Posts made by Jack Waugh

How should ABC voting treat unmarked candidates?

Who knows how the Gibbard theorem applies to ABC voting? In optimizing my vote, how do I take into account the stances of the other voters? Assume I know them perfectly. Do I maybe exaggerate support for a compromise candidate from D to C, with a metered probability?

which leads to https://github.com/Election-Tech-Initiative/electionguard/blob/main/docs/concepts/Verifiability.md

Suppose a tallying algo is enacted that requires a preference matrix. What grounds could be cited to convince the public that each precinct correctly sums up the preferences in the votes to build the precinct-level addend to the preference matrix?

I first heard of ABC Voting when Beloved Comrade @Ex-dente-leonem posted about it and another system (Score B2R). I think that ABC Voting is so interesting that I'm starting the present post dedicated to just it.

I'm hereby running up the flagpole an alternative name IRV-Llull in contrast to IRV-Ware.

I heard arguments from two individuals who push for three-valued Score with the default being in the middle and the numbers set so that the middle is zero and the bottom is -1. Both of these advocates opine that voters need an explicit way to express impassioned opposition to a candidate. I think they may be right about that need, but I don't like their proposed solution. I hereby suggest that ABC voting fulfills that need not only psychologically, but better, by acting in the tally in a way that honors that impassioned opposition to the max. In this system, a voter can clearly oppose a candidate with any of the grades D, E, or F, because these grades deny the candidate a positive point for the initial ordering. Within that, the system still provides a way to express a preference for the lesser evil over the greater evil, and it provides that without compromising the effect of the voter's expression concerning the voter's preferred candidates, whom the voter will naturally place them up in the A, B, and C region. For these reasons, I want to sing the praises of this system.

For readers coming on this system here for the first time, I'll repeat how it works:

• Voters assign A, B, C, D, E or F to each candidate. Unmentioned candidates, I suggest, get D. This is my sop to those who think such should get the middle in a score system.

• A, B, or C confers a point of tolerance for the candidate; D, E, and F do not.

• The tally starts by arranging the candidates in order of how many tolerating votes they got, with the most tolerated candidate by that measure at the top of the list.

• The Llull stage of the tally begins, as though the candidates had entered the church in the order determined above (or maybe it's the reverse of that order). In any event, the bottom two candidates on the list are compared first, according to how many voters expressed a preference for one over the other minus how many expressed the opposite preference. A candidate who receives an "E" from a given voter, for example, is understood to be preferred by that voter over a candidate who receives an "F". Whichever candidate loses that comparison is removed from the list.

• The bottom-two comparisons according to voter counts who prefer one candidate over the other vs. the opposite preference are repeated until only one candidate remains on the list. That is the winner.

@cfrank I agree that for purposes of "one person, one vote", two approvals or two disapprovals should not count as two votes. However, legislators and judges and juries seem to lack for an ability to think straight.

Someone responded to me by pointing out that voting for-or-against is horrible. I would respond to that response instead of posting a new topic, but the search mechanism in this forum software (NodeBB) doesn't seem to be of much help in digging that conversation up again. And anyway, I want to promote my position to a larger audience if I can.

I predict that the benefits of for-or-against as opposed to FPtP would be revolutionary. I know how I would vote in it for President of the United Snakes of America in 2024.

In the Green-party forum I recently posted about, one of the listeners mentioned a State in which she or he said that the law prohibits voting systems that allow more than one "vote" per voter. This would probably outlaw then Approval and the like. Vote For-or-against would meet that constraint.

When grift organizations e-mail me for donations and mention "democracy", I respond along the following lines. In most cases, no one responds to my response. Their e-mail is only in broadcast mode.

You need to campaign for decoupling money from electoral outcomes via changing to a voting system that gives equal weight to the voters, one voter to another, instead of First Past the Post or RCV-Ware.

• https://www.rangevoting.org/Cash3.html
• https://1787regime.wordpress.com/2024/04/07/not-allowed-to-vote-ones-values/

Equality advocates should pile on, and bring references to systems that are better than strictly-ranked Ware and be ready to explain why they are better and cost no more.

https://mailchi.mp/cb90c9c0ae48/green-party-of-st-louis-meeting-wed-april-19-2017-recent-and-future-elections-6417718?e=321ac5f05b

I think I like it.

ABC || DEF

A Stein
B Williamson
C West
D Kennedy
E Harris
F Trump

@wolftune, any number of times, I have responded to PropagandaName's Tweets in favor of not-further-specified RCV by asking "Why do you oppose bottom-two runoff?". There has not been much response. Maybe more voices repeating the same question would help draw attention to it.

If equal-ranking is permitted at the bottom (e. g., by leaving candidates unmarked) but not permitted at the top, then the Frohnmayer equality constraint is violated.

@casimir, what are the grounds to say that balance, as a sine qua non requirement, is TOOOOOOOOO restrictive? I say no harm will come from making it a hard requirement. Anything noncompliant is suspicious, as having toooooooooooo much resemblance to choose-one plurality.

In response to your mention of for-or-against, let me point out that I usually argue that balance is necessary, not that it is sufficient.

As compared to Score{0-10 by 1}, Approval may require more effort to apply the proper strategy/tactic. I think it requires using probability. This is too hard to explain to typical voters. It may be too hard to recruit activists for the system because of false intuition about the system that would be bypassed with Score{0-10 by 1}. People perceive a dilemma about whether to approve compromise candidates, as though the only answers were to approve them always or never, when the correct answer is to approve them with a probability depending on your Score vote for them. If the voter is going to figure out her Score vote, why not use Score{0-10 by 1} voting?

An attempted counterargument may say that Approval is easy to count by hand, but Score {0-10 by 1} would take too long. The counter to that is to say the initial election can be done with computers, the winner placed in office, and a hand check done afterward.

@casimir, thanks, that's a very interesting write-up.

At the end, you suggest that the equal-vote criterion would be too restrictive as a minimal standard. But all four of those systems pass it, I believe. They are additive and Frohnmayer-balancing.

In summary, does anyone have a serious objection to the idea of all of us promoting this system for all cases where there isn't already a big investment in Score (including Approval) or STAR?

We who want to eliminate vote-splitting and spoiler effects have grounds to choose a system that the public can easily understand. Does anyone think this system passes muster in that regard?