Re: [EM] Voting without ballots (was: Minimax)
That method could be said to be problematic since it is not symmetric. But on the other hand it could be an excellent method for many elections where we the asymmetry has some explanation.
If we have an incumbent president that we might or might not change, that president could be the default option. In addition to voting on who is the best candidate, we would be voting also on whether to change the president or not. If the current president is no longer a candidate, he could nominate a proposed successor who would be the default option. Or in a two-party system one could pick the (first) candidate of the opposition party as the default option, since that's the way things tend to go anyway (at least after the current president has run out of allowed terms). Or maybe a parliament or other central but wider group would pick the default option. The default option should not be a weak option but one that is a plausible winner (the current president should thus nominate a candidate of some other party as the default option if it is obvious that he is currently popular).
This method would allow also having a threshold that would be required to not elect the incumbent or default option.
This method is a nice way to solve the problem of cyclic group preferences if one can use time to distribute the victory to multiple candidates. If there is a permanent top cycle A>B>C>A, C would win A, B would win C in the next election after maybe a year etc. No need to "break" the cycle since time will (maybe) allow all the cyclic candidates to participate. They just need to wait for their turn. (Not allowing the Minmax winner outside of the top cycle to win could be considered a disadvantage of this method.)
The ballot could be simply an approval ballot, asking "which options do you prefer to the default option". This method could thus be seen also as an improved Approval method. Adding a (strong) default option to Approval makes it more sensible than it is without that default option.
This approach could offer also a smooth path away from Plurality.
Juho
On 12 May 2017, at 02:48, Kevin Venzke stepjak@yahoo.fr wrote:
I think this is an interesting topic... Basically how to make a decision in a meeting when you don't have ballots and people aren't inclined to talk to each other. There's probably a whole field on this topic and I'm just not aware of it...
My idea here would be one I've mentioned before (but as an election method), that somehow a default option is initially selected, and then this option wins unless a full majority thinks another option is better. The largest such majority would immediately win in that situation; we would not chase around any cycles. So, there is both an advantage and disadvantage to being the default option. The advantage is obvious. The disadvantage is that the default option could lose when, had it not been the default, it could have won.
To rephrase a bit, I'm saying that for candidates A...Z where A is the default option, you only collect the votes for B...Z against A. Nobody's voting A over anything.
By not collecting the other pairwise information you might hopefully reduce the potential to make arguments against the winner's legitimacy. (A bit like the difficulty of undermining the winner under FPP or Approval, when the only sure data available is the counts that produced the winner.)
You could pick the default option using an election method (and even have an instant-win criterion if one is sensible). The mechanism does not enforce Plurality, only minimal defense.
Sorry for lack of other responses. I can't find enough time to write.
Kevin
De : Juho Laatu juho.laatu@gmail.com
À : Election Methods election-methods@lists.electorama.com
Cc : Kevin Venzke stepjak@yahoo.fr
Envoyé le : Dimanche 7 mai 2017 5h32
Objet : Re: [EM] Ordering defeats in Minimax
On 07 May 2017, at 12:39, Kristofer Munsterhjelm <km_elmet@t-online.de mailto:km_elmet@t-online.de> wrote:
On 05/07/2017 10:58 AM, Juho Laatu wrote:
On 06 May 2017, at 23:26, Kevin Venzke <stepjak@yahoo.fr mailto:stepjak@yahoo.fr
<mailto:stepjak@yahoo.fr mailto:stepjak@yahoo.fr>> wrote:
Legislators usually don't use election
methods. I would say essentially they figure out among themselves what
the outcome likely must be and then "elect" it, typically by a majority
vote. If voters could do that directly I think that would be the ideal
situation, as far as minimizing the need for strategy.
That's an interesting viewpoint. I'd like to see Condorcet tested in
some parliaments. Although negotiations and Plurality can be used to
solve many cases, also Condorcet and the idea of having multiple
candidates could have some benefits. The first one in my mind is
openness. Even when the "big boys" have agreed something behind the
screens, and then bring that decision into a majority vote, some smaller
groupings could add some alternatives in the election that they consider
better. This would open up the available alternatives to the world
(media, public), and all voters would have to take position on the
presented alternatives. Small parties or groupings could thus make
sensible compromise proposals that might even win, if they are good enough.
I think some parliaments use the following procedure for multiple proposals or amendments:
Some group decides the agenda (I'm not sure who; it probably depends on the parliament in question).
Then the parliament discusses the first two options, and does a majority vote between the two. It then does a majority vote between the winner so far and the next option, and so on until all options have been considered. The winner at the end of the procedure is the option that is accepted.
This method is Condorcet (because the CW beats everybody pairwise, and so whenever it becomes the winner so far, it'll stay the winner so far; and whenever it's the challenger, it will win against the winner so far and become the next winner so far). Furthermore, it's Smith. How cycles are broken depends on the agenda.
Yes, this is some sort of "serial Condorcet". The biggest problem of this approach is probably the agenda based cycle breaking that you mentioned.
If there is a Condorcet winner, it will win. But if there is any chance of having a cycle, then the chairman (if he decides the agenda) may arrange the agenda so that his favourite option is the last one in the queue of options. With three candidates this already determines the order of all pairwise votes.
With three candidates the cycle could be either A>B>C>A or A>C>B>A. The chairman wants A to win. The first vote will therefore be between B and C. If we have cycle A>B>C>A, B will win the first round, and A will win the second round (against B). If we have cycle A>C>B>A, C will win the first round, and A will win the second round (against C). The chairman could thus pick the preferred winner in the case that there is no Condorcet winner.
The voters could also try to use some strategies. This gives us some quite interesting strategy patterns, since after the agenda has been set, the method is no more symmetric with respect to the different options. A and B supporters could also join forces and vote against A at the last round, if they don't like the chairmans idea of constructing a strategic agenda.
Juho
Election-Methods mailing list - see http://electorama.com/em http://electorama.com/emfor list info
I like this idea, Juho. If there is an incumbent, put all the challengers
on one approval ballot and ask "Which ones do you like better than the
incumbent?" It is a nice solution to the what-does-approval-mean problem.
Like you said, the threshold doesn't have to be 50%, either. It could go
down, in a gradual sort-of term limit. After one term, the mayor will
remain if no challenger gets over 60%. After two terms, it's 50%. After
three terms, it's 40%. After four terms, it's 30%.
In the past, I haven't been sure what to do if there is no incumbent.
Letting the incumbent pick a successor is an interesting solution. As is
letting the largest opposition party choose someone (if the election is
partisan).
Andy
On Thu, May 11, 2017 at 11:30 PM, Juho Laatu juho.laatu@gmail.com wrote:
That method could be said to be problematic since it is not symmetric. But
on the other hand it could be an excellent method for many elections where
we the asymmetry has some explanation.
If we have an incumbent president that we might or might not change, that
president could be the default option. In addition to voting on who is the
best candidate, we would be voting also on whether to change the president
or not. If the current president is no longer a candidate, he could
nominate a proposed successor who would be the default option. Or in a
two-party system one could pick the (first) candidate of the opposition
party as the default option, since that's the way things tend to go anyway
(at least after the current president has run out of allowed terms). Or
maybe a parliament or other central but wider group would pick the default
option. The default option should not be a weak option but one that is a
plausible winner (the current president should thus nominate a candidate of
some other party as the default option if it is obvious that he is
currently popular).
This method would allow also having a threshold that would be required to
not elect the incumbent or default option.
This method is a nice way to solve the problem of cyclic group preferences
if one can use time to distribute the victory to multiple candidates. If
there is a permanent top cycle A>B>C>A, C would win A, B would win C in the
next election after maybe a year etc. No need to "break" the cycle since
time will (maybe) allow all the cyclic candidates to participate. They just
need to wait for their turn. (Not allowing the Minmax winner outside of the
top cycle to win could be considered a disadvantage of this method.)
The ballot could be simply an approval ballot, asking "which options do
you prefer to the default option". This method could thus be seen also as
an improved Approval method. Adding a (strong) default option to Approval
makes it more sensible than it is without that default option.
This approach could offer also a smooth path away from Plurality.
Juho
On 12 May 2017, at 02:48, Kevin Venzke stepjak@yahoo.fr wrote:
I think this is an interesting topic... Basically how to make a decision
in a meeting when you don't have ballots and people aren't inclined to talk
to each other. There's probably a whole field on this topic and I'm just
not aware of it...
My idea here would be one I've mentioned before (but as an election
method), that somehow a default option is initially selected, and then this
option wins unless a full majority thinks another option is better. The
largest such majority would immediately win in that situation; we would not
chase around any cycles. So, there is both an advantage and disadvantage to
being the default option. The advantage is obvious. The disadvantage is
that the default option could lose when, had it not been the default, it
could have won.
To rephrase a bit, I'm saying that for candidates A...Z where A is the
default option, you only collect the votes for B...Z against A. Nobody's
voting A over anything.
By not collecting the other pairwise information you might hopefully
reduce the potential to make arguments against the winner's legitimacy. (A
bit like the difficulty of undermining the winner under FPP or Approval,
when the only sure data available is the counts that produced the winner.)
You could pick the default option using an election method (and even have
an instant-win criterion if one is sensible). The mechanism does not
enforce Plurality, only minimal defense.
Sorry for lack of other responses. I can't find enough time to write.
Kevin
De : Juho Laatu juho.laatu@gmail.com
À : Election Methods election-methods@lists.electorama.com
Cc : Kevin Venzke stepjak@yahoo.fr
Envoyé le : Dimanche 7 mai 2017 5h32
Objet : Re: [EM] Ordering defeats in Minimax
On 07 May 2017, at 12:39, Kristofer Munsterhjelm km_elmet@t-online.de
wrote:
On 05/07/2017 10:58 AM, Juho Laatu wrote:
On 06 May 2017, at 23:26, Kevin Venzke <stepjak@yahoo.fr
mailto:stepjak@yahoo.fr> wrote:
Legislators usually don't use election
methods. I would say essentially they figure out among themselves what
the outcome likely must be and then "elect" it, typically by a majority
vote. If voters could do that directly I think that would be the ideal
situation, as far as minimizing the need for strategy.
That's an interesting viewpoint. I'd like to see Condorcet tested in
some parliaments. Although negotiations and Plurality can be used to
solve many cases, also Condorcet and the idea of having multiple
candidates could have some benefits. The first one in my mind is
openness. Even when the "big boys" have agreed something behind the
screens, and then bring that decision into a majority vote, some smaller
groupings could add some alternatives in the election that they consider
better. This would open up the available alternatives to the world
(media, public), and all voters would have to take position on the
presented alternatives. Small parties or groupings could thus make
sensible compromise proposals that might even win, if they are good
enough.
I think some parliaments use the following procedure for multiple
proposals or amendments:
Some group decides the agenda (I'm not sure who; it probably depends on
the parliament in question).
Then the parliament discusses the first two options, and does a majority
vote between the two. It then does a majority vote between the winner so
far and the next option, and so on until all options have been considered.
The winner at the end of the procedure is the option that is accepted.
This method is Condorcet (because the CW beats everybody pairwise, and
so whenever it becomes the winner so far, it'll stay the winner so far; and
whenever it's the challenger, it will win against the winner so far and
become the next winner so far). Furthermore, it's Smith. How cycles are
broken depends on the agenda.
Yes, this is some sort of "serial Condorcet". The biggest problem of this
approach is probably the agenda based cycle breaking that you mentioned.
If there is a Condorcet winner, it will win. But if there is any chance of
having a cycle, then the chairman (if he decides the agenda) may arrange
the agenda so that his favourite option is the last one in the queue of
options. With three candidates this already determines the order of all
pairwise votes.
With three candidates the cycle could be either A>B>C>A or A>C>B>A. The
chairman wants A to win. The first vote will therefore be between B and C.
If we have cycle A>B>C>A, B will win the first round, and A will win the
second round (against B). If we have cycle A>C>B>A, C will win the first
round, and A will win the second round (against C). The chairman could thus
pick the preferred winner in the case that there is no Condorcet winner.
The voters could also try to use some strategies. This gives us some quite
interesting strategy patterns, since after the agenda has been set, the
method is no more symmetric with respect to the different options. A and B
supporters could also join forces and vote against A at the last round, if
they don't like the chairmans idea of constructing a strategic agenda.
Juho
Election-Methods mailing list - see http://electorama.com/em for list info
Election-Methods mailing list - see http://electorama.com/em for list info
I have concerns about dropping the threshold below 50%, even if the incumbent has been in power so long that we want a bias in favor of getting rid of him. Strategically the method starts to turn into Approval, with people voting with that mindset rather than voting as though it is a pairwise comparison. (Even at 50% this could be an issue; imagine for instance the situation that the default option is so terrible that voters expect that many candidates will hit the 50% threshold.)
One of my main ideas is to address the area where Approval makes me nervous: The situation that the winner doesn't have a majority, but it looks like there could have been a majority among a divided opposition, if one posed the question right. I hope that clarifying the stakes would reduce the chicken dilemma potential.
Another motivation is to try to keep more options available for the last step of the election. It's easy to imagine this as a revision of the top-two runoff or jungle primary, where we take note of just the single frontrunner of the first round, and avoid eliminating the losing options for the second round.
I certainly think the contests against the default should use an approval ballot, but I think there should probably be an additional checkbox for those who don't want to vote for anybody as superior to the default. I just think you would want some distinction between a pro-default ballot and a blank ballot.
Kevin
De : Andy Jennings <elections@jenningsstory.com>
À : Juho Laatu juho.laatu@gmail.com
Cc : Kevin Venzke stepjak@yahoo.fr; Election Methods election-methods@lists.electorama.com
Envoyé le : Vendredi 12 mai 2017 7h32
Objet : Re: [EM] Voting without ballots (was: Minimax)
I like this idea, Juho. If there is an incumbent, put all the challengers on one approval ballot and ask "Which ones do you like better than the incumbent?" It is a nice solution to the what-does-approval-mean problem.
Like you said, the threshold doesn't have to be 50%, either. It could go down, in a gradual sort-of term limit. After one term, the mayor will remain if no challenger gets over 60%. After two terms, it's 50%. After three terms, it's 40%. After four terms, it's 30%.
In the past, I haven't been sure what to do if there is no incumbent. Letting the incumbent pick a successor is an interesting solution. As is letting the largest opposition party choose someone (if the election is partisan).
Andy
On Thu, May 11, 2017 at 11:30 PM, Juho Laatu juho.laatu@gmail.com wrote:
That method could be said to be problematic since it is not symmetric. But on the other hand it could be an excellent method for many elections where we the asymmetry has some explanation.
If we have an incumbent president that we might or might not change, that president could be the default option. In addition to voting on who is the best candidate, we would be voting also on whether to change the president or not. If the current president is no longer a candidate, he could nominate a proposed successor who would be the default option. Or in a two-party system one could pick the (first) candidate of the opposition party as the default option, since that's the way things tend to go anyway (at least after the current president has run out of allowed terms). Or maybe a parliament or other central but wider group would pick the default option. The default option should not be a weak option but one that is a plausible winner (the current president should thus nominate a candidate of some other party as the default option if it is obvious that he is currently popular).
This method would allow also having a threshold that would be required to not elect the incumbent or default option.
This method is a nice way to solve the problem of cyclic group preferences if one can use time to distribute the victory to multiple candidates. If there is a permanent top cycle A>B>C>A, C would win A, B would win C in the next election after maybe a year etc. No need to "break" the cycle since time will (maybe) allow all the cyclic candidates to participate. They just need to wait for their turn. (Not allowing the Minmax winner outside of the top cycle to win could be considered a disadvantage of this method.)
The ballot could be simply an approval ballot, asking "which options do you prefer to the default option". This method could thus be seen also as an improved Approval method. Adding a (strong) default option to Approval makes it more sensible than it is without that default option.
This approach could offer also a smooth path away from Plurality.
Juho
On 12 May 2017, at 02:48, Kevin Venzke stepjak@yahoo.fr wrote:
I think this is an interesting topic... Basically how to make a decision in a meeting when you don't have ballots and people aren't inclined to talk to each other. There's probably a whole field on this topic and I'm just not aware of it...
My idea here would be one I've mentioned before (but as an election method), that somehow a default option is initially selected, and then this option wins unless a full majority thinks another option is better. The largest such majority would immediately win in that situation; we would not chase around any cycles. So, there is both an advantage and disadvantage to being the default option. The advantage is obvious. The disadvantage is that the default option could lose when, had it not been the default, it could have won.
To rephrase a bit, I'm saying that for candidates A...Z where A is the default option, you only collect the votes for B...Z against A. Nobody's voting A over anything.
By not collecting the other pairwise information you might hopefully reduce the potential to make arguments against the winner's legitimacy. (A bit like the difficulty of undermining the winner under FPP or Approval, when the only sure data available is the counts that produced the winner.)
You could pick the default option using an election method (and even have an instant-win criterion if one is sensible). The mechanism does not enforce Plurality, only minimal defense.
Sorry for lack of other responses. I can't find enough time to write.
Kevin
De : Juho Laatu <juho.laatu@gmail.com>
À : Election Methods <election-methods@lists. electorama.com>
Cc : Kevin Venzke stepjak@yahoo.fr
Envoyé le : Dimanche 7 mai 2017 5h32
Objet : Re: [EM] Ordering defeats in Minimax
On 07 May 2017, at 12:39, Kristofer Munsterhjelm km_elmet@t-online.de wrote:
On 05/07/2017 10:58 AM, Juho Laatu wrote:
On 06 May 2017, at 23:26, Kevin Venzke <stepjak@yahoo.fr
mailto:stepjak@yahoo.fr> wrote:
Legislators usually don't use election
methods. I would say essentially they figure out among themselves what
the outcome likely must be and then "elect" it, typically by a majority
vote. If voters could do that directly I think that would be the ideal
situation, as far as minimizing the need for strategy.
That's an interesting viewpoint. I'd like to see Condorcet tested in
some parliaments. Although negotiations and Plurality can be used to
solve many cases, also Condorcet and the idea of having multiple
candidates could have some benefits. The first one in my mind is
openness. Even when the "big boys" have agreed something behind the
screens, and then bring that decision into a majority vote, some smaller
groupings could add some alternatives in the election that they consider
better. This would open up the available alternatives to the world
(media, public), and all voters would have to take position on the
presented alternatives. Small parties or groupings could thus make
sensible compromise proposals that might even win, if they are good enough.
I think some parliaments use the following procedure for multiple proposals or amendments:
Some group decides the agenda (I'm not sure who; it probably depends on the parliament in question).
Then the parliament discusses the first two options, and does a majority vote between the two. It then does a majority vote between the winner so far and the next option, and so on until all options have been considered. The winner at the end of the procedure is the option that is accepted.
This method is Condorcet (because the CW beats everybody pairwise, and so whenever it becomes the winner so far, it'll stay the winner so far; and whenever it's the challenger, it will win against the winner so far and become the next winner so far). Furthermore, it's Smith. How cycles are broken depends on the agenda.
Yes, this is some sort of "serial Condorcet". The biggest problem of this approach is probably the agenda based cycle breaking that you mentioned.
If there is a Condorcet winner, it will win. But if there is any chance of having a cycle, then the chairman (if he decides the agenda) may arrange the agenda so that his favourite option is the last one in the queue of options. With three candidates this already determines the order of all pairwise votes.
With three candidates the cycle could be either A>B>C>A or A>C>B>A. The chairman wants A to win. The first vote will therefore be between B and C. If we have cycle A>B>C>A, B will win the first round, and A will win the second round (against B). If we have cycle A>C>B>A, C will win the first round, and A will win the second round (against C). The chairman could thus pick the preferred winner in the case that there is no Condorcet winner.
The voters could also try to use some strategies. This gives us some quite interesting strategy patterns, since after the agenda has been set, the method is no more symmetric with respect to the different options. A and B supporters could also join forces and vote against A at the last round, if they don't like the chairmans idea of constructing a strategic agenda.
Juho
Election-Methods mailing list - see http://electorama.com/em for list info
Election-Methods mailing list - see http://electorama.com/em for list info
Election-Methods mailing list - see http://electorama.com/em for list info
On 13 May 2017, at 01:07, Kevin Venzke stepjak@yahoo.fr wrote:
I have concerns about dropping the threshold below 50%, even if the incumbent has been in power so long that we want a bias in favor of getting rid of him. Strategically the method starts to turn into Approval, with people voting with that mindset rather than voting as though it is a pairwise comparison. (Even at 50% this could be an issue; imagine for instance the situation that the default option is so terrible that voters expect that many candidates will hit the 50% threshold.)
Yes, the default option should be strong (one of the potential winners, or stronger). Otherwise we will have all the problems of basic Approval.
One of my main ideas is to address the area where Approval makes me nervous: The situation that the winner doesn't have a majority, but it looks like there could have been a majority among a divided opposition, if one posed the question right. I hope that clarifying the stakes would reduce the chicken dilemma potential.
Another motivation is to try to keep more options available for the last step of the election. It's easy to imagine this as a revision of the top-two runoff or jungle primary, where we take note of just the single frontrunner of the first round, and avoid eliminating the losing options for the second round.
Yes. One could use the first round just to define the default option. Approval could be good enough for the first round. The first round and the combined two round method should be simple, since if it gets complex, we might as well be use some basic Condorcet method.
I certainly think the contests against the default should use an approval ballot, but I think there should probably be an additional checkbox for those who don't want to vote for anybody as superior to the default. I just think you would want some distinction between a pro-default ballot and a blank ballot.
I guess that would be for the statistics, i.e. that difference would not influence the outcome. One (not necessarily good) approach would be to allow all options to be ticked either as "worse" or "better" than the default option.
Juho
Kevin
De : Andy Jennings elections@jenningsstory.com
À : Juho Laatu juho.laatu@gmail.com
Cc : Kevin Venzke stepjak@yahoo.fr; Election Methods election-methods@lists.electorama.com
Envoyé le : Vendredi 12 mai 2017 7h32
Objet : Re: [EM] Voting without ballots (was: Minimax)
I like this idea, Juho. If there is an incumbent, put all the challengers on one approval ballot and ask "Which ones do you like better than the incumbent?" It is a nice solution to the what-does-approval-mean problem.
Like you said, the threshold doesn't have to be 50%, either. It could go down, in a gradual sort-of term limit. After one term, the mayor will remain if no challenger gets over 60%. After two terms, it's 50%. After three terms, it's 40%. After four terms, it's 30%.
In the past, I haven't been sure what to do if there is no incumbent. Letting the incumbent pick a successor is an interesting solution. As is letting the largest opposition party choose someone (if the election is partisan).
Andy
On Thu, May 11, 2017 at 11:30 PM, Juho Laatu <juho.laatu@gmail.com mailto:juho.laatu@gmail.com> wrote:
That method could be said to be problematic since it is not symmetric. But on the other hand it could be an excellent method for many elections where we the asymmetry has some explanation.
If we have an incumbent president that we might or might not change, that president could be the default option. In addition to voting on who is the best candidate, we would be voting also on whether to change the president or not. If the current president is no longer a candidate, he could nominate a proposed successor who would be the default option. Or in a two-party system one could pick the (first) candidate of the opposition party as the default option, since that's the way things tend to go anyway (at least after the current president has run out of allowed terms). Or maybe a parliament or other central but wider group would pick the default option. The default option should not be a weak option but one that is a plausible winner (the current president should thus nominate a candidate of some other party as the default option if it is obvious that he is currently popular).
This method would allow also having a threshold that would be required to not elect the incumbent or default option.
This method is a nice way to solve the problem of cyclic group preferences if one can use time to distribute the victory to multiple candidates. If there is a permanent top cycle A>B>C>A, C would win A, B would win C in the next election after maybe a year etc. No need to "break" the cycle since time will (maybe) allow all the cyclic candidates to participate. They just need to wait for their turn. (Not allowing the Minmax winner outside of the top cycle to win could be considered a disadvantage of this method.)
The ballot could be simply an approval ballot, asking "which options do you prefer to the default option". This method could thus be seen also as an improved Approval method. Adding a (strong) default option to Approval makes it more sensible than it is without that default option.
This approach could offer also a smooth path away from Plurality.
Juho
On 12 May 2017, at 02:48, Kevin Venzke <stepjak@yahoo.fr mailto:stepjak@yahoo.fr> wrote:
I think this is an interesting topic... Basically how to make a decision in a meeting when you don't have ballots and people aren't inclined to talk to each other. There's probably a whole field on this topic and I'm just not aware of it...
My idea here would be one I've mentioned before (but as an election method), that somehow a default option is initially selected, and then this option wins unless a full majority thinks another option is better. The largest such majority would immediately win in that situation; we would not chase around any cycles. So, there is both an advantage and disadvantage to being the default option. The advantage is obvious. The disadvantage is that the default option could lose when, had it not been the default, it could have won.
To rephrase a bit, I'm saying that for candidates A...Z where A is the default option, you only collect the votes for B...Z against A. Nobody's voting A over anything.
By not collecting the other pairwise information you might hopefully reduce the potential to make arguments against the winner's legitimacy. (A bit like the difficulty of undermining the winner under FPP or Approval, when the only sure data available is the counts that produced the winner.)
You could pick the default option using an election method (and even have an instant-win criterion if one is sensible). The mechanism does not enforce Plurality, only minimal defense.
Sorry for lack of other responses. I can't find enough time to write.
Kevin
De : Juho Laatu <juho.laatu@gmail.com mailto:juho.laatu@gmail.com>
À : Election Methods <election-methods@lists. electorama.com mailto:election-methods@lists.electorama.com>
Cc : Kevin Venzke <stepjak@yahoo.fr mailto:stepjak@yahoo.fr>
Envoyé le : Dimanche 7 mai 2017 5h32
Objet : Re: [EM] Ordering defeats in Minimax
On 07 May 2017, at 12:39, Kristofer Munsterhjelm <km_elmet@t-online.de mailto:km_elmet@t-online.de> wrote:
On 05/07/2017 10:58 AM, Juho Laatu wrote:
On 06 May 2017, at 23:26, Kevin Venzke <stepjak@yahoo.fr mailto:stepjak@yahoo.fr
<mailto:stepjak@yahoo.fr mailto:stepjak@yahoo.fr>> wrote:
Legislators usually don't use election
methods. I would say essentially they figure out among themselves what
the outcome likely must be and then "elect" it, typically by a majority
vote. If voters could do that directly I think that would be the ideal
situation, as far as minimizing the need for strategy.
That's an interesting viewpoint. I'd like to see Condorcet tested in
some parliaments. Although negotiations and Plurality can be used to
solve many cases, also Condorcet and the idea of having multiple
candidates could have some benefits. The first one in my mind is
openness. Even when the "big boys" have agreed something behind the
screens, and then bring that decision into a majority vote, some smaller
groupings could add some alternatives in the election that they consider
better. This would open up the available alternatives to the world
(media, public), and all voters would have to take position on the
presented alternatives. Small parties or groupings could thus make
sensible compromise proposals that might even win, if they are good enough.
I think some parliaments use the following procedure for multiple proposals or amendments:
Some group decides the agenda (I'm not sure who; it probably depends on the parliament in question).
Then the parliament discusses the first two options, and does a majority vote between the two. It then does a majority vote between the winner so far and the next option, and so on until all options have been considered. The winner at the end of the procedure is the option that is accepted.
This method is Condorcet (because the CW beats everybody pairwise, and so whenever it becomes the winner so far, it'll stay the winner so far; and whenever it's the challenger, it will win against the winner so far and become the next winner so far). Furthermore, it's Smith. How cycles are broken depends on the agenda.
Yes, this is some sort of "serial Condorcet". The biggest problem of this approach is probably the agenda based cycle breaking that you mentioned.
If there is a Condorcet winner, it will win. But if there is any chance of having a cycle, then the chairman (if he decides the agenda) may arrange the agenda so that his favourite option is the last one in the queue of options. With three candidates this already determines the order of all pairwise votes.
With three candidates the cycle could be either A>B>C>A or A>C>B>A. The chairman wants A to win. The first vote will therefore be between B and C. If we have cycle A>B>C>A, B will win the first round, and A will win the second round (against B). If we have cycle A>C>B>A, C will win the first round, and A will win the second round (against C). The chairman could thus pick the preferred winner in the case that there is no Condorcet winner.
The voters could also try to use some strategies. This gives us some quite interesting strategy patterns, since after the agenda has been set, the method is no more symmetric with respect to the different options. A and B supporters could also join forces and vote against A at the last round, if they don't like the chairmans idea of constructing a strategic agenda.
Juho
Election-Methods mailing list - see http://electorama.com/em http://electorama.com/emfor list info
Election-Methods mailing list - see http://electorama.com/em http://electorama.com/em for list info
Election-Methods mailing list - see http://electorama.com/em http://electorama.com/emfor list info
I agree Approval would be good enough for the first round. But I suggest that majority approval on the first round should be an instant win. I agree in advance that that suggestion seems far from ideal. However, I'm concerned about the possibility that the second round may appear to ask a question that seemingly has already been resolved in the first round.
Example: In the first round, one candidate on the Left gets majority approval and so we take him as the default for the second round. But in the second round it could happen that candidates on the Right secure majorities against the Left candidate. (Maybe the first round result motivated the opposition to participate in greater numbers.) One could argue at the very least that the first round seems to have been pointless in that case. I wouldn't want candidates to feel there is little reason to contest the first round.
Regarding ballot format: You're right that I didn't mean it should make a difference. It's partly that I wonder what the default candidate tells his supporters to do? "Support me on election day by submitting a blank ballot"? That seems odd.
I don't have a nice name for the general mechanism. "Oppositional approval" sounds like a contradiction and "approval vs. a default option" is bulky. "Contrastive approval" produces an overused acronym... Arguably it isn't approval at all and should be named for being a "single-column" pairwise method.
Kevin
De : Juho Laatu <juho.laatu@gmail.com>
À : EM list election-methods@electorama.com
Cc : Kevin Venzke stepjak@yahoo.fr
Envoyé le : Vendredi 12 mai 2017 19h02
Objet : Re: [EM] Voting without ballots (was: Minimax)
On 13 May 2017, at 01:07, Kevin Venzke stepjak@yahoo.fr wrote:
I have concerns about dropping the threshold below 50%, even if the incumbent has been in power so long that we want a bias in favor of getting rid of him. Strategically the method starts to turn into Approval, with people voting with that mindset rather than voting as though it is a pairwise comparison. (Even at 50% this could be an issue; imagine for instance the situation that the default option is so terrible that voters expect that many candidates will hit the 50% threshold.)
Yes, the default option should be strong (one of the potential winners, or stronger). Otherwise we will have all the problems of basic Approval.
One of my main ideas is to address the area where Approval makes me nervous: The situation that the winner doesn't have a majority, but it looks like there could have been a majority among a divided opposition, if one posed the question right. I hope that clarifying the stakes would reduce the chicken dilemma potential.
Another motivation is to try to keep more options available for the last step of the election. It's easy to imagine this as a revision of the top-two runoff or jungle primary, where we take note of just the single frontrunner of the first round, and avoid eliminating the losing options for the second round.
Yes. One could use the first round just to define the default option. Approval could be good enough for the first round. The first round and the combined two round method should be simple, since if it gets complex, we might as well be use some basic Condorcet method.
I certainly think the contests against the default should use an approval ballot, but I think there should probably be an additional checkbox for those who don't want to vote for anybody as superior to the default. I just think you would want some distinction between a pro-default ballot and a blank ballot.
I guess that would be for the statistics, i.e. that difference would not influence the outcome. One (not necessarily good) approach would be to allow all options to be ticked either as "worse" or "better" than the default option.
Juho
Kevin
De : Andy Jennings <elections@jenningsstory.com>
À : Juho Laatu juho.laatu@gmail.com
Cc : Kevin Venzke stepjak@yahoo.fr; Election Methods election-methods@lists.electorama.com
Envoyé le : Vendredi 12 mai 2017 7h32
Objet : Re: [EM] Voting without ballots (was: Minimax)
I like this idea, Juho. If there is an incumbent, put all the challengers on one approval ballot and ask "Which ones do you like better than the incumbent?" It is a nice solution to the what-does-approval-mean problem.
Like you said, the threshold doesn't have to be 50%, either. It could go down, in a gradual sort-of term limit. After one term, the mayor will remain if no challenger gets over 60%. After two terms, it's 50%. After three terms, it's 40%. After four terms, it's 30%.
In the past, I haven't been sure what to do if there is no incumbent. Letting the incumbent pick a successor is an interesting solution. As is letting the largest opposition party choose someone (if the election is partisan).
Andy
On Thu, May 11, 2017 at 11:30 PM, Juho Laatu juho.laatu@gmail.com wrote:
That method could be said to be problematic since it is not symmetric. But on the other hand it could be an excellent method for many elections where we the asymmetry has some explanation.
If we have an incumbent president that we might or might not change, that president could be the default option. In addition to voting on who is the best candidate, we would be voting also on whether to change the president or not. If the current president is no longer a candidate, he could nominate a proposed successor who would be the default option. Or in a two-party system one could pick the (first) candidate of the opposition party as the default option, since that's the way things tend to go anyway (at least after the current president has run out of allowed terms). Or maybe a parliament or other central but wider group would pick the default option. The default option should not be a weak option but one that is a plausible winner (the current president should thus nominate a candidate of some other party as the default option if it is obvious that he is currently popular).
This method would allow also having a threshold that would be required to not elect the incumbent or default option.
This method is a nice way to solve the problem of cyclic group preferences if one can use time to distribute the victory to multiple candidates. If there is a permanent top cycle A>B>C>A, C would win A, B would win C in the next election after maybe a year etc. No need to "break" the cycle since time will (maybe) allow all the cyclic candidates to participate. They just need to wait for their turn. (Not allowing the Minmax winner outside of the top cycle to win could be considered a disadvantage of this method.)
The ballot could be simply an approval ballot, asking "which options do you prefer to the default option". This method could thus be seen also as an improved Approval method. Adding a (strong) default option to Approval makes it more sensible than it is without that default option.
This approach could offer also a smooth path away from Plurality.
Juho
On 12 May 2017, at 02:48, Kevin Venzke stepjak@yahoo.fr wrote:
I think this is an interesting topic... Basically how to make a decision in a meeting when you don't have ballots and people aren't inclined to talk to each other. There's probably a whole field on this topic and I'm just not aware of it...
My idea here would be one I've mentioned before (but as an election method), that somehow a default option is initially selected, and then this option wins unless a full majority thinks another option is better. The largest such majority would immediately win in that situation; we would not chase around any cycles. So, there is both an advantage and disadvantage to being the default option. The advantage is obvious. The disadvantage is that the default option could lose when, had it not been the default, it could have won.
To rephrase a bit, I'm saying that for candidates A...Z where A is the default option, you only collect the votes for B...Z against A. Nobody's voting A over anything.
By not collecting the other pairwise information you might hopefully reduce the potential to make arguments against the winner's legitimacy. (A bit like the difficulty of undermining the winner under FPP or Approval, when the only sure data available is the counts that produced the winner.)
You could pick the default option using an election method (and even have an instant-win criterion if one is sensible). The mechanism does not enforce Plurality, only minimal defense.
Sorry for lack of other responses. I can't find enough time to write.
Kevin
De : Juho Laatu <juho.laatu@gmail.com>
À : Election Methods <election-methods@lists. electorama.com>
Cc : Kevin Venzke stepjak@yahoo.fr
Envoyé le : Dimanche 7 mai 2017 5h32
Objet : Re: [EM] Ordering defeats in Minimax
On 07 May 2017, at 12:39, Kristofer Munsterhjelm km_elmet@t-online.de wrote:
On 05/07/2017 10:58 AM, Juho Laatu wrote:
On 06 May 2017, at 23:26, Kevin Venzke <stepjak@yahoo.fr
mailto:stepjak@yahoo.fr> wrote:
Legislators usually don't use election
methods. I would say essentially they figure out among themselves what
the outcome likely must be and then "elect" it, typically by a majority
vote. If voters could do that directly I think that would be the ideal
situation, as far as minimizing the need for strategy.
That's an interesting viewpoint. I'd like to see Condorcet tested in
some parliaments. Although negotiations and Plurality can be used to
solve many cases, also Condorcet and the idea of having multiple
candidates could have some benefits. The first one in my mind is
openness. Even when the "big boys" have agreed something behind the
screens, and then bring that decision into a majority vote, some smaller
groupings could add some alternatives in the election that they consider
better. This would open up the available alternatives to the world
(media, public), and all voters would have to take position on the
presented alternatives. Small parties or groupings could thus make
sensible compromise proposals that might even win, if they are good enough.
I think some parliaments use the following procedure for multiple proposals or amendments:
Some group decides the agenda (I'm not sure who; it probably depends on the parliament in question).
Then the parliament discusses the first two options, and does a majority vote between the two. It then does a majority vote between the winner so far and the next option, and so on until all options have been considered. The winner at the end of the procedure is the option that is accepted.
This method is Condorcet (because the CW beats everybody pairwise, and so whenever it becomes the winner so far, it'll stay the winner so far; and whenever it's the challenger, it will win against the winner so far and become the next winner so far). Furthermore, it's Smith. How cycles are broken depends on the agenda.
Yes, this is some sort of "serial Condorcet". The biggest problem of this approach is probably the agenda based cycle breaking that you mentioned.
If there is a Condorcet winner, it will win. But if there is any chance of having a cycle, then the chairman (if he decides the agenda) may arrange the agenda so that his favourite option is the last one in the queue of options. With three candidates this already determines the order of all pairwise votes.
With three candidates the cycle could be either A>B>C>A or A>C>B>A. The chairman wants A to win. The first vote will therefore be between B and C. If we have cycle A>B>C>A, B will win the first round, and A will win the second round (against B). If we have cycle A>C>B>A, C will win the first round, and A will win the second round (against C). The chairman could thus pick the preferred winner in the case that there is no Condorcet winner.
The voters could also try to use some strategies. This gives us some quite interesting strategy patterns, since after the agenda has been set, the method is no more symmetric with respect to the different options. A and B supporters could also join forces and vote against A at the last round, if they don't like the chairmans idea of constructing a strategic agenda.
Juho
Election-Methods mailing list - see http://electorama.com/em for list info
Election-Methods mailing list - see http://electorama.com/em for list info
Election-Methods mailing list - see http://electorama.com/em for list info
Election-Methods mailing list - see http://electorama.com/em for list info
On 13 May 2017, at 21:18, Kevin Venzke stepjak@yahoo.fr wrote:
I agree Approval would be good enough for the first round. But I suggest that majority approval on the first round should be an instant win. I agree in advance that that suggestion seems far from ideal. However, I'm concerned about the possibility that the second round may appear to ask a question that seemingly has already been resolved in the first round.
In that situation we could have only one round and a separate system (e.g. a parliament) for picking the default option. One benefit of the method is that the default option need not be the absolute favourite of the second round voters, so using some other (but reasonably similar) voter group to determine the default option is not a problem.
Example: In the first round, one candidate on the Left gets majority approval and so we take him as the default for the second round. But in the second round it could happen that candidates on the Right secure majorities against the Left candidate. (Maybe the first round result motivated the opposition to participate in greater numbers.) One could argue at the very least that the first round seems to have been pointless in that case. I wouldn't want candidates to feel there is little reason to contest the first round.
Regarding ballot format: You're right that I didn't mean it should make a difference. It's partly that I wonder what the default candidate tells his supporters to do? "Support me on election day by submitting a blank ballot"? That seems odd.
Yes, it makes sense to allow all voters to use the pen. It is also ok to collect additional statistical data in the ballots (if that doesn't add too much complexity to the ballots). One could ask e.g. how good the default candidate is in numerical scale, or why not have a numerical scale (range) for indicating how much better or worse the other candidates are. I don't expect any serious conflicts like winner having only very little support in the numeric votes.
I don't have a nice name for the general mechanism. "Oppositional approval" sounds like a contradiction and "approval vs. a default option" is bulky. "Contrastive approval" produces an overused acronym... Arguably it isn't approval at all and should be named for being a "single-column" pairwise method.
Maybe one could use words "comparison", "change", "proposal", "improve". Approval doesn't sound quite right since we are not asking for approval but comparison with respect to the default option.
Juho
Kevin
De : Juho Laatu juho.laatu@gmail.com
À : EM list election-methods@electorama.com
Cc : Kevin Venzke stepjak@yahoo.fr
Envoyé le : Vendredi 12 mai 2017 19h02
Objet : Re: [EM] Voting without ballots (was: Minimax)
On 13 May 2017, at 01:07, Kevin Venzke <stepjak@yahoo.fr mailto:stepjak@yahoo.fr> wrote:
I have concerns about dropping the threshold below 50%, even if the incumbent has been in power so long that we want a bias in favor of getting rid of him. Strategically the method starts to turn into Approval, with people voting with that mindset rather than voting as though it is a pairwise comparison. (Even at 50% this could be an issue; imagine for instance the situation that the default option is so terrible that voters expect that many candidates will hit the 50% threshold.)
Yes, the default option should be strong (one of the potential winners, or stronger). Otherwise we will have all the problems of basic Approval.
One of my main ideas is to address the area where Approval makes me nervous: The situation that the winner doesn't have a majority, but it looks like there could have been a majority among a divided opposition, if one posed the question right. I hope that clarifying the stakes would reduce the chicken dilemma potential.
Another motivation is to try to keep more options available for the last step of the election. It's easy to imagine this as a revision of the top-two runoff or jungle primary, where we take note of just the single frontrunner of the first round, and avoid eliminating the losing options for the second round.
Yes. One could use the first round just to define the default option. Approval could be good enough for the first round. The first round and the combined two round method should be simple, since if it gets complex, we might as well be use some basic Condorcet method.
I certainly think the contests against the default should use an approval ballot, but I think there should probably be an additional checkbox for those who don't want to vote for anybody as superior to the default. I just think you would want some distinction between a pro-default ballot and a blank ballot.
I guess that would be for the statistics, i.e. that difference would not influence the outcome. One (not necessarily good) approach would be to allow all options to be ticked either as "worse" or "better" than the default option.
Juho
Kevin
De : Andy Jennings <elections@jenningsstory.com mailto:elections@jenningsstory.com>
À : Juho Laatu <juho.laatu@gmail.com mailto:juho.laatu@gmail.com>
Cc : Kevin Venzke <stepjak@yahoo.fr mailto:stepjak@yahoo.fr>; Election Methods <election-methods@lists.electorama.com mailto:election-methods@lists.electorama.com>
Envoyé le : Vendredi 12 mai 2017 7h32
Objet : Re: [EM] Voting without ballots (was: Minimax)
I like this idea, Juho. If there is an incumbent, put all the challengers on one approval ballot and ask "Which ones do you like better than the incumbent?" It is a nice solution to the what-does-approval-mean problem.
Like you said, the threshold doesn't have to be 50%, either. It could go down, in a gradual sort-of term limit. After one term, the mayor will remain if no challenger gets over 60%. After two terms, it's 50%. After three terms, it's 40%. After four terms, it's 30%.
In the past, I haven't been sure what to do if there is no incumbent. Letting the incumbent pick a successor is an interesting solution. As is letting the largest opposition party choose someone (if the election is partisan).
Andy
On Thu, May 11, 2017 at 11:30 PM, Juho Laatu <juho.laatu@gmail.com mailto:juho.laatu@gmail.com> wrote:
That method could be said to be problematic since it is not symmetric. But on the other hand it could be an excellent method for many elections where we the asymmetry has some explanation.
If we have an incumbent president that we might or might not change, that president could be the default option. In addition to voting on who is the best candidate, we would be voting also on whether to change the president or not. If the current president is no longer a candidate, he could nominate a proposed successor who would be the default option. Or in a two-party system one could pick the (first) candidate of the opposition party as the default option, since that's the way things tend to go anyway (at least after the current president has run out of allowed terms). Or maybe a parliament or other central but wider group would pick the default option. The default option should not be a weak option but one that is a plausible winner (the current president should thus nominate a candidate of some other party as the default option if it is obvious that he is currently popular).
This method would allow also having a threshold that would be required to not elect the incumbent or default option.
This method is a nice way to solve the problem of cyclic group preferences if one can use time to distribute the victory to multiple candidates. If there is a permanent top cycle A>B>C>A, C would win A, B would win C in the next election after maybe a year etc. No need to "break" the cycle since time will (maybe) allow all the cyclic candidates to participate. They just need to wait for their turn. (Not allowing the Minmax winner outside of the top cycle to win could be considered a disadvantage of this method.)
The ballot could be simply an approval ballot, asking "which options do you prefer to the default option". This method could thus be seen also as an improved Approval method. Adding a (strong) default option to Approval makes it more sensible than it is without that default option.
This approach could offer also a smooth path away from Plurality.
Juho
On 12 May 2017, at 02:48, Kevin Venzke <stepjak@yahoo.fr mailto:stepjak@yahoo.fr> wrote:
I think this is an interesting topic... Basically how to make a decision in a meeting when you don't have ballots and people aren't inclined to talk to each other. There's probably a whole field on this topic and I'm just not aware of it...
My idea here would be one I've mentioned before (but as an election method), that somehow a default option is initially selected, and then this option wins unless a full majority thinks another option is better. The largest such majority would immediately win in that situation; we would not chase around any cycles. So, there is both an advantage and disadvantage to being the default option. The advantage is obvious. The disadvantage is that the default option could lose when, had it not been the default, it could have won.
To rephrase a bit, I'm saying that for candidates A...Z where A is the default option, you only collect the votes for B...Z against A. Nobody's voting A over anything.
By not collecting the other pairwise information you might hopefully reduce the potential to make arguments against the winner's legitimacy. (A bit like the difficulty of undermining the winner under FPP or Approval, when the only sure data available is the counts that produced the winner.)
You could pick the default option using an election method (and even have an instant-win criterion if one is sensible). The mechanism does not enforce Plurality, only minimal defense.
Sorry for lack of other responses. I can't find enough time to write.
Kevin
De : Juho Laatu <juho.laatu@gmail.com mailto:juho.laatu@gmail.com>
À : Election Methods <election-methods@lists. electorama.com mailto:election-methods@lists.electorama.com>
Cc : Kevin Venzke <stepjak@yahoo.fr mailto:stepjak@yahoo.fr>
Envoyé le : Dimanche 7 mai 2017 5h32
Objet : Re: [EM] Ordering defeats in Minimax
On 07 May 2017, at 12:39, Kristofer Munsterhjelm <km_elmet@t-online.de mailto:km_elmet@t-online.de> wrote:
On 05/07/2017 10:58 AM, Juho Laatu wrote:
On 06 May 2017, at 23:26, Kevin Venzke <stepjak@yahoo.fr mailto:stepjak@yahoo.fr
<mailto:stepjak@yahoo.fr mailto:stepjak@yahoo.fr>> wrote:
Legislators usually don't use election
methods. I would say essentially they figure out among themselves what
the outcome likely must be and then "elect" it, typically by a majority
vote. If voters could do that directly I think that would be the ideal
situation, as far as minimizing the need for strategy.
That's an interesting viewpoint. I'd like to see Condorcet tested in
some parliaments. Although negotiations and Plurality can be used to
solve many cases, also Condorcet and the idea of having multiple
candidates could have some benefits. The first one in my mind is
openness. Even when the "big boys" have agreed something behind the
screens, and then bring that decision into a majority vote, some smaller
groupings could add some alternatives in the election that they consider
better. This would open up the available alternatives to the world
(media, public), and all voters would have to take position on the
presented alternatives. Small parties or groupings could thus make
sensible compromise proposals that might even win, if they are good enough.
I think some parliaments use the following procedure for multiple proposals or amendments:
Some group decides the agenda (I'm not sure who; it probably depends on the parliament in question).
Then the parliament discusses the first two options, and does a majority vote between the two. It then does a majority vote between the winner so far and the next option, and so on until all options have been considered. The winner at the end of the procedure is the option that is accepted.
This method is Condorcet (because the CW beats everybody pairwise, and so whenever it becomes the winner so far, it'll stay the winner so far; and whenever it's the challenger, it will win against the winner so far and become the next winner so far). Furthermore, it's Smith. How cycles are broken depends on the agenda.
Yes, this is some sort of "serial Condorcet". The biggest problem of this approach is probably the agenda based cycle breaking that you mentioned.
If there is a Condorcet winner, it will win. But if there is any chance of having a cycle, then the chairman (if he decides the agenda) may arrange the agenda so that his favourite option is the last one in the queue of options. With three candidates this already determines the order of all pairwise votes.
With three candidates the cycle could be either A>B>C>A or A>C>B>A. The chairman wants A to win. The first vote will therefore be between B and C. If we have cycle A>B>C>A, B will win the first round, and A will win the second round (against B). If we have cycle A>C>B>A, C will win the first round, and A will win the second round (against C). The chairman could thus pick the preferred winner in the case that there is no Condorcet winner.
The voters could also try to use some strategies. This gives us some quite interesting strategy patterns, since after the agenda has been set, the method is no more symmetric with respect to the different options. A and B supporters could also join forces and vote against A at the last round, if they don't like the chairmans idea of constructing a strategic agenda.
Juho
Election-Methods mailing list - see http://electorama.com/em http://electorama.com/emfor list info
Election-Methods mailing list - see http://electorama.com/em http://electorama.com/em for list info
Election-Methods mailing list - see http://electorama.com/em http://electorama.com/emfor list info
Election-Methods mailing list - see http://electorama.com/em http://electorama.com/emfor list info