Vote #1: Poll Choice Limits

[Jase]

Actually, I think Mike was questioning the case where you might have a vote spread of say, exaggerating it:

10: 50
20: 49
30: 49
40: 49
50: 49

etc. Where FPTP would say 10 wins, but obviously, the desire for a higher number is much better supported. It's just its supporters aren't in agreement on how much higher it should be.

[Dark Pulse]

Well, my issue with this is if 50 people pick 10, 40 pick 50 and another 40 pick 45. Clearly the people who picked 50 would prefer 45 over 10, and yet 10 would have the most votes due to the vote being split between 45 and 50, which to me seems pretty absurd.

It doesn't seem like it's going to go that way, fortunately, but it easily could have.

Doesn't matter. When one has multiple choices, one has to win. It 50% of people vote for a centrist government, 35% vote for right-wing, and 15% vote for left-wing, does that mean the government should be center-right because there were more right-wingers than left-wingers?

I'm pretty sure that the BL limit was higher than that, actually, because we had polls with a lot more than 10 characters in them.

It's possible I'm misremembering. Regardless, the uses of needing it would still be slim. At the same time, acceptable limits must be achieved, because if a spammer got on here and found unlimited options... he'd have a field day.

Well, the point is, if people don't use it, so what? What have you lost? If, on the other hand, people actually do need it and you won't raise it, it causes real problems.

See above. Do you want to be the idiot who does Damage Control when something like that happens?

Well, we've already found one case, which is Lantz's poll for choosing which route he should write next. Which, IMO, is enough by itself to justify increasing it.

There's always exceptions to the rule, but I'd say probably 90% of the polls that will ever be created will use 10 or less. Regardless, I've put it up to vote. When the poll ends, the winner is implemented. That's just the way it works.

Actually, I think Mike was questioning the case where you might have a vote spread of say, exaggerating it:

10: 50
20: 49
30: 49
40: 49
50: 49

etc. Where FPTP would say 10 wins, but obviously, the desire for a higher number is much better supported. It's just its supporters aren't in agreement on how much higher it should be.

While possible, that's so incredibly unlikely that in that case I'd probably just do a 3-day runoff or something between 20-50, and then a 3-day runoff between 10 and the winner of that.

If they STILL can't decide, 10 would win, because it's obvious they can't make up their fucking minds.

[Mike1984]

Doesn't matter. When one has multiple choices, one has to win. It 50% of people vote for a centrist government, 35% vote for right-wing, and 15% vote for left-wing, does that mean the government should be center-right because there were more right-wingers than left-wingers?

Yes, it should be, actually

But, in this case, the situation I was thinking of is more ;like 30% voting for the right wing, 34% for the centre and 36% for the left wing. The left wing would win despite the fact that 64% of the people would clearly prefer the centrist candidate.

It's possible I'm misremembering. Regardless, the uses of needing it would still be slim. At the same time, acceptable limits must be achieved, because if a spammer got on here and found unlimited options... he'd have a field day.

Well, of course (and, if you look at my previous posts, you'll see that I said exactly that several times).

See above. Do you want to be the idiot who does Damage Control when something like that happens?

Well, obviously not, but I can't see why it would be a problem if you set the limit to e.g. 100 (as an example of something which would be beyond anything anyone will ever need). I agree that making it unlimited is not plausible, but setting the limit higher than anything we're ever possibly going to need is.

There's always exceptions to the rule, but I'd say probably 90% of the polls that will ever be created will use 10 or less.

That's probably true, yes, but for the 10% that won't be it's a serious hassle.

Regardless, I've put it up to vote. When the poll ends, the winner is implemented. That's just the way it works.

Well, fair enough.

While possible, that's so incredibly unlikely that in that case I'd probably just do a 3-day runoff or something between 20-50, and then a 3-day runoff between 10 and the winner of that.

It's not anything like as unlikely as you think, actually, at least not in a slightly less absurd form. You could easily get, for instance, 10 votes for 10 and then 5 for each of 25, 30, 35 and 40.

This seems like a reasonable idea, though.

If they STILL can't decide, 10 would win, because it's obvious they can't make up their fucking minds.

What do you mean by "can't decide"...?

[Dark Pulse]

Yes, it should be, actually

But, in this case, the situation I was thinking of is more ;like 30% voting for the right wing, 34% for the centre and 36% for the left wing. The left wing would win despite the fact that 64% of the people would clearly prefer the centrist candidate.

Wrong. 34% voted for the centrist. 30% voted for the right. You can't - and don't - combine them like that. By this logic, the opposition should win every election, since the votes of the "losers" would out-combine the votes of the winner.

Well, obviously not, but I can't see why it would be a problem if you set the limit to e.g. 100 (as an example of something which would be beyond anything anyone will ever need). I agree that making it unlimited is not plausible, but setting the limit higher than anything we're ever possibly going to need is.

And I don't think we need more than 10, and definitely not more than 50. The community seems to be settling on 30 right now, but there's still several days left to vote. Until the poll expires, it's still technically anyone's game.

That's probably true, yes, but for the 10% that won't be it's a serious hassle.

In your opinion. Some other people may not care as much as you do. That's also what this vote will resolve.

What do you mean by "can't decide"...?

Simply put, if even after a runoff they can't come to a consensus, then it's clear they won't get anywhere with their votes because they're too busy bickering and bitching. The whole point of a runoff is to come to some mutually-agreeable choice. If they can't do that not once, but twice, they've effectively proved their futility at swaying people towards one choice or another - and in that case a more concrete vote would win. You can't have runoff after runoff or the whole system gets held up.

Think of it like papal elections: They shut all the cardinals in, and they don't go anywhere until they agree on a candidate by a majority. The runoff is basically the "shut-in period." If you can't agree on one within 3 days, then you're just going to keep fighting, so there's little point in continuing.

[Mike1984]

Wrong. 34% voted for the centrist. 30% voted for the right. You can't - and don't - combine them like that. By this logic, the opposition should win every election, since the votes of the "losers" would out-combine the votes of the winner.

However, when you're voting on a clearly numerical question (like this one) it's absurd to use that logic. If someone votes for 50, they'd obviously support 45 over 10, and to assume that they're neutral when comparing 45 to 10 (especially with non-retractable votes, which means that someone who voted for 50 can't change to 45 if they see that 50 can't win) is just plain absurd, and somewhat unfair.

And I don't think we need more than 10, and definitely not more than 50. The community seems to be settling on 30 right now, but there's still several days left to vote. Until the poll expires, it's still technically anyone's game.

Well, we definitely do need more than 10. I doubt we need more than 50 (or even 30), honestly, but there's no harm in it, so I don't see why not to set it as high as possible.

Simply put, if even after a runoff they can't come to a consensus, then it's clear they won't get anywhere with their votes because they're too busy bickering and bitching. The whole point of a runoff is to come to some mutually-agreeable choice. If they can't do that not once, but twice, they've effectively proved their futility at swaying people towards one choice or another - and in that case a more concrete vote would win. You can't have runoff after runoff or the whole system gets held up.

What, so if one option gets 45% of the vote in the run-off, then you'll stick with 10 by default even though that option is clearly more popular than it?

Think of it like papal elections: They shutter all the cardinals in, and they don't go anywhere until they agree on a candidate by a majority. The runoff is basically the "shut-in period." If you can't agree on one within 3 days, then you're just going to keep fighting, so there's little point in continuing.

Well, the problem with that is that, if the cardinals don't agree initially, they change their votes until they do. We can't do that, so even if everyone would mutually accept 30 as an option, the poll can't adjust to reflect that.

[ItsaRandomUsername]

Over Nine Thousand for me.

Granted, I'm pretty sure most polls won't need to go higher than 50 or even 40, but I'd rather be safe than sorry.

[Theocrass]

I voted for 30, 'cuz I wanna be popular. ;_;

[Jase]

Well, the problem with that is that, if the cardinals don't agree initially, they change their votes until they do. We can't do that, so even if everyone would mutually accept 30 as an option, the poll can't adjust to reflect that.

Which is when you do what DP said he'd do: start a new thread with a new poll to sort that out.

[Dark Pulse]

However, when you're voting on a clearly numerical question (like this one) it's absurd to use that logic. If someone votes for 50, they'd obviously support 45 over 10, and to assume that they're neutral when comparing 45 to 10 (especially with non-retractable votes, which means that someone who voted for 50 can't change to 45 if they see that 50 can't win) is just plain absurd, and somewhat unfair.

Wrong. The question isn't "Do you support a higher number over a lower number." The question is "Which of these choices do you support." That's not how this poll works, despite what you think. You're voting for one specific, exact choice - not a range. If 10 gets 50 votes, and 45 gets 49 votes, and the others are scattered, 10 and 45 get a runoff. If there's ties, I would then runoff the tie and compare the tie to the nominal winner. If the tie can't be broken because nobody's willing to budge, then the nominal vote wins.

Well, we definitely do need more than 10. I doubt we need more than 50 (or even 30), honestly, but there's no harm in it, so I don't see why not to set it as high as possible.

Because there is no "limit." It's a number I enter.

What, so if one option gets 45% of the vote in the run-off, then you'll stick with 10 by default even though that option is clearly more popular than it?

Winner of the tied run-off will face the nominal winner. Basically, if any option gets more than 50% in the first round, it wins automatically. If no option gets a 50% majority, the top, say, 2-3 choices do a shorter run-off. Obviously, ties need to be broken as well.

Well, the problem with that is that, if the cardinals don't agree initially, they change their votes until they do. We can't do that, so even if everyone would mutually accept 30 as an option, the poll can't adjust to reflect that.

Partially why I said I was looking into seeing if there's mods that allow vote retraction. Even if not, I can still edit the poll results if asked. In the case of runoffs, votes are anonymous; the reason the first round is public is so that people can't say "I voted, this rule sucks" and they obviously didn't.

[Mike1984]

Wrong. The question isn't "Do you support a higher number over a lower number." The question is "Which of these choices do you support." That's not how this poll works, despite what you think. You're voting for one specific, exact choice - not a range. If 10 gets 50 votes, and 45 gets 49 votes, and the others are scattered, 10 wins, period. The only way it wouldn't win is if there's ties, as I would then runoff the tie and compare the tie to the nominal winner. If the tie can't be broken because nobody's willing to budge, then the nominal vote wins.

But that's clearly not true, though. If I vote for 50, whilst I prefer 50, I still would rather 45 over 10, and to claim otherwise is just plain stupid.

Although, the way you're doing it seems to resolve this issue.

Because there is no "limit." It's a number I enter.

Well, OK, then just something far higher than the alternative.

Winner of the tied run-off will face the nominal winner. Basically, if any option gets more than 50% in the first round, it wins automatically. If no option gets a 50% majority, the top, say, 2-3 choices do a shorter run-off. Obviously, ties need to be broken as well.

Ah, OK.

So, basically, if no one option gets an absolute majority first time around, you're going to do a run-off between the top few options?

That makes sense, and resolves most of the issue I have with the vote being split.

Partially why I said I was looking into seeing if there's mods that allow vote retraction. Even if not, I can still edit the poll results if asked. In the case of runoffs, votes are anonymous; the reason the first round is public is so that people can't say "I voted, this rule sucks" and they obviously didn't.

Well, yeah, but the mod isn't here now, and that means that it's impossible to reach a "consensus" because most people will have voted before that consensus is reached.

[Jase]

People could just cheat and chat in the thread before voting.

[Mike1984]

People could just cheat and chat in the thread before voting.

Well, yeah, but that's not necessarily reliable....

[Dark Pulse]

But that's clearly not true, though. If I vote for 50, whilst I prefer 50, I still would rather 45 over 10, and to claim otherwise is just plain stupid.

Although, the way you're doing it seems to resolve this issue.

Well, it's simply not that kind of vote. If you vote someone for elected office, you're voting that person, not someone "similar to them" or "like them." That's not how these sorts of votes work. Also, since you quoted my post pre-edit, I'm just clarifying that if there is no vote with >50%, then yes, a runoff of the top handful of choices (2-4) will be undertaken.

Ah, OK.

So, basically, if no one option gets an absolute majority first time around, you're going to do a run-off between the top few options?

That makes sense, and resolves most of the issue I have with the vote being split.

Yes. At the same time, if something is >50% when the voting period ends, it instantly wins - no questions asked.

Well, yeah, but the mod isn't here now, and that means that it's impossible to reach a "consensus" because most people will have voted before that consensus is reached.

So you simply debate, and reach a consensus, before the vote is put in. Of course, nothing's stopping the tally from happening earlier, but the voting will be anonymous in tiebreakers (NOT runoffs!) so that this way, there's no finger-pointing, accusations of backstabbing, etc.

[Mike1984]

Well, it's simply not that kind of vote. If you vote someone for elected office, you're voting that person, not someone "similar to them" or "like them." That's not how these sorts of votes work.

Yeah, but we're not voting for an elected office, we're voting for a number in a range of numbers, where it is clearly obvious that someone who votes for 50 will prefer 45 to 10.

Also, since you quoted my post pre-edit, I'm just clarifying that if there is no vote with >50%, then yes, a runoff of the top handful of choices (2-4) will be undertaken.

Well, in this case it's not too much of an issue (although it may be in the case of the run-off).

Yes. At the same time, if something is >50% when the voting period ends, it instantly wins - no questions asked.

Well, of course.

But, in that case there's no question of "vote-spreading" causing the effect, so it's not a problem.

So you simply debate, and reach a consensus, before the vote is put in. Of course, nothing's stopping the tally from happening earlier, but the voting will be anonymous in tiebreakers (NOT runoffs!) so that this way, there's no finger-pointing, accusations of backstabbing, etc.

Well, that makes sense, but it's not always easy to work out what the consensus actually is.

[Dark Pulse]

Yeah, but we're not voting for an elected office, we're voting for a number in a range of numbers, where it is clearly obvious that someone who votes for 50 will prefer 45 to 10.

For the third time, no we're not. We're voting for a number. If you want 45, you're voting for 45. You're not voting for 40, 41, 42, 43, 44, 46, 47, 48, 49, or 50. You're voting for 45.

I really don't know how to make this any more obvious. That's not how I'm running this poll. You're voting for an EXACT number. No more, no less. End of story.

Well, that makes sense, but it's not always easy to work out what the consensus actually is.

This is the problem of the tied voters, not mine. They have to resolve this - or cede their vote.

[Mike1984]

For the third time, no we're not. We're voting for a number. If you want 45, you're voting for 45. You're not voting for 40, 41, 42, 43, 44, 46, 47, 48, 49, or 50. You're voting for 45.

I really don't know how to make this any more obvious. That's not how I'm running this poll. You're voting for an EXACT number. No more, no less. End of story.

And, again, I don't care about an exact number, and I doubt that anyone else does, either. I just have a general idea of how high it should be.

Not that it matters much, given that you're holding a run-off to resolve the issue if no one option gets 50% the first time around. That's basically all I was advocating anyway.

This is the problem of the tied voters, not mine. They have to resolve this - or cede their vote.

I was talking more about the run-off, actually.

What happens if no option gets 50% in that? Does the one with the most votes win?

[Dark Pulse]

And, again, I don't care about an exact number, and I doubt that anyone else does, either. I just have a general idea of how high it should be.

Not that it matters much, given that you're holding a run-off to resolve the issue if no one option gets 50% the first time around. That's basically all I was advocating anyway.

Actually, I'm pretty sure you're the only person thinking of a vote as being "ranged." But oh well; it hardly matters. You advocate no limit at all, as do a few other people.

We'll see where it lies in six days. It's that simple.

I was talking more about the run-off, actually.

What happens if no option gets 50% in that? Does the one with the most votes win?

In that case, yes. In the run-off, simple majority wins, since by that point it's pared down to 2-4 options.

[Cruor]

I am, though I didn't appreciate being made out to be a dick simply because I didn't change the vBulletin default of 10, plus IIRC the Beast's limit was 10 votes anyway and nobody ever seemed to have a problem with that. At the same time, I do see the logical argument for increasing it, but at the same time, I question how many people would actually make use of it. With the exception of "Favorite Character" polls, can you really think of any sort of question that'd need more than 10 choices? I can't. Admittedly, it's somewhat foolish to also say "Well, why raise it if only a few polls will ever need it?" but at the same time, like I said, people tend to get indecisive and a bit antsy when a whole load of choices are thrown at their face. It's psychology, basically - too many options and the mind shuts down.

This was the average poll size.

[Dark Pulse]

This was the average poll size.

It was also a character poll, and thus - as I've noted - not your "average" poll. Your "average" poll would be ones that have nothing to do with characters, because, simply put, character polls are bound to have lots of choices. Character Polls are really the only ones that will benefit from a higher amount of choices, and honestly, in my opinion, bumping up the number of choices seems to be people just being too lazy to do it in rounds.

But, personal feelings aside, I will adjust it to whatever gets voted in.

[Cruor]

How about, it was the average poll size inside Events & Gatherings.