Slashdot Mirror

Condorcet voting can avoid Arrow's paradox too, and is superior to Approval in many ways.

The EC does not need a complete overhaul, unless you can come up with a better system to represent the notion that the US are a federal union of sovereign states (as the two house Congress does) for a singular office. It would probably be good if states awarded EC votes proportionally or by district, but the EC institution itself is pretty sound. My solution would be to award EC votes by district and use the two at-large votes to adjust those results toward the proportional result. This allows people to organize geographically (which is obviously an efficient way to do so) to win their neighbors without completely disenfranchising the losers state-wide.

Condorcet voting can avoid Arrow's paradox too, and is superior to Approval in many ways.

The EC does not need a complete overhaul, unless you can come up with a better system to represent the notion that the US are a federal union of sovereign states (as the two house Congress does) for a singular office. It would probably be good if states awarded EC votes proportionally or by district, but the EC institution itself is pretty sound. My solution would be to award EC votes by district and use the two at-large votes to adjust those results toward the proportional result. This allows people to organize geographically (which is obviously an efficient way to do so) to win their neighbors without completely disenfranchising the losers state-wide.

Oh please for the love of Pete, NO! I've said this many times on \. already too, but this is LJ post is the only recent one I can find. IRV is a provably flawed system, please stop advocating it! Pushing for voting reform is great, but we need Condorcet voting, not IRV.

And BTW, we need to keep the EC.

Instant runoff voting is no good.

The link is currently down so I'll try to sum up briefly:

1) It 'works' fine if the 3rd party candidate has no chance of beating a candidate from the other two parties, but then who cares?

2) Let's say a third party does arise, and you end up with a situation like this (the letters below signify political parties and are of course random):

33%: G, D, R ("Behold the G revolution!")
14%: D, G, R ("Anyone but R, but D before G!")
18%: D, R, G ("We sober judges prefer R to G")
35%: R, D, G ("Pluralities should win")

IRV decides D is bumped (with 32%), then G (with 47%): so party R wins by 6% with 53%. What a clear mandate. But hark! Some small sect of the G electorate really despises R and decides that they can shift their vote to "D, G, R". So now it is:

31%: G, D, R ("IRV would never betray us!")
16%: D, G, R ("We feel so bolstered/dirty!")
18%: D, R, G ("We're still more important than DGR, right?")
35%: R, D, G ("No no, vote your conscience")

Now, G is eliminated first (with 31%), and then R (with only 35%) leaving D as the winner with 65% of the electorate, a 30-point victory! A mega-landslide!

Do you see the problem? Putting aside how 2% of the population engineered a 36 point swing in the tally, the bottom line is that even with IRV you still should worry about electability and how to vote strategically if you really want your preferences respected.

In this case the D and R parties are aligned, and most of the voters want one of those two candidates to win.

What makes you assume the parties are aligned? Just because the majority votes for one of those two? That's because we are effectively in a two party duopoly; there's no other practical choices. They are most certainly not aligned.

But I digress.

The purpose of a voting system is to determine the majority preference. If the majority prefer X but the system elects Y then the system is broken. The politics of the parties is irrelevant. IRV is one such broken system.

All you're saying is that the parties can try and adapt to the broken voting system by modifying thier politics. That doesn't change the fact that the system is broken. An effective democracy needs an accurate election process. IRV isn't it. It produces erratic and unpredictible results. You really should read this page for an analysis of different methods. Out of all the different methods, IRV comes last. There's no real point in me reproducing that content of that site here.

I've heard that IRV is not the best idea. Condorcet Voting is actually better - or Approval voting. See http://www.electionmethods.org

Reading the article, I was appalled that IRV failed the monotonicity criteria -- that under IRV it is possible that increasing preference for a candidate can cause him/her to lose, and decreasing preference for a candidate can cause her/him to win.

This is nonsense. It means that it's impossible to understand the true implications of your vote, and expressing your true preferences can produce a contrary outcome.

Just for that, IRV is grotesque and an unacceptable voting system.

So I found a brief discussion of arrow's theorem on the website .

It's good to see that they considered this important theorem but given my understanding of arrow's theorem suggests they give it far too short shift. In particular they seem to be under the impression that it only causes problems because the introduction of another canidate might effect the election. This might be mathematically equivalent but it sweeps a very important point under the rug.

Another implication of arrow's theorem is that voters have an incentive to misrepresent their preferences. I believe this is equivalent to the irrelevant alternatives statement mathematically. Imagine an election where the introduction of an extra party would shift the election supposing everyone voted honestly, now this can only happen if that party is part of the smith set (i.e. beats at least one other party). This means that certain people would have incentive not to vote honestly so as to guarantee this party either does or doesn't make it into the smith set.

In short I don't think we can so blithely dismiss the consequences of arrow's theorem as, "well yah sometimes adding a new party affects the election" The result of this is that concordant voting, just like IRV also has situations which encourage insincer voting.

This means that the choice of voting system must be made by more practical pragmatic concerns. For instance what types of elections are common, simplicity for the average voter and the political system. Quite frankly the US is a two party system and this system has many advantages. For a two party system a IRV is a very good system.

The question of getting rid of the two party system is another matter entierly, but most of our governmental institutions are designed to work in a two party system and changing this requires much more than a change in voting system.

Frankly, I don't think a truly multi-party system is a good idea given our constitution. If we truly wanted three or more serious parties I think we would be compelled to move to a parlimentary system. Also the constitution was designed to make things difficult to do and it is only the oil of the party machinery which can get things done in congress. I think there is a real danger that any serious third party would simply bring congress to a grinding halt.

What's wrong with "If one candidate beats all the others in a head-to-head, that candidate wins"? I'm pretty sure most people would understand that idea.

The difficult part, IMHO, would be convincing a politically-motivated media to run their fact files and commentary on why such-and-such a method would meet the goal(s) of a fair election, so enough people actually understood what was going on that the general population would accept it (the two not being the same thing at all).

The problem with Condorcet, for example, is going to be explaining how they break a tie -- not really an issue if you're used to a duopoly, but rather important here! What you need to make it successful is a system that is sound, which will stand up to critical examination from the few, but that can be summed up in a nutshell so the many understand how to vote. Fortunately, most systems meet the latter criterion: you say "list the candidates in order of preference", "vote for the guy you want", "tick all the guys you'd be happy with", etc.

BTW, if anyone hasn't looked, the linked site (electionmethods.org) is very well done. As a mathematician and someone who cares about elections, I found the page on technical evaluations of the various methods most interesting. The kind of criteria it presents for a good system -- the one-liner sound-bites -- are the sort of thing that should be fed to the general public, with the accompanying reasoning available for those critical enough to examine the details.

"Condorcet doesn't require that you rank candidates differently."

"You're right that Condorcet favours moderate, inclusive candidates. That's a feature. It's inbuilt protection against the absurdities of swing politics."

"You don't know what "disenfranchise" means.

"It's a cure to the self-perpetuating two-party duopoly."

It is an erratic voting system because ranking a candidate higher can actually cause the candidate to lose, and ranking a candidate lower can cause the candidate to win. As if that weren't bad enough, it can also fail to elect a candidate who is preferred over each of the other candidates by a majority of the voters.

Are there web resources that clearly explain the technical details of how to arrive at this conclusion? While I suspect you are correct, it would be nice to have a URL that I could point people toward.

The arguments against the Electoral College simply dismiss the College out of hand, without staring the best reason for having them in the first place. Our government is a federation of sovereign states (e.g., "Federal" government and "United" States). We are not a monolithic governmental body. The purpose of the Electoral College is to intentionally skew the numerical advantage of smaller states to make them more equal in power to the larger states when the states are acting as equals, such as in the Senate and Presidential elections. This is a negotiated settlement of state vs. state power that is fair and balanced and has stood the test of time.

Properly speaking, our President is elected by the states, not by individual voters. (In fact, in the beginning, the President was directly elected by the state legislatures, and there was no direct, popular vote.) There's nothing "un-democratic" about the Electoral College. It's just the states' way of voting.

In my opinion, any proposed change in a voting mechanism must address the need for state vs. state balance of power, or it simply won't fly. The reason the Electoral College is in the Consitution has to do with the way our Union is is organized, not with some supposed desire to "keep women and minorities down" as electionmethods.org would argue (*sigh*).

It may be that changing the voting mechanism could help states select Electors better, especially in a tight race with more than two close contenders. But in the end, it will always be very much to each State's advantage to award the Electors as winner-take-all, because this maximizes their leverage against the other states in the Union.

In fact, without the Electoral College, the effect of winner-take-all would be even more pronounced, only, it would be the winners of just a handful of states.

There seems to be a reply to this guy at http://www.electionmethods.org/IRVing.htm. Just a heads-up for people following this thread.

Daniel

What each of the reforms discussed (Borda, Condorcet, IRV) has in common is that it incorporates voter preference between all the candidates rather that just their first choice. In that sense, each method is better than our current one. They differ in how the voter's preferences are aggregated.

Each of these voting systems has "problem scenarios" in which an unexpected or undesired outcome is produced-- that is, one that runs counter to intuition in some way. The observation that all reasonable voting systems allow for problem scenarios is crystalized in Arrow's impossibility theorem.

One of the usual responses to the impossibility theorem has been to develop various desirable criteria other than those required in Arrow's theorem, and see which voting method satisfies the most criteria. You'll see an analysis like this at Election Methods that argues for the Condorcet method.

An alternative analysis has been developed in recent years by mathematicians that identifies the voting method with the least number of problem scenarios. We've known for many years that every voting method has problem scenarios but very few people have asked how many problem scenarios there are with each voting method. It turns out that this question is rather hard and requires some sophisticated math to answer.

Say we have 5 candidates and 1000 voters. On each of the 1000 ballots, the candidates can be ranked in 5!=120 different ways. To each possible "1000-ballot-profile," a voting method assigns a ranking of the candidates-- that is, a winner, and the runners-up. For each method, we want to find how many of the 1000-ballot-profiles result in unexpected outcomes. The upshot is that the analysis shows that the Borda count has the least number of problem scenarios and is robust in another sense. (The Borda count is when the candidates ranked first through fifth by a voter get 5 through 1 points respectively. It should be familiar to many people because it is often used in athletic contexts.)

I recommend looking at Chaotic Elections! A Mathematician Looks at Voting for an expository account and Basic Geometry of Voting for the details.

In case you didn't explore the site fully, this page explains their arguments against IRV. Personally, I find them very compelling.

What you say here leads into their arguments:

Yes there are contrived conditions where you can show that some mathematically disproportionate fraction of the populace would be "happier" with a different candidate, but look at the reality of voting in the US. 90-99% of the voters split their votes relatively evenly between the two major parties. The rest split them fairly unevenly between the remaining minor contenders.

As long as the minor parties are quite minor, IRV will just provide more interesting protest votes. And there is probably some value in that, but it isn't enough. As soon as a party or candidate becomes big enough to challenge the main two, the spoiler effect comes right back into play.

But please read the above linked page for a much clearer and more thorough explanation.

Personally I like it better than concordant voting because it reflects my feelings well - I may not be able to accurately rank how I feel about each candidate (so many issues were one's better than others) - but I sure know if I'd be OK with each candidate very quickly.

The demand of election reform over our current plurality voting method in order for the non-trivial roles of third parties is epitomized by Duverger's Law. However, IRV isn't necessarily the only or best solution. Both Approval voting and Condorcet voting are generally considered more accurate. This is a good explanation as to IRV's shortcomings, as well as an insight into other methodologies.

Almost any of the alternative voting systems is vastly superior to plurality voting. Among them Approval voting is the simplest and easiest to understand and implement.

It's true that Condorcet has some advantages over Approval but these are mostly theoretical and are greatly offset by its complexity.

Almost any of the alternative voting systems is vastly superior to plurality voting. Among them Approval voting is the simplest and easiest to understand and implement.

It's true that Condorcet has some advantages over Approval but these are mostly theoretical and are greatly offset by its complexity.

There's a fairly good rebuttal of this on the electionmethods.org website.

Rob

1. Bush
2. Badnarik

I'm not sure how a strongly supported candidate can be marginalized. If the candidate is marginalized, then he has lost support. Can you give an example of a strongly supported candidate that loses to a less supported candidate using the condorcet methond, but not in another voting system (Approval, IRV)? Picking random votes doesn't count as an example because randomly choosing votes can't distinguish between the most popular candidates and the least popular candidates.

1. Bush
2. Badnarik

I'm not sure how a strongly supported candidate can be marginalized. If the candidate is marginalized, then he has lost support. Can you give an example of a strongly supported candidate that loses to a less supported candidate using the condorcet methond, but not in another voting system (Approval, IRV)? Picking random votes doesn't count as an example because randomly choosing votes can't distinguish between the most popular candidates and the least popular candidates.

I think the biggest problem with Condorcet is explaining it to Joe and Jane six-pack.

Approval Voting is beautifully simple, is easy to count by hand and I'm told that it requires little or no change to existing voting machines.

This makes a lot of sense.

Moreover, IRV does NOT eliminate the spoiler problem! E.g., imagine we use IRV the Greens get lots of additional support. Suppose there are 3 major voting blocks in an election and they rank their candidates as follows:
40% rank 1. Republican, 2. Democrat, 3. Green
25% rank 1. Democrat, 2. Republican, 3. Green
35% rank 1. Green, 2. Democrat, 3. Republican

Who should win? Imagine if you asked, "If the election were only between Repub. and Dem, who would you want to win?" 60% would say "D" and 40% R. Now ask "G vs. D." 35% say G, and 65% say D. "R vs G": 65% say R, 35% say G. So the Dems win all the pairs they are in. They should win the election. (Technically, they are called the Condorcet winner and any election system that always selects the Condorcet winner (so long as once exists) is called a Condorcet system.) The Repubs win only if we take the Dems out of the race.

But that is exactly what IRV does. The Dems are eliminated because they have the fewest number of 1st place choices, and their votes are transferred to the Repubs, who then win. The fact the Green voters preferred the Dems over the Repubs never entered into the calculation!

If there were another "third" party, you could even construct a more likely situation where the Dems lost even though they had MORE 1st place votes than the Greens.

The more parties there are, and the more support that third parties have in an election, the more random the results that IRV gives. Essentially, as long as third parties remain nice and small, then it eliminates the "spoiler" threat of the current system. But if they ever gain significant support, all hell breaks loose. You just can't tell who will win, even if by some reasonable means, like Condorcet, there is someone who SHOULD win.

It's a REALLY, REALLY bad system. There are some ways that it is even worse than the current one. (As the link in the parent pointed out, you can also have a case where a set of voters ranking a certain candidate LOWER, with no other changes, causes that candidate to WIN instead of lose! This doesn't happen in the current system.)

The good thing about IRV is that once people get used to ranking their preferences, you can change to a better system just by changing the backend. The actual process of voting doesn't change. But I fear that if IRV got enacted in a bunch of places, we would soon have a paradox occur and the negative publicity would take the wind out of all kinds of electoral reform. I think the main reason that IRV is the most popular alternative method is because it's easy to explain.

Approval voting is also easy to explain, but I doubt if it will ever gain any traction since it violates the "one person-one vote" philosophy; it just seems, intuitively, unfair, even though it is much fairer than the current system.

People who advocate IRV should really check out http://electionmethods.org/ for some good discussion of its horrors. Wikipedia also some good info.

I'm not sure what you're worked up about

Do you understand non-monotonicity? The idea that a few people changing their vote from G to R (aka "push-overs") changes the winner from D to G? Does that eensy bit of insanity not bother you?

Among other flaws, IRV also isn't Summable: you can't count the ballots precinct-by-precinct and add it up hierarchically. You can only calculate the winner after you have ALL of the raw data (which is O(n!), BTW) in a single computer. Manual recounts would take MONTHS to collate.

I would be a Green, but the push for IRV is the wrong solution to the right problem. The problem of forming a government that expresses the will of the people throught voting is a hard one (see Florida, 2000). I don't have the answer, but IRV isn't it.

I usually feel very close to the Green Party. I like their stance on many issues (not least the environment!), but the American Green Party insistence on using IRV disappoints me a lot.

We do need a better voting method like Approval and Condorcet, but Instant Runoff Voting is not a good system.

Unfortunately, Cobb puts politics before policies, nice talk before issues. When told about the problems with IRV, he replied that good luck, go out and organize, but IRV is the only alternative voting method that has a base of support, so that what he's going to push.

That's why I ended up endorsing Libertarian candidate Badnarik instead, because Badnarik supports Approval (so he told Slashodt last week).

The need for a better voting algorithm is obvious, but Instant Run-off Voting (IRV) isn't it. IRV is a particular voting algorithm that produces some unpredicatable (to the voter) results. There are much better methods available, such as approval voting and the Condorcet method.

IRV is little more than a snappy name covering bad math. It makes a lousy poster-child for the movement to adopt an alternative voting method. How bad is the math on IRV? Under certain circumstances, you can benefit your candidate less by ranking him highest than if you had ranked him lower. That is not a result we want adopted. That's actually worse than the current situation where if you cast your single vote for your true favorite, the candidate you dislike most may win.

The need for a better voting algorithm is obvious, but Instant Run-off Voting (IRV) isn't it. IRV is a particular voting algorithm that produces some unpredicatable (to the voter) results. There are much better methods available, such as approval voting and the Condorcet method.

IRV is little more than a snappy name covering bad math. It makes a lousy poster-child for the movement to adopt an alternative voting method. How bad is the math on IRV? Under certain circumstances, you can benefit your candidate less by ranking him highest than if you had ranked him lower. That is not a result we want adopted. That's actually worse than the current situation where if you cast your single vote for your true favorite, the candidate you dislike most may win.

The need for a better voting algorithm is obvious, but Instant Run-off Voting (IRV) isn't it. IRV is a particular voting algorithm that produces some unpredicatable (to the voter) results. There are much better methods available, such as approval voting and the Condorcet method.

IRV is little more than a snappy name covering bad math. It makes a lousy poster-child for the movement to adopt an alternative voting method. How bad is the math on IRV? Under certain circumstances, you can benefit your candidate less by ranking him highest than if you had ranked him lower. That is not a result we want adopted. That's actually worse than the current situation where if you cast your single vote for your true favorite, the candidate you dislike most may win.

For Democrats, Greens are the party which champions what Democrats used to: support for working people and people of color and protection of the environment.

This quote smacks of party politics. No party should champian "people of color." Is black a color? Is white a color? How about yellow-brown? Or red-brown? Saying "I support blacks" is just as racist as saying "I hate blacks" simply by the nature that you are separating them into a group. Equality does not mean adjusting the scales to be even - it means getting rid of the scale entirely.

...we really need to address is the corruption in the White House and in Congress...

The democrats say that too. And the republicans. It's easy for a party who is outside the system to say that, but what is the plan for doing it?

We need to replace it with Instant Runoff Voting.

I said the same thing last week. Someone from Slashdot corrected me. IRV is worse than our current system - the problems are subtle to see but very significant. Here is why. I didn't believe it until I read it.

Question: ...Electoral votes from a state be split proportional to the popular vote... Response:I believe we should move rapidly towards Instant Runoff Voting, as outlined above, rather than tinker with an anachronistic relic.

This is a naive response. You can't just say "okay, let's replace all the state election systems and change all the state constitutions all at once, and forget the steps that get us there." This country's system is an anachronistic relic. Good call there. But you must tinker with it until you get what you want.

One common thread amongs the smaller party replies is that they are often ideologically good, but realistically bad. I heard a Green party spokesperson on NPR say that if the Green party wins, they will immediately withdraw all troops from Iraq and apologize. That's beautiful, but it would also plunge Iraq into civil war, cause the UN to hate us even more, and kill millions of Iraqis. Great in concept, but unrealistic. We need people who realize that politics is compromise, and that small steps are what move us forward.

I would like to see the process streamlined so that undocumented workers, who are here and are paying taxes and contributing to our society, can obtain citizenship more simply and easily

This is the most political of all the answers. "Undocumented workers" is a nice way to say criminals who illegally tresspassed, dodged or lied to border police, and/or forged identities to get here. They are criminals and should be sent home. There are people who wait patiently for work visas for years to get into the US. But since the illegal immigrants have gotten good enough at forging IDs to vote, they are now a constituent base and must be appealed to.

50% of my coworkers are immigrants, and I respect every one of them. I went for lunch today, and I was served by immigrants. I respect every one of these people from IT professionals to minimum-wage workers. But it is really scary when we decide that we need to give voting rights to people who shouldn't even be allowed to walk the streets. I really hope it is just some massive trick to have them all come out, admit it, and ship them home. What's the unemployment rate right now?

Just so everyone understand where I am coming from, I am an independent who voted for Greens, Libertarians, Democrats, and Republicans. I look at candidate's qualifications first, and the party has no bearing on my decision. I'm not anti-green, I merely question some of these responses. I do fear that some of these Green part

The best voting method I've seen is Condorcet voting. But even that isn't perfect.

Anyone interested in reading more about Condorcet voting should go to electionmethods.org.

Actually, IRV suffers from MANY problems. ElectionMethods.org has an excellent summary of many "multiple selection" voting methods. IRV is the worst of all of them, as it can end up selecting the candidate who does not have the most votes. Educate yourself about the dangers of IRV and the many superior alternatives!

Condorcet's method avoids Duverger's Law by providing a ranked voting system in which all preferences are simultaneously pairwise evaluated. It can be seen as a generalization of plurality voting and approval voting. However it is unlike Instant Runoff Voting, which also uses ranking and pairwise comparisons, because IRV evaluates preferences sequentially, discarding some as it goes. So though the vote casting method is the same between the two, the vote counting method is different.*

To get a conceptual grasp for what Condorcet tries to do, ask yourself what the definition of "election winner" is. Plurality voting says it is the guy with the most votes. This has a strategic problem whenever there are more than two candidates, however - are you really voting for candidate A, or against candidate B when you really prefer dark horse C? Since the system doesn't allow you to express your full preferences, there is a temptation to abandon your principles and put your vote where you think it will have the most effect - behind one of the two perceived front-runners. They're perceived that way because everyone else faces the same dilemma you do.

OTOH, Condorcet voting says the winner is the candidate who would win a clear majority of head-to-head elections against all the other candidates individually. If you're the best candidate, then this is something you ought to be able to do - right? When clear 1-2-3 preferences are given, it is easy to tally this. Time-consuming (you're not counting ballots but "wins" - so in an n-way race there will be (n^2 - n)/2 "wins" on every ballot) but conceptually easy.

Note this is related to how the media looks at 3rd parties as "spoilers" - they ask people "if Nader wasn't in the race, would you vote Bush or Gore?" Why don't they ask "if Gore weren't in the race, would you vote Bush or Nader?" Maybe Gore was a spoiler for Nader! If you can vote for who you really want (Nader, for example) while still making your preference of Gore over Bush clear, the spoiler problem is eliminated. With Condorcet, you can vote your conscience without sacrificing the result!

The Condorcet winner tends to be the concensus winner - he might not be the first choice of the largest voting bloc, but the average voter satisfaction will be higher. Take the 1912 presidential race. Together Roosevelt and Taft got 51% of the vote, and differed little. But Democrat Wilson won with only 42% because the other two split the conservative vote. I think it's fairly obvious that the 51% would have been happier had either Roosevelt or Taft won, though only about 25% would have listed either as their first choice.

I like to think of the Libertarians as the parallel today. Fiscally conservative and socially liberal (to generalize), they can be seen as "between" the Democrats and Republicans. Though they might not win many first-place rankings, I tend to believe that most Ds and Rs would prefer them over the "other guy". This large number of second-over-third-place wins could put them over the top. And conceptually this makes sense - you get a president that a clear majority can live with, instead of a polarizing factor in the Oval Office.

FWIW, approval voting comes somewhere between these models. You can give multiple candidates a "thumbs up" instead of just one like plurality, and the guy with the most votes still wins, but there is no differentiation between them like Condorcet. Thus there is still a temptation to vote strategically. If you approve of A and B but actually preferred A, you might not give an approval to B in order to sabotage his chances. Then we're right back to plurality voting and two major candidates.

*Sidebar: IRV is dangerous beca

Condorcet's method avoids Duverger's Law by providing a ranked voting system in which all preferences are simultaneously pairwise evaluated. It can be seen as a generalization of plurality voting and approval voting. However it is unlike Instant Runoff Voting, which also uses ranking and pairwise comparisons, because IRV evaluates preferences sequentially, discarding some as it goes. So though the vote casting method is the same between the two, the vote counting method is different.*

To get a conceptual grasp for what Condorcet tries to do, ask yourself what the definition of "election winner" is. Plurality voting says it is the guy with the most votes. This has a strategic problem whenever there are more than two candidates, however - are you really voting for candidate A, or against candidate B when you really prefer dark horse C? Since the system doesn't allow you to express your full preferences, there is a temptation to abandon your principles and put your vote where you think it will have the most effect - behind one of the two perceived front-runners. They're perceived that way because everyone else faces the same dilemma you do.

OTOH, Condorcet voting says the winner is the candidate who would win a clear majority of head-to-head elections against all the other candidates individually. If you're the best candidate, then this is something you ought to be able to do - right? When clear 1-2-3 preferences are given, it is easy to tally this. Time-consuming (you're not counting ballots but "wins" - so in an n-way race there will be (n^2 - n)/2 "wins" on every ballot) but conceptually easy.

Note this is related to how the media looks at 3rd parties as "spoilers" - they ask people "if Nader wasn't in the race, would you vote Bush or Gore?" Why don't they ask "if Gore weren't in the race, would you vote Bush or Nader?" Maybe Gore was a spoiler for Nader! If you can vote for who you really want (Nader, for example) while still making your preference of Gore over Bush clear, the spoiler problem is eliminated. With Condorcet, you can vote your conscience without sacrificing the result!

The Condorcet winner tends to be the concensus winner - he might not be the first choice of the largest voting bloc, but the average voter satisfaction will be higher. Take the 1912 presidential race. Together Roosevelt and Taft got 51% of the vote, and differed little. But Democrat Wilson won with only 42% because the other two split the conservative vote. I think it's fairly obvious that the 51% would have been happier had either Roosevelt or Taft won, though only about 25% would have listed either as their first choice.

I like to think of the Libertarians as the parallel today. Fiscally conservative and socially liberal (to generalize), they can be seen as "between" the Democrats and Republicans. Though they might not win many first-place rankings, I tend to believe that most Ds and Rs would prefer them over the "other guy". This large number of second-over-third-place wins could put them over the top. And conceptually this makes sense - you get a president that a clear majority can live with, instead of a polarizing factor in the Oval Office.

FWIW, approval voting comes somewhere between these models. You can give multiple candidates a "thumbs up" instead of just one like plurality, and the guy with the most votes still wins, but there is no differentiation between them like Condorcet. Thus there is still a temptation to vote strategically. If you approve of A and B but actually preferred A, you might not give an approval to B in order to sabotage his chances. Then we're right back to plurality voting and two major candidates.

*Sidebar: IRV is dangerous beca

Whatever your opinion after either set of debates, why not express your opinion like you never have before...

In the weeks leading up to the Presidential elections, many Slashdot users as well as some third party candidates (Badnarik) have advocated the use of better voting systems such as Instant Runoff Voting, Approval Voting and Condorcet Voting. There is a new web site which has just started a worldwide campaign to promote such alternatives, and it has just come up with an unexpected endorsement for the US 2004 election. What's much more interesting is a new service which is, as far as I am aware, unique on the web: a MOD of the popular software phpBB allows users not only to post messages like on any other Bulletin Board, but also to create their own polls, with their own choice of candidates, using a variety of voting methods. Users can thus create their own polls, cast their ballots and talk about the results all in the same forum.

American voters should use this unique opportunity to cast the alternative ballot that they will not be able to cast in November, and maybe also create new polls corresponding to other local elections in their states. It would be very interesting to see what the result would be if voters were given the opportunity to voice their true preferences.

Whatever your opinion after either set of debates, why not express your opinion like you never have before...

In the weeks leading up to the Presidential elections, many Slashdot users as well as some third party candidates (Badnarik) have advocated the use of better voting systems such as Instant Runoff Voting, Approval Voting and Condorcet Voting. There is a new web site which has just started a worldwide campaign to promote such alternatives, and it has just come up with an unexpected endorsement for the US 2004 election. What's much more interesting is a new service which is, as far as I am aware, unique on the web: a MOD of the popular software phpBB allows users not only to post messages like on any other Bulletin Board, but also to create their own polls, with their own choice of candidates, using a variety of voting methods. Users can thus create their own polls, cast their ballots and talk about the results all in the same forum.

American voters should use this unique opportunity to cast the alternative ballot that they will not be able to cast in November, and maybe also create new polls corresponding to other local elections in their states. It would be very interesting to see what the result would be if voters were given the opportunity to voice their true preferences.

Well, this article submission was too much (or not enough) pro-whomever for their taste:

In the weeks leading up to the Presidential elections, many Slashdot users as well as some third party candidates (Badnarik) have advocated the use of better voting systems such as Instant Runoff Voting, Approval Voting and Condorcet Voting. There is a new web site which has just started a worldwide campaign to promote such alternatives, and it has just come up with an unexpected endorsement for the US 2004 election. What's much more interesting is a new service which is, as far as I am aware, unique on the web: a MOD of the popular software phpBB allows users not only to post messages like on any other Bulletin Board, but also to create their own polls, with their own choice of candidates, using a variety of voting methods. Users can thus create their own polls, cast their ballots and talk about the results all in the same forum.

American voters should use this unique opportunity to cast the alternative ballot that they will not be able to cast in November, and maybe also create new polls corresponding to other local elections in their states. It would be very interesting to see what the result would be if voters were given the opportunity to voice their true preferences.

Well, this article submission was too much (or not enough) pro-whomever for their taste:

In the weeks leading up to the Presidential elections, many Slashdot users as well as some third party candidates (Badnarik) have advocated the use of better voting systems such as Instant Runoff Voting, Approval Voting and Condorcet Voting. There is a new web site which has just started a worldwide campaign to promote such alternatives, and it has just come up with an unexpected endorsement for the US 2004 election. What's much more interesting is a new service which is, as far as I am aware, unique on the web: a MOD of the popular software phpBB allows users not only to post messages like on any other Bulletin Board, but also to create their own polls, with their own choice of candidates, using a variety of voting methods. Users can thus create their own polls, cast their ballots and talk about the results all in the same forum.

American voters should use this unique opportunity to cast the alternative ballot that they will not be able to cast in November, and maybe also create new polls corresponding to other local elections in their states. It would be very interesting to see what the result would be if voters were given the opportunity to voice their true preferences.

They may be spoilers, but the American election process is so out of whack that third party candidates are simply not going to have any realistic chance at being elected. You're complaining about the effect rather than the cause -- once third-parties are viable, they will be included in the debates. Including them now would just be a waste of time.

Take a look at electionmethods.org, and in particular the page on Why the Electoral College Should be Abolished.

They may be spoilers, but the American election process is so out of whack that third party candidates are simply not going to have any realistic chance at being elected. You're complaining about the effect rather than the cause -- once third-parties are viable, they will be included in the debates. Including them now would just be a waste of time.

Take a look at electionmethods.org, and in particular the page on Why the Electoral College Should be Abolished.

The question is a good one, but the specific term "instant-runoff voting" should be replaced with "approval voting or other improved voting methods".

Instant Run-off Voting (IRV) is a particular voting algorithm that produces some unpredicatable (to the voter) results. There are much better methods available, such as approval voting and the Condorcet method.

IRV is little more than a snappy name covering bad math. It makes a lousy poster-child for the movement to adopt an alternative voting method. How bad is the math on IRV? Under certain circumstances, you can benefit your candidate less by ranking him highest than if you had ranked him lower. That is not a result we want adopted. I'm very interested to hear the candidates' responses to this question, but please revise it before we ask it.

The question is a good one, but the specific term "instant-runoff voting" should be replaced with "approval voting or other improved voting methods".

Instant Run-off Voting (IRV) is a particular voting algorithm that produces some unpredicatable (to the voter) results. There are much better methods available, such as approval voting and the Condorcet method.

IRV is little more than a snappy name covering bad math. It makes a lousy poster-child for the movement to adopt an alternative voting method. How bad is the math on IRV? Under certain circumstances, you can benefit your candidate less by ranking him highest than if you had ranked him lower. That is not a result we want adopted. I'm very interested to hear the candidates' responses to this question, but please revise it before we ask it.

The question is a good one, but the specific term "instant-runoff voting" should be replaced with "approval voting or other improved voting methods".

Instant Run-off Voting (IRV) is a particular voting algorithm that produces some unpredicatable (to the voter) results. There are much better methods available, such as approval voting and the Condorcet method.

IRV is little more than a snappy name covering bad math. It makes a lousy poster-child for the movement to adopt an alternative voting method. How bad is the math on IRV? Under certain circumstances, you can benefit your candidate less by ranking him highest than if you had ranked him lower. That is not a result we want adopted. I'm very interested to hear the candidates' responses to this question, but please revise it before we ask it.

IRV is certainly not the best solution.

For other alternatives such as Approval and Condorcet, look here.

Also, you don't need to wait for Kerry/Bush to get elected to cast a Condorcet or Approval ballot!.

OK, did you nap through that whole USA PATRIOT thing? The one with the FBI going into libraries checking up on peoples' reading habits? The one where we've got American citizens detained without benefit of counsel or due process?

If you don't think Bush is more conservative than Kerry, you're a loony. For the record: I think Bush and Kerry are both loonies.

Yes: I was using a hyperbole. Again with the tongue in the cheek.

When I said "I", I meant "I as a member of We the People". It was a rhetorical thing...musta gone a little fast for you. My bad. I would like my voting preferances to be tallied in a more-accurate-than-binary way.

I dislike the Republicans. I dislike the Democrats. I

It certainly wouldn't take 100 run off elections. It would simply take switching to a tallying method like Approval or Condorcet voting. Learn.

How is it fair that the vote of somebody in Wyoming worth three to four times the vote of somebody in California?

Shouldn't the President represent the largest possible number of Americans?

Of course, getting rid of the electoral college is only the teeniest step. We really need new ways of scoring elections.

How about the now century-plus old campaign to convince Americans there are only two choices?

Easy. There are only two choices. The system is locked. You can only express one preference, so there's no point in voting for someone if you're pretty sure they're not going to get elected as long as you have a preference between two (or more) candidates who are doing well in the polls. For various reasons (yay statistics), you generally expect to see only a few candidates doing well enough in the polls that people will vote for them, or that (almost as important) corps will send them money to campaign with.

This is not just a social problem, it's a mathematical one. Basic game theory tells you that in terms of who's going to get elected this November, there's just no point in voting for Nader unless you truly do not give a damn whether Bush or Kerry wins.

Therefore, the solution must be a mathematical one instead of a social one; that is, we should change the election system. We should do this anyway, because the Electoral College is pretty much obsolete.

Condorcet Voting, for instance, does a much better job of allowing multiple parties to compete, for the reason that putting your favorite party (which you believe to be unlikely to win) at the top won't really hurt your favorite realistic candidate's chances. (There is a theorem that this can't be perfectly true in any voting system, but it's pretty close.)

Unfortunately, there is no analog that works this way for election funding. You only have so much money to give. Perhaps some sort of campaign finance reform could bring the system closer to sane, but enforcing mathematical principles on funding doesn't really work very well....

It is not his fault if the US electoral system is completely and utterly broken.

Very true. Unfortunately the chance of it ever changing seems extremely slim. Its my observation that electoral reform is something most americans have never thought about and are really not interested. There is plenty of stuff on the net like FairVote.org, ElectionMethods.org but they don't seem to be having much success in bringing it into public debate.

As a foreigner in the US, my few attempts at discussing this with people who I know are intelligent and politically aware has been a dismal failure. The suggestion that they don't have the best system in the world is just not something they are willing to consider.