KM
Kristofer Munsterhjelm
Fri, Apr 12, 2024 2:51 PM
On 2024-04-12 15:58, Richard Lung wrote:
This poll, I gather from the unfamiliar sounding names, is of
single-member systems. But there need not be two classes of election
systems, single-member and multi-member. There are not two truths of
electoral system. There is (one truth) general electoral system.
It is indeed possible to have a multi-winner method that reduces to a
particular single-winner method when the number of seats is set to one.
Such a method need not even use elimination. Schulze STV doesn't use
elimination, nor do optimization-based cardinal methods like Phragmen.
-km
On 2024-04-12 15:58, Richard Lung wrote:
>
> This poll, I gather from the unfamiliar sounding names, is of
> single-member systems. But there need not be two classes of election
> systems, single-member and multi-member. There are not two truths of
> electoral system. There is (one truth) general electoral system.
It is indeed possible to have a multi-winner method that reduces to a
particular single-winner method when the number of seats is set to one.
Such a method need not even use elimination. Schulze STV doesn't use
elimination, nor do optimization-based cardinal methods like Phragmen.
-km
TP
Toby Pereira
Fri, Apr 12, 2024 3:19 PM
I will mention why I nominated Smith//Score. This method uses rated ballots, but uses them to infer a ranking. If there is a Condorcet winner, they are elected. Otherwise, elect the score winner of the Smith Set (the top cycle). I previously put why I think rated ballots work well for Condorcet (see below). And given a rated ballot, electing the highest scored candidate given no Condorcet winner seems the most simple and logical option, and shouldn't damage independence of clones or monotonicity. It also sidesteps any worry/complications over whether margins/winning votes etc. are the best thing to look at.
Toby
I think one problem of burial-resistant methods is that they assume the electorate are aware of the consequences of it and will act accordingly. I think it might be a bit optimistic to expect the average voter to behave any differently using any method that uses a specific ballot type. Using a ranked ballot, if A and B are the frontrunning candidates, then supporters of A might rank B bottom because it's the obvious thing to do (which has been pointed out on here before I believe). Do you think the adoption of a specific Condorcet method will prevent that? I'm not convinced.
Also, if there are two frontrunning candidates, A and B, it's quite likely anyway that supporters of A will see B as the worst candidate anyway, below the ones they know very little about. So it wouldn't really even be an act of burial, and therefore honest voting behaviour could cause a non-entity to win, because this is what burial-resistant methods do.
I've said this before, but possibly the best solution for a Condorcet method would to be to use rated ballots. In this case B is less likely to be buried by the A supporters, because they would be likely to score the non-entity candidates 0 as well.
On Thursday, 11 April 2024 at 08:37:54 BST, Michael Ossipoff <email9648742@gmail.com> wrote:
On Wed, Apr 10, 2024 at 23:45 Chris Benham cbenhamau@yahoo.com.au wrote:
Have the nominations closed?
Yes, at 5:14:59 GMT, April 12th.
Not that I want to nominate another method.
There has been very little electioneering, with I think most of the
nominated methods not even being mentioned.
Yes, we weren’t given any information on most of the nominees. People should have told the advantages/merits of their nominees.
Some haven't even been explained let alone discussed or promoted.
Exactly. That’s why I ranked most of them all together at the same rank-position. The ones I know about & like ranked in order of merit, then the ones that I don’t know equal-ranked, & then, below them, the ones I know that I don’t like. Pluraity was at bottom, as everyone agrees, & so there was no reason to rank it.
I will mention why I nominated Smith//Score. This method uses rated ballots, but uses them to infer a ranking. If there is a Condorcet winner, they are elected. Otherwise, elect the score winner of the Smith Set (the top cycle). I previously put why I think rated ballots work well for Condorcet (see below). And given a rated ballot, electing the highest scored candidate given no Condorcet winner seems the most simple and logical option, and shouldn't damage independence of clones or monotonicity. It also sidesteps any worry/complications over whether margins/winning votes etc. are the best thing to look at.
Toby
>I think one problem of burial-resistant methods is that they assume the electorate are aware of the consequences of it and will act accordingly. I think it might be a bit optimistic to expect the average voter to behave any differently using any method that uses a specific ballot type. Using a ranked ballot, if A and B are the frontrunning candidates, then supporters of A might rank B bottom because it's the obvious thing to do (which has been pointed out on here before I believe). Do you think the adoption of a specific Condorcet method will prevent that? I'm not convinced.
>Also, if there are two frontrunning candidates, A and B, it's quite likely anyway that supporters of A will see B as the worst candidate anyway, below the ones they know very little about. So it wouldn't really even be an act of burial, and therefore honest voting behaviour could cause a non-entity to win, because this is what burial-resistant methods do.
>I've said this before, but possibly the best solution for a Condorcet method would to be to use rated ballots. In this case B is less likely to be buried by the A supporters, because they would be likely to score the non-entity candidates 0 as well.
On Thursday, 11 April 2024 at 08:37:54 BST, Michael Ossipoff <email9648742@gmail.com> wrote:
On Wed, Apr 10, 2024 at 23:45 Chris Benham <cbenhamau@yahoo.com.au> wrote:
Have the nominations closed?
Yes, at 5:14:59 GMT, April 12th.
Not that I want to nominate another method.
There has been very little electioneering, with I think most of the
nominated methods not even being mentioned.
Yes, we weren’t given any information on most of the nominees. People should have told the advantages/merits of their nominees.
Some haven't even been explained let alone discussed or promoted.
Exactly. That’s why I ranked most of them all together at the same rank-position. The ones I know about & like ranked in order of merit, then the ones that I don’t know equal-ranked, & then, below them, the ones I know that I don’t like. Pluraity was at bottom, as everyone agrees, & so there was no reason to rank it.
RL
Richard Lung
Fri, Apr 12, 2024 5:42 PM
From the K.M. quote, I gather that the answer was "yes," single member
systems. Perhaps Schulze STV was on the list. From Wikipedia it is said
to be similar to CPO-STV which uses Condorcet pairing. This however does
seem to be an elimination procedure of sorts.
At any rate conventional STV, whether or not including Schulze, do not
appear to be in the poll, which hardly suggests a unitary approach to
election method. Besides still not affording a general election method.
Cardinal voting methods (the cumulative voting family) confound the vote
(of individual choice) with the count (of community choice.
Regards,
Richard Lung.
On 12/04/2024 15:51, Kristofer Munsterhjelm wrote:
On 2024-04-12 15:58, Richard Lung wrote:
This poll, I gather from the unfamiliar sounding names, is of
single-member systems. But there need not be two classes of election
systems, single-member and multi-member. There are not two truths of
electoral system. There is (one truth) general electoral system.
It is indeed possible to have a multi-winner method that reduces to a
particular single-winner method when the number of seats is set to one.
Such a method need not even use elimination. Schulze STV doesn't use
elimination, nor do optimization-based cardinal methods like Phragmen.
-km
Election-Methods mailing list - see https://electorama.com/em for list
info
From the K.M. quote, I gather that the answer was "yes," single member
systems. Perhaps Schulze STV was on the list. From Wikipedia it is said
to be similar to CPO-STV which uses Condorcet pairing. This however does
seem to be an elimination procedure of sorts.
At any rate conventional STV, whether or not including Schulze, do not
appear to be in the poll, which hardly suggests a unitary approach to
election method. Besides still not affording a general election method.
Cardinal voting methods (the cumulative voting family) confound the vote
(of individual choice) with the count (of community choice.
Regards,
Richard Lung.
On 12/04/2024 15:51, Kristofer Munsterhjelm wrote:
> On 2024-04-12 15:58, Richard Lung wrote:
>>
>> This poll, I gather from the unfamiliar sounding names, is of
>> single-member systems. But there need not be two classes of election
>> systems, single-member and multi-member. There are not two truths of
>> electoral system. There is (one truth) general electoral system.
>
> It is indeed possible to have a multi-winner method that reduces to a
> particular single-winner method when the number of seats is set to one.
>
> Such a method need not even use elimination. Schulze STV doesn't use
> elimination, nor do optimization-based cardinal methods like Phragmen.
>
> -km
> ----
> Election-Methods mailing list - see https://electorama.com/em for list
> info
RL
Richard Lung
Fri, Apr 12, 2024 5:46 PM
"there’s no one universal set of principles that govern the perfect
electoral system."
Whoever said there was.
On 12/04/2024 15:03, Michael Garman wrote:
“Truth” is perhaps not the appropriate word here — there’s no one
universal set of principles that govern the perfect electoral system.
On Fri, Apr 12, 2024 at 3:58 PM Richard Lung voting@ukscientists.com
wrote:
This poll, I gather from the unfamiliar sounding names, is of
single-member systems. But there need not be two classes of
election systems, single-member and multi-member. There are not
two truths of electoral system. There is (one truth) general
electoral system
It has been believed that the lack of proportional representation
in single-member systems set them apart. But this is not
necessarily the case. It is possible to have a general election
system that works the same way for single-member systems as
multi-member systems. Indeed (pardon me) I have invented it! The
way it is done is to hold both an election count and an exclusion
count, based on keep values, and average them, whether in a single
or multi-member system.
This redeems single-member systems for a rational count, as in
multi-member systems. Without this, single-member systems, unlike
multi-member systems, are stuck at a sub-rational level of
accuracy, and are basically (“last past the post”) elimination counts.
Regards,
Richard Lung.
On 11/04/2024 07:45, Chris Benham wrote:
Have the nominations closed? Not that I want to nominate
another method.
There has been very little electioneering, with I think most of
the nominated methods not even being mentioned.
Some haven't even been explained let alone discussed or promoted.
When does voting close?
Chris Benham
----
Election-Methods mailing list - see https://electorama.com/em for
list info
----
Election-Methods mailing list - see https://electorama.com/em for
list info
"there’s no one universal set of principles that govern the perfect
electoral system."
Whoever said there was.
On 12/04/2024 15:03, Michael Garman wrote:
> “Truth” is perhaps not the appropriate word here — there’s no one
> universal set of principles that govern the perfect electoral system.
>
> On Fri, Apr 12, 2024 at 3:58 PM Richard Lung <voting@ukscientists.com>
> wrote:
>
>
> This poll, I gather from the unfamiliar sounding names, is of
> single-member systems. But there need not be two classes of
> election systems, single-member and multi-member. There are not
> two truths of electoral system. There is (one truth) general
> electoral system
>
> It has been believed that the lack of proportional representation
> in single-member systems set them apart. But this is not
> necessarily the case. It is possible to have a general election
> system that works the same way for single-member systems as
> multi-member systems. Indeed (pardon me) I have invented it! The
> way it is done is to hold both an election count and an exclusion
> count, based on keep values, and average them, whether in a single
> or multi-member system.
>
> This redeems single-member systems for a rational count, as in
> multi-member systems. Without this, single-member systems, unlike
> multi-member systems, are stuck at a sub-rational level of
> accuracy, and are basically (“last past the post”) elimination counts.
>
> Regards,
>
> Richard Lung.
>
>
>
> On 11/04/2024 07:45, Chris Benham wrote:
>>
>> Have the nominations closed? Not that I want to nominate
>> another method.
>>
>> There has been very little electioneering, with I think most of
>> the nominated methods not even being mentioned.
>>
>> Some haven't even been explained let alone discussed or promoted.
>>
>> When does voting close?
>>
>> Chris Benham
>>
>> ----
>> Election-Methods mailing list - see https://electorama.com/em for
>> list info
> ----
> Election-Methods mailing list - see https://electorama.com/em for
> list info
>
MG
Michael Garman
Fri, Apr 12, 2024 5:47 PM
That’s what you’re implying when you describe it in terms of “truth.”
On Fri, Apr 12, 2024 at 7:46 PM Richard Lung voting@ukscientists.com
wrote:
"there’s no one universal set of principles that govern the perfect
electoral system."
Whoever said there was.
On 12/04/2024 15:03, Michael Garman wrote:
“Truth” is perhaps not the appropriate word here — there’s no one
universal set of principles that govern the perfect electoral system.
On Fri, Apr 12, 2024 at 3:58 PM Richard Lung voting@ukscientists.com
wrote:
This poll, I gather from the unfamiliar sounding names, is of
single-member systems. But there need not be two classes of election
systems, single-member and multi-member. There are not two truths of
electoral system. There is (one truth) general electoral system
It has been believed that the lack of proportional representation in
single-member systems set them apart. But this is not necessarily the case.
It is possible to have a general election system that works the same way
for single-member systems as multi-member systems. Indeed (pardon me) I
have invented it! The way it is done is to hold both an election count and
an exclusion count, based on keep values, and average them, whether in a
single or multi-member system.
This redeems single-member systems for a rational count, as in
multi-member systems. Without this, single-member systems, unlike
multi-member systems, are stuck at a sub-rational level of accuracy, and
are basically (“last past the post”) elimination counts.
Regards,
Richard Lung.
On 11/04/2024 07:45, Chris Benham wrote:
Have the nominations closed? Not that I want to nominate another
method.
There has been very little electioneering, with I think most of the
nominated methods not even being mentioned.
Some haven't even been explained let alone discussed or promoted.
When does voting close?
Chris Benham
Election-Methods mailing list - see https://electorama.com/em for list
info
Election-Methods mailing list - see https://electorama.com/em for list
info
That’s what you’re implying when you describe it in terms of “truth.”
On Fri, Apr 12, 2024 at 7:46 PM Richard Lung <voting@ukscientists.com>
wrote:
>
> "there’s no one universal set of principles that govern the perfect
> electoral system."
>
> Whoever said there was.
>
>
> On 12/04/2024 15:03, Michael Garman wrote:
>
> “Truth” is perhaps not the appropriate word here — there’s no one
> universal set of principles that govern the perfect electoral system.
>
> On Fri, Apr 12, 2024 at 3:58 PM Richard Lung <voting@ukscientists.com>
> wrote:
>
>>
>> This poll, I gather from the unfamiliar sounding names, is of
>> single-member systems. But there need not be two classes of election
>> systems, single-member and multi-member. There are not two truths of
>> electoral system. There is (one truth) general electoral system
>>
>> It has been believed that the lack of proportional representation in
>> single-member systems set them apart. But this is not necessarily the case.
>> It is possible to have a general election system that works the same way
>> for single-member systems as multi-member systems. Indeed (pardon me) I
>> have invented it! The way it is done is to hold both an election count and
>> an exclusion count, based on keep values, and average them, whether in a
>> single or multi-member system.
>>
>> This redeems single-member systems for a rational count, as in
>> multi-member systems. Without this, single-member systems, unlike
>> multi-member systems, are stuck at a sub-rational level of accuracy, and
>> are basically (“last past the post”) elimination counts.
>>
>> Regards,
>>
>> Richard Lung.
>>
>>
>>
>> On 11/04/2024 07:45, Chris Benham wrote:
>>
>>
>> Have the nominations closed? Not that I want to nominate another
>> method.
>>
>>
>> There has been very little electioneering, with I think most of the
>> nominated methods not even being mentioned.
>>
>> Some haven't even been explained let alone discussed or promoted.
>>
>> When does voting close?
>>
>> Chris Benham
>>
>> ----
>> Election-Methods mailing list - see https://electorama.com/em for list
>> info
>>
>> ----
>> Election-Methods mailing list - see https://electorama.com/em for list
>> info
>>
>
KM
Kristofer Munsterhjelm
Fri, Apr 12, 2024 10:50 PM
On 2024-04-12 19:42, Richard Lung wrote:
From the K.M. quote, I gather that the answer was "yes," single member
systems. Perhaps Schulze STV was on the list. From Wikipedia it is said
to be similar to CPO-STV which uses Condorcet pairing. This however does
seem to be an elimination procedure of sorts.
Yes, the current poll is about single-member systems.
On your other point: while CPO-STV is an elimination method, Schulze
STV is not. Schulze STV works by comparing outcomes (ways to elect the
required number of winners) that differ by one candidate. It then
generalizes the Schulze method to find the best outcome based on these
comparisons. I'm not quite sure how the comparison function works: it
uses linear optimization. But there's no sequential elimination of
candidates.
At any rate conventional STV, whether or not including Schulze, do not
appear to be in the poll, which hardly suggests a unitary approach to
election method. Besides still not affording a general election method.
I'm not saying that the poll is about unified methods, just that if
(say) Schulze wins the poll, then that doesn't immediately preclude one
from considering multi-winner methods since anybody who thinks Schulze
is the best single-winner method could view it as just a special case of
Schulze STV.
Cardinal voting methods (the cumulative voting family) confound the vote
(of individual choice) with the count (of community choice.
I don't know what that means, unfortunately. You're using terminology
that I don't understand.
-km
On 2024-04-12 19:42, Richard Lung wrote:
>
> From the K.M. quote, I gather that the answer was "yes," single member
> systems. Perhaps Schulze STV was on the list. From Wikipedia it is said
> to be similar to CPO-STV which uses Condorcet pairing. This however does
> seem to be an elimination procedure of sorts.
Yes, the current poll is about single-member systems.
On your other point: while CPO-STV *is* an elimination method, Schulze
STV is not. Schulze STV works by comparing outcomes (ways to elect the
required number of winners) that differ by one candidate. It then
generalizes the Schulze method to find the best outcome based on these
comparisons. I'm not quite sure how the comparison function works: it
uses linear optimization. But there's no sequential elimination of
candidates.
> At any rate conventional STV, whether or not including Schulze, do not
> appear to be in the poll, which hardly suggests a unitary approach to
> election method. Besides still not affording a general election method.
I'm not saying that the poll is about unified methods, just that if
(say) Schulze wins the poll, then that doesn't immediately preclude one
from considering multi-winner methods since anybody who thinks Schulze
is the best single-winner method could view it as just a special case of
Schulze STV.
> Cardinal voting methods (the cumulative voting family) confound the vote
> (of individual choice) with the count (of community choice.
I don't know what that means, unfortunately. You're using terminology
that I don't understand.
-km
CB
Chris Benham
Sat, Apr 13, 2024 2:13 AM
Toby,
What range of scores do you have in mind? Proposals have varied between
0-2 (i.e. 3 rating slots) and 0-99 (100 rating slots).
I think your suggestion has merit as a response to Score fans that boast
of Score's "high Condorcet efficiency". You are in effect responding:
"Well if that is such a good thing, why not make the method a little bit
more complicated and raise that efficiency to 100% ?"
I agree that this much better than SCORE or STAR. But I don't like
deciding things by just adding up (or averaging) raw Score scores, even
just
within the Smith set, because of the same reason I prefer Approval to
Score. Naive sincere voters are unfairly disadvantaged compared to
strategic
exaggerators.
Here is a variant I very much like but didn't nominate it because I
didn't think it was "propose-able" enough in say the US.
Voters score the candidates on a range large enough to strictly rank
all the candidates plus leave several large-ish gaps. Default score is zero.
Eliminate all the candidates not in the Smith set.
Interpreting ballots as approving remaining candidates they score above
average (of the scores they give to remaining candidates), elect
the most approved member of the Smith set.
So this is like a "Declared Strategy" method that simulates: first use
rankings to identify the members of the Smith set and then hold an Approval
election among those candidates.
What do you think?
Chris
On 13/04/2024 12:49 am, Toby Pereira wrote:
I will mention why I nominated Smith//Score. This method uses rated
ballots, but uses them to infer a ranking. If there is a Condorcet
winner, they are elected. Otherwise, elect the score winner of the
Smith Set (the top cycle). I previously put why I think rated ballots
work well for Condorcet (see below). And given a rated ballot,
electing the highest scored candidate given no Condorcet winner seems
the most simple and logical option, and shouldn't damage independence
of clones or monotonicity. It also sidesteps any worry/complications
over whether margins/winning votes etc. are the best thing to look at.
Toby
I think one problem of burial-resistant methods is that they assume
the electorate are aware of the consequences of it and will act
accordingly. I think it might be a bit optimistic to expect the
average voter to behave any differently using any method that uses a
specific ballot type. Using a ranked ballot, if A and B are the
frontrunning candidates, then supporters of A might rank B bottom
because it's the obvious thing to do (which has been pointed out on
here before I believe). Do you think the adoption of a specific
Condorcet method will prevent that? I'm not convinced.
Also, if there are two frontrunning candidates, A and B, it's quite
likely anyway that supporters of A will see B as the worst candidate
anyway, below the ones they know very little about. So it wouldn't
really even be an act of burial, and therefore honest voting behaviour
could cause a non-entity to win, because this is what burial-resistant
methods do.
I've said this before, but possibly the best solution for a Condorcet
method would to be to use rated ballots. In this case B is less likely
to be buried by the A supporters, because they would be likely to
score the non-entity candidates 0 as well.
On Thursday, 11 April 2024 at 08:37:54 BST, Michael Ossipoff
email9648742@gmail.com wrote:
On Wed, Apr 10, 2024 at 23:45 Chris Benham cbenhamau@yahoo.com.au wrote:
Have the nominations closed?
Yes, at 5:14:59 GMT, April 12th.
Not that I want to nominate another method.
There has been very little electioneering, with I think most of the
nominated methods not even being mentioned.
Yes, we weren’t given any information on most of the nominees. People
should have told the advantages/merits of their nominees.
Some haven't even been explained let alone discussed or promoted.
Exactly. That’s why I ranked most of them all together at the same
rank-position. The ones I know about & like ranked in order of merit,
then the ones that I don’t know equal-ranked, & then, below them, the
ones I know that I don’t like. Pluraity was at bottom, as everyone
agrees, & so there was no reason to rank it.
Toby,
What range of scores do you have in mind? Proposals have varied between
0-2 (i.e. 3 rating slots) and 0-99 (100 rating slots).
I think your suggestion has merit as a response to Score fans that boast
of Score's "high Condorcet efficiency". You are in effect responding:
"Well if that is such a good thing, why not make the method a little bit
more complicated and raise that efficiency to 100% ?"
I agree that this much better than SCORE or STAR. But I don't like
deciding things by just adding up (or averaging) raw Score scores, even
just
within the Smith set, because of the same reason I prefer Approval to
Score. Naive sincere voters are unfairly disadvantaged compared to
strategic
exaggerators.
Here is a variant I very much like but didn't nominate it because I
didn't think it was "propose-able" enough in say the US.
*Voters score the candidates on a range large enough to strictly rank
all the candidates plus leave several large-ish gaps. Default score is zero.
Eliminate all the candidates not in the Smith set.
Interpreting ballots as approving remaining candidates they score above
average (of the scores they give to remaining candidates), elect
the most approved member of the Smith set*.
So this is like a "Declared Strategy" method that simulates: first use
rankings to identify the members of the Smith set and then hold an Approval
election among those candidates.
What do you think?
Chris
On 13/04/2024 12:49 am, Toby Pereira wrote:
> I will mention why I nominated Smith//Score. This method uses rated
> ballots, but uses them to infer a ranking. If there is a Condorcet
> winner, they are elected. Otherwise, elect the score winner of the
> Smith Set (the top cycle). I previously put why I think rated ballots
> work well for Condorcet (see below). And given a rated ballot,
> electing the highest scored candidate given no Condorcet winner seems
> the most simple and logical option, and shouldn't damage independence
> of clones or monotonicity. It also sidesteps any worry/complications
> over whether margins/winning votes etc. are the best thing to look at.
>
> Toby
>
>
> >I think one problem of burial-resistant methods is that they assume
> the electorate are aware of the consequences of it and will act
> accordingly. I think it might be a bit optimistic to expect the
> average voter to behave any differently using any method that uses a
> specific ballot type. Using a ranked ballot, if A and B are the
> frontrunning candidates, then supporters of A might rank B bottom
> because it's the obvious thing to do (which has been pointed out on
> here before I believe). Do you think the adoption of a specific
> Condorcet method will prevent that? I'm not convinced.
>
> >Also, if there are two frontrunning candidates, A and B, it's quite
> likely anyway that supporters of A will see B as the worst candidate
> anyway, below the ones they know very little about. So it wouldn't
> really even be an act of burial, and therefore honest voting behaviour
> could cause a non-entity to win, because this is what burial-resistant
> methods do.
>
> >I've said this before, but possibly the best solution for a Condorcet
> method would to be to use rated ballots. In this case B is less likely
> to be buried by the A supporters, because they would be likely to
> score the non-entity candidates 0 as well.
>
>
> On Thursday, 11 April 2024 at 08:37:54 BST, Michael Ossipoff
> <email9648742@gmail.com> wrote:
>
>
>
>
> On Wed, Apr 10, 2024 at 23:45 Chris Benham <cbenhamau@yahoo.com.au> wrote:
>
>
> Have the nominations closed?
>
>
> Yes, at 5:14:59 GMT, April 12th.
>
>
> Not that I want to nominate another method.
>
> There has been very little electioneering, with I think most of the
> nominated methods not even being mentioned.
>
>
> Yes, we weren’t given any information on most of the nominees. People
> should have told the advantages/merits of their nominees.
>
>
> Some haven't even been explained let alone discussed or promoted.
>
>
> Exactly. That’s why I ranked most of them all together at the same
> rank-position. The ones I know about & like ranked in order of merit,
> then the ones that I don’t know equal-ranked, & then, below them, the
> ones I know that I don’t like. Pluraity was at bottom, as everyone
> agrees, & so there was no reason to rank it.
>
>
TP
Toby Pereira
Sat, Apr 13, 2024 12:15 PM
I was thinking at least 0 to 9, although 0 to 99 would be better in terms of being able to distinguish between all candidates. Most approved among the Smith Set definitely has merit, but it's a question of how you would determine whether a candidate is approved. I'm not sure an above average score seems like the right measure. I think I'd prefer something more explicit. Voters could even enter the score that every candidate scoring that score or above counts as approved. If left blank, maybe highest scored candidates only - I know you seem to have moved in that direction for candidates implicitly approved on ranked ballots.
Toby
On Saturday, 13 April 2024 at 03:13:14 BST, Chris Benham cbenhamau@yahoo.com.au wrote:
Toby,
What range of scores do you have in mind? Proposals have varied between 0-2 (i.e. 3 rating slots) and 0-99 (100 rating slots).
I think your suggestion has merit as a response to Score fans that boast of Score's "high Condorcet efficiency". You are in effect responding:
"Well if that is such a good thing, why not make the method a little bit more complicated and raise that efficiency to 100% ?"
I agree that this much better than SCORE or STAR. But I don't like deciding things by just adding up (or averaging) raw Score scores, even just
within the Smith set, because of the same reason I prefer Approval to Score. Naive sincere voters are unfairly disadvantaged compared to strategic
exaggerators.
Here is a variant I very much like but didn't nominate it because I didn't think it was "propose-able" enough in say the US.
Voters score the candidates on a range large enough to strictly rank all the candidates plus leave several large-ish gaps. Default score is zero.
Eliminate all the candidates not in the Smith set.
Interpreting ballots as approving remaining candidates they score above average (of the scores they give to remaining candidates), elect
the most approved member of the Smith set.
So this is like a "Declared Strategy" method that simulates: first use rankings to identify the members of the Smith set and then hold an Approval
election among those candidates.
What do you think?
Chris
On 13/04/2024 12:49 am, Toby Pereira wrote:
I will mention why I nominated Smith//Score. This method uses rated ballots, but uses them to infer a ranking. If there is a Condorcet winner, they are elected. Otherwise, elect the score winner of the Smith Set (the top cycle). I previously put why I think rated ballots work well for Condorcet (see below). And given a rated ballot, electing the highest scored candidate given no Condorcet winner seems the most simple and logical option, and shouldn't damage independence of clones or monotonicity. It also sidesteps any worry/complications over whether margins/winning votes etc. are the best thing to look at.
Toby
I think one problem of burial-resistant methods is that they assume the electorate are aware of the consequences of it and will act accordingly. I think it might be a bit optimistic to expect the average voter to behave any differently using any method that uses a specific ballot type. Using a ranked ballot, if A and B are the frontrunning candidates, then supporters of A might rank B bottom because it's the obvious thing to do (which has been pointed out on here before I believe). Do you think the adoption of a specific Condorcet method will prevent that? I'm not convinced.
Also, if there are two frontrunning candidates, A and B, it's quite likely anyway that supporters of A will see B as the worst candidate anyway, below the ones they know very little about. So it wouldn't really even be an act of burial, and therefore honest voting behaviour could cause a non-entity to win, because this is what burial-resistant methods do.
I've said this before, but possibly the best solution for a Condorcet method would to be to use rated ballots. In this case B is less likely to be buried by the A supporters, because they would be likely to score the non-entity candidates 0 as well.
On Thursday, 11 April 2024 at 08:37:54 BST, Michael Ossipoff <email9648742@gmail.com> wrote:
On Wed, Apr 10, 2024 at 23:45 Chris Benham cbenhamau@yahoo.com.au wrote:
Have the nominations closed?
Yes, at 5:14:59 GMT, April 12th.
Not that I want to nominate another method.
There has been very little electioneering, with I think most of the
nominated methods not even being mentioned.
Yes, we weren’t given any information on most of the nominees. People should have told the advantages/merits of their nominees.
Some haven't even been explained let alone discussed or promoted.
Exactly. That’s why I ranked most of them all together at the same rank-position. The ones I know about & like ranked in order of merit, then the ones that I don’t know equal-ranked, & then, below them, the ones I know that I don’t like. Pluraity was at bottom, as everyone agrees, & so there was no reason to rank it.
I was thinking at least 0 to 9, although 0 to 99 would be better in terms of being able to distinguish between all candidates. Most approved among the Smith Set definitely has merit, but it's a question of how you would determine whether a candidate is approved. I'm not sure an above average score seems like the right measure. I think I'd prefer something more explicit. Voters could even enter the score that every candidate scoring that score or above counts as approved. If left blank, maybe highest scored candidates only - I know you seem to have moved in that direction for candidates implicitly approved on ranked ballots.
Toby
On Saturday, 13 April 2024 at 03:13:14 BST, Chris Benham <cbenhamau@yahoo.com.au> wrote:
Toby,
What range of scores do you have in mind? Proposals have varied between 0-2 (i.e. 3 rating slots) and 0-99 (100 rating slots).
I think your suggestion has merit as a response to Score fans that boast of Score's "high Condorcet efficiency". You are in effect responding:
"Well if that is such a good thing, why not make the method a little bit more complicated and raise that efficiency to 100% ?"
I agree that this much better than SCORE or STAR. But I don't like deciding things by just adding up (or averaging) raw Score scores, even just
within the Smith set, because of the same reason I prefer Approval to Score. Naive sincere voters are unfairly disadvantaged compared to strategic
exaggerators.
Here is a variant I very much like but didn't nominate it because I didn't think it was "propose-able" enough in say the US.
*Voters score the candidates on a range large enough to strictly rank all the candidates plus leave several large-ish gaps. Default score is zero.
Eliminate all the candidates not in the Smith set.
Interpreting ballots as approving remaining candidates they score above average (of the scores they give to remaining candidates), elect
the most approved member of the Smith set*.
So this is like a "Declared Strategy" method that simulates: first use rankings to identify the members of the Smith set and then hold an Approval
election among those candidates.
What do you think?
Chris
On 13/04/2024 12:49 am, Toby Pereira wrote:
I will mention why I nominated Smith//Score. This method uses rated ballots, but uses them to infer a ranking. If there is a Condorcet winner, they are elected. Otherwise, elect the score winner of the Smith Set (the top cycle). I previously put why I think rated ballots work well for Condorcet (see below). And given a rated ballot, electing the highest scored candidate given no Condorcet winner seems the most simple and logical option, and shouldn't damage independence of clones or monotonicity. It also sidesteps any worry/complications over whether margins/winning votes etc. are the best thing to look at.
Toby
>I think one problem of burial-resistant methods is that they assume the electorate are aware of the consequences of it and will act accordingly. I think it might be a bit optimistic to expect the average voter to behave any differently using any method that uses a specific ballot type. Using a ranked ballot, if A and B are the frontrunning candidates, then supporters of A might rank B bottom because it's the obvious thing to do (which has been pointed out on here before I believe). Do you think the adoption of a specific Condorcet method will prevent that? I'm not convinced.
>Also, if there are two frontrunning candidates, A and B, it's quite likely anyway that supporters of A will see B as the worst candidate anyway, below the ones they know very little about. So it wouldn't really even be an act of burial, and therefore honest voting behaviour could cause a non-entity to win, because this is what burial-resistant methods do.
>I've said this before, but possibly the best solution for a Condorcet method would to be to use rated ballots. In this case B is less likely to be buried by the A supporters, because they would be likely to score the non-entity candidates 0 as well.
On Thursday, 11 April 2024 at 08:37:54 BST, Michael Ossipoff <email9648742@gmail.com> wrote:
On Wed, Apr 10, 2024 at 23:45 Chris Benham <cbenhamau@yahoo.com.au> wrote:
Have the nominations closed?
Yes, at 5:14:59 GMT, April 12th.
Not that I want to nominate another method.
There has been very little electioneering, with I think most of the
nominated methods not even being mentioned.
Yes, we weren’t given any information on most of the nominees. People should have told the advantages/merits of their nominees.
Some haven't even been explained let alone discussed or promoted.
Exactly. That’s why I ranked most of them all together at the same rank-position. The ones I know about & like ranked in order of merit, then the ones that I don’t know equal-ranked, & then, below them, the ones I know that I don’t like. Pluraity was at bottom, as everyone agrees, & so there was no reason to rank it.
CB
Chris Benham
Sat, Apr 13, 2024 3:22 PM
Voters could even enter the score that every candidate scoring that
score or above counts as approved.
If the scores are only used to imply ratings, with the voters manually
entering their approval cutoffs, then what is the point of them versus
simple ranking??
My suggestion perfectly simulates the voters using rational zero-info
approval strategy among the Smith set candidates. What is wrong with that?
With the approvals manually entered and fixed as you suggest, you are
doing nothing to address the unfairness of voters who happen to make
some approval
distinction among the candidates who make it into the Smith set having
more influence on the result than those that don't.
One crude and simple solution is to have a grading or score ballot with
enough slots to accommodate twice the number of candidates, and the top
half of them
are labelled "Approved" and the bottom half labelled "Not Approved" and
the scores given interpreted accordingly.
These ballots could be used for Smith//Approval or Margins Sorted Approval.
Another idea that has been suggested it to have the approval cutoff
represented by a virtual candidate on a ranked ballot. Ballots are
interpreting as approving
the candidates they rank above this "candidate". I don't mind that.
Chris
On 13/04/2024 9:45 pm, Toby Pereira wrote:
I was thinking at least 0 to 9, although 0 to 99 would be better in
terms of being able to distinguish between all candidates. Most
approved among the Smith Set definitely has merit, but it's a question
of how you would determine whether a candidate is approved. I'm not
sure an above average score seems like the right measure. I think I'd
prefer something more explicit. Voters could even enter the score that
every candidate scoring that score or above counts as approved. If
left blank, maybe highest scored candidates only - I know you seem to
have moved in that direction for candidates implicitly approved on
ranked ballots.
Toby
On Saturday, 13 April 2024 at 03:13:14 BST, Chris Benham
cbenhamau@yahoo.com.au wrote:
Toby,
What range of scores do you have in mind? Proposals have varied
between 0-2 (i.e. 3 rating slots) and 0-99 (100 rating slots).
I think your suggestion has merit as a response to Score fans that
boast of Score's "high Condorcet efficiency". You are in effect
responding:
"Well if that is such a good thing, why not make the method a little
bit more complicated and raise that efficiency to 100% ?"
I agree that this much better than SCORE or STAR. But I don't like
deciding things by just adding up (or averaging) raw Score scores,
even just
within the Smith set, because of the same reason I prefer Approval to
Score. Naive sincere voters are unfairly disadvantaged compared to
strategic
exaggerators.
Here is a variant I very much like but didn't nominate it because I
didn't think it was "propose-able" enough in say the US.
Voters score the candidates on a range large enough to strictly rank
all the candidates plus leave several large-ish gaps. Default score is
zero.
Eliminate all the candidates not in the Smith set.
Interpreting ballots as approving remaining candidates they score
above average (of the scores they give to remaining candidates), elect
the most approved member of the Smith set.
So this is like a "Declared Strategy" method that simulates: first use
rankings to identify the members of the Smith set and then hold an
Approval
election among those candidates.
What do you think?
Chris
On 13/04/2024 12:49 am, Toby Pereira wrote:
I will mention why I nominated Smith//Score. This method uses rated
ballots, but uses them to infer a ranking. If there is a Condorcet
winner, they are elected. Otherwise, elect the score winner of the
Smith Set (the top cycle). I previously put why I think rated ballots
work well for Condorcet (see below). And given a rated ballot,
electing the highest scored candidate given no Condorcet winner seems
the most simple and logical option, and shouldn't damage independence
of clones or monotonicity. It also sidesteps any worry/complications
over whether margins/winning votes etc. are the best thing to look at.
Toby
I think one problem of burial-resistant methods is that they assume
the electorate are aware of the consequences of it and will act
accordingly. I think it might be a bit optimistic to expect the
average voter to behave any differently using any method that uses a
specific ballot type. Using a ranked ballot, if A and B are the
frontrunning candidates, then supporters of A might rank B bottom
because it's the obvious thing to do (which has been pointed out on
here before I believe). Do you think the adoption of a specific
Condorcet method will prevent that? I'm not convinced.
Also, if there are two frontrunning candidates, A and B, it's quite
likely anyway that supporters of A will see B as the worst candidate
anyway, below the ones they know very little about. So it wouldn't
really even be an act of burial, and therefore honest voting behaviour
could cause a non-entity to win, because this is what burial-resistant
methods do.
I've said this before, but possibly the best solution for a Condorcet
method would to be to use rated ballots. In this case B is less likely
to be buried by the A supporters, because they would be likely to
score the non-entity candidates 0 as well.
On Thursday, 11 April 2024 at 08:37:54 BST, Michael Ossipoff
email9648742@gmail.com mailto:email9648742@gmail.com wrote:
On Wed, Apr 10, 2024 at 23:45 Chris Benham <cbenhamau@yahoo.com.au
mailto:cbenhamau@yahoo.com.au> wrote:
Have the nominations closed?
Yes, at 5:14:59 GMT, April 12th.
Not that I want to nominate another method.
There has been very little electioneering, with I think most of the
nominated methods not even being mentioned.
Yes, we weren’t given any information on most of the nominees. People
should have told the advantages/merits of their nominees.
Some haven't even been explained let alone discussed or promoted.
Exactly. That’s why I ranked most of them all together at the same
rank-position. The ones I know about & like ranked in order of merit,
then the ones that I don’t know equal-ranked, & then, below them, the
ones I know that I don’t like. Pluraity was at bottom, as everyone
agrees, & so there was no reason to rank it.
Toby,
> Voters could even enter the score that every candidate scoring that
> score or above counts as approved.
If the scores are only used to imply ratings, with the voters manually
entering their approval cutoffs, then what is the point of them versus
simple ranking??
My suggestion perfectly simulates the voters using rational zero-info
approval strategy among the Smith set candidates. What is wrong with that?
With the approvals manually entered and fixed as you suggest, you are
doing nothing to address the unfairness of voters who happen to make
some approval
distinction among the candidates who make it into the Smith set having
more influence on the result than those that don't.
One crude and simple solution is to have a grading or score ballot with
enough slots to accommodate twice the number of candidates, and the top
half of them
are labelled "Approved" and the bottom half labelled "Not Approved" and
the scores given interpreted accordingly.
These ballots could be used for Smith//Approval or Margins Sorted Approval.
Another idea that has been suggested it to have the approval cutoff
represented by a virtual candidate on a ranked ballot. Ballots are
interpreting as approving
the candidates they rank above this "candidate". I don't mind that.
Chris
On 13/04/2024 9:45 pm, Toby Pereira wrote:
> I was thinking at least 0 to 9, although 0 to 99 would be better in
> terms of being able to distinguish between all candidates. Most
> approved among the Smith Set definitely has merit, but it's a question
> of how you would determine whether a candidate is approved. I'm not
> sure an above average score seems like the right measure. I think I'd
> prefer something more explicit. Voters could even enter the score that
> every candidate scoring that score or above counts as approved. If
> left blank, maybe highest scored candidates only - I know you seem to
> have moved in that direction for candidates implicitly approved on
> ranked ballots.
>
> Toby
>
> On Saturday, 13 April 2024 at 03:13:14 BST, Chris Benham
> <cbenhamau@yahoo.com.au> wrote:
>
>
> Toby,
>
> What range of scores do you have in mind? Proposals have varied
> between 0-2 (i.e. 3 rating slots) and 0-99 (100 rating slots).
>
> I think your suggestion has merit as a response to Score fans that
> boast of Score's "high Condorcet efficiency". You are in effect
> responding:
>
> "Well if that is such a good thing, why not make the method a little
> bit more complicated and raise that efficiency to 100% ?"
>
> I agree that this much better than SCORE or STAR. But I don't like
> deciding things by just adding up (or averaging) raw Score scores,
> even just
> within the Smith set, because of the same reason I prefer Approval to
> Score. Naive sincere voters are unfairly disadvantaged compared to
> strategic
> exaggerators.
>
> Here is a variant I very much like but didn't nominate it because I
> didn't think it was "propose-able" enough in say the US.
>
> *Voters score the candidates on a range large enough to strictly rank
> all the candidates plus leave several large-ish gaps. Default score is
> zero.
> Eliminate all the candidates not in the Smith set.
> Interpreting ballots as approving remaining candidates they score
> above average (of the scores they give to remaining candidates), elect
> the most approved member of the Smith set*.
>
> So this is like a "Declared Strategy" method that simulates: first use
> rankings to identify the members of the Smith set and then hold an
> Approval
> election among those candidates.
>
> What do you think?
>
> Chris
>
> On 13/04/2024 12:49 am, Toby Pereira wrote:
> I will mention why I nominated Smith//Score. This method uses rated
> ballots, but uses them to infer a ranking. If there is a Condorcet
> winner, they are elected. Otherwise, elect the score winner of the
> Smith Set (the top cycle). I previously put why I think rated ballots
> work well for Condorcet (see below). And given a rated ballot,
> electing the highest scored candidate given no Condorcet winner seems
> the most simple and logical option, and shouldn't damage independence
> of clones or monotonicity. It also sidesteps any worry/complications
> over whether margins/winning votes etc. are the best thing to look at.
>
> Toby
>
>
> >I think one problem of burial-resistant methods is that they assume
> the electorate are aware of the consequences of it and will act
> accordingly. I think it might be a bit optimistic to expect the
> average voter to behave any differently using any method that uses a
> specific ballot type. Using a ranked ballot, if A and B are the
> frontrunning candidates, then supporters of A might rank B bottom
> because it's the obvious thing to do (which has been pointed out on
> here before I believe). Do you think the adoption of a specific
> Condorcet method will prevent that? I'm not convinced.
>
> >Also, if there are two frontrunning candidates, A and B, it's quite
> likely anyway that supporters of A will see B as the worst candidate
> anyway, below the ones they know very little about. So it wouldn't
> really even be an act of burial, and therefore honest voting behaviour
> could cause a non-entity to win, because this is what burial-resistant
> methods do.
>
> >I've said this before, but possibly the best solution for a Condorcet
> method would to be to use rated ballots. In this case B is less likely
> to be buried by the A supporters, because they would be likely to
> score the non-entity candidates 0 as well.
>
>
> On Thursday, 11 April 2024 at 08:37:54 BST, Michael Ossipoff
> <email9648742@gmail.com> <mailto:email9648742@gmail.com> wrote:
>
>
>
>
> On Wed, Apr 10, 2024 at 23:45 Chris Benham <cbenhamau@yahoo.com.au
> <mailto:cbenhamau@yahoo.com.au>> wrote:
>
>
> Have the nominations closed?
>
>
> Yes, at 5:14:59 GMT, April 12th.
>
>
> Not that I want to nominate another method.
>
> There has been very little electioneering, with I think most of the
> nominated methods not even being mentioned.
>
>
> Yes, we weren’t given any information on most of the nominees. People
> should have told the advantages/merits of their nominees.
>
>
> Some haven't even been explained let alone discussed or promoted.
>
>
> Exactly. That’s why I ranked most of them all together at the same
> rank-position. The ones I know about & like ranked in order of merit,
> then the ones that I don’t know equal-ranked, & then, below them, the
> ones I know that I don’t like. Pluraity was at bottom, as everyone
> agrees, & so there was no reason to rank it.
>
>
TP
Toby Pereira
Sat, Apr 13, 2024 7:26 PM
Chris
Primarily, the scores are there to determine the ranks and therefore the Condorcet winner, or the Smith Set if there's no Condorcet winner. The reason for scores is what I've said previously - I think it works as protection against burial, and also I think with a large number of candidates it's easier for the voter. With rankings, they really have to write in the ranks in numerical order, and if they accidentally miss someone out, it can mess up the whole ordering. Scores can be done in any order as long they have a general idea of the best and worst candidates.
I don't think your suggestion is too bad in theory, but it might be a bit opaque to most voters. So I'm not saying I dislike it.
Yes, based on how people approve, some will have more influence in the approval run-off under what I proposed. This would also be the same as for your proposal with double the number of ratings slots, as I understand it. There is also some merit in allowing voters to choose who they approve, even if there are some drawbacks.
Toby
On Saturday, 13 April 2024 at 16:22:10 BST, Chris Benham cbenhamau@yahoo.com.au wrote:
Toby,
Voters could even enter the score that every candidate scoring that score or above counts as approved.
If the scores are only used to imply ratings, with the voters manually entering their approval cutoffs, then what is the point of them versus simple ranking??
My suggestion perfectly simulates the voters using rational zero-info approval strategy among the Smith set candidates. What is wrong with that?
With the approvals manually entered and fixed as you suggest, you are doing nothing to address the unfairness of voters who happen to make some approval
distinction among the candidates who make it into the Smith set having more influence on the result than those that don't.
One crude and simple solution is to have a grading or score ballot with enough slots to accommodate twice the number of candidates, and the top half of them
are labelled "Approved" and the bottom half labelled "Not Approved" and the scores given interpreted accordingly.
These ballots could be used for Smith//Approval or Margins Sorted Approval.
Another idea that has been suggested it to have the approval cutoff represented by a virtual candidate on a ranked ballot. Ballots are interpreting as approving
the candidates they rank above this "candidate". I don't mind that.
Chris
On 13/04/2024 9:45 pm, Toby Pereira wrote:
I was thinking at least 0 to 9, although 0 to 99 would be better in terms of being able to distinguish between all candidates. Most approved among the Smith Set definitely has merit, but it's a question of how you would determine whether a candidate is approved. I'm not sure an above average score seems like the right measure. I think I'd prefer something more explicit. Voters could even enter the score that every candidate scoring that score or above counts as approved. If left blank, maybe highest scored candidates only - I know you seem to have moved in that direction for candidates implicitly approved on ranked ballots.
Toby
On Saturday, 13 April 2024 at 03:13:14 BST, Chris Benham cbenhamau@yahoo.com.au wrote:
Toby,
What range of scores do you have in mind? Proposals have varied between 0-2 (i.e. 3 rating slots) and 0-99 (100 rating slots).
I think your suggestion has merit as a response to Score fans that boast of Score's "high Condorcet efficiency". You are in effect responding:
"Well if that is such a good thing, why not make the method a little bit more complicated and raise that efficiency to 100% ?"
I agree that this much better than SCORE or STAR. But I don't like deciding things by just adding up (or averaging) raw Score scores, even just
within the Smith set, because of the same reason I prefer Approval to Score. Naive sincere voters are unfairly disadvantaged compared to strategic
exaggerators.
Here is a variant I very much like but didn't nominate it because I didn't think it was "propose-able" enough in say the US.
Voters score the candidates on a range large enough to strictly rank all the candidates plus leave several large-ish gaps. Default score is zero.
Eliminate all the candidates not in the Smith set.
Interpreting ballots as approving remaining candidates they score above average (of the scores they give to remaining candidates), elect
the most approved member of the Smith set.
So this is like a "Declared Strategy" method that simulates: first use rankings to identify the members of the Smith set and then hold an Approval
election among those candidates.
What do you think?
Chris
On 13/04/2024 12:49 am, Toby Pereira wrote:
I will mention why I nominated Smith//Score. This method uses rated ballots, but uses them to infer a ranking. If there is a Condorcet winner, they are elected. Otherwise, elect the score winner of the Smith Set (the top cycle). I previously put why I think rated ballots work well for Condorcet (see below). And given a rated ballot, electing the highest scored candidate given no Condorcet winner seems the most simple and logical option, and shouldn't damage independence of clones or monotonicity. It also sidesteps any worry/complications over whether margins/winning votes etc. are the best thing to look at.
Toby
I think one problem of burial-resistant methods is that they assume the electorate are aware of the consequences of it and will act accordingly. I think it might be a bit optimistic to expect the average voter to behave any differently using any method that uses a specific ballot type. Using a ranked ballot, if A and B are the frontrunning candidates, then supporters of A might rank B bottom because it's the obvious thing to do (which has been pointed out on here before I believe). Do you think the adoption of a specific Condorcet method will prevent that? I'm not convinced.
Also, if there are two frontrunning candidates, A and B, it's quite likely anyway that supporters of A will see B as the worst candidate anyway, below the ones they know very little about. So it wouldn't really even be an act of burial, and therefore honest voting behaviour could cause a non-entity to win, because this is what burial-resistant methods do.
I've said this before, but possibly the best solution for a Condorcet method would to be to use rated ballots. In this case B is less likely to be buried by the A supporters, because they would be likely to score the non-entity candidates 0 as well.
On Thursday, 11 April 2024 at 08:37:54 BST, Michael Ossipoff <email9648742@gmail.com> wrote:
On Wed, Apr 10, 2024 at 23:45 Chris Benham cbenhamau@yahoo.com.au wrote:
Have the nominations closed?
Yes, at 5:14:59 GMT, April 12th.
Not that I want to nominate another method.
There has been very little electioneering, with I think most of the
nominated methods not even being mentioned.
Yes, we weren’t given any information on most of the nominees. People should have told the advantages/merits of their nominees.
Some haven't even been explained let alone discussed or promoted.
Exactly. That’s why I ranked most of them all together at the same rank-position. The ones I know about & like ranked in order of merit, then the ones that I don’t know equal-ranked, & then, below them, the ones I know that I don’t like. Pluraity was at bottom, as everyone agrees, & so there was no reason to rank it.
Chris
Primarily, the scores are there to determine the ranks and therefore the Condorcet winner, or the Smith Set if there's no Condorcet winner. The reason for scores is what I've said previously - I think it works as protection against burial, and also I think with a large number of candidates it's easier for the voter. With rankings, they really have to write in the ranks in numerical order, and if they accidentally miss someone out, it can mess up the whole ordering. Scores can be done in any order as long they have a general idea of the best and worst candidates.
I don't think your suggestion is too bad in theory, but it might be a bit opaque to most voters. So I'm not saying I dislike it.
Yes, based on how people approve, some will have more influence in the approval run-off under what I proposed. This would also be the same as for your proposal with double the number of ratings slots, as I understand it. There is also some merit in allowing voters to choose who they approve, even if there are some drawbacks.
Toby
On Saturday, 13 April 2024 at 16:22:10 BST, Chris Benham <cbenhamau@yahoo.com.au> wrote:
Toby,
Voters could even enter the score that every candidate scoring that score or above counts as approved.
If the scores are only used to imply ratings, with the voters manually entering their approval cutoffs, then what is the point of them versus simple ranking??
My suggestion perfectly simulates the voters using rational zero-info approval strategy among the Smith set candidates. What is wrong with that?
With the approvals manually entered and fixed as you suggest, you are doing nothing to address the unfairness of voters who happen to make some approval
distinction among the candidates who make it into the Smith set having more influence on the result than those that don't.
One crude and simple solution is to have a grading or score ballot with enough slots to accommodate twice the number of candidates, and the top half of them
are labelled "Approved" and the bottom half labelled "Not Approved" and the scores given interpreted accordingly.
These ballots could be used for Smith//Approval or Margins Sorted Approval.
Another idea that has been suggested it to have the approval cutoff represented by a virtual candidate on a ranked ballot. Ballots are interpreting as approving
the candidates they rank above this "candidate". I don't mind that.
Chris
On 13/04/2024 9:45 pm, Toby Pereira wrote:
I was thinking at least 0 to 9, although 0 to 99 would be better in terms of being able to distinguish between all candidates. Most approved among the Smith Set definitely has merit, but it's a question of how you would determine whether a candidate is approved. I'm not sure an above average score seems like the right measure. I think I'd prefer something more explicit. Voters could even enter the score that every candidate scoring that score or above counts as approved. If left blank, maybe highest scored candidates only - I know you seem to have moved in that direction for candidates implicitly approved on ranked ballots.
Toby
On Saturday, 13 April 2024 at 03:13:14 BST, Chris Benham <cbenhamau@yahoo.com.au> wrote:
Toby,
What range of scores do you have in mind? Proposals have varied between 0-2 (i.e. 3 rating slots) and 0-99 (100 rating slots).
I think your suggestion has merit as a response to Score fans that boast of Score's "high Condorcet efficiency". You are in effect responding:
"Well if that is such a good thing, why not make the method a little bit more complicated and raise that efficiency to 100% ?"
I agree that this much better than SCORE or STAR. But I don't like deciding things by just adding up (or averaging) raw Score scores, even just
within the Smith set, because of the same reason I prefer Approval to Score. Naive sincere voters are unfairly disadvantaged compared to strategic
exaggerators.
Here is a variant I very much like but didn't nominate it because I didn't think it was "propose-able" enough in say the US.
*Voters score the candidates on a range large enough to strictly rank all the candidates plus leave several large-ish gaps. Default score is zero.
Eliminate all the candidates not in the Smith set.
Interpreting ballots as approving remaining candidates they score above average (of the scores they give to remaining candidates), elect
the most approved member of the Smith set*.
So this is like a "Declared Strategy" method that simulates: first use rankings to identify the members of the Smith set and then hold an Approval
election among those candidates.
What do you think?
Chris
On 13/04/2024 12:49 am, Toby Pereira wrote:
I will mention why I nominated Smith//Score. This method uses rated ballots, but uses them to infer a ranking. If there is a Condorcet winner, they are elected. Otherwise, elect the score winner of the Smith Set (the top cycle). I previously put why I think rated ballots work well for Condorcet (see below). And given a rated ballot, electing the highest scored candidate given no Condorcet winner seems the most simple and logical option, and shouldn't damage independence of clones or monotonicity. It also sidesteps any worry/complications over whether margins/winning votes etc. are the best thing to look at.
Toby
>I think one problem of burial-resistant methods is that they assume the electorate are aware of the consequences of it and will act accordingly. I think it might be a bit optimistic to expect the average voter to behave any differently using any method that uses a specific ballot type. Using a ranked ballot, if A and B are the frontrunning candidates, then supporters of A might rank B bottom because it's the obvious thing to do (which has been pointed out on here before I believe). Do you think the adoption of a specific Condorcet method will prevent that? I'm not convinced.
>Also, if there are two frontrunning candidates, A and B, it's quite likely anyway that supporters of A will see B as the worst candidate anyway, below the ones they know very little about. So it wouldn't really even be an act of burial, and therefore honest voting behaviour could cause a non-entity to win, because this is what burial-resistant methods do.
>I've said this before, but possibly the best solution for a Condorcet method would to be to use rated ballots. In this case B is less likely to be buried by the A supporters, because they would be likely to score the non-entity candidates 0 as well.
On Thursday, 11 April 2024 at 08:37:54 BST, Michael Ossipoff <email9648742@gmail.com> wrote:
On Wed, Apr 10, 2024 at 23:45 Chris Benham <cbenhamau@yahoo.com.au> wrote:
Have the nominations closed?
Yes, at 5:14:59 GMT, April 12th.
Not that I want to nominate another method.
There has been very little electioneering, with I think most of the
nominated methods not even being mentioned.
Yes, we weren’t given any information on most of the nominees. People should have told the advantages/merits of their nominees.
Some haven't even been explained let alone discussed or promoted.
Exactly. That’s why I ranked most of them all together at the same rank-position. The ones I know about & like ranked in order of merit, then the ones that I don’t know equal-ranked, & then, below them, the ones I know that I don’t like. Pluraity was at bottom, as everyone agrees, & so there was no reason to rank it.
