[EM] Best Single-Winner Method

[Toby Pereira]

The best single-winner method probably depends on the particular election and also the limitations of how it can be run. If, for example, an election for some position could be run over the internet, then what I would regard to be the best method would not be the same as for an election for public office where people turn up to vote in person on a particular day. For such an internet election, I like the following:
Score voting, but where the current total/average is available to see and continuously updated, and where people are able to change their vote as many times as they like. Obviously you could end up with a lot of people withholding their vote until right near the end or completely changing it to prevent people from optimising their tactical vote, so I would suggest a couple of things to limit the effect of this. The deadline to set your initial vote could be earlier than the deadline to change your vote, and the end time for changes could be non-deterministic.
For example, voters might have nine days to cast their initial vote. Then there would be one guaranteed extra day where votes can be freely changed (they can also be changed within the initial nine days). After that the election would end randomly with a half life of two hours. So if you were organising such an election there would be three variables to set.
This avoids the problem of score voting that people wouldn't know how to rate their second favourite candidate etc. because they can see how the candidates are doing. And it would have the advantages of score voting that the results are clear and easy to understand (as it's just a total or average) and that it isn't subject to weird unexpected paradoxes that most voting methods we talk about have (e.g. Condorcet methods can have cycles and fail participation). Strength of preference would matter but with voters more aware of what to do so that later-no-harm failures wouldn't be at so much risk of happening. In the following example involving honest score ballots (scores out of 10):
49 voters: A=10, B=0, C=149 voters: A=0, B=10, C=12 voters: A=9, B=8, C=10
C would be the Condorcet winner but is really a non-entity. Under normal score voting (where voters cannot change votes), C would have no chance of winning. Under this method, it's very unlikely that there would be anywhere near enough tactical voting to get C anywhere near close enough to winning, so it keeps the advantages of score voting and enhances it in terms of giving the voters knowledge of how to vote tactically, but without throwing the baby out with the bathwater and going full pairwise.
Toby

      From: steve bosworth <stevebosworth at hotmail.com>
 To: EM list <election-methods at electorama.com> 
 Sent: Monday, 20 May 2019, 20:46
 Subject: Re: [EM] Best Single-Winner Method
   
 #yiv0883059675 P {margin-top:0;margin-bottom:0;}Re: Best Single- Winner Method
Sennet Williams, Forest Simmons, Robert Bristow-Johnson, Abd dul Raman Lomax, and Chris Benham have recently addressed each others’ claims about IRV, 3-slot Methods, IBIFA, and Asset.  This discussion prompts me to request some help later after I have clarified several issues.Firstly, please correct me if I am mistaken but currently I am assuming that we all would ideally want the Best Single-Winner Method:   
   - To be simple enough so voters can both use it and understand how it is counted;
   - To minimize the wasting of citizens’ votes (see below), and
   - To guarantee that the winner among 3 or more candidates is the candidate most supported by at least 50% plus one (an absolute majority) of all the citizens voting, and
   - To offer as few incentives and possibilities for voting tactical.

   
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20190529/8e2b3c9b/attachment.html>