This seems to be pretty consistent with previous Range Voting polls I have created.

My most recent one had just the "big fish", and like this poll, it has Gore with a respectable lead, followed by Edwards, and then Obama trailing shortly behind. It uses 1-5 ratings:

Al Gore : 3.15 John Edwards : 2.75 Barack Obama : 2.62 Hillary Clinton : 1.68

A much older poll that I began last year, with around 4,000 participants at this point, has it:

Barack Obama : 3.10 (WINNER, by far) Al Gore : 2.76 John Edwards : 2.59 Hillary Clinton : 2.34 Wesley Clark : 2.26 Dennis Kucinich : 2.17 Bill Richardson : 2.08 Joe Biden : 2.00 Christopher Dodd : 1.68 Harry Reid : 1.60

Note: There were 47 candidates in this poll; these are just the major Dems pulled out.

Obama was the overall winner in my older poll. So these more recent results may reflect that his "aura" has begun to dim as people start to look past his iconic nature and become more skeptical about him as a realistic choice for President.

And for those of you who think this mock election did a good job of representing the electorate's real sentiment, consider the project to get Range Voting used in the Iowa DP caucuses.

range voting is fine for straw-poll type ballots that mean very little.

And it is also the best election method for political elections, as it has the highest voter satisfaction index.

when my vote actually matters, though, i'm not going to jeopardize my candidate's chances by rating a strong opponent highly.

Well then you're very "altrustic" voter, because you are willing to sacrifice your own happiness with the election result in order to help your very favorite candidate. Of course we know that some voters will throw themselves into the fire for a candidate they love - just look at the 97,488 Florida voters who voted for Nader in 2000.

But the fact is that most voters care about themselves more than the candidates, and so they will wisely not behave in the way you describe.

in this ballot, i rated both Edwards and Clinton pretty high, though not as high as Obama, of course. if this had been the real thing, i would have slapped big fat zeros on both their asses.

And then you would have been left regretting that decision if it turned out that Edwards and Clinton were the actual front-runners, and the one that you liked the least of the two won. You just effectively made your vote irrelevant by giving both front-runners a zero.

a better way to escape the one vote to the death method of voting is, i think instant run-off voting -- IRV.

You are wrong, and in fact have it completely backward. With Range Voting, voting for just one candidate is almost always a horrible strategy, whereas with IRV it "best" feasible generalized strategy to top-rank your favorite front-runner. This behavior ends up causing IRV to degrade toward plurality.

But more to the point, Range Voting yields more satisfying results than IRV. So you Mr. Voter, if you care enough to vote to get a satisfying result, have a big incentive to get Range Voting. You also have a big incentive not to foolishly give your favorite candidate a 10 and the rest 0's.

in IRV i pick Obama 1, Clinton 2 and Edwards 3. nothing there hurts my candidate.

Incorrect. Consider this IRV election.

% of voters - their vote 26% Obama> Edwards> McCain 23% Edwards> Obama> McCain 2% Edwards> McCain> Obama 49% McCain> Edwards> Obama

Edwards has 25% of the first round votes, and is eliminated. McCain then wins 51% to 49% against Obama. But Edwards is preferred to Obama by 74% of the voters, and preferred to McCain by 51%. IRV says Edwards is better than each of his rivals, yet it fails to elect him when all three run at once. The result is not just the wrong candidate wins, but the wrong party. So Obama voters, and most of the Edwards voters, are punished for being honest about their first preferences. Of course, in the real world, voters quickly wise up and learn to betray their favorites in order to top-rank their favorite of the front-runners, causing IRV to degrade toward plurality, and to achieve two-party duopoly (sorry third parties - no room for you).

In our second example, not only are voters punished for their sincerity, but their sincerity also hurts the candidate that they attempt to help.

% of voters - their vote 33% McCain > Gore > Dean > Nader 29% Gore > McCain > Dean > Nader 24% Dean > Gore > McCain > Nader 14% Nader > Dean > Gore > McCain <— ?!

McCain wins. But if the Nader voters in the bottom row had instead dishonestly voted

McCain > Nader > Dean > Gore

then Gore would have won (which they'd prefer to the old winner McCain) despite the fact this just raised their opinion of McCain from last to first place, and lowered Gore to last place. So by honestly saying that they liked Gore better than Bush, voters not only hurt themselves but Gore as well. So much for the myth that IRV encourages honest voting.

Still not convinced? Maybe you think these "hypothetical" examples are somehow unrealistic, and wouldn't happen in real life. In that case you are wrong, because they do happen, and then voters wise up and learn to vote strategically by top-ranking their favorite front-runners, not necessarily their sincere favorites; and the result is two-party domination in all IRV-elected posts in all four countries where IRV has seen long-term widespread use. If this strategic behavior wasn't going on, we would not expect IRV to lead to two party domination. Yet if it was going on, then two-party domination is exactly what we would expect.

And even if we are wrong, and IRV would somehow encourage voters to be honest angels, whereas Rangers would be strategic devils, Range Voting still outperforms IRV, making more voters more satisfied.

nothing in IRV hurts my candidate, while RV forces me to be dishonest about my real feelings.

As I showed above, this is simply false.

it's also a lot easier.

On the contrary, Range Voting is easier in some important objective ways.

1) It is simpler to count, because it only requires one round of tallying.

2) It is simpler for voters to use, because they may just score a few candidates, and "abstain" on all the ones they do not care to ponder about; whereas leaving someone unranked on an IRV ballot cannot be disregarded, because it is effectively the same thing as marking a candidate tied-for-last (bad for him). With Range Voting, an abstention simply does not affect a candidate's score.

3) Range Voting is algorithmically simpler in that a voting process can entail simply going down the list and scoring each candidate (a linear algorithmic complexity), whereas any rank order system requires one pass through the entire list to choose each subsequent pick (or a Quicksort, for the computer science wiz with a calculator for a brain). To put this into perspective, consider my 2008 Presidential election with 47 candidates. Going down the list and clicking on 1-5 stars for each candidate is a cinch. But try ordering them if you think that's easier. You'll soon realize that it is not.

Range Voting is a concept where gut instinct tends to drastically betray reality. I myself began my experience with the Center for Range Voting as an ardent detractor. But when you put initial intuition and bias aside, and look at the facts, Range Voting is the clearly superior option, and IRV reveals itself to be far too problematic to be practical for elections of any importance.

Wes Clark hasn't even announced and he comes in 5th. I like that.

Run Wes, run!

Gore is right up there in my book also, just don't think he will run.

close italics I hope.

jeez, Clay, you are a true believer.

Jus like a believe in evolution, and other things that science supports, but some doubters will never believe for reasons I do not entirely understand.

and you really won me over with that. no doubt you'll have great success with that kind of rhetorical tactic.

I don't attempt to convince people with rhetorical tactics. I intend to convince people with evidence, working in a scientific framework. If there's something about the evidence I've presented that leaves you less than convinced, please discuss it. But my sentence, as I wrote it, is not mean-spirited - it is simply a fact, which I supported with copious evidence.

What happens to write-in votes under range voting?

As someone from an entirely different part of the political landscape, who happens to think that Kucinich is the best of the Democrats (because of his opposition to the Iraq war from the beginning, and for having the cojones to start impeachment proceedings against Cheney), I'm wondering what people find objectionable about Kucinich. Why did he fare poorly in this poll?

Now that the first debate has happened, my votes would change drastically. I would give Gravel and Kucinich 10's and all the others zero. Gravel is Kucinich with balls. Kucinich is correct on impeachment. How about a re-vote?

What happens to write-in votes under range voting?

They are treated just like any other candidate. But remember, in order for a candidate to be legitimate, he must get at least half as much total score as the one who got the most. So don't think you can win with 3 "perfect 10" write-ins from your friends. You'll have 30 total points, and that won't put you in the running by any means.

You can read more about the quorum rule here.

I'm wondering what people find objectionable about Kucinich. Why did he fare poorly in this poll?

I compiled a rather long list of his positions on various issues, which I think make him a horrible candidate. I see him as a merciless crook, who has no problem regulating private business and stealing money from the public to fund various socialist projects - rather than letting them decide for themselves how to spend their own money. As if it wasn't already clear that he hated freedom, he even voted for the Flag Desecration Amendment, to stop the "atrocity" of people burning their own private property if it happens to have red and white strips, and stars. Disgusting. Get this guy out of America. He hates freedom.

Here's a sample:

• Reparations for slavery should take form of social programs. (Nov 2003)

• Affirmative action is necessary & right & must be preserved. (Aug 2003)

• Voted YES on constitutional amendment prohibiting flag desecration. (Jun 2003) <-- !!!!!!!!!

• Rated 64% by the ACLU, indicating a mixed civil rights voting record. (Dec 2002)

• Enron: Dems should become party of re-regulation. (May 2002)

• Rated 15% by the US COC, indicating an anti-business voting record. (Dec 2003)

• Establish universal pre-kindergarten programs. (Aug 2003)

• Voted NO on allowing vouchers in DC schools. (Aug 1998)

• Voted NO on vouchers for private & parochial schools. (Nov 1997)

• Voted YES on increasing AMTRAK funding by adding $214M to $900M. (Jun 2006)

• Fund 2,500 Boys and Girls Clubs in underserved areas. (Dec 1997)

• Voted NO on reforming the UN by restricting US funding. (Jun 2005)

• Voted YES on keeping Cuba travel ban until political prisoners released. (Jul 2001)

• Voted NO on implementing CAFTA, Central America Free Trade. (Jul 2005)

• Voted NO on implementing US-Australia Free Trade Agreement. (Jul 2004)

• Voted NO on implementing US-Singapore free trade agreement. (Jul 2003)

• Voted NO on implementing free trade agreement with Chile. (Jul 2003)

What I find most revealing in this poll, as well as every other on-line poll of Democrats that I have seen is how poorly Hillary does versus the national phone polls. In fact this Blue Oregon poll matches my personal discussions with my peers off-line.

Exactly. I believe this is a reflection of the fact that Range Voting eliminates the importance of electability, and thus greatly diminishes the importance of cash.

With almost all voting methods, including plurality and IRV, voters are strategically forced to vote as if their favorite candidate was their favorite front-runner, or at least the candidate who would have the best chance against the opposition candidate. Thus winning becomes, bizarrely, not just an issue of convincing people that you are the best candidate, but that you can win - that is, that others think you are one of the two front-runners.

Say you think that Edwards is your favorite, but since he doesn't have the money and apparent electability of Clinton or Obama, you know he will be seen as a loser, and thus a vote for him is like a vote for Nader - a total waste. So even if Edwards really was the favored candidate, he isn't elected. Clinton, by having her large war chest, and connection with her husband William, creates an aura of inevitability that "scares" people into voting for her - even when it's just a poll and nothing's on the line. Voter psychology is indeed rather bizarre.

But with Range Voting, this problem disappears. Go ahead and give Edwards a 10 if you like. You are still free to give high scores to other similar candidates, so as to maximize the chance that some candidate you would be happy with is elected. Top-rating your sincere favorite can never hurt you or that candidate with Range Voting. A "vote for Nader" is NEVER a "vote for Bush", as it easily can be with plurality voting or IRV. Range Voting is the voting method where we really can say, "Vote your hopes, not your fears".

Generally voters like us on BlueOregon aren't leading indicators of public sentiment (we're liberal outliers), but in the case of an election 18 months out with three credible, centrist candidates, we may be this time around.

Actually I'm an ardent libertarian. I oppose all taxes and economic regulations such as tariffs, rent control, and minimum wage laws. And I think Obama and Edwards, despite their socialist tendencies, are decent men with a lot of integrity - for politicians. But I loathe the sculpted, scripted, plastic-ness of Hillary Clinton. So don't think that this is just part of being a liberal outlier. We can all unify behind our Hillary hatred.

See, Hillary is bringing diverse people together. She's a force of unity. By dropping out of the race, she has the power to change the world. :)

Range Voting just doesn't seem suitable for elections--MAYBE a primary, but certainly not a general.

The facts say otherwise. The best voting method is the one that makes the most voters the most satisfied. The entire reason you show up at the polls is to push the election toward a favorable outcome for yourself. Range Voting gives you the greatest expected satisfaction from the election outcome.

You're not discussing how much you like them; you are choosing who you want to win. You can't want two people to win, and you can't want someone to win a certain amount more than another.

There is no such thing as "wanting someone to win". There is only how much you want someone to win. When you say you "want Obama to win", for instance, what you are really saying is, "I want Obama to win more than I want Edwards or Clinton or any one else to win".

There are a couple of easy ways to prove this, using an economics concept called "revealed preference". For instance, say that I, a corrupt government official tell you that you may have whichever candidate you want, because I will rig the election. You tell me that you prefer Obama over Edwards over Clinton. I tell you that I am going to rig the election so that Clinton wins. But if you will simply pay me 1,000 dollars, I will see to it that Edwards, your second choice, wins. You analyze this proposition and assess your wealth, and determine how much utility you sacrifice by giving up 1,000 dollars, and decide that it is worth it. The difference in value between Edwards and Clinton is greater than 1,000 dollars, in your eyes. Now you get Edwards, but I say, "Hey pal, for another 200 bucks, I'll go ahead and just make Obama the winner." Now you consider that proposition and think, "Do I like Obama so much more than Edwards that I would sacrifice 200 dollars?"

This may sound like a contrived example, but these preferences come into play realistically for you in the voting booth with Range Voting. Say these three candidates appear, based on the polls, to be virtually tied. You begin by strategically giving Obama a 10 and Clinton a 0, so as to maximize your favorite candidate's chances over Clinton. But now you have to think about where to place Edwards. Your intuition, which is incorrect, is that you should just give Edwards a 0 without a second thought, in order to have the best shot at getting your favorite candidate. But this is economically imprudent, because it may well be that Obama comes out in third place, and Clinton wins by a hair over Edwards. Now you are kicking yourself that you did not also give Edwards a 10.

Your best decision is to decide which is greater, the difference between how much you like Obama and Edwards, or the difference between how much you like Edwards and Clinton. If the latter is greater, then you are strategically wise to give both Obama and Edwards a 10, whereas if the former is greater, then you want to give Edwards and Clinton both 0's.

Here's a little page about how making wise strategic assumptions will make Range voters pleasantly surprised.

IRV mimics these features of the final tabulation and requires no translation, as range voting does.

IRV certainly does not "mimick" the effects of Range Voting, and it produces far less favorable outcomes. Here are some nice graphics which show that.

I do not know what you mean by "translation", but IRV is substantially more process-intensive than Range Voting, often requiring multiple rounds of summation, whereas Range Voting is objectively simpler to use and tabulate.

Range Voting can also be conducted on all standard plurality voting machines, where as IRV requires costly upgrades.

Range Voting also increases the number of spoiled ballots (by a factor of 7, here in San Francisco) where as Range Voting experimentally reduces them.

IRV also increases the risk of ties, whereas Range Voting reduces it.

IRV is also non-additive, meaning that two precincts can declare A the winner, but when the ballots are summed together, B is the winner. This is more bizarre than bizarre.

IRV also "requires" an ordering of all candidates - those not listed are treated as equally tied for last, which is extremely harmful. With Range Voting, a voter can simply score the handful of candidates he is knowledgeable about, and abstain from the rest, neither helping nor hurting them. Giving voters far less work to do is certainly simpler.

This is expressed more formally by Princeton math Ph.D. Warren D. Smith:

Range Voting is simpler than IRV. If you don't believe me, try writing a computer program to do both. The range voting program will be shorter. Range voting also is simpler in the sense that it requires fewer operations to perform an election. In a V-voter, N-candidate election, range voting takes roughly 2VN operations. However, IRV voting takes roughly that many operations every 2 rounds. In a 135-candidate election like California Gubernatorial 2003, IRV would require about 67 times as many operations. (In fact, range voting is simple enough that it could be done with hand calculators, if necessary.)

The way the individual votes--no ties, ranked in order regardless of the differences in sentiment between candidates--is the way the votes are counted.

In poor voting methods such as IRV, you are right. But not with Range Voting. And the result is that Range Voting has a much higher social utility efficiency.

And the primary flaw of range voting--that a candidate who is not preferred as the best choice can nonetheless win by being less polarizing than another candidate--does not exist in IRV.

It most certainly does exist in IRV, and the effect is significantly more severe. The strategy under Range Voting is essentially to pick some utility threshold, and give a maximum score to all candidates who are at least that valuable, and give 0's to the rest. This is effectively Approval Voting, which is still an excellent voting method.

With IRV, on the other hand, the general strategy is to top-rank your favorite of the two front-runners - as if you were using plurality voting! This would be like giving a 10 to your favorite front-runner in a Range Voting election, and a 0 to everyone else - which is not a good strategy. The result is that the winner tends to be the same one who would have won under plurality, or one who is almost as bad. This is also why IRV enforces two-party duopoly.

The results of IRV's extreme susceptibility to strategic voting are apparent in social utility efficiency values. For example:

Range (honest voters) 96.71% Range (strategic voters) 78.99% IRV (honest voters) 78.49% Plurality (honest voters) 67.63% IRV (strategic voters) 39.07% Plurality (strategic voters) 39.07%

So you can stay loyal to IRV, but if you use it, you are going to be less happy than you would with Range Voting.

Yo Clay -- I'm enjoying your discussions about range voting, but I gotta tell ya, you're not doing yourself any favors with your ranting about the "socialist tendencies" of Edwards and Obama.

But the thing that is so great about Range Voting is that people with diverse views can all embrace it. It will make us all more happy on average with the results of elections.

I am putting my neck out on this one. If we have range voting money and name reg will be even more important. You wont rank all the candidates you dont know and in the case of the presidental election thats going to be a lot of candidates. So the 3rd part and I candidates get screwed. Then you have to decide between Obama and Mccain for example. You dont know much about one so you give him a low score and the other a higher score. When I took the test for example I gave Clark and a few others really low scores because I know little about them. I probably shouldnt have done it but most voters will do it. They vote based on headlines they read and many times dont actually know what they are doing. todays Oregonian has a doonsbury comic that protarys that very well.

What is being elected to office, anything more than winning? I want that woman to win, and if not her then him, her, him, him, that guy and finally her over there.

But, put more formally, what you want is to be as satisfied as possible. I don't know how familiar you are with the economic principle of "expected value", but it is calculated like this.

Say your odds of winning the lottery are 1 in 1 million, and the payoff is 2 million dollars. Your expected value then, is 2 million dollars divided by 1 million = 2 dollars.

Say I offer you a 20% chance of receiving 10 dollars, or a 1% chance of winning 1000 dollars. Your expected value in the first case is 2 dollars, whereas in the second case it is 10 dollars. Choosing the second option gives you a greater expected value. This is analogous to the fact that Range Voting produces the greatest social utility efficiency. It is going to, on average, make you much happier than you would be with plurality, or IRV, or Borda, or Condorcet, or any of the myriad of other methods that have been proposed and tested against it.

But it gets even better. Suppose in the scenario above, you don't care about the greatest expected value, you just want to have the greatest chance of getting enough money to, say, satisfy your immediate hunger with a candy bar. In that case, you'd take the greater chance of getting less money. A lot of voters are like this too. They care more about getting someone they would generally approve of than they do about vying for one of their absolute favorites. In that case, they could use a threshold of, say, an honest 5, and give 10's to all candidates to whom they'd honestly give a 5 or above, maximizing the chances of getting a satisfactory candidate. This is why you are wrong to think it wise to "bullet vote" for one candidate.

Many people then ask, "Why not simply use Approval Voting, which would be identical to giving every candidate either a 0 or a 10 - disapprove or a approve?"

Well, because having those extra intermediate options, even if they might not be strategically wise to use, will cause a lot of honest voters to use them anyway, and as a result, the overall societal satisfaction with the results will go up - meaning better winners are picked. We recently explored this issue in depth.

Another important ramification of the extra options is that, by using them honestly, voters are actually served fairly well. If they want to be strategic, that takes more work to plan how to do well - you could easily mis-polarize, and hurt yourself. So simply going with that honesty is, for many voters, worth it. And the results are that society as a whole is better off.

And yet another benefit of Range Voting is the Nursery Effect, that people will tend to give a no-hoper like Nader an honest 4 or 5 or 6 if they think he's "okay", because they know he has no chance of winning anyway, and they don't feel the need to be strategic with him. The result is that fledgling candidates and third parties receive massively better initial results, which gives them the momentum they need to be taken seriously and gain popularity down the road. Once they do get that popularity, then they are "adults" ready to leave the nest and fend for themselves in the harsh world of politics. The Nursery Effect just helps them get started.

In IRV that's what you choose, and in the election that's how they count them up.

IRV is but one of an innumerable set of ordinal (rank order) voting methods, that all take ordered preferences, but compute the result differently. Consider Condorcet (condor-SAY) methods for instance. A Condorcet voting method looks at the orderings and elects the candidate who wins when paired head-to-head against all of his rivals. This initially seems like a logical way to pick the "right winner", but it is not, as the following example shows:

% of voters - their ordered preferences 35% A > B > C 33% B > C > A 32% C > A > B

Here we are forced to ask, who is the right winner? But the problem is that it doesn't matter who you pick, a contradiction ensues. Say that you claim A is best candidate, for instance. But if that is true, then it means that A is better than both B and C. But notice that if B were to drop out, C would beat A with a whopping 65% "majority". Thinking in majoritarian terms forces you to contradict yourself, and the mathematician Kenneth Arrow famously won the Nobel prize essentially for highlighting this type of scenario in a mathematical statement commonly known as "Arrow's theorem".

The problem with ordinal voting methods (Range Voting is cardinal) is that they all fail to satisfy certain basic logical assumptions of how a voting method should behave. For instance, monotonicity says that it should never hurt my candidate to rank him higher. Yet IRV can fail this, such that a group of like-minded voters can, by moving candidate B from 4th to 1st place, cause him to go from winning to losing.

Range Voting not only avoids this nonsense, but it is objectively superior using the one ultimate measure of a voting method's quality - social utility efficiency. That is, after voting in order to "get your way", how much of "your way" did you get?

So your intuitive sense that IRV somehow just feels better, are betrayed by a strict assessment of the related math. However most of the major IRV advocates are not voting method experts and either do not know/understand these concepts, or (in the case of Rob Richie) simply deliberately pretend not to know better, and repeat false statements about IRV, such as "IRV elects majority winners" and "IRV prevents spoilers/vote splitting/wasted votes".

No, it doesn't. But now is the time when the tide starts changing, and society wakes up. The Center for Range Voting was founded in 2005 by Princeton math Ph.D. Warren D. Smith, and Colorado engineer Jan Kok, and serves to promote and educate on important advancements to social organization, a filed in which elections play a major part.

The Libertarian Reform Caucus uses and endorses Range Voting, and calls IRV a "bullet in the foot". Range Voting has been discussed with increasing frequency on various web sites and political blogs. It is being promoted by computer scientists and political activists as they are starting to be awakened to its advantages. Paradigm shifts take time. IRV was invented 137 years ago, and the advances in voting theory that have allowed these insights into Range Voting have only come about in the past 10 years or so. This will not be an overnight process. And as simple as Range Voting is to use (I have conducted a successful RV exit poll in Beaumont, TX) it is clear that recognition of its superior properties is often stifled by false intuition. I began my first two weeks of experience with Range Voting as an ardent detractor, criticizing Warren D. Smith for even having the audacity to put up a web site for such an apparently foolish concept. I think of myself as a pretty mathematically astute person, having completed the majority of a computer engineering degree. Yet it took me a lot of debate before I finally understood that he was, after all, correct. So I have nothing but the utmost empathy with doubters. All I can say is that the obviousness of the need for Range Voting becomes progressively clearer the more you learn about it.

Range Voting also saves lives.

Here are some other inventions that can help humanity.

1. Hurriquake nails – a better nail
2. Detoxified cottonseeds

Time for a reality check. Is the goal an academic debate, or is the goal to run elections where ordinary people (who may not pay attention to politics in an odd numbered year) are going to be voting?

Elections are not decided by people who use language like "Range voting is a far better decision-making algorithm" or talk about strategy in IRV voting.

Face it, folks, many voters are very busy people who want to talk about WHY they choose a particular candidate, not about the procedure for voting. Mail voting made life easier for those with work and family considerations, strange work hours, long commute. Statisticians and others may not like that result, but this is about real world voters. And real world election workers---how would the 36 county clerks be trained to tabulate the election results? Should they hire 36 election mathematicians? Trust someone's computer program because the advocates say so?

Elections are sometimes close, and many elections have only 2 candidates (state reps. Roblan, Richardson, Galizio, Cowan, Girod, Berger, Minnis, Clem, Dallum, for instance). Many people work very hard at challenging jobs (medical, safety, construction, janitorial, retail--try standing on a sales floor all day!, education, etc.). Why should they applaud a change in voting system just because theory says so?

Seldom does discussion of topics like this address such 2 candidate elections as the state reps mentioned above. Even if you think you have the best voting theory and/or the best algorithm, there will be no change in voting procedures without convincing hard working ordinary people that they need to change voting procedure because people skilled in math say they should.

Those of you who advocate such systems should hit the speaking circuit: Rotary and other civic organizations, parent clubs, Grange, local party meetings, etc. If you talk to live audiences, answer their questions, and then come back on here and tell us what you learned from those face to face decisions, that would be interesting.

But don't expect those of us who think campaign finance reform (including ending pass-through as discussed by the Public Comm. on the Legislature), or nonpartisan legislature, or open primary, or similar ideas make more sense, to applaud your ideas simply because someone is excited about an algorithm.

I have discussed this face to face with an IRV advocate who thanked me for my ideas. My example was this: in Hse. Dist. 20 the 2 candidates were Berger and Garcia. Unless it can be explained to voters how they would be allowed to vote in a 2 candidate race (2 votes for either Berger and Garcia, one vote each, or are such systems only for multicandidate races?) ordinary folks aren't likely to discuss this among their friends.

As an old friend said when she burned out on politics, 5% of the population live and breathe politics. They may think they decide elections, but the folks who actually decide elections are the 95% more likely to be discussing work, family, sports, or other topics than theories of how elections are run.

Range voting does not give the highest voter satisfaction just because it represents the highest mathematical utility on a scale that you have shown to manipulate to serve your opinions about range voting anyway.

Dr. Warren D. Smith, who created the open source software used to calculate these figures, has shown me no indication that he has manipulated it to get the desired results. On the contrary, he founded the Center for Range Voting long after he had performed the simulations, and wanted some avenue for telling the world about the increasing evidence he was accumulating that Range Voting was amazing improvement to democracy.

But in any case, you are free to download the source code and point out the bias you are claiming.

The point I was making is that RV is counter-intuitive: it's not how the election is counted, and it's not how people actually vote.

But we suggest it should be the way we vote, since it is so far superior to the way we do vote. And actually there are several organizations with thousands of members who use Approval Voting, the simplest form of Range Voting.

You act like voting for most people is as if they're on Survivor, plotting strategy. That's nonsense.

It's funny that you should say that, because most critics of Range Voting say that strategic voting will be its downfall, and we have to make the case to them that

A) There is reason to believe a LOT of voters will vote honestly, and B) Even if 100% of voters are strategic, Range Voting still behaves very well. In fact Range Voting's margin of benefit over IRV doubles if voters go from being 100% honest to 100% strategic.

So I'm glad that you are now changing your tune, and admitting that many Range Voting users will NOT be strategic. That is indeed our sincere hope.

People's aim in voting is to vote for the person they want

Which is a great reason to use Range Voting, because with Range Voting, you are never penalized for giving a 10 to your sincere favorite candidate (because you can still give a 10 to your favorite front-runner). Whereas with methods such as plurality voting and IRV, voters are strategically forced to not vote for the person they most want, but instead to vote for the front-runner that they most want.

Of course voters will occasionally throw themselves in front of traffic for their favorite candidates, which we saw en masse in Florida back in 2000, when 97,488 Nader voters actually did vote for the candidate they most wanted - Nader - to their massive detriment.

So you are somewhat right. Most of the time, voters vote for their favorite front-runner, but sometimes they do as you suggest, and vote for the candidate the most want. With Range Voting, they can do both. No more dilemma.

and also in a number of cases seek to prevent the election of someone they don't want.

Which you can do in Range Voting by giving the candidate a low score, or even strategically giving him a 0.

Your assessment that IRV becomes a top-two is strange

And based on decades of historical fact from the four major IRV using countries (mainly Australia and Ireland). But here's a more formal analysis that I would kindly ask you to check out.

what it most often does is allow the possibility of a major candidate being ranked third, behind a minor candidate.

Sure, I might say "Gore > X > Bush", so X is ahead of Bush. But I won't say "X > Gore > Bush", because X either

1) Will lose to Gore, and so won't win anyway, or 2) Will defeat Gore, but will lose to Bush, whereas Gore could likely have beaten Bush, or 3) Will defeat Gore and go on to defeat Bush.

Outcome 3 is the only one that makes it advantageous to top-rank X, whereas outcome 2 is disadvantageous. So a rational voter has to ask himself, which is the more likely outcome if I top-rank a non-front-runner, 2 or 3? If 3, then go for it. But if 2, then DO NOT go for it.

It doesn't take a rocket scientist to see which is the more likely outcome, and why it is therefore irrational in IRV to top-rank anyone besides a front-runner, no matter who is your sincere favorite is.

The obvious result of this behavior is that third parties almost never win in IRV, just like with plurality - not because they are not the true favored candidate (which they sometimes might actually be) but because they are not seen to be electable by the massively strategizing electorate.

And what's more, this strategy is simple and has a big payoff, whereas forming a good strategy with Range Voting takes a lot more thought, which an average lazy voter might avoid. And even if voters are strategic with Range Voting, they get a greater average satisfaction with the outcome, by a rather large factor - even if Range Voting were to somehow encourage otherwise 100% honest voters to be 50% strategic. Even when we give IRV an enormous "head start" (or "benefit of the doubt") it fails to compare to Range Voting.

one of the main drawbacks of both IRV and RV is that low information voters totally corrupt the system beyond the first 2 or 3 candidates, since they have no idea who they are.

Ah, but Range Voting avoids this problem, by allowing abstentions (which a fair amount of voters seem to use with unknowns like Mr. Gravel here), which minimize the harm caused by voter ignorance/laziness.

In range voting you are asking the voter to place a strategic overlay on their process, when by IRV you accomplish the same thing in an intuitive fashion that mimics the way the votes are counted.

No, we are not asking voters to be strategic. We would prefer it if they were honest. And no, IRV does not "accomplish the same thing", or anywhere near the same thing, as Range Voting. The strategies and behaviors are completely different. This is evident in the social utility efficiency figures, which you claim we biased, although you have not produced any evidence of that.

Your issue with the number of rounds that need to take place has no practical relevance; it's done by machine automatically, not by the voter.

Ah, but that is another problem! You see, for years and years, IRV was done with manual counting, laboriously. But because of its complexity to count, it has begun to appear almost hand-in-hand with electronic voting machines. If you know much about election fraud, you should know that nothing makes it as simple, large-scale, and undetectable as electronic voting. The other day I spent 2 hours and 7 minutes on the phone with Rebecca Mercuri, one of the world's foremost authorities on electronic voting and elections, and I repeat her view that it is absolutely crucial that we completely eliminate all forms of voting machines, in favor of 100% paper ballots, manually counted before any member of the public who cares to watch. Beyond just being an exceptionally poor voting method, IRV is extremely conducive to the implementation of (electronic!) voting machines. Ever see HBO's Hacking Democracy? I highly suggest it. It is a sobering look at the prevalence of electoral fraud in our government. And in all honesty, I believe the problem to be far more extensive than they say it is.

Range Voting does take longer to manually count, because you now have to do one count for every candidate, instead of one for every race. But this is a linear, not an exponential increase in complexity, as I previously mentioned, and should not serve to get people hooked on electronic voting.

But RV makes the voter spend time assessing each candidates position on a spectrum--and flatly, we don't elect people by spectrum; we do it by number of votes.

Yes, and our current election method is ridiculously terrible. Range Voting is vastly better. So we should use Range Voting.

You say RV would make me more happy--but I certainly would not be happy with John Edwards winning an election in your example, when actually he placed third and shouldn't even be part of a runoff in a hypothetical.

So you are trying to say that you would not be happy with having your favorite candidate win with Range Voting if he would come in 3rd with IRV? Okay, I can make the same argument that I would not be happy having my sincere favorite candidate win with IRV, if he would actually come in 3rd with Range Voting. And since Range Voting is a much better voting method than IRV, my argument is much stronger.

It's simply fundamentally wrong to elevate a candidate when more people prefer another one

This is a common, and false, axiom that I call the "majority fallacy". It can be stated as follows:

A is a better winner than B if more than half the voters prefer A to B.

Now you say you believe this, but in a moment, you will not. Consider the following example:

% of voters - their preferences 35% A > B > C 33% B > C > A 32% C > A > B

Who is the best candidate here? Whom does your logic say we should elect? Well let's see.

67% of voters say A > B, so A is a better candidate than B 68% of voters say B > C, so B is a better candidate than C 65% of voters say C > A, so C is a better candidate than A

Uh oh! Your axiom fails, because it contradicts itself. No matter whom you claim to be the rightful winner, there is another candidate who is preferred to him by a HUGE majority, who must therefore be a better candidate, but who does not win. So clearly the best candidate is not necessarily preferred to each of his individual competitors by a majority, but rather, the one who makes society the most happy. The one who makes the most voters the most satisfied. That's where social utility efficiency comes into play. It does not fall prey to the internal paradox of the majority axiom.

Incidentally, you are apparently unaware that IRV often fails to elect the Condorcet winner (when there even is one). Here's a simple example:

26% Obama> Edwards> McCain 23% Edwards> Obama> McCain 2% Edwards> McCain> Obama 49% McCain> Edwards> Obama

Here Edwards is preferred by 51% of the voters over McCain. And he is preferred by 74% of the voters to Obama. If he ran against either one of them alone, IRV would say he was the best candidate. But when all three run at once, IRV changes its mind and says McCain is best. So IRV has the same "flaw" that you cite as an irreconcilable difference with Range Voting.

Now, if Obama, an irrelevant alternative, were to drop out the night before the election, and voters were to keep their exact same preferences, Edwards would beat McCain. Somehow having Obama in or out of the race changes IRV's opinion of who is better between Edwards and McCain. I call that just a bit bizarre. But really it is not that bizarre, since almost all voting methods fail this criterion, called "independence of irrelevant alternatives". But not Range Voting. It passes it.

and that's exactly what RV does--it makes a candidate with solid 7s but no 9s or 10s win over one with a bunch of 10s and also quite a few 0s. That's dumb, IMO.

Why is it dumb to bring one person's happiness down from a 9 to a 7 to bring another person up from a 0 to a 10? That is a net increase? If that is "dumb", please explain for me what is smart. Please present me with a single-winner voting method with a higher social utility efficiency than Range Voting.

I am putting my neck out on this one. If we have range voting money and name reg will be even more important.

We actually make the case that Range Voting will significantly reduce the importance of money. And I think it's a very convincing case indeed.

You wont rank all the candidates you dont know and in the case of the presidental election thats going to be a lot of candidates.

That's less of a problem with Range Voting than with any other voting method that I know of, because Range Voting allows abestentions. Notice here that a lot of people did not know who Gravel was, and so apparently that is why they abstained. This reduces the degree to which I candidate who is potentially good is harmed simply by not being very famous. IRV offers nothing like this, for if you fail to rank any candidate, that is effectively the same as ranking him "tied for last", and severely punishes him for not being famous. That makes money extremely important - because to have a chance, you have to pay to get your face on T.V. and billboards, so you can be famous.

So the 3rd part and I candidates get screwed.

Exactly the opposite happens. Range Voting produces what is called the Nursery Effect, in which candidates like Nader receive approximately 40 times as support from Range Voting. This is why the Libertarian Reform Caucus recently published an essay by me entitled "Why Third Parties Must Get Range Voting or Die". Yes, it's perhaps overly dramatic, but the point was to get the readers' attention, and the LRC had already called IRV a "bullet in the foot" before I even found them.

When I took the test for example I gave Clark and a few others really low scores because I know little about them.

Then you presumably would have ordered them very low with IRV as well, whereas with Range Voting you could have taken advantage of the "no opinion" option, but chose to "play it safe". Range Voting at least gives you that option.

I probably shouldnt have done it but most voters will do it.

1) Empirical evidence, like my small but interesting exit poll in Beaumont, TX, shows that many voters will actually use the abstention option, especially if it is placed near the zero.

2) If you think most voters will do it, then they'll do it with IRV as well; but in IRV it

3) Leads to lower voter satisfaction

So your arguments actually turn out to work in Range Voting's favor, and against IRV. Thank you for making my case for me. :)

Clay,

I appreciate your dedication. Really, I do. But for the average voter (like me), this is snoozeville. If you can't figure out an easier way to explain it or a better way to run it, it doesn't matter.

I think everybody talking here needs to take a look in the mirror. Geeks with ice cream or just geeks.

How about something simple for a change?

Clay's yammering about social utility theory, wild-ass hypotheticals, and other BS is tending to obscure the simplicity of range voting. I think you could explain range-voting in under 3 minutes to a classroom of third-graders, and they'd all get it

But you are missing the point Kari. If the goal is simply to explain how Range Voting works, then citing social utility efficiencies would be the last thing I would want to do. But when it comes to proving that Range Voting is the best voting method, then citing those figures is crucial. Social utility efficiency is the be-all-end-all metric of voting method quality. It is unfortunate that it is rather esoteric, but that does not diminish its significance one bit, and I am a bit disappointed that you would make some rather insightful comments, but then go on to hand wave like this about the most crucial issue in all of election theory.

Simply put, if using Range Voting worked exactly the same, but it produced a poor social utility efficiency, then it would be a bad voting method. That is not "BS", that is reality.

Furthermore, much of what you call "wild *** hypotheticals", are severe and serious pathologies that do happen in real life. The net effect of these pathologies is their probability of occurring times the harm they do when they occur - just like how one calculates expected value. That is what Bayesian regret measures. Voter satisfaction index is a scaled expression of utility on an opposite axis. So in effect, one tells us how bad a voting method is, while the other tells us how good it is. It's two different ways of looking at the same information.

Clay: " That is what Bayesian regret measures. Voter satisfaction index is a scaled expression of utility on an opposite axis."

Clay, my friend,

You seem to not be getting the message, which causes me great Bayesian. I suggest you put that on your axis and express it.

Is the goal an academic debate, or is the goal to run elections where ordinary people (who may not pay attention to politics in an odd numbered year) are going to be voting?

Absolutely. And there's reason to believe that Range Voting will cause them to pay a lot more attention - because if you look at trends in voter turnout, nothing keeps it down like the feeling that the winner is a foregone conclusion, whose incumbency and funding will make him nearly impossible to dethrone. But since Range Voting takes away the favorite betrayal incentive, and provides for dramatically greater expression of preferences for a host of alternatives, it can be expected to increases the sense that elections are contentious, and thus increase turnout.

Elections are not decided by people who use language like "Range voting is a far better decision-making algorithm" or talk about strategy in IRV voting.

Indeed, they are decided people like those Texans I performed a Range Voting exit poll on - who had no problem using it. And how about all those politically astute folks using Range Voting at HotOrNot.com. :)

Face it, folks, many voters are very busy people who want to talk about WHY they choose a particular candidate, not about the procedure for voting.

Yeah but just consider that Oakland and Davis (California), and Peirce County (Washington), and Minneapolis, all just voted to enact Instant Runoff Voting, which is worse and more complicated than Range Voting, and doesn't even have the support of a lot of third party members, because they know it is just as lethal to them as plurality. But imagine presenting a simpler and better method that should have the support of all third party members (once they have been properly educated about it). I think there's good reason to believe that enacting Range Voting will be less work than many believe. And besides, it's so important that people should try anyway, because it could well be the difference between life and death for billions of people within a few decades or centuries.

Mail voting made life easier for those with work and family considerations, strange work hours, long commute.

But then you have proof of how you voted, and can sell your vote, which I think should be perfectly legal, but I'm just saying.

Statisticians and others may not like that result, but this is about real world voters. And real world election workers---how would the 36 county clerks be trained to tabulate the election results? Should they hire 36 election mathematicians? Trust someone's computer program because the advocates say so?

Ah, the "complexity" argument. Please read this before you go on down this road.

Elections are sometimes close

So it is handy that Range Voting significantly reduces the probability of ties.

and many elections have only 2 candidates (state reps. Roblan, Richardson, Galizio, Cowan, Girod, Berger, Minnis, Clem, Dallum, for instance).

In which case voters can just give out 0's and 10's, or be honest if they want better results for society. And this isn't as superfluous as you may think - don't forget about write-ins.

Many people work very hard at challenging jobs (medical, safety, construction, janitorial, retail--try standing on a sales floor all day!, education, etc.). Why should they applaud a change in voting system just because theory says so?

Because they'll be dramatically happier with election results, and have fewer wars, and have laws which better represent the will of the people, and elections which are less dependent on money, more lives saved, and for a myriad of other reasons to numerous to even list.

Furthermore I would ask you to be really careful before say such dismissive things as "because theory says so". Those social utility calculations were RIGOROUS, and employed hundreds of millions of trials, using 720 unique combinations of 5 separate parameters. And Range Voting won in ALL of them, typically by a HUGE margin. So pending some uncovering of some massive flaw in the source code which could account for the appearance of this huge quality disparity, I think there is an overwhelming scientific case that these utilities are very meaningful, and that you are doing a disservice by speaking about them so dismissively. Especially when it appears that you have not really taken any time to research them. If you were to research the software, procedure, etc. and spot some glaring error, we would really like to know, so that we could improve the results. But I do not think it helps anyone to dismiss something out of hand.

Seldom does discussion of topics like this address such 2 candidate elections as the state reps mentioned above. Even if you think you have the best voting theory and/or the best algorithm, there will be no change in voting procedures without convincing hard working ordinary people that they need to change voting procedure because people skilled in math say they should.

It all starts with you. One handy way to start is to point out to third parties that Range Voting is life or death for them. If you can gently explain that to them, you have the potential to win over an enormous number of converts. I've certainly turned a great deal of Libertarians onto RV.

Those of you who advocate such systems should hit the speaking circuit: Rotary and other civic organizations, parent clubs, Grange, local party meetings, etc.

Co-founder Warren D. Smith gave a lecture a few months back at Carnegie-Mellon, and is currently in Maryland doing the same thing before hundreds of people. I'm working hard to get Range Voting used in the elections at San Francisco State University. I have a major Seattle food co-op ready to test out Range Voting for its internal elections. I'm working to schedule a presentation for my local San Francisco Green Party, as well as the Libertarian Party. But I'm also primarily a musician, with a full-time hospital office day job, working to try to make a career out of music. This Range Voting stuff means I typically get about 3-4 hours of sleep per night, and feel like a zombie at work all day. So why do I do it? Because I think it's far and away the single most important issue in the world. And I think that as more and more people start to slowly peel away the layers of the case to be made for Range Voting (which is severely hindered by the unfortunate false intuition that tends to plague it - which happened to me at first too!), the amount of work I have to do will diminish, and this will become more of a movement. It has to, otherwise I don't think humanity makes it very much longer.

If you talk to live audiences, answer their questions, and then come back on here and tell us what you learned from those face to face decisions, that would be interesting.

I learned about Range Voting last August. By November, I was doing Range Voting exit polls in Texas, and had an essay published on the home page of the Libertarian Reform Caucus. On my way to and from work each day, and during my breaks and lunch I am on the phone and conducting emails with various academics, union leaders, and university officials trying to further this cause. Believe me, I know what public reaction to it is generally like. It is always the same process, over and over and over again. People always have the same misconceptions, that are usually brought up in the same order, and usually take about the same number of days for them to finally grasp. It's like clockwork. But I put up with it, because I think it really is very very important. My sincere hope is that, so will others here, and then they will devote their time to this enormously important cause. It may not happen, but at least I can die knowing I did my best to positively effect the world, if even in vain. Life is really quite short. What else can we possibly do but try to leave the world better than we found it?

But don't expect those of us who think campaign finance reform (including ending pass-through as discussed by the Public Comm. on the Legislature), or nonpartisan legislature, or open primary, or similar ideas make more sense, to applaud your ideas simply because someone is excited about an algorithm.

Range Voting is as much of an improvement over plurality voting as plurality voting is over non-democratic random selection of the winner. So if you think those issues are more important than getting Range Voting, then you think they are more important than the advent of democracy itself. I think that's beyond silly.

I have discussed this face to face with an IRV advocate who thanked me for my ideas. My example was this: in Hse. Dist. 20 the 2 candidates were Berger and Garcia. Unless it can be explained to voters how they would be allowed to vote in a 2 candidate race (2 votes for either Berger and Garcia, one vote each, or are such systems only for multi-candidate races?) ordinary folks aren't likely to discuss this among their friends.

I don't know how this is a question. With 2 candidates, IRV is identical to plurality. With Range Voting, most people would probably just (sensibly) give their favorite a "10" and the other a "0". Probably many would not, and would actually give, say, a 7 and a 3 if that's how they honestly felt - and that altruism would actually be better for society. And since Range Voting makes it permissible to vote for more than one candidate in each race (indeed, it is encouraged) it dramatically lowers the number of spoiled ballots. And since it can just treat an illegible vote for a particular candidate as an abstention for that candidate (which doesn't help OR hurt him) Range Voting also virtually eliminates the harm done by what little spoilage there is. So Range Voting is great for 2-candidate races, and some voters may even choose to give both candidates a high score, so as to make sure no write-in steals the show. Perfectly okay to do.

As an old friend said when she burned out on politics, 5% of the population live and breathe politics. They may think they decide elections, but the folks who actually decide elections are the 95% more likely to be discussing work, family, sports, or other topics than theories of how elections are run.

I think that's basically nonsense, since the it is the activists out there, like Bev Harris, who fight to investigate electoral fraud, and waged political warfare to get IRV passed in several municipalities recently, and started blogs like Daily Kos, which now have enormous impact on the direction of elections. This is especially true in the Internet age.

Granted, a lot of activists don't do much, and I must admit that before I discovered Range Voting, I never particularly pictured myself as any kind of political activist - because I just didn't see what little ol' me could ever do, unless I wanted to work my fingers to the bone being a lawyer or banker or something terribly mundane like that (to a computer geek and musician like myself), in order to have the money to make any activism purposeful. But I think I happen to have walked into a movement which has such huge potential because it is so desirable to...well...everybody.

Now compare that to my fondness for libertarianism. If I go out and work as hard as I possibly can to eliminate taxes and economic regulations, and laws against drugs and prostitution, and flag burning, and all of that - where am I really going to get? There are people on the other side of the fence who are pushing just as hard in the opposite direction. So it's really just a waste of resources, unless you are so rich that you have money to burn.

But with Range Voting we have concept which I can be so optimistic about, because it is almost 100% non-partisan. I cannot think of a single good reason that any rational person who understands it would not want it. None. Both the most ardent libertarians and the most ardent authoritarians will be happier with election results simply by switch to Range Voting - and the improvement will be comparable to that achieved by the advent of democracy in the first place. There are just no words to describe how significant that is.

So maybe I am optimistic to a fault, but I also think Range Voting is so important that, even if you think it only has a one in a billion shot of catching on, you - a rational person who presumably does not want humanity to destroy itself - should be zealously working to get it.

As for issues like campaign finance reform, I will fight you tooth and nail, because I don't believe there should be any restrictions or regulations on how much money is given to candidates. I think you should be able to sell your vote if you like. It's yours to sell.

But we can BOTH be better off by getting Range Voting, so it's something that brings people together who might otherwise be brutal enemies, if they can both understand why it is so beneficial to them, and work to make it happen.

So, I hope that gives you a better idea of why I am so optimistic and concerned about Range Voting.

Regards, Clay

Posted by: t.a. barnhart | Apr 29, 2007 11:16:40 PM

with range voting, this is what i see the average voter doing:

stares at ballot stares at 8 candidates stares at sliding scale thinks: wtf? screw it. Seinfeld's on in 10. i like ... him most. i want him to win. the rest of these -- 0. 0. 0. etc

Your intuition turns out to be wrong. In my exit poll in Beaumont, TX (I cannot emphasize enough that this is where I was) there were 5 candidates for governor, I just casually asked people walking by if they'd give me a moment of their time to score the candidates, and almost all of theme scored all 5 candidates.

What's more, even lazy voters might choose to score 2 or more of the options, which is still a benefit over plurality voting.

And what's more is that the races that voters will be more lazy about are the ones that are less significant. When it comes to things like governor, or president, you can bet your life that voters will take that extra few seconds to score a few additional candidates...otherwise why would they have taken 30 minutes out of their lives to stop by the polling place, knowing that the odds their vote would break a tie are infinitesimal?

elections don't need to me made more rational. we need to make them idiot-proof and theft-proof.

And Range Voting does make them more idiot proof, because it, perhaps counter-intuitively, lowers the number of spoiled ballots, and allows for abstentions to reduce the harm of voter ignorance ("Me dumb voter...me not know these three guys...me just give them "no opinion"), and picks VASTLY better candidates.

As for making elections theft-proof, I just spent 2 hours and 7 minutes on the phone on Friday with Rebecca Mercuri, one of the world's leading experts on that very issue (and our own Warren D. Smith just co-authored a formal math paper with MIT's Ron Rivest about "fraud proof" election methods). Her position, for which she made a very compelling argument, is that the one and only solution is

• PAPER BALLOTS
• Manual counting
• Counted in front of any member of the public who wishes to observe

Her point is, it's stupid to use these incredibly complicated secure voting concepts that use encryption and other advanced mathematics concepts, when nothing much is ever done about fraud even when it is detected. So we have to stop the fraud in the first place. So we need to get rid of voting machines COMPLETELY and have transparent elections (including transparent ballot boxes like most respectable democracies have - but not the USA's).

in time, we can worry about voters giving a damn. half of them don't. simplicity is key, and RV ain't.

Simplicity is key, to an extent. But what we find is that

A) Range Voting is plenty simple, because lots of real world polls on average Joe's (not college students) show that it is.

B) The "complexity" of Range Voting - that you are supposed to score all the candidates - actually is beneficial because it both lowers spoiled ballots, and picks enormously better winners.

BTW, regular people have LOTS of experience understanding range voting. Ever watch figure skating in the Olympics?

Hah! I'm currently trying to get the US Figure Skating Organization to use Range Voting for their internal elections for this very reason. :)

I should go after the gymnastics organizations next.

Clay -- All that wonderfulness to all concerned is STILL JUST YOUR OPINION.

It does not prove anything for anyone else more worldly-wise than you.

And we will not be conned into an election system that seems to give more weight to extremists than to people who actually consider things.

Why am I not surprised that you are a Libertarian ... i.e., you believe in grandiose schemes that rely on fairy tales on human nature, and which in reality empower the most greedy, most ruthless, most deceitful members of what was formerly a community ...

Human beings don't work on logic, never have, never will. So-called logic is generally a distraction from the reality of actual human systems.

Kari - that's not the best example to give. Due to the inherent problems with Range Voting, Figure Skating has recently moved to an "Ordinal System". Which is another word for Instant Runoff Voting.

I was not aware the system had changed, but this article both suggests the new system may be worse, and does not describe it as an ordinal method. It sounds to me like scores are still being used.

Furthermore, it would be wrong to say that an ordinal method had been used "because of problems" with scoring, because there is no known ordinal voting method with a higher generally social utility efficiency than Range Voting (certain, rather strange, methods can do slightly better in a limited set of cases). Whatever problems they are having with traditional scoring would be better remedied with a larger panel of judges. Instead they are trying to curb the problem with some intricate system of random eliminations, or something along those lines.

Finally "ordinal method" is not another word for "instant runoff voting". IRV is but one of an infinite number of ordinal voting methods, and it is nowhere near the best. Virtually any Condorcet method, as well as Borda, DMC and many others are known to be significantly better than IRV.

The problem with Range Voting is obvious to anyone who actually looks at it, different power is given to voters based on their own internal scaling.

This is not a problem, but a benefit, as shown by social utility efficiency calculations. And if you as a voter do not like this, you are more than free to exaggerate. If 100% of voters employ that strategy, we get Approval Voting, which is still the second best of the common voting methods, and still hugely superior to IRV. In fact the improvement that Range Voting offers over IRV doubles if the electorate goes from being 100% honest to 100% strategic. And Range Voting outperforms IRV even if we make gross accomodations for IRV, pretending that it will mysteriously encourage voters to be more honest than with Range Voting (when in fact it seems that the opposite is true).

In Ice Skating for example, some judges tended to vote with a range of 7.2 to an absolute top of 9.8. Others scaled the same exact group on a range of 6.0 to 10.0. The result? The first judge's opinions ended up mattering less than the judge who used the larger range.

Yes, it strikes me as particularly bizarre that judges do not employ their full range. Ideally they should see all the contestants, and then give the worst a 0, and the best a 10, and then score the others with respect to those two.

But this is a poor criticism because

1) This is nothing like the way voters rate candidates, where they go into the voting booth having already seen all the contestants (and real world polls, including this one show that real world voters do not vote like those judges)

2) I am willing to bet that if you ran a social utility efficiency calculation using such disparate ranges, Range Voting would still outperform most any ordinal method. I will ask Warren D. Smith to do this, hopefully sometime soon - because I simply do not have the time to modify his code right now when I'm spending all this time with you guys. :)

And that is in a public vote setting. Every vote is immediately evaluated by the public. When votes are anonymous, gamesmanship like T.J. proposes would allow voters to disenfranchise anyone who didn't vote the extremes. That would tend to be the casual voters and the ignorant.

But Range Voting is more resistant to strategic voting than methods like plurality and IRV, so the more voters vote strategically, the bigger the improvement that Range Voting offers over the other methods, which violates your intuition of course. I would start by reading this.

I'd also like to say that Clay Shentrup seriously needs several more years in an institution, hopefully one that provides instruction on reasoning. His website is full of logical fallacies (largely either begging the question or informal logic), and his writings here are plainly contradictory with the science of economics.

Most of the material at RangeVoting.org was written by Princeton applied math Ph.D. Warren D. Smith, who got his undergrad at M.I.T., and just co-authored a paper on fraud-resistant voting techniques with M.I.T.'s Ron Rivest, the 'R' in RSA encryption, which is used by just about every secure online banking transaction in the world.

So I would suggest you attempt to replace these vague unsupported claims with a good dose of evidence, because we've heard it all before.

His post on "expected value" pretty plainly shows that he thinks that people don't buy lottery tickets (since their expected value is far less than what you pay for them). But guess what? They do.

No, I never claimed that people don't buy lottery tickets or anything remotely like that. The point is that using Range Voting is like having a better chance of winning more money, so a wise voter wants Range Voting. And unless you are as dumb as people who buy lottery tickets, you should too.

I'd like to go on, but it's late. I'll leave the discussion of what hypothetical voting system we're never going to adopt in the U.S. for some later time.

Ah yes, since it's unlikely that we'll radically improve the world, why even try, right? Hey, I'm just as convinced as the next guy that humanity is rapidly headed toward inevitable doom, but at least I'm trying to prevent it. Meanwhile you speak dismissively about a subject on which you have very little knowledge, making more than a few apparently false claims.

I would ask that you take more time to get to understand the arguments and the subject before being so dismissive. And maybe be a little more optimistic. There's no reason to think that with hard work, this isn't possible. Human history is chock-full of people doing things that others said were impossible. Why don't you spend some time helping to promote Range Voting, and see how successful you are, before you give up on it.

Regards, Clay

Clay -- All that wonderfulness to all concerned is STILL JUST YOUR OPINION.

On the contrary, I have cited a profusion of objective facts, such as social utility figures. So my opinion has played virtually no part in the vast majority of what I have said.

It does not prove anything for anyone else more worldly-wise than you.

That is simply false. But if you'd like to address a specific claim, I can address it.

And we will not be conned into an election system that seems to give more weight to extremists than to people who actually consider things.

ALL election systems are susceptible to strategic behavior. But Range Voting is the least susceptible. In fact it is almost exactly half as damaged by a switch from 100% honest to 100% strategic voting as IRV.

So your argument here is actually an argument for Range Voting, not against it.

Why am I not surprised that you are a Libertarian

Imagine someone who cares so much about democracy also caring about freedom. What are the odds! :)

you believe in grandiose schemes that rely on fairy tales on human nature, and which in reality empower the most greedy, most ruthless, most deceitful members of what was formerly a community

On the contrary, I believe that government, with its ability to take our money by force and use it to blow the arms off of children in Iraq, is an empowerment of the most greedy, most ruthless, and most deceitful people - e.g. Dick Cheney, George W. Bush, etc. I believe it is government organizations such as the FDA which interfere with private economic transactions, to prolong the release of and inflate the cost of life saving drugs, that thousands of people die waiting for. I believe it is economic interference such as tariffs that benefit the most corrupt business owners, who use their weight in government to apply discriminating economic policies against their competitors. I believe it is government interference in the free market that causes inflation, recessions, and Great Depressions. I believe that it is Jack Abramhoff and Alberto Gonzales who cause us the greatest harm possible purely because government exists. Without it, they would be subject to the demands of the freemarket, and subject to immediate "firing" at any time. With it, they are able to go on far too long before being caught, and we only know their names because they are the unlucky few who did get caught. Government is their sword and shield and armor. Take it away and they are mere mortals.

But this is really not the appropriate avenue for this discussion, and I would invite you to contact me personally if you would like to discuss this issue further.

Human beings don't work on logic, never have, never will. So-called logic is generally a distraction from the reality of actual human systems.

Well, some do not, hence the prevalence of social diseases such as creationism in my native Kansas. But there are many exceptions to this rule. There are rational logical people out there - such as Richard Dawkins and Stephen Hawking, and the deceased Carl Sagan, and many many others. And it is the world's best and brightest to whom we should turn for guidance, moreso than to politicians, generally speaking. I believe it really is a life-or-death issues for society as we know it to start thinking of solutions to problems like peak oil and global warming. But without a good decision-making process, we are...well, I won't use the word. So we need Range Voting. If we're going to have to deal with democracy instead of liberty, it should at least be as utilitarian as possible. We need to decrease the importance of cash, and cast aside this horrid excuse for an election system that we call "plurality voting". That's where it all begins.

Range Voting also increases the number of spoiled ballots (by a factor of 7, here in San Francisco) where as Range Voting experimentally reduces them.

The first "Range Voting" should have been IRV, if that wasn't obvious enough.

I'm going to sign off now, but I want to thank you Kari for performing this poll, and all of you who have participated in the discussion. For my part I have spent far too much time trying to dispell many of the myths that were stubling blocks for myself as well, when I was new to Range Voting. I hope most of you have found it informative and interesting. I some of you will become believers, and join the movement.

Regards, Clay Shentrup San Francisco, CA 415.240.1973

Well, Clay, you've made a lot of bold statements here.

This is not range voting

Sure it is. You rate the faces, and attractive people win the distinction of being acknowledged as attractive. They only thing is, there's no significant incentive for strategic voting. But since Range Voting still operates exceptionally under extreme strategic voting, that is of little concern.

I think Clay's confused about why it's not a good electoral system.

I'm not confused. Range Voting is the best single-winner voting method known of, and I have cited the social utility figures which prove that.

Clay, we had cyclical boom and bust economies in this country until we had a progressive tax code and the implementation of regulated core industries

And most any libertarian would argue that the most severe booms and busts happen because of government intervention in the market, mainly via the Federal Reserve. And we would also tell you that the ends do not justify the means. So even if economic regulations make the world better in your opinion, that doesn't give you the right to steal from others. You don't get to tell me how much I have to pay my employees any more than I get tot tell you how to dress, or whom to marry.

While libertarianism is great in the vacuum of the abstract, it fails utterly to address the real issues and dynamics of the economy.

And we say exactly the opposite, and I think economists like Friedman and von Mises make a much better argument for our case than any pro-socialism economist. And again, even if you were right, and we CAN make the economy "better" via regulations, that is still wrong. Violating the rights of others is wrong. It's one of the first things you learn as a child. You don't steal, you don't hit.

Your fantasies about how wonderful a "pure" free market would be is just that, fantasy. Raw laissez faire economic policy leads directly to systemic crushing poverty to most of the population, always has.

And if you do not like that, you are free to donate your money to help the impoverished. But NOT to steal from others, effectively forcing them to live in the way wish.

That said, I do appreciate most libertarian views when it comes to civil liberties, but on economics, no thanks.

Yup, it's the ultimate irony. Liberals get downright fuming when they are told they cannot say what they want, or dress how they want, or marry whom they want. But when it comes to stealing from people, or disregarding their right to hire whomever they want for whatever reason, they are perfectly fine with it. Freedom is okay, as long as it's the freedoms you say we can have. It's sort of the mirror image of Republicanism - equal but opposite disregard for human rights.

Don't wish to further veer OT, but do appreciate you talking and arguing about the potentials of range-voting.

Thank you.

Clay, I will not further derail this thread, but progressive tax codes are not "stealing"

"Progressive" is a clever euphemism for "increasing". But taxes are indeed stealing - they are enforced, not optional. I like businesses, like restaurants, where my human rights are respected, and if I want the pancakes, I shell out my hard-earned cash for some pancakes, and if I don't want to, I don't have to. If the restaurant were to hand me some pancakes whether I wanted them or not, and then hold me hostage until I paid for them, that would be stealing - an that is exactly what taxes are, no matter how much you deny it. They are taken by force.

you seem to not grasp the basics of market history or principles

I spend hours reading The Economist and LewRockwell.com each month. I think I know quite a bit about economics and history actually. But having lived in San Francisco, Seattle, and Lawrence (Kansas) the past few years (all bastions of fiscal anti-freedomism), I have noticed an overwhelming trend among socialist-leaning people. Thy are generally highly ignorant of economics. They are rarely able to answer basic questions like "What is a fiat currency" or "What is stagflation?" or even "What does the Federal Reserve do?" Or, "Who is the current chairman of the board of directors of the Federal Reserve?" Even more hilarious, try asking a socialist what a Federal Reserve Note is. Give him a hint later and tell him he might find some in his wallet.

Socialists by and large strike me as being incapable of learning from history, or understanding basic mathematical principles at the foundation of economics. Consider the minimum wage, for example. How many of these San Francisco hipsters I live amongst understand that increasing the minimum wage leads to higher unemployment among young and unskilled workers? And how to people no comprehend that if we leave the market alone, the worker and employer achieve a more utilitarian result, because they alone negotiate the pay? How can a rational person suggest interfering with a decision that two people make with each other - it is sheer lunacy. But it happens because socialists do not think about logic, and long-term effects. They think about emotional things like, "Well, that poor worker needs more money, and the businessman has so much that, even if he earned it fair and square, we are just going to disregard his right to do with it as he pleases, and force him to give more of it to that worker, even if the worker had agreed to work for less."

I can't say that socialists have evil intentions, but nevertheless, their actions are as pernicious and dangerous as those perpetrated by the likes of lunatics such as George W. Bush, over in that desert Vietnam we have going on.

Now by comparison, if you ask most libertarians about economics, they seem to know a lot, and have some opinions that they've spent a lot of time mulling over in their heads, because libertarians tend to be highly analytical, and are often computer scientists and engineers, and the like. Not that this proves anything by itself, but it is a noteworthy observation.

And the point I always try to come back to when socialists want to talk about their adamant faith that socialism will lower poverty and other problems, is this. If I steal 10,000 dollars from a wealthy person to help a poor person in need, that is STILL A CRIME. That is the basic underlying truth that socialists just can't seem to grasp. For them, if they want an outcome, but do not have the means to execute it, they are okay with forcing others to do their work for them (or just hand over the cash to do it). That is why I wear a shirt that says "Taxation Equals Slavery". And I mean it.

I suggest some educational material for socialists.

http://lewrockwell.com/paul/paul364.html http://lewrockwell.com/orig6/sanandaji1.html

and thus makes me wonder as to the validity of your arguing the minutia of statistical validity of range-voting (which I admittedly am not well versed in).

This is a fallacy on multiple levels. The biggest problem is that you suggest my position is based on a lack of understanding of economics. Now of course, I would bet that I am far more learned on the subject of economics than 99.9% of people in your camp. But that is mere speculation, and is irrelevant. It is irrelevant because, I did not CHOOSE my position on the basis of "what will lead to the least poverty homelessness" or some such criterion. I do not believe in committing crimes against others in order to further causes important to me. So even if you could show that socialism lead to a wealthier society with less poverty, that would not change my position.

But in any case, you need to research social utility efficiency for yourself, and not take my word for it. It is troubling that this is such a major obstacle for so many who do not appreciate it because they prematurely dismisses it before having the faintest understanding about it. I hope that by analyzing it for yourself, and maybe even running some trials after compiling it on your machine, you will understand why it is the be-all-end-all measure of "good voting method".

I would love for you to post about some of these topics as a guest article (not that I have any control or sway over what gets published here at BlueOregon) to discuss and argue some of the wide-ranging topics you have tried to inject here, but don't want to continue to veer decidedly OT.

I agree with you about the second part, but not the first part. I don't think there's any changing liberals. They have to want to change. They have to inherently find their guilty conscience for stealing the freedoms of others, and find the irony in their want of certain freedoms for themselves in light of that. I think back to the teenage years I spent as an ardent socialist, and the hackneyed arguments I used to bolster my opinion, having required such incredibly expensive surgery that I dropped out of school to spend years working to prepare and pay for. But I think back now and I realize what it was. I was desperate. And when you get desperate, your morals get loose. Suddenly breaking into a house to steal money to feed your young child seems tempting. But it's still a crime. And that's what liberals just can't get through their heads.

You can have the last word if you like.

It's not about "last words". It's about human freedom. You are free to spend as much time discussing this with me as you like. My number is 415.240.1973.

Regards, Clay

Yeah, Clay, this is the part where you sign off and let the conversation happen. You're using up too much oxygen.

Fair enough.

My last request would be to see the full ballot information, to study it. I think we could all benefit from seeing how different voters used their ranges, and how many appeared to vote tactically instead of honestly, etc.

I sure do, if you're going to profit from a business that uses the streets, utility hookups, fire, police, military, Social Security, Medicare---and all the other things the rest of us pay for that make running a business possible.

That's the same thing as a restaurant telling me, "Look, I just gave you dinner - which is a benefit to you whether you wanted it or not, so pay up." I'm happy to pay my own private fire, security, power, retirement, insurance, etc. I would be happy to opt out of Medicare and not have to pay for it. Therefore by offering me those things, a government has no right to tax me for them. I did not ASK for them. I do not WANT them.

What's even more hilarious is that the idea of a central bank such as a Federal Reserve, was brought to the US by...socialists. Eye-ro-nee!

My point exactly. So it sounds like we agree on at least one thing. Now how do we abolish it? It's a behemoth.

Can Clay Shentrup say anything in fewer than 500 words?

I found the use of Range Voting for this poll very interesting and intriguing. It took a little more time and effort than simply "voting for only one" (Plurality Voting) or even "voting for several" (Approval Voting) but was ultimately more satisfying since I could be more expressive in my vote, or indication of preferences. It made me think a little more about my choices, since I could now indicate both how MUCH better I thought some candidates were than others and even which candidates I thought were rough equivalents of each other.

Range Voting seems perfect for single-winner elections in which there are more than two candidates, and particularly when some of the candidates are similar, especially partisan primaries. Other voting methods, such as Instant Runoff Voting or other ranked voting methods, instead would eliminate the voting expressiveness advantages of Range Voting.

I voted a much higher score for one candidate than any of the others, based on recommendations by others. But I didn't vote anybody a 1 or a 10 (I gave 2 "no opinion" scores, based on my ignorance). Since I'm not a Democrat, my personal expected voter satisfaction is not that much different for most of the Democratic choices.

Since this is one of the first times I have used Range Voting in a political poll, I simply forgot to exaggerate my preferences (vote strategically) and expand my expressed range to 1-10, but will probably remember to do so in the future, expecially if I have strong preferences in the race in question.

A big THANK YOU to BlueOregon.com for keeping an open mind and being willing to experiment with Range Voting.

The use of range voting for public elections would be a disaster. I imagine electioneers handing out cards telling people to give 10s to their top choices and zeroes to everyone else.

The built-in incentive to vote strategically is just as pronounced in range as in approval voting. The difference is: approval voting, under strategic conditions, simply degenerates into winner-take-all. With range, strategic conditions fundamentally threaten majority rule. A few more tens for Gravel and people would be saying very different things about the result here.

The use of range voting for public elections would be a disaster. I imagine electioneers handing out cards telling people to give 10s to their top choices and zeroes to everyone else.

1) A huge number of people will vote "honestly", regardless - this is supported by empirical evidence.

2) A huge number of people will vote strategically, whereas voting for just one is extremely unstrategic.

3) There's no reason to think a person why honestly preferred a 3rd party candidate (e.g. Nader) wouldn't vote for him in addition to his favorite front-runner (e.g. Gore).

4) This would be no worse than what we already do.

5) This is essentially what happens under alternatives like IRV already - which is why it leads to two-party duopoly just like plurality voting.

6) Even if a lot of voters did this, we can expect that in successive elections they would wise up and stop, seeing how incredibly strategically stupid it is.

The built-in incentive to vote strategically is just as pronounced in range as in approval voting.

Actually this is false. There is a much smaller payoff for voting strategically with Approval Voting.

But either way, both of these methods are substantially more resistant to strategic voting than poor methods like IRV and plurality.

The difference is: approval voting, under strategic conditions, simply degenerates into winner-take-all.

This is a very confusing statement. Range Voting and Approval Voting are identical if 100% of voters are strategic. I am used to hearing the term "winner-take-all" used to describe single-winner elections. One winner "takes it all". Range, Approval, plurality, and IRV are all single-winner election methods.

With range, strategic conditions fundamentally threaten majority rule.

1) Majority rule is a well-refuted concept, since Condorcet cycles disprove the "majority axiom". So threatening majority rule (in favor of highest social utility) is good.

2) Under strategic conditions, Range/Approval voting is arguably more likely to yield the Condorcet winner than real Condorcet methods. So your claim is quite off the mark.

A few more tens for Gravel and people would be saying very different things about the result here.

No, not "very different things". If the scores are close enough that "a few tens" would make the difference, then society clearly doesn't have a very big preference between the front-runners.

And ultimately, Range Voting has a higher social utility efficiency than other potential methods, so that's the bottom line.

Other voting methods, such as Instant Runoff Voting or other ranked voting methods, instead would eliminate the voting expressiveness advantages of Range Voting.

It is important, however, to distance ourselves from the arbitrary nature of such criteria, and focus instead on utility efficiency.

"Can Clay Shentrup say anything in fewer than 500 words?"

The answer would appear to be no.

I spoke to soon, the answer would appear to be "sometimes." Thank goodness.

But I didn't vote anybody a 1 or a 10

Ideally you should always give your favorite candidate the max score, and your least favorite a min, otherwise the scale is rather arbitrary.

Although I actually have a theory (which I hope we will eventually test) that if we use such an inherently-logarithmically-scaled score vector (where a "10" has meaning in that score approaches 10 as honest utility value approaches infinity), we actually get an even better social utility efficiency.

I spoke to soon, the answer would appear to be "sometimes." Thank goodness.

If you can find segments of my comments that I could have excised without reducing the completeness of my responses, I invite you to donate your time as my editor. Otherwise, you are free not to read them. That's a nice free market solution from this libertarian.

Excised and completeness are mutally exclusive, so you can start with that dilemma. Better, you were asked to give some air to this thread, which you agreed to do. Is commitment not a libertarian principle ;)

You did ask for them, Clay. You're part of a democracy which called for them, and part of a country which has as one of its founding principles to provide for the common defense, promote the general welfare and secure the blessings of liberty.

No I am not. I never consented to participate in giving money to whatever a democratic system chose, and I never signed anything saying I would acknowledge the U.S. Constitution (or any government).

"Every man for himself" is conspicuously absent from the text of the Constitution.

The Constitution is an irrelevant piece of paper. Laws once said slavery was acceptable - but it was still a crime. And taxation is just a less severe form of slavery.

A few more tens for Gravel and people would be saying very different things about the result here.

I regret not having fully realized the wrongness of this statement the first time. Let's take a closer look.

Gravel had 174 votes, with an average of 3.47, for a total score of 604 points. Gore had 3,753! To even pass the quorum rule, Gravel would need another 1,273 points! That's 127 perfect 10's. Sorry, that's not "close" by any means. Also, Dodd would not have been eligible (even with a better average) because he did not meet the quorum.

Now as for how much differently things would have had to be in order for the winner to change? Well, his closest competitor was Edwards, who would have needed a total of 253 additional points out of the 500 voters who scored them. I do not call this close.

Now had Gore not run, and Edwards would have won it, Obama would have needed an additional 25 points to win. But this would simply mean that society is almost equally happy with either candidate.

And what's more, Range Voting makes ties/near-ties less likely than they are with plurality voting and IRV (which makes them even more likely than with plurality).

You're really not addressing the issue here torridjoe. A Russian could have said something similar about communism, "It's just the way it is." But that doesn't make it right.

This comment by Shentrup speaks directly to the raison d'etre of his range voting "movement."

5) This is essentially what happens under alternatives like IRV already - which is why it leads to two-party duopoly just like plurality voting.

Behind all the Bayesian regret and social utility jargon is a desire to elect winners who have not won the most votes. Range voting lets that happen by changing the rules of the game. You no longer need a majority or plurality to win. You just need a few folks willing to give you a ten and everyone else a zero.

This is an intuitively undemocratic way to empower minor parties. Imagine the public outcry the first time this happens in a mayoral or gubernatorial race.

If you want to empower minor parties, you need a voting system that awards them representation in proportion to their share of the vote. You need proportional voting in multi-member districts.

Behind all the Bayesian regret and social utility jargon is a desire to elect winners who have not won the most votes.

Of course, because that's a terribly bad way to hold on election.

Range voting lets that happen by changing the rules of the game.

Exactly! The current rules are a travesty.

You no longer need a majority or plurality to win.

And there's no reason you should. I have already discussed this extensively here.

I'm not sure you're aware of this, or even what it means, but Range Voting picks the Condorcet winner more often than plurality voting. That makes your point quite ironic.

You just need a few folks willing to give you a ten and everyone else a zero.

1) This is what EVERYONE does with plurality voting already.

2) The negative effect of strategy is far worse with plurality voting and IRV.

3) There's no reason to think that one candidate's voters would be disproportionately full of selflessly strategic (willing to hurt their own expected utility to bullet vote for a candidate, altruistically) voters as compared with another candidate.

4) If the average of candidate B is so close to candidate A's average that "a few" bullet votes could make the difference, society clearly isn't much worse off with candidate B anyway.

5) Just look at the figures in this tiny election. It would take more than "a few" Edwards voters to zero Gore out before he'd lose - and a lot of them would never want to do that, because what if that made Clinton beat Gore, for instance.

I discussed this above as well. I encourage you to read the numerous cases where I've already addressed these issues, before forming strong opinions.

This is an intuitively undemocratic way to empower minor parties.

On the contrary, Range Voting is by far the best voting method for minorities. Maybe that's why the Libertarian Reform Caucus uses it and endorses it.

Imagine the public outcry the first time this happens in a mayoral or gubernatorial race.

Look at these ordered preferences

34% Clinton > Edwards > Obama 33% Edwards > Obama > Clinton 33% Obama > Edwards > Clinton

In a plurality election where only your favorite candidate gets a vote, Clinton wins, despite the fact that 66% of the voters would prefer either of her rivals. Time to rethink your beliefs about our absurdly terrible voting method.

If you want to empower minor parties, you need a voting system that awards them representation in proportion to their share of the vote. You need proportional voting in multi-member districts.

We're not talking about multi-winner elections, we're talking about single-winner elections, such as mayor, governor, etc.

If you want "proportional representation", we have Proportional ("Reweighted") Range Voting, and Asset Voting - both of which are far and away superior to methods like STV.

http://RangeVoting.org/RRV.html http://RangeVoting.org/Asset.html