I'm no advocate of STAR, but interestingly it did come out well in Jameson Quinn's VSE (Voter Satisfaction Efficiency - basically utility) simulations. https://electionscience.github.io/vse-sim/VSEbasic So while it might not exactly pass a lot of criteria, it seems that is does perform generally well overall, as long as you accept that utility is a useful measure. Obviously all simulations contain simplifications and assumptions and that has to be taken into account as well. But then there is the question of what exactly one is after in a voting method - whether it's Condorcet at all cost and then whatever you can get hold of after that, utility, ticking off a list of criteria, or something else.
I do think failing clone independence is quite a black mark against STAR in any case. One way to fix it is to have the election method "clone" all the candidates anyway. Then run a two-winner sequential proportional election (take your pick of the methods) to find the two candidates for the run-off. The two candidates could just be a candidate and their clone, in which case that candidate automatically wins without a run-off.
Toby
On Saturday, 12 August 2023 at 03:45:25 BST, C.Benham cbenham@adam.com.au wrote:

I see from the "STAR Voting" advocates' website they propose using 0-5 scoring ballots.

STAR Voting it seems to me is just awful. It fails almost every desirable criterion you can think of.

It meets Condorcet Loser and Plurality and that's about it.  Their propaganda that it is somehow better
than IRV is very very dumb and/or dishonest.

Forest wrote:

But Copeland suffers from two fatal defect that STAR does not have ... Copeland is neither Decisive nor Clone
Independent.

What gives you the idea that STAR is Clone Independent?   It obviously fails Clone-Loser.  Say  the score winner pairwise loses to
the score runner-up. If we add a clone of the score-winner then the previous winner will be displaced out of the run-off.

One of the silly things about it is that all the major factions will have incentive to field two candidates (in the hope of capturing
both run-off spots).

Chris Benham

https://www.starvoting.org/

Date: Thu, 10 Aug 2023 17:19:15 -0700
From: Forest Simmons forest.simmons21@gmail.com
To: EM Election-methods@lists.electorama.com
Subject: [EM] STAR
Message-ID:
CANUDvfoOeBgZAgWiKPFG+UU0fcoDi771ZHHRTmkCEzVx-mLVyQ@mail.gmail.com
Content-Type: text/plain; charset="utf-8"

Score Then Automatic Runoff (STAR) elects the pairwise winner between the
two candidates with the highest score totals.

One of the biggest problems with this method is that there is an
appreciable likelihood that the winner W will be a candidate that is
pairwise dominated by some other candidate C, which means that C not only
scores higher than W on more ballots than not, but if there even exists  a
beatpath from W back to C, it will take at least three steps.

Most other extant methods have this same defect, but almost all of them are
hard to fix compared to STAR. This fact makes it easy for a tweaked version
of STAR to become arguably superior to any of these other methods.

1. Initialize a set S of candidates with the STAR winner.
2. If any candidate pairwise dominates the newest member of S, from among
   such candidates add in to S the one with the highest score.
3. Repeat step 2 until the set S cannot be enlarged any further in this way.
4. Elect the last candidate to be added to the set.

Usually step 2 will be invoked only one or two times if at all ... so this
is not a big tweak.

With this tweak STAR becomes arguably superior to any method currently in
use.

The only other method currently in use that always elects pairwise
undominated candidates is Copeland.  But Copeland suffers from two fatal
defect that STAR does not have ... Copeland is neither Decisive nor Clone
Independent.

Will STAR proponents take advantage of this opportunity? ... or will they
pass it up?

fws

In this clone-proof STAR method, you run a two-winner sequential proportional election (using a method that elects the score winner first) to determine the two candidates in the run-off. So the first candidate in the run-off is obviously the score winner, as would be the case in STAR. You then determine the second winner but among the remaining candidates would be an exact clone of the score winner. Essentially you allow that candidate to be elected to the run-off twice if they have enough support. The run-off is then the head-to-head that STAR does anyway. If a candidate and their clone make the run-off, that candidate just gets elected without the need for a run-off. I'm not putting this forward as a "good" method necessarily, but it is one way to cloneproof STAR.
I agree that it's not necessarily the case that greater utility can be achieved only through reduced criterion compliance. However, I do think that utility is still a useful thing to look at, even if it isn't as clear-cut as mathematical criterion compliance, and relies on accurate assumptions about candidates and voters.
Toby

On Saturday, 12 August 2023 at 18:04:24 BST, C.Benham <cbenham@adam.com.au> wrote:  

Jameson Quinn used to participate in discussion here.  I am not a fan of his or her ideas on voting methods.

I am very sceptical about claims that some method is great despite being crap on criterion compliances, based purely on computer simulations.

I refuse to believe that having fewer criterion compliances is needed for the sake of "greater utility".

"Condorcet at all cost" and then "ticking off a list of criteria" seems like a fine approach (or at least start) to me.  But some criteria are more desirable than others (and opinions can vary
on which) and some are incompatible with each other and we can invent or suggest new ones.

I do think failing clone independence is quite a black mark against STAR in any case. One way to fix it is to have the election method "clone" all the candidates anyway.

How would it do that?

Chris B.

On 13/08/2023 12:17 am, Toby Pereira wrote:

I'm no advocate of STAR, but interestingly it did come out well in Jameson Quinn's VSE (Voter Satisfaction Efficiency - basically utility) simulations. https://electionscience.github.io/vse-sim/VSEbasic So while it might not exactly pass a lot of criteria, it seems that is does perform generally well overall, as long as you accept that utility is a useful measure. Obviously all simulations contain simplifications and assumptions and that has to be taken into account as well. But then there is the question of what exactly one is after in a voting method - whether it's Condorcet at all cost and then whatever you can get hold of after that, utility, ticking off a list of criteria, or something else.
I do think failing clone independence is quite a black mark against STAR in any case. One way to fix it is to have the election method "clone" all the candidates anyway. Then run a two-winner sequential proportional election (take your pick of the methods) to find the two candidates for the run-off. The two candidates could just be a candidate and their clone, in which case that candidate automatically wins without a run-off.
Toby
On Saturday, 12 August 2023 at 03:45:25 BST, C.Benham cbenham@adam.com.au wrote:

I see from the "STAR Voting" advocates' website they propose using 0-5 scoring ballots.

STAR Voting it seems to me is just awful. It fails almost every desirable criterion you can think of.

It meets Condorcet Loser and Plurality and that's about it.  Their propaganda that it is somehow better
than IRV is very very dumb and/or dishonest.

Forest wrote:

But Copeland suffers from two fatal defect that STAR does not have ... Copeland is neither Decisive nor Clone
Independent.

What gives you the idea that STAR is Clone Independent?   It obviously fails Clone-Loser.  Say  the score winner pairwise loses to
the score runner-up. If we add a clone of the score-winner then the previous winner will be displaced out of the run-off.

One of the silly things about it is that all the major factions will have incentive to field two candidates (in the hope of capturing
both run-off spots).

Chris Benham

https://www.starvoting.org/

Date: Thu, 10 Aug 2023 17:19:15 -0700
From: Forest Simmons forest.simmons21@gmail.com
To: EM Election-methods@lists.electorama.com
Subject: [EM] STAR
Message-ID:
CANUDvfoOeBgZAgWiKPFG+UU0fcoDi771ZHHRTmkCEzVx-mLVyQ@mail.gmail.com
Content-Type: text/plain; charset="utf-8"

Score Then Automatic Runoff (STAR) elects the pairwise winner between the
two candidates with the highest score totals.

One of the biggest problems with this method is that there is an
appreciable likelihood that the winner W will be a candidate that is
pairwise dominated by some other candidate C, which means that C not only
scores higher than W on more ballots than not, but if there even exists  a
beatpath from W back to C, it will take at least three steps.

Most other extant methods have this same defect, but almost all of them are
hard to fix compared to STAR. This fact makes it easy for a tweaked version
of STAR to become arguably superior to any of these other methods.

1. Initialize a set S of candidates with the STAR winner.
2. If any candidate pairwise dominates the newest member of S, from among
   such candidates add in to S the one with the highest score.
3. Repeat step 2 until the set S cannot be enlarged any further in this way.
4. Elect the last candidate to be added to the set.

Usually step 2 will be invoked only one or two times if at all ... so this
is not a big tweak.

With this tweak STAR becomes arguably superior to any method currently in
use.

The only other method currently in use that always elects pairwise
undominated candidates is Copeland.  But Copeland suffers from two fatal
defect that STAR does not have ... Copeland is neither Decisive nor Clone
Independent.

Will STAR proponents take advantage of this opportunity? ... or will they
pass it up?

fws

I think this is an interesting point. We can ask at a philosophical level what makes a good voting method. Is it just one that ticks the most boxes, or is it one that most reliably gets the "best" result? And what do we mean by best result? Are we working from the assumption that there is a best result to begin with (e.g. nearest the median, highest utility, closest possible to Condorcet winner) and looking for a method that most reliably picks that? I think generally while passing certain criteria is a good thing, occasional failures on multiple criteria isn't necessarily worse than strictly passing more criteria but also having worse failures than others. This is very hypothetical of course, and it does depend on the specific criteria.
However, if someone was naively looking at a list of criteria in isolation without knowing what was compatible with what, I think they would be more likely to put participation ahead of Condorcet than vice versa in a list of importance. But because participation is very hard to achieve (in simple terms you can have score, approval and some other awful methods), it's largely brushed under the carpet. Whereas, on the other hand, Condorcet is largely held up as the biggest deal of the lot on this mailing list. It's not that I think Condorcet is such a bad thing - I just don't think it's the be-all-and-end-all. In terms of voting reform, it can make sense to push for it because it makes intuitive sense, but in an ideal world where everyone was enlightened, I don't think it would have to be a deal-breaker. And that's partly because the premise of Condorcet is essentially built on a logical fallacy - basically that if A is preferred to B on more ballots that vice versa then electing A must be a better result than electing B.
Toby
On Sunday, 13 August 2023 at 12:41:36 BST, Colin Champion colin.champion@routemaster.app wrote:

I find Chris's position unintelligible. People who rely on evaluations can say what it is which makes a certain candidate a good choice of winner from a given election, namely - in the case of spatial models - his being closest to the centre of the distribution of voters (usually understood as the median). But people who reason from criteria never say what makes a criterion valid or invalid, or important rather than unimportant. So the criteria come across as postulates pulled from thin air - supposedly intuitive truths whose truth conditions cannot be identified.

CJC

On 12/08/2023 18:04, C.Benham wrote:

Jameson Quinn used to participate in discussion here.  I am not a fan of his or her ideas on voting methods.

I am very sceptical about claims that some method is great despite being crap on criterion compliances, based purely on computer simulations.

I refuse to believe that having fewer criterion compliances is needed for the sake of "greater utility".

"Condorcet at all cost" and then "ticking off a list of criteria" seems like a fine approach (or at least start) to me.  But some criteria are more desirable than others (and opinions can vary
on which) and some are incompatible with each other and we can invent or suggest new ones.

I do think failing clone independence is quite a black mark against STAR in any case. One way to fix it is to have the election method "clone" all the candidates anyway.

How would it do that?

Chris B.

On 13/08/2023 12:17 am, Toby Pereira wrote:

I'm no advocate of STAR, but interestingly it did come out well in Jameson Quinn's VSE (Voter Satisfaction Efficiency - basically utility) simulations. https://electionscience.github.io/vse-sim/VSEbasic So while it might not exactly pass a lot of criteria, it seems that is does perform generally well overall, as long as you accept that utility is a useful measure. Obviously all simulations contain simplifications and assumptions and that has to be taken into account as well. But then there is the question of what exactly one is after in a voting method - whether it's Condorcet at all cost and then whatever you can get hold of after that, utility, ticking off a list of criteria, or something else.
I do think failing clone independence is quite a black mark against STAR in any case. One way to fix it is to have the election method "clone" all the candidates anyway. Then run a two-winner sequential proportional election (take your pick of the methods) to find the two candidates for the run-off. The two candidates could just be a candidate and their clone, in which case that candidate automatically wins without a run-off.
Toby
On Saturday, 12 August 2023 at 03:45:25 BST, C.Benham cbenham@adam.com.au wrote:

I see from the "STAR Voting" advocates' website they propose using 0-5 scoring ballots.

STAR Voting it seems to me is just awful. It fails almost every desirable criterion you can think of.

It meets Condorcet Loser and Plurality and that's about it.  Their propaganda that it is somehow better
than IRV is very very dumb and/or dishonest.

Forest wrote:

But Copeland suffers from two fatal defect that STAR does not have ... Copeland is neither Decisive nor Clone
Independent.

What gives you the idea that STAR is Clone Independent?   It obviously fails Clone-Loser.  Say  the score winner pairwise loses to
the score runner-up. If we add a clone of the score-winner then the previous winner will be displaced out of the run-off.

One of the silly things about it is that all the major factions will have incentive to field two candidates (in the hope of capturing
both run-off spots).

Chris Benham

https://www.starvoting.org/

Date: Thu, 10 Aug 2023 17:19:15 -0700
From: Forest Simmons forest.simmons21@gmail.com
To: EM Election-methods@lists.electorama.com
Subject: [EM] STAR
Message-ID:
CANUDvfoOeBgZAgWiKPFG+UU0fcoDi771ZHHRTmkCEzVx-mLVyQ@mail.gmail.com
Content-Type: text/plain; charset="utf-8"

Score Then Automatic Runoff (STAR) elects the pairwise winner between the
two candidates with the highest score totals.

One of the biggest problems with this method is that there is an
appreciable likelihood that the winner W will be a candidate that is
pairwise dominated by some other candidate C, which means that C not only
scores higher than W on more ballots than not, but if there even exists  a
beatpath from W back to C, it will take at least three steps.

Most other extant methods have this same defect, but almost all of them are
hard to fix compared to STAR. This fact makes it easy for a tweaked version
of STAR to become arguably superior to any of these other methods.

1. Initialize a set S of candidates with the STAR winner.
2. If any candidate pairwise dominates the newest member of S, from among
   such candidates add in to S the one with the highest score.
3. Repeat step 2 until the set S cannot be enlarged any further in this way.
4. Elect the last candidate to be added to the set.

Usually step 2 will be invoked only one or two times if at all ... so this
is not a big tweak.

With this tweak STAR becomes arguably superior to any method currently in
use.

The only other method currently in use that always elects pairwise
undominated candidates is Copeland.  But Copeland suffers from two fatal
defect that STAR does not have ... Copeland is neither Decisive nor Clone
Independent.

Will STAR proponents take advantage of this opportunity? ... or will they
pass it up?

fws

Election-Methods mailing list - see https://electorama.com/em for list info

Election-Methods mailing list - see https://electorama.com/em for list info

To be clearer I should have said "logical impossibility" rather than "logical fallacy".
Toby
On Sunday, 13 August 2023 at 23:17:44 BST, Toby Pereira tdp201b@yahoo.co.uk wrote:

I think this is an interesting point. We can ask at a philosophical level what makes a good voting method. Is it just one that ticks the most boxes, or is it one that most reliably gets the "best" result? And what do we mean by best result? Are we working from the assumption that there is a best result to begin with (e.g. nearest the median, highest utility, closest possible to Condorcet winner) and looking for a method that most reliably picks that? I think generally while passing certain criteria is a good thing, occasional failures on multiple criteria isn't necessarily worse than strictly passing more criteria but also having worse failures than others. This is very hypothetical of course, and it does depend on the specific criteria.
However, if someone was naively looking at a list of criteria in isolation without knowing what was compatible with what, I think they would be more likely to put participation ahead of Condorcet than vice versa in a list of importance. But because participation is very hard to achieve (in simple terms you can have score, approval and some other awful methods), it's largely brushed under the carpet. Whereas, on the other hand, Condorcet is largely held up as the biggest deal of the lot on this mailing list. It's not that I think Condorcet is such a bad thing - I just don't think it's the be-all-and-end-all. In terms of voting reform, it can make sense to push for it because it makes intuitive sense, but in an ideal world where everyone was enlightened, I don't think it would have to be a deal-breaker. And that's partly because the premise of Condorcet is essentially built on a logical fallacy - basically that if A is preferred to B on more ballots that vice versa then electing A must be a better result than electing B.
Toby
On Sunday, 13 August 2023 at 12:41:36 BST, Colin Champion colin.champion@routemaster.app wrote:

I find Chris's position unintelligible. People who rely on evaluations can say what it is which makes a certain candidate a good choice of winner from a given election, namely - in the case of spatial models - his being closest to the centre of the distribution of voters (usually understood as the median). But people who reason from criteria never say what makes a criterion valid or invalid, or important rather than unimportant. So the criteria come across as postulates pulled from thin air - supposedly intuitive truths whose truth conditions cannot be identified.

CJC

On 12/08/2023 18:04, C.Benham wrote:

Jameson Quinn used to participate in discussion here.  I am not a fan of his or her ideas on voting methods.

I am very sceptical about claims that some method is great despite being crap on criterion compliances, based purely on computer simulations.

I refuse to believe that having fewer criterion compliances is needed for the sake of "greater utility".

"Condorcet at all cost" and then "ticking off a list of criteria" seems like a fine approach (or at least start) to me.  But some criteria are more desirable than others (and opinions can vary
on which) and some are incompatible with each other and we can invent or suggest new ones.

I do think failing clone independence is quite a black mark against STAR in any case. One way to fix it is to have the election method "clone" all the candidates anyway.

How would it do that?

Chris B.

On 13/08/2023 12:17 am, Toby Pereira wrote:

I'm no advocate of STAR, but interestingly it did come out well in Jameson Quinn's VSE (Voter Satisfaction Efficiency - basically utility) simulations. https://electionscience.github.io/vse-sim/VSEbasic So while it might not exactly pass a lot of criteria, it seems that is does perform generally well overall, as long as you accept that utility is a useful measure. Obviously all simulations contain simplifications and assumptions and that has to be taken into account as well. But then there is the question of what exactly one is after in a voting method - whether it's Condorcet at all cost and then whatever you can get hold of after that, utility, ticking off a list of criteria, or something else.
I do think failing clone independence is quite a black mark against STAR in any case. One way to fix it is to have the election method "clone" all the candidates anyway. Then run a two-winner sequential proportional election (take your pick of the methods) to find the two candidates for the run-off. The two candidates could just be a candidate and their clone, in which case that candidate automatically wins without a run-off.
Toby
On Saturday, 12 August 2023 at 03:45:25 BST, C.Benham cbenham@adam.com.au wrote:

I see from the "STAR Voting" advocates' website they propose using 0-5 scoring ballots.

STAR Voting it seems to me is just awful. It fails almost every desirable criterion you can think of.

It meets Condorcet Loser and Plurality and that's about it.  Their propaganda that it is somehow better
than IRV is very very dumb and/or dishonest.

Forest wrote:

But Copeland suffers from two fatal defect that STAR does not have ... Copeland is neither Decisive nor Clone
Independent.

What gives you the idea that STAR is Clone Independent?   It obviously fails Clone-Loser.  Say  the score winner pairwise loses to
the score runner-up. If we add a clone of the score-winner then the previous winner will be displaced out of the run-off.

One of the silly things about it is that all the major factions will have incentive to field two candidates (in the hope of capturing
both run-off spots).

Chris Benham

https://www.starvoting.org/

Date: Thu, 10 Aug 2023 17:19:15 -0700
From: Forest Simmons forest.simmons21@gmail.com
To: EM Election-methods@lists.electorama.com
Subject: [EM] STAR
Message-ID:
CANUDvfoOeBgZAgWiKPFG+UU0fcoDi771ZHHRTmkCEzVx-mLVyQ@mail.gmail.com
Content-Type: text/plain; charset="utf-8"

Score Then Automatic Runoff (STAR) elects the pairwise winner between the
two candidates with the highest score totals.

One of the biggest problems with this method is that there is an
appreciable likelihood that the winner W will be a candidate that is
pairwise dominated by some other candidate C, which means that C not only
scores higher than W on more ballots than not, but if there even exists  a
beatpath from W back to C, it will take at least three steps.

Most other extant methods have this same defect, but almost all of them are
hard to fix compared to STAR. This fact makes it easy for a tweaked version
of STAR to become arguably superior to any of these other methods.

1. Initialize a set S of candidates with the STAR winner.
2. If any candidate pairwise dominates the newest member of S, from among
   such candidates add in to S the one with the highest score.
3. Repeat step 2 until the set S cannot be enlarged any further in this way.
4. Elect the last candidate to be added to the set.

Usually step 2 will be invoked only one or two times if at all ... so this
is not a big tweak.

With this tweak STAR becomes arguably superior to any method currently in
use.

The only other method currently in use that always elects pairwise
undominated candidates is Copeland.  But Copeland suffers from two fatal
defect that STAR does not have ... Copeland is neither Decisive nor Clone
Independent.

Will STAR proponents take advantage of this opportunity? ... or will they
pass it up?

fws

Election-Methods mailing list - see https://electorama.com/em for list info

Election-Methods mailing list - see https://electorama.com/em for list info