what the justification (in terms of criterion
      compliances) is for<br>
      the bad failure of Condorcet Loser.<br>
      <br>
      Chris Benham<br>
      <br>
      On 6/6/2015 4:06 AM, Forest Simmons wrote:<br>
    </div>
    <blockquote
cite="mid:CAP29one832eAigjsDP_JcrMtPyujLotjw_4KRfeq-mG_Xs9kJw@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div>
          <div>
            <div>Chris,<br>
              <br>
            </div>
            in the second example the symmetrically completed ballots
            are<br>
            <br>
            <div>27 A<br>
            </div>
            <div>11 A>C<br>
            </div>
            <div>11 C>A<br>
            </div>
            <div>02 C<br>
            </div>
            <div>11 C>B<br>
            </div>
            <div>11 B>C<br>
            </div>
            27 B<br>
            <br>
          </div>
          Candidate C, the IA-MPO winner, is ranked on 46 ballots. 
          Neither A nor B is top ranked on more than 38 ballots.  So it
          seems to me that Plurality is not violated.<br>
          <br>
        </div>
        <div>In summary, IA-MPO does not violate the strong version of
          Plurality on the symmetrically completed ballots, and does not
          violate the (original) weaker version of Plurality (what I
          called Plurality') when applied to the original ballots.  I
          think it is too early to throw it out.<br>
          <br>
        </div>
        <div>Forest<br>
        </div>
        <div><br>
        </div>
        Forest<br>
      </div>
      <div class="gmail_extra"><br>
        <div class="gmail_quote">On Fri, Jun 5, 2015 at 9:46 AM,
          C.Benham <span dir="ltr"><<a moz-do-not-send="true"
              href="mailto:cbenham@adam.com.au" target="_blank">cbenham@adam.com.au</a>></span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div bgcolor="#FFFFFF" text="#000000">
              <div>Forest,<span class=""><br>
                  <br>
                  <blockquote type="cite">"Symmetrical completion
                    normally would replace  16 A=C with 8 A>C and 8
                    C>A .  I understand why you didn't do it that
                    way:  you didn't want to go outside the category of
                    two slot ballots.  But just because the voters have
                    to vote two slot ballots doesn't mean that we are
                    prohibited from using a counting method that creates
                    auxiliary data structures like matrices or three
                    slot rankings."</blockquote>
                  <br>
                </span> Your presumption about my motive is wrong. I did
                it that way because (perhaps because of lack of sleep)
                that was the only way that occurred to me.<br>
                I don't like 2-slot ballots and if they are used I can't
                take seriously the idea that anything other than
                Approval should be used to determine the winner.<br>
                <br>
                Also I wasn't suggesting or contemplating using the
                symmetric completion at the top to modify IA-MPO, rather
                I was just suggesting using it to test<br>
                whether or not the result is in compliance with the
                Plurality criterion.<br>
                <br>
                Unfortunately your second example shows that even the
                newly modified version of IA-MPO (that works on the
                ballots symmetrically completed at<br>
                the top) miserably fails Plurality.<br>
                <br>
                Chris Benham
                <div>
                  <div class="h5"><br>
                    <br>
                    <br>
                    On 6/5/2015 8:15 AM, Forest Simmons wrote:<br>
                  </div>
                </div>
              </div>
              <blockquote type="cite">
                <div>
                  <div class="h5">
                    <div dir="ltr"><br>
                      <div class="gmail_extra"><br>
                        <div class="gmail_quote">On Wed, Jun 3, 2015 at
                          7:43 PM, C.Benham <span dir="ltr"><<a
                              moz-do-not-send="true"
                              href="mailto:cbenham@adam.com.au"
                              target="_blank">cbenham@adam.com.au</a>></span>
                          wrote:<br>
                          <blockquote class="gmail_quote"
                            style="margin:0px 0px 0px
                            0.8ex;border-left:1px solid
                            rgb(204,204,204);padding-left:1ex">
                            <div bgcolor="#FFFFFF" text="#000000">
                              <div>...<br>
                                <span><br>
                                </span> Forest, I'm not sure that this
                                isn't the same as the normal  Plurality
                                criterion.  The reference to "first
                                preference" in the Plurality criterion
                                definition I think refers to exclusive
                                first preference.<br>
                                <br>
                                (I gather that Woodall's criteria are
                                only about strict rankings from the top,
                                which may or may not be truncated,) I
                                suppose it could and should be extended
                                to applying to ballots<br>
                                that are symmetrically "completed" only
                                at the top. Doing that to your example
                                gives:<br>
                                <br>
                                41 A<br>
                                18 C<br>
                                41 B<br>
                                <br>
                                Electing C on these ballots is insane
                                and I don't see how electing C on the
                                original ballots (where some of the
                                votes are given half to one candidate
                                and half to another) is<br>
                                really any more justified.<br>
                                <br>
                                Yes, this convinces me that the
                                Plurality criterion should definitely be
                                applied to to the ballots symmetrically
                                completed at the top and that we can
                                without regret<br>
                                kiss  IA-MPO  goodbye.<br>
                              </div>
                            </div>
                          </blockquote>
                          <div><br>
                             Symmetrical completion normally would
                            replace  16 A=C with 8 A>C and 8 C>A
                            .  I understand why you didn't do it that
                            way:  you didn't want to go outside the
                            category of two slot ballots.  But just
                            because the voters have to vote two slot
                            ballots doesn't mean that we are prohibited
                            from using a counting method that creates
                            auxiliary data structures like matrices or
                            three slot rankings.<br>
                            <br>
                          </div>
                          <div>If we did this (I think more appropriate)
                            kind of symmetric completion, the working
                            ballots would become<br>
                            <br>
                            33 A<br>
                          </div>
                          <div>08 A>C<br>
                          </div>
                          <div>08 C>A<br>
                            <div>02 C<br>
                            </div>
                            08 C>B<br>
                          </div>
                          <div>08 B>C<br>
                          </div>
                          <div> 33 B<br>
                          </div>
                          <div>The resulting respective IA-MPO scores
                            for A, B, and C would become  49-49, 49-49,
                            and 34-41, so this version of IA-MPO with a
                            front end of symmetric completion at the top
                            would give a tie to A and B, the only
                            candidates with a non-negative score.<br>
                          </div>
                          <div><br>
                          </div>
                          <div>Let's try it on<br>
                            <br>
                          </div>
                          <div>27 A<br>
                          </div>
                          <div>22 A=C<br>
                          </div>
                          <div>02 C<br>
                          </div>
                          <div>22 B=C<br>
                          </div>
                          <div>27 B<br>
                            <br>
                          </div>
                          <div>Candidates A and B are tied for Approval
                            Winner with 49 approvals each against 46 for
                            C, making C the ballot Condorcet Loser.<br>
                            <br>
                          </div>
                          <div>Let's do the natural symmetric completion
                            to see the likely sincere ballots that would
                            be voted if equal ranking at top were not
                            allowed (nor practically required,as in
                            Approval):<br>
                            <br>
                            <div>27 A<br>
                            </div>
                            <div>11 A>C<br>
                            </div>
                            <div>11 C>A<br>
                            </div>
                            <div>02 C<br>
                            </div>
                            <div>11 C>B<br>
                            </div>
                            <div>11 B>C<br>
                            </div>
                            27 B<br>
                          </div>
                          <div><br>
                          </div>
                          <div>The respective IA-MPO scores for A, B,
                            and C are  49-49, 49-49, and 46-38, the only
                            positive difference. So C wins.  Note that C
                            is still the ballot Condorcet Loser.<br>
                            <br>
                          </div>
                          <div>Whether or not we like this result
                            probably reflects how much we prefer a
                            centrist over an extremist, all else being
                            equal.<br>
                          </div>
                          <div> </div>
                          <blockquote class="gmail_quote"
                            style="margin:0px 0px 0px
                            0.8ex;border-left:1px solid
                            rgb(204,204,204);padding-left:1ex">
                            <div bgcolor="#FFFFFF" text="#000000">
                              <div> <br>
                                Another version of the criterion is
                                "Pairwise Plurality"  (suggested a while
                                ago by Kevin or me): If candidate X's
                                lowest pairwise score is higher than
                                candidate Y's highest<br>
                                pairwise score, then Y must not be
                                elected".<br>
                                <br>
                                I like this. Both IA-MPO and  SMD,TR
                                fail it, as in the two examples.  <br>
                              </div>
                            </div>
                          </blockquote>
                          <div><br>
                          </div>
                          <div>Nice idea! <br>
                          </div>
                        </div>
                        <br>
                      </div>
                    </div>
                  </div>
                </div>
                <p color="#000000" align="left"><span class="">No virus
                    found in this message.<br>
                    Checked by AVG - <a moz-do-not-send="true"
                      href="http://www.avg.com" target="_blank">www.avg.com</a><br>
                  </span> Version: 2015.0.5961 / Virus Database:
                  4355/9941 - Release Date: 06/04/15</p>
              </blockquote>
              <br>
            </div>
          </blockquote>
        </div>
        <br>
      </div>
      <p class="" avgcert""="" color="#000000" align="left">No virus
        found in this message.<br>
        Checked by AVG - <a moz-do-not-send="true"
          href="http://www.avg.com">www.avg.com</a><br>
        Version: 2015.0.5961 / Virus Database: 4355/9950 - Release Date:
        06/05/15</p>
    </blockquote>
    <br>
  </body>
</html>