<div dir="auto">That's pretty good ... not too bad for a deterministic method, if in a good faith, no pressure context. <div dir="auto"><br></div><div dir="auto">Some people would resort to sortition, especially if the decision is one that arises frequently ... every Friday, say.</div><div dir="auto"><br></div><div dir="auto">Other  lottery methods exist that reduce the amount of randomness (the entropy) but remain proportional (like random ballot).</div><div dir="auto"><br></div><div dir="auto">During a  ten year period winding down in 2015, Jobst and I intensively explored how much you can limit entropy (maximize the degree of consensus) under the constraint of minimal acceptable fairness ... which for us was the same kind of long term proportionality that random ballot favorite would ensure ... so min randomness required for proportional fairness ... or expressed differently ... max consensus with zero sacrifice  of strict statistically proportional representation.</div><div dir="auto"><br></div><div dir="auto">Our exploration culminated in Jobst's invention of MaxParC , a Maximum Partial Consensus method based on ballots where the voters specify (for each alternative) how many other voters would have to be on board before they would be willing to join with them in support of that alternative.</div><div dir="auto"><br></div><div dir="auto">From that information you can easily figure out which alternative has the potential for the most partial consensus... and that option will be awarded a lottery probability in proportion to that max potential support ... etc.</div><div dir="auto"><br></div><div dir="auto">That's the best known solution for maximising consensus without compromising proportional fairness.</div><div dir="auto"><br></div><div dir="auto">But in any context where MaxParC would give 100 percent consensus, there is always a simpler lottery method that would ensure the same result among rational voters who are already aware of the other voters' preferences ...  so they don't need to fill out MaxParC ballots.</div><div dir="auto"><br></div><div dir="auto">If rational, well informed voters already know among themselves that at least one candidate is a shoo-in for neing a 100 percent implicit approval winner (or a 100 percent consensus MaxParC winner), then a score ballot is all we need ...</div><div dir="auto"><br></div><div dir="auto">... among the candidates with 100 percent implicit approval, elect the one with the greatest score total.</div><div dir="auto"><br></div><div dir="auto">The projected MaxParC winner C should be among those 100 percent Implici Approval candidates.</div><div dir="auto"><br></div><div dir="auto">However, if (by some rare fluke) neither C nor any other candidate amassed100 percent implicit approval, then the winner must be chosen by random ballot favorite, in order to ensure fair proportionality.</div><div dir="auto"><br></div><div dir="auto">Does that seem to make sense?</div><div dir="auto"><br></div><div dir="auto">fws</div><div dir="auto"><br></div><div dir="auto"><br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Fri, Aug 18, 2023, 1:58 PM Colin Champion <<a href="mailto:colin.champion@routemaster.app">colin.champion@routemaster.app</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
  
    
  
  <div>
    <font face="Helvetica, Arial, sans-serif">Forest - I read fpdk's
      post as an implicit argument for cardinal voting (which was why it
      was relevant to STAR). Each friend states the utility of each
      topping to himself or herself: which topping do they choose
      collectively? And the answer is the one whose sum of utilities is
      greatest. I don't think there's a better answer.<br>
         CJC<br>
    </font><br>
    <div>On 18/08/2023 21:50, Forest Simmons
      wrote:<br>
    </div>
    <blockquote type="cite">
      
      <div dir="auto">He posed a pizza choice among friends pronlem.. a
        problem of consensus as opposed to "tyranny of the majority" ...
        how to find the best consensus decision when a simple majority
        first place preference would not be ideal.</div>
      <br>
      <div class="gmail_quote">
        <div dir="ltr" class="gmail_attr">On Fri, Aug 18, 2023, 10:41 AM
          Colin Champion <<a href="mailto:colin.champion@routemaster.app" target="_blank" rel="noreferrer">colin.champion@routemaster.app</a>>
          wrote:<br>
        </div>
        <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
          <div> <font face="Helvetica, Arial, sans-serif">Forest – I
              may be being slow, but... what problem are you trying to
              solve? The problem which fpdk (quite plausibly, to my
              mind) said was optimally solved by cardinal voting? Or the
              problem which I claimed was optimally solved by decision
              theory? Or something to do with tactical voting?<br>
                 CJC<br>
            </font><br>
            <div>On 18/08/2023 18:32, Forest Simmons wrote:<br>
            </div>
            <blockquote type="cite">
              <div dir="auto">It's been a while since I thought about
                this but here's something that somebody with some number
                crunching resources should experiment with ... a lottery
                method that I used to call "the ultimate lottery" back
                before Jobst invented MaxParC, which arguably has at
                least an equal claim to ultimateness:
                <div dir="auto"><br>
                </div>
                <div dir="auto">Ballots are positive homogeneous
                  functions of the candidate probability variables. The
                  homogeneity degree doesn't matter as long as all of
                  the  ballots are of the same degree.</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">The candidate probabilities are chosen
                  to maximize the product of the ballots.</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">This candidate probability distribution
                  can be realized as a spinner. The spinner is spun to
                  determine the winner.</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">How would this work for our pizza
                  example?</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">For example, each voter's ballot could
                  be her pizza desirability [score] expectation as a
                  function of the lottery probabilities.</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">Then each A faction voter would submit
                  the same ballot ... namely the function given by the
                  expression</div>
                <div dir="auto">100pA+80pC, while each B faction voter
                  would submit the expression</div>
                <div dir="auto">100pB+80pC.</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">When these ballots are multiplied
                  together, we get the product</div>
                <div dir="auto">(<span style="font-family:sans-serif">100pA+80pC)^60×(</span><span style="font-family:sans-serif">100pB+80pC)^40.</span></div>
                <div dir="auto"><span style="font-family:sans-serif"><br>
                  </span></div>
                <div dir="auto"><span style="font-family:sans-serif">The
                    p values that maximize this product (subject to the
                    constraint that they are non-negative and sum to 100
                    percent) are pA=pB=0, and pC=100%.</span></div>
                <div dir="auto"><span style="font-family:sans-serif"><br>
                  </span></div>
                <div dir="auto"><span style="font-family:sans-serif">The
                    lottery that maximizes the expectation product is
                    called the Nash lottery after John Nash who first
                    used this idea for efficient allocation of limited
                    resources.</span></div>
                <div dir="auto"><span style="font-family:sans-serif"><br>
                  </span></div>
                <div dir="auto"><font face="sans-serif">Since
                    expectations are linear combinations of the
                    probabilities, they are homogeneous of degree one
                    ... one person, on vote. Their product is
                    homogeneous of degree n ... so n people, n votes.</font></div>
                <div dir="auto"><font face="sans-serif"><br>
                  </font></div>
                <div dir="auto"><font face="sans-serif">Instead of using
                    voter expectations for their ballots, the voters
                    could have used other homogeneous expressions ...
                    for example, by simply replacing </font><span style="font-family:sans-serif">each sum of products
                    by a max of the same products.</span></div>
                <div dir="auto"><font face="sans-serif"><br>
                  </font></div>
                <div dir="auto"><font face="sans-serif">The product of
                    these modified ballots would be ...</font></div>
                <div dir="auto"><font face="sans-serif"><br>
                  </font></div>
                <div dir="auto"><span style="font-family:sans-serif">[max(100pA,80pC)]^60</span></div>
                <div dir="auto"><span style="font-family:sans-serif">×[max(</span><span style="font-family:sans-serif">100pB,80pC)]^40.</span><font face="sans-serif"><br>
                  </font></div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">Maximization of this product with the
                  same constraints as before, yields the same consensus
                  distribution ... pC=100%.</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">This information is new in the sense
                  that it has never been submitted for official
                  publication ... it's an exclusive bonus of Rob
                  Lanphier's EM list archive... first posted to this
                  list back in 2011 after Jobst and I published our 2010
                  paper on the use of mixed strategies for achieving
                  consensus.</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">Anyway, it turns out that using the Max
                  operator in place of the Sum operator yields a
                  distribution with less entropy whenever the two
                  distributions are not identical.</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">Less entropy means less randomness,
                  which means less chance, which in this context, means
                  more consensus.</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">In our example, the candidate
                  distribution turned out to be 100 percent  candidate C
                  ... zero randomness ... zero entropy ... 100 percent
                  consensus.</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">Now you can see why I mentioned the need
                  for number crunching capability ... experimenting with
                  these ballot product maximizations requires some
                  serious number crunching.</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">The field is wide open. Is the Ultimate
                  Lottery Method strongly monotonic? For that matter,
                  how about even the Nash Lottery?</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">Can MaxParC be formulated in terms of
                  the Ultimate Lottery?</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">Somebody with some grad students should
                  get them going on this!</div>
                <div dir="auto"><br>
                </div>
                <div dir="auto">fws</div>
              </div>
              <br>
              <div class="gmail_quote">
                <div dir="ltr" class="gmail_attr">On Thu, Aug 17, 2023,
                  11:10 AM Forest Simmons <<a href="mailto:forest.simmons21@gmail.com" rel="noreferrer noreferrer" target="_blank">forest.simmons21@gmail.com</a>>
                  wrote:<br>
                </div>
                <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
                  <div dir="auto">Suppose voter utilities for three
                    kinds of pizza are
                    <div dir="auto"><br>
                    </div>
                    <div dir="auto">60 A[100]>C[80]>>B[0]</div>
                    <div dir="auto">40 B[100]>C[80]>>A[0]</div>
                    <div dir="auto"><br>
                    </div>
                    <div dir="auto">Suppose the voters must choose by
                      majority choice between pizza C and the favorite
                      pizza of a voter to be determined by randomly
                      drawing a voter name from a hat.</div>
                    <div dir="auto"><br>
                    </div>
                    <div dir="auto">The random drawing method would give
                      voter utility expectations of</div>
                    <div dir="auto"><br>
                    </div>
                    <div dir="auto">60%100+40%0 for each A groupie, and</div>
                    <div dir="auto">40%100+60%0 for each B groupie.</div>
                    <div dir="auto"><br>
                    </div>
                    <div dir="auto">The max utility expectation would be
                      60.</div>
                    <div dir="auto"><br>
                    </div>
                    <div dir="auto">On the other hand, if voters decide
                      to go with the sure deal C, the assured utility fo
                      every voter will be 80.</div>
                    <div dir="auto"><br>
                    </div>
                    <div dir="auto">Every rational voter faced with this
                      choice will choose C.</div>
                    <div dir="auto"><br>
                    </div>
                    <div dir="auto">Here we have an ostensibly random
                      method that is sure to yield a consensus decision
                      when voters vote ratkonally.</div>
                    <div dir="auto"><br>
                    </div>
                    <div dir="auto">More on this topic at</div>
                    <div dir="auto"><br>
                    </div>
                    <div dir="auto"><a href="https://www.researchgate.net/figure/Properties-of-common-group-decision-methods-Nash-Lottery-and-MaxParC-Solid-and-dashed_fig3_342120971" style="font-family:sans-serif" rel="noreferrer
                        noreferrer noreferrer" target="_blank">https://www.researchgate.net/figure/Properties-of-common-group-decision-methods-Nash-Lottery-and-MaxParC-Solid-and-dashed_fig3_342120971</a><span style="font-family:sans-serif"> </span><br>
                    </div>
                    <div dir="auto"><span style="font-family:sans-serif"><br>
                      </span></div>
                    <div dir="auto"><span style="font-family:sans-serif">fws</span></div>
                  </div>
                  <br>
                  <div class="gmail_quote">
                    <div dir="ltr" class="gmail_attr">On Thu, Aug 17,
                      2023, 1:18 AM Forest Simmons <<a href="mailto:forest.simmons21@gmail.com" rel="noreferrer noreferrer noreferrer noreferrer" target="_blank">forest.simmons21@gmail.com</a>>
                      wrote:<br>
                    </div>
                    <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
                      <div dir="auto">The best methods that I know of
                        for the friends context are minimum entropy
                        lottery methods characterized by max possible
                        consensus (min entropy) consistent with a
                        proportional lottery method with higher entropy
                        fallback to disincentivize  gratuitous
                        defection.
                        <div dir="auto"><br>
                        </div>
                        <div dir="auto">Jobst's MaxParC (Max Partial
                          Consensus) is the best example.</div>
                        <div dir="auto"><br>
                        </div>
                        <div dir="auto">Too late to elaborate tonight.</div>
                        <div dir="auto"><br>
                        </div>
                        <div dir="auto">fws<br>
                          <div dir="auto"><br>
                          </div>
                          <div dir="auto">I'll </div>
                        </div>
                      </div>
                      <br>
                      <div class="gmail_quote">
                        <div dir="ltr" class="gmail_attr">On Wed, Aug
                          16, 2023, 10:01 AM <<a href="mailto:fdpk69p6uq@snkmail.com" rel="noreferrer noreferrer noreferrer
                            noreferrer noreferrer" target="_blank">fdpk69p6uq@snkmail.com</a>>
                          wrote:<br>
                        </div>
                        <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
                          <div dir="ltr"><br>
                            <div class="gmail_quote">
                              <div dir="ltr" class="gmail_attr">On Mon,
                                Aug 14, 2023 at 12:09 AM C.Benham wrote:<br>
                              </div>
                              <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
                                >   I think this is an interesting
                                point. We can ask at a philosophical
                                level what makes a good voting method.
                                Is it just one that ticks the most
                                boxes, or is it one that most reliably
                                gets the "best" result?<br>
                              </blockquote>
                              <div><br>
                              </div>
                              <div>The one that most reliably gets the
                                best result in the real world. The
                                difficulty with this approach is
                                accurately modeling human voting
                                behavior and the consequent utility
                                experienced from the winner, but it's
                                still the better answer philosophically.</div>
                              <div><br>
                              </div>
                              <div>(Note that VSE predates Jameson Quinn
                                by decades, and has had several
                                different names: <a href="https://en.wikipedia.org/wiki/Social_utility_efficiency" rel="noreferrer noreferrer noreferrer
                                  noreferrer noreferrer noreferrer" target="_blank">https://en.wikipedia.org/wiki/Social_utility_efficiency</a>)</div>
                              <div><br>
                              </div>
                              <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"> >
                                And that's partly because the premise of
                                Condorcet is essentially built on a
                                logical fallacy - basically that if A is
                                preferred to B on more ballots that vice
                                versa then electing A must<br>
                                > be a better result than electing B.<br>
                                <br>
                                I'd be interested in reading your
                                explanation of why you think that is a <br>
                                "logical fallacy".  What about if there
                                are only two candidates?<br>
                              </blockquote>
                              <div><br>
                              </div>
                              <div>Ranked ballots can't capture strength
                                of preference. It's possible for a
                                majority-preferred candidate to be very
                                polarizing (loved by 51% and hated by
                                49%), while the minority-preferred
                                candidate is broadly-liked and has a
                                much higher overall
                                approval/favorability rating.  Which
                                candidate is the rightful winner?<br>
                              </div>
                              <div><br>
                              </div>
                              <div><a href="https://leastevil.blogspot.com/2012/03/tyranny-of-majority-weak-preferences.html" rel="noreferrer noreferrer noreferrer
                                  noreferrer noreferrer noreferrer" target="_blank">https://leastevil.blogspot.com/2012/03/tyranny-of-majority-weak-preferences.html</a></div>
                              <div><br>
                              </div>
                              <div>"Suppose you and a pair of friends
                                are looking to order a pizza. You, and
                                one friend, really like mushrooms, and
                                prefer them over all other vegetable
                                options, but you both also really, <i>really</i>
                                like pepperoni. Your other friend also
                                really likes mushrooms, and prefers them
                                over all other options, but they're also
                                vegetarian. What one topping should you
                                get?
                                <p>Clearly the answer is mushrooms, and
                                  there is no group of friends worth
                                  calling themselves such who would
                                  conclude otherwise. It's so obvious
                                  that it hardly seems worth calling
                                  attention to. So why is it, that if we
                                  put this decision up to a vote, do so
                                  many election methods, which are
                                  otherwise seen as perfectly reasonable
                                  methods, fail? Plurality, <a href="http://en.wikipedia.org/wiki/Two-round_system" rel="noreferrer noreferrer
                                    noreferrer noreferrer noreferrer noreferrer" target="_blank">top-two
                                    runoffs</a>, <a href="http://en.wikipedia.org/wiki/Instant-runoff_voting" rel="noreferrer noreferrer
                                    noreferrer noreferrer noreferrer noreferrer" target="_blank">instant
                                    runoff voting</a>, all variations of
                                  <a href="http://en.wikipedia.org/wiki/Condorcet_method" rel="noreferrer noreferrer
                                    noreferrer noreferrer noreferrer noreferrer" target="_blank">Condorcet's
                                    method</a>, even <a href="http://en.wikipedia.org/wiki/Bucklin_voting" rel="noreferrer noreferrer
                                    noreferrer noreferrer noreferrer noreferrer" target="_blank">Bucklin
                                    voting</a>; all of them,
                                  incorrectly, choose pepperoni."</p>
                              </div>
                              <div>(And strength of preference is
                                clearly a real thing in our brains.  If
                                you prefer A > B > C, and are
                                given the choice between Box 1, which
                                contains B, and Box 2, which has a 50/50
                                chance of containing A or C, which do
                                you choose?  What if the probability
                                were 1 in a million of Box 2 containing
                                C?  By varying the probability until
                                it's impossible to decide, you can
                                measure the relative strength of
                                preference for B > C vs A > C.)<br>
                              </div>
                            </div>
                          </div>
                          ----<br>
                          Election-Methods mailing list - see <a href="https://electorama.com/em" rel="noreferrer noreferrer noreferrer
                            noreferrer noreferrer noreferrer noreferrer" target="_blank">https://electorama.com/em</a>
                          for list info<br>
                        </blockquote>
                      </div>
                    </blockquote>
                  </div>
                </blockquote>
              </div>
              <br>
              <fieldset></fieldset>
              <pre>----
Election-Methods mailing list - see <a href="https://electorama.com/em" rel="noreferrer noreferrer" target="_blank">https://electorama.com/em</a> for list info
</pre>
            </blockquote>
            <br>
          </div>
          ----<br>
          Election-Methods mailing list - see <a href="https://electorama.com/em" rel="noreferrer noreferrer noreferrer" target="_blank">https://electorama.com/em</a>
          for list info<br>
        </blockquote>
      </div>
    </blockquote>
    <br>
  </div>

----<br>
Election-Methods mailing list - see <a href="https://electorama.com/em" rel="noreferrer noreferrer" target="_blank">https://electorama.com/em</a> for list info<br>
</blockquote></div>