<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="Generator" content="Microsoft Exchange Server">
<!-- converted from text --><style><!-- .EmailQuote { margin-left: 1pt; padding-left: 4pt; border-left: #800000 2px solid; } --></style>
</head>
<body>
<div>
<div>Kristofer,</div>
<div><br>
</div>
<div>your points are all well taken, well articulated,..and illuminating.</div>
<div><br>
</div>
<div> I am abandoning an effort to distinguish "sincere strategy" from "manipulative strategy," since "sincere strategy" is just an oxymoron from one point of view, and from another it just means "following a ballot strategy that sincerely optimizes one's expected
outcome under the rules of the game," as opposed to bluffs, threats, intimidations, biased polls, and similar manipulations designed to deceive the other voters into voting sub-optimal ballots. [not to mention outright voter suppression still common here in
many states] That would be a pretty low bar to meet in more civil democracies!</div>
<div><br>
</div>
<div>I would be very content to have an optimum (i.e. unique & optimal) strategy method for electing the clearly defined sincere winner of the same ballot set under normal rules.</div>
<div><br>
</div>
<div>That is what "choose a branch at each juncture" has going for it: the optimum perfect-info strategy under top-down rules yields (with complete certainty) the same winner that the same ballots would if they were counted under the (more natural) bottom-up
rules under zero info (i.e. sincere) voting.</div>
<div><br>
</div>
<div>This gives us a clean way of side-stepping Satterthwaite's result without contradicting it in the least!</div>
<div><br>
</div>
<div><br>
</div>
<div><br>
</div>
<div id="x_composer_signature">
<div dir="auto" style="font-size:85%; color:#575757">Sent from my MetroPCS 4G LTE Android Device</div>
</div>
<div>-------- Original message --------</div>
<div>From: Kristofer Munsterhjelm <km_elmet@t-online.de> </div>
<div>Date: 7/21/21 2:36 AM (GMT-08:00) </div>
<div>To: Susan Simmons <suzerainsimmons@outlook.com>, election-methods@lists.electorama.com
</div>
<div>Subject: Re: [EM] Manipulation Resistant Voting </div>
<div><br>
</div>
</div>
<font size="2"><span style="font-size:11pt;">
<div class="PlainText">On 7/19/21 1:01 AM, Susan Simmons wrote:<br>
> >Now suppose there's a typical Condorcet cycle A>B>C>A and the agenda<br>
> ordering is A>B>C; so the method proceeds by matching B and C, and then<br>
> matching the winner with A, and then the outcome of this is the final<br>
> winner.<br>
> <br>
> Good point ... just concatenating binary choices together is not <br>
> enough. It is important that the outcome of each binary decision be <br>
> final, not just the last decision in the sequence.<br>
> <br>
> Suppose, for example, that the method, at every stage, either (1) <br>
> accepts the next remaining agenda item X as the final winner or (2) <br>
> eliminates X and applies the method recursively to the remainder of the <br>
> list?<br>
> <br>
> Then every C supporter would vote sincerely to stick with C, and the A <br>
> supporters would vote sincerely to eliminate C. The B>C supporters would <br>
> sincerely vote to keep C, while the B>A supporters would sincerely vote <br>
> to eliminate C.<br>
> <br>
> So the winner would be A or C depending on whether or not the B>A <br>
> faction was larger than the B>C faction.<br>
<br>
I think Benham can be defined that way. At every stage, either X is <br>
accepted as the final (because he beats every candidate ranked higher <br>
than him), or X is eliminated and the method is recursively applied to <br>
the rest of the list.<br>
<br>
The list, then, consists of the winners in IRV order, last man standing <br>
ranked first.<br>
<br>
But Benham is just IRV if all the candidates are in a top cycle. So if <br>
we have IRV order A>B>C and an ABCA cycle, wouldn't the point stand above?<br>
<br>
Before strategy: the method checks whether C is the CW. It is not, so <br>
then C is eliminated. Then it checks if B is the CW among the remaining <br>
candidates. It isn't, so B is eliminated, then A wins.<br>
<br>
If the method is manual, there must be something the voters can report <br>
on in the agenda method itself. If pairwise preferences are hoisted from <br>
the IRV ballots, then the agenda method is entirely automatic, so there <br>
would be nothing to report.<br>
<br>
So since the method has to ask the voters *something* in the agenda <br>
method itself, that something must be whether the last placed candidate <br>
is a CW. The manual Benham method would take form of first calculating <br>
the IRV order, then asking the voter "does the last placed candidate <br>
beat everybody else pairwise?".<br>
<br>
Now suppose you're a B>C>A voter. In the first round, if you report your <br>
true preference ("B>C>A"), this would indicate, along with everybody <br>
else's true preferences, that C is not the CW. So C is eliminated, after <br>
which A wins. So you should instead falsify your preference to C>B>A so <br>
that C's chance of becoming a CW is maximized.<br>
<br>
That's what you say above ("the B>C voters would sincerely vote to keep <br>
C"), but it doesn't strike me as very sincere. See more below.<br>
<br>
> So given the agenda, and the special finality rule (once an option is <br>
> chosen the winner must be a member of that option) all of the rational <br>
> choices will be sincere in the sense that they are the choices whose <br>
> rational outcomes are preferred.<br>
> .....<br>
> <br>
> >In reference to an earlier post of mine, suppose we define honesty as<br>
> what a Random Ballot type method would return. In your method, this<br>
> would be a trace down the tree that ends at the candidate who the voter<br>
> prefers most of the candidates at the bottom level of the tree.<br>
> <br>
> But a voter is not at liberty to choose an entire trace. You probably <br>
> mean that as long as Favorite is a descendant of the current node, if <br>
> you do not choose the branch that leads to Favorite, then your vote is <br>
> insincere.<br>
> <br>
> If that's your definition of "sincere" in this context, then you are <br>
> right ... sincere strategy is not rational.<br>
> <br>
> But it seems to me that when presented with only two choices, branchA or <br>
> branchB, leading almost surely to final wins for X and Y, respectively, <br>
> then voting for branchA should not be impunged as "insincere strategy," <br>
> if you truly prefer X over Y.<br>
> <br>
> It's a matter of definition. So perhaps we could distinguish between <br>
> "naive sincerity" and "rational sincerity."<br>
<br>
This is a bit too reminiscent of the cardinal supporters' arguments that <br>
Burr dilemmas can't happen because you'd be using polls to anticipate <br>
the support of the relative candidates, so that adjusting your Approval <br>
vote to either Approve of Best and Good or only Best is not a tactical <br>
burden on the voter.<br>
<br>
That said, I seem to have misunderstood your method. So I'll describe it <br>
(and my insincerity argument) and then let's see if I understood it <br>
properly.<br>
<br>
Your method consists of two stages. In the first stage, what's <br>
essentially a hierarchical clustering is constructed as a binary tree <br>
with all the candidates at the leaves. Then the tree is published and <br>
the voters either participate in log(n) runoffs (if it's a manual <br>
method) or submit the necessary information to automatically do those <br>
runoffs (if it's an automatic one).<br>
<br>
To revise my example, the tree is something like:<br>
<br>
----- A<br>
-----+<br>
| ----- D<br>
----+<br>
| ----- B<br>
-----+<br>
----- C<br>
<br>
Now suppose that a voter's mental preference ordering of the candidates <br>
is A>B>C>D. In a Random Ballot/Dictator situation, the voter would first <br>
choose the top branch, then the top branch again to arrive at A.<br>
<br>
Now suppose there's an election, and the voter has to choose whether to <br>
support the top branch or the bottom branch from the root. Suppose <br>
furthermore that the voter knows that if the top branch is chosen, the <br>
outcome is very likely to be D. Now that voter has an incentive to lie <br>
about his decision relative to a random dictator situation, by throwing <br>
his support in favor of the bottom branch so that the lesser evil wins.<br>
<br>
That, I would say, is strategy. It may not *seem* like strategy because <br>
the ballot format fails universal domain. Like Approval, the voter is <br>
limited in what he can express by the state of the rest of the <br>
electorate; and there exists a choice that is consistent with his <br>
preferences, conditioned upon what the rest of the electorate will let <br>
him do. The voter is being honest under the constraint (i.e. he's not <br>
reversing any preferences, because his choice is consistent with trying <br>
to optimize the expected outcome). However, he still has to adapt his <br>
choice to what the other voters are doing.<br>
<br>
The comparison to Approval is that in a Burr dilemma, a voter isn't <br>
preference-reversing (e.g. Approving both Best and Bad when his <br>
preference order is Best > Good > Bad). But the failure of universal <br>
domain exhibited by the Approval ballot means that he's still <br>
constrained by the rest of the voters: the proper way to vote is Best <br>
only if Bad is far behind, or Best and Good if Bad is close.<br>
<br>
Now, there is a difference to Approval. A voter who truly values honesty <br>
above all else can behave as if under Random Ballot in your method; <br>
that's impossible in Approval. But your method and Approval are only <br>
strategy-proof under a definition of honesty where there are multiple <br>
sincere votes, and the voter adapts to the electorate by choosing one of <br>
the many sincere votes.<br>
<br>
In the stricter Gibbardian sense that "it's strategy if you have to <br>
adapt your answer to what other people are doing", then both Approval <br>
and your method are susceptible to strategy. This stricter sense is what <br>
stops feedback cycles, regret after anticipating wrong, etc.<br>
<br>
As I don't like the sort of manual DSV that's implied by Approval, I'm <br>
inclined to use the stricter definition, and so I probably wouldn't like <br>
the method above, either. (Well, it *might* be a good method if paired <br>
with DSV and accepting a ranked ballot; I don't know whether it would.)<br>
<br>
<br>
<br>
Consider this analogy: Say we have a two-step method where, in the first <br>
round, the Benham and Ranked Pairs winners are chosen, and if they <br>
differ, there's a manual runoff for the second round.<br>
<br>
Now suppose for the sake of the analogy that you *know* everybody else's <br>
honest pairwise preferences so you can perfectly predict the outcome of <br>
the second round. In particular, in a runoff between A and D, D will <br>
win; and in a runoff between B and C, C will win.<br>
<br>
Furthermore, your honest preference is A>B>C>D, and you know that if you <br>
vote A>B>C>D, then A and D will be the Benham and RP winners <br>
respectively, whereas if you vote B>C>A>D, B and C will be the winners.<br>
<br>
Isn't then voting B>C>A>D in the first round strategy? If so, why is <br>
taking the bottom branch any different?<br>
<br>
It seems to me that the only difference is that B>C>A>D feels wrong <br>
because it's indisputably tactical voting in either Benham or RP alone, <br>
so we know it's dishonest. But it's easier to conceal the strategic <br>
behavior when your hand is forced by the constraints placed upon you by <br>
the method, so in the context of a two-round system, voting B>C>A>D is <br>
also "rational sincerity" in the sense that choosing the bottom branch is.<br>
<br>
> So one can say naive sincerity is violated by this choose-a-branch <br>
> method, but (it seems to me) realistic, rational, strategic sincerity <br>
> that accepts the premise of the method is not violated.<br>
> <br>
> The "premise" is that when you choose a branch, you are showing a <br>
> preference for the likely winner of that branch over the likely winner <br>
> of the other branch ... NOT implying support for any other preference. <br>
> So it cannot be dishonest or "falsification" unless you actually prefer <br>
> the likely winner of the other branch over the one you chose.<br>
> <br>
> <br>
> I think this notion of sincerity gives us some traction ... the other <br>
> one is too stringent to lead to a manipulation proof Condorcet method.<br>
<br>
Right. I'm perfectly content to say that no Condorcet method (for that <br>
matter, no deterministic *election method!*) is strategy-proof, and <br>
point at Gibbard. In the strict sense "you don't have to anticipate <br>
other people's responses", it's perfectly true.<br>
<br>
But methods that fail universal domain and so collapse some subset of <br>
strategic (in the Gibbard sense) voting into sincere ballots leave too <br>
much of a burden on the minds of the voter. If the voters' model of <br>
reality is off too much, then they may end up regretting making the <br>
choices they did.<br>
<br>
And I would like mental burden to be a game you choose: if you're <br>
honest, just go ahead and submit your true preference. Otherwise, if you <br>
want to play (devise a strategy) then you accept taking the heat (chance <br>
of backfire, regret, etc.) but that's a deliberate choice you're making.<br>
<br>
Now, you *could* be naively sincere in your method, but I have the <br>
intuition that it would perform rather badly in that case, because it <br>
doesn't have contingency logic.<br>
</div>
</span></font>
</body>
</html>