@jack-waugh @ChocoPi in BTR, it is already impossible for a Condorcet winner to be eliminated, yes?

That's what I thought, so sticking a Condorcet check on should make no difference.

I’m also considering what @SaraWolk suggests, namely that another method like RCV (IRV) might be a more practical conduit for change, even though it is significantly less ideal.

I'm not recommending RCV (IRV). I think it's oversold, broken, and extremely damaging to the voting reform movement.

I recommend STAR, Approval, and Condorcet, and support a number of others.

I suggest that Approval makes it more work than necessary to vote responsibly as one has to resort to a stochastic technique.

Mixed strategies can be optimal in any electoral system.

@jack-waugh said in Approval Voting as a Workable Compromise:

I suggest that Score makes it more work than necessary to vote responsibly as one has to exaggerate support for ones compromise candidate.

…except, ironically, score makes randomization less important by letting you give partial scores instead of requiring you to vote at random.

@jack-waugh said in Approval Voting as a Workable Compromise:

Then there goes Score-B2R.

That's fine. STAR, Smith//Score, ranked pairs, etc. are all probably good enough. IRV isn't.

There's nothing really wrong with approval,

I suggest that Approval makes it more work than necessary to vote responsibly as one has to resort to a stochastic technique.

score,

I suggest that Score makes it more work than necessary to vote responsibly as one has to exaggerate support for ones compromise candidate.

or STAR

Same as Score except that the candidates will come in pairs of near-clones.

except that people heard about IRV first. As I've mentioned before, the easiest way to deal with that is to get a case to the Supreme Court that strikes down any nonmonotonic system as unconstitutional (a strategy that's already been used successfully in Germany).

Then there goes Score-B2R.

But you can do even simpler and better than BTR. You just slap a Condorcet check on it, the end.

Please lay this out as a separate topic.

First, I'll point out that Fargo is raw Approval and St. Louis is Approval-into-Runoff. These behave very differently. (Hint: the latter is way better, and is basically STAR's twin brother.)

Meanwhile, I'm extremely skeptical. The average number of approvals per ballot in St. Louis is way lower than in Fargo, and the approval curve is much less sharply-peaked. That suggests to me that some voters are turkey-raising. Though luckily, not enough are doing it for it something to have gone wrong… yet.

Maybe our best marketing strategy is to try slide B2R IRV rating-ballots into the place of Hare IRV RCV in popular mindshare as a solution with the same flash but without the disappointing and counterintuitive results that Hare often delivers.

People have been trying this for as long as IRV has been proposed. The answer to "why does IRV keep gianing ground" is "because billionaires keep pouring money into FairVote and FairVote keeps lying about IRV".

There's nothing really wrong with approval, score, or STAR except that people heard about IRV first. As I've mentioned before, the easiest way to deal with that is to get a case to the Supreme Court that strikes down any nonmonotonic system as unconstitutional (a strategy that's already been used successfully in Germany).

First, I'll point out that Fargo is raw Approval and St. Louis is Approval-into-Runoff. These behave very differently. (Hint: the latter is way better, and is basically STAR's twin brother.)

Second, I am in complete agreement that the best path forward is simple improvements to IRV.

But you can do even simplier and better than BTR. You just slap a Condorcet check on it, the end.

Your ballots are now precinct summable and monotonic outside of cycles. The results are incredibly straightforward, ("Here's how much the winner beats everyone else by:") yet it also produces a complete pairwise matrix for any campaigns or political scientists who want it. It's 100% Condorcet efficient; no more center-squeeze, maximimally resistant to polarization. It's fully cloneproof.

And best of all, it exhibits the highest strategy resistance of any method. (Only Baldwin's comes close.)

BTR is good, but it should be regarded as strictly inferior to the more straightforward Condorcet//Hare alternatives unless you are somehow at a loss for computational speed.

This is sort of a seperate question, even if it is a logical next one.

Think of it this way: We are talking about electing legislators with a new voting system. We are not talking about the legislators using a new voting system to vote on bills. That might also be a good idea, but it's an entirely seperate matter.

The electoral college is similar. Each electorate elects a (single) desired target for their state's delegation to vote for. We're not talking about how the delegations themselves vote. That might also be a good idea, but it's an entirely seperate matter.

Of course, there is more reason to address the latter than the former; legislature are deliberative bodies, while the Electoral College as currently set up is basically not. But the point is they are legally and functionally an independent system. In both examples, there is little reason for the representative body to conduct itself in the exact same way as the preceeding public vote.

$0.

I think we all know this figure is missing a few zeroes. /s

More importantly, this still leaves the issue of the Electoral College. How would we translate the EC district votes into EC votes under this system? Approval voting would be an immediate improvement for electing Congress Critters and state/local officials, but at some point we'll have to think through how to reform the Electoral College. The way I see it, abolishing the EC is too radical a proposition to have associated with any advocacy for an Approval Voting reform, so a plausible idea for how the EC could work is worth having. Simply leaving the EC the way it is while selecting Electors using Approval Voting would probably lead to more problems.

You just flip a switch. Uncheck the "discard overvotes" box, rewrite one sentence on the instructions line (in each language), and make sure the LEO guidance on tabulation reflects all this. The end. $0.

Extending the proposal to anything that actually requires a new ballot/process/everything defeats the entire point of the exercise. For that implementation cost, you could do practically any constitutionally valid single-winner system.

First, based on what I’m reading, it was mostly a hereditary, aristocratic oligarchy without broad-based democratic participation. It did not have a written constitution, or a formal bill of rights. Second, it was not a confederation of sovereign states but a centralized unitary government. Third, it had a much smaller and more homogenous population (in terms of culture) than the United States, which was spread out over far less land and simultaneously had a lower population density.

These things make it much easier for the small elite political class to settle affairs in whatever way they might agree upon amongst themselves, including using a chosen voting system. The political elite class was also much more highly educated than the rest of the population. When it comes to voting systems for us, there are strict laws and regulations that need to be satisfied, in addition to the need for support from a much more numerous and heterogeneous population, many of whom are relatively uneducated or misinformed, compounded with deliberate efforts against technical voting reform by our own political elite (and associated media). This is why I’m advocating for a simple reform, like approval voting. I’m also considering what @SaraWolk suggests, namely that another method like RCV (IRV) might be a more practical conduit for change, even though it is significantly less ideal.

@k98kurz What you mean "combined"?

By "combined approval voting", I mean score voting with the options Disapprove (-1), Neutral (0), and Approve (+1); a candidate's score is the sum of "Approve" votes minus the sum of "Disapprove" votes. Iirc, I read that the Republic of Venice used CAV in its elections for around 500 years, though the name of the method is of much more recent origin.

call it IRV but mention that some people call it RCV in brackets

I further qualify it "Hare".

It's bad enough that there isn't a single clear term for Choose One Voting/ Plurality/FPTP.

My advice is to call the ranked ballot family of voting methods "Ranked Voting", call IRV "Ranked Choice Voting" and call FPTP "Choose One Voting".

In academic contexts, we should also use these common words but then have the technical names in parenthesis. The target audience for this stuff is groups like LWV who form non-expert volunteer committees to get up to speed and study reforms before making recommendations that carry massive weight.

This is another reason to not rebrand Approval, which already has a clear and self explanatory name.

@cfrank I say Ranked Voting to be more inclusive of ranked methods. (Ranked Choice/RCV is IRV and also sometimes STV).

Sidenote: Since advocates call IRV RCV, but academics and electoral theorists call it IRV, they have a firewall between the method and its scientific criticisms.

That's a problem and I strongly encourage us to all start calling it what they call it so the algorithms and search engines connect the two.

I think a better compromise might be to call it IRV but mention that some people call it RCV in brackets, so it's still linked for search engine purposes.

Sidenote: Since advocates call IRV RCV, but academics and electoral theorists call it IRV, they have a firewall between the method and its scientific criticisms.

That's a problem and I strongly encourage us to all start calling it what they call it so the algorithms and search engines connect the two.