[zestyping]

Approval, not ranked-choice.

Ranked-choice reduces transparency and understanding of the vote-counting process, disenfranchises an alarming percentage of lower-income voters, obstructs risk-limiting audits (which are essential for security), and is non-monotonic (increasing voter support for a candidate can make them lose). Further, ranked-choice doesn't actually fix the spoiler problem and won't eliminate two-party dominance.

Approval voting is cheap and easy to implement, dead simple to explain, count, and audit. Not only does it eliminate the spoiler problem, it is easy to see why it does so: your ability to vote for any candidate is independent of your ability to vote for any other.

[saulpw]

I've heard the arguments for approval voting, and I'm sure it's all the things you mention and more, but people don't get it. I don't get it. I don't want to vote for both Hillary and Bernie. I want to vote for Bernie, and then only if Bernie can't win, would I let my vote go to Hillary. You can explain to me until you're blue in the face why approval is strictly better even in this situation, but I am emotionally attached to my vote counting for Bernie more than any other candidate, so reason isn't going to work on my lizard brain.

I know, it sucks. Politics is terrible. But we have some momentum behind RC/IRV so we should use it and stop the single-vote FPTP system that's plagued us for centuries. Anything is better than that. So let's join forces and get behind whatever has momentum even if it's not technically the best.

[Gormo]

Approval voting seems to me to be worse on all counts that the previous commenter was levying against ranked-choice. To your point, the spoiler effect seems like it would be much worse with approval than with a ranked ballot, since highly partisan voters would have little reason to approve of any candidate other than the single candidate they want in office. Approving of anyone else lessens their candidate's chance of winning.

A ranked choice ballot at least requires you to assign a unique value to every candidate on the ballot: you can honestly rank your second choice without being concerned that doing so undermines your first.

[nobody9999]

>A ranked choice ballot at least requires you to assign a unique value to every candidate on the ballot: you can honestly rank your second choice without being concerned that doing so undermines your first.

That's highly implementation dependent. Where I live we have ranked-choice ballots for local primary elections, while the local general elections are FPTP. State and Federal elections are all FPTP for primary and general elections.

While I am free to rank up to five candidates when filling out my ballot, I am not required to use all five choices.

I can just ignore all that if I choose and just rank one candidate first and leave the rest of the ballot blank. Or I can rank multiple candidates, but I'm not required to "assign a unique value to every candidate on the ballot."

In fact, if there are more than five candidates for a particular office, I can only rank five of them.

All that said, I'm absolutely in favor of RCV and wish we had it for all elections, not just local primary elections.

[Gormo]

It sounds like the local ranked-choice implementation is unnecessarily complex and constrained. A simple "rank all candidates from 1 (most preferred) to n (least preferred)" for n candidates seems like the better solution.

[nobody9999]

>It sounds like the local ranked-choice implementation is unnecessarily complex and constrained. A simple "rank all candidates from 1 (most preferred) to n (least preferred)" for n candidates seems like the better solution.

I'm sure you're right. Unfortunately, I'm not the person you'd need to convince.

Here's contact information[0] for the relevant folks, and thanks for taking an interest. I'm sure my fellow townspeople will be grateful for your guidance. You have my thanks for stepping up to help us improve our voting systems!

For your reference, here's some background on the how the process came to be[1][2][3][4]

[0] https://www.vote.nyc/page/contact-us

[1] https://ballotpedia.org/New_York_City_Ballot_Question_1,_Ele...

[2] https://apnews.com/article/nyc-ranked-choice-voting-explaine...

[3] https://rankthevotenyc.org/history-of-rcv-in-nyc/

[4] https://rankthevotenyc.org/what-we-learned-from-new-york-cit...

[ClayShentrup]

it's the worst of the commonly discussed alternatives.

https://www.rangevoting.org/BayRegsFig

[nobody9999]

Thank you for your expert opinion. Unlike yourself and your colleague[0], I am not an expert on voting systems and infrastructure.

I am just a consumer of such things and have exactly zero say in my town's approach to voting.

I do know that RCV is better than FPTP, even more so if we don't, at least, require a majority, and am glad my town is at least making a start at such things.

That said, I'd love to make it even better.

As I suggested[1] to your colleague, it would be terrific if your expertise could be used to improve the voting system where I live.

I'd expect that the folks[2] who make such decisions could be convinced to re-frame things in another referendum based upon the recommendations of you and your organization. I know I'd certainly appreciate it!

[0] https://news.ycombinator.com/item?id=47035812

[1] https://news.ycombinator.com/item?id=47036908

[2] https://www.vote.nyc/page/contact-us

[ClayShentrup]

LOL, people get it just fine. fargo adopted it by a 64% supermajority and st louis adopted it by a 68% supermajority.

https://approval.vote/

> You can explain to me until you're blue in the face why approval is strictly better even in this situation, but I am emotionally attached to my vote counting for Bernie more than any other candidate, so reason isn't going to work on my lizard brain.

but your actual strategy is to rank hillary in 1st because bernie can't win. or, in the case of my aunt, she preferred warren but voted biden to beat trump. she would have ranked them biden>warren>trump in a ranked election for that very reason. this is called "compromise strategy".

bro, approving both of them is better than being strategically forced to say that you prefer clinton to bernie or biden to warren.

[saulpw]

Welp, you just proved my point. I still don't get it. I want to vote my preference and I don't want to vote Hillary and Bernie equally. shrug.

[vant1]

> Approval voting is cheap and easy to implement, dead simple to explain, count, and audit.

Not so dead simple to vote, though. If you're a sincere voter and you prefer Alice to Bob and Bob to Charlie, do you approve of Alice, or both Alice and Bob?

That choice has to be either strategic or very noisy.

There seems to be some unavoidable complexity to voting methods: letting the voter deal with the complexity leads to a method with a very simple algorithm but that's tricky to use. Letting the method itself deal with it leads to more complex algorithms, but makes it easier to vote.

That said, the alternative vote is a bad ranked voting method; with that I do agree. Just beware of the complexity hidden in the system, whether that's Approval or Ranked Pairs.

[eszed]

I agree with this. Ranked choice is easy to explain to a naive voter: everyone understands how a preference order works, and the result is "the candidate more people like the most". Counting the votes is (a bit) complicated, but I think the (minority of) people who get excited by implementation details out-smart themselves, by worrying that most people won't understand the details. Of course most people won't understand the details, because they don't care about the details. They don't know how votes are tallied now!

My position admittedly breaks down when people lie to low-information voters about the fairness of the process - but, in my defence, people will lie about any system that doesn't produce the results they want. I'd prefer they lodge their objections to a better system than first-past-the-post.

[ClayShentrup]

massively false based on actual data. https://clayshentrup.medium.com/star-voting-is-simpler-than-...

[eszed]

That's fair. RCV does break down with a large number of candidates. Though doesn't star voting have some odd corner cases? Regardless, every alternative scheme I've seen seriously proposed would be a massive improvement over FPTP.

[ClayShentrup]

this is dead simple. thousands of voters have had no problem. https://approval.vote/

if you really want to get into the game theory, here it is. https://www.rangevoting.org/RVstrat6

[ClayShentrup]

it's ironic that approval voting is better, but people therefore often wrongly think it's less optimal, when it's actually _more_ optimal.

https://web.archive.org/web/20190219005733/https://sites.goo...

you used to live across the street from me on harper street in berkeley by the way.