[EM] Did anyone pay attention to what he said about Approval vs Condorcet?

[Michael Ossipoff]

Replying below:

On Wed, May 29, 2024 at 21:16 robert bristow-johnson <
rbj at audioimagination.com> wrote:

> > On 05/29/2024 11:00 PM EDT Michael Ossipoff <email9648742 at gmail.com>
> wrote:
> >
> >
> > Just one more astronomical reference:
> >
> > Next time you look at the Milky-Way, or a long-exposure photo of it or
> any galaxy:
> >
> > Realize that, for every one of its estimated half-trillion stars, there
> are 12 instances of having to find a candidate-name in a ranking of 135
> candidate-names, when doing a handcount of Condorcet with 135 candidates
> (the number that a recent California governor-election had), & 300 million
> voters.
> >
>
> > BTW, when I posted my reply to one of KM’s last posts before he
> temmporarily left this list, I worded it as a reply to him, even though I’d
> removed his name from the post’s “To:” field, as he’d requested.
> >
> > In that reply, I often said that I didn’t know what he meant. In
> philosophy it isn’t unusual for what one person says to be completely
> meaningless to someone else.
> >
> > That doesn’t mean that one of those people is wrong. It’s just that one
> person’s philosophical framework & topic isn’t the other person’s.
> >
> > KM’s philosophy of voting systems, & the philosophical framework in
> which he was discussing them can rightly, validly & reasonably influence
> KM’s preferences & choices among voting-systems.
> >
> > But it doesn’t apply to the objective matter-of-fact Approval-properties
> that I’d stated.
> >
>

Reply directly below this comment:


> I dunno, different countries and different states do different things.
> But where I am from, the state has actual ballot access requirements.  A
> certain number of signatures.  I think the legislature had someone do
> research on some probabilistic correlation of how many signatures required
> and how many candidates make it onto the ballot.  How much they settled on,
> I don't know but there are usually 4 or 5 candidates on the November ballot
> for Governor of Vermont.
>
> Now for IRV, 5 candidates is 205 tallies.  For Condorcet-Plurality or
> Condorcet-TTR, it's 25 tallies.
>

Not clear what you mean by a “tally”, but a Condorcet handcount is
humungously longer than an IRV count. But that’s irrelevant, because my
comparison was between Approval & Condorcet.

A Condorcet exhaustive pairwise-count with 5 candidates needs counting of
20 pairwise vote-totals. (2 for each pair of candidates).

More reply below:

But you heard of "The Paradox of Choice"?  We don't want just two
> candidates, we want more choices than that.  But we don't want 135
> candidates either on even an electronic ballot.  (And I will never give up
> on paper ballots and optical-scan technology, so I want all these races to
> fit on a single ballot, front and back.)


There were 135 candidates in the California election that Chris referred to.


>
> We just don't want more than the 4 or 5 most credible candidates.  The way
> you get sufficient credibility to be on the ballot is to get lots of valid
> signatures on a petition.  The number of signatures needed is a function of
> the population of the district of the elected office.


…greatly favoring corporate-funded candidates who can hire all the paid
signature-gatherers they want.  … resulting in economic ballot-exclusion of
honest parties.

Maybe you like that, but not all of us want it.

With N candidates, the number of pairwise-votes that must be counted in
Condorcet’s exhaustive pairwise count is 2(N-1) times the number of
approvals that must be counted in Approval.

With 5 candidates, that ratio is 8.

If that doesn’t sound like much, consider the number of voters in a
precinct.

Adequate count-security auditing is 8 times more infeasible with Condorcet
than with Approval.

5 candidates? What would that be this time?

Jull Stein, a Libertarian, RFK Jr, Biden & Trump?

One progressive? No Cornell West, or any other progressive?

…for which you gain “only” 8 times more infeasibility for an adequate
count-security audit.

>
>
> So hyped-up numbers for a wild-assed hypothetical doesn't persuade.


135 candidates isn’t “hyped-up” or “wild-assed hypothetical”. That’s how
many candidates there were in the California governor-election that Cris
mentioned.

Sure, an Approval-election could reduce the field to 5.  …if you really
want a handcount-audit to be 9 times more infeasible than with Approval.

>
>
> I'm missing KM already.
>

Your fawning a**-kissing adulation is getting tiresome.

>
> --
>
> r b-j . _ . _ . _ . _ rbj at audioimagination.com
>
> "Imagination is more important than knowledge."
>
> .
> .
> .
> ----
> 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/20240531/485b8d69/attachment-0001.htm>