Request for help: List of Pathological Elections
The past few months I've been knuckling down and rewriting large swathes of
election-methods Wikipedia, which has been taking up a lot of my time. This
includes redoing the articles on Arrow's theorem, Spoiler effects, and
Favorite betrayal almost completely from scratch.
That said, there's only so much I can do. So, I'd like to request help on a
new article I think is important, which I'm drafting here:
https://en.wikipedia.org/wiki/Draft:List_of_pathological_elections
Useful resources are Warren D. Smith's list of pathological elections and
the Wikipedia article on vote-splitting.
It is my opinion that "Plurality" and "First-Past-The-Post" are synonymous.
But looking at your Wiki page, I am concerned in how you would ever know for certain an election is spoiled with FPTP. I mean, like Florida 2000, perhaps the Ralph Nader voters would have split even up between George W Bush and Al Gore. Not likely, but we don't know without the ranked ballot.
But for the RCV elections in which the Condorcet Winner is not elected (whether the CW exists or not), we know from the ranked ballot data that the election was spoiled and we know exactly how.
But we can't know it with FPTP without speculating how voters for the ostensible spoiler would have voted had their candidate not run.
BTW don't let FairVote mislead you about "redefining" the spoiler effect. A "spoiler" is a candidate who loses an election but, simply because of they were a candidate in the election, they materially change the winner of the election. Burlington 2009 and Alaska August 2022 and Minneapolis Ward 2 in 2021 and Oakland School Bard District 4 in 2022 were all objectively spoiled elections because the Condorcet Winner was not elected. But in the latter two, there was no CW and, in the cycle, the spoiler is always the candidate that winner would defeat in a head-to-head runoff.
--
r b-j . _ . _ . _ . _ rbj@audioimagination.com
"Imagination is more important than knowledge."
.
.
.
On 05/06/2024 3:30 PM EDT robert bristow-johnson rbj@audioimagination.com wrote:
It is my opinion that "Plurality" and "First-Past-The-Post" are synonymous.
But looking at your Wiki page, I am concerned in how you would ever know for certain an election is spoiled with FPTP. I mean, like Florida 2000, perhaps the Ralph Nader voters would have split even up between George W Bush and Al Gore. Not likely, but we don't know without the ranked ballot.
But for the RCV elections in which the Condorcet Winner is not elected (whether the CW exists or not), we know from the ranked ballot data that the election was spoiled and we know exactly how.
But we can't know it with FPTP without speculating how voters for the ostensible spoiler would have voted had their candidate not run.
BTW don't let FairVote mislead you about "redefining" the spoiler effect. A "spoiler" is a candidate who loses an election but, simply because of they were a candidate in the election, they materially change the winner of the election. Burlington 2009 and Alaska August 2022 and Minneapolis Ward 2 in 2021 and Oakland School Bard District 4 in 2022 were all objectively spoiled elections because the Condorcet Winner was not elected. But in the latter two, there was no CW and, in the cycle, the spoiler is always the candidate that winner would defeat in a head-to-head runoff.
Forgot to reference the FairVote bullshit: https://fairvote.org/defining-the-spoiler-effect/
Don't go for this shit.
--
r b-j . _ . _ . _ . _ rbj@audioimagination.com
"Imagination is more important than knowledge."
.
.
.
I’m not so sure that you should be rewriting & redoing all the those
articles.
You’re likely just ruining them, & someone else will later find the damage
& have to restore the articles. One fortunate feature of Wikipedia is that
damaged to articles can be “reverted” without having to rewrite the
articles.
On Mon, May 6, 2024 at 11:47 Closed Limelike Curves <
closed.limelike.curves@gmail.com> wrote:
The past few months I've been knuckling down and rewriting large swathes
of election-methods Wikipedia, which has been taking up a lot of my time.
This includes redoing the articles on Arrow's theorem, Spoiler effects, and
Favorite betrayal almost completely from scratch.
That said, there's only so much I can do. So, I'd like to request help on
a new article I think is important, which I'm drafting here:
https://en.wikipedia.org/wiki/Draft:List_of_pathological_elections
Useful resources are Warren D. Smith's list of pathological elections and
the Wikipedia article on vote-splitting.
Election-Methods mailing list - see https://electorama.com/em for list
info
Hi rb-j, I agree with you on that point. A spoiler effect is an IIA failure.
I addressed some of your concerns here, including the ones about FPTP. For
plurality elections, standard Wiki rules apply; you need a reliable source
saying "Ralph Nader spoiled the election", backed up with evidence like
polling, rather than literal mathematical proof in the form of the exact
ballots.
https://en.wikipedia.org/wiki/Draft_talk:List_of_pathological_elections#What_exactly_is_a_pathology
?
On Mon, May 6, 2024 at 12:32 PM robert bristow-johnson <
rbj@audioimagination.com> wrote:
On 05/06/2024 3:30 PM EDT robert bristow-johnson <
rbj@audioimagination.com> wrote:
It is my opinion that "Plurality" and "First-Past-The-Post" are
synonymous.
But looking at your Wiki page, I am concerned in how you would ever know
for certain an election is spoiled with FPTP. I mean, like Florida 2000,
perhaps the Ralph Nader voters would have split even up between George W
Bush and Al Gore. Not likely, but we don't know without the ranked ballot.
But for the RCV elections in which the Condorcet Winner is not elected
(whether the CW exists or not), we know from the ranked ballot data that
the election was spoiled and we know exactly how.
But we can't know it with FPTP without speculating how voters for the
ostensible spoiler would have voted had their candidate not run.
BTW don't let FairVote mislead you about "redefining" the spoiler
effect. A "spoiler" is a candidate who loses an election but, simply
because of they were a candidate in the election, they materially change
the winner of the election. Burlington 2009 and Alaska August 2022 and
Minneapolis Ward 2 in 2021 and Oakland School Bard District 4 in 2022 were
all objectively spoiled elections because the Condorcet Winner was not
elected. But in the latter two, there was no CW and, in the cycle, the
spoiler is always the candidate that winner would defeat in a head-to-head
runoff.
Forgot to reference the FairVote bullshit:
https://fairvote.org/defining-the-spoiler-effect/
Don't go for this shit.
--
r b-j . _ . _ . _ . _ rbj@audioimagination.com
"Imagination is more important than knowledge."
.
.
.
On 2024-05-07 01:19, Michael Ossipoff wrote:
I’m not so sure that you should be rewriting & redoing all the those
articles.
You’re likely just ruining them, & someone else will later find the
damage & have to restore the articles. One fortunate feature of
Wikipedia is that damaged to articles can be “reverted” without having
to rewrite the articles.
That can be more difficult than it seems. There's a very long path of
talk page actions, dispute resolution, etc. that someone can do to
oppose the reversion. Someone with superior endurance may well outlast
other editors who become tired of or worn down by the process and leave.
Even if they don't, it may take years before things come to a head like
in
https://en.wikipedia.org/wiki/Wikipedia:Requests_for_arbitration/Cold_fusion.
Wikipedia tries to be a level playing field and seek consensus. But it's
far from perfect.
Apart from that, though, I agree: WP:BOLD should definitely be tempered
with caution.
-km
On 2024-05-06 21:32, robert bristow-johnson wrote:
Forgot to reference the FairVote bullshit: https://fairvote.org/defining-the-spoiler-effect/
Oh, I see! Spoiler effect equals mutual majority failure. Let me try,
with tongue planted firmly in cheek...
"A spoiler effect happens when a majority, who could have got someone
they preferred to the current winner elected by coming together, is
instead splintered so that their preferred candidate does not win."
In other words: a Condorcet failure. That was fun, but arbitrary
redefinitions of a spoiler effect don't really go anywhere :-)
-km
On 2024-05-06 21:30, robert bristow-johnson wrote:
It is my opinion that "Plurality" and "First-Past-The-Post" are synonymous.
But looking at your Wiki page, I am concerned in how you would ever
know for certain an election is spoiled with FPTP. I mean, like Florida
2000, perhaps the Ralph Nader voters would have split even up between
George W Bush and Al Gore. Not likely, but we don't know without the
ranked ballot.
I suppose we could use polls to argue that the possibility exists, if
those polls ask for pairwise or ranked/rated preference information. It
would depend on the accuracy of the poll, e.g. if it asks both "Who
would you vote for in the election" and "out of X and Y, who would you
rather see win", and the former matches the actual outcome, then that
would suggest the latter is accurate too.
It won't be ironclad, of course. But without the ballots, nothing is :-)
-km