<div dir="auto"><div><br><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">El mar., 15 de mar. de 2022 5:19 a. m., Kristofer Munsterhjelm <<a href="mailto:km_elmet@t-online.de">km_elmet@t-online.de</a>> escribió:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On 13.03.2022 07:02, Forest Simmons wrote:<br>
<br>
> Warren and I gave an example an of a set of four factions in a single<br>
> plane that induces a cycle geometrically. This cannot happen with only<br>
> three factions ... if there is a cycle, then the preferences are<br>
> inconsistent with the lengths of the sides of the triangle. <br>
> <br>
> A three dimensional issue space can give rise to a cycle but those<br>
> preferences are not based on metrics/distances between factions.<br>
<br>
I think you're using a more constrained model than me. </blockquote></div></div><div dir="auto"><br></div><div dir="auto">Yes, in your example, apparently the candidates are not voters, or if they are, then there are six faction locations, not three.</div><div dir="auto"><br></div><div dir="auto">My assumption is that the geometric proximity of the three factions (fo each other) determines their preferences.</div><div dir="auto"><br></div><div dir="auto"><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Here's a spatial<br>
model example with three candidates and three faction centers that<br>
should work (unless I've miscalculated):<br>
<br>
Let A be at (0.85, 0.8), B at (0.2, 0.8), and C at (0.5, 0.4).<br>
<br>
Then let voter faction 1 be at (1.1, 1.5). The distance to A is ~0.75.<br>
The distance to B is ~1.14. The distance to C is ~1.25. So A>B>C.<br>
<br>
Let voter faction 2 be at (-0.65, 0.9). The distance to B is ~0.86. The<br>
distance to C is ~1.25. The distance to A is ~1.50. So B>C>A.<br>
<br>
Let voter faction 3 be at (0.8, -0.1). The distance to C is ~0.58. The<br>
distance to A is ~0.90. The distance to B is ~1.08. So C>A>B.<br>
<br>
See also <a href="https://www.rangevoting.org/BlackSingle.html" rel="noreferrer noreferrer" target="_blank">https://www.rangevoting.org/BlackSingle.html</a>. Quoting:<br>
<br>
> Myth #2 busted: It also is not the case that, in two dimensions with<br>
> voters who prefer candidates closer to them in L2 (or L1) distance, a<br>
> Condorcet winner necessarily exists at all.<br>
<br>
The page then proceeds to give an example with three voters (faction<br>
centers in your example) for the Manhattan distance.<br>
<br>
But my point is that even if it were impossible to create such examples<br>
in the plane, then we couldn't as a consequence assume thet the voters<br>
aren't voting in some more generalized space where it *is* possible.<br>
That would be like saying that just because a cycle can't exist with<br>
single-peaked preferences on a line, then all Condorcet cycles must be<br>
the result of strategy.<br>
<br>
In addition, there may be differential constraints that imply that if<br>
<br>
33: A>B>C<br>
33: B>C>A<br>
33: C>A>B<br>
<br>
is a perfect tie, then increasing the support for A and decreasing the<br>
support for C should not make C win. It may be (though I haven't proven<br>
so) that making the method elect C will induce some kind of<br>
monotonicity failure as a consequence.<br>
<br>
(From a spatial model, there's a similar argument: suppose we have 1/3<br>
of the voters at each point, then we move some voters from the point<br>
closest to C to the point closest to A. Changing the initial tie into a<br>
C win as a result seems somewhat strange.)<br>
<br>
So a choice to elect C would have to be about strategy resistance rather<br>
than honest performance.<br>
<br>
> If the cycle is false, though, then any faction could have done the<br>
> burial. You say the A faction is the only group that has anything to<br>
> win<br>
> by conducting the burial, so they must have done it. However, there's a<br>
> bit of battle-of-wits logic here. Suppose that the method did elect<br>
> C by<br>
> this logic. Then it's possible that the C voters, knowing this,<br>
> engineered the cycle (honest is C>B>A) in order to push their winner<br>
> from B (their second choice) to C (their first). <br>
> <br>
> <br>
> The burial of C by A isn't the only possibility ... just the most likely<br>
> to succeed ... so the one most in need of checking.<br>
> <br>
> If the check confirms B>C as sincere, then B wins, so the burial of B<br>
> did not pay off for C.<br>
> <br>
> Okay, so C can't win<br>
> because of second-order reasoning.<br>
> <br>
> <br>
> C can and will win if it was buried by A, because it is a finalist in<br>
> the B vs C honesty check.<br>
<br>
I was making a point about unrestricted domain. If the method passes<br>
unrestricted domain, then there's no such check, and you can't determine<br>
who the burying coalition is. Because it's a simple Condorcet cycle, no<br>
matter who the winner ends up being, there could have been a burying<br>
coalition who benefitted.<br>
<br>
What threw me, I think, is that you were arguing that C should be<br>
elected the winner of the single-winner ranked voting method, based on<br>
the given ballots.<br>
<br>
But on rereading, it seems that you're proposing that there's an<br>
(essentially automated) top-two runoff after the ordinary ranked method<br>
is finished, between the winner and the runner-up; and that the ranked<br>
method should elect C so that C is one of the two finalists in the<br>
second round, thus deterring first-round burial.<br>
<br>
In that case, the first-round method is not trying to figure out who the<br>
single winner of the election should be, but rather which set of two<br>
candidates will be the finalists for the second round - for either a<br>
manual runoff, or an automated one.<br>
<br>
And I would suspect that a single-winner (ranked, UD) method has a<br>
different objective than one that's intended to pick finalists for a<br>
runoff. A runoff finalist method can prioritize covering all the bases<br>
while a single-winner method is constrained to that the winner must be a<br>
reasonable compromise, and that the runner-up has these properties in<br>
case the winner is ineligible.<br>
<br>
<br>
A note about the automatic runoff mechanism: it's interesting in that it<br>
has different properties than Approval/Range, as far as breaking UD<br>
goes. It shows that not all UD-failing methods have to be ambiguous (the<br>
way ratings are) or lead to inherent instrumental voting even by honest<br>
voters. And that, in turn, means that not all extensions beyond UD are<br>
susceptible to the same problems as rated voting.[1]<br>
<br>
I would think manual runoff is better, though, for two reasons. First,<br>
it's a very subtle point that makes the honesty for the runoff ballot<br>
incentive compatible, and ordinary voters might not get why it's safe to<br>
be honest.<br>
<br>
Second, having separate rounds makes it possible to have finalist<br>
debates to examine their differences in more detail. The voters can then<br>
change their minds between rounds based on what they see - e.g. a weak<br>
CW could be exposed as a weak CW between the rounds if the runner-up's<br>
platform withstands greater scrutiny. This might be part of the reason<br>
why plurality + top two doesn't degrade into two-party rule the way<br>
ordinary plurality, contingent vote, and IRV does. (Then again, the<br>
necessary honest voting in the final round might be what makes TTR<br>
better than IRV.)<br>
<br>
Since the three-candidate method with a runoff is comparable to a "loser<br>
election" (where the loser is disqualified), it should be possible to<br>
rewrite my minimal manipulability finder to construct a minimally<br>
manipulable method in this context. It would take some work, however.<br>
<br>
-km<br>
<br>
[1] Maybe the relevant distinction is whether the proposed<br>
extension/breaking of UD would allow the method to pass IIA. If it does,<br>
*then* the blurring of context and method may follow, as it does in<br>
Range (e.g. IIA no longer implying that the winner doesn't change if<br>
candidates enter or exit). But this is just an idea.<br>
</blockquote></div></div></div>