An election system

[mike-l]

And even if you do accept all the axioms, it still isn't a mathematical proof. You have the additional axiom that "Optimizing social welfare" means "picking the state of the world all individuals would prefer if they were in a state of uncertainty about their identity.", and further, that Individuals would prefer to maximize their expected value, as opposed to say, minimizing their variance, or maximizing their worst outcome, or minimizing the chances of being below expected value, or maximizing the chances of being above it, or maximizing the maximum they could be, etc etc etc. And then more axioms that a normalization exists.

[Qaanol]

We would argue that the way you model irrationality is through ignorance factors. People have an imperfect estimate of their own utility. Tactical voters act rationally with regard to maximizing that estimate. That model is good because it doesn't care why their estimate is wrong. It could be irrational behavior, or just miseducation from lying TV ads. Ignorance is ignorance.

No, rational voters act tactically with regard to maximizing that estimate. Most people are not rational, most people can be put in a situation and shown a mathematical proof of optimal rational behavior and still choose an irrational course of action. Most people do not want to act like perfectly rational agents. There are plenty of tactical voters whose tactics derive from things other than rational behavior.

Different strategies is realistic. I don't know what "interpretations and reactions" means. I assume interpretation means there's some additional "ignorance factor" applied to the poll results. So one voter could take "35% chance Romney wins" to mean "29% chance Romney wins", and another could read that differently. But then, different "reactions"?

Okay, suppose we have four rational voters. They’re looking at the same polling data, where Romney’s at 35% and Santorum’s at 30%. One of them might interpret that to mean “Romney has a 35% chance to win, and Santorum has a 30% chance to win.” The other three might interpret that to mean “Romney has a 70% chance to win, and Santorum has a 20% chance to win.”

Of these last three, one might react by saying, “I was going to vote for Romney, but he’s going to win without my help so I’m not going to waste my time casting a redundant vote.” One might react by saying, “I was going to vote for Santorum, but he’s going to lose regardless so I’m not going to waste my time casting an irrelevant vote.” And one might react by saying, “I was going to vote for Dr. Paul, but he has no chance so I’m going to vote for Obama.”

The added tabulation complexity pales in comparison to the net utility increase seen e.g. here:
http://ScoreVoting.net/BayRegsFig.html

Au contraire. What that chart shows is that introducing more values in the range has almost no potential to improve the outcomes: approval voting already provides so much of a benefit over FPTP and IRV, that putting more values in the range has only negligible room to improve it further.

Having a non-uniform issues-space is exactly how things work in real life.

Can you be more clear about what that means? In the simplest case, I'm assigning issue-space coordinates on two dimensions. I pick X=5, Y=-2. If you offset that, you just move the origin, which does nothing. If you "compress" one dimension, relative to another (e.g. odds of X=4 are the same as odds Y=2), I'm not even sure that would have any effect in terms of relative outcomes among the different voting systems. I'd be curious to see.

I’ve been updating my model to use a system where the distribution of people for each issue has a random number N different peaks. The peaks are simple triangle bumps, as (left end of peak) + (width of peak)*(rand + rand)/2, where the width is (rand) and the left endpoint is (rand)*(1 - width). Each time a person is drawn from that distribution, first one of the N peaks is chosen (each peak has its own probability of being chosen, which sum to unity) and then the person’s opinion is drawn from that peak’s distribution.

I just simply cannot understand why in God's name you'd do this. It sounds totally arbitrary, not a way of modeling any real-world phenomenon.

If you look at actual ideological positions on a Nolan chart (e.g. the aggregate results of the OK Cupid politics quiz), you see what looks pretty much like a Gaussian distribution. If you're trying to be realistic, it would seem you want to find examples of real preference distributions, and try to create a utility generator which mimics those realistic distributions. It's not clear to me how your idea above has anything to do with modeling reality.

You’re telling me that people’s opinions on “When should abortion be legal?” looks like a Gaussian distribution? Bullshit. It looks like the sum of two sharply peaked Gaussians with vastly different means, and perhaps a smaller, broader third Gaussian with peak near the middle.

I’m approximating gaussians with triangles because it’s simpler for me to code right now. If/when I get better data and/or feel like improving the model, I’ll switch to a more realistic set of issue distributions.

[broken lader]

It is mathematically proven that the social utility function is just the sum of the individual utilities.
http://ScoreVoting.net/UtilFoundns.html

That page does not say what you think it says.

Seeing as how I co-founded a non-profit with the author of the page, I think I know what it says.

It says "it would be nice for social utility to be the average/sum of individual utilities. What do we have to assume to make that the case?"

Incorrect. It shows how, based on a series of seemingly irrefutable premises, the only social utility function that can possibly be correct is summing individuals' cardinal utilities.

It also lists various problems with and objections to the assumptions.

It states some potential objections, but then refutes them. E.g.

Anti-Continuist: Any probability p>0 that my child dies (no matter how small p is), is worse than me losing 1 penny of money. Therefore the Continuity axiom is false.
Continuist: I know you are lying, because I know you bent down to pick up a penny and therefore stopped watching your child for 30 milliseconds. Also, you are not presently tossing pennies over your child's head to protect her against incoming meteorites.

It does not say "it is mathematically proven that social utility is just the same of individual utilities" by any stretch of the imagination.

That's indeed what it says.

[homunq]

While I have great respect for readers of xkcd as a group, could I suggest that this is probably not the best forum for this discussion? There's already a listserv for this, with a number of people with diverse opinions and expertise. Sure, there's a couple of people who are near-cranks, but probably no more than there are here; and overall the intelligence and expertise is quite high. (Actually, in the time I've participated there, I've seen several people, including myself, mature in their opinions; so there's hope even for the cranks. I've also seen others who were too stubbornly crank-ish, give up and leave when people continued to push them on their questionable statements.)

See http://wiki.electorama.com/wiki/Election-methods_mailing_list for more info. I strongly encourage people to take it over there.

Cheers.

[mike-l]

Au contraire. What that chart shows is that introducing more values in the range has almost no potential to improve the outcomes: approval voting already provides so much of a benefit over FPTP and IRV, that putting more values in the range has only negligible room to improve it further.

I would like to see how the results look with our population bias added in, I expect IRV to do much better than before, as the qualitative difference between my original simulation and yours was an on average skewed population. In any case, I agree, and given that rational voters will never give a non min/max range, it seems pointless.

Bullshit.

I agree

@homunq thanks for the link. I imagine qaanol and I will keep posting our thoughts here, but I think I shall subscribe to this list as well as a quick perusal seems to be pretty objective, unlike certain other sites.

[broken lader]

Au contraire. What that chart shows is that introducing more values in the range has almost no potential to improve the outcomes: approval voting already provides so much of a benefit over FPTP and IRV, that putting more values in the range has only negligible room to improve it further.

It is not "negligible" by any stretch of the imagination. We're talking 1/12th to 1/6th as much improvement as the INVENTION OF DEMOCRACY.
http://scorevoting.net/vsi.html

[homunq]

You're welcome.

Again, I'd encourage you to move this discussion wholesale. You're free to post things here, of course. But if your goal is to find and spread the truth, you should seek the broadest forum possible; both to spread the truth you have, and to expose and discard any falsehoods you hold. (And if you're worried about distractions, you could of course just ignore any threads which you didn't start.)

[broken lader]

a quick perusal seems to be pretty objective, unlike certain other sites.

You won't find a more objective and accurate site on this topic than ScoreVoting.net and electology.org. If you're insinuating that we're not "objective", I'd love to see your evidence for that. As far as I can tell, your metric for "objective" is "adding in a bunch of misguided support for worse voting methods", which is precisely why I don't use that discussion forum.

If you operate rationally, you see that Bayesian Regret is the right metric. Which shows you that Score Voting is the best voting method (among those simple enough to be practical for political use).

[mike-l]

Alright, been doing some reading on electowiki and election-methods. There's definitely a fair bit of support for approval voting there, but there's also quite a bit of support for a seemingly strictly better method, called Majority Choice Approval, which is similar to range voting in that you partition candidates into scores, but instead of just adding up the score, which causes range voting to strategically degenerate into approval voting, it instead sequentially checks all votes above X until one candidate has a majority. If at some point multiple winners attain a majority there are a number of ways to resolve it, eg taking the person with the most 'top votes (highest ranking)' or the person with the most 'approval votes (not lowest ranking)' or anything in between (of course, always within the group that attained the majority at the same time), or a 2 candidate runoff can be run amongst top candidates picked with one or more of those ways.

I'm looking to see if they've run any simulations, or I may just run some myself, but this system alleviates pretty much all of my concerns with most voting systems.

[Qaanol]

MCA and Range voting can be described, succinctly, as “Ways of making Approval Voting more complicated”. I admit the 3-level system, whether you call the levels “Preferred, Approved, Unapproved” or “Excellent, Acceptable, Unacceptable” or “Approve, Neutral, Disapprove” or “2, 1, 0”, has a certain appeal.

But when you get down to implementing it, and you have to explain to Joe the Plumber, and to Granny down in Florida, and all the other voters how it works, anything complicated is going to sound like shenanigans. If it’s not immediately and trivially obvious to an average person how their vote will be counted, and how the winner will be determined, then the system is too complicated. If it requires more than a one-sentence explanation, most people will not bother to try to understand.

And that is where FPTP shines. “Vote for one person, whoever gets the most votes wins.” Everyone can understand that. It has no trickses, it has no subtlety, it doesn’t hide anything. The directions are simple, the tallying method is simple, and anyone can look at the results and instantly see who won.

The only other voting system that even comes close in terms of simplicity, is Approval Voting. “Vote for or against each candidate, whoever gets the most votes wins.” It works just like all the yes-or-no questions people are used to on ballots: each candidate is a yes-or-no question, either you approve of that candidate or you don’t. The tallying is equally simple, just count up all the votes.

And remarkably, Approval Voting is a gigantic improvement over FPTP. In fact, to my knowledge the only time any voting system outperforms Approval Voting, is under the assumption that voters are naïvely honest to their own detriment. On the other hand, under the assumption that voters tactically cast ballots to maximize their own impact on the election, no voting system does better than Approval, and most do a lot worse.

In order to get a change from FPTP to a better voting system, we need a supermajority of election system experts to agree on one single option to use. Consensus among experts would be even better. If there’s a divisive split among the people who have studied the matter, it is almost guaranteed that neither side will win and we’ll be stuck with FPTP, which I think all of us agree is very poor. Once we get our consensus choice implemented, then we’re all immediately much better off.

As such, I propose Approval Voting. In the limit of tactical voters, both Range and MCA reduce to Approval Voting anyway. In practical terms, everybody can understand how Approval Voting works. And it can be implemented with almost no change to existing ballots: just make each candidate their own line item, so people can vote for or against them.

Approval Voting is a better gauge of public opinion than either IRV or Condorcet, and drastically simpler to implement, utilize, tally, and verify. Approval Voting is on par with MCA and Range voting in terms of reflecting the will of the people, and much simpler to explain and use.

Furthermore, all the objections that critics throw at alternative voting systems, like “Some people get more votes than others” or “Some people’s votes count for more than others” are easily rebuffed by Approval Voting. After all, when each candidate is their own line item, then every voter gets the same number of votes: one per candidate. Either a yes or a no, just like every other question on the ballot.

[mike-l]

I'm more than willing to agree that approval is the best step forward from FPTP, both because of its plausibility and ease of implementation

MCA is not tactically AV though. Let's use the preferred approved unapproved variant, where ties are broken by approved+ votes. Suppose among 3 candidates I really want A and really don't want C, and I'd be ok with B. voting approved for B can help A win if close to half the people prefer B. It can only help C win if c is prefers by a majority and more people approve of B than C And b is not preferred by a majority. I have no idea what the odds are of that, but there are certainly situations where the chances of helping A win are much greater than th chances of helping C win, so in those cases people would rationally and sincerely vote at all 3 levels

Also, by your own data condorcet outperforms AV by your metric, and you have only tested 4 systems under admittedly unrealistic conditions, so it's pretty disingenuous to say that no system to your knowledge outperforms AV

[Annihilist]

I dislike any system whereby individual politicians who have power in parliament owe more to their party than to the people who elected them. The point of having a bunch of parliamentarians isn't merely counting which party has more points, but rather those parliamentarians are supposed to do stuff, and each individually have power to keep the government from pulling a fast one.

Under a party list PR system, if you disagree with your parties position for whatever reason, you are basically screwed. You might as well be an entry on a spreadsheet.

Under an individually elected system, if you disagree with your parties position for whatever reason, and can get enough of the same people who voted for you last time to support you anyhow, you aren't screwed. You get to keep your job in parliament.

I want my parliamentarians to be have leverage against their parties.

Here in Australia our politicians live and breathe this notion. They don't represent you to the party. They represent the party to you. All they do is try to sell you their political party so you can vote for them and give them power. They hold no regard for the people who elected them.

[Derek]

Here in Australia our politicians live and breathe this notion. They don't represent you to the party. They represent the party to you. All they do is try to sell you their political party so you can vote for them and give them power. They hold no regard for the people who elected them.

If the American political system does anything well, I think it's that it strongly avoids this problem. There is a large degree of variance within parties, and primaries are often hotly contested elections themselves (see: Current Republican primaries).

[Annihilist]

Here in Australia our politicians live and breathe this notion. They don't represent you to the party. They represent the party to you. All they do is try to sell you their political party so you can vote for them and give them power. They hold no regard for the people who elected them.

If the American political system does anything well, I think it's that it strongly avoids this problem. There is a large degree of variance within parties, and primaries are often hotly contested elections themselves (see: Current Republican primaries).

I am quite fond of this system. In Australia, the party leaders are voted in by the parliamentary representatives of that party. In America, they are voted in by the people (as far as I can tell). I think we should change to primary voting in Australia so we get better and more direct representation.

Not sure what you know about Australian politics, but theres a guy I'm fond of in the right-wing party called Malcolm Turnbull. I don't normally support the right wing party, but this guy stands up for his personal beliefs and has defied the party numerous times. He's a credible and reliable personality, rather than just another faceless person regurgitating party lines and shit. And he's more liberal than most of the left-wing major party. the Ronpaul kind of reminds me of him.

Yeah, he is the preferred leader of the right-wing party here, but within the party he was voted out. So we the people don't get a say in this.

[Mechanicus]

Here in Australia our politicians live and breathe this notion. They don't represent you to the party. They represent the party to you. All they do is try to sell you their political party so you can vote for them and give them power. They hold no regard for the people who elected them.

If the American political system does anything well, I think it's that it strongly avoids this problem. There is a large degree of variance within parties, and primaries are often hotly contested elections themselves (see: Current Republican primaries).

I am quite fond of this system. In Australia, the party leaders are voted in by the parliamentary representatives of that party. In America, they are voted in by the people (as far as I can tell). I think we should change to primary voting in Australia so we get better and more direct representation.

Not sure what you know about Australian politics, but theres a guy I'm fond of in the right-wing party called Malcolm Turnbull. I don't normally support the right wing party, but this guy stands up for his personal beliefs and has defied the party numerous times. He's a credible and reliable personality, rather than just another faceless person regurgitating party lines and shit. And he's more liberal than most of the left-wing major party. the Ronpaul kind of reminds me of him.

Yeah, he is the preferred leader of the right-wing party here, but within the party he was voted out. So we the people don't get a say in this.

In Britain, we're still mostly like Australia in that respect, but we're moving more to having MPs whose views not only vary within their parties, but they vote and speak out. We had some primaries for party candidates and it produced some rather good and independent people. So independent that the plan to roll out open primaries for all candidacies was silently dropped.

Australia's got a good check-and-balance system though. The Senate's quite active and powerful with its permanently hung membership.

[Annihilist]

Here in Australia our politicians live and breathe this notion. They don't represent you to the party. They represent the party to you. All they do is try to sell you their political party so you can vote for them and give them power. They hold no regard for the people who elected them.

If the American political system does anything well, I think it's that it strongly avoids this problem. There is a large degree of variance within parties, and primaries are often hotly contested elections themselves (see: Current Republican primaries).

I am quite fond of this system. In Australia, the party leaders are voted in by the parliamentary representatives of that party. In America, they are voted in by the people (as far as I can tell). I think we should change to primary voting in Australia so we get better and more direct representation.

Not sure what you know about Australian politics, but theres a guy I'm fond of in the right-wing party called Malcolm Turnbull. I don't normally support the right wing party, but this guy stands up for his personal beliefs and has defied the party numerous times. He's a credible and reliable personality, rather than just another faceless person regurgitating party lines and shit. And he's more liberal than most of the left-wing major party. the Ronpaul kind of reminds me of him.

Yeah, he is the preferred leader of the right-wing party here, but within the party he was voted out. So we the people don't get a say in this.

In Britain, we're still mostly like Australia in that respect, but we're moving more to having MPs whose views not only vary within their parties, but they vote and speak out. We had some primaries for party candidates and it produced some rather good and independent people. So independent that the plan to roll out open primaries for all candidacies was silently dropped.

Australia's got a good check-and-balance system though. The Senate's quite active and powerful with its permanently hung membership.

I suppose so. We're still pretty shitty in other aspects though.

[Ixtellor]

I think a change as suggested in this thread would only serve to confuse voters and create new problems both forseen and unforseen, and more importantly I don't see the new system resulting in any benefit to society or our democracy.

We have an ignorant population and a non-perfect system of electing officials on our behalf, yet we still get the paradox of mass politics where our government seems to function well inspite of these flaws.

My last ballot was 17 pages long and basically impossible for the average voter to have any rational opinion or knowledge of all those candidates. But moving to the new system would amplify the problem.

Instead of allowing ignorant voters the ability to choose between "party I tend to agree with versus party I tend not to agree with" you would be expecting them to vote arbitrarily for hundreds of people they don't know, whose positions they dont' know, and for many jobs they don't understand.

I envision a lot more "Alvin Green" situations than the best case situations described by proponants of changing the system. (google it)

I wish I lived in a nation that cared more about Congressional representation than Jersey Shore, but we don't. Therefore, Simpler serves a better and more reasonable expectation for the electorate.

[Роберт]

I think a change as suggested in this thread would only serve to confuse voters and create new problems both forseen and unforseen, and more importantly I don't see the new system resulting in any benefit to society or our democracy.

We have an ignorant population and a non-perfect system of electing officials on our behalf, yet we still get the paradox of mass politics where our government seems to function well inspite of these flaws.

My last ballot was 17 pages long and basically impossible for the average voter to have any rational opinion or knowledge of all those candidates. But moving to the new system would amplify the problem.

Instead of allowing ignorant voters the ability to choose between "party I tend to agree with versus party I tend not to agree with" you would be expecting them to vote arbitrarily for hundreds of people they don't know, whose positions they dont' know, and for many jobs they don't understand.

I envision a lot more "Alvin Green" situations than the best case situations described by proponants of changing the system. (google it)

I wish I lived in a nation that cared more about Congressional representation than Jersey Shore, but we don't. Therefore, Simpler serves a better and more reasonable expectation for the electorate.

I think the onus is on you to prove that approval voting would lead to MORE Alvin Green situations, rather than fewer.

[Qaanol]

I think a change as suggested in this thread would only serve to confuse voters and create new problems both forseen and unforseen, and more importantly I don't see the new system resulting in any benefit to society or our democracy.

Wait, what? Approval voting will instantly and single-handedly eliminate the “spoiler effect” of vote-splitting (which cost Gore the 2000 presidential election, recently cost Eliot Cutler the 2010 Maine gubernatorial election, and presumably happens a lot more often than I’m aware of since I only know the major incidents that directly affect me.)

We have an ignorant population and a non-perfect system of electing officials on our behalf, yet we still get the paradox of mass politics where our government seems to function well inspite of these flaws.

“…and therefore we should not make any attempt to improve the voting system”? Are you being deliberately dense?

My last ballot was 17 pages long and basically impossible for the average voter to have any rational opinion or knowledge of all those candidates. But moving to the new system would amplify the problem.

Wait, what? Approval Voting would not change the ballot itself whatsoever. You’d still have a line for each candidate, the only difference is you could vote “Yes” for each candidate you approve of, and “No” for the rest.

Instead of allowing ignorant voters the ability to choose between "party I tend to agree with versus party I tend not to agree with" you would be expecting them to vote arbitrarily for hundreds of people they don't know, whose positions they dont' know, and for many jobs they don't understand.

Wait, what? Candidates would still have their party affiliations listed on the ballot. You could still bullet-vote for just one candidate if that’s what you want. And where do you get “hundreds of people” on the ballot for “positions they [voters] don’t know”? You’d still be voting for the same positions as you currently do. Each party would still be limited to just 1 candidate per seat (or perhaps 2 or 3 candidates?), and serious candidates are going to have campaigned for name recognition anyway.

I envision a lot more "Alvin Green" situations than the best case situations described by proponants of changing the system. (google it)

Wait, what? Are you talking about his surprise primary win? How does that have anything to do with the voting system used, and how on earth do you get from there to the idea that such a thing would be more common in Approval Voting, rather than essentially eliminated from ever happening?

I wish I lived in a nation that cared more about Congressional representation than Jersey Shore, but we don't. Therefore, Simpler serves a better and more reasonable expectation for the electorate.

Um, that’s one of the central arguments I’ve put forth in favor of Approval Voting. It is equally simple as the current FPTP system, and on par with the best theoretical voting systems for choosing winners that accurately reflect the will of the people.

By switching from FPTP: “Vote for one candidate, whoever gets the most votes wins.”
To Apprroval Voting: “Vote for or against each candidate, whoever gets the most votes wins.”

You get, immediately and for free:

• Complete elimination of the “spoiler effect”
• Accurate measurement of support for third-parties

Also, if you let each party field 2 or 3 candidates, then voters have an incentive to approve the “least bad” of the opposing major parties. That means you’ll often get really high approval, like 70+%, for the candidates that are actually in the middle of the spectrum for the district they represent. As a result of representing the middle ground, those legislators will be better able to work together and compromise.

At the national level, this will result in a congress that is not nearly so polarized as we have now, with much higher approval ratings. At the state level, you’ll get legislatures that include more different parties, since some individual towns are apt to elect candidates from outside the parties that dominate the national stage.

To the best of my understanding, there is literally no possible way for Approval Voting to ever be worse than FPTP for any reason. It is equally simple to use as FPTP is, it eliminates the “spoiler effect” wherein individual voters who approve of two candidates currently have to guess which one is “more likely to win” (and if too many guess wrong they all lose), and it allows the true spread of voter opinions to show up on the final tallies, because voting for a third party is no longer a “wasted vote”.

[Yakk]

We have an ignorant population and a non-perfect system of electing officials on our behalf, yet we still get the paradox of mass politics where our government seems to function well inspite of these flaws.

“…and therefore we should not make any attempt to improve the voting system”? Are you being deliberately dense?

No, Ixtellor doesn't behave deliberately dense.

[Ixtellor]

Approval voting will instantly and single-handedly eliminate the “spoiler effect” of vote-splitting

Yes it will do that one "good" thing. Not a logical reason to do something. Closing all the coal plants tomorrow would reduce global warming, but that doesn't mean its a smart thing.

which cost Gore the 2000 presidential election

Over simplification of a problem you don't understand.
Gore lost for numerous reasons,
#1 was probably the fact he told Clinton to hide in a closet and not go around the country talking about the economical miracle and popular reforms he had done as President.
#2 Electoral College, he got between 500,000 and 150,000 more votes than bush (Absentee ballots not counted).
#3 Eye rolling at the debates.
#4 Butterfly Ballot
#5 SCOTUS overstepping its bounds and saying "Sure counting votes is good, but there isn't enough time" (the basically acknowledged their intrusion in their ruling.
#6 Brooks Brother Riot.
.... Way at the bottom... spoiler effect.

recently cost Eliot Cutler the 2010 Maine gubernatorial election,

#1 Run off elections.
#2 Run a better campaign.
#3 You can't prove Eliot would have won with your system.

“…and therefore we should not make any attempt to improve the voting system”?

Opinon, Conjecture, Bias, Supposition.

I never said we shouldn't improve the voting system, so don't straw man me.

I am saying your "solution" maybe isn't the best one and might cause more problems than it solves.

So after not recognizing that your spouting opinions and assigned me a position I don't hold... who is being dense? More importantly, quit being intellectually dishonest.

Approval Voting would not change the ballot itself whatsoever

Also, if you let each party field 2 or 3 candidates,

So you say "would not change the ballot" followed with a plan to change the ballot.

How does that [Alvin Green]have anything to do with the voting system used,

Alvin Green was a situation of too many people on the ballot coupled with the fact that voters are ignorant and outside of a few offices (President, Gov) don't really know the people running.
In your proposed system of adding MORE candidates (you both deny and admit it in your posts), I see more Alvin Green situations. Where people get surprise wins due to insignifant things like 1) Being listed first 2) having a catchy name 3) Not having a 'bad' name.

I realize this is speculation, but there is a history of too many choices causing electoral
problems.

You get, immediately and for free:
Accurate measurement of support for third-parties

Conjecture.

To the best of my understanding,

Rather than go through the rest of you post, lets stop on this. Its the problem.

To the best of your limited knowledge, bias, inability to forsee problems, ignorance of the voting population and why they really vote, overestimation that people are voting for anything as noble as ideas policy or political ideology.... your idea sounds good... to you.

My favorite line:

there is literally no possible way for Approval Voting to ever be worse than FPTP for any reason.

Thats not very sciency.

No, Ixtellor doesn't behave deliberately dense.

We have an ignorant population and a non-perfect system of electing officials on our behalf, yet we still get the paradox of mass politics where our government seems to function well inspite of these flaws.

“…and therefore we should not make any attempt to improve the voting system”? Are you being deliberately dense?

No, Ixtellor doesn't behave deliberately dense.

This is Serious Business. Hold yourself to a higher standard.

[Yakk]

We have an ignorant population and a non-perfect system of electing officials on our behalf, yet we still get the paradox of mass politics where our government seems to function well inspite of these flaws.

“…and therefore we should not make any attempt to improve the voting system”? Are you being deliberately dense?

No, Ixtellor doesn't behave deliberately dense.

This is Serious Business. Hold yourself to a higher standard.

Meh. You don't behave deliberately dense. If you are making a mistake and not understanding it, as far as I can tell, it is because you don't understand it, not because you are holding your eyes closed and saying "I don't want to see this".

So if Qaanol wants to say "are you being an idiot, or just pretending to be one" as a way of "hiding" a personal attack, he should just make the personal attack.

(And yes, it also contained snark, because I was cranky and thought the wording was amusing. Sorry for the snark.)

[Роберт]

How does that have anything to do with the voting system used,

Alvin Green[sic] was a situation of too many people on the ballot

This just in: Ixtellor thinks ballots should have no more than 1 person on them.

[broken lader]

I think a change as suggested in this thread would only serve to confuse voters and create new problems both forseen and unforseen

Actually, Score Voting, and especially Approval Voting, have been shown to reduce the number of spoiled ballots. In that sense, they are simpler than Plurality Voting. Voters literally screw up their ballots less often.

Also, I once conducted a small exit poll with Score Voting. In Beaumont, Texas. While it was a relatively small sample size, I saw no one give any indication that they didn't understand what to do.

I agree that ranked voting systems can be confusing (and tend to have a lot more spoiled ballots than Plurality). That seems partially due to the fact that people don't know what the tabulation algorithm is, and tend to assume it's Borda.

I don't see the new system resulting in any benefit to society or our democracy.

The benefit is massive. You can see the difference in Bayesian Regret between Approval Voting and Plurality Voting here. It roughly doubles the benefit of having democracy in the first place.

It also solves the spoiler problem. E.g. look at this simplified example election:

35% Conservative > Progressive 1 > Progressive 2
33% Progressive 1 > Progressive 2 > Conservative
32% Progressive 2 > Progressive 1 > Conservative

If everyone votes for their favorite with Plurality Voting, then the conservative wins. But a huge 65% majority of the voters would have taken either of the progressive candidates over the conservative. With Approval Voting, it's very likely that one of the progressives would win in this scenario, leading to a vastly more representative outcome. The average voter would be much better off.

We have an ignorant population and a non-perfect system of electing officials on our behalf, yet we still get the paradox of mass politics where our government seems to function well inspite of these flaws.

So we shouldn't try to make things vastly better?! That's simply irrational. And moreover, we are certainly not functioning "quite well". We are getting ourselves into a staggering amount of debt. We are ill prepared for peak oil. Global warming looks to be an increasingly significant threat. Last decade, we went into a war with the wrong country, leading to the deaths of something like hundreds of thousands of civilians. Uhh.. no, things are not functioning nearly as well as they could be.

My last ballot was 17 pages long and basically impossible for the average voter to have any rational opinion or knowledge of all those candidates. But moving to the new system would amplify the problem.

I agree that this can be a problem. We use Instant Runoff Voting in San Francisco, and we had 16 candidates for mayor last fall. You could rank up to three. But I found it challenging to research all the options. I did a lot of research, and worked on two candidates' campaigns, and ultimately made what I thought was a pretty smart choice. But when I went door knocking on those campaigns, I talked to so many voters who just had no real knowledge of the differences between the candidates. They had kind of picked out "their guy"(s) based on previous loyalty, or some views stated in some pamphlet, or some endorsement by their labor union, or whatever. And they generally couldn't give you any policy details that could really justify their choice.

That being said, the problem with Plurality Voting is much worse. Because essentially you only have two choices. And they are the two party nominees. Anyone else who runs might be highly qualified, but generally will not win, because people think they can't win, and won't want to waste their vote on them. So just about the only people who can get elected are those who've spent years climbing the ladder inside the established parties, making deals with special interests, etc. And with two choices, it doesn't matter all that much how informed or ignorant the electorate is. The most you can possibly do by informing the electorate is to switch the winner from one to the other.

With e.g. Approval Voting, you get a lot of more viable options, many of whom are better than the best electable option you'd have with Plurality Voting. Winners tend to be candidates with the best name-recognition-TIMES-approval-ratio. So it does help to have connections and have the money to promote yourself, but also having much stronger support among those voters who've heard of you (aren't "ignorant" of your platform) can allow you to defeat a less appealing candidate who would traditionally be considered much more "electable".

It seems like you kind of think people will make smarter choices if you just give them two choices. But that's relative. Their choice can only be as good as the two options they get.

Instead of allowing ignorant voters the ability to choose between "party I tend to agree with versus party I tend not to agree with" you would be expecting them to vote arbitrarily for hundreds of people they don't know, whose positions they dont' know, and for many jobs they don't understand.

If there are hundreds of candidates running, people will just ignore the vast majority of them. The greater name recognition a candidate has, the greater his or her chances of having voters spend time considering his or her views. But in the end, maybe 5-10 candidates will meet the threshold of being strong enough to earn the attention of enough voters to potentially win. Here in San Francisco, there was definitely a point where we were able to see from polls and other indicators, that only about 7 candidates really had a realistic chance. And for people who followed really closely, it was more like 3-4 who seemed to have a chance. So as the election drew near, voters were indeed able to focus their attention on a smaller pool of candidates.

Compared to the old system of basically having two or three legitimate choices, and possibly facing vote splitting, I think it has been a great thing to have more options.

I envision a lot more "Alvin Green" situations than the best case situations described by proponants of changing the system. (google it)

If you're letting this stop you from supporting an improvement to the horrendously bad Plurality Voting system, I strongly urge you to think more about situations like this.

One system you might actually like is Asset Voting. You get some number of votes ("asset"), and the candidates have some amount of time after the election in which to redistribute their votes. In my hypothetical example above, the two progressives may have each had less asset than the conservative, but they could have made some agreement where one would give his asset to the other, in exchange for fighting for some of his policies. Then the progressive majority (or conservative majority, if we flip the example) would be reflected. Some don't like the idea of entrusting their vote to a candidate, but you could say that if you don't trust a candidate with your vote, how can you trust him to be elected to office and represent your values there? One obvious benefit of this system is that you only have to find one candidate that you strongly support and trust. You trust him, as someone who knows the political landscape, to have a very good idea of which other candidates share his values, and thus to know how best to redistribute your vote, should he not have enough votes to win. The idea is, this is easier than having to figure out for yourself which other candidates you might want to support out of a large pool of candidates.

[broken lader]

There's definitely a fair bit of support for approval voting there, but there's also quite a bit of support for a seemingly strictly better method, called Majority Choice Approval

Yeah, MCA does a little better than Approval Voting, if you look at the green table here.
http://scorevoting.net/StratHonMix.html

Worth the extra complexity? Probably not. And Score Voting kills MCA, assuming a reasonable fraction of the electorate votes honestly with Score Voting. I expect 10% to 50% of voters to be generally honest with Score Voting, based on all kinds of evidence I've seen over the past six years.

[broken lader]

Wait, what? Approval Voting would not change the ballot itself whatsoever. You’d still have a line for each candidate, the only difference is you could vote “Yes” for each candidate you approve of, and “No” for the rest.

But more candidates would run. That's his point I think.

But I agree with pretty much everything you said.

[Yakk]

You can independently control the number of candidates who run.

The size of US ballots has more to do with (A) multiple levels of governments "crowding in" on the same election system, (B) voting for the dog-catcher equivalent habit of US governments.

The result is a single ballot with:
US President
US Senate
US Congress
State Senate
State Congress
State appeals court judge (3 of them)
State district attorney
5 State ballot initiatives
1 State constitutional amendment
Mayor
City council
City Chief of Police
Independent Deputies
City court Judge (7 of them)
City chief dog catcher
4 City ballot initiatives
County Sheriff
County Reeve
County dog catcher
17 County ballot initiatives

But it is worse than this. You get a ballot with Candidates that you aren't allowed to vote for on them. On the ballot it tells you "if you live in district X, vote only here", but it goes and includes huge lists of people you aren't supposed to vote for.

Texas sample ballots:
http://www.co.travis.tx.us/county_clerk ... sample.asp
and you'll see how bad it actually is.

Here is an elections Canada sample ballot:
http://www.elections.ca/content.aspx?se ... &lang=e#8c

Provincial and Municipal ballots look similar. You get a list of Candidates on one side, and a place to put a mark on the other side. They then fold up in half, so you can carry them back to the ballot box to drop them in without people seeing what marks you made.

You get the ballots that correspond to who you are voting for. City, Provinces and Federal levels of government hold elections separately, each roughly every 4 years.

[mike-l]

To be fair, Canadian municipal ballots are slightly more complicated, you vote for mayor, counsellors, and school board chairs. But nothing like the linked Texas ballot.

There's really no doubt though that approval voting is significantly better than FPTP, and by far the easiest system to change to from FPTP. People in this thread and elsewhere have done many simulations, and AV beats FPTP in every case. FPTP is better than random selection and not much else. The debate is only whether AV is the finish line or if it's just a stepping stone to a better system.

Also, do you seriously vote for dog catchers? Edit, ok that was a joke, but the texas ballot was crazy enough that I believed it.

[broken lader]

Qaanol,

Did you run those voting method simulations? If so, I'm quite curious to see what kind of results you got. I would expect that Score Voting would come out in first place (unless you added in some of the weird "exotic" methods that are seemingly too complex for real public elections). But I could be surprised. In any case, I'm sure we'd all find the results interesting.

Thanks