Navigation

    Voting Theory Forum

    • Register
    • Login
    • Search
    • Recent
    • Categories
    • Tags
    • Popular
    • Users
    • Groups

    Advocacy Tailored to Location

    Advocacy
    2
    5
    361
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • J
      Jack Waugh last edited by Jack Waugh

      I'm suggesting this strategy for changing the laws around single-winner elections:

      for a given locality, if they have already taken on the expense of the logistics of IRV, then promote Score{1, .99, .01, 0} else promote Approval.

      Approval-ordered Llull (letter grades) [10], Score // Llull [9], Score, STAR, Approval, other rated Condorcet [8]; equal-ranked Condorcet [4]; strictly-ranked Condorcet [3]; everything else [0].

      J 2 Replies Last reply Reply Quote 0
      • J
        Jack Waugh @Jack Waugh last edited by Jack Waugh

        For non-equally-spaced Score schemes, I'm suggesting the following system of letter grades:

        • A = 1
        • B = .99
        • C = .9
        • D = .5
        • E = .1
        • F = .01
        • G = 0

        A system omitting C, D, and E would use letter grades ABFG. The middle letters would be reserved for possible expansion of the system for if people complained that it had insufficiently fine granularity.

        Approval-ordered Llull (letter grades) [10], Score // Llull [9], Score, STAR, Approval, other rated Condorcet [8]; equal-ranked Condorcet [4]; strictly-ranked Condorcet [3]; everything else [0].

        1 Reply Last reply Reply Quote 0
        • N
          NevinBR last edited by NevinBR

          @Jack-Waugh said in Advocacy Tailored to Location:

          for a given locality, if they have already taken on the expense of the logistics of IRV, then promote Score{1, .99, .01, 0} else promote Approval.

          Without detracting from your goals here, I think it is worth considering a much smaller incremental change for places that already have IRV, which nonetheless solves significant problems and improves outcomes of elections.

          The IRV halting condition is: “If one of the remaining candidates has more than half of the remaining votes, they win.”

          That could be modified to: “If one of the remaining candidates would defeat all of the others head-to-head, they win.”

          This small change turns IRV into not only a Condorcet method, but in fact a Smith method. The winner is guaranteed to be in the topologically highest strongly-connected-component of the pairwise results graph.

          From an advocacy and education perspective, I suggest using the term “clear winner” in place of “Condorcet winner”. Then the new halting condition can be expressed succinctly as, “If one of the remaining candidates is a clear winner, they win.”

          This phrasing also makes it easier to explain the problem being solved: the standard IRV method can fail to elect a clear winner. With this change to the halting condition, we can guarantee that won’t happen.

          J 1 Reply Last reply Reply Quote 0
          • J
            Jack Waugh @NevinBR last edited by

            @NevinBR, where I said "they have already taken on the expense of the logistics of IRV", I should have said "they have already taken on the expense of the logistics of IRV, but have repealed IRV."

            Approval-ordered Llull (letter grades) [10], Score // Llull [9], Score, STAR, Approval, other rated Condorcet [8]; equal-ranked Condorcet [4]; strictly-ranked Condorcet [3]; everything else [0].

            1 Reply Last reply Reply Quote 1
            • J
              Jack Waugh @Jack Waugh last edited by Jack Waugh

              I'm becoming more and more convinced that people can't be convinced that Approval is expressive enough. I think it probably is, but it's extremely hard to get across. And I am not 100% sure that Approval is really as expressive as I have been arguing that it is. Maybe there is a hole in the argument that I have been advancing. So I'm thinking we should advocate for Score with finer granularity in the single-winner cases, unless the local activists argue that it's just a nonstarter logistically. How about using a range with seven values? I suspect that spacing them according to a logistic function would work best strategically, but that's impossible to explain to people, so how about seven values equally spaced then?

              Approval-ordered Llull (letter grades) [10], Score // Llull [9], Score, STAR, Approval, other rated Condorcet [8]; equal-ranked Condorcet [4]; strictly-ranked Condorcet [3]; everything else [0].

              1 Reply Last reply Reply Quote 0
              • First post
                Last post