CONUNDRUM: Can a 2-candidate election with a "Majority Winner" produce the most Bayesian Regret??

[George Sanders]

CONUNDRUM: Can a 2-candidate election with a "Majority Winner" produce the most Bayesian Regret??  

Let’s consider a polarizing demagogue @ 51% vs. the loser who was loved by 49% but was the near 2^nd choice of the other 51% (The ‘loser’ actually has 100% Approval here!!!). Isn't this the common-sense/intuitive reason why Warren Smith has spent so much of his life trying to prove that Score Voting produces the least Bayesian Regret?…and it would do this all-at-once--insofar as Score Voting expands the field of viable candidates, thereby undermining simple poll predictions and ‘strategic voting’ (as would happen to score voting in a 2-candidate race). In other words, Score Voting will select the one winner from a large field of candidates who would likely provide the greatest net satisfaction to the collective group--the electorate. (Every young person already understands score voting intuitively via internet reviews for products and services.) And the above example may also be the common-sense/intuitive reason why Steve Brams and now Clay Shentrup have spent so much of their lives trying to prove that Approval Voting is the easiest-to-implement form of ‘simplified’ Score Voting. 

An interesting note: If you consider this 2-candidate race here as a top-two runoff, then notice that the CW/‘extremist’ won(!).

Which is why I don't want any voting method selecting the “majority-winner”-polarizing-demagogue in the above example, does ANYONE? And if you don’t, then why would you want a runoff or a ranking to select the "majority winner" in ANY  voting method--where TWO polarizing demagogues could capture ALL the first-place votes?! 

[Jameson Quinn]

My contention is that in the real world, Majority Judgment will actually handle situations like this just as well, and closely-related situations significantly better, than Score.

Let's look at some scenarios. First, the scenario where Score does better, which I claim is unrealistic:

51: X100, Y75

49: X0, Y75

(Note: I'm using 100/75/50/25/0 for A/B/C/D/F in MJ)

In this situation, Score would correctly choose Y an MJ would incorrectly choose X. But this would never happen. Why? Because in the real world, polarizing candidates tend to have polarized supporters. Here's a more realistic version:

40: X100, Y0 (honest 50), Z0

11: X100, Y75, Z0

49: Z100, Y75

X is a polarizing candidate. Generally, this means a radical; either on the right (like either of the Le Pens in France, Duke in Louisiana, or Sólyom of Fidesz in Hungary, Rios Montt in Guatemala, or even You-Know-Who in 1930s Germany) or on the left (like Allende in Chile 1970). X's supporters are a bare majority, but their support is strong; only a few of them are willing to vote for any other candidates. Y, the honest utility winner, loses in both Score and MJ.

So far, that's one scenario (which I call unrealistic) where Score is better, and one (which I call simplified but essentially realistic in its outline) where both Score and MJ fail. So why do I think that MJ is better in cases like this? Because if you take that last scenario, and slightly modify just 2 of the X voters, then MJ gets it right but Score fails badly:

40: X100, Y0 (honest 50), Z0

9: X100, Y75, Z0

2: X75, Y100, Z0

49: Z100, Y75

X and Y both have a median of 75 or B, but both GMJ and MJ will give the win to Y because there are fewer 0s for Y.

In essence, this is a chicken dilemma. Y and Z have split their votes, causing Score to elect X; but MJ still elects Y. Note that this can happen even with all-honest voters. Here's a scenario with votes in bold and honest utilities in parentheses:

45: X100(10), Y0(0), Z0(0)

30: X0(0), Y100(11), Z82(9)

25: X0(0), Y82(9), Z100(11)

The Y and Z voters have effectively split their votes by normalizing their utilities to express their honest view of the difference between those two candidates; while the X voters, with no superior alternative, stay unified behind X. X is the utility loser, but wins in Score; while MJ, even if the normalization effect is exaggerated by rounding the second choice candidate down to a 75/B, still correctly elects the utility winner Y.

I find the above scenarios essentially realistic. I've also played with less-simplified versions of them, and I find that the MJ advantage is pretty robust. Yes, it is possible for MJ to fall prey to the chicken dilemma; but in general, score begins to fail chicken even when all of the majority voters are being moderately cooperative, while MJ doesn't fail until at least some voters are being strongly uncooperative. 

I find MJ's behavior better here, because if some voters are strongly uncooperative, it is not the job of the voting system to read their minds and imagine that deep down, they really the other near-clone to be better than the non-clone. Another way of saying that is: Score frequently elects the (honest and voted) Condorcet Loser in chicken dilemma situations, while MJ does not elect that candidate unless they are at least the voted Condorcet winner.

This, to me, is an important advantage for MJ. So let me emphasize that, and close this mail by falling prey to Godwin's law:

ADOLPH HITLER was clearly a minority candidate and would have probably lost under both score and MJ. But in a hypothetical later election, where ADOLPH HITLER had gotten slightly more support but not quite a majority, it is possible that, because of a chicken-dilemma scenario, SCORE VOTING WOULD HAVE ELECTED ADOLPH HITLER while MJ or GMJ would have kept him from power.

(Caps above intended as self-mocking satire, but I actually believe that the above is true. Argument not intended for reproduction on a non-specialist forum.)

Jameson

2013/5/25 George Sanders <geov...@gmail.com>

--
You received this message because you are subscribed to the Google Groups "The Center for Election Science" group.
To unsubscribe from this group and stop receiving emails from it, send an email to electionscien...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Stephen Unger]

Ths argument makes sense to me.

Steve
............

Stephen H. Unger
Professor Emeritus
Computer Science and Electrical Engineering
Columbia University
............

On Sat, 25 May 2013, George Sanders wrote:

>
> CONUNDRUM: Can a 2-candidate election with a "Majority Winner" produce the

> most Bayesian Regret??��
>
> Let�s consider a polarizing demagogue @ 51% vs. the loser who was loved by
> 49%�but was the near 2nd�choice of the other 51%�(The �loser� actually has
> 100% Approval here!!!).�Isn't�this the common-sense/intuitive reason why
> Warren Smith has spent so much of�his�life trying to prove that Score Voting
> produces the least Bayesian Regret?�and it would do this�all-at-once--insofar
> as Score Voting�expands the field of viable candidates, thereby undermining
> simple poll predictions and �strategic voting� (as would happen to score

> voting in a 2-candidate race). In other words, Score Voting will select the
> one winner from a large field of candidates who would likely provide the
> greatest net satisfaction to the collective group--the electorate. (Every
> young person already understands score voting intuitively via internet
> reviews for products and services.) And the above example may also be

> the�common-sense/intuitive reason why Steve Brams and now Clay Shentrup have
> spent so much of�their lives trying to prove that Approval Voting is the
> easiest-to-implement form of �simplified� Score Voting.�
>
> An interesting note: If you consider this 2-candidate race here as a�top-two
> runoff, then notice that the CW/�extremist� won(!).

>
> Which is why I don't want any voting method selecting

> the��majority-winner�-polarizing-demagogue�in the above example, does ANYONE?
> And if you don�t, then why would you want a runoff or a ranking to select the
> "majority winner" in�ANY��voting method--where TWO polarizing demagogues
> could capture ALL the first-place votes?!�

>
> --
> You received this message because you are subscribed to the Google Groups
> "The Center for Election Science" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to electionscien...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

> �
> �
>
>

[Jameson Quinn]

2013/5/25 Stephen Unger <un...@cs.columbia.edu>

Ths argument makes sense to me.

Which one? George's or mine? In either case, what problems do you see with the other one?

Jameson

[Warren D Smith]

see my reply under new title

Range voting (also called "Score voting"), Majority Judgment, and
"Burr dilemma" (also called "chicken dilemma"), and Bayesian Regret



--
Warren D. Smith
http://RangeVoting.org <-- add your endorsement (by clicking
"endorse" as 1st step)

[George Sanders]

EGADS, Warren, you killed my thread! But, not to worry, some other fool will undoubtedly come to my rescue.

Actually, I started this "CONUNDRUM" thread in order to hopefully get responses from Abd &/or Ossipoff--insofar as each of them has, at times, extolled the virtues of “majority wins” as an important concept not to be lightly dismissed. I hope I made my case against “majority wins” as sociologically simple and compelling as possible--the ‘layman’ that I am as I try to wade through Abd’s political tomes in order to glean some of the interesting and thought-provoking concepts hidden within. So-o-o: Abd and Ossipoff: Are you still out there? 

P.S.: Does anyone know just where and how the concept of "majority wins" historically came about?

[Abd ul-Rahman Lomax]

At 08:56 AM 5/25/2013, Jameson Quinn wrote:
>My contention is that in the real world,
>Majority Judgment will actually handle
>situations like this just as well, and
>closely-related situations significantly better, than Score.
>
>Let's look at some scenarios. First, the
>scenario where Score does better, which I claim is unrealistic:
>
>51: X100, Y75
>49: X0, Y75

Darned tooting it's unrealistic. Two candidate
election, voters are being asked to choose
between two options. Score does "better," at the
cost of violating the Majority Criterion, which
is a Bad Idea. Yes, if those are really the
absolute utilities, then violating the Criterion
improves overall utility. It's a very strange and
artificially polarized electorate.

>(Note: I'm using 100/75/50/25/0 for A/B/C/D/F in MJ)

You might not see votes like this in a million
years. *Voting is making choices, not expressing
judgments.* Both those preference strengths are
significant. Voters will have no problem
understanding who their favorite is. What is
totally weird here is half of the electorate that
thinks the utility of the *favorite* of the other half is zero.

>In this situation, Score would correctly choose
>Y an MJ would incorrectly choose X. But this
>would never happen. Why? Because in the real
>world, polarizing candidates tend to have polarized supporters.

And why do the Y voters not vote for their clear
favorite with 100%? Notice that they have *higher preference strength.*

Now, if there are other candidates in the
election, that could explain it. But that's misleading as an example, then.

This is exactly the kind of situation where I
consider that the majority criterion should be
tested, with a literal pairwise election. The Y
voters will be, if the utilities are realistic as
absolute utilities, more highly motivate to turn out. Y will win.

>Here's a more realistic version:
>
>40: X100, Y0 (honest 50), Z0
>11: X100, Y75, Z0
>49: Z100, Y75

Arggh. Please folks, an alarm should go off when
you don't allow "honest" for votes that are
sincere and that don't reverse preferences.

This "more realistic version" has the voter vote
zero for Y, it's a bullet vote. Why? A voter will
bullet vote *because they have a strong
preference.* The vote of 25 is available. Why
would the voter not use it? I can say: because
their preference of X over Y is strong enough
that the vote expresses the result they want,
i.e. their expected utility *is expressed* within
the resolution of the system. If "50" means
"approved," then we know that the voter would
supposedly prefer to vote for Y than to (1) see
the vote be wasted or (2) cause a runoff.

If you are stating absolute utilities, say so,
and for a complete explanation of an election
scenario, stating underlying absolute utilities
allows predicting expected voter behavior. Here,
you have, it seems, confused absolute utilities
with presumed utility for ratings.

How voters will vote will depend on the method
and on how they perceive the frontrunners. No
voters have Y as a favorite which should also set
off the "unrealistic" alarm. Now, I'd like to
consider MJ as a method for use in a runoff
system. 50 or above is approval (C or better). If
50% approval is not found in the primary, then
there will be a runoff. I would have the runoff
be the top two with an extra condition: if the
votes show a pairwise victory for a third
candidate, there *is* a runoff between the
primary leader plus the pairwise winner.

In the primary, this is what I expect as votes --
for most voters in each defined faction:

15: X A, Y D, Z 0 (mildly strategic vote)
10: X A, Y C, Z 0 (accurate vote)
15: X A, Y 0, Z 0 (bullet vote)
11: X A, Y C, Z 0
49: X F, Y B, Z A

MJ: Median grade: X, A, Y, C, Z, F. X wins.
However, with a pairwise trigger for a runoff, Y still loses.

>X is a polarizing candidate.

Why? Z could also be considered polarizing. The
middle candidate, Y, is closer to X than to Z. So
why not describe Z as polarizing?

> Generally, this means a radical; either on the
> right (like either of the Le Pens in France,
> Duke in Louisiana, or Sólyom of Fidesz in
> Hungary, Rios Montt in Guatemala, or even
> You-Know-Who in 1930s Germany) or on the left
> (like Allende in Chile 1970). X's supporters
> are a bare majority, but their support is
> strong; only a few of them are willing to vote
> for any other candidates. Y, the honest utility
> winner, loses in both Score and MJ.

That was not the case with YouKnowWho in Germany.
I.e, he did not have the support of a majority.
The support of X by X voters is *not* strong. The
"real preference strength" is 50% for most of
them. That's middling support; for about 20% of
them, it is weak support, but still significant.
These voters know whom they favor. All of them strongly dislike Z.

Notice: the Z supporters, by inadequate support
for a compromise candidate, cause the loss of
that candidate, and the loss of the election to X, whom they strongly dislike.

What would Range do? The Range votes might be the
same (Range 4), with equivalent runoff triggers.

15: X 4, Y 1, Z 0 (mildly strategic vote)
10: X 4, Y 2, Z 0 (accurate vote)
15: X 4, Y 1, Z 0 (bullet vote has become a mildly strategic vote)
11: X 4, Y 2, Z 0
49: X 0, Y 3, Z 4

Totals:
X: 204
Y: 234
Z: 196

Range winner: Y, though by a relatively small
margin. However, X beats Y pairwise. Runoff. What
will happen in the runoff? This is where
classical voting system analysis has fallen on
its face, by failing to consider that the runoff
will have a *different* electorate. More strongly
motivated voters turn out in larger numbers. We
really don't know what the *rest of the
population considers.* Setting them aside, what
if the runoff is simply a reduced set of *these voters*?

Sincere utilities:
40: X100, Y50
11: X100, Y75
49: X0, Y75

There are 49% of voters who are highly motivated to turn out and vote for Y.
The other 51% of voters are weakly motivate to
turn out and vote for X. It is *highly likely*
that X supporter turnout will be suppressed to
some degree. Y is likely to win. And this is
neglecting the *rest of the population,* that did
not vote in the primary. From the first round
votes, from first preference supporters, who are
more visible, the rest of the population may have
thought that Y did not have a chance.

We are seeing here why top two runoff does
produce comeback winners a third of the time,
entirely aside from the additional campaigning
that may have been done. This is better than top
two runoff, it's a range runoff system. It will
get a condorcet winner into the runoff.

Notice, however, what allowed that: a few X
supporters who voted to show some utility for Y.
If not for that, it could not have happened.
I.e., Later No Harm failure. So this is also a
demonstration of how Later No Harm failure may be
necessary in a system that will optimize absolute social utility.

>So far, that's one scenario (which I call
>unrealistic) where Score is better, and one
>(which I call simplified but essentially
>realistic in its outline) where both Score and MJ fail.

The "failure" is to take the "real utilities" of
the X voters into consideration, beause they were
not voted. The X voters chose to conceal those
utilities, and that behavior is caused by a
strong preference. But, wait, the definition here
is that the preference is weak. Real-world
decision-making process has been neglected for an
assumption that people "want their favorite to
win," so they suppress their "real utilities."
But that's a contradiction: if they have a strong
preference for their favorite to win, *that is their real utility.*

Now, something indeed can go awry in the
translation of absolute utilities to range votes
(or MJ votes). Hence I propose methods that
detect conditions indicating that possibility and
then test it. Runoff voting tests *absolute
utilities* through differential turnout. (That
same process is operational in the primary
election or single elections, it's just not
visible in effect. In a primary that is held as a
special election, rather than as part of a
general election, where a choice is seen as *very
important* to voters, they will turn out in
unusual numbers. In a runoff that is a special
election, voters will turn out in major numbers
to defeat an Awful candidate. A general election
runoff, used in some places, will present a
reduced set of candidates to a broader
electorate, motivated by other elections, perhaps.

[Jameson Quinn]

I think we have a failure of communication. Here's my view of what just happened. (If I'm reading you wrong, please correct me.)

George: 

MJ can do bad things.

Jameson:

In theory, MJ can do bad things, such as X, but that's an unlikely scenario and it's unrealistic to imagine that Score would fix it anyway; and here's how X relates to some good things (Y) that MJ can do.

Stephen: 

 I agree [with George]. 

Jameson:

With George? 

 Warren:

Jameson is wrong about X being unlikely and a probable score failure, but he has a potentially interesting point about Y.

Abd: 

 Jameson:

>  MJ can do bad things, such as X,

 No, Jameson, X is unlikely. And Score would actually fix it, because when I say "Score" I mean "Runoff between Score and Condorcet winners".

Do you see what's wrong with that last response?

Also: when I say "honest" in a scenario, I'm referring to underlying utilities. I think that there should be a certain amount of leeway for assuming certain utilities when building a scenario, especially if those utilities are compatible with some n-dimensional ideological positions. I don't think it's helpful to define "honest" in such a way that all semi-honest votes are tautologically honest.

2013/5/26 Abd ul-Rahman Lomax <a...@lomaxdesign.com>

At 08:56 AM 5/25/2013, Jameson Quinn wrote:

...

Here's a more realistic version:

40: X100, Y0 (honest 50), Z0
11: X100, Y75, Z0
49: Z100, Y75

Arggh. Please folks, an alarm should go off when you don't allow "honest" for votes that are sincere and that don't reverse preferences.

Arggh, yourself. Please see above.

 

This "more realistic version" has the voter vote zero for Y, it's a bullet vote. Why? A voter will bullet vote *because they have a strong preference.* The vote of 25 is available. Why would the voter not use it? I can say: because their preference of X over Y is strong enough that the vote expresses the result they want, i.e. their expected utility *is expressed* within the resolution of the system. If "50" means "approved," then we know that the voter would supposedly prefer to vote for Y than to (1) see the vote be wasted or (2) cause a runoff.

If you are stating absolute utilities, say so, and for a complete explanation of an election scenario, stating underlying absolute utilities allows predicting expected voter behavior. Here, you have, it seems, confused absolute utilities with presumed utility for ratings.

Why does it seem that way? When I say "honest", I mean absolute utility.

 

...

X is a polarizing candidate.

Why? Z could also be considered polarizing. The middle candidate, Y, is closer to X than to Z. So why not describe Z as polarizing?

Both X and Z are polarizing candidates, in that they are rated overwhelmingly at either top or bottom, with few middle ratings. My claim here is (α) that the supporters of such candidates will mostly tend to bullet vote. A similar claim (β) is that if Z's partisans give Y some support, then at least some of Y's partisans will usually give Z some support. Neither α nor β imply the other, but either will be sufficient to support my point here, which is that this kind of MJ failure is unlikely.

Jameson

[Abd ul-Rahman Lomax]

At 06:59 AM 5/26/2013, George Sanders wrote:

>EGADS, Warren, you killed my thread! But, not to worry, some other
>fool will undoubtedly come to my rescue.
>
>Actually, I started this "CONUNDRUM" thread in order to hopefully
>get responses from Abd &/or Ossipoff--insofar as each of them has,
>at times, extolled the virtues of "majority wins" as an important
>concept not to be lightly dismissed.

Not "wins." Decides, chooses. Mere majority rule is the *minimum*,
not the goal. The goal, in fact, is unanimity. There are currently
some interesting arguments in http://en.wikipedia.org/wiki/Majority_rule

> I hope I made my case against "majority wins" as sociologically
> simple and compelling as possible--the 'layman' that I am as I try
> to wade through Abd's political tomes in order to glean some of the
> interesting and thought-provoking concepts hidden within. So-o-o:
> Abd and Ossipoff: Are you still out there?

Ossipoff has promised to stop writing here. I haven't.

Notice my Favorite Voting System. Asset Voting. Multiwinner, it can
set a standard for the election of a seat in an Assembly at
*Unanimity* of a quota, i.e., a specific proportion the votes
matching the proportion of the voters being represented.

I.e., Asset reassembles votes under the intelligent (hopefully!)
direction and control of the "candidates" -- they are really public
voters, "electors," -- to create a seat with a specific number of
voters represented by each seat, all of whom have either voted for
that seat or the person they voted for has reassigned the vote
(directly, probably, but indirect reassingnment is possible) to it.

Asset can thus use the Hare quota. The equivalent of the Hare quota
for a single winner election is unanimity. Unanimity in a single
secret ballot poll is almost entirely impossible, but it could be
possible in some Asset single-winner elections. More attainable, of
course, would be the Droop quota, where the quota is V/(N+1), with V
being the number of voters and N being the number of winners. The
Hare quota is V/N.

>P.S.: Does anyone know just where and how the concept of "majority
>wins" historically came about?

Congratulations on asking a question that could be of great value.

Majority rule is ancient. It has been used in deliberative bodies for
a very, very long time, and remains practically ubiquitous. It is
widely recognized, especially recently, that majority rule can be a
fallback, a safeguard against *minority rule* where the status quo
favors a minority, but it is also recognized that decisions with a
mere majority can be divisive.

In informal decision-making, where a group of friends vote on a
restaurant to go to, say, they may begin with either vote for one or
an approval election, i.e., indicate approval or disapproval for each
nomination. They will tentatively ratify the result by proposing a
choice, and well generally seek unanimous acceptance. This is *not*
necessarily first preference. Ultimate acceptance is indicated by the
friends actually going to the restaurant, sitting down, and ordering.

In such a process, it can happen that someone is an outlier, the only
person opposed to a choice. If that opposition is strong, the person
is likely to express that, and a group will normally reconsider.
There are many possible solutions, from switching to a slightly less
acceptable alternative that can be accepted by the outlier, to
somehow compensating the outlier (i.e., the rest of us want to eat
here, but we will ask the management to allow you to bring in a meal
from another restaurant, so you can join us -- and if they don't
accept that, we will walk out.) Or they could blame the outlier for
being "antisocial," or could respond in other dysfunctional ways.
Human society. Don't leave home without it.

Majority rule was *assumed* in the Federalist papers and in other
discourse and practice at the time. A majority of electors was
required for the election of the President. If that failed, the
election was resolved by Congress according to certain majority rules.

Supermajority is a rule that has been required for fundamental
changes. It's commonly assumed that cloture requires a supermajority.
That's not *entirely* true. The default process requires a
supermajority, but a mere majority can overturn that by effectively
appealing a ruling of the chair and confirming the opposite ruling.

In private organizations following Robert's Rules, changes to the
bylaws may require a two-thirds majority. But an absolute majority of
members eligible to vote can change any bylaw, even without notice.
(The supermajority rule applies to a quorum voting at a meeting
called with notice, etc.)

It's obvious why, and if you don't understand it on reflection, ask again.

A quick summary, though, is that where the status quo favors a
minority, requirements for a super majority -- up to and including
consensus -- effetively create, not the intended consensus rule, but
minority rule. And I've seen this happening in real organizations
that had become enamored of consensus rule. They had found that the
exhileration, and sense of unity created, by consensus process, was
worth the substantial time that it can take. And then, after some
years ... the other side bites. I saw a situation where members were
being injured by a rule that had been established when there were
fewer members, who were wealthier. At the meeting I was at, it was
clear that a strong majority favored changing the rule, but ... those
who had established it were adamant, and refused to stand aside. The
rule was "right," they claimed. And they had built nothing into their
process to handle this, they assumed that a minority like that would
naturally stand aside.

Sometimes, yes. Sometimes, not.

So the majority, it seemed, was helpless. They were not actually
helpless, the majority never is, but they were unwilling to do what
it would take to confront that minority. That confrontation would
violate their expectations of themselves, and they were still holding
on to the exhileration they had found when the group was truly united.

I saw how, after years, participation in community meetings in
cohousing communities declined, as people no longer felt they could
afford the very substantial time it takes to find 100% consensus. Yet
they were unwilling to consider the concept of proxy voting, they
thought of it as if it were "remote voting," only. I.e., someone
would vote specifically on behalf of of an absent member, who had not
heard the deliberation and negotiations, but who was nevertheless
being allowed to make a fixed and determined vote. The concept of an
enabled proxy, who would simply vote their own conscience, being
trusted to do that, was completely off their radar. They had fixed on
a moral concept, it is "better" if everyone is there, struggling with
the process. And people who don't come are "bad." And, sure enough,
these people drift away, participate less and less, and put their
home up for sale....

Majority rule is simple and fast, but also far more flexible than one
might think. Above, I described a group of friends picking a
restaurant. They may quickly settle on a majority choice, but members
who dislike that choice can and will speak up, and the group will
*informally* consider preference strength.

It's like the ancient Spartan voting that Warren has written about:
elections were by the volume of a shout in a public meeting. Range
Voting. Approval, properly implemented, is simply a way of speeding
up the process of repeated elections that is how decisions are
routinely made. Range is great as a polling method.

Now, what about public elections? Plurality actually works most of
the time, under certain circumstances, and those circumstances
include nonpartisan elections with only a few candidates. Usually
more sophisticated methods won't change the outcome, or, if they do,
the plurality result was *close.* There is no major injustice.
Because naive plurality and zero-knowledge plurality voting strategy
is simple for most voters, it was probably used, almost
automatically, for public elections. We got stuck with it.

Robert's Rules of Order *never* recommends election by a plurality.
It does generally recommend vote-for-one, for all voting, and only
allows voting for multiple choices in narrow circumstances,
safeguarded by the right of a majority to reject an election. Most
questions are Yes/No.

Real deliberative process considers preference strength *informally*.
That is, it actually functions much more like Range Voting than some
voting systems theoriests, without experience with deliberative
process -- or with some negative experience at some point in their
lives -- realize.

Basically, members who don't know the procedures, in the presence of
a chair who doesn't get how important it is for the chair to *enable*
and *empower* all members to be heard properly, and in the absence of
members who will use the rules to stand up for minorities, even if
they disagre with them, may find their experience constricted and may
feel oppressed. But this is the bottom line: if a majority of members
of a voluntary organization aren't willing to support you, why are
you trying to support the organization?

Generally, Robert's Rules, operating as to most decisions by majority
rule, requires a supermajority to take certain actions that can
impact minorities. It normally takes a two-thirds vote to proceed to
a vote on a motion. That is a decision that can prevent a minority
from presenting an argument they consider important. (That's a
"cloture" rule, it closes debate.)

Abuse of this by a minority that simply wants to stop the majority
from making a decision is "dilatory," which any chair can rule out of
order. And that's why it's possible to move around filibusters. The
U.S. Senate got stuck in a tradition while forgetting the reason for
the tradition. It became purely rule-bound. It was never the
tradition that a minority could block voting, only that a minority
had the right to complete debate. As soon as minority members start
reading novels into the record, etc., obviously to simply fill the
time, as soon as arguments become completely repetitive, etc., it
could be stopped.

And that is a proper exercise of decision-making power by the
majority. After all, the majority has the right to decide it doesn't
want to hear any more, it's ready to decide. That is *routinely* a
two-thirds decision, which was fairly recently lowered to
three-fifths in the Senate, but filibusters are not routine
completion of debate, by definition. They are delaying tactics.

I came to this in my study of consensus organizations: consensus is
highly desirable, but rules that *require* it readily become
oppressive. How does a group decide when it is ready to decide and
approve or reject a motion? I came to the conclusion that if the
required level, other than as a routine procedural setting of the
standard at a higher fraction, is more than half, the majority is
being coerced by the minority, into continuing a discussion without
their consent. The majority will, wisely, not jump to cut off debate
that allows a community to seek and find broader consensus, but will
cut it off when the debate starts to do more harm than good. There is
no substitute for an awake and thoughtful community. If the majority
is wise, then, the community will be wise. If not, well, there is
little hope for community wisdom.