[EM] Exactly what is the non-monotonicity demonstrated in the Alaska 2022 special election.

[Closed Limelike Curves]

When I say Begich got too many votes, I'm referring to the participation
failure: if you removed a bunch of Palin > Begich > Peltola votes, which
are votes for Begich over Peltola, the winner switches from Peltola to
Begich. The nonmonotonicity is that if Peltola had won more votes from
Palin supporters, she would have lost.

It seems like some of the complaints are about length, so how about
restoring it mostly to where it was after my most recent edit, and then
adding a bit more about center-squeeze to the body? Or maybe it can go into
Draft: Center Squeeze so people can read it from the link?

A major issue is the editor is trying to add in the word "some" to make it
sound like this was a vocal minority, rather than the general consensus
among people who have studied social choice theory; they're also removing
the intuitive description of a participation failure as a situation where a
candidate can lose as a result of winning too many votes.

On Mon, Jul 8, 2024 at 4:23 PM robert bristow-johnson <
rbj at audioimagination.com> wrote:

>
> So Limey, I have edited the
> https://en.wikipedia.org/wiki/2022_Alaska%27s_at-large_congressional_district_special_election
> page.  I am trying to synthesize the most concise but complete description
> of the pathologies demonstrated in that specific election with the specific
> vote tallies that we have.  (There are notable differences between the
> Graham-Squire & McCune results and those that I got from Robbie Robinette,
> but I am not worrying about that for now.)
>
> Now, I understand exactly and quantitatively the Center Squeeze and
> exactly how Palin has the role of spoiler.  That's not a problem.  But I
> cannot see non-monotonicity demonstrated at all in this election.  I have
> trouble with this specific sentence: "The election was also notable as a
> negative vote weight event, a pathology where a candidate (Begich) is
> eliminated as a result of winning too many votes."  I do not see that
> Begich was eliminated as a result of winning too many votes.
>
> I understand how a hypothetical non-monotonicity was demonstrated in the
> Burlington 2009 IRV election.  If Bob Kiss campaigned hard in Kurt Wright
> territory and somehow won over 741 votes that otherwise had gone to Wright,
> that would have caused Kiss to lose, because Wright would have been
> eliminated before the final round and Kiss would lose to Montroll.  But
> that's a big "what if" hypothetical.  That actually did not happen, and
> that is the reason I left out non-monotonicity out of my paper.  (Thwarted
> majority, spoiler effect, negative reward for voting for your favorite
> candidate.)
>
> I had thought that IIA and Favorite Betrayal were not the same as
> non-monotonicity.  Non-monotonicity means, strictly, if more people vote
> for (1st choice) a specific candidate, that causes that candidate, who
> would have otherwise won, to lose.
>
> Can you spell out the numbers?  Or someone else?  You can use these
> numbers:
> https://drive.google.com/file/d/1y32bPVmq6vb6SwnMn6vwQxzoJfvrv6ID/view or
> those from the Graham-Squire/McCune report or whatever good numbers from
> the cast vote records you get.
>
> I just don't see non-monotonicity demonstrated in the August 2022 Alaska
> election.
>
> Also, we need to team up to preserve these edits at Wikipedia.  I'm gonna
> hit 3RR real soon.
>
> --
>
> 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/20240708/d1b013c2/attachment.htm>