Preferential Voting and the VOTE-123 system

Preferential Voting means you get to rank your choices for your vote. You mark a "1" next to your first choice, and then you can put a "2", "3", etc. next to your backup votes in case that person can't win.

1.2. What are the advantages of Preferential Voting against the current system (Biggest Minority Rules)?

Here they are:

1.2.1. - Majority rules, every vote counts.: In the current US system, all someone needs in order to win is the biggest minority. You can win without a majority - both Clinton and Dubya won even though the majority voted against them. That means it's possible for most of the voters to get the candidate they like least! This is impossible with Preferential Voting; even when there's no majority favorite, the system makes sure that whoever wins who is preferred by the majority. Over 50%, or you lose.
The current system fails to be majority rules because it's throwing away votes! When a third-party candidate (a Perot or a Nader) enters the race, there's a lot of argument about whether voting for them is "throwing away your vote". In reality, it's not the voters who throw away those votes, but the system. Preferential Voting is a way of making sure that every person can have a vote in the decisive contest between the final two strongest candidates.
1.2.2. - Honest Voting: In every election under the current system, many voters face a lose-lose choice. They can make a symbolic (but essentially ineffective) vote for their true preference; or they can vote for the "lesser of two evils" (essentially lying about your preferences). With Preferential Voting, you get the best of both worlds. You can vote your first choice honestly, then the "lesser evil" as a backup choice. Whoever wins, you know your vote counted. There may still be tough choices to make, or occasions when a strategic rather than an honest vote might happen to pay off (more on this later), but that would be the exception. The system would no longer systematically disenfranchise huge numbers of people year after year.
1.2.3. - Less Negative Campaigning: Since candidates would need second-choice votes from their opponent's supporters to win, simply insulting your opponent would be a losing strategy. (That goes double for political allies - no more wasted energy on trash-talking between, for instance, Nader and Gore).
1.2.4. - Break the duopoly (the two-party monopoly).: All the two parties do all day (when they're not fundraising) is squabble over a few "swing" voters at the center. But even if you're one of those much-sought-after swing voters, you would benefit from a wider debate. There's plenty of issues that get no attention just because the two parties agree about them. If they can't use an issue to score points against the opposition, they ignore it - and the media mostly follow their lead. With more parties bringing issues to the table, nothing would ignored.
1.2.5. - Gives Room for Campaign-Finance Reform: The current system has the effect of magnifying the early advantage that money buys. Money is the fastest way out of the catch-22 of new campaigns - the fact that no one will pay attention to you if they know you're "a waste of a vote". This system rewards corruption.
Preferential voting would give the little guys a better chance because you can vote based on the message, not who's winning the crooked horse-race. And if a few of the little guys who won were from little parties, they wouldn't owe any loyalty to the party hacks who would start shooting them in the back as soon as they stood up for real campaign-finance reform.
1.2.6. - Improves Turnout: All of the above advantages will work together to improve turnout. Fewer people will feel disenfranchised; more will be happy with our democracy; there will be more meaningful choices; and you'll be able to say more with your vote.
1.2.7. - Cheaper and More Accurate than Two-Round Runoff: One old half-fix for Biggest Minority Rules is the two-round system where the top two vote-getters from the first round advance to the second round. Running a whole extra round of voting is expensive; turnout often drops; and sometimes the wrong two advance. France saw this problem in their 2002 elections. The nation reacted in horror when a very split vote led to a neo-fascist advancing to the second round - only to prove what everyone knew, that he had no broad support. Under PV, he would never have made it into the top two. Preferential Voting acheives more reasonable, more democratic results, with less hassle.

1.3. What are the arguments against Preferential Voting, and how would you answer them?

First off, if you're campaigning for some form of PV, don't get stuck on defense. The opposition will simply try to muddy the waters and promote fear, uncertainty, and doubt. They know that when people get confused, they usually vote "no".

So if they're just blatantly lying, don't get caught up in a technical argument about why; just point out who they are and why they're lying - "You're a professional lobbyist, you're afraid this will put you out of work, and you're not even trying to convince people rationally to vote no, you're just trying to scare and confuse them!".

For instance, in the campaign against measure A, the opposition sent out fliers claiming that "your right to vote is in danger... [from a system that] has no track record". They draped these blatant lies in rousing images of democratic crusaders. Fliers to Chinese-Americans had pictures of Tienanmen Square and the ones to women had pictures of sufragettes and NOW rallies (even though both NOW and the League of Women Voters both endorsed measure A). In a Republican town, they would have probably used a picture of George Washington saying "No taxation withoute Representation." ... even though measure A saved money.

But as long as you can keep from getting more technical than your opponent, here are the more specific rebuttals for their arguments:

1.3.1. - "Too Complicated"

For the voters, Preferential Voting is as easy as 1-2-3. It's easy to put your preferences in order, and you only have to vote for the candidates you care about. Opponents will often conjure up an insulting portrait of some hypothetical hardworking immigrant whose English is not good enough to count to 3. In fact, nonpartisan immigrant groups support Preferential Voting because it is a fairer system that counts every vote. Preferential voting systems in the US have consistently elected more minorities, not less, giving a fairer representation of the electorate.

Counting the votes is also a clear, straightforward process (whichever system you choose VOTE-123 or IRV) that can be done as easily by computer as the current system. If you have to fall back on a hand recount, that is also clearly feasible.

Sure, I make things complicated when I try to answer any question anyone could ask. So just read the answers you care about.

1.3.2. - "A computer picks the winner"

The voters pick the winner, the computer just counts the votes.

It's true, the fastest way to count PV is electronically. It's also the fastest way to count regular elections, and many places already use electronic vote-counting machines. (As we saw in Florida 2000, there's problems and potential frauds with mechanical and hand counts too!) But as long as you keep a paper trail of ballots and you can check the results, the courts have the power to overthrow or punish any fraud. Especially with VOTE-123 (a particular kind of PV, explained below), there are even fewer possible kinds of fraud and just as many chances to hand-check the machine results as with the current system.

1.3.3. - "One Man, One Vote"

Whenever Preferential Voting comes up, you can rely on opponents to say this as if it's a counter-argument. It's not. It's just a trick to get you defensively explaining things. "One person, one vote" means that everybody get's the same chance to vote, and that every vote is counted equally. Preferential Voting satisfies this. Each voter has their vote count for one and only one candidate at a time.

...shoot, I fell for the trick. If your opponent says "But that's not one man one vote", the right answer is just "That's a lie. Yes it is."

1.3.4. - "Un-American", "No Track Record"

Preferential voting has a rich history in the US. It's been used in local elections in many different parts of the country and there's even language that favors it in an several Supreme Court decisions (Thomas & Scalia, in Holder v. Hall, and an older decision I've lost the reference to - email if you can help me). And in many parts of the world, Preferential Voting is the norm.

But more importantly, improving our democracy is the American ideal. In the 18th century, Alexis de Tocqueville said in Democracy in America: "Democracy is like a rising tide; it only ebbs to flood back with greater force, and soon one sees that for all its fluctuation it is always gaining ground." And he was right. Year by year, decade by decade, we've won rights for the landless, for African-Americans, for women, for 18-year-olds; and other democratic rights such as the right to directly elect our senators, citizen initiatives in many states, term limits, campaign finance limits, public access to the airwaves, better voter registration, recall elections, etc. Every 30 or 40 years there's a major advance, and even with the backlashes that sometimes come in between we've still made steady progress. It's been about 35 years since the last major advance - the 1968 Civil Rights act - so it's time for the tide to rise again.

1.3.5. - "Give third parties a chance, and the loonies will take over."

Preferential Voting means majority rules; it's not going to somehow let a minority take over. Instead, the majority - dissatisfied people in both parties, independents, and nonvoters - would have real options and a real voice. All political parties would have have something to offer the mainstream. If a "third" party was out of touch with the voters, it would stay "third". But if the "major" parties didn't get in better touch, they'd end up out in the cold too.

1.3.6. - "There's no perfect system, so why try to improve?"

It's true that there's no perfect system. There's a mathematical proof (Arrow's theorem; see the "Technicalities" section) that any voting system has flaws; and even if there were a flawless voting system, human nature wouldn't change. There will always be corruption and political pandering from the top; there will always be voters looking for a scapegoat, and voters who want to have their cake and eat it. Politics will always involve shifting alliances that give extra leverage to certain minorities. And politicians will never really keep their promises unless the people stay alert and organized to force them to.

This reform doesn't fix all of that, or even most of it. But it lets us out of the straitjacket we're in so we can start to deal with those problems. And, just as important, it's an issue where people can get together and organize across "party lines" and without needing politicians to lead us, which is the healthiest thing we can do for our democracy.

1.3.7. - "Supported by Special Interests"

This kind of cheap personal attack says a lot more about the people making it. In fact, the biggest opponents of Preferential Voting are the ultimate special interest - political consultants and palm-greasing political action committees.

1.3.8. - "Too much work to rank so many candidates"

Remember, you only have to rank the candidates you care about, so you can ignore the wacko candidates if you want to. But, that said, it's still a little more work to listen to different arguments and make real choices. That's democracy. If making choices is too much work for you, I'd be happy to tell you how to vote :).

But seriously. If you really can only handle making one vote, you'd still be able to do that. Your vote would still count just as long as your favorite had a chance - just the way it does now. Just because you like to limit your options, is no reason to limit everyone else's.

1.3.9. - ""

Sometimes the problem is not opponents, but apathy. Then, you have to put the issue in context. Whatever other issue you care about - taxes, rights, tranquility, or public programs - a livelier democracy is at the heart of getting things done, and its the only sure way to protect the progress you make. Talk to people in language that they understand about how better democracy would matter to the issues they care about.

2. The politics of PV (or: Get Involved)

2.1. Is PV a liberal idea, or a conservative one?: Neither. It is supported by people across the political spectrum - Libertarians and Greens, Republicans and Democrats. So we have to learn to work together. For instance, if you're a Green who advocates PV, please don't imply that it would give any special advantages to progressives. It would let new actors onto the political stage, but nobody would become a star unless they could sway the American mainstream.
2.2. Who would logically support Preferential Voting?: Just about everyone who cares about democracy. The strongest supporters would probably be members of third parties, because they're the most disenfranchised by the current system. Then there would be rank-and-file Republicans and Democrats who are tired of seeing the irrational effects of vote-splitting. If you live in Washington state right now [late 2002], the current system has given you the wrong president (a Republican), the wrong senator (a Democrat), and a few other wrong leaders.
One group of supporters that it would be fatal to discount are the officeholders. It feels good to pass sensible legislation that makes your constituents happy. Most incumbents are glad to see the logic of this reform, if they get a strong message from their constituents; only a few are so wedded to the current corrupt system that they truly oppose this.
2.3. Who would logically oppose Preferential Voting (PV)?: Logically, the only people against PV would be the people who benefit from distorting the democratic process. First and foremost, that would be the political consultants who make their living crafting and test-marketing negative campaigns for candidates. Second in line against PV would be rich lobbying groups, who want political campaigns to be as expensive as possible, so that the candidates are ever more desperate for their money.
Third, there would be the party hacks for the two major parties. Most everyday Republicans and Democrats alike would love the better democracy that PV would bring, but there are two kinds who wouldn't. There the zealots who have confused their ends (passing legislation they like in order to make this a better country) with their means (electing "more Democrats" or "more Republicans", not just "better people"). And there are a minority of career politicians who actually like the high-rolling, corrupt campaigns of today, who don't want real democracy keeping them honest. Luckily, most politicians would rather win a fair election. Even the ones who sell out a little to win would rather not have to.
2.4. But my friend thinks it's a bad idea, and she's not a crooked polititian...: True, some people oppose PV just because they're suspicious of anything new. I can't blame them; there are so many people out there trying to make bad ideas look good, that an actual good idea can be the last thing you expect. The best thing to do with these people is to explain who you are and why you support PV. When they realize that you're not selling snake oil, they might not decide to support it, but they'll at least stand out of the way.
2.5. Can you show any evidence that the opposition to Preferential Voting (PV) really is so corrupt?: Sure, I have 2 examples:
2.6. How do I get involved in activism for Preferential Voting (PV)?: Here's how:

3. Specific versions of Preferential Voting (or, IRV and VOTE-123: the Tomorrow Twins)

3.1. What are the different systems for counting Preferential Voting?: There's VOTE-123; Instant Runoff Voting (IRV); and, for electing many people at a time to a representative body such as a school board or Congress, there's different forms of Proportional Representation (PR), not treated in this FAQ.
3.2. What does VOTE-123 stand for?: Virtual One-on-one Tournament Elections using 1st, 2nd, 3rd choices. All those words describe how to count up who wins, but to vote in this system the simple name is enough: you just vote 1, 2, 3, putting the candidates in order of preference.
3.3. I've never heard of VOTE-123 before, and when I do a web search on that nothing comes up. What gives?: VOTE-123 is more commonly known as "Condorcet Voting" after the man who invented it. There are several minor variations of this style of voting, and VOTE-123 can refer to any of them.
3.4. Why doesn't this FAQ talk much about Proportional Representation (PR)?: I think that PR is an important reform, but I can't do everything. Also, the picture with PR is more complicated - you have to weigh the advantages and the disadvantages before you see what a good idea it is. (I'm too honest to leave out the disadvantages, so I'd only confuse you; you should hear the story first from someone who's better at simplifying).
Anyway even if you had PR, you'd still need VOTE-123 or IRV for single-winner races. And with luckily, with both VOTE-123 and IRV, there are NO REAL DISADVANTAGES compared to the current system, so I don't have to lie about anything.
3.5. Where can I learn more about Proportional Representation (PR)?: I have some good places to start in the links section
3.6. Describe the system of Virtual One-on-one Tournament Elections (VOTE-123), in 50 words or less.: Voters just rank the candidates in the order they prefer. Any candidate who is preferred by the majority of voters over each other candidate individually, wins. In the rare case that there is no-one who can beat everyone else one-on-one, whoever comes closest to doing so wins.
3.7. Describe the system of Instant Runoff Voting (IRV).: Once the voters have ranked their preferences on preferential ballots, the votes are counted a number of times. The first round, your vote goes to your first-choice candidate. If nobody gets a majority, the last-place candidate is eliminated, and if you had voted for them your vote gets shifted to your next choice. Candidates are eliminated until one has a clear majority.

4. Which version is best (or: Grudgeless Match between the Tomorrow Twins)

4.1. Would VOTE-123 and IRV ever come up with a different winner? When, and why?

Let's take the 1992 Bush Sr./Clinton/Perot election as an example. You have a conservative, a liberal, and a wild card. Say that in the hypothetical state of North Virginia, 40% voted Clinton as their first choice, 35% went for Bush Sr., and 25% went for Perot. But say nearly all the second choices, from both Clinton and Bush voters, go for Perot instead of to the "opposing" party.

Under IRV (after all the piddly also-rans were weeded out) Perot would be last place and would be eliminated. Then, the Perot voters would decide the election. Say a solid majority of Perot's voters - 16 of the 25% - put Bush Sr. as their second choice, so Bush has his 35% + 16% from Perot = 51%, and Bush wins it.

How does VOTE-123 decide this? In the Perot/Bush race, Perot is preferred by most of the Clinton voters (say, 35 of the 40%) so he gets his 25% + that 35% = 60% and wins it. And in the Perot/Clinton race, Perot again gets (say) his 25% + 30% from Bush voters = 55% for a win. So since Perot can beat both Clinton and Bush Sr in head-to-head races, he wins the "Virtual One-on-one Tournament" and wins the election. Who wins the Bush/Clinton race - which under IRV was the final, decisive runoff - doesn't end up mattering.

(Note: this page is nonpartisan and nothing here is intended to be a comment on the actual politics of any real candidates, examples are illustrations only, and all numbers are made up.)

4.1.1. In the above example of VOTE-123 and IRV disagreeing, which is right?

(Again, I have to emphasize that this is an example and not meant to comment on the merits of these specific candidates). There are basically two possibilities. If the voters got a good look at all three candidates and voted their true feelings, then VOTE-123 got the right answer with Perot. The Bushies hate Clinton, the Clintonites hate Bush, but both sides agree that Perot is a fine compromise. But if Perot is more of an unknown; if both Republicans and Democrats only marked him as a second choice because "what the hell, he can't be worse than my enemy", then IRV got the right answer by choosing Bush. "What the hell", "Anybody's better than a Democrat", and "You can't do worse than a Republican" are all understandable sentiments, but they're dangerous reasons to put somebody in charge.

4.1.2. How can you claim that VOTE-123 is ever right to pick a compromise? What if I think that compromises are morally weak?

Some people argue that IRV right to avoid a compromise, even if the voters have voted thoughtfully. You may feel that a compromise is morally weak, or that a candidate with fewer strong supporters should never win. That's your moral choice to make, but unfortunately IRV won't work the way you want it to in this situation. If the voters really are thoughtfully choosing, many will be reading the polls and will see the tight race coming. People voting for an extreme candidate will realize that, if the compromise candidate is eliminated, the election may go to the other extreme. People who fear that too much will have to dishonestly vote for the compromise candidate. It's the same old "lesser of two evils" arm-twisting that causes so much trouble today.

4.1.3. Can you summarrize that for me? When do VOTE-123 and IRV disagree? When is VOTE-123 right? When is IRV right?

The two systems disagree when, between the top three candidates, there is a close election with a "compromise" candidate, who ranks number 1 for few voters but comes in number 2 for a broad spectrum of voters. VOTE-123 would pick the compromise candidate because of their broad support. IRV would eliminate the "compromise" candidate because they lack strong supporters, thus choosing a candidate with both stronger support and stronger opposition.

VOTE-123 is right if the voters have made a thoughtful decision. If the voters on either side really prefer a compromise over the other extreme, then that's what they should get.

IRV is right if the compromise candidate has not been well-scrutinized by the voters. A candidate should win for what they are, not for what they aren't. If voters mark their second choice only because that person "couldn't possibly be worse" than the "real opponent", that person should not win.

4.2. Which is better, VOTE-123 or IRV?

They're both a whole lot better than the current system - they share all the generic advantages of Preferential Voting. VOTE-123 has a few clear technical advantages, and gets the right answer when votes are based on responsible decision-making. IRV is better known and has a longer track record, and is safer if voters irresponsibly mark preferences without adequate scrutiny. On balance, I'd say VOTE-123 is clearly better, but advocates of either one should be able to work together.

4.3. OK, so according to you VOTE-123 (Virtual One-on-one Tournament Elections) is better. What are the technical advantages over IRV (Instant Runoff Voting)?

Here are the three advantages:

4.3.1. - IRV is still vulnerable to the "lesser of two evils" trap.

When IRV eliminates a third-place "centrist compromise" candidate, the winner will be one of the extremes. The voters for the losing extreme may be so afraid of that result that they decide to dishonestly mark the compromise candidate as first choice, above their true favorite, to prevent the compromise candidate from getting eliminated. Since VOTE-123 can pick a compromise the majority supports, instead of eliminating people, nobody would be faced with that choice.

For instance, if under IRV Nader had made a much stronger run, and threatened to actually pass Gore without catching up to Bush, those Nader supporters might still have had choose between voting their hopes (Nader) or their fears (Gore, because if Gore is eliminated the centrist Gore votes would go over to Bush and Bush would win.) (Note: It's fine for Bush to win, if he actually has majority support. This isn't about Republicans, Democrats, and Greens, but about real majority-rules democracy. A truly democratic system wouldn't scare people into voting dishonestly.

*Current System (Biggest Minority Rules):*	This picture represents the current situation. New parties start out with such an uphill battle against the "lesser of two evils" phenomenon that they can never get beyond square 1. The fix is in, the game's not worth watching.
*IRV:*	This picture represents the situation under IRV. Parties can get a fair start, but they have a tough hump to get over from about 25% (where they pass the centrist party on their side) to the 50% line (where they beat out both of the centrist parties). Within that zone, they still have to deal with all the old-fashioned bickering with people who should be their allies. Still, they have room to develop their offense and a good pass (a single strong campaign) could put them into winnable territory.
*VOTE-123:*	This picture represents the situation under VOTE-123. A fair playing field for new and old parties alike.

This extra advantage for the main parties may not seem like much. But if the goal is to keep the candidates honest, any unfair advantage just gives them that much more room for corruption.

4.3.2. - It is never safe to vote dishonestly. Period.

As you saw in the prior example, under IRV there are times when you might have to pick the "lesser of two evils" instead of your true preference. This is a real problem in itself, and it's even worse when voters blow it out of proportion. In Australia, where IRV is being used, there are a lot of funny (false) myths like that about how to vote strategically, and the big parties consistently reap the unfair benefits. (I report this based on individual discussions with Australians - my mother's family is from Australia. If you know of any studies or articles about this phenomenon, I'd really love it if you'd send me a reference.

In order to stop such myths, you need to be able to make clear, categorical statements. If you get some talking head on TV saying "you should always vote your true preferences - unless the situation satisfies the Pareto conditions for Condorcet-unfairness as detailed in my paper on toroidal instabilities in assymmetric voting blah blah blah," you're lost. Under VOTE-123, there is no dishonest strategy that isn't dangerous. (Even if that doesn't discourage you, the payoff is small, and the difficulty and public relations danger inherent in planning and organizing strategies is huge). It's impossible to predict whether your strategically dishonest vote will backfire and end up helping someone you despise. The math professor on TV can say it, the 2nd grade teacher can say it to their class, and everyone can understand it. And if anyone comes around saying "A vote for [Nader] is a vote for [Bush]", they'll get laughed off stage or (if they say it in a paid ad) prosecuted for false advertising.

4.3.3. - More Easily Counted, better protection from fraud

Under VOTE-123, each of the two-way races in the Virtual One-on-one Tournament can be counted locally, at the precinct level, and then to find the overall winner you could just add up the local results for each race. There's no need to ship the ballots around and maybe "lose" or "find" some. If there's a close race and you need to do a recount, only two or (very rarely) three candidates will be close to winning, so you only have to recount the few close matches between the "finalist" candidates. Since you're only recounting one or two close races, it can easily be done even by hand.

Under IRV, you have to add up the first-round totals from all the precincts before you decide who gets eliminated. Then you have to recount all the ballots for the eliminated candidate, add it all up, and decide who is eliminated again. Either you ship all the votes to a central location, or you need several rounds of counting and communication between the center and the precincts. Either way, there is more opportunity for fraud, it can slow things down, and it makes a hand-recount (or random spot-checks on the validity of the computer-counted result) more difficult.

4.4. Are there any advantages to IRV over VOTE-123?

There's a couple of them:

4.4.1. Better known, with more political momentum.: For instance, there are websites and mailing lists devoted to advocating IRV and getting the system adopted in the US. (see the links section at the bottom of this FAQ)There's nothing similar for VOTE-123, but I'm working on it - if you're interested in helping, please contact me.
4.4.2. Doesn't pick compromise candidates.: As I said above, this can be both a disadvantage and an advantage for IRV. But in the interests of fairness, I'm listing it here with the advantages.

4.5. So you're saying IRV sucks, and we should crush the IRVies? OR: If you don't like IRV, I hate you!

The whole point of these systems is that you can be honest about who your favorite candidate is and still also support other good candidates. Shouldn't we be able to do the same for the systems themselves?

5. Technicalities (or: I'm a geek)

5.1. Give me an example of a 3-way tie under VOTE-123. Discuss.

Lets imagine a race with four candidates - Alejandro, Beth, Chris, and Doofus. Say there are the following groups of voters:

Blocs of voters:	"Rural She-Males": 25%	"Basketball Uncles": 30%	"Young and/or Restless": 35%	"Dupes": 10%
First choice:	Chris	Alejandro	Beth	Doofus
Second Choice:	Alejandro	Beth	Chris
Third Choice:	Doofus	Doofus	Doofus
Blank (last choice):	Beth	Alejandro	Alejandro	Alejandro, Beth, Chris

Who wins? Lets get Doofus out of the way first. Nobody but the 10% of Dupes really like him, but against Alejandro, for instance, he gets 25% from the "Rural She-Males" voting bloc (who dislike Alejandro more). Still, with only 35%, he loses. In similar ways, Doofus loses against Beth and Chris.

But look at Alejandro vs. Beth. The "Basketball Uncles" and the "Rural She-Males" both have a slight preference for Alejandro over Beth, so Alejandro wins with broader support (55%). But notice that Beth actually has deeper support, because the 35% of "Young/Restless" love her and hate Alejandro. And the same kind of thing happens with Beth/Chris (Beth wins with 65% support, but the 25% "Rural She-Males" are especially unhappy with that) and with Chris/Alejandro (Chris's 60% of mild supporters win, even though he's hated by Alejandro's bloc of 30% "Basketball Uncles").

So A beats B beats C beats A, there's no clear winner. This is called a "Condorcet tie" and the candidates involved (everyone but Doofus) are called the "Condorcet set". Broad support balances against deep support; with three strong candidates, nobody has enough of both to win. The Condorcet tiebreaker says that whoever lost by the slimmest margin wins; that's Beth, who lost to Alejandro by only 55 to 35.

How would IRV have handled this? It would eliminate first Doofus, then Chris, leaving the final race between Alejandro and Beth. As we saw, Alejandro can beat Beth head-to-head so Alejandro would win. IRV papered over this weird tie situation by ignoring the fact that Alejandro would lose to Chris. But whether you ignore it or not, this situation is a real paradox. (If this were a soap opera, and IRV were being used, on the night of voting day, when she realizes how the votes are going, Beth would commit suicide. She'd be eliminated, her votes would swing to Chris, and Chris would beat her nemesis Alejandro. The loser dropping out would, paradoxically, give her voters more power and change who won! And here's another paradox: if you held an election for "most hated candidate", IRV would always pick the same winner among )

How would the current system of Biggest Minority Rules handle this? Well, if everyone voted honestly, Beth would win with 35%. But if one candidate had an advantage or disadvantage - something that most voters see as irrelevant to their own views but think other people might notice; advantages like a flashier, more expensive campaign or a more established party, or disadvantages like coming from a minority group - that could change the results. For instance, with enough money, Alejandro might be able to convince the Rural She-Males that "A vote for Chris is a vote for Beth", and win enough of them to take the election.

Reality check: this whole three-way-tie scenario is pretty implausible. It's based on the fact that candidate B appeals to A's voters but A doesn't appeal to B's voters. In the real world, two candidates will always appeal more-or-less equally to each others voters. If A and B are similar enough for a strong majority A voters to pick B second, then something like the same number of B voters will probably want to pick A second, too. With a reasonable amount of that kind of symmetry, VOTE-123 won't tie (unless the situation is so close that you'd need a recount no matter which system you used).

5.2. But the US has that weird electoral college to elect the president... how could we possibly use PV for presidential races?

Actually, stupid as it is, the electoral system is actually a big advantage for reforming the presidential elections. Each state chooses its own electors, and within certain parameters of democracy they are free to decide how to make that choice. Instead of having to change the whole country, all at once, we could move towards PV on a state-by-state basis.

Since the electoral system is already complicated two-round election, with voters electing electors who elect the president, things would not be simple. A state could proportionally allocate its electors to the top two vote-getters. Within the electoral college, if no candidate has a majority, there are actually many rounds of quasi-runoff voting. Electors could be bound by oath to treat that "non-instant runoff" as if it were instant runoff, with each elector standing for a given order of preferences and bound to change to the next preference if and only if the prior preference comes in last. Theoretically, they could break those vows, but the current system of choosing people who really belong to the appropriate party and fining oathbreakers a couple of thousand dollars seems to work 99.9% of the time.

5.3. Could PV actually fix the problems of the electoral college?

Yes. You could refine this system of "two-stage quasi-IRV" even further. A large state like California could even out the electoral leverage. The two "senatorial" electors would go to whoever got the majority. As for the rest of the electors, they'd start out split 50/50 the electors for a 50/50 vote. But then for every 1% extra of popular vote, the candidate would get 3% more electors, so that by 67% they'd have 100% of the electors. That would exactly balance out the 3-electors-for-one-district that the smallest states have. Each elector would represent the same number of voters, whether they were from a big state or a small state. The electoral system would still be in place, just as it is today, giving the an advantage to the underlying preferences of smaller states. But you'd win an extra elector whenever you could change the mind of (say) 10,000 voters, whether they lived in California or Alaska, so candidates could afford to campaign wherever the voters were. A voter would be a voter, Republican or Democratic, Texas or New Hampshire; you wouldn't get ignored just for being in the wrong place or the wrong party for your place.

In fact, with this last refinement, PV would have extra benefits for the voters in whatever state adopted it first. Your state wouldn't be one big chunk, so it would never be "safe" for any candidate to ignore.

So actually, the very complications that make the electoral college such a hassle the way it is, just make it easier and more beneficial to adopt PV!

5.4. How could you do a hand count for VOTE-123?

First, of course, separate the ballots in piles. Each pile is blindly double-checked by a random pair of counters; any counter who makes too many errors is disqualified.

To count, each counter has a table of n x n cells, where N is the number of possibly winning candidates (as noted above, N is down below 10 within hours of the election). For 5 candidates, you can make a 5 by 5 table with room for tallies in each cell and 5 dangling strips that fold over to cover each row. A tally in column X row Y means a vote for X over Y; if there are more tallies there than in column Y row X, then X is beating Y. To sum up the tallies for two piles, just add the tallies in each cell. To tally one ballot, check the first preference on the ballot - say, C. Mark a tally for each cell down column C, and then fold the strip across row C. Now check the next preference, mark a tally in each uncovered cell down that column, and cover that row. Continue until there are no preferences left on the ballot. Obviously, in a real election, you wouldn't have dangling strips of paper, you'd a cheap plastic gizmo to put your tally paper into, and you'd press a button at the bottom or top of a given column to cover the corresponding row and then swipe a slider to reset it for the next ballot. Yes, it's 2 or 3 times slower than counting a plurality vote, but it's 10 times more democratic.

5.5. How could you do a hand count for IRV?

If you have all the ballots in one place, it's easy. Separate the ballots into counted piles according to the first choice marked. If no one pile is 50%, take the smallest pile, cross out the first choice on those ballots, and redistribute. If any have no choices left, put them in an "exhausted" pile. Continue the process until one candidate's pile is over 50% of the ballots NOT in the "exhausted" pile.

The problem is that you have to have all the ballots in one place. You can transmit them physically or electronically, but you need every individual ballot. However you transmit them, it's an opportunity for fraud or incompetence.

If you don't have the ballots in one place, you can do the first count, transmit the results to a central location, wait until everyone else finishes and the results are all in, recieve the word of who to eliminate, recount those ballots, retransmit the results, wait again... It's a lot of chatter back and forth, and a lot of waiting for the slowest counter, not just once, but several times.

5.6. 2003 California Recall Election

5.6.1. Would PV be a good idea for the 2003 California Recall Election?: YES! With over 100 candidates, the winner is practically guaranteed to win with less than 50% of the votes - with maybe over 75% of the electorate voting against him or her (or, given the breadth of the candidate field, it...). PV would fix that.
5.6.2. Is it possible to implement statewide PV on a short time scale - ie, between the time a recall election is inevitable and the time it happens?: For IRV, the answer is sadly, no. One county should take not more than 6 months to get ready for an IRV election. But IRV on a broad scale takes one, unified system. A single county with a sweetheart contract with a corrupt and inept voting machines provider, and the whole thing is thrown off schedule. That was the case with San Francisco, which has as of this writing taken over 14 months to get ready for IRV after the passage of Measure A.
For VOTE-123, the answer is a qualified yes. VOTE-123 allows heterogeneous systems for vote counting, and you just add up the results. Thus, those counties which can perform the simple task of updating their technology could use it, those counties which cannot would just have to do a hand count.
5.6.3. Wait a minute... the hand count process for VOTE-123 that you described above would get pretty unweildy with over 100 candidates.: You're right. For situations where the number of candidates could get out of control, the law should state that only the top five candidates in terms of first-choice votes are eligible to win. Not once in a thousand years will this make any difference, but it makes the hand-count feasible. Each precinct first counts the first choice votes and reports their preliminary totals. Starting when 95% of all the votes everywhere have been reported, the secretary of state keeps a public list of which candidates could possibly be in the top 5 given the current totals. This list will be available within hours, using current technology, and will have under 10 candidates, shrinking quickly down to 5.

5.7. If VOTE-123 is like a round-robin tournament, what is the current system? What is IRV?

The current system would be like putting 7 teams on the court, then calling time after 30 seconds and declaring the winner as whoever has the ball. Sure, the very worst team is probably not going to win, but it's nothing near fair.

IRV would start out the same way, with 7 teams on the court. Every 5 minutes you'd kick one team off the court, whichever one had scored the fewest points. When you got down to 2 teams, you'd finish out a normal game. So overall, the last part would be fair. But during the first part a good team with lots of "assists" could get kicked off just for lack of baskets. So you'd end up fighting almost more against your "allies" (the other teams going the same direction) than your opponents.

VOTE-123 is a round-robin tournament. Each team plays each team, one-on-one, fair and square. But because of the magic of Preferential Voting, you could run all the games at once "instantly", so the teams don't get tired out.

5.8. Hey, wait a minute. In your three "advantages" for VOTE-123 over IRRV, the first two are really just the same thing, restated twice. What's the real third advantage for VOTE-123?

Here, does this count?:

5.8.1. -VOTE-123 lets more voters vote on more candidates more easily.

Under VOTE-123, if you think that two candidates are equally good (or bad) you can rank them equally. You don't have to split hairs. But under IRV, you either have to number each candidate in strict order, or just not vote on them. If you give up before you get to either of the two winning candidates, your voice won't count in the decisive runoff.

This is especially important on certain voting equipment. For instance, on ballots designed for the Optech Eagle (one modern vote-scanning machine), there's only room to mark first, second, and third choices. Under IRV, that means voting for only three people. Under VOTE-123, you can mark a few people at each level. For example, you might mark one person as your favorite, four candidates tied for your second choice, and two candidates for your third choice. It's almost certain that your vote would count, that one of those 7 choices would be among the top two candidates. [Note: I am not 100% sure that the Optech Eagle will allow this. From the ballot design it looks as though it should be possible but I am looking into the technical specifications further.]

This also means that under VOTE-123, no ballots will be counted out as invalid just for double-markings. It's one less rule that a semi-literate voter can mistakenly break. Remember all those double-punched ballots in Florida 2000? Wouldn't be nice if that were never a problem?

(To count the ballots with tied votes would be easy. Remember that VOTE-123 works like a Virtual One-on-one Tournament, with two-way contests between each pair of candidates. On the contests between two candidates that you marked as tied with each other, you don't care which one of those two wins, so your vote wouldn't be counted. Your vote would still count in the races you cared about, where one of the tied candidates was facing someone you ranked as better or worse than them.)

5.8.2. (But IRV can be rejiggered to do much the same)

In some versions of IRV, a "tie" vote is divided into equal parts among the candidates. So if I vote 3 people for my second choice, once my first choice is eliminated each of my second choices get 1/3 of a vote. If one of them is eliminated, the other two go up to 1/2 of a vote, and then when there is only one left my reconstituted vote rests with them. To me, this seems artificial and ugly. Essentially, two of my three choices will always be eliminated, and I have no say in which two; one of the eliminated ones may have been able to win it if they had had a full vote. But I have to admit this is a valid option, and so I can't list this as a VOTE-123 advantage except here in the geeky section.

5.9. You say "over 50%, or you lose", but in any real election there will be people who don't vote on the top candidates, so their ballot won't count in the decisive race. Counting those "exhausted" ballots, the winner could have less than 50% of the vote.

"Over 50%" means "over 50% of those who care". If you don't bother to rank various candidates, or if you rank them tied, or if you don't bother to vote at all, you are sending the message that you don't care which of these candidates win. This is no different from the current situation of low voter turnout (especially the dismal turnout in second-round runoff elections). In fact, since turnout would improve, you would be a lot closer to having a real 50% of the population, a true mandate.

5.10. Is a perfect voting system possible?

No. There's a mathematical proof, "Arrow's Theorem", that shows that no system of voting among a list of options can satisfy some basic criteria. Arrow showed that you there's no system which doesn't fail to meet at least one of the following goals (actually, my rewording of the goals makes them even tougher than Arrow's wording. The Arrow names for each criterion are included in parentheses): 1. Democratic (non-dictatorship): the system should try to count votes equally, not be a dictatorship where one person's vote is "more equal than everyone else's. 2. Always works (universality): the system should always be able to pick a winner. 3. Unbiased (citizen sovereignty): the system shouldn't rule out certain results beforehand. 4. Cannot be manipulated by the voters (monotonicity): you should never be able to make X win by dishonestly ranking X lower. 5. Cannot be manipulated by the candidates (independence of irrelevant alternatives):a candidate should not be able to change the results by entering or dropping out of the race, unless they can win.

5.11. So using the same mathematical language in Arrow's theorem, what are the flaws in IRV? In VOTE-123?

Both of them have problems with independence of irrelevant alternatives (IIAC). This means that they can be manipulated by controlling the pool of candidates. It also means that the voters can change the results by strategically ignoring (down-ranking) certain candidates who have no chance.

5.12. Wait a minute. Up above you say that under VOTE-123 "You can always vote honestly, period." But here you seem to be contradicting that by saying that no voting system can be perfect that way.

OK, you caught me. There's a theorem related to Arrow's theorem (the 'Gibbard-Satterthwaite theorem') that shows that there is no voting system without some strategy - that is, some situation where an omniscient voter could vote dishonestly and thus increase their satisfaction. But just as you can mathematically prove that there's always a situation like that, you can mathematically prove that (especially given the right tiebreaker) under VOTE-123 it is always too risky to try it. Specifically:

Strategic voting only works under VOTE-123 if, either before or after the strategic changing of votes, there is a "Condorcet tie" - what I've been calling a three-way tie. This situation is very rare - as I said, it involves nearly-balanced voting blocs which not only disagree about which candidate is better, but also about which candidates are similar to each other. Most especially, it is very, very hard to predict accurately using polls. Most people will never make the effort to reliably tell pollsters who their second choice will be, and in such a closely-balanced situation the tiniest polling error could throw off the poll's results. And if you switch around your vote based on a poll that's wrong, you're more likely to end up hurting than helping yourself - especially since your plans will probably leak out and people can plan a counterstrategy.

So, you can still say it categorically: under VOTE-123, there is never a good reason to vote strategically.

5.13. What is the "perfect" tiebreaker for VOTE-123?

As I say above, the practical differences between tiebreakers are so miniscule that the best tiebreaker is the one you can explain most easily and simply. That said, the most logical, mathematically most widely-accepted tiebreaker is Ranked Pairs (aka Tideman's), which breaks circular ties at their weakest link (that is, it gives preference to higher-margin victories. The implications of that definition are explained in the glossary). The argument as to why it is the best given honest voters has to do with bayesian statistics (treating each voter as an independent probablistic measure of who's better) and is somewhere on condorcet.org.

5.14. But you're a geek. Surely you must have some more complicated tiebreaker to propose.

Well, I'm glad you asked. I have my own tiebreaker which I call RASP (risk-averse-strategy prevention).

The most strategy-proof (for the voters) tiebreaker, and the one I personally would choose if I were the oxymoronic Democracy Dictator, is to draw lots among the Smith set (the ones tied to win). But then, a party could run a number of identical candidates, and, since we're dreaming, voters would be patient enough to rank all those clones, and all the clones would get into the tie, and then that party would have a better chance of winning. So to prevent this hypothetical, read the fine print:

If there is a strict subset E of the Smith set such that all members of E are "equivalent", that is, some member of E beats all the candidates that any member of E beats, then first (recursively) choose a winner from E and eliminate all the other members of E. For instance, if there are 4 members of the smith set, there will always be a two-member E (all 4-point digraphs which contain a spanning cycle are isomorphic), and so you'll be able to reduce the Smith set to two people. (There trivially cannot be two different maximal overlapping E's; handle all nonoverlapping ones simultaneously).

There are other possible tiebreakers, and any one of them has its logic. That's part of why I like the random tiebreaker. I feel you could really argue that any one of the Condorcet winners is the "right" answer, so drawing at random is not a bad choice. However, randomness doesn't sit well with the American populace (even though it is often the official tiebreaker in the current system), so I'll agree with RP as the consensus choice.

5.15. Wouldn't you get the same result as RASP just by picking a random ballot and using that ballot's top choice out of the Smith set?

Yes, and that's a whole lot simpler, but that doesn't work for my proof below.

5.16. Can you prove that VOTE-123, drawing lots to break ties, allows no risk-free voting strategies besides honesty?

Yes. First, you can look at this page for examples of strategies and counterstrategies under VOTE-123, and you can see that all of strategies there are tricky to plan and easy to foil.

Incomplete explanation of my proof:

Statement 1: No effective strategy exists where there is a single Condorcet winner both before and after strategizing.

Proof: If A is the Condorcet winner before strategizing and B is the Condorcet winner after strategizing, then A must have lost votes and/or B gained them in the A/B race. But if the strategizer prefers B to A, they would have already voted B over A, so no change would help this happen.

Statement 2: Thus, the only three possible kinds of strategy are 1. to change the result of a tie; 2. to cause a tie; or 3. to break a tie

Statement 3: Given a lot-drawing tiebreaker: In each of the above 3 cases, a strategic voter whose honest preferences are fully-ordered (ie, no two candidates rank equal in their own preferences) stands a chance of getting a WORSE result than they would have if they voted honestly.

Case 1: Because of our choice of tiebreaker, the results within a tie are random, so no strategy can work here.

Case 2: To cause a tie, you must cause at least one candidate Z to beat the honest winner A. If the strategic voter already preferred Z to A there is no strategy that will do this any more than an honest vote. So the set of possible strategic winners (A, B,...(0 or more extras here), Z) includes at least one possible random result (Z) which the strategic voter sees as worse than A.

Case 3: To break a tie, you must cause the new strategic winner A to beat their honest nemesis Z. But if the strategic voter had already liked A better than Z, then they would already have voted that, and no strategy helps. So the strategic winner A is worse than at least one of the possible nonstrategic winners Z.

Note that this does not contradict Gibbard-Satterthwaite ONLY because our assumption of fully-ordered honest choices may be untrue. It is possible that the strategic voters actually saw A and Z equally good/bad, or at least see some candidate B's superiority as more important than the A/Z distinction. But even in that case, in a three-way tie (the only kind likely in a thousand years) you have someone going to all the trouble of organizing a strategic bloc (with, given a free press, near-certain discovery of their plans and even counterstrategies possible) all for just a 1/3 chance of improving their outcome.

I think it unlikely that the electorate will accept a random tiebreaker (even though in most states they already accept one without knowing it). But the whole Condorcet tie situation is so finicky. For the purposes of making it impossible to plan a strategy, almost any tiebreaker is unpredictable enough to be essentially as good as random. (Good tiebreakers in genereal tend to make strategy impossible in one, but not both, of Case 2 and Case 3. For instance, RP makes case 2 strategies ineffective, and, if my calculations are correct, provides a safe counterstrategy for Case 1 strategies, but allows Case 3 strategies. But don't take my word for it, if you've even read this far it's worth working it out for yourself).

5.17. All right. Stop trying to convince me. Talk about the difference between IRV and VOTE-123 without pushing one side.

I'll do my best.

If everyone votes honestly, IRV tends to go somewhat more for the extremes, whereas VOTE-123 goes for a compromise. But this has a contradictory effect! There are some situations where human nature leads most people to prefer compromise, where as in others most of us will not want a compromise. So people may end up using strategy to try to make one system function like the other. Because of strategy, the final results of either system may be backwards from what you'd expect.

In the case of IRV, when there's a good compromise available, people may be scared into switching their first preference. This is analogous to the current "lesser of two evils situation. In the case of VOTE-123, when people on either side see the compromise option as only slightly better than the other extreme, they may vote their honest first choice, but strategically refuse to give their second choice compromise option (*). By cutting out the middle, they would be taking a gamble that their side might win.

These strategies, on the face of them, might not seem so horrible. Instead of being held hostage to strategy as under Biggest Minority Rules, voters are voluntarily choosing strategy to express their meta-preferences (whether they prefer compromise or steadfastness). But strategy introduces another set of problems. If people vote dishonestly because of some strategy, the parties that would have won those honest votes lose some of the "media clout" they deserve. Remember, politics is not just about power, but also about voice. The most admirable movements and parties are not focused towards to taking power, but towards communicating a message - in giving power to the truth - and elections should serve that goal. Therefore, strategy is always an overcorrection. It may elect the right person in the short term, but in the long term it will distort the debate.

Thus IRV's strategies will tend to cut the extremes out of the debate, whereas VOTE-123 will tend to cut out the middle. (Note that in this context, "extreme" is not the same as "strident". You can be thoughtful, or foolishly overconfident, no matter where you are on the political spectra). My view is that more of the creativity comes from the extremes, so VOTE-123 is better.

(*)NOTE: Actually, strategy should be practically nonexistent under VOTE-123. The only reason I include this answer, even though it may seem to undercut my other points, is that I think that it's funny. Two big systems, one pushing for compromise, and one pushing for steadfastness; but if pesky little humans push back, the results are actually backwards from what you'd expect. It's a good parable for the dangers of big government in general, and to share the joke, I had to ignore the asymmetries for one answer.

5.18. In the "Progressive era" of the 1930's, many parts of the US actually had Proportional Representation, a form of Preferental Voting. What happened, why was it repealed?

Proportional representation was unrelentingly opposed by the old corrupt political machines. They mostly played on people's fear of the kind of candidates that PR started to elect. That included racist pandering whenever minorities finally won some representation. Also, PR actually elected a few Communist candidates into local city councils. Eventually, in the 40's, the red-baiting and race-baiting worked, and PR was repealed.

The whole opposition to PR was based on lies, though. It's true that, because it works by electing a whole legislative body at once, PR will give a fair representation to minority parties like Communists. (These candidates could not have won for IRV or VOTE-123, where it takes a majority to win because there is only one winner at a time). But there was no danger that the Communists would take over; a majority of Americans oppose communism, and so a democratic system like PR would not have put Communists in charge unless that changed. The right answer was not to shut the Communists out to breed in the darkness, but to shine the light of democracy on them. Under such a light, the voters would have seen how they used the small, proportional power they were granted. They would have had to actually improve the lives of the workers, as they claimed to. If they just advanced the interests of the Soviet empire, as their opponents claimed they were doing, they would simply lose votes.

5.19. But I don't want to give any seats at all to Communists (or whatever people I don't like.) I'm afraid they'll stage a coup and take over!

The more democratic a system is, the less vulnerable it is to coups and revolutions. It's easy to find examples of drastically undemocratic, centralized systems that totally collapsed at the tiniest push (the Russian Czars, the Inca empire), or of democracies that have weathered great storms, but I think the best argument comes from a halfway case: Iraq. As long as Saddam Hussein's power was weak and his opponents provided real democratic threats, he was actually a decent politician, pushing popular and thoughtful reforms. But Iraq didn't have a strong democracy, so at a certain point he could just grab all the power and hold it. Power went to his head and he became a murderous megalomaniac.

Power corrupts, democracy and open debate prevent corruption. Preferential Voting improves democracy, which is the best recipe for avoiding a loony in charge.

5.20. I disagree with you. I think that VOTE-123 is not as good as [IRV, Approval Voting, Modified Party List Hokey Pokey, the system I worked out myself on the back of an envelope].

Great, another geek! Let's work together to get something that's better than the current system, and then we can part ways.

5.21. This stuff is fascinating. How can I learn more?

See the links section at the bottom.

5.22. Who is this FAQ written for? Do you really expect people to read the whole thing?

It's for the curious, or newcomers to voting reform, to browse a few questions. I hope it might convince a few IRV advocates to keep up the good work, but also raise the possibility of VOTE-123 within their organization. And I hope the exhaustive list of arguments and counterarguments are all presented in a sharp enough form that they're ready for anyone developing a campaign to pick and choose from. And sure I think people will read the whole thing, if they want a cookie.

5.23. I just read the whole darn FAQ. Do I get a cookie?

If you really read the whole thing, sure. Email me your address and I'll send you a homemade cookie (or three). Offer good while supplies last (ie, until I get really tired of it.) They're vegan.

6. Glossary (or: Alphabet Soup)

6.1. Approval Voting:: A system where you can vote for as many or as few candidates as you want, only 1 vote each. (Actually, you can do the same thing with VOTE-123 if you want to, but VOTE-123 is more flexible). Not actually a form of PV, but shares some of the advantages. Better than the current system of BMR. Especially good for electing people for a single, clear job, like school board or assessor; but if the person will wear multiple hats (like a president) it can be agonizing (and too important) deciding how good is "good enough for my vote".
6.2. BMR:: Biggest Minority Rules, my term for Plurality Voting, which is the current system in most of the US. The opposite of PV.
6.3. Borda Count:: A system of PV which assigns points to each candidate for each ballot - so many points for being first preference, so many for second, etc. Mostly just of academic interest. Works well if everybody gives an honest ranking of their preferences, but it's vulnerable to manipulation by both candidates and voters. Since your lower choices can help beat your top choice, most people would just end up voting only their top choice and we'd be back where we started.
6.4. Condorcet:: Another word for Virtual One-on-one Tournament Elections (VOTE-123). Condorcet invented this system.
6.5. Condorcet-Black:: A version of VOTE-123 with a specific tiebreaker. Clever, but somewhat complicated.
6.6. CVD:: Center for Voting and Democracy, a US organization that advocates for PV and PR.
6.7. Cumulative Voting:: A multi-winner system where each voter gets a certain number of votes, and they can spread them among candidates or pile them on one candidate as they choose. Not a kind of PV, but shares many of the advantages of PR.
6.8. FPTP:: First Past The Post, the British term for Plurality Voting (or Biggest Minority Rules), which is the current system in most of the US. I don't understand why they call it that, it probably has something to do with cricket. The opposite of PV.
6.9. Hare's method:: Another label for IRV.
6.10. IRV:: Instant Runoff Voting, a single-winner form of PV. Last-place candidates are eliminated and their votes are recounted, using backup votes, for remaining candidates, until someone has a majority.
6.11. IRRV:: Instant Round Robin Voting, another term for VOTE-123.
6.12. MPV:: Majority Preference Voting, another name for IRV.
6.13. Path voting:: Another name for RP. (A good version of VOTE-123.)
6.14. PR:: Proportional Representation, a general term for similar multi-winner systems which try to elect representatives in proportion to their support, like a cross-section of voters. Works for School Boards, City Councils, State Senates, or Congress; you'd still need a single-winner system such as VOTE-123 or IRV for mayors, governors, presidents, etc. Most PR systems use PV. This FAQ does not deal much with PR, but PR is a good idea.
6.15. PV:: Preferential Voting, a general term which includes VOTE-123, IRV, and most systems of PR. Voters rank candidates 1, 2, 3, etc. in order of preference.
6.16. Ranked Ballots:: Another word for Preferential Voting (PV)
6.17. RP:: Ranked Pairs, aka Tideman's method. A specific version of VOTE-123 with one sensible way to break circular ties. Ranked Pairs gives the ranking of the candidates that always reflects the majority preference between any two candidates, except in order to reflect majority preferences with greater margins.
It works like this: First, you eliminate completely all the candidates not tied to win, and all the one-on-one matches and scores involving those candidates. Then you make a list of the remaining matches, from widest margin of victory to narrowest margin. You go down that list, eliminating the loser of each match, EXCEPT when to eliminate the loser would cause a circular tie with people already eliminated. Before the end of the list, there will be only one winner left.
6.18. STV:: Single Transferable Vote, system that includes both PR (for multi-winner) and IRV (for single-winner).
6.19. Tideman's method:: Another name for RP. (A good version of VOTE-123.)
6.20. VOTE-123:: "Virtual One-on-one Tournament Elections using 1st, 2nd, 3rd choices". Also known as Condorcet Voting. A single-winner form of PV. Votes are counted in one-on-one races, like a round-robin tournament. Whoever scores a majority against each other candidate, wins. This is a general term for all systems which use pairwise results to get a winner. For my purposes, VOTE-123 can be considered as a single system, since specific VOTE-123 systems differ only in how they break circular ties - which, in real-world political elections, are very uncommon

7. Links

7.1. Discussion and Advocacy

7.1.1. What can I find at http://condorcet.org/ ?: A good site advocating VOTE-123 (referred to as Condorcet here). Lots of examples. More technical than this site.
7.1.2. What can I find at http://electionmethods.org/ ?: Yet another good site that advocates VOTE-123 (and calls it Condorcet). Mostly recaps the points I make here. A little too sectarian (ie, anti-IRV) for my taste.
7.1.3. What can I find at http://demochoice.org/ ?: A chance to use IRV and PV in web polls. They call them by the general term "preference voting". Also, you can download the software and run your own polls!
7.1.4. What can I find at http://www.fairvoteca.org/learn/quotedump.asp ?: On the Californians for Electoral Reform site, a large list of relevant (but unclassified) quotes.

7.2. Networking and Activism

7.2.1. What can I find at http://www.fairvote.org/ ?: Home page of the Center for Voting and Democracy. The definitive site for practical advocacy of Preferential Voting systems in the US. They are very pragmatically focused, which means that currently they focus on IRV and mostly ignore VOTE-123; but if we could (politely) convince them that VOTE-123 is not just better, it has a real political chance, they'd happily switch over.
7.2.2. What can I find at http://www.calirv.org/ ?: As above, for California.
7.2.3. What can I find at http://groups.yahoo.com/groups/instantrunoffca ?: Here's where you can sign up for an email list about IRV activism in California. Hopefully, one day they'll work for VOTE-123 too. To sign up for the national equivalent (instantrunoff instead of instantrunoffca), send an email to the address nate on the server instantrunoff.com (I don't use the @ sign here so he doesn't get spam). The national list is moderated for pragmatic activism relevancy and so there's not more than one short digest message a day. Also check out instantrunofffreewheeling on yahoogroups, it allows freer discussion.
7.2.4. ??? electionmethods list: An email list about election methods in general. I've never subscribed myself but calumnious rumor has it that there is a certain quota of holy wars between people who think they have the One Right Method.

7.3. Meta-links

7.3.1. What can I find at http://dmoz.org/Society/Politics/Campaigns_and_Elections/Voting_Systems/ ?: A page of links about voting systems in general. The same page with google rankings.

7.4. History

7.4.1. What can I find at http://www.fairvote.org/library/history/true_experiences.htm ?: PV was too democratic for Ohio in the 30's and 40's. Good account of practical results, analysis of effect on turnout, etc.
7.4.2. What can I find at http://www.nyu.edu/gsas/dept/history/public_history/PR/ ?: PV was too democratic for New York in the 30's and 40's. Good account of opposition strategies.

This list of questions and answers was generated by makefaq.

Any text herein may be used under the terms of the Python 2.2 license (with copyright holder changed to "Jameson Quinn"). Basically, use it however you want, as long as you give me credit somewhere.