4 ms·

So many plugs for approval voting in these threads, but it just seems so terrible for what people really want to express. For example, in 2016, plenty of peopl

So many plugs for approval voting in these threads, but it just seems so terrible for what people really want to express. For example, in 2016, plenty of people wanted Sanders more than Clinton. Who are they supposed to "approve?" Many of them would note their preference by just approving Sanders and not Clinton, which is the opposite of what a voting system is supposed to help achieve.

obelos 5y ago
People voting for the only candidate they want just doesn’t strike me as a failure. The failure comes when someone has to vote for someone who is not their preferred. That said, the 2016 Dem primary doesn’t seem like a useful example for testing the constraints of any voting system since there were really only two viable candidates. All methods work equally well under those conditions. Why would someone approve both of the only two candidates when only one of them could win?
- lostdog 5y ago
  I was talking about the general election not the primary. (If we used a decent voting system, then all popular candidates could go to the general).
adjkant 5y ago
Exactly. What about if it was Sanders, Clinton, Kasich, and Trump? As a far left voter, do I have to approve Kasich if I really don't want Trump to win? In RCV, it's simple: Sanders, Clinton, maybe Kasich to ensure Trump doesn't win. There are 100% problems in IRV that Clinton or Kasich may be the first elimination there, meaning moderates can get screwed even with preference, but it's still way better than Approval IMO. The answer is a better RCV system than IVF, but let's start by getting rid of FPTP. I'd love to see some nuanced RCV methods that combine IRV, Ranked Pairs, Condorcet Winners, etc to basically improve IRV eliminations and take "shortcuts" like steps that consider what % of ballots you're on. Basically, approval voting has its place, but it should be within RCV. Does it make understanding the tally harder for the layman? Yes. But if designed right, it simplifies the voting end so that people can essentially approval and rank vote together. Rank everyone you approve/would accept, and rank in your true preference order. That's what we should be solving for IMO.
- postalrat 5y ago
  Vote Sanders and Clinton. Or Sanders, Clinton, and Kasich. Or just Sanders. Vote for as many as you approve of. Simple.
  - bootlooped 5y ago
    What's not simple is where to draw the approve/disapprove line.
    - postalrat 5y ago
      You could be honest and pick the ones you approve of. Or if you just dont want some candidate(s) to win pick everyone else.
      - bootlooped 5y ago
        I don't think it's as simple of a question as you suggest. Think of a food you disapprove of. Ok, now if you're stranded on a desert island, do you approve of it now? If you're at your usual grocery store you might disapprove of certain foods, but if you have to go to a much smaller one you may switch some of those to approve.
- cosmojg 5y ago
  > Basically, approval voting has its place, but it should be within RCV. An "approval-within-RCV" system would replicate multiple score voting (aka range voting). Why not skip the charade and advocate for a score voting system like STAR instead?
- ClayShentrup 5y ago
  Approval voting is superior to IRV in every way we can measure. https://electionscience.org/library/approval-voting-versus-irv/ https://electionscience.org/library/approval-voting-versus-i... Particularly noteworthy is their relative performance in computer simulations of accuracy against tactical voting. https://www.rangevoting.org/BayRegFig https://www.rangevoting.org/BayRegFig
ajmurmann 5y ago
That's where something like STAR voting is even better. Just score people 0 to 5 on how much you like them. Gets rid of strategic voting, spoiler effect and works super well for large sets of choices as well. It's in essence like we score movies or restaurant reviews which makes it familiar as well.
- lostdog 5y ago
  I would take STAR voting! The only downside is having to hear politicians ask for 5 stars in every speech!
- MichaelApproved 5y ago
  Let’s be honest, it’s hard enough getting Americans to the ballot on Election Day, we don’t need to introduce more complexity. Ranking requires far less brain power than a rating system would. Maybe the energy expenditure between the two is small for the HN group but for the general public, the aversion to doing the mental work would be enough to dramatically reduce turnout. This goes for all countries, not just America. Not to mention how easy it is to spread lies for something as simplistic as FPTP. The more complicated a system is, the easier it’ll be to spread lies about it.
  - ajmurmann 5y ago
    I'm not sure why you claim that ranking requires less brain power. Ranking becomes especially burdensome as you have more candidates/options. On top of that ranked voting seems much simpler on the surface than it actually is. The actual calculation is quite opaque in comparison.
  - goatse-4-u 5y ago
    .
  - ClayShentrup 5y ago
    > Ranking requires far less brain power than a rating system would. Rating is simpler than ranking in every way we can measure, from rates of ballot spoilage, to average time taken to complete a ballot. https://medium.com/election-science/star-voting-is-simpler-than-irv-84b8990986f2 https://medium.com/election-science/star-voting-is-simpler-t... https://www.rangevoting.org/Complexity https://www.rangevoting.org/Complexity Most obviously, ranking is effectively a bubble sort, О(n^2), whereas rating is just two passes: 1. Get the max and min. 2. Normalize, thus O(n). And Approval Voting, technically rating on a 0-1 binary scale, is arguably even simpler than the status quo.
aimor 5y ago
They are supposed to approve of each candidate individually. This avoids problems such as the one you allude to in your example where Sanders may lose to Clinton despite more total people approving of Sanders, simply because Clinton supporters (the larger group) rank Clinton>Sanders while Sanders supporters (a minority of voters) only rank Sanders.
ClayShentrup 5y ago
> it just seems so terrible for what people really want to express. Intuition would say so, but intuition is wrong. https://electionscience.org/library/expressiveness-in-approval-vs-ranked-ballots/ https://electionscience.org/library/expressiveness-in-approv... > Who are they supposed to "approve?" Everyone they prefer to the expected utility of the winner. https://www.rangevoting.org/RVstrat6 https://www.rangevoting.org/RVstrat6 And indeed, Sanders (and Warren) did quite well in multiple approval voting polls like this one. https://electionscience.org/press-releases/new-poll-74-of-democratic-primary-voters-would-support-warren-for-president/ https://electionscience.org/press-releases/new-poll-74-of-de...

obelos 5y ago

People voting for the only candidate they want just doesn’t strike me as a failure. The failure comes when someone has to vote for someone who is not their preferred. That said, the 2016 Dem primary doesn’t seem like a useful example for testing the constraints of any voting system since there were really only two viable candidates. All methods work equally well under those conditions. Why would someone approve both of the only two candidates when only one of them could win?

lostdog 5y ago
I was talking about the general election not the primary. (If we used a decent voting system, then all popular candidates could go to the general).

adjkant 5y ago

Exactly. What about if it was Sanders, Clinton, Kasich, and Trump? As a far left voter, do I have to approve Kasich if I really don't want Trump to win? In RCV, it's simple: Sanders, Clinton, maybe Kasich to ensure Trump doesn't win. There are 100% problems in IRV that Clinton or Kasich may be the first elimination there, meaning moderates can get screwed even with preference, but it's still way better than Approval IMO. The answer is a better RCV system than IVF, but let's start by getting rid of FPTP. I'd love to see some nuanced RCV methods that combine IRV, Ranked Pairs, Condorcet Winners, etc to basically improve IRV eliminations and take "shortcuts" like steps that consider what % of ballots you're on. Basically, approval voting has its place, but it should be within RCV. Does it make understanding the tally harder for the layman? Yes. But if designed right, it simplifies the voting end so that people can essentially approval and rank vote together. Rank everyone you approve/would accept, and rank in your true preference order. That's what we should be solving for IMO.

postalrat 5y ago
Vote Sanders and Clinton. Or Sanders, Clinton, and Kasich. Or just Sanders. Vote for as many as you approve of. Simple.
- bootlooped 5y ago
  What's not simple is where to draw the approve/disapprove line.
  - postalrat 5y ago
    You could be honest and pick the ones you approve of. Or if you just dont want some candidate(s) to win pick everyone else.
    - bootlooped 5y ago
      I don't think it's as simple of a question as you suggest. Think of a food you disapprove of. Ok, now if you're stranded on a desert island, do you approve of it now? If you're at your usual grocery store you might disapprove of certain foods, but if you have to go to a much smaller one you may switch some of those to approve.
cosmojg 5y ago
> Basically, approval voting has its place, but it should be within RCV. An "approval-within-RCV" system would replicate multiple score voting (aka range voting). Why not skip the charade and advocate for a score voting system like STAR instead?
ClayShentrup 5y ago
Approval voting is superior to IRV in every way we can measure. https://electionscience.org/library/approval-voting-versus-irv/ https://electionscience.org/library/approval-voting-versus-i... Particularly noteworthy is their relative performance in computer simulations of accuracy against tactical voting. https://www.rangevoting.org/BayRegFig https://www.rangevoting.org/BayRegFig

ajmurmann 5y ago

That's where something like STAR voting is even better. Just score people 0 to 5 on how much you like them. Gets rid of strategic voting, spoiler effect and works super well for large sets of choices as well. It's in essence like we score movies or restaurant reviews which makes it familiar as well.

lostdog 5y ago
I would take STAR voting! The only downside is having to hear politicians ask for 5 stars in every speech!
MichaelApproved 5y ago
Let’s be honest, it’s hard enough getting Americans to the ballot on Election Day, we don’t need to introduce more complexity. Ranking requires far less brain power than a rating system would. Maybe the energy expenditure between the two is small for the HN group but for the general public, the aversion to doing the mental work would be enough to dramatically reduce turnout. This goes for all countries, not just America. Not to mention how easy it is to spread lies for something as simplistic as FPTP. The more complicated a system is, the easier it’ll be to spread lies about it.
- ajmurmann 5y ago
  I'm not sure why you claim that ranking requires less brain power. Ranking becomes especially burdensome as you have more candidates/options. On top of that ranked voting seems much simpler on the surface than it actually is. The actual calculation is quite opaque in comparison.
- goatse-4-u 5y ago
  .
- ClayShentrup 5y ago
  > Ranking requires far less brain power than a rating system would. Rating is simpler than ranking in every way we can measure, from rates of ballot spoilage, to average time taken to complete a ballot. https://medium.com/election-science/star-voting-is-simpler-than-irv-84b8990986f2 https://medium.com/election-science/star-voting-is-simpler-t... https://www.rangevoting.org/Complexity https://www.rangevoting.org/Complexity Most obviously, ranking is effectively a bubble sort, О(n^2), whereas rating is just two passes: 1. Get the max and min. 2. Normalize, thus O(n). And Approval Voting, technically rating on a 0-1 binary scale, is arguably even simpler than the status quo.

aimor 5y ago

They are supposed to approve of each candidate individually. This avoids problems such as the one you allude to in your example where Sanders may lose to Clinton despite more total people approving of Sanders, simply because Clinton supporters (the larger group) rank Clinton>Sanders while Sanders supporters (a minority of voters) only rank Sanders.

ClayShentrup 5y ago

> it just seems so terrible for what people really want to express. Intuition would say so, but intuition is wrong. https://electionscience.org/library/expressiveness-in-approval-vs-ranked-ballots/ https://electionscience.org/library/expressiveness-in-approv... > Who are they supposed to "approve?" Everyone they prefer to the expected utility of the winner. https://www.rangevoting.org/RVstrat6 https://www.rangevoting.org/RVstrat6 And indeed, Sanders (and Warren) did quite well in multiple approval voting polls like this one. https://electionscience.org/press-releases/new-poll-74-of-democratic-primary-voters-would-support-warren-for-president/ https://electionscience.org/press-releases/new-poll-74-of-de...