Hi Robert,

I don't think jump ball is a real voting term. I assume he wants to talk about a
tied situation so that he doesn't need to specify what Condorcet method is being
compared.

However, if the IRV advocate wants to point out that Condorcet methods are
susceptible to burial while IRV is not, he will always be able to find some
scenario to show it.

I think we should say the strategy "works" (i.e. for both of the 40% blocs) only
if they both prefer the coin flip over the election of C.

And actually, as you show below, it may be pessimistic to assume the alternative
to the coin flip is the election of C.

Yes, that has high likelihood of backfiring and electing A. In fact if the method
satisfies Schwartz, there is no tie. A is the only candidate in the Schwartz set:
A has a beatpath to both B and C, which neither can return.

In general a line of thinking like "C is probably going to win, let's try to make
C lose" seems a little suspect, or not fully fleshed out.

Kevin

Hi Kevin, Robert, et al,

No offense to what r b-j wrote, but I think I want to respond mainly
to what Kevin Venzke wrote.

More inline....

On Sat, Apr 9, 2022 at 8:52 PM Kevin Venzke stepjak@yahoo.fr wrote:

The scenarios that FairVote folks like to point out are the classic
Burlington-2009-style center squeeze elections.  The "two parties"
that emerge in the final round of an RCV/IRV election are not really
parties, but rather two coalitions of extremists ready to do battle
against each other.  So, in Burlington in 2009, it was Kurt Wright and
the Republicans versus Bob Kiss and the Vermont Progressives in the
final round of voting, and of course Bob Kiss won.  Can you imagine
the largest city in Vermont electing a Republican?  A state
represented in the United States Senate by Senator Bernie Sanders, the
former mayor of Burlington?  A state whose capital is a 90 minute bus
ride from Burlington?  It seems ludicrous, doesn't it?  But Andy
Montroll (the Democratic Party candidate) almost certainly should have
won the 2009 election, and replaced Bob Kiss (the Vermont Progressive
candidate)

It could be that the Republicans try again in 2024, against Democratic
Party incumbent.  The 2021 election was a close one:
https://en.wikipedia.org/wiki/2021_Burlington_mayoral_election

However, Mayor Weinberger nearly lost to Vermont Progressive rather
than a Republican.  2024 is going to be a very interesting election.
It seems that the Vermont Progressives are the useful idiots helping
to keep the two party system in place.  I'm really glad I didn't vote
for Bernie Sanders in any Democratic Party primary.

I'm betting most people think of "B" as the middle candidate in an ABC
election.  Thus this seems like a more intuitive way of expressing
that election:
ABC - 40%
BAC - 10%
BCA - 10%
CBA - 40%

A and C are the two candidates that get 40% and they are the extremes.
B is caught in the middle with only 20%.  Instead of the electorate
compromising on B (which has 20% first place support, but 80% second
place support), we have a very close election in RCV/IRV between A and
C.  Guess what friends: close elections are going to be close.  There
will be a recount.  This example just demonstrates why IRV/RCV fails
terribly in close elections, and the center squeeze is a real thing:
https://electowiki.org/wiki/Center_squeeze

But yes, put "C" as the middle candidate, and then convince people
that electoral reform is complicated and mumble something about
Arrow's theorem, and make the example more difficult to read.  Totally
expressed in good faith.  Thanks Rob (Richie).  TWO PARTY SYSTEM
FOREVER!!!!!!!!!!  POLARIZATION RULEZZZZ!!!!!!!1!!!1!

So many of the theoretical strategies are (as you say, Kevin), "not
fully fleshed out".

Rob Lanphier
[ who may not have been "Rob" first, but has always been "Rob" best
:-P  Also, Rob Lanphier doesn't actually want the two party system
forever, but may be willing to work for the Democratic party in 2022
to avoid what happened in 2017 through 2021.  It is CRAZY that Justice
Ketanji Brown Jackson's senate confirmation was 53-47, and that was
considered a courageous bipartisan vote by Senators Collins,
Murkowski, and Romney,  For my colleagues in the United States: the
2022 election MATTERS ].

On 10.04.2022 03:49, robert bristow-johnson wrote:

I don't know what "jump ballot" is, either. A plain IRV calculation from
Rob LeGrand's calculator
https://web.archive.org/web/20200813191652/http://www.cs.angelo.edu/~rlegrand/rbvote/calc.html
says:

40: A>C>B
40: B>C>A
10: C>A>B
10: C>B>A

C is eliminated, then A and B are tied and some tiebreaker must be employed.

IMHO, Rob should have broken the symmetry, say

41: A>C>B
40: B>C>A
10: C>A>B
9: C>B>A

to muddy the waters as little as possible. Perhaps "jump ballot" is his
word for "needs a tiebreaker"?

In any case, this looks a bit like center squeeze (A and B are the wing
candidates and C is the center).

I.e.

40: A>B>C (honest A>C>B)
40: B>A>C (honest B>C>A)
10: C>A>B (no change)
10: C>B>A (no change)

This is a DH3-ish situation. C is the Condorcet loser, so in every
Condorcet method that passes Condorcet loser, the result is a perfect
tie between A and B, further suggesting that what he means by "jump
ballot" is "a tie that must be broken".

My first observation is that if we assume that C is the proper winner in
this particular example, then Condorcet dominates IRV in the game
theoretical sense here, because Condorcet gets it wrong in the presence
of strategy, but IRV gets it wrong even without strategy. Or as Warren
put it:

- The pathology can occur with certain patterns of votes *and* plotting.
- IRV has problems that can occur without plotting.
- So, NO SALE!

Anyway, since it is a DH3 situation, we need to first check what happens
when one of the factions bury, because if that backfires, there's no
incentive to bury in the first place.

So suppose that we use BTR-IRV and the B-voters bury first...

i.e.

40: A>C>B (no change)
40: B>A>C (honest B>C>A)
10: C>A>B (no change)
10: C>B>A (no change)

For BTR-IRV, the first round Plurality loser is C. The runner-up has to
be chosen at random from {A, B}. Since the example is completely
symmetrical, there's a 50-50 chance of either. So the burial works if
the B voters consider C worse than a 50-50 chance at either A or B.[1]

More generally, with any method that elects from the Schwartz set, A
wins, as Kevin pointed out. So then there's no incentive for the B group
to bury because that only changes the winner from C (someone they rank
second) to A (someone they rank third).

Since the example is completely symmetrical, the same holds for the A
faction. So neither faction has an incentive to bury in methods that
pass Schwartz.

So if there's anything to take from this (from a Condorcet proponent
perspective), I'd say it's these two things:

1. A method that passes Schwartz eliminates the bury incentive for this

particular election.
2. BTR-IRV does no worse than IRV for this election when there's
strategy, and does strictly better with honesty.

Let's hope it will be Condorcet.

-km

[1]  Of course, if C is closer to B than A, the B faction wouldn't bury,
but since the example is symmetrical, you could then argue that it's the
A faction who buries, because now they have an incentive... as long as
they're not too risk averse.

This discussion about IRV and FairVote (not VoteFair) is motivating me
to point out my different perspective of the Burlington VT mayoral
election result.

IMO the biggest failure in the Burlington was an IIA -- independence of
irrelevant alternatives -- failure.  Not the Condorcet failure.

If the Republican candidate had not been in the race, the Democrat would
have won instead of the Progressive.  That didn't happen because the
Republicans voted sincerely.  If they had voted tactically by ranking
the Republican after the Democrat and before the Progressive, their
ballots would not have gotten "stuck" supporting the Republican during
the top-three counting round where the Democrat was eliminated.

Of course the major weakness of IRV is that the counting does not look
deeply into all the marks on all the ballots.

This weakness can be reduced by removing any clearly unpopular
candidates, such as the Republican in the Burlington election.  I
suggest eliminating the "pairwise losing candidate" -- which is the
"Condorcet loser" among the remaining candidates.

When the FairVote folks defend the Burlington result they spin it to be
a Condorcet failure and point out that a candidate who is second-ranked
on all the ballots doesn't really deserve to win.  That's somewhat
reasonable.

But their only defense against the IIA failure criticism might be that
it rarely happens.  OK, but it does happen.  And it happens because the
method does not look deeply enough into all the marks on all the
ballots.  And that's an obvious flaw in IRV.

Many fans of STAR voting initially supported IRV until this flaw (and
another one I'll introduce below) became obvious.  Alas, instead of
choosing to support a better way to count ranked ballots, the STAR fans
created their own kind of ballot that isn't compatible with anything
else.  Now that STAR voting momentum is slowing down, it's time to
consider resolving the conflicts between IRV, STAR, and better
ranked-choice ballot methods.

That's why I recently suggested a way to improve the Ranked Robin method
that some STAR fans created.

Yet ultimately what's needed is for Rob Richie -- the leader of the
FairVote organization -- to admit that IRV and STV are not acceptable as
they are currently counted, yet can be improved in simple ways to make
them widely acceptable.

The other flaw in IRV is that the FairVote counting method does not
allow a voter to mark two or more candidates at the same preference
level.  This is easy to fix (as I've discussed here before).

These simple improvements to IRV and STV would yield a compromise that I
believe would be acceptable to STAR fans, math-savvy Condorcet fans, the
FairVote organization, and me, the VoteFair guy.

Richard Fobes
The VoteFair guy

On 4/9/2022 6:49 PM, robert bristow-johnson wrote:

This is a good example for the method that successively removes pairs of
the two most disparate ballots until only one ballot remains.

The distance between ballots is the Kendall-tau (or swap cost) distance.

In this case diametrically opposites ACB and BCA are maximally far apart
... three swaps to get from one to the other. . so these polar opposite
ballots are removed two at a time until only the CBA and CAB ballots are
left.  These ballots are also removed in pairs until only one pair is left,
which means that the two finish orders CAB and CBA are tied for winning
finish order.

El dom., 10 de abr. de 2022 2:35 a. m., Kristofer Munsterhjelm <
km_elmet@t-online.de> escribió:

In this case the most disparate pairs are ...
(ABC,CBA) and (BAC, CAB)

After removing ten of each of the pairs, we are left with

30 ABC
30 BAC

These two finish orders are tied.

Do the A or B supporters prefer a coin flip between these orders over a
coin flip between the sincere tied finish orders CAB and CBA?

Look at the geometry of the rank orders:

abc, ACB , CAB, CBA,  BCA, bac

The insincere orders are in lower case.

The question is how likely is it that the A supporters, for example, would
prefer a coin toss between the two outer most orders over a coin toss
between the two inner most orders?

Hi Richard (F.),

Le dimanche 10 avril 2022, 12:34:37 UTC−5, Richard, the VoteFair guy electionmethods@votefair.org a écrit :

I somewhat view it this way, although I think it won't ring true for many people
to call the first-round IRV leader an "irrelevant alternative" or even spoiler.
IIA also sounds a little abstract, and doesn't make it clear that specific voters
can have a grievance, that it's not just a candidate or a theoretician who is
complaining.

So I prefer to focus on the compromise incentive on the Republican voters, as you
mention.

The relationship with Condorcet is that strategy criteria (whether nomination or
voting strategy) can provide the motivation for Condorcet. One doesn't have to
adopt Condorcet as a fundamental principle. And Condorcet also need not be the
last word on whichever properties it's intended to optimize.

So I don't look at a scenario and say "the problem here is that the CW lost." It
should be possible to say more than that. Although perhaps when speaking with
other Condorcet advocates, this statement is good enough as a shorthand.

Kevin

Hi Richard,

Le dimanche 10 avril 2022, 03:21:38 UTC−5, Richard Lung voting@ukscientists.com a écrit :

This is a reasonable stance, but the "1 in 135" argument can be questioned given
that every RCV election is conducted under RCV's specific voting and nomination
incentives. That is, we can't be sure that the candidate selection would be the
same between different methods, or that voters would vote the same way after
accounting for strategic incentives.

Compare the argument that every known FPTP election also elects the Condorcet
winner, as far as anyone can really tell.

You recently shared how to solve Binomial STV ("BSTV") for the three-candidate
one-winner case. Based on winner agreement rates, BSTV seems to be about as
different from RCV as possible. I wouldn't say all methods sit squarely between
them, but it's almost true that RCV and BSTV are closer to any other method than
each other.

This makes it hard for me to imagine what properties there could really be in
common between these two methods.

So I wonder, if you were to have the opportunity to conduct test elections for
Binomial STV, what standards would you use to assess whether the tests were
successful?

Kevin

Thank you, Kevin,

I would not expect to see exactly the same result with Binomial STV as 
traditional STV, especially in the single member case. This is because
single memer preference voting, in all its variations, does not have a
rational count, but only a less accurate ordinal count of more or less.
BSTV uses keep value ratios, for both an election count and an exclusion
count (of preferences in reverse). In the case of (multi-member) STV
there is a rational election count but not a rational exclusion count,
so there is less dearth of preference information, there, but still a
dearth, compared to BSTV.

The difference this makes to the voters (for whom the systenm is for) is
that they are informed:   Vote for the candidates, in order of greatest
choice. Vote also for the candidates in reverse order of least choice.
Your last preference (according to the number of candidates) counts as
much against a candidate as your first preference counts toward a candidate.

In short, the more preferential information obtained gives voters more
power of choice.

In test elections, I would look for the onset of an entropy effect: In
Irish STV elections, the use of preferences rapidly falls off. This
might be less marked with BSTV because of the power to express dislikes,
as well as likes. The comparison I am thinking of is the arrow of time
(pun not intended). Laws of physics make time reversible. Just as BSTV
makes preferences reversible. But times arrow prevails, and this is
accounted for by entropy. Some BSTV elections might not have this
"entropic" effect, so it  might be necessary to use significance tests
of how far candidates are from a quota, without solely relying on an
over-all BSTV result based on a geometric mean candidate keep values of
the election count and the exclusion count.

Regards,

Richard Lung.

On 10/04/2022 23:03, Kevin Venzke wrote: