• Steve@communick.news
      link
      fedilink
      English
      arrow-up
      51
      arrow-down
      3
      ·
      5 months ago

      Just be careful you’re not letting the perfect be the enemy of the good.

      FPTP is absolutely the worst. Anything else is a massive improvement.

      • Deceptichum@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        17
        arrow-down
        9
        ·
        edit-2
        5 months ago

        Do it right and do it once, don’t half-arse it and pretend you’ll get around to finishing it off later.

        Momentum is hard to get, be careful aiming for the lesser option because you may not be able to ever get the public to care enough again to push for the better option.

        tl;dr: 1 change is easier to pull off than 2

        • Null User ObjectOP
          link
          fedilink
          arrow-up
          14
          arrow-down
          1
          ·
          edit-2
          5 months ago

          That’s the very definition of letting perfect be the enemy of good. We can have really good now, or we can debate ad nauseum for decades about what would be perfect, never reach an agreement, and have done nothing.

        • intelisense@lemm.ee
          link
          fedilink
          arrow-up
          12
          arrow-down
          2
          ·
          5 months ago

          Wrong. We did exactly what you suggest in the U.K. Any kind of PR is off the table for at least a generation here.

    • Maggoty@lemmy.world
      link
      fedilink
      arrow-up
      22
      ·
      5 months ago

      RCV is leagues ahead of FPTP. Star is literally just FPTP with a run off. And Ranked Robin is complex.

      That’s why RCV keeps getting the nod. It’s not the best scientific system, it’s the one at the crux of improvement and ease of understanding.

    • intelisense@lemm.ee
      link
      fedilink
      arrow-up
      20
      ·
      5 months ago

      The last time you posted this, I commented about all the reddit users that disappeared the day after the vote. I hate repeating myself, so here is a different anecdote.

      During the campaign, many people - including yours truly - pointed out that if the admitedly mediocre change was rejected, the powers that be would argue that ‘the people’ had voted against change - any change. And that is exactly what happened https://www.bbc.com/news/uk-politics-13297573.

      …the referendum had delivered a “resounding answer that settles the question” over electoral change and people now wanted the government…

      David Cameron (Prime Minister)

      I personally believe that this result will settle the debate over changing our electoral system for the next generation.

      Matthew Eliot (Leader of the No campaign)

      You get exactly one chance at this. As Ru Paul likes to say, Do Not Fuck It Up.

    • nondescripthandle@lemmy.dbzer0.com
      link
      fedilink
      arrow-up
      17
      ·
      edit-2
      5 months ago

      STAR voting encourages single candidate votes too much if you’re voting for the underdog. Lets say it’s Biden Trump and Sanders. If I want Bernie, assuming hes the underdog to Biden, I have incentive to not rank Biden at all because if I do he gets points that could push him ahead of Bernie. Its in my best interest to make sure Bernie gets the biggest point differential from my ballot which means only voting for Bernie and no one else.

      So for anyone who’s going to vote what would be third party, the best they can do is not rank anyone else, defeating the whole point and ensuring what people call the spoiler effect will still exist.

      ALL the systems you proposed have problems, not just RCV. You have to pick your poison so to speak.

    • Cornelius_Wangenheim@lemmy.world
      link
      fedilink
      arrow-up
      2
      arrow-down
      1
      ·
      5 months ago

      The problem with any scoring style system is that there is no answer for when a person gives 2 finalists the same score. They want to vote, but the system did not gather enough information to know where to allocate it.