The election methods trade-off paradox/impossibility theorems paradox.
The election methods trade-off paradox/impossibility theorems paradox.
For the sake of argument, suppose a trade-off theory of elections that
there is no consistently democratic electoral system: the impossibility
supposition.
That supposition implies some conception (albeit non-existent) of a
consistently derived right election result.
If there is no such measure, then there is no standard even to judge
that there is a trade-off between electoral systems.
Suppose there is a consistent theory of choice, setting a standard by
which electoral systems can be judged for their democratic consistency.
It follows that the election result will only be as consistent as the
electoral system, and there is no pre-conceivably right election result,
because that presupposes a perfection not given to science as a
progressive pursuit.
--
Richard Lung.
http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:
https://plus.google.com/106191200795605365085
E-books in epub format:
https://www.smashwords.com/profile/view/democracyscience
I kinda don't accept this paradox. Just to compare the form of a election
method paradox statement: Arrow's theorem was that given a set of desired
properties and the constraint of rankings ballots, those set of desirable
properties could not all be simultaneously fulfilled. One can almost
trivially step outside of that paradox by eliminating the constraint of the
rankings ballot.
My model of understanding people and elections is a utilitarian one. A
person derives some amount of utility from the outcome of an election and
everyone is apportioned the same share of utility which we might count as
0..1 or -1..1 . These model persons can be summed up and and a global
social utility calculated. The ideal election method perfectly knows every
person and elects the true global social utility maximizing candidate. This
sounds an awful lot like score voting. But then we have to start to
complicate the model with imperfect knowledge of a voter's utility, the
imperfect expression of that on a ballot, strategic ballot casting rather
than honest, messy computation and practical administration issues of
running an election in the real world, and so on. So we might wind up with
a best practical method that isn't just simple score voting.
But I still believe there is a pragmatic 'best' method, we have techniques
for evaluating that, and we should do this and put something up in the real
world. Personally I'll take a rankings ballot that's Condorcet counted with
any cycle resolution method as 'good enough' and practically applicable;
and tinkering around the edges for a slightly better method is fun
mathematical curiosity but I'd also like to get some laws passed.
What do you think of my model statement?
Is there a more formal statement of limitations you were heading towards?
On Thu, Jun 22, 2017 at 2:30 AM, Richard Lung voting@ukscientists.com
wrote:
The election methods trade-off paradox/impossibility theorems paradox.
For the sake of argument, suppose a trade-off theory of elections that
there is no consistently democratic electoral system: the impossibility
supposition.
That supposition implies some conception (albeit non-existent) of a
consistently derived right election result.
If there is no such measure, then there is no standard even to judge that
there is a trade-off between electoral systems.
Suppose there is a consistent theory of choice, setting a standard by
which electoral systems can be judged for their democratic consistency.
It follows that the election result will only be as consistent as the
electoral system, and there is no pre-conceivably right election result,
because that presupposes a perfection not given to science as a progressive
pursuit.
--
Richard Lung.http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:https://plus.google.com/106191200795605365085
E-books https://plus.google.com/106191200795605365085E-books in epub format:https://www.smashwords.com/profile/view/democracyscience
Election-Methods mailing list - see http://electorama.com/em for list info
Brian Olson,
Where we differ is that I do not see ranked choice as a constraint.
Single order choice, the x-vote is the constraint on ranked voting as a
multiple-order choice.
The problem with election methods, practical and theoretical is that
they impose constraints on the voters freedom of choice, in one way or
another, and so are that much less true election methods.
(A minor example, the classic objections to cumulative voting seem to
apply to some apprently modern versions or variations.)
When Condorcet and Borda, disagreed on the best way to conduct a count
of preference voting, Laplace decided in favor of Borda. (I grant you
that Condorcet has information value, when weighted. But I am not well
informed on this approach and know of no convincing reason why it should
be adopted or how you would persuade the public of that.) JFS Ross
explained that Laplace favored Method Borda because higher preferences
were more important and should count more. The Gregory method removes
the objection to Borda of "later harm." This is the direction I have
followed (weighted count of ranked choice), following on from where Meek
method STV leaves off.
Richard Lung
On 22/06/2017 15:01, Brian Olson wrote:
I kinda don't accept this paradox. Just to compare the form of a
election method paradox statement: Arrow's theorem was that given a
set of desired properties and the constraint of rankings ballots,
those set of desirable properties could not all be simultaneously
fulfilled. One can almost trivially step outside of that paradox by
eliminating the constraint of the rankings ballot.
My model of understanding people and elections is a utilitarian one. A
person derives some amount of utility from the outcome of an election
and everyone is apportioned the same share of utility which we might
count as 0..1 or -1..1 . These model persons can be summed up and and
a global social utility calculated. The ideal election method
perfectly knows every person and elects the true global social utility
maximizing candidate. This sounds an awful lot like score voting. But
then we have to start to complicate the model with imperfect knowledge
of a voter's utility, the imperfect expression of that on a ballot,
strategic ballot casting rather than honest, messy computation and
practical administration issues of running an election in the real
world, and so on. So we might wind up with a best practical method
that isn't just simple score voting.
But I still believe there is a pragmatic 'best' method, we have
techniques for evaluating that, and we should do this and put
something up in the real world. Personally I'll take a rankings ballot
that's Condorcet counted with any cycle resolution method as 'good
enough' and practically applicable; and tinkering around the edges for
a slightly better method is fun mathematical curiosity but I'd also
like to get some laws passed.
What do you think of my model statement?
Is there a more formal statement of limitations you were heading towards?
On Thu, Jun 22, 2017 at 2:30 AM, Richard Lung <voting@ukscientists.com
mailto:voting@ukscientists.com> wrote:
The election methods trade-off paradox/impossibility theorems paradox.
For the sake of argument, suppose a trade-off theory of elections
that there is no consistently democratic electoral system: the
impossibility supposition.
That supposition implies some conception (albeit non-existent) of
a consistently derived right election result.
If there is no such measure, then there is no standard even to
judge that there is a trade-off between electoral systems.
Suppose there is a consistent theory of choice, setting a standard
by which electoral systems can be judged for their democratic
consistency.
It follows that the election result will only be as consistent as
the electoral system, and there is no pre-conceivably right
election result, because that presupposes a perfection not given
to science as a progressive pursuit.
--
Richard Lung.
http://www.voting.ukscientists.com <http://www.voting.ukscientists.com>
Democracy Science series 3 free e-books in pdf:
https://plus.google.com/106191200795605365085
E-books <https://plus.google.com/106191200795605365085E-books> in epub format:
https://www.smashwords.com/profile/view/democracyscience <https://www.smashwords.com/profile/view/democracyscience>
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
Richard Lung.
http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:
https://plus.google.com/106191200795605365085
E-books in epub format:
https://www.smashwords.com/profile/view/democracyscience
Compared to a rankings ballot, a ratings ballot contains more information
about a voter's preference or utility for the available choices.
I can say
A > B > C > D
or I can say with more detail
A = 1.0; B = 0.9; C = 0.1; D = 0.0
If there is more information available, it is possible for an election
algorithm to use that information and more accurately represent the voters
and find the greater global utility winner.
If we had a specially insightful rational bayesian voting populace we might
ask them for their confidence interval of how sure they are are about each
choice and get more information still.
A = 1.0 (e=.3); B = 0.9 (e=.1), C = 0.1 (e=0.5), D = 0.0 (e=0.1)
And then we'd work out an election algorithm to maximize the expected value
of the global utility over the known voter utility distributions.
On Thu, Jun 22, 2017 at 2:47 PM, Richard Lung voting@ukscientists.com
wrote:
Brian Olson,
Where we differ is that I do not see ranked choice as a constraint. Single
order choice, the x-vote is the constraint on ranked voting as a
multiple-order choice.
The problem with election methods, practical and theoretical is that they
impose constraints on the voters freedom of choice, in one way or another,
and so are that much less true election methods.
(A minor example, the classic objections to cumulative voting seem to
apply to some apprently modern versions or variations.)
When Condorcet and Borda, disagreed on the best way to conduct a count of
preference voting, Laplace decided in favor of Borda. (I grant you that
Condorcet has information value, when weighted. But I am not well informed
on this approach and know of no convincing reason why it should be adopted
or how you would persuade the public of that.) JFS Ross explained that
Laplace favored Method Borda because higher preferences were more important
and should count more. The Gregory method removes the objection to Borda of
"later harm." This is the direction I have followed (weighted count of
ranked choice), following on from where Meek method STV leaves off.
Richard Lung
On 22/06/2017 15:01, Brian Olson wrote:
I kinda don't accept this paradox. Just to compare the form of a election
method paradox statement: Arrow's theorem was that given a set of desired
properties and the constraint of rankings ballots, those set of desirable
properties could not all be simultaneously fulfilled. One can almost
trivially step outside of that paradox by eliminating the constraint of the
rankings ballot.
My model of understanding people and elections is a utilitarian one. A
person derives some amount of utility from the outcome of an election and
everyone is apportioned the same share of utility which we might count as
0..1 or -1..1 . These model persons can be summed up and and a global
social utility calculated. The ideal election method perfectly knows every
person and elects the true global social utility maximizing candidate. This
sounds an awful lot like score voting. But then we have to start to
complicate the model with imperfect knowledge of a voter's utility, the
imperfect expression of that on a ballot, strategic ballot casting rather
than honest, messy computation and practical administration issues of
running an election in the real world, and so on. So we might wind up with
a best practical method that isn't just simple score voting.
But I still believe there is a pragmatic 'best' method, we have techniques
for evaluating that, and we should do this and put something up in the real
world. Personally I'll take a rankings ballot that's Condorcet counted with
any cycle resolution method as 'good enough' and practically applicable;
and tinkering around the edges for a slightly better method is fun
mathematical curiosity but I'd also like to get some laws passed.
What do you think of my model statement?
Is there a more formal statement of limitations you were heading towards?
On Thu, Jun 22, 2017 at 2:30 AM, Richard Lung voting@ukscientists.com
wrote:
The election methods trade-off paradox/impossibility theorems paradox.
For the sake of argument, suppose a trade-off theory of elections that
there is no consistently democratic electoral system: the impossibility
supposition.
That supposition implies some conception (albeit non-existent) of a
consistently derived right election result.
If there is no such measure, then there is no standard even to judge that
there is a trade-off between electoral systems.
Suppose there is a consistent theory of choice, setting a standard by
which electoral systems can be judged for their democratic consistency.
It follows that the election result will only be as consistent as the
electoral system, and there is no pre-conceivably right election result,
because that presupposes a perfection not given to science as a progressive
pursuit.
--
Richard Lung.http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:https://plus.google.com/106191200795605365085
E-books https://plus.google.com/106191200795605365085E-books in epub format:https://www.smashwords.com/profile/view/democracyscience
Election-Methods mailing list - see http://electorama.com/em for list
info
--
Richard Lung.http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:https://plus.google.com/106191200795605365085
E-books in epub format:https://www.smashwords.com/profile/view/democracyscience
so many discussions here are more arcane than i can grok. �but i can grok most of this.
---------------------------- Original Message ----------------------------
Subject: Re: [EM] The election methods trade-off paradox/impossibility theorems paradox.
From: "Brian Olson" bql@bolson.org
Date: Thu, June 22, 2017 4:31 pm
To: "Richard Lung" voting@ukscientists.com
Cc: "EM" election-methods@lists.electorama.com
Compared to a rankings ballot, a ratings ballot contains more information
about a voter's preference or utility for the available choices.
I can say
A > B > C > D
or I can say with more detail
A = 1.0; B = 0.9; C = 0.1; D = 0.0
If there is more information available, it is possible for an election
algorithm to use that information and more accurately represent the voters
and find the greater global utility winner.
but voters can skew that information quantitatively and, if they want to be tactical, insincerely.
how does the voter know that by rating Candidate B with 0.9 and not lower, that he is not helping this candidate beat his favorite, Candidate A? �maybe it should be 0.8. �or
0.5.
while Score voting requires too much information from the voters (making them act as a trained expert and consider quantitatively how candidates should be rated as if they are an Olympic ice skating judge) and while Approval doesn't get enough information from voters (does not
differentiate preference between two "approved" candidates), there has always been the problem of either Score voting or Approval voting facing the voter about what to do with their second choice (Candidate B). �how much juice should the voter give to Candidate B if they want B to
beat C but do not want B to beat A? �(and for Approval, the question is, having the same concern, shall the voter "approve" B or not?)
Score and Approval cannot answer that question in any simple manner. �but with Ranked-Choice the answer is clear.
and the other problem with Score is the "One-Person-One-Vote" standard. �if i really, really, really like Candidate A over Candidate B and you only sorta like B
over A by just a little bit, it should not matter to what disparate degree we like our candidates. �your vote for B should weigh just as much as my vote for A, even if my excitement for A exceeds your excitement for B. �that's what "one-person-one-vote" means.
i still just
do not get why Score and Approval (in application to governmental elections) have the following they do have.
bestest,
r b-j
If we had a specially insightful rational bayesian voting populace we might
ask them for their confidence interval of how sure they are are about each
choice and get more information still.
A = 1.0 (e=.3); B = 0.9 (e=.1), C = 0.1 (e=0.5), D = 0.0 (e=0.1)
And then we'd work out an election algorithm to maximize the expected value
of the global utility over the known voter utility distributions.
�
and, except for making examples to show how a system breaks (like a Proof by Contradiction in mathematics), i just cannot see how judging the comparative value of systems with simulated assumptions is helpful. �there are not enough simulated cases to be able to consider how any
system will work under all conditions.
�
On Thu, Jun 22, 2017 at 2:47 PM, Richard Lung voting@ukscientists.com
wrote:
Brian Olson,
Where we differ is that I do not see ranked choice as a constraint. Single
order choice, the x-vote is the constraint on ranked voting as a
multiple-order choice.
The problem with election methods, practical and theoretical is that they
impose constraints on the voters freedom of choice, in one way or another,
and so are that much less true election methods.
(A minor example, the classic objections to cumulative voting seem to
apply to some apprently modern versions or variations.)
When Condorcet and Borda, disagreed on the best way to conduct a count of
preference voting, Laplace decided in favor of Borda. (I grant you that
Condorcet has information value, when weighted. But I am not well informed
on this approach and know of no convincing reason why it should be adopted
or how you would persuade the public of that.) JFS Ross explained that
Laplace favored Method Borda because higher preferences were more important
and should count more. The Gregory method removes the objection to Borda of
"later harm." This is the direction I have followed (weighted count of
ranked choice), following on from where Meek method STV leaves off.
Richard Lung
On 22/06/2017 15:01, Brian Olson wrote:
I kinda don't accept this paradox. Just to compare the form of a election
method paradox statement: Arrow's theorem was that given a set of desired
properties and the constraint of rankings ballots, those set of desirable
properties could not all be simultaneously fulfilled. One can almost
trivially step outside of that paradox by eliminating the constraint of the
rankings ballot.
My model of understanding people and elections is a utilitarian one. A
person derives some amount of utility from the outcome of an election and
everyone is apportioned the same share of utility which we might count as
0..1 or -1..1 . These model persons can be summed up and and a global
social utility calculated. The ideal election method perfectly knows every
person and elects the true global social utility maximizing candidate. This
sounds an awful lot like score voting. But then we have to start to
complicate the model with imperfect knowledge of a voter's utility, the
imperfect expression of that on a ballot, strategic ballot casting rather
than honest, messy computation and practical administration issues of
running an election in the real world, and so on. So we might wind up with
a best practical method that isn't just simple score voting.
But I still believe there is a pragmatic 'best' method, we have techniques
for evaluating that, and we should do this and put something up in the real
world. Personally I'll take a rankings ballot that's Condorcet counted with
any cycle resolution method as 'good enough' and practically applicable;
and tinkering around the edges for a slightly better method is fun
mathematical curiosity but I'd also like to get some laws passed.
What do you think of my model statement?
Is there a more formal statement of limitations you were heading towards?
On Thu, Jun 22, 2017 at 2:30 AM, Richard Lung voting@ukscientists.com
wrote:
The election methods trade-off paradox/impossibility theorems paradox.
For the sake of argument, suppose a trade-off theory of elections that
there is no consistently democratic electoral system: the impossibility
supposition.
That supposition implies some conception (albeit non-existent) of a
consistently derived right election result.
If there is no such measure, then there is no standard even to judge that
there is a trade-off between electoral systems.
Suppose there is a consistent theory of choice, setting a standard by
which electoral systems can be judged for their democratic consistency.
It follows that the election result will only be as consistent as the
electoral system, and there is no pre-conceivably right election result,
because that presupposes a perfection not given to science as a progressive
pursuit.
--
Richard Lung.http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:https://plus.google.com/106191200795605365085
E-books https://plus.google.com/106191200795605365085E-books in epub format:https://www.smashwords.com/profile/view/democracyscience
--
Richard Lung.http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:https://plus.google.com/106191200795605365085
E-books in epub format:https://www.smashwords.com/profile/view/democracyscience
--
r b-j � � � � � � � � �rbj@audioimagination.com
"Imagination is more important than knowledge."
Ah yes, the old criticism against cumulative voting (made for instance
by Enid Lakeman) is relevant here. Cumulative votes count against each
other.
I suppose the basic objection is that voters cannot be also their own
counters, because that is a function of peoples combined votes, not of
individual voters.
BTW I didn't answer your previous question about a formal basis for my
election method. It is based on work by SS Stevens on the scales of
measurement, published in "Science" in the 1940s. Long ago, when I was
a young man, Unesco gave me a copyright for this.
Richard Lung.
On 22/06/2017 21:31, Brian Olson wrote:
Compared to a rankings ballot, a ratings ballot contains more
information about a voter's preference or utility for the available
choices.
I can say
A > B > C > D
or I can say with more detail
A = 1.0; B = 0.9; C = 0.1; D = 0.0
If there is more information available, it is possible for an election
algorithm to use that information and more accurately represent the
voters and find the greater global utility winner.
If we had a specially insightful rational bayesian voting populace we
might ask them for their confidence interval of how sure they are are
about each choice and get more information still.
A = 1.0 (e=.3); B = 0.9 (e=.1), C = 0.1 (e=0.5), D = 0.0 (e=0.1)
And then we'd work out an election algorithm to maximize the expected
value of the global utility over the known voter utility distributions.
On Thu, Jun 22, 2017 at 2:47 PM, Richard Lung <voting@ukscientists.com
mailto:voting@ukscientists.com> wrote:
Brian Olson,
Where we differ is that I do not see ranked choice as a
constraint. Single order choice, the x-vote is the constraint on
ranked voting as a multiple-order choice.
The problem with election methods, practical and theoretical is
that they impose constraints on the voters freedom of choice, in
one way or another, and so are that much less true election methods.
(A minor example, the classic objections to cumulative voting seem
to apply to some apprently modern versions or variations.)
When Condorcet and Borda, disagreed on the best way to conduct a
count of preference voting, Laplace decided in favor of Borda. (I
grant you that Condorcet has information value, when weighted. But
I am not well informed on this approach and know of no convincing
reason why it should be adopted or how you would persuade the
public of that.) JFS Ross explained that Laplace favored Method
Borda because higher preferences were more important and should
count more. The Gregory method removes the objection to Borda of
"later harm." This is the direction I have followed (weighted
count of ranked choice), following on from where Meek method STV
leaves off.
Richard Lung
On 22/06/2017 15:01, Brian Olson wrote:
I kinda don't accept this paradox. Just to compare the form of a
election method paradox statement: Arrow's theorem was that given
a set of desired properties and the constraint of rankings
ballots, those set of desirable properties could not all be
simultaneously fulfilled. One can almost trivially step outside
of that paradox by eliminating the constraint of the rankings ballot.
My model of understanding people and elections is a utilitarian
one. A person derives some amount of utility from the outcome of
an election and everyone is apportioned the same share of utility
which we might count as 0..1 or -1..1 . These model persons can
be summed up and and a global social utility calculated. The
ideal election method perfectly knows every person and elects the
true global social utility maximizing candidate. This sounds an
awful lot like score voting. But then we have to start to
complicate the model with imperfect knowledge of a voter's
utility, the imperfect expression of that on a ballot, strategic
ballot casting rather than honest, messy computation and
practical administration issues of running an election in the
real world, and so on. So we might wind up with a best practical
method that isn't just simple score voting.
But I still believe there is a pragmatic 'best' method, we have
techniques for evaluating that, and we should do this and put
something up in the real world. Personally I'll take a rankings
ballot that's Condorcet counted with any cycle resolution method
as 'good enough' and practically applicable; and tinkering around
the edges for a slightly better method is fun mathematical
curiosity but I'd also like to get some laws passed.
What do you think of my model statement?
Is there a more formal statement of limitations you were heading
towards?
On Thu, Jun 22, 2017 at 2:30 AM, Richard Lung
<voting@ukscientists.com <mailto:voting@ukscientists.com>> wrote:
The election methods trade-off paradox/impossibility theorems
paradox.
For the sake of argument, suppose a trade-off theory of
elections that there is no consistently democratic electoral
system: the impossibility supposition.
That supposition implies some conception (albeit
non-existent) of a consistently derived right election result.
If there is no such measure, then there is no standard even
to judge that there is a trade-off between electoral systems.
Suppose there is a consistent theory of choice, setting a
standard by which electoral systems can be judged for their
democratic consistency.
It follows that the election result will only be as
consistent as the electoral system, and there is no
pre-conceivably right election result, because that
presupposes a perfection not given to science as a
progressive pursuit.
--
Richard Lung.
http://www.voting.ukscientists.com <http://www.voting.ukscientists.com>
Democracy Science series 3 free e-books in pdf:
https://plus.google.com/106191200795605365085
E-books <https://plus.google.com/106191200795605365085E-books> in epub format:
https://www.smashwords.com/profile/view/democracyscience <https://www.smashwords.com/profile/view/democracyscience>
----
Election-Methods mailing list - see http://electorama.com/em
for list info
--
Richard Lung.
http://www.voting.ukscientists.com <http://www.voting.ukscientists.com>
Democracy Science series 3 free e-books in pdf:
https://plus.google.com/106191200795605365085 <https://plus.google.com/106191200795605365085>
E-books in epub format:
https://www.smashwords.com/profile/view/democracyscience <https://www.smashwords.com/profile/view/democracyscience>
--
Richard Lung.
http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:
https://plus.google.com/106191200795605365085
E-books in epub format:
https://www.smashwords.com/profile/view/democracyscience
I was speaking only of ballots, and and in the abstract that some election
algorithm could take that information and make a good outcome of it.
I don't favor raw Score summation. It's strategy prone. For choices where
my honest vote might be [1.0, 0.8, 0.6, 0.4, 0.2, 0.0] I should probably
vote strategically [1.0, 1.0, 1.0, 0.0, 0.0, 0.0].
And if you don't like that and the varying vote power depending on how you
vote: I have a system for you!
"Instant Runoff Normalized Ratings" (IRNR)
Each ballot is normalized so that all ballots have the same magnitude. The
modified ballots are summed, and the choice with the lowest sumarry rating
is disqualified. Each ballot is then normalized again as if the
disqualified choice was not there, redistributing the vote across the
choices in proportion to the original ballot. The new modified ballots are
summed and the process is repeated until there are two choices remaining
and one choice wins over the other.
I think this works better with an honest ballot in the case where you like
some choice more than another 'just a little bit' or by whatever margin.
/Brian
On Fri, Jun 23, 2017 at 1:02 AM, robert bristow-johnson <
rbj@audioimagination.com> wrote:
so many discussions here are more arcane than i can grok. but i can grok
most of this.
---------------------------- Original Message ----------------------------
Subject: Re: [EM] The election methods trade-off paradox/impossibility
theorems paradox.
From: "Brian Olson" bql@bolson.org
Date: Thu, June 22, 2017 4:31 pm
To: "Richard Lung" voting@ukscientists.com
Cc: "EM" election-methods@lists.electorama.com
Compared to a rankings ballot, a ratings ballot contains more information
about a voter's preference or utility for the available choices.
I can say
A > B > C > D
or I can say with more detail
A = 1.0; B = 0.9; C = 0.1; D = 0.0
If there is more information available, it is possible for an election
algorithm to use that information and more accurately represent the
voters
and find the greater global utility winner.
but voters can skew that information quantitatively and, if they want to
be tactical, insincerely.
how does the voter know that by rating Candidate B with 0.9 and not lower,
that he is not helping this candidate beat his favorite, Candidate A?
maybe it should be 0.8. or 0.5.
while Score voting requires too much information from the voters (making
them act as a trained expert and consider quantitatively how candidates
should be rated as if they are an Olympic ice skating judge) and while
Approval doesn't get enough information from voters (does not differentiate
preference between two "approved" candidates), there has always been the
problem of either Score voting or Approval voting facing the voter about
what to do with their second choice (Candidate B). how much juice should
the voter give to Candidate B if they want B to beat C but do not want B to
beat A? (and for Approval, the question is, having the same concern, shall
the voter "approve" B or not?)
Score and Approval cannot answer that question in any simple manner. but
with Ranked-Choice the answer is clear.
and the other problem with Score is the "One-Person-One-Vote" standard.
if i really, really, really like Candidate A over Candidate B and you only
sorta like B over A by just a little bit, it should not matter to what
disparate degree we like our candidates. your vote for B should weigh just
as much as my vote for A, even if my excitement for A exceeds your
excitement for B. that's what "one-person-one-vote" means.
i still just do not get why Score and Approval (in application to
governmental elections) have the following they do have.
bestest,
r b-j
If we had a specially insightful rational bayesian voting populace we
might
ask them for their confidence interval of how sure they are are about
each
choice and get more information still.
A = 1.0 (e=.3); B = 0.9 (e=.1), C = 0.1 (e=0.5), D = 0.0 (e=0.1)
And then we'd work out an election algorithm to maximize the expected
value
of the global utility over the known voter utility distributions.
and, except for making examples to show how a system breaks (like a Proof
by Contradiction in mathematics), i just cannot see how judging the
comparative value of systems with simulated assumptions is helpful. there
are not enough simulated cases to be able to consider how any system will
work under all conditions.
On Thu, Jun 22, 2017 at 2:47 PM, Richard Lung voting@ukscientists.com
wrote:
Brian Olson,
Where we differ is that I do not see ranked choice as a constraint.
Single
order choice, the x-vote is the constraint on ranked voting as a
multiple-order choice.
The problem with election methods, practical and theoretical is that
they
impose constraints on the voters freedom of choice, in one way or
another,
and so are that much less true election methods.
(A minor example, the classic objections to cumulative voting seem to
apply to some apprently modern versions or variations.)
When Condorcet and Borda, disagreed on the best way to conduct a count
of
preference voting, Laplace decided in favor of Borda. (I grant you that
Condorcet has information value, when weighted. But I am not well
informed
on this approach and know of no convincing reason why it should be
adopted
or how you would persuade the public of that.) JFS Ross explained that
Laplace favored Method Borda because higher preferences were more
important
and should count more. The Gregory method removes the objection to
Borda of
"later harm." This is the direction I have followed (weighted count of
ranked choice), following on from where Meek method STV leaves off.
Richard Lung
On 22/06/2017 15:01, Brian Olson wrote:
I kinda don't accept this paradox. Just to compare the form of a
election
method paradox statement: Arrow's theorem was that given a set of
desired
properties and the constraint of rankings ballots, those set of
desirable
properties could not all be simultaneously fulfilled. One can almost
trivially step outside of that paradox by eliminating the constraint of
the
rankings ballot.
My model of understanding people and elections is a utilitarian one. A
person derives some amount of utility from the outcome of an election
and
everyone is apportioned the same share of utility which we might count
as
0..1 or -1..1 . These model persons can be summed up and and a global
social utility calculated. The ideal election method perfectly knows
every
person and elects the true global social utility maximizing candidate.
This
sounds an awful lot like score voting. But then we have to start to
complicate the model with imperfect knowledge of a voter's utility, the
imperfect expression of that on a ballot, strategic ballot casting
rather
than honest, messy computation and practical administration issues of
running an election in the real world, and so on. So we might wind up
with
a best practical method that isn't just simple score voting.
But I still believe there is a pragmatic 'best' method, we have
techniques
for evaluating that, and we should do this and put something up in the
real
world. Personally I'll take a rankings ballot that's Condorcet counted
with
any cycle resolution method as 'good enough' and practically applicable;
and tinkering around the edges for a slightly better method is fun
mathematical curiosity but I'd also like to get some laws passed.
What do you think of my model statement?
Is there a more formal statement of limitations you were heading
towards?
On Thu, Jun 22, 2017 at 2:30 AM, Richard Lung voting@ukscientists.com
wrote:
The election methods trade-off paradox/impossibility theorems paradox.
For the sake of argument, suppose a trade-off theory of elections that
there is no consistently democratic electoral system: the impossibility
supposition.
That supposition implies some conception (albeit non-existent) of a
consistently derived right election result.
If there is no such measure, then there is no standard even to judge
that
there is a trade-off between electoral systems.
Suppose there is a consistent theory of choice, setting a standard by
which electoral systems can be judged for their democratic consistency.
It follows that the election result will only be as consistent as the
electoral system, and there is no pre-conceivably right election
result,
because that presupposes a perfection not given to science as a
progressive
pursuit.
--
Richard Lung.http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:
epub format:https://www.smashwords.com/profile/view/democracyscience
--
Richard Lung.http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:https://plus.google.com/
106191200795605365085
E-books in epub format:https://www.smashwords.com/profile/view/
democracyscience
--
r b-j rbj@audioimagination.com
"Imagination is more important than knowledge."
Election-Methods mailing list - see http://electorama.com/em for list info
---------------------------- Original Message ----------------------------
Subject: Re: [EM] The election methods trade-off paradox/impossibility theorems paradox.
From: "Brian Olson" bql@bolson.org
Date: Fri, June 23, 2017 8:35 am
To: "EM" election-methods@lists.electorama.com
I was speaking only of ballots, and and in the abstract that some election
algorithm could take that information and make a good outcome of it.
No, we should not make the voters cook up that information. �all we should ask the voters is "whom do you prefer A or B?" �and "if you can't get your favorite, whom is your next
preference?"
I don't favor raw Score summation. It's strategy prone.
of course it is. �scoring is strategy prone if you use the scores in any algorithm other than simple ranking. �and then don't use scores. �just rank.
� For choices where
my honest vote might be [1.0, 0.8, 0.6, 0.4, 0.2, 0.0] I should probably
vote strategically [1.0, 1.0, 1.0, 0.0, 0.0, 0.0].
but then you're not helping your first choice beat your second or third choice.
And if you don't like that and the varying vote power depending on how you
vote: I have a system for you!
"Instant Runoff Normalized Ratings" (IRNR)
Each ballot is normalized so that all ballots have the same magnitude.
pfffft! �way too complicated.
The
modified ballots are summed, and the choice with the lowest sumarry rating
is disqualified. Each ballot is then normalized again as if the
disqualified choice was not there, redistributing the vote across the
choices in proportion to the original ballot. The new modified ballots are
summed and the process is repeated until there are two choices remaining
and one choice wins over the other.
I think this works better with an honest ballot in the case where you like
some choice more than another 'just a little bit' or by whatever margin.
�
no reason to use that over Condorcet (with a simple method to deal with cycles, ranked-pairs is still a lot easier to explain than Schulze and they pick the same winner if there are 3 in the Smith set, so let's use the simpler method)
�
--
�
r b-j � � � � � � � � �rbj@audioimagination.com
�
"Imagination is more important than knowledge."
I also have found it hard to believe that Score voting and Approval
voting are taken seriously.
There is a tradition of cumulative voting in one of the states, that,
not surprisingly, would not have compared badly with FPTP.
An American political science association uses Approval Voting. I take
that choice in itself to be a form of strategic voting. Neither looking
too bad with FPTP, nor looking too good, in the eyes of (gerrymandering)
politicians.
Richard Lung.
On 23/06/2017 17:09, robert bristow-johnson wrote:
---------------------------- Original Message ----------------------------
Subject: Re: [EM] The election methods trade-off paradox/impossibility
theorems paradox.
From: "Brian Olson" bql@bolson.org
Date: Fri, June 23, 2017 8:35 am
To: "EM" election-methods@lists.electorama.com
I was speaking only of ballots, and and in the abstract that some
election
algorithm could take that information and make a good outcome of it.
No, we should not make the voters cook up that information. all we
should ask the voters is "whom do you prefer A or B?" and "if you
can't get your favorite, whom is your next preference?"
I don't favor raw Score summation. It's strategy prone.
of course it is. scoring is strategy prone if you use the scores in
any algorithm other than simple ranking. and then don't use scores.
just rank.
For choices where
my honest vote might be [1.0, 0.8, 0.6, 0.4, 0.2, 0.0] I should probably
vote strategically [1.0, 1.0, 1.0, 0.0, 0.0, 0.0].
but then you're not helping your first choice beat your second or
third choice.
And if you don't like that and the varying vote power depending on
how you
vote: I have a system for you!
"Instant Runoff Normalized Ratings" (IRNR)
Each ballot is normalized so that all ballots have the same magnitude.
pfffft! way too complicated.
The
modified ballots are summed, and the choice with the lowest sumarry
rating
is disqualified. Each ballot is then normalized again as if the
disqualified choice was not there, redistributing the vote across the
choices in proportion to the original ballot. The new modified
ballots are
summed and the process is repeated until there are two choices remaining
and one choice wins over the other.
I think this works better with an honest ballot in the case where
you like
some choice more than another 'just a little bit' or by whatever margin.
no reason to use that over Condorcet (with a simple method to deal
with cycles, ranked-pairs is still a lot easier to explain than
Schulze and they pick the same winner if there are 3 in the Smith set,
so let's use the simpler method)
--
r b-j rbj@audioimagination.com
"Imagination is more important than knowledge."
Election-Methods mailing list - see http://electorama.com/em for list info
--
Richard Lung.
http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:
https://plus.google.com/106191200795605365085
E-books in epub format:
https://www.smashwords.com/profile/view/democracyscience
I do agree that quantitative counts are better than qualitative counts
of "greater than". How much quantitative measurement extracted from the
voting data (rather than imposed on it) is a simple way to assess a
voting method's efficiency. Sooner or later, tho, the method has to
resort to less precisely based decisions of greater or lesser. This, by
the way, proves or shows that election results are generally
probabilistic rather than deterministic, and therefore cannot be
decisively judged by deductive standards of electoral efficiency (like
social choice theory).
My own election method (Binomial Transferable Vote) does not resort to
elimination of candidates during the count, to avoid a more or less
arbitrary plurality count (greater than) decision.
Richard Lung.
On 23/06/2017 13:35, Brian Olson wrote:
I was speaking only of ballots, and and in the abstract that
/some/ election algorithm could take that information and make a good
outcome of it.
I don't favor raw Score summation. It's strategy prone. For choices
where my honest vote might be [1.0, 0.8, 0.6, 0.4, 0.2, 0.0] I should
probably vote strategically [1.0, 1.0, 1.0, 0.0, 0.0, 0.0].
And if you don't like that and the varying vote power depending on how
you vote: I have a system for you!
"Instant Runoff Normalized Ratings" (IRNR)
Each ballot is normalized so that all ballots have the same magnitude.
The modified ballots are summed, and the choice with the lowest
sumarry rating is disqualified. Each ballot is then normalized again
as if the disqualified choice was not there, redistributing the vote
across the choices in proportion to the original ballot. The new
modified ballots are summed and the process is repeated until there
are two choices remaining and one choice wins over the other.
I think this works better with an honest ballot in the case where you
like some choice more than another 'just a little bit' or by whatever
margin.
/Brian
On Fri, Jun 23, 2017 at 1:02 AM, robert bristow-johnson
<rbj@audioimagination.com mailto:rbj@audioimagination.com> wrote:
so many discussions here are more arcane than i can grok. but i
can grok most of this.
---------------------------- Original Message
----------------------------
Subject: Re: [EM] The election methods trade-off
paradox/impossibility theorems paradox.
From: "Brian Olson" <bql@bolson.org <mailto:bql@bolson.org>>
Date: Thu, June 22, 2017 4:31 pm
To: "Richard Lung" <voting@ukscientists.com
<mailto:voting@ukscientists.com>>
Cc: "EM" <election-methods@lists.electorama.com
<mailto:election-methods@lists.electorama.com>>
--------------------------------------------------------------------------
Compared to a rankings ballot, a ratings ballot contains more
information
about a voter's preference or utility for the available choices.
I can say
A > B > C > D
or I can say with more detail
A = 1.0; B = 0.9; C = 0.1; D = 0.0
If there is more information available, it is possible for an
election
algorithm to use that information and more accurately represent
the voters
and find the greater global utility winner.
but voters can skew that information quantitatively and, if they
want to be tactical, insincerely.
how does the voter know that by rating Candidate B with 0.9 and
not lower, that he is not helping this candidate beat his
favorite, Candidate A? maybe it should be 0.8. or 0.5.
while Score voting requires too much information from the voters
(making them act as a trained expert and consider quantitatively
how candidates should be rated as if they are an Olympic ice
skating judge) and while Approval doesn't get enough information
from voters (does not differentiate preference between two
"approved" candidates), there has always been the problem of
either Score voting or Approval voting facing the voter about what
to do with their second choice (Candidate B). how much juice
should the voter give to Candidate B if they want B to beat C but
do not want B to beat A? (and for Approval, the question is,
having the same concern, shall the voter "approve" B or not?)
Score and Approval cannot answer that question in any simple
manner. but with Ranked-Choice the answer is clear.
and the other problem with Score is the "One-Person-One-Vote"
standard. if i really, really, really like Candidate A over
Candidate B and you only sorta like B over A by just a little bit,
it should not matter to what disparate degree we like our
candidates. your vote for B should weigh just as much as my vote
for A, even if my excitement for A exceeds your excitement for B.
that's what "one-person-one-vote" means.
i still just do not get why Score and Approval (in application to
governmental elections) have the following they do have.
bestest,
r b-j
If we had a specially insightful rational bayesian voting
populace we might
ask them for their confidence interval of how sure they are are
about each
choice and get more information still.
A = 1.0 (e=.3); B = 0.9 (e=.1), C = 0.1 (e=0.5), D = 0.0 (e=0.1)
And then we'd work out an election algorithm to maximize the
expected value
of the global utility over the known voter utility distributions.
and, except for making examples to show how a system breaks (like
a Proof by Contradiction in mathematics), i just cannot see how
judging the comparative value of systems with simulated
assumptions is helpful. there are not enough simulated cases to
be able to consider how any system will work under all conditions.
On Thu, Jun 22, 2017 at 2:47 PM, Richard Lung
<voting@ukscientists.com <mailto:voting@ukscientists.com>>
wrote:
Brian Olson,
Where we differ is that I do not see ranked choice as a
constraint. Single
order choice, the x-vote is the constraint on ranked voting as a
multiple-order choice.
The problem with election methods, practical and theoretical is
that they
impose constraints on the voters freedom of choice, in one way
or another,
and so are that much less true election methods.
(A minor example, the classic objections to cumulative voting
seem to
apply to some apprently modern versions or variations.)
When Condorcet and Borda, disagreed on the best way to conduct
a count of
preference voting, Laplace decided in favor of Borda. (I grant
you that
Condorcet has information value, when weighted. But I am not
well informed
on this approach and know of no convincing reason why it should
be adopted
or how you would persuade the public of that.) JFS Ross
explained that
Laplace favored Method Borda because higher preferences were
more important
and should count more. The Gregory method removes the objection
to Borda of
"later harm." This is the direction I have followed (weighted
count of
ranked choice), following on from where Meek method STV leaves off.
Richard Lung
On 22/06/2017 15:01, Brian Olson wrote:
I kinda don't accept this paradox. Just to compare the form of
a election
method paradox statement: Arrow's theorem was that given a set
of desired
properties and the constraint of rankings ballots, those set of
desirable
properties could not all be simultaneously fulfilled. One can
almost
trivially step outside of that paradox by eliminating the
constraint of the
rankings ballot.
My model of understanding people and elections is a utilitarian
one. A
person derives some amount of utility from the outcome of an
election and
everyone is apportioned the same share of utility which we
might count as
0..1 or -1..1 . These model persons can be summed up and and a
global
social utility calculated. The ideal election method perfectly
knows every
person and elects the true global social utility maximizing
candidate. This
sounds an awful lot like score voting. But then we have to start to
complicate the model with imperfect knowledge of a voter's
utility, the
imperfect expression of that on a ballot, strategic ballot
casting rather
than honest, messy computation and practical administration
issues of
running an election in the real world, and so on. So we might
wind up with
a best practical method that isn't just simple score voting.
But I still believe there is a pragmatic 'best' method, we have
techniques
for evaluating that, and we should do this and put something up
in the real
world. Personally I'll take a rankings ballot that's Condorcet
counted with
any cycle resolution method as 'good enough' and practically
applicable;
and tinkering around the edges for a slightly better method is fun
mathematical curiosity but I'd also like to get some laws passed.
What do you think of my model statement?
Is there a more formal statement of limitations you were
heading towards?
On Thu, Jun 22, 2017 at 2:30 AM, Richard Lung
<voting@ukscientists.com <mailto:voting@ukscientists.com>>
wrote:
The election methods trade-off paradox/impossibility theorems
paradox.
For the sake of argument, suppose a trade-off theory of
elections that
there is no consistently democratic electoral system: the
impossibility
supposition.
That supposition implies some conception (albeit non-existent)
of a
consistently derived right election result.
If there is no such measure, then there is no standard even to
judge that
there is a trade-off between electoral systems.
Suppose there is a consistent theory of choice, setting a
standard by
which electoral systems can be judged for their democratic
consistency.
It follows that the election result will only be as consistent
as the
electoral system, and there is no pre-conceivably right
election result,
because that presupposes a perfection not given to science as
a progressive
pursuit.
--
Richard Lung.http://www.voting.ukscientists.com
<http://www.voting.ukscientists.com>
Democracy Science series 3 free e-books in
pdf:https://plus.google.com/106191200795605365085
<https://plus.google.com/106191200795605365085>
E-books <https://plus.google.com/106191200795605365085E-books
<https://plus.google.com/106191200795605365085E-books>> in epub
format:https://www.smashwords.com/profile/view/democracyscience
<https://www.smashwords.com/profile/view/democracyscience>
--
Richard Lung.http://www.voting.ukscientists.com
<http://www.voting.ukscientists.com>
Democracy Science series 3 free e-books in
pdf:https://plus.google.com/106191200795605365085
<https://plus.google.com/106191200795605365085>
E-books in epub
format:https://www.smashwords.com/profile/view/democracyscience
<https://www.smashwords.com/profile/view/democracyscience>
--
r b-j rbj@audioimagination.com <mailto:rbj@audioimagination.com>
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for
list info
Election-Methods mailing list - see http://electorama.com/em for list info
--
Richard Lung.
http://www.voting.ukscientists.com
Democracy Science series 3 free e-books in pdf:
https://plus.google.com/106191200795605365085
E-books in epub format:
https://www.smashwords.com/profile/view/democracyscience