[EM] Voting-System Choice for Polls (Just one more thing I want to say)

[Juho Laatu]

> On 20 Dec 2016, at 01:33, Michael Ossipoff <email9648742 at gmail.com> wrote:

> We have no reason to worry about offensive strategy in CIVS polls.
> 
> And so MAM is the best choice for CIVS polls. MAM is one of the rank-counts that CIVS supports.

Ok, MAM is a quite decent method. If strategic voting is minimal, I guess you pick the method that gives best results with sincere votes. I'm still lacking good understanding of why exactly MAM would produce best winners with sincere votes.

>> When offensive strategy is used against the CWs, s/he'll be in the voted Smith-set. That's a good reason to elect from the voted Smith-set, so that the CWs can maybe be rescued.
> 
> In theory that might save the CWs in some very rare cases.
> 
> I can't say why it's important to guarantee that the winner is in the voted Smith-set, other than that it confers compliance with various criteria, including Mutual-Majority.

That's a good answer since there are situations where one can justify electing from a Smith set that consists of very similar minded clones. The top cycle can be sincere or strategic. The same matrix can however probably be also a result of sincere votes where the Smith set candidates are not clones but a much more competitive group. In such cases it could make sense to elect outside the Smith set. One key point here is that information in the matrix is limited, and it is impossible to say if there are clones, and which preferences are strong and which ones weak.

> But, because one member of the voted Smith set will be the CWs, if truncation or burial is being attempted against it (neuter gender because I'm talking about poll-alternatives) then that means that disqualifying non-members of the voted Smith-set narrows the field in a way favorable to the CWs.

It is not possible to tell if there was a CWs. The top loop can be sincere as well (or a result of various good and bad strategies). And the best sincere winner might be also outside the Smith set.

> Note that electing outside the Smith set may sometimes also help us, e.g. by making strategic plans void by not electing the (possibly strongly top-looped) favourite of the strategists.
> 
> Maybe, but, by pairwise-count standards, electing from the Smith set keeps the winner among the pairwise publicly-favored candidates.

I don't understand your definition of "pairwise publicly-favored candidates". Why can't candidates outside the Smith set be such candidates? I mean situations where the leading candidate outside the Smith set could lose by one vote to all the Smith set members. Losing only marginally to many candidates could be a better result than losing a lot to fewer candidates.

> Of course, in an election, when Approval elects the candidate approved, considered satisfactory by the most people, is probably more important than electing the CWs or from the sincere Smith-set.

Also pairwise comparisons can sometimes lead to situations where best candidate (based on those pairwise comparisons) is found outside the Smith set. (see the marginal loss example above)

> MAM, by itself would surely, it seems to me, be the best way to find the CWs, for sincere voters.

All Condorcet methods can find the CWs with sincere votes.

> I'm sure there will always be some irrational behaviour, but that might be just general noise, and would not influence the election. If some large grouping will use some irrational strategy, and as a result will make the outcome of the election worse from their point of view, maybe they will learn and will vote rationally (sincerely or strategically) in the next election.
> 
> If the CWs is  in your strong bottom-set, then burial, even when it risks electing a worse member of your strong bottom-set, isn't irrational, if there's _any_ chance of successfully changing the winner to someone in your strong top-set.
> 
> If we were using Condorcet, and if the expected CWs were a Repugnocrat, I'd routinely use burial strategy, every time, in hopes of electing a progressive. That's good strategy.

Ok, if your preferences are A >>>>> B > C, and your guess is that B is a CWs, then it would be quite risk free to try to create cycles. It is however quite possible that your vote will be part of the "general noise". Some C supporters could also compromise (= top rank B) if they believe that B's victory is not certain, and some B supporters could rank A lowest if they believe that many A supporters plan to bury B. Many A supporters would not like the idea of increasing the chances of C winning the election. Many voters would vote sincerely since that's what they believe is the right thing to do. There are so many parameters and uncertainty in polls that coherent strategic voting may not be common. It may not be easy to get sufficient number of strategic voters to follow the plan.

I have been asking for generic strategic guidance for Condorcet elections. I think you are close to proposing one rule here. Maybe something like "if you strongly dislike the expected winner, bury it under all such candidates that are not much worse than it". Maybe add "if some of your favourite candidates are expected to beat those candidates". Maybe you can improve my formulation of that possible strategic advice. Maybe simulations could tell if that strategy does more good than harm, or nothing at all (assuming some nice distribution of preference strengths among voters).

> One more observation on how I would approach the problem of finding a Condorcet method that is sufficiently strategy proof. I would pick first the Condorcet method that produces best possible results with sincere votes. Then I'd let EM members watch one episode of some soap opera, and then vote on who was the best actor. That would be the first poll. Then they would watch another episode, and vote again in the second poll. Then they would watch one more episode, and cast their final vote. People would be allowed to discuss about their preferences, so that similar minded people would find each others, and they could plan strategies for the final election (and why not also for the polls). People would also be encouraged to publish strategic advice to others. People could also attach information on how they voted strategically in their ballot (note, not after the election but only in the ballot when voting). This whole process would be repeated few times.
> 
> But people would then be using counterstrategy agains the declared offensive strategy, and that would make it unrealistic, because in real elections, the voters might not be as adept at counterstrategy.

The strategy would be declared only if the strategists consider that useful. If they don't publish, other voters may not be able to follow the strategy. If they publish, other voters might react to that in many ways. Just like in real elections.

Of course we are talking about a very knowledgeable and possibly also strategic voter community here. In real life elections everything would be one step milder. If some method can survive this test, real life elections could be an easier case.

> Why not just, instead, work out hypothetical results out on paper, instead of using actual voting?

Yes, that works in theory. In real life I have seen too many times scenarios where the claimed vulnerabilities are based on having exact information on what the preferences will be at the election day, and ability to change the voting behaviour of all those voters (100%) that would have the strategic interest, and all other voters not changing their behaviour in any way, and not applying any other strategies. That's what I would call hypothetical and unrealistic :-). I created that testing method to reflect the uncertainties that are present in real elections. If you are disciplined enough to put all that information on paper, with all the uncertainties, that's ok too. But the normal line of discussion at EM is not disciplined enough to achieve that.

> But, ever since EM started, I've always felt that poll are super useful & valuable, as the only way to find out what it's like to use the various voting-systems. You probably know that I started lots of polls at EM. They've fallen out of fashion, and I consider that regrettable, because you can't fully evaluate  & compare voting-systems without actually trying them. Voting in polls, and counting polls. 
> 
> So yes, I'm in favor of polls at EM. Regrettably no one else is.

Yes, polls can be very helpful. And paper work sometimes too, if disciplined enough.

>  I have great trust on the skills of people on this mailing list, but so far I have not seen any good strategic guidance for voters or parties in real life Condorcet elections.
> 
> Well, I mentioned, above, that I'd routinely use burial strategy when the expected CWs is in my strong bottom-set, in an official Condorcet election.

Yes. If you can, please upgrade my description above to something that you believe to be valid guidance to regular voters. (Alternatively you can give also strategic rules for parties, that may then give further guidance to their supporters).

Here's another one that may work e.g. against some minmax style Condorcet methods. "If some unwanted party/grouping has numerous equally disliked candidates, rank them in a loop in alphabetical order, starting from a random candidate".

BR, Juho


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20161220/f00e94d9/attachment-0001.htm>