[EM] Regarding RCV, in poll & in November

[Michael Ossipoff]

Though I said that “ I don’t give a shit what the losers want”, & that the
Mutual-Majority, rightfully, should be the ones to choose which of their
candidates should be the ones who choose which if their candidate is
elected…

… & though I said that Condorcet is unnecessarily considerate [to those
outside of the Mutual-Majority…

Nonetheless I’ve wanted that unnecessary considerateness. In fact I’d still
like it, to displease as few as possible.

…if Condorcet didn’t have its problem of an overly
computationally-intensive count.

That’s why Condorcet ( in its best versions)  has been my ranked-method
proposal.

But Condorcet *does* have that problem, & so yes not having that
unnecessary considerateness for the losers is perfectly okay.


On Sat, Jul 20, 2024 at 15:34 Michael Ossipoff <email9648742 at gmail.com>
wrote:

> When I reported preliminary poll-results I didn’t report about RCV or STV,
> because there were small-electorate ties that prevent a typical count, &
> require tiebreaking rules (which would have to have been stated before the
> voting).
>
> But I applied two plausible nonrandom tiebreakers, &, by both of them, RCV
> elected Biden.
>
> Note that, while RCV elected Biden, Approval returned an unresolvable tie
> between Biden & the CW, Marianne Williamson.
>
> Approval thereby came closer to electing the CW, & more likely would have,
> with a larger, but similarly-constituted, electorate.
>
> Approval doesn’t share rank-methods’ drastic compromise-strategy of voting
> someone the voter like less, over someone s/he likes more.
>
> Therefore there’s certainly a strong case for not bothering with the
> implementation-expense of rank-balloting, & instead just using Approval.
> …which doesn’t have rank-methods’ worst-case strategy.
>
> Yes, defensive order-reversal needn’t happen in rank-methods.
>
> In the best Condorcet methods, no one needs defensive order-reversal (or,
> really, due to autodeterence, *any* defensive strategy).
>
> But those methods are disqualified from public political elections, by
> their enormous computational need, with consequent correspondingly-high
> count-fraud vulnerability.
>
> Yes there are some Condorcet-Criterion complying methods that don’t need
> the exhaustive pairwise count, but they haven’t been shown to have wv
> Condorcet’s excellent strategy-properties.
>
> But RP(wv), MinMax(wv) or Smith//MinMax(wv) would be good for
> single-winner choices in Parliament.
>
> For that purpose, might as well use the very best: RP(wv).
>
> ..& I like them for some polls, like the current one here, & Christina
> Tobin’s Free-&-Equal polls.
>
> Not all polls. Some polls are better with RCV. Some are better with
> Approval.
>
> What about RCV, in public political elections?
>
> For members of the Mutual-Majority, there’s absolutely no need for any
> strategy. Just rank sincerely, & RCV is guaranteed to elect a candidate of
> the Mutual-Majority.
>
> …the one with the best broad-support among the Mutual-Majority.
>
> Then why do the Republicans dislike RCV? Well, they’ve been finding that
> they’re not in the Mutual-Majority.
>
> Which way did the transfers, & the win, go in Burlington & Alaska? The
> Republicans can’t have not noticed that.
>
> There’s been a lot of criticism of RCV, because the situation isn’t as
> good for those *not* in the Mutual-Majority.
>
> But:
>
> 1. Those outside the Mutual-Majority are still no worse off than they
> were, with Plurality. …& aren’t the nonprogressives the ones who think that
> Plurality is good enough?
>
> 2. Who says that a good method has to elect someone majority-unpreferred?
>
> Yes, when the Condorcet-complying methods elect the CW, they’re electing
> the Mutual-Majority candidate most liked by those outside the
> Mutual--Majority. That’s mighty considerate!  It’s unnecessarily
> considerate.
>
> Who ever said that people not in the mutual-majority should gays say in
> which Mutual-Majority candidate is elected?
>
> That’s the business of Mutual-Majority. *They* should choose which if
> their candidates to elect.
>
> I don’t give a shit what the losers want.
>
> RCV always elects the candidate of the Mutual-Majority who has the best
> broad-support among the Mutual-Majority.
>
> …& guess what: Most of the people wan what’s best for most of the people.
>
> i.e. The Progressives are the Mutual-Majority.
>
> They can’t lose in RCV.
>
> Then what about in my poll?
>
> A public election, in a jurisdiction where the Progressives have enacted
> RCV is different.
>
> They certainly didn’t enact RCV so that they could bury someone they like
> more, under someone they like less. …& so they won’t.
>
> Have I been critical of RCV? No!
>
> I’ve merely been critical of what I now (more politely) call the
> FairVote-Fib (FF).
>
> The FF has given a good method a bad-name, a bad reputation.  …& has
> turned a lot of people against a good method.
>
> The FF is Indy’s fib to Republicans (…&, I claim, to all Republocratic
> candidates …& their minority of preferrers.
>
> But, to Progressives, it isn’t a fib at all.
>
> A sincere-ranking Mutual-Majority can’t lose in RCV.
>
> Yes I like Approval, & it has advantages.
>
> Approvals remains my main proposal to jurisdictions where there’s no
> already-existent RCV proposal or even big public interest in one. …places
> where there’s nothing happening yet with singl-winner reform.
>
> But, where there’s an RCV ballot-measure, & RCV is what’s on the ballot…or
> anywhere where RCV is already what people want & are talking about , then I
> support the RCV proposal.
>
> In the 35 years of the current RCV project, how many states use it
> statewide: 2
>
> Hopefully that will soon be 4.
>
> But even then, &  if RCV keeps gaining 2 seats every 2 years, it will take
> 46 years to succeed nationwide.
>
> Too slow.
>
> Approval with is minimalness, it’s minimal change from the status-quo, &
> it’s zero implementation-expense, Approval could take off a LOT faster.
>
> Plurality lets you give a point to one candidate…& requires you to give
> zero points to each of the others.
>
> It lets you rate one candidate “Yes”,  & requires you to rate all the
> others “No”.
>
> What, it isn’t intended as points or ratings?
>
> Irrelevant—That’s the count result.
>
> Plurality forces you to ( usually insincerely, unwillingly &
> complainingly) make some compromise that you despise beat your favorite.
>
> How Democratic is that coercion?
>
> With 3 or more candidates, democracy requires letting you express, &
> have-counted, relative preference or merit among & across those candidates.
>
> If it forbids that expression, it isn’t democracy.
>
> Approval consists of no change other than dropping the current
> antidemocratic forbidding of expression of merit & preference among &
> across the candidates…dropping the artificial antidemocratic requirement to
> rate all but one at zero, thereby forcing falsification of preferences.
>
>
>
>
>
>
>
>
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20240720/32784f1d/attachment.htm>