Vote for 2 then Instant Runoff



  • @robertpdx
    I've been trying to find a simple way let voters choose between "later no harm" and "no friend betrayal" while maintaining preferences between candidates. Here is the result:
    http://www.classicalmatter.org/Election Science/Vote2ThenInstantRunoff.html
    Constructive feedback would be welcome. I am new at this.



  • "Approval voting allows only binary levels of support rather than a range of preferences." For large elections, I disagree, because the voter can:

    • decide her strategic vote as though faced with a fine-grained Score election.
    • normalize the scores to the probability-like range [0, 1].
    • approve the candidate(s) whose score is 1.
    • disapprove the candidate(s) whose score is zero.
    • for the remaining candidate(s), consult a source of random numbers and approve the candidate with the probability given by that candidate's score.

    If many voters do this, their effect on each candidate's total will be as though fine-grained Score were in effect.

    If you say voters won't do this, I respond that not doing it would amount to voluntarily giving up voting power. Experience with FPtP shows that voters do not voluntarily give up power.



  • I believe the system doesn't provide balance.



  • @Jack-Waugh I am not questioning voters' understanding of probability theory. I am saying that for each candidate, a voter has two choices: approve or not approve. That fits the definition of a binary choice.



  • @Jack-Waugh There is "tit-for tat" balance in the runoff stage but not in the "vote for 2" stage. Do you think that would cause voter dissatisfaction?



  • @robertpdx said in Vote for 2 then Instant Runoff:

    @Jack-Waugh I am not questioning voters' understanding of probability theory. I am saying that for each candidate, a voter has two choices: approve or not approve. That fits the definition of a binary choice.

    But in practice, their choice is not binary. They can choose to use a randomized process, which is a different choice than a definite decision to approve or not approve, with a different outcome overall. Consider for example an Approval election with candidates Nader, Bush, and Gore. If the Nader supporters all approve Gore, Nader cannot win. But if they approve Gore at .99 relative frequency, Nader can win.



  • @robertpdx said in Vote for 2 then Instant Runoff:

    @Jack-Waugh There is "tit-for tat" balance in the runoff stage but not in the "vote for 2" stage. Do you think that would cause voter dissatisfaction?

    When people ask me that, I ask them for their analysis of why FPtP elects oligarchs.



  • @Jack-Waugh I'm not happy with FPtP either. That's why I got interested in this field. But I am trying to be practical rather than theoretical. "Vote for 2" is a standard, well-accepted method for choosing two winners. And a runoff is the standard, well-accepted method for deciding between two candidates. So I propose to combine those two steps into a single election. I think the result is demonstrably better than RCV/IRV since 2nd-choice votes are counted from the start. I wouldn't say that this method is necessarily better than scored or STAR voting (or approval voting), but it does have the advantage that voters can express preferences between additional candidates without harming their first choices. There should not be a cost for voters to participate in the runoff stage.
    Also, I have previously read that a hand count of RCV/IRV requires re-reading ballots to perform the iterative elimination scheme. That can be avoided using the table that I proposed, although I suppose it is debatable if that would be faster or more reliable.



  • @robertpdx said in Vote for 2 then Instant Runoff:

    "Vote for 2" is a standard, well-accepted method for choosing two winners. And a runoff is the standard, well-accepted method for deciding between two candidates. So I propose to combine those two steps into a single election.

    Right, well, that's an interesting point. If Vote for 2 then Instant Runoff can be sold to Joe Sixpack on account of its being based on well-accepted methods, but Approval cannot, because it is novel, then getting something better than FPtP/IRV in place may be better than nothing, in helping to replace the omnicidal regime with a people's regime.


Log in to reply