Actually, Malibu Stacey was riffing on Barbie. Sad but true.

I've got a comment about preference voting for "seconds" (and quorums, for that matter). On a normal ballot -- say, with 10 different candidates for an office -- there's not an obvious order on the options, but that's not true here. Say I think 6 is the optimum quorum size. Then my opinions on 8 and 10 are unclear, but it's almost definite that I'll rank them 6, 4, 2, 0 (with 8 and 10 somewhere -- probably either high or low.)

I really can't imagine someone ranking the options, say, 6, 2, 0, 4. It doesn't make sense.

You'd get nearly the same results as preference voting if you simple had people vote for the largest number of seconders that they would support (so, by voting for "6", you're saying that you're ok with 6 or below.). Then start counting at 2. Would over half the people be happy with 2? If not, 0 wins. If so, check 4. Would over half the people be happy with 4? If not, 2 wins. And so on.

This is easy to vote on, but hard to explain. It also doesn't let someone say "I think 6 is best, but 8's almost as good," but you'll get results that are almost the same as preference voting, with much less confusion on the part of voters.

I expect no one followed my argument at all, but as the mathematician lurking in the corner, I felt I had to pipe up...

(following up myself) In Wolfram's case:

I just want to be sure that out of 100 voters, if 25 want 2 Seconds, and 20 want 3 Seconds and 20 want 4 Seconds and 35 want 10 Seconds, the 35 don't win it. That's very disparate and unfair

the win would go to 4 seconds, as only 45 voters are happy with 2 (or fewer), and 65 are happy with 4 or fewer.

MayaP's math doesn't hurt.

(so, by voting for "6", you're saying that you're ok with 6 or below.).

But what if I want to go the other way? What if my ideal choce is 6, but if that gets eliminated, I want something higher not lower?

Malibu Stacey was riffing on Barbie.

I stand corrected. Sorry Gandalfe.

You'd get nearly the same results as preference voting if you simple had people vote for the largest number of seconders that they would support (so, by voting for "6", you're saying that you're ok with 6 or below.).

Very interesting.

I have a point about the validity of votes that needs addressing.

If you give me four things to vote for, and tell me to vote for them as my first, second, third and fourth choices, what happens to votes which are "incorrect", for instance:

I vote them all "1".

I slip and vote them as my first, second, second and third choices.

I vote "1" for the choice I want and leave the rest blank.

In those cases, what gets counted? The usable portion of the vote? Or does it get thrown out?

If Jon B or anyone else is happy to code PHP that keeps saying "nuh-uh, try again, bozo" then that problem will be solved, but it'll be a pain.

Or, is it my constitutional right to assign a single option a "1" because I believe so strongly that that's the only option I could bear to live with?

Or for a better example (from my post a ways back), Let's say for votor turnout my first choice is 30. My second and third choices would probably be 40 and 20 (or vice versa), not 20 and 10. I'd want something near my first choice, not necessarily less than my first choice. Your logic may vary. Which is why I like preferential voting.

Math is scary, that's what math is.

All I gotta say is that it's called a simple majority because it's simple!

If Jon B or anyone else is happy to code PHP that keeps saying "nuh-uh, try again, bozo" then that problem will be solved, but it'll be a pain.

I've been thinking about that very issue. I don't think it will be a pain to code, and I'm happy to do it.

So you'd rank them 6 8 4 10 2 0? (I assume you'd still rather have 2 than 0, after all.) Then you can vote for 10, knowing that if enough people like 6-or-fewer, it'll win; if not, it'll go up to 8, and if not, it'll go all the way to 10. (And if 4 winds, then over half of the people don't like any option higher than 4, so it doesn't matter if you prefered 6, 8, or 10).

I really think this works for both seconding and voter turnout, unless you think anyone would prefer, in order, 6-2-4, or (for turnout), 50-20-30.