[EM] Sequential-Pairwise offensive & defensive strategy?

Thu Sep 21 14:03:13 PDT 2023

Hi everyone,

Oh Michael...where do I begin?  Your apparent move to the dark side makes
me sad.  I realize that this intro may sound condescending, but I truly
don't mean it that way.  I deeply respect your opinion. YOU were the one
who taught me about "center squeeze" in 1995 or so, and made me rethink
AV/PV/IRV/RCV (or whatever the name of the week is).  I just think you're
incorrect about FairVote.

More inline below.....

On Wed, Sep 20, 2023 at 10:32 PM Michael Ossipoff <email9648742 at gmail.com>
wrote:

> Yes, that fancier Sequential Pairwise ordering method would make it harder
> for a strategist to guess anything about the comparison-order…when, instead
> of the candidate’s top-count, it uses the sum of the top-counts of those
> above him on each ballot.
>
> An extra layer of unpredictability for a would-be strategist.
>
> The vulnerability of the simplest-defined procedure would have to be
> weighed against what polls & focus-groups say about people’s requirement
> for definition-simplicity & brevity.
>

RCV is already poorly understood.  When I moved to San Francisco in 2011, I
expected to grudgingly like voting in RCV elections, and I expected to
enjoy ranking my choices   What I found instead was that very few people
here understand how votes are counted, and many folks in my lefty political
tribe here take great pride in their ignorance of math and the inner
workings of their electoral system, trusting that the powers-that-be will
count things correctly.

As "exhibit A", I will point to the recent clown show in Alameda County
(i.e. just a few miles east of me, on the other side of a puddle known as
the "San Francisco Bay"):
https://www.sfchronicle.com/bayarea/article/Alameda-County-admits-tallying-error-in-17682520.php

It would seem that they had been counting RCV elections wrong for DECADES,
and only noticed the problem in 2022.  Simplicity and precinct summability
matters.

Undeniably the more multi-layered & unpredictable ordering procedure is
> more strategy-proof & better.
>
> Strategy-evaluation for Condorcet-complying pairwise-count methods has
> proven to be complicated & more difficult than one would expect.
>

This I will agree with. That is why I've hopped on the approval voting
bandwagon for single-winner reform.  I've been (more-or-less) aligned with
FairVote when it comes to multi-winner reform, since the many of the
problems with STV dissipate as the number of seats being selected for
rises.  For example, using STV to proportionally select members of a 5-seat
board helps ensure diverse representation.  I'm not aware of anyplace in
the SF Bay Area that is doing that, though.  What they do is divide the
land inside the city/county/whatever into districts, and then performs
single-winner elections in each district.  It's horrific.

But it now seems for sure that there are such methods that are sufficiently
> offensive-strategy-proof or well-protected from offensive-strategy.  It’s
> only a question of how many & which ones.
>
> The Condorcet-IRV-Runoff hybrids hold promise, with merit to always be
> balanced with what is heard in polls & focus groups about
> definition-brevity.
>
> So I’m sure that I’ll propose & recommend good Condorcet versions (even if
> I don’t yet know which ones & how many) over IRV.
>
> …but I’ll nonetheless include IRV among the methods that I offer, because
> it’s better than a lot of people believe.   …though its merit & workability
> strongly depend on the electorate & the candidate-lineup.
>
> I.e. Because it isn’t Condorcet-complying, it’s necessary that the
> electorate aren’t timid lesser-evil giveaway voters.
>

Back in 2018, I was heavily involved in lefty politics.  The mayor (Ed Lee)
dropped dead while grocery shopping at a Safeway that I had been to many
times.  He wasn't 500 years old; and he seemed in good health.  It was
truly a surprise to everyone,  Suddenly, we had nearly a dozen local
politicians with almost no name recognition competing to be the next mayor:
https://en.wikipedia.org/wiki/2018_San_Francisco_mayoral_special_election

The good news: the Condorcet winner (London Breed) won the election.  The
bad news: it was a close election, almost exclusively centered on identity
politics (though YIMBY vs NIMBY also played a factor).  There were three
well-funded candidates (Breed: who was "the establishment" candidate, born
and raised in the Fillmore, which USED TO BE an affordable neighborhood in
SF, and she would become our first Black woman serving as mayor), Mark Leno
(who would have become our first openly gay mayor), and Jane Kim (who
promised to be our first Korean-American mayor).

Having only lived here since 2011, I still considered myself new(-ish) to
the political scene.  I asked a friend (who I met in a lefty political org
here in SF) who I should vote for.  He said "None of them. They're all
corrupt.  Breed's funded by Conway."  I didn't know who George Conway was
at the time, but that didn't matter.  I was looking for a mainstream(-ish)
candidate to put SOMEWHERE in my ranking, so that I could evaluate the
niche candidates relative to my mainstream anchor.  I pressed him: "if
someone was holding a gun to your head, and you had to choose between
Breed, Kim, or Leno, who would you choose?".  He refused to answer, despite
how hard I pressed on the issue.

The polls leading up to the election were spotty and difficult to
decipher.  Most of the news coverage was about the Breed/Leno/Kim food
fight.  Breed had served as mayor for a day or two after Ed Lee died, but
there was a political food fight over her running for mayor after getting
to be the incumbent, so she was removed as mayor.  I think that helped her,
since (at least for me) it made her a bit more sympathetic candidate, and
given that the rules said she should be mayor (as previous President of the
Board of Supervisors).  Many SF voters get high and mighty about "rule of
law", and yet, the law was changed just as soon as a Black woman took
office.

My point: I'm guessing that my activist friend was one of the dreaded
"bullet voters" that FairVote misinforms people about.  I find FairVote a
flawed organization, and I specifically think that founder/leader Rob
Richie to be a deeply unethical (perhaps even Machiavellian) political
player.  Approval voting finds consensus candidates in a way that I think
the electorate can understand.  "Mark all of the candidates you approve of,
and the candidate with the highest approval rating wins"  I suspect that
approval voting is better at rooting out corruption than RCV, and I have
some more anecdotal evidence to back me up.

St. Louis switched their mayoral elections to approval top-two.  The
primary election selects the two candidates with the highest approval
rating, and the general election decides between the last two.   All eyes
in the electoral reform community were on St. Louis, and St. Louis also
elected their first Black woman to be their mayor:
https://en.wikipedia.org/wiki/2021_St._Louis_mayoral_election

If you've never been to St. Louis, I'll break it down for you.  The north
part of town is the flood plain at the confluence of the Missouri and
Mississippi Rivers.  It's where the poor people live, and many of them are
Black.  South of Interstate 44 (which replaced "Route 66") is where one
finds the rich people, where many houses were built for slaveowners (back
in the day).  The St. Louis area has long been the fulcrum of race
relations in this country.  For example, many folks know about the
predominantly Black suburb Ferguson in the flight path of STL, also known
as "Lombard".  Famous Nazi sympathizer and aviator Charles Lindbergh flew
the "Spirit of St. Louis" when he made his first solo transcontinental
flight in 1927.

As an outsider who has mainly been through St. Louis (many times), and who
has only stayed overnight in St. Louis a few nights, I only have a
superficial understanding of the place.  But I watched the Wikipedia
articles about the election with an eagle eye.  It would seem that Lewis
Cass (<https://en.wikipedia.org/wiki/Lewis_E._Reed>) was the sore loser,
and it appeared he was starting to lead an anti-approval campaign in St.
Louis.  However, Reed (and a couple of his cronies on the Board of
Aldermen) were caught on video accepting bribes from undercover FBI
agents.  They're in prison now.

I don't know much about Tishaura Jones, but all outside evidence suggests
she's a good-government type who is cleaning up the politics of St. Louis.
I'm hopeful she gets re-elected in 2025, and that FairVote doesn't try to
get approval voting replaced with RCV with the help of useful idiots that
will take money in exchange for supporting the "correct" policies.

But an electorate that has just enacted IRV in a referendum didn’t do so
> because they want to rank Lesser-Evil over their favorite. They enacted it
> because they want to rank sincerely, to express & fully help their favorite.
>
> Let’s support Oregon’s IRV (RCV) referendum next year!
>
> Though IRV doesn’t meet the Condorcet Criterion, it meets Mutual-Majority:
>
> IRV always elects the candidate of the largest faction of the
> Mutual-Majority.  …The favorite candidate of the Mutual-Majority.
>
> IRV didn’t “fail” in Burlington & Alaska. It did what it’s supposed to do.
>

RCV elected the extreme candidates, rather than the compromise.  Sometimes,
it elects the "good guy/gal" candidate (like Mary Peltola), but sometimes,
it elects an incompetent candidate (like Bob Kiss), and it seems quite
likely to me that Alaska will elect an extreme right-wing Republican
candidate when Peltola runs for reelection in 2024.  Approval (and probably
STAR) will elect candidates toward the center of public opinion, but RCV
gets random when elections get close:
https://electowiki.org/wiki/Yee_diagram

Under approval and STAR, those of us who think of ourselves as "left-wing"
are not as likely to get lucky every so often and get a minority candidate
like Mary Peltola (who sadly wasn't quite able to appeal to the median
voter in Alaska, it seems), but we're also not playing Russian roulette and
potentially getting a right-wing reactionary that could be funded by
foreign adversaries.  I've heard (from someone, I can't remember who) that
one can see Russia from Alaska.  :-)

> We don’t yet have a big Condorcet organization, or referenda, initiatives
> or strong lobbying for it, but let’s support the already ongoing enactment
> efforts for IRV, now named RCV.
>

Let's not.  FairVote isn't that big (only $4.5MM/year in annual revenue,
based on my cursory investigation).  The Center for Election Science pulls
in over $1MM/year, and is looking for new executive director (or "CEO" as
they say: https://electionscience.org/about/careers-and-board/ ), and I'm
optimistic that they may just hire someone who is more sympathetic to
Condorcet consistency than the prior executive director.  The Equal Vote
Coaltion is a small scrappy startup, but seems to be using their money
better than the prior two orgs (and seems far more amenable and adaptable
to Internet feedback than either of the larger orgs).  Having helped a
small-ish non-profit ($10MM/year in 2010 when I joined) become a larger
organization ($60MM/year to $70MM/year or so these days, I think), I've
learned not to get too enamored of (or intimidated by) "big non-profits".

If Rob Richie can prove that he's not unethical, and come to the table with
the rest of the electoral reform community, and debate openly and honestly,
then maybe I'd consider teaming up with them.  Perhaps the FairVote Board
can fire Richie, since he promised effective steps toward proportional
representation, and hasn't achieved that after three decades.  He's only
burned bridges and salted the earth for effective reform efforts through
backroom dealing (e.g. what FairVote did in Seattle in 2022).

As of right now, there's someone with the username "RRichie" that doesn't
disclose their clear conflict-of-interest on English Wikipedia, but often
makes very pro-RCV arguments (and pro-FairVote marketing) under that name:
https://en.wikipedia.org/wiki/Special:Contributions/RRichie

Rob

>
> On Wed, Sep 20, 2023 at 18:24 Forest Simmons <forest.simmons21 at gmail.com>
> wrote:
>
>> So called BTR-IRV, "Bottom Two Runoff IRV" goes along those lines.
>>
>> You probably remember "Benham" that runs IRV elimination until there
>> remains a candidate undefeated by any of the other remaining candidates.
>>
>> This reminds me of basing the Sequential Pairwise Elimination agenda
>> order on Top preferences ... by using those preferences to "de-clone" the
>> Borda agenda order:
>>
>> The agenda order is given by SB(X), the Sum over all ballots B of the
>> first place votes of the candidates ranked above X on B.
>>
>> The larger SB(X), the later X is (on average) in the rankings, and the
>> rearlier X is in the elimination agenda.
>>
>>
>>
>> On Wed, Sep 20, 2023, 4:56 AM Michael Ossipoff <email9648742 at gmail.com>
>> wrote:
>>
>>> If, using voted rankings, Sequential-Pairwise’s comparison-order is
>>> determined by giving, to the candidates with higher top-count score, a
>>> later position in the comparison-order, so that voters don’t know what the
>>> comparison-order will be…
>>>
>>> …
>>>
>>> …& if voters’ knowledge of eachother’s preferences is no better than it
>>> is now in political-elections…
>>>
>>> …
>>>
>>> …Does that Sequential-Pairwise election have an offensive strategy with
>>> gain-expectation comparable to what it would have in MinMax, RP & CSSD?
>>>
>>>
>>>
>>> …And, if so, is there a defensive strategy to thwart or deter that
>>> offensive strategy?
>>>
>>>
>>>
>>> …That seems of interest because Sequential-Pairwise is so much less
>>> computationally-demanding than the other pairwise-count methods.
>>>
>>> If, using voted rankings, Sequential-Pairwise’s comparison-order is
>>> determined by giving, to the candidates with higher top-count score, a
>>> later position in the comparison-order, so that voters don’t know what the
>>> comparison-order will be…
>>>
>>> …
>>>
>>> …& if voters’ knowledge of eachother’s preferences is no better than it
>>> is now in political-elections…
>>>
>>> …
>>>
>>> …Does that Sequential-Pairwise election have an offensive strategy with
>>> gain-expectation comparable to what it would have in MinMax, RP & CSSD?
>>>
>>>
>>>
>>> …And, if so, is there a defensive strategy to thwart or deter that
>>> offensive strategy?
>>>
>>>
>>>
>>> …That seems of interest because Sequential-Pairwise is so much less
>>> computationally-demanding than the other pairwise-count methods.
>>>
>> ----
>>> 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/20230921/f0fcfa3c/attachment-0001.htm>