<div dir="auto">Richard, I believe Double defeat Hare is also known as RCIPE.<div dir="auto"><br></div><div dir="auto"><a href="https://electowiki.org/wiki/Ranked_Choice_Including_Pairwise_Elimination">https://electowiki.org/wiki/Ranked_Choice_Including_Pairwise_Elimination</a><br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Sat, Jun 22, 2024, 07:43 Richard, the VoteFair guy <<a href="mailto:electionmethods@votefair.org">electionmethods@votefair.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On 6/22/2024 1:31 AM, Chris Benham wrote:<br>
> I think those of us who are extra hung-up about frustrating Burial<br>
> should look more closely at Double Defeat, Hare. ...<br>
<br>
What is Double Defeat Hare?<br>
<br>
I didn't try to find it on Electowiki because lately that website has <br>
not been responding when I click on links to it.<br>
<br>
Richard Fobes<br>
the VoteFair guy<br>
(and now the only Richard here)<br>
<br>
<br>
On 6/22/2024 1:31 AM, Chris Benham wrote:<br>
> Kevin,<br>
> <br>
> Something I didn't address in my previous email:<br>
> <br>
>>> And what is wrong with your "Improved<br>
>>> Condorcet Approval" method ? I think it would be good using<br>
>>> unrestricted ranking ballots with an explicit approval cutoff.<br>
>> ICA or C//A (implicit) are not bad. They don't satisfy SFC. In my <br>
>> recent simulations<br>
>> on frontrunner truncation strategy, C//A is among the best Condorcet <br>
>> methods. In<br>
>> random elections I am disturbed that ICA and C//A are worse than WV <br>
>> methods at<br>
>> strong FBC (i.e. what I call compromise incentive).<br>
>><br>
>> You've asked me many times about C//A(explicit) and I still think it's <br>
>> bad. The<br>
>> entire notion of C//A(implicit) being good at deterring burial is <br>
>> based on the<br>
>> fact that if you use burial to prevent there from being a Condorcet <br>
>> winner, then in<br>
>> the "cycle resolution" you cannot prefer any candidate to the one you <br>
>> raised<br>
>> insincerely.<br>
>><br>
>> In C//A(explicit), burial only backfires if it actually creates a fake <br>
>> CW. Creating<br>
>> a fake cycle is never bad for your favorite.<br>
> <br>
> In these mass public elections the chance that an individual voter's <br>
> ballot will be pivotal is negligible. I would think that if the method <br>
> is making some attempt to minimise the number of "wasted votes", then <br>
> many voters would want to be able to express their full sincere ranking <br>
> and also would at least not mind giving their sincere or semi-sincere <br>
> approval cutoff. For example in a Hare (aka IRV) election (with <br>
> unrestricted strict ranking from the top) I can't imagine it ever <br>
> crossing my mind to vote insincerely (and I never have).<br>
> <br>
> Also I think the method needs to have some justification on the <br>
> assumption that the voters are sincere. So if voters complain "Why <br>
> aren't I allowed to rank among the candidates I don't approve?" and "How <br>
> do we know that the voted CW is really the CW if we have been forced to <br>
> truncate our rankings?", we need some answer that isn't just about <br>
> "deterring burial". We could say "Well this is more simple, and it's <br>
> not strictly a Condorcet method but rather an Approval-Condorcet hybrid <br>
> that is trying to produce a high SU winner" but I expect that a lot of <br>
> voters would not be fully satisfied with that answer. (I wouldn't be.)<br>
> <br>
> I have become firmer in my support for Double Defeat (explicit) methods <br>
> and my favourite of these is Margins-Sorted Approval. Those method allow <br>
> voters to rank however many candidates they wish and also give a <br>
> approval cutoff, and no candidate that is pairwise-beaten by a more <br>
> approved candidate is allowed to win. Much of the time this is by itself <br>
> decisive.<br>
> <br>
> I think those of us who are extra hung-up about frustrating Burial <br>
> should look more closely at Double Defeat, Hare. I suspect that it has <br>
> most of the Condorcet efficiency of an actual Condorcet method while <br>
> retaining most of the Burial resistance of Hare, while possibly tending <br>
> to give higher SU winners than both.<br>
> <br>
> Chris B.<br>
> <br>
> On 21/06/2024 9:32 pm, Kevin Venzke wrote:<br>
>> Hi Chris,<br>
>><br>
>>> On Mono-add-Plump as a weak version of Participation:<br>
>>>> Yes but almost all proposals fail Participation, so we will be in a <br>
>>>> lot of trouble<br>
>>>> if we insist on this kind of thinking.<br>
>>> What sort of "trouble"? I don't see how your conclusion follows from<br>
>>> your premise. Why do "almost all proposals fail Participation"? It<br>
>>> isn't because there is anything inherently wrong with "that kind of<br>
>>> thinking". It is because it just happens that Participation is very<br>
>>> expensive (in terms of other desirable criterion compliances, such as<br>
>>> Condorcet). But in that way Mono-add-Plump is very very cheap (if not<br>
>>> free), and some of us are currently "in trouble" due to disregarding<br>
>>> "this kind of thinking".<br>
>> What I'm saying is that if we pursue criteria in the vein of <br>
>> Participation (or<br>
>> monotonicity), we cut down the list of methods we can consider, and we <br>
>> aren't<br>
>> necessarily getting anything of value except that fewer people can <br>
>> call the method<br>
>> absurd. What I call inherently of value would be things like sincere <br>
>> Condorcet<br>
>> efficiency or reduced strategic incentives.<br>
>><br>
>>> Suppose a mini-bus with a driver is contracted to pick up a group of<br>
>>> people and take then on a trip to one of X, Y or Z after polling the<br>
>>> passengers on their ranking-preferences among these alternative<br>
>>> destinations. After the bus is nearly full it is mistakenly assumed that<br>
>>> there will be no more passengers and the driver applies some algorithm<br>
>>> to the rankings of those present and announces that winning alternative<br>
>>> is X.<br>
>>> Then it is learned that there are two more passengers to come to fill up<br>
>>> the bus. They do so and the driver says to them "I've polled all the<br>
>>> other passengers and at the moment the winning destination is X. Where<br>
>>> would you like to go?" and they reply "X is our first preference and Y<br>
>>> wouldn't be too bad and we are very glad we aren't gong to Z".<br>
>>> The driver replies "You prefer Y to Z? In that case the new winning<br>
>>> alternative is Y". Now if these two voters (and perhaps others whose<br>
>>> first preference was X) were enlightened election-method experts, they<br>
>>> might think "Obviously this fellow's election-method algorithm fails<br>
>>> Participation (and presumably Later-no-Harm). Perhaps it meets<br>
>>> Condorcet, which we know is incompatible with both Participation and<br>
>>> Later-no-Harm. Perhaps before we showed up there was a top cycle and our<br>
>>> Y>Z preferences turned Y into the Condorcet winner.<br>
>>> But we know that Condorcet is also incompatible with Later-no-Help so us<br>
>>> revealing our second preferences could have just as likely helped us, so<br>
>>> I suppose we were just unlucky."<br>
>>> Or if they were not experts but charitably minded they might think "I<br>
>>> suppose it is possible that this fellow made an honest mistake due to<br>
>>> him being thick and us confusing him with too much information".<br>
>>> Now replay this scenario except this time the new passengers just say<br>
>>> "Great! We just really want to go to X and we don't know or care about<br>
>>> any other destination." And then the driver says "In that case the<br>
>>> winning alternative changes from X to Y".<br>
>>> The response could only be that the destination-decider (supposedly<br>
>>> purely based on the passengers' stated preferences) is insane (or<br>
>>> malevolent, in any case illegitimate) and that Y is obviously an<br>
>>> illegitimate winner.<br>
>>> Did you notice a very different vibe from the first case, which was a<br>
>>> failure of Participation and Mono-add-Top but not Mono-add-Plump?<br>
>> The difference in vibe is quite similar to your own difference in vibe <br>
>> when you<br>
>> compare these situations.<br>
>><br>
>> In one case, some voters are willing to say "I guess there were other <br>
>> considerations<br>
>> in play; we were unlucky" but in the other case they won't go there. <br>
>> And that's<br>
>> fine, that is their right.<br>
>><br>
>>> In December 2008 on EM I argued that Schulze's Generalised Majority<br>
>>> Criterion is a mistaken standard because the concept is vulnerable to<br>
>>> Mono-add-Plump.<br>
>> But given their compatibility, isn't that a strange thing to say?<br>
>><br>
>>> Your new MDDA 2 method fails the example I gave:<br>
>>> 25 A>B<br>
>>> 26 B>C<br>
>>> 23 C>A<br>
>>> 04 C<br>
>>> (78 ballots, majority threshold = 40)<br>
>>> Implicit approval scores: C 53, B 51, A 48. No candidate is<br>
>>> disqualified due to sub-majority approval.<br>
>>> B>C 51-27, C>A 53-25, A>B 48-26. All candidates have a<br>
>>> "majority" strength defeat, so it "isn't possible" to disqualify any<br>
>>> candidate on that basis. So, according to the rules of MDDA 2, we elect<br>
>>> the most approved candidate, C.<br>
>>> Now say we add 22 ballots that plump for C to give:<br>
>>> 25 A>B<br>
>>> 26 B>C<br>
>>> 23 C>A<br>
>>> 26 C<br>
>>> (100 ballots, majority threshold = 51)<br>
>>> Implicit approval scores: C 75, B 51, A 48. Now A has sub-majority<br>
>>> approval and so is disqualified.<br>
>>> B>C 51-49, C>A 75-25, A>B 48-26. Now C and A have<br>
>>> majority-strength defeats and B doesn't, so (according to the rules of<br>
>>> MDDA 2), A (again) and C are disqualified leaving B as the new winner.<br>
>>> The contention that C is the right winner when there were just 78<br>
>>> ballots but when we add 22 ballots that plump (bullet vote) for C the<br>
>>> right winner is no longer C is .... completely crazy.<br>
>> The voters' behavior had a side effect of strengthening B. All sorts <br>
>> of monotonicity<br>
>> failures take such an appearance.<br>
>><br>
>> And again, there could be differences in severity, e.g. what percent <br>
>> of voters think<br>
>> a given phenomenon is absurd. But I don't find that very interesting <br>
>> because it<br>
>> doesn't tell us about the merits of the method. It's basically <br>
>> marketability.<br>
>><br>
>>>> Well, in an environment where the concept of "median voter" is likely<br>
>>>> to be meaningful,...<br>
>>> What "environment" is that? And why is that the environment the one we<br>
>>> should primarily focus on?<br>
>> One where voter and candidate preferences can be explained by an <br>
>> underlying issue<br>
>> space. In this case if you could project everyone onto a plane or <br>
>> spectrum it would<br>
>> be a bit easy to find the median voter and their preferred candidate.<br>
>><br>
>> I think this usually describes public elections, but it probably <br>
>> wouldn't cover a<br>
>> vote on what color is the best, or a vote on what cuisine to have <br>
>> delivered. So I<br>
>> think we should probably have IIB for those cases.<br>
>><br>
>>> I think that is the sort of thinking that<br>
>>> leads some people to support Median Ratings methods, which we know are<br>
>>> garbage because they fail Dominant Candidate and Irrelevant Ballots<br>
>>> Independence, and the voters have a strong incentive to just submit<br>
>>> approval ballots (giving the same result as Approval). And it has led<br>
>>> you to the absurdity of suggesting a method that fails Mono-add-Plump.<br>
>> Not at all, median rating methods aren't motivated by the notion of a <br>
>> single median<br>
>> voter. There are multiple median voters on different posed questions, <br>
>> and that's<br>
>> true on a pairwise matrix as well.<br>
>><br>
>>> I think for the purposes of properly analysing single-winner election<br>
>>> methods and inspiring the invention of new ones, we can and should do<br>
>>> without criteria that refer to irrelevant ballots dependent "majority"<br>
>>> thresholds or pairwise defeats. Those have almost no positive point<br>
>>> aside from marketing.<br>
>> You're saying that criteria directly specifying "majority" and not <br>
>> something else<br>
>> is what lacks positive points aside from marketing? That could be true.<br>
>><br>
>>> My suggestion for something as close as possible to Minimal Defense:<br>
>>> *If the number of ballots that vote X above bottom and Y no higher than<br>
>>> equal-bottom is greater than Y's maximum pairwise support, then Y can't<br>
>>> win.*<br>
>> I don't hate that. I don't know what you gain from using "max pairwise <br>
>> support"<br>
>> instead of "votes in total."<br>
>><br>
>>> I propose Double Defeat (Implicit) as something that can substitute for<br>
>>> the votes-only versions of Minimal Defense and SFC and also Plurality.<br>
>>> *Interpreting ranking (or ranking above equal bottom) as approval, no<br>
>>> candidate that is pairwise-beaten by a more approved candidate is<br>
>>> allowed to win.*<br>
>> It's interesting but it doesn't cover SFC. In an SFC failure scenario the<br>
>> disqualified candidate might very well have more approval than the <br>
>> candidate who<br>
>> disqualifies them. The concern is that supporters of the latter gave <br>
>> the election<br>
>> away.<br>
>><br>
>>> That already inspires a simple method suggestion: DDI,MMM: *Elect the<br>
>>> candidate not disqualified by Double-Defeat (Implicit) that is highest<br>
>>> ordered by MinMax(Margins).*<br>
>>> What do you think of that?<br>
>> I don't like it but it might be fine.<br>
>><br>
>>> And what is wrong with your "Improved<br>
>>> Condorcet Approval" method ? I think it would be good using<br>
>>> unrestricted ranking ballots with an explicit approval cutoff.<br>
>> ICA or C//A (implicit) are not bad. They don't satisfy SFC. In my <br>
>> recent simulations<br>
>> on frontrunner truncation strategy, C//A is among the best Condorcet <br>
>> methods. In<br>
>> random elections I am disturbed that ICA and C//A are worse than WV <br>
>> methods at<br>
>> strong FBC (i.e. what I call compromise incentive).<br>
>><br>
>> You've asked me many times about C//A(explicit) and I still think it's <br>
>> bad. The<br>
>> entire notion of C//A(implicit) being good at deterring burial is <br>
>> based on the<br>
>> fact that if you use burial to prevent there from being a Condorcet <br>
>> winner, then in<br>
>> the "cycle resolution" you cannot prefer any candidate to the one you <br>
>> raised<br>
>> insincerely.<br>
>><br>
>> In C//A(explicit), burial only backfires if it actually creates a fake <br>
>> CW. Creating<br>
>> a fake cycle is never bad for your favorite.<br>
>><br>
>> Kevin<br>
>> <a href="http://votingmethods.net" rel="noreferrer noreferrer" target="_blank">votingmethods.net</a><br>
>><br>
> ----<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>
----<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>