Disapproval voting may seem functionally equivalent to approval voting. However, disapproval voting would make it more difficult to "bullet vote" because that would require disapproving every candidate except one's favorite. It would also be more awkward for candidates to urge their supporters to disapprove every candidate but themselves. One promising way to narrow down large fields of candidates may be to combine approval with disapproval. Voters could approve any candidates they found at least somewhat acceptable and disapprove candidates they found unacceptable. Many voters would leave candidates they were uncertain about unmarked. Furthermore, if voters were given short descriptions of all candidates and their political views, most voters would likely not have much trouble deciding which candidates were acceptable, which ones were unacceptable, and which ones they were uncertain about. If I were me, I would disapprove not only candidates I found clearly unacceptable but also candidates who described themselves so vaguely as to make it impossible to make a reasonable guess as to how they would perform if elected. Any thoughts? -Ralph Suter > Date: Fri, 25 Aug 2023 22:51:40 +0200 > From: Colin Champion <colin.champion@routemaster.app> > To: EM <election-methods@lists.electorama.com> > Subject: Re: [EM] Condorcet meeting > Message-ID: <d337b1f5-45e6-e7fa-7f78-0c28cddce0e7@routemaster.app> > Content-Type: text/plain; charset="utf-8"; Format="flowed" > > I?m not persuaded of approval voting. My guess is that voters will > bullet-vote for the candidate they like best among those they know > about; candidates will encourage bullet voting on their own behalf; > pundits will have nothing better to say, and voters will have no motive > to award more than the minimum number of approvals. > ?? A danger of quadratic and entropic measures is that they don't > impose the constraint that the number of survivors has to be small > enough to make ranked voting effective on the second round. > ?? CJC > > On 25/08/2023 15:05, Kristofer Munsterhjelm wrote: >> On 2023-08-25 01:50, Forest Simmons wrote: >>> I agree with Kristofer that Approval is plenty good for the narrowing >>> down phase. >>> >>> Your favorite pundits and candidates will definitely make known their >>> recommendations.? Trust your own judgment and gut, as you collate and >>> cull out their llists of recommendations. >>> >>> If there are going to be only six finalists, that doesn't mean you >>> can only approve six or that you have to approve more than one. >>> >>> My rule is to approve my favorite as well as everybody else that I >>> like almost as much. >>> >>> Here's an idea for deciding on n, the number of finalists after the >>> approval ballots have been tallied: >>> >>> For this purpose, temporarily count the ballots fractionally, and let >>> f(X) be the fraction of the total that X gets in this tally ... so >>> that the f(X) values sum to unity. >>> >>> The value of n should be the reciprocal of the sum of the squares of >>> the f(X) vslues... the standard formula for the minimum number of >>> seats that would be acceptable for proportional representation of a >>> diverse population. >> Another option is to use the exponential of the Shannon entropy: >> https://electowiki.org/wiki/Effective_number_of_parties#Entropy_measure >> >> -km >

I think I prefer approval/disapproval to plain approval, since it elicits more information from voters and they will be delighted to get their antipathies off their chest.    CJC On 26/08/2023 19:46, Ralph Suter wrote: > Disapproval voting may seem functionally equivalent to approval > voting. However, disapproval voting would make it more difficult to > "bullet vote" because that would require disapproving every candidate > except one's favorite. It would also be more awkward for candidates to > urge their supporters to disapprove every candidate but themselves. > > One promising way to narrow down large fields of candidates may be to > combine approval with disapproval. Voters could approve any candidates > they found at least somewhat acceptable and disapprove candidates they > found unacceptable. Many voters would leave candidates they were > uncertain about unmarked. Furthermore, if voters were given short > descriptions of all candidates and their political views, most voters > would likely not have much trouble deciding which candidates were > acceptable, which ones were unacceptable, and which ones they were > uncertain about. If I were me, I would disapprove not only candidates > I found clearly unacceptable but also candidates who described > themselves so vaguely as to make it impossible to make a reasonable > guess as to how they would perform if elected. > > Any thoughts? > > -Ralph Suter >> Date: Fri, 25 Aug 2023 22:51:40 +0200 >> From: Colin Champion <colin.champion@routemaster.app> >> To: EM <election-methods@lists.electorama.com> >> Subject: Re: [EM] Condorcet meeting >> Message-ID: <d337b1f5-45e6-e7fa-7f78-0c28cddce0e7@routemaster.app> >> Content-Type: text/plain; charset="utf-8"; Format="flowed" >> >> I?m not persuaded of approval voting. My guess is that voters will >> bullet-vote for the candidate they like best among those they know >> about; candidates will encourage bullet voting on their own behalf; >> pundits will have nothing better to say, and voters will have no motive >> to award more than the minimum number of approvals. >> ?? A danger of quadratic and entropic measures is that they don't >> impose the constraint that the number of survivors has to be small >> enough to make ranked voting effective on the second round. >> ?? CJC >> >> On 25/08/2023 15:05, Kristofer Munsterhjelm wrote: >>> On 2023-08-25 01:50, Forest Simmons wrote: >>>> I agree with Kristofer that Approval is plenty good for the narrowing >>>> down phase. >>>> >>>> Your favorite pundits and candidates will definitely make known their >>>> recommendations.? Trust your own judgment and gut, as you collate and >>>> cull out their llists of recommendations. >>>> >>>> If there are going to be only six finalists, that doesn't mean you >>>> can only approve six or that you have to approve more than one. >>>> >>>> My rule is to approve my favorite as well as everybody else that I >>>> like almost as much. >>>> >>>> Here's an idea for deciding on n, the number of finalists after the >>>> approval ballots have been tallied: >>>> >>>> For this purpose, temporarily count the ballots fractionally, and let >>>> f(X) be the fraction of the total that X gets in this tally ... so >>>> that the f(X) values sum to unity. >>>> >>>> The value of n should be the reciprocal of the sum of the squares of >>>> the f(X) vslues... the standard formula for the minimum number of >>>> seats that would be acceptable for proportional representation of a >>>> diverse population. >>> Another option is to use the exponential of the Shannon entropy: >>> https://electowiki.org/wiki/Effective_number_of_parties#Entropy_measure >>> >>> -km >>

On second thoughts, would this not lead to tactical voting? I imagine that a voter who both approved and disapproved would have more influence than one who only approved or only disapproved.    CJC On 27/08/2023 09:52, Colin Champion wrote: > I think I prefer approval/disapproval to plain approval, since it > elicits more information from voters and they will be delighted to get > their antipathies off their chest. >    CJC > > On 26/08/2023 19:46, Ralph Suter wrote: >> Disapproval voting may seem functionally equivalent to approval >> voting. However, disapproval voting would make it more difficult to >> "bullet vote" because that would require disapproving every candidate >> except one's favorite. It would also be more awkward for candidates >> to urge their supporters to disapprove every candidate but themselves. >> >> One promising way to narrow down large fields of candidates may be to >> combine approval with disapproval. Voters could approve any >> candidates they found at least somewhat acceptable and disapprove >> candidates they found unacceptable. Many voters would leave >> candidates they were uncertain about unmarked. Furthermore, if voters >> were given short descriptions of all candidates and their political >> views, most voters would likely not have much trouble deciding which >> candidates were acceptable, which ones were unacceptable, and which >> ones they were uncertain about. If I were me, I would disapprove not >> only candidates I found clearly unacceptable but also candidates who >> described themselves so vaguely as to make it impossible to make a >> reasonable guess as to how they would perform if elected. >> >> Any thoughts? >> >> -Ralph Suter >>> Date: Fri, 25 Aug 2023 22:51:40 +0200 >>> From: Colin Champion <colin.champion@routemaster.app> >>> To: EM <election-methods@lists.electorama.com> >>> Subject: Re: [EM] Condorcet meeting >>> Message-ID: <d337b1f5-45e6-e7fa-7f78-0c28cddce0e7@routemaster.app> >>> Content-Type: text/plain; charset="utf-8"; Format="flowed" >>> >>> I?m not persuaded of approval voting. My guess is that voters will >>> bullet-vote for the candidate they like best among those they know >>> about; candidates will encourage bullet voting on their own behalf; >>> pundits will have nothing better to say, and voters will have no motive >>> to award more than the minimum number of approvals. >>> ?? A danger of quadratic and entropic measures is that they don't >>> impose the constraint that the number of survivors has to be small >>> enough to make ranked voting effective on the second round. >>> ?? CJC >>> >>> On 25/08/2023 15:05, Kristofer Munsterhjelm wrote: >>>> On 2023-08-25 01:50, Forest Simmons wrote: >>>>> I agree with Kristofer that Approval is plenty good for the narrowing >>>>> down phase. >>>>> >>>>> Your favorite pundits and candidates will definitely make known their >>>>> recommendations.? Trust your own judgment and gut, as you collate and >>>>> cull out their llists of recommendations. >>>>> >>>>> If there are going to be only six finalists, that doesn't mean you >>>>> can only approve six or that you have to approve more than one. >>>>> >>>>> My rule is to approve my favorite as well as everybody else that I >>>>> like almost as much. >>>>> >>>>> Here's an idea for deciding on n, the number of finalists after the >>>>> approval ballots have been tallied: >>>>> >>>>> For this purpose, temporarily count the ballots fractionally, and let >>>>> f(X) be the fraction of the total that X gets in this tally ... so >>>>> that the f(X) values sum to unity. >>>>> >>>>> The value of n should be the reciprocal of the sum of the squares of >>>>> the f(X) vslues... the standard formula for the minimum number of >>>>> seats that would be acceptable for proportional representation of a >>>>> diverse population. >>>> Another option is to use the exponential of the Shannon entropy: >>>> https://electowiki.org/wiki/Effective_number_of_parties#Entropy_measure >>>> >>>> >>>> -km >>>

On 2023-08-27 15:55, Colin Champion wrote: > On second thoughts, would this not lead to tactical voting? I imagine > that a voter who both approved and disapproved would have more influence > than one who only approved or only disapproved. Yes; for every Range-style single-winner voting method, the optimal strategy is to give everybody above some threshold rank the highest possible score, and everybody below it the lowest possible score. (This doesn't necessarily hold for a proportional multiwinner method, though, even if it reduces to Range in the single-winner case.) -km