-
Single-winner voting system.
Ballots:
The ballot rates every candidate "Endorse", "Compromise", or "Oppose". The default is Oppose.
Tally:
The first stage of the tally is to find a collection of semifinalists. The count of semifinalists can be calculated and published as soon as the field of candidates is fixed.
Let Nc be the count of candidates. Then the count of semifinalists is min(Nc, max(3, ceil(.708 + ln(Nc)))).
Under this formula, with 3 through 9 candidates, there will be 3 semifinalists (as in 3-2-1). 10 through 26 candidates would lead to 4 semifinalists. It would take 27 candidates to get up to 5 count of semifinalists.
The semifinalists are the top finishers from a Score tally based on scoring Endorse at 100, Compromise at 99, and Oppose at zero.
In terms of power relations, I had rather use a PR method than the tops. But that's too hard to get people to understand, so I'm going for a simplification by saying it's just the top scorers.
Find two finalists: the semifinalists with the fewest Oppose ratings.
Find one winner: the finalist who is rated above the other on more ballots.
Motor-voter motivation:
The first qualification is some more or less fair indication of support, for which Score seems to have the power to achieve that, close enough for government work.
The second and last qualifications are motivated similarly to 3-2-1. But the present proposal addresses what appears to me as a defect of 3-2-1, namely that the first stage of it looks a lot like FPtP.
Strategy:
Compromise to the lesser evil between the heavily fronted candidates??
Naming:
The "S" stands for "Score", and the "2-1" stand for the number of finalists and winners, and for similarity to "3-2-1."
Evolution
In the event that the money parties are substantially hurt and people's candidates draw significant support, the score associated to Compromise could be ratcheted down to 90, then 50. But to whom to give this decision of when, is a question.
-
@jack-waugh Ok the names are a bit goofy but I assume that was the intention.
Why is "hostile stare" given a 99? I don't get that.
Voting behavior in the US is mostly based on hatred and fear, as it should be.
I don't know if you are serious here. Are we supposed to be considering this with any seriousness? This seems more trollish than anything.
-
This post is deleted! -
@jack-waugh said in S-2-1:
So that's why I think that out of 100, the compromise should get 99 and the most-hated should get zero. To put strong differentiation between them under the control of the voter.
Are you going to tell the voters that "love" and "dislike" are assigned nearly identical scores? While expecting them to vote as if the names accurately reflect how they are to be interpreted?
Your suggestion makes me want to just go with ranking, rather than scoring, because it demonstrates just how meaningless the scores are. Giving them numbers that don't in any way correspond to the names really highlights the exact reason score voting rubs me the wrong way.
I like score ballots just fine, I just don't like using all the information in them so directly, especially not in ways that are so "subject to interpretation" on the part of each voter. Among other things, that makes the system highly unstable.
I think there are plenty of good ways that arrive at those middle ground candidates that bring out neither hatred nor worshipful devotion and blind loyalty. Ways that are game theoretically stable, and don't rely fooling people as to how their votes are to be interpreted.
BTW, I think we'd be much better off taking discussions of fascism, genocide, etc to forums other than this one. Sure, it's very relevant that our country is highly polarized, and we'd like to reduce that through better voting methods. But getting more specific than that will only drive away the kind of people we'd like to have here, and attract partisans and, frankly, conspiracy-minded crackpots. Can I request we tone that sort of stuff down?
-
Now that thanks to the work of @Sass, I am finally able to understand what 3-2-1 is (which I never could get near to understanding it based on the original description), I want to offer an improvement over it in regard to the first stage. This first stage winnows down the field of N candidates to just 3. But in 3-2-1, this stage depends only on first choices. This, I feel, smacks too much of IRV and FPtP, so carries risk of emulating the behavior of those systems. Why not take advantage of the ingenious final two stages of 3-2-1 without messing up in the design of the first stage?
-
@rob It's my opinion that strategy in Score merits a difference between the score used in the tally for the compromise candidate and the voter's true feelings.
-
The whole idea of giving an "honest" rating in Score represents a failure to think realistically. This idea should be fought at every turn.
-
This post is deleted! -
This post is deleted! -
@jack-waugh said in S-2-1:
he whole idea of giving an "honest" rating in Score represents a failure to think realistically.
Ok..... I'd say this is exactly why purists seem to prefer ranking. When ranking candidates, there is zero ambiguity as to what is meant by an honest ballot.
Well I should probably say "near zero", since someone will argue that. But I think most people will agree that "I like A more than B" is not ambiguous, while "I like A twice as much as B" is.
If you look at what 99% of partisans say on Facebook and the comments of Youtube, you have to notice that it is motivated by hatred and fear of the "other side".... I suggested the marketing names I did to let these people know that we understand how they feel and we have the solution to their needs.
And yet you say "Voting behavior in the US is mostly based on hatred and fear, as it should be."
I don't get it. You WANT to embrace that mindset?
Facebook YouTube etc is just as broken as our voting system. People who post the sort of comments you describe aren't representative of typical people, since the vast majority of typical people steer clear from participating in the cesspool that is Facebook and YouTube comments, at least in places where things get political.
Regardless, if you consider "Tongue kiss" to be a good way of marketing a voting system.... uhhh. Ok.
-
"I like A twice as much as B"
This meaning in no way attaches itself to a score that figures in only additively and comparatively.
Scores are really an interval scale, not a ratio scale.
-
@rob Endorse/Compromise/Oppose
-
@jack-waugh said in S-2-1:
Scores are really an interval scale, not a ratio scale.
Intervals are problematic as well.
Endorse/Compromise/Oppose
I don't see how "compromise" makes sense in that context. If you are a someone who likes Nader best, Gore second best, and Bush the least, giving Gore a score of "compromise" doesn't make sense to me.
These instructions (and "naming") work for me, for cardinal ballots on the left and ranked ballots on the right. I particularly like the way, for the cardinal ballots, they tell the voter to give their favorite 5 and least favorite a 0, so there is no implication that the scores have absolute meaning. They also say "best" and "worst" rather than something like "like" and "dislike".... again, indicating that the scores are relative to the field of candidates rather than absolute expressions of how the voter feels about each candidate.
Nonetheless, for a ton of reasons, I think cardinal ballots should be treated as is they are ordinal, except in the (hopefully) rare exception of having no Condorcet winner.
-
@rob So you would prefer either your own Condorcet proposal or Sass's Condorcet proposal over 3-2-1?
-
@jack-waugh I prefer Condorcet methods over non-Condorcet methods. And for Condorcet methods, the simpler the better.
3-2-1 isn't particularly interesting to me, although if it started getting traction I'd give it more of a look.
-
@jack-waugh It's really weird reading a Score advocate claim that voting behavior should be based on hatred and fear. Score is all about consensus.
Also, "Tongue Kiss" is super f****** gross. I'm genuinely repulsed and knowing that it's from the person who manages this site makes me want leave the entire forum.
Anyway, 3-2-1 was really designed with with the delegation case in mind, not the undelegated case. Quinn is expecting many voters to rate a single candidate "Good" and then let most of their ballot be filled out by that favorite. You seem to argue in favor of favorites anyway, so I'm not sure what you issue with it is.
Voters tend to use the scores 100, 99, 50, 1, and 0. That corresponds to favorite, backup, meh, lesser evil, greater evil. Your scale is super lopsided. Really, you should just use those 5 terms and then find the two semi-finalists with the fewest "evil" ratings, regardless of whether they're lesser or greater.
-
@sass, I went back and changed the terms for the ratings.
I thought I understood 3-2-1 from the new description in Electowiki that I suppose you either wrote or influenced. But as I read it, it looks as though delegation is a design option.
The main description says << Find 3 semifinalists: the candidates with the most “good” ratings.² >>.
-
It's really weird reading a Score advocate claim that voting behavior should be based on hatred and fear. Score is all about consensus.
Agree completely. I mean I'm not a huge fan of Score but good voting systems should be about bringing people together toward something they can all be relatively happy with.
Instead, this sounds like someone is just raging against the machine rather than approaching it with a positive vision of a better future. Makes me feel like I'm in the wrong place.
Also, "Tongue Kiss" is super f****** gross. I'm genuinely repulsed and knowing that it's from the person who manages this site makes me want leave the entire forum.
Sorry Jack but this was my impression too. I'd recommend giving thought to how that sort of cranky-ranty-weird-angry-gross stuff comes off to people visiting here and considering participating. Rkjoyce did a lot of that on the old forum (example https://votingtheory.com/archive/posts?where={"topic_id"%3A342} ), but he wasn't the one actually running the forum.
