Just a Fun Idea

Here’s just an interesting way for the forum to conduct elections of whatever sort: We could have periodic elections to decide on the voting method we will be using as a community over the following period, starting with something I think we will all be equally unhappy with: plurality!

The idea is that even starting off as poorly as possible, as the process continues, the voting systems we tend to elect might over time reflect what systems people like or feel like work for them. It’s something I’d like to see happen, anyway. I think it would also be interesting to see what systems people tend to prefer over others, especially in a community of people who are more knowledgeable about these systems than, say, your average bear.

Any thoughts? I’m not sure if that’s the sort of thing @RobBrown had in mind when he mentioned a “voting system of the month,” at first I thought maybe that was more of just a play on “employee of the month,” but maybe I’m just dense.

1 Like

It’s clever, although I admit that I’d rather we submit “rich ballots” from the get go, and evaluate them using lots of methods.

Ok, fine. Plurality it is.

Plurality
IRV/Ranked Choice
Approval
Score/Range
Majority Judgement
SODA
STAR
STLR
STELLA
DV
:heavy_check_mark: Cardinal Baldwin
Condorcet (ranked pairs)
Condorcet (ranked pairs) blended with Cardinal Baldwin

I reserve the right to come back and edit my post when I see how others have voted. :slight_smile:

“Rich ballots” could also be interesting. I also would propose to include 3-2-1 Voting, STV, Bucklin Voting, Borda Count, and Copeland on the ballot. Also maybe it would be more interesting to omit the method we are currently using so that we get more variety. Although it would also be interesting to see which methods tend to elect themselves…

Haha also we gotta have a way to stop that from happening if the votes are on the forum. So probably they should be collected externally in a document with public time stamps on editing.

Feel free to post your own choice ,and add all those options to the ballots. I just did the ones off the top of my head, or that I had heard mentioned here recently.

I think being allowed to edit your vote actually is rather interesting. I’m convinced that my method of choice would be more or less immune to it… at least if we allow 0-100 ballots as opposed to 1-5.

Alright here’s my ballot, despite the fact that I’m not entirely well-informed:

Plurality
IRV/Ranked Choice
Approval
Score/Range
Majority Judgement
SODA
STAR
STLR
STELLA
DV
Cardinal Baldwin
:heavy_check_mark: Condorcet (ranked pairs)
Condorcet (ranked pairs) blended with Cardinal Baldwin
3-2-1 Voting
STV
Bucklin
Borda
Copeland

Alright the polls are closed. I flipped a coin and I won. jk.

1 Like

Plurality
IRV/Ranked Choice
Approval
Score/Range
:heavy_check_mark: Net Approval (score but on a -1 to +1 scale so it’s effectively approval but with the option to give candidates you don’t know a neutral score)
Majority Judgement
SODA
STAR
STLR
STELLA
DV
Cardinal Baldwin
Condorcet (ranked pairs)
Condorcet (ranked pairs) blended with Cardinal Baldwin

Hopefully there are more people who are a fan of pure rated methods (the only methods that truly pass IIA, fav betrayal, participation, consistency, etc.) so I’m not just splitting the vote in this very established Condorcet/Cardinal Baldwin 2 party system \s

What about multi-winner systems? That’s an interesting topic that doesn’t get much attention. If the new forum is going to be self governing then it might need both single winner and multi-winner positions and that would be a great opportunity to experiment with different multi-winner methods.

So after this round we use the winner to elect the next method? And then keep going recursively until we hit a steady state? Cool idea. Can we limit it to methods which have electowiki pages. I do not know all of these. If somebody feels strongly for a system then they can make the page.

Plurality
IRV/Ranked Choice
Approval
Score/Range
Net Approval (score but on a -1 to +1 scale so it’s effectively approval but with the option to give candidates you don’t know a neutral score)
Majority Judgement
SODA
STAR
:heavy_check_mark: STLR
STELLA
DV
Cardinal Baldwin
Condorcet (ranked pairs)
Condorcet (ranked pairs) blended with Cardinal Baldwin
3-2-1 Voting
STV
Bucklin
Borda
Copeland

Do we have any rules for when one round ends and a new one begins? Perhaps we should make each round end when there isn’t currently a tie and at-least 8 participated in that round? Or should we make it a time thing?

Also here’s my super sleek tally:

Plurality
IRV/Ranked Choice
Approval
Score/Range
image Net Approval
Majority Judgement
SODA
STAR
image STLR
STELLA
DV
image Cardinal Baldwin
image Condorcet (ranked pairs)
Condorcet (ranked pairs) blended with Cardinal Baldwin
3-2-1 Voting
STV
Bucklin
Borda
Copeland

Adding my vote to the tally:

Plurality
IRV/Ranked Choice
Approval
Score/Range
image Net Approval
Majority Judgement
SODA
image STAR
image STLR
STELLA
DV
image Cardinal Baldwin
image Condorcet (ranked pairs)
Condorcet (ranked pairs) blended with Cardinal Baldwin
3-2-1 Voting
STV
Bucklin
Borda
Copeland

There is definitely no chance that vote-splitting will affect the result of this vote in any way.

No set in stone rules yet, I think we’ll develop those kinds of rules as we get a feel for whether people will participate and how often people check the forum. I think a combination of participation and time makes most sense, as long as people are made aware that a vote is happening. After a certain threshold is met, we can announce when the polls will close, or something. We just have to agree on some rules about that too, like should the poll be open for a whole day after the threshold is met? Or what? How do we decide the threshold and the timing? I’m not sure, whatever we can agree on.

Lol the problem is sort of a chicken and egg thing. How do we vote on how to do the vote? Yeah… I think if we want something like that, it’ll just have to get started somehow, and if people are empassioned about some aspect or another, we’re all fairly reasonable people for the most part, they can just talk to us and I’m sure we can agree on something.

So what are we actually voting for? Are we saying that plurality is the current voting system of the month and we’re voting with plurality to find the next one? And do we need to list all the poll options for every vote? I mean, if I vote, I could just name one method rather than list them all and put a tick by one of them! And can we add options?

By the way, I’ve always felt that the best method for voting online is score voting but where the current result is always visible and you’re allowed to change your vote (though you’d need some software for it - it wouldn’t work with people just posting their votes).

1 Like

Basically yes. The reason I want it to start with plurality is because everyone hates it lol. It’s just a suggestion but I think it’s the best way for us to agree on a starting point, since a lot of folks have their favorite method and if we can’t start somewhere it may just turn into a debate about which voting system is best. Plurality is bound to elect some other method that’s significantly better than itself anyway, so we’ll just have to have faith in the recursion itself over time to converge on a collection of systems that we all generally like.

It’s also interesting to consider candidate selection. I was thinking we could initiate a multi-winner system that selects the candidates. That has the same issue though. We’ll all just have to hash it out and come to a general agreement. Since there would be a single multi-winner system that selects candidates, we should start with single-winner and then develop the system from there.

To get started, we will have to choose a single winner system and a multi-winner system, somehow. Then we will have the precedence in place to continue in a reasonable way. Eventually we would use the single-winner system to select a multi-winner system, and the multi-winner system to establish candidates for both systems. I’m not sure how to do the nominations.

If you prefer, we could also leave some aspects of these initial decisions up to a uniform random process.

Also I think it makes sense to copy the poll and add your own tally. I’m not sure, this is all up in the air. We could establish a shared Google doc and save copies of the ballots of our own. But I feel like there might be a problem with the votes being out in the open for everyone to see, maybe I’m wrong. Otherwise we’ll need to set up a system that keeps the votes a secret until the polls are closed. Some kind of encryption maybe.

I started adding choices so it would make it easier for people to cut and paste and pick one.

We’re starting with plurality because it is kind of funny. It doesn’t mean anything. Nothing is official here. Eventually I’d like to do them better with a voting widget and we look at results under different systems etc, but at this point especially, it’s just for fun.

1 Like

Lol if there are only 18 participants it’s possible we end up with everybody voting for their own system. We’ll see what happens. Also while it doesn’t matter I’m changing my vote:

Plurality
IRV/Ranked Choice
Approval
Score/Range
image Net Approval
Majority Judgement
SODA
image STAR
image STLR
STELLA
DV
image Cardinal Baldwin
Condorcet (ranked pairs)
Condorcet (ranked pairs) blended with Cardinal Baldwin
image 3-2-1 Voting
STV
Bucklin
Borda
Copeland

I think that first it would be better if everyone proposed some methods of voting, and then you only vote on those.

As an initial voting method I propose a wide range such as [0,100] without equal scores.
The vote can then be converted into a single choice, multiple choice, rank, and range in order to apply all (or almost) the initial proposed methods, hoping that there is an obvious winner who will be the first method used.

With plurality I just have to wait for others to vote for their preferred method, and in the end I (instead of voting for my favorite) vote the best from the others… seeing other people’s votes in advance makes no sense.

I found this site that seems to me valid for our purpose:
https://esurv.org/
it’s completely free.
To distinguish the votes you can add a text field in which everyone will indicate his name (or email).
Obviously, whoever manages the poll (the only one who can see the results before closing) will not be able to vote, thus avoiding tactics based on the votes of others.

In any case, if things aren’t done right, I will abstain from this “game”.

1 Like

You don’t have to play, but I’d prefer we do it here and use our own tools to tabulate (although, for plurality, we can just count, right?). Doing it here makes sure it is the people who they say they are.

Regardless, it was @cfranks post, why not just do it his way? Nobody is stopping you from starting your own.

I was joking about changing my vote after seeing how others vote, btw.

Ok, but you can only use the single choice here.
I proposed that external site only to obtain the votes on which our own tools are then used to find the winner.

You’re right, I will do so. Here use plurality.

Gotcha. If the new forum actually happens (who knows at this point, I’m at the point of bailing following after seeing a bit of EqualVote’s colors), we’ll build nice tools of our own, with our Codepen work being a nice starting point. (I know you’ve done some nice JS work elsewhere too, as have I) We could host the server side of them at the same domain as the forum.

1 Like

OK I will put my vote back, lol whoops. That does mean that late-comers have a strategic advantage, but that’s fine for now. I think @Essenzia has a good idea using the external site, especially if it keeps the ballots secret until the polls close.