<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix">On 12/9/2016 10:05 AM, Michael Ossipoff
      wrote:<br>
      <br>
      <blockquote type="cite">
        <div class="gmail_quote">The fact of it being better to
          equal-rank the set that is important to you instead of
          choosing among them is true with other rank methods too.<br>
          <br>
        </div>
        <div class="gmail_quote">In particular, it's true of Condorcet
          & Bucklin. It would be true of IRV too, if IRV allowed
          equal-ranking.<br>
          <br>
        </div>
        I don't have proof that it's, in principle, a property of _all_
        ranking-methods, but I don't know of an exception.</blockquote>
      <br>
      Mike,<br>
      <br>
      If  IRV allows equal-ranking, it should definitely be the
      "fractional" version (so that in every round each ballot gives a
      single vote or fractions of a vote<br>
      that sum to 1).<br>
      <br>
      If instead a ballot gives a full vote (approval style) to each
      candidate it highest ranks then the method is much more vulnerable
      to Pushover and fails<br>
      Mutual Dominant Third.   I gave an example of that a while
      (probably years) ago.<br>
      <br>
      Here is an example of Push-over that is also a failure of 
      Unburiable Mutual Dominant Third:<br>
      <br>
      45: A=C  (sincere is A>B)<br>
      35: B>A<br>
      20: C>B<br>
      <br>
      B is the sincere MDT candidate, and so with normal IRV is
      strategically invulnerable, but under  ER-IRV(Whole) loses to A.<br>
      <br>
      The fractional version also makes Push-over strategising a bit
      easier, so for IRV and Benham I'm opposed to allowing above-bottom<br>
      equal-ranking.<br>
      <br>
      But if that is insisted on I suggest that to decide which
      candidate is next eliminated we first order the candidates by the
      fractional method,<br>
      and then have ballots that equal-highest rank more than one
      candidate give a full vote to the one of those that is highest in
      the order (and <br>
      nothing to any other candidate).  Then eliminate the candidate
      with the fewest votes.<br>
      <br>
      So in the above example the initial order is  C 40.5  >  B 35
      >  A 22.5.   The 45 A=C ballots give a full vote to C, giving 
      C 65 > B 35 > A 0.  We eliminate<br>
      A and then B and C wins.   (There are better examples where the
      device does some good).<br>
      <br>
      You wrote "I don't have proof that it's, in principle, a property
      of _all_ ranking-methods, but I don't know of an exception."<br>
      <br>
      It's true of  Winning Votes and Bucklin and any of the proposed
      methods that meet FBC, but I don't see how it is of 
      ER-IRV(fractional).<br>
      <br>
      Or MinMax Margins or  ER-Benham(fractional)  or even 
      Smith//Approval (implicit). <br>
      <br>
      In the latter case obviously the voter should truncate hir
      bottom-set, but couldn't it be the case that if the voter strictly
      ranks hir top set then one <br>
      of them will be the CW while the most approved candidate is in hir
      bottom set but if the voter equal top-ranks hir top set there will
      be a top-cycle <br>
      that includes the most approved candidate?<br>
      <br>
      Chris Benham<br>
      <br>
      <br>
      On 12/9/2016 10:05 AM, Michael Ossipoff wrote:<br>
    </div>
    <blockquote
cite="mid:CAOKDY5DsiMkffSak0tVssNwjFVmB3=5T-R9qH6nNWd2KkMhtLw@mail.gmail.com"
      type="cite">
      <div dir="ltr"><br>
        <div class="gmail_extra"><br>
          <div class="gmail_quote">When I used the example of MDDA &
            MDDAsc, to illustrate that it's better to equal-top-rank
            your strong top-set, rather than choosing among them by
            ranking them in order of preference--That wasn't intended as
            criticism of MDDA & MDDAsc.<br>
            <br>
          </div>
          <div class="gmail_quote">The fact of it being better to
            equal-rank the set that is important to you instead of
            choosing among them is true with other rank methods too.<br>
            <br>
          </div>
          <div class="gmail_quote">In particular, it's true of Condorcet
            & Bucklin. It would be true of IRV too, if IRV allowed
            equal-ranking.<br>
            <br>
          </div>
          <div class="gmail_quote">I don't have proof that it's, in
            principle, a property of _all_ ranking-methods, but I don't
            know of an exception.<br>
            <br>
          </div>
          <div class="gmail_quote">One fairly obvious thing that can be
            said for MDDA & MDDAsc is that your protection for your
            strong top-set, even when ranking them (and no one else) in
            order of preference, and approving them all (as is the
            default), is as good as your protection of them in Approval,
            when you approve only lthem.<br>
            <br>
          </div>
          <div class="gmail_quote">A majority doing so in approval will
            elect one of them.<br>
            <br>
          </div>
          <div class="gmail_quote">A majority doing so in MDDA or MDDAsc
            will give a majority-disqualification to everyone else. And
            if preferrers of one of your strong bottom-set try burial or
            truncation, and if they thereby manage to make everyone
            majority-disqualified, then someone in your strong set will
            win the Approval count.<br>
            <br>
          </div>
          <div class="gmail_quote">That suggests that MDDA & MDDAsc
            let you choose among your strong top-set, and still protect
            them from your strong bottom-set just as well as Approval
            would have let you. That's an improvement over Approval.<br>
            <br>
          </div>
          <div class="gmail_quote">Of course an additional improvement
            is that MDDA & MDDAsc give you an easy, convenient,
            & reliable way to avoid chicken-dilemma (by denying
            approval to the candidate of the distrusted faction.<br>
            <br>
          </div>
          <div class="gmail_quote">It's just that MDDA & MDDAsc
            allow you to further enhance the protection of your strong
            top-set, by top-ranking them all. If a majority do that,then
            it would be quite impossible for buriers or truncators to
            majority-disqualify them. Of course if any significant
            number of voters similar to you top-rank those candidates,
            that makes it much more difficult, or impossible, for
            buriers or triuncators to majority-disqualify them.<br>
            <br>
          </div>
          <div class="gmail_quote">If you use the chicken-dilemma
            defense of denying approval to the candidate of the
            distrusted faction, and that candidate is someone whom you
            top-rank, then you're still protecting hir from burial &
            truncation, for the reason described above.<br>
            <br>
          </div>
          <div class="gmail_quote">If the candidate to whom you deny
            approval is someone you rank below top, then that is no
            longer true. If the method is MDDA, that candidate still has
            the full truncation-proofness protection that any ranked
            candidate has. If the method is MDDAsc, that is no longer
            guaranteed. But, if Mono-Add-Plump is necessary for public
            acceptance, then the cutting-loose of that middle-ranked
            candidate of the distrusted faction is a regrettable but
            justifiable action resulting from reasons that that faction
            has given you for defection-deterrence.<br>
            <br>
          </div>
          <div class="gmail_quote">Likewise, though MDDA protects your
            middle-ranked candidates from truncation by eachother's
            factions, that protection isn't essential, because reliably
            choosing _among_ your strong top-set isn't the important
            thing. <br>
            <br>
          </div>
          <div class="gmail_quote">In MDDAsc, you're still fully
            protecting your top-ranked candidates against everyone else,
            and you're still fully protecting all of your rannked &
            approved candidates against your unranked, unapproved
            candidates. That's what's important.<br>
            <br>
          </div>
          <div class="gmail_quote">MDDA & MDDAsc are the rank
            methods that best deliver the benefits that are available
            from ranking-methods.<br>
            <br>
          </div>
          <div class="gmail_quote">Now, to resum my reply:<br>
             <br>
          </div>
          <div class="gmail_quote">
            <blockquote class="gmail_quote" style="margin:0 0 0
              .8ex;border-left:1px #ccc solid;padding-left:1ex">
              <div dir="ltr">
                <div class="gmail_extra">
                  <div class="gmail_quote">
                    <div><br>
                    </div>
                    <div>I like to remind people that, very often, "Good
                      enough is better than best."  That is, a voting
                      system (or a candidate) that is "good enough" may
                      very likely better than one that is "best". <br>
                    </div>
                  </div>
                </div>
              </div>
            </blockquote>
            <div><br>
            </div>
            <div>Exactly. Eecting one that is good enough is much more
              important than reducing the probability of doing so, by
              trying to choose among the ones that are good enough. <br>
              <br>
            </div>
            <div>[Replying farther down] :<br>
              <br>
            </div>
            <blockquote class="gmail_quote" style="margin:0 0 0
              .8ex;border-left:1px #ccc solid;padding-left:1ex">
              <div dir="ltr">
                <div class="gmail_extra">
                  <div class="gmail_quote">
                    <div><br>
                      <br>
                    </div>
                    <span class="">
                      <blockquote class="gmail_quote" style="margin:0 0
                        0 .8ex;border-left:1px #ccc
                        solid;padding-left:1ex">
                        <div dir="ltr">
                          <div>
                            <div>
                              <div>
                                <div>
                                  <div>
                                    <div>
                                      <div>
                                        <div>
                                          <div>
                                            <div>
                                              <div>
                                                <div><br>
                                                </div>
                                                1. In this country, for
                                                the 99%, a progressive
                                                government would be
                                                incomparably better than
                                                a Republocrat government
                                                (like we've had for a
                                                long time, and still
                                                have). <br>
                                                <br>
                                                If you don't believe it,
                                                look at some progressive
                                                party platforms (Greens,
                                                etc.), and compare them
                                                to the things that
                                                people are saying that
                                                they want, or that they
                                                want changed.<br>
                                                <br>
                                              </div>
                                              So, for the 99%, _any_
                                              progressive would support
                                              better policies than_any_
                                              republocrat.<br>
                                              <br>
                                            </div>
                                            That means that, for the
                                            99%, there's a strong
                                            top-set and a strong
                                            bottom-set.<br>
                                            <br>
                                          </div>
                                          ...And, when there is,
                                          Approval voting is really
                                          simple: <br>
                                          <br>
                                        </div>
                                        Approve (only) all of your
                                        strong top-set.<br>
                                        <br>
                                      </div>
                                      2. Suppose we're talking about a
                                      better world, in a better future,
                                      in which the 99% don't have a
                                      bottom-set. Or suppose we're
                                      talking about some other country,
                                      or some entirely different
                                      non-political <a
                                        moz-do-not-send="true"
                                        href="http://voting-situation.in"
                                        target="_blank">voting-situation.in</a>
                                      which you don't have strong top
                                      & bottom sets.<br>
                                      <br>
                                    </div>
                                    There are various ways that you
                                    could vote.<br>
                                    <br>
                                  </div>
                                  a) If you wanted to, and if any
                                  reliable predictive information is
                                  available, then you could use it for
                                  tactical voting. (We're talking about
                                  voting in Approval).<br>
                                  <br>
                                </div>
                                b) If not, you could, if you wanted to,
                                try to estimate where, in the candidate
                                lineuup, your merit-expectation is, and
                                approve down to there, as an
                                expectation-maximizing strategy.
                                Depending on what is known or felt about
                                the relation between the distributions
                                of voters & candidates, you could
                                approve down to the mean, the mid-range,
                                or the median, of the candidates'
                                merits. <br>
                                <br>
                              </div>
                              Of course the median & midrange would
                              be easiest: The midrange is the point
                              halfway between the worst & the best.
                              But easiest of all is the median. You'd
                              approve the best half of the candidates.
                              That could be regarded as a rough estimate
                              for the other two central-tendency
                              measures, when they're difficult to
                              estimate.<br>
                            </div>
                          </div>
                        </div>
                      </blockquote>
                      <div><br>
                      </div>
                    </span>
                    <div>"Approve about half" is a good enough, easy to
                      remember guideline.  It would seem to maximize
                      your impact as well.  </div>
                  </div>
                </div>
              </div>
            </blockquote>
            <div><br>
            </div>
            <div>Yes, you're voting between the maximum number of
              candidate-pairs.<br>
              <br>
            </div>
            <div>[Replying farther down] :<br>
              <br>
              <br>
            </div>
            <div> <br>
            </div>
            <blockquote class="gmail_quote" style="margin:0 0 0
              .8ex;border-left:1px #ccc solid;padding-left:1ex">
              <div dir="ltr">
                <div class="gmail_extra">
                  <div class="gmail_quote">
                    <div> <br>
                      <br>
                      Whether "about half" is good enough does depend
                      where the frontrunners are in each voter's
                      ordering of candidates, but given that the
                      frontrunners are likely to be close to the median
                      across all voters anyway, then they will likely be
                      positioned near the median of most voters'
                      ordering.  <br>
                      <br>
                       <br>
                    </div>
                    <span class="">
                      <blockquote class="gmail_quote" style="margin:0 0
                        0 .8ex;border-left:1px #ccc
                        solid;padding-left:1ex">
                        <div dir="ltr">
                          <div>c) But you needn't bother with a) or b).
                            <br>
                          </div>
                          <div><br>
                          </div>
                          <div>Even without strong top &
                            bottom-sets, you can still take a guess
                            about which set you'd like to elect instead
                            of the other candidates. <br>
                            <br>
                            <br>
                          </div>
                          <div>Maybe, though you don't have strong top
                            & bottom sets, you have _ordinary_ top
                            & bottom sets, meaning that the merit
                            difference between the sets is greater (even
                            if not incomparably greater) than the merit
                            differences within those 2 sets.<br>
                          </div>
                          <div>If so, you likely will feel like
                            approving (only) all of your (ordinary)
                            top-set.<br>
                            <br>
                          </div>
                          <div>Or maybe even that isn't so, and you
                            don't have any kind of top & bottom
                            sets. Maybe the merit gradation is uniform,
                            without any gaps or natural dividing-lines.
                            What then? <br>
                            <br>
                          </div>
                          <div>Well, then you don't know where to make
                            your approval cutoff. You don't have an
                            obvious way to choose which set you want to
                            approve over the other. <br>
                            <br>
                          </div>
                          <div>No problem! If you don't know which set
                            approve, then it doesn't matter!<br>
                            <br>
                          </div>
                          <div>Just approve as you feel like. Maybe just
                            guess. Maybe flip a coin, or draw a number
                            from a bag. Or have the candidates' names in
                            a bag, and draw one to choose which one to
                            approve down to. If you don't know which set
                            you want to approve, then it doesn't matter
                            which set you approve.<br>
                            <br>
                          </div>
                          <div>Any such set that you choose by guessing
                            will include the best, and won't include the
                            worst, and will be within the range that you
                            feel that the approval cutoff should be in.
                            That's good enough! Don't worry about it.<br>
                          </div>
                          <div><br>
                          </div>
                          <div>Another thing: If, by guessing or drawing
                            from a bag, you make a choice of what set to
                            approve, but, when you start to actually do
                            so, you don't feel good about it, then don't
                            do it.<br>
                            <br>
                          </div>
                          <div>Maybe you'll say to yourself, "This is
                            _disgusting_ !"   Then of course don't do
                            it. Don't approve down that far. Go by your
                            feelings.<br>
                            <br>
                          </div>
                          <div>People who assume, as a starting premise,
                            that it's necessary to get the best
                            candidate possible are making things
                            unnecessarily difficult for themselves. Even
                            the more elaborate methods, the
                            ranking-methods, do do that as reliabliably
                            automcatically as their advocates sometimes
                            seem to believe.<br>
                            <br>
                          </div>
                          <div>By approving (only) your strong top-set,
                            or your ordinary top-set, or (absent either
                            of those) a set that is a good guess, within
                            the range where you feel that the approval
                            cutoff should be--By approving that set,
                            you're maximizing the probability of
                            electing from that set. <br>
                            <br>
                          </div>
                          <div>And that's good enough. <br>
                            <br>
                          </div>
                          <div>My message to those who complain that
                            Approval doesn't automatically elect the
                            best candidate that you can get is: You
                            worry too much.<span
                              class="m_-4678168771722237146HOEnZb"><font
                                color="#888888"><br>
                              </font></span></div>
                        </div>
                      </blockquote>
                      <div><br>
                      </div>
                    </span>
                    <div>I'm not so worried about electing the best.  I
                      would worry about electing a much worse candidate
                      in a surprising upset.  </div>
                  </div>
                </div>
              </div>
            </blockquote>
            <div><br>
            </div>
            <div>Then, in Approval, approve all of your strong top-set.<br>
            </div>
            <div><br>
               </div>
            <blockquote class="gmail_quote" style="margin:0 0 0
              .8ex;border-left:1px #ccc solid;padding-left:1ex">
              <div dir="ltr">
                <div class="gmail_extra">
                  <div class="gmail_quote">
                    <div>Elections really ought to be much more boring,
                      but not enough to put us to sleep.<br>
                    </div>
                  </div>
                </div>
              </div>
            </blockquote>
            <div><br>
            </div>
            <div>With honest elections and honest media, elections
              wouldn't be boring, because you'd be choosing among
              various versions of the very best. The choice among them,
              the discussion regarding their different approaches to the
              best policies and directions, would be anything but
              boring.<br>
              <br>
            </div>
            <div>What's boring is when the media keep claiming your
              choice is between two criminallyi-corrupt, bought
              candidates, and when people believe it.<br>
              <br>
            </div>
            <div>Michael Ossipoff<br>
            </div>
            <div><br>
              <br>
            </div>
            <div> </div>
            <blockquote class="gmail_quote" style="margin:0 0 0
              .8ex;border-left:1px #ccc solid;padding-left:1ex">
              <div dir="ltr">
                <div class="gmail_extra">
                  <div class="gmail_quote">
                    <div> <br>
                    </div>
                    <blockquote class="gmail_quote" style="margin:0 0 0
                      .8ex;border-left:1px #ccc solid;padding-left:1ex">
                      <div dir="ltr">
                        <div><span class="m_-4678168771722237146HOEnZb"><font
                              color="#888888"><br>
                            </font></span></div>
                        <span class="m_-4678168771722237146HOEnZb"><font
                            color="#888888">
                            <div>Michael Ossipoff <br>
                            </div>
                          </font></span></div>
                    </blockquote>
                    <div><br>
                    </div>
                    <div>I'm still planning to reply to a couple of your
                      earlier messages with a couple more comments. <br>
                    </div>
                  </div>
                  <span class="HOEnZb"><font color="#888888"><br
                        clear="all">
                      <br>
                      -- <br>
                      <div class="m_-4678168771722237146gmail_signature"
                        data-smartmail="gmail_signature">Daniel
                        LaLiberte<br>
                        <a moz-do-not-send="true"
                          href="mailto:daniel.laliberte@gmail.com"
                          target="_blank">daniel.laliberte@gmail.com</a><br>
                      </div>
                    </font></span></div>
              </div>
            </blockquote>
          </div>
          <br>
        </div>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">----
Election-Methods mailing list - see <a class="moz-txt-link-freetext" href="http://electorama.com/em">http://electorama.com/em</a> for list info
</pre>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <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: 2016.0.7924 / Virus Database: 4728/13557 - Release
        Date: 12/08/16</p>
    </blockquote>
    <p><br>
    </p>
  </body>
</html>