[EM] Characterization of some voting-systems & their results

[Closed Limelike Curves]

>
> Which is fine by me if that’s truly who the voters want, but it seems like
> a stretch to me to say that approval solves the lesser-evil problem.
>
The reason it solves the lesser-evil problem is because it satisfies the
honest favorite (sincere favorite, no favorite betrayal) criterion.
Essentially, what this means is that there is no reason not to place your
favorite candidate first on the ballot. That means you don't have to
compromise your values to have an effective vote.

I suppose you're right that if most people continue to approve of a lesser
evil, that evil would still win. But then we have to ask, "evil according
to who?" Approval solves the lesser evil problem, *according to the voters
themselves*: the voters never have an incentive to pretend a lesser evil is
their favorite candidate.

It also means that if most voters agree that a candidate Bob is non-evil,
but all of his opponents are different degrees of evil, Bob is guaranteed
to win: you just have to approve of him. This doesn't conflict with
approving of some lesser-evils as well. At that point, voters will see that
Bob is picking up steam and can win, and will drop their approval of
non-Bob candidates as soon as he looks like a frontrunner.

Unfortunately, approval voting can't solve the issue of bad voters who
approve of evil (no voting system can). But what it *can* do is guarantee
that if voters are unhappy with their options, they can choose another one
that's better, without being afraid of wasting their vote as a result.

See also Myerson and Weber's 1990 paper, showing that score/approval voting
always have a Condorcet (most-loved, majority-rule) winner as a stable Nash
equilibrium. (This is unlike plurality, IRV, or other systems that fails
sincere favorite: for these systems, a popular candidate can end up being
crushed, just because everyone thinks they have no chance of winning. This
actually happened just 2 years after the paper was published, with polls
showing Ross Perot could have beat either Bush or Gore one-on-one, but he
lost because everyone thought he had no hope of winning. Ironically, that
means Bush was a spoiler for Perot, not the other way around.)

On Sun, Mar 24, 2024 at 1:12 AM Michael Garman <
michael.garman at rankthevote.us> wrote:

>
> Perhaps…but this assumes that everyone values the same candidates the
> same. The polls that show that near-majorities of Americans want a third
> party are often taken out of context by people who don’t take into account
> that everyone wants a different third party — I want a socialist party,
> some people want a centrist party, others want a libertarian party, still
> others want a full-fledged Nazi party, etc etc. If most people approve one
> third party candidate and one “lesser evil,” we get lesser evil again.
>
> Which is fine by me if that’s truly who the voters want, but it seems like
> a stretch to me to say that approval solves the lesser-evil problem.
> On Sun, Mar 24, 2024 at 2:42 AM Michael Ossipoff <email9648742 at gmail.com>
> wrote:
>
>> Sorry, one more comment, in answer to a frequent criticism:
>>
>> In Approval:
>>
>> If some or many people approve an unfavorite lesser-evil “compromise”
>> (many are going to make the mistake of voting for one in November), then
>> pretty much everyone who approves him also approves better candidates too.
>>
>> But those better candidates will also be approved by people, including
>> me, who would never approve that evil.
>>
>> …& so a better candidate wins & the evil loses.
>>
>>
>> On Sat, Mar 23, 2024 at 18:30 Michael Ossipoff <email9648742 at gmail.com>
>> wrote:
>>
>>> Addendum:
>>>
>>> In the Approval-result, I should have added that there’s no
>>> sincere-vs-strategic issue.
>>>
>>> On Sat, Mar 23, 2024 at 18:27 Michael Ossipoff <email9648742 at gmail.com>
>>> wrote:
>>>
>>>> Approval maximizes the number of people pleasantly-surprised, &/or the
>>>> number who get something they like.
>>>>
>>>> Condorcet:
>>>>
>>>> “This complicated automatic-machine will legalistically tell us who has
>>>> the right to get their way.”
>>>>
>>>> Score:
>>>>
>>>> Robert is right, that people can’t be expected to rate sincerely
>>>> (merit-proportionate) in Score when something is at-stake.
>>>>
>>>> But in polls where nothing material is at-stake, & when people would
>>>> rather find out how liked their favorite is—instead of falsifying to make
>>>> people believe that their favorite is most-liked—then people will rate
>>>> merit-proportionate, & score will meaningfully measure overall merit.
>>>>
>>> ----
>> Election-Methods mailing list - see https://electorama.com/em for list
>> info
>>
> ----
> Election-Methods mailing list - see https://electorama.com/em for list
> info
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20240324/cc943a7e/attachment.htm>