Portland Green Party 2012 Endorsements and promoting Ranked Choice Instant Runoff Voting

PBA endorsement of Brady was "unexpected"?

I find that hard to believe.

And was that "hostile environment" Red&Black?

"Hostile"? To any progressive that environment would be HOME!

It greatly disturbs me that the Greens continue to promote IRV. That voting method certainly is backed by Rockefeller Brothers Fund grants, the Carnegie Mellon Fund, and the Ford Foundation, and I have found web links to prove it over the years, but the pages tend to disappear.

Many mathematicians consider IRV to be a menace to democracy. For example:

/ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ
Kathy Dopp
 http://kathydopp.com/wordpress/?cat=8

IRV/STV remove the right cast a vote with a positive effect on a candidate's chances of winning because it is non-monotonic (increasing the number of votes for a candidate may cause that candidate to lose, whereas the candidate otherwise would have won).
\____________________

/ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ
MPRnews
 link to minnesota.publicradio.org

St. Paul, Minn. — A mathematician who's spent years analyzing voting systems told a Minneapolis audience Tuesday that instant runoff voting is not the answer.

IRV allows voters to rank the candidates for a given office in order of preference. Winners are determined in a series of simulated runoff elections. IRV supporters say it's a better reflection of public opinion than so-called "plurality" elections, where whoever gets the most votes wins.

But Donald Saari, a mathematics professor at the University of California Irvine, told an audience at the University of Minnesota's Institute for Math and its Applications that IRV suffers from many of the same problems as traditional plurality elections.
\____________________

And so on.

But it's not just the mathematicians who condemn IRV. Many communities that have adopted IRV have later abolished it.

/ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ
REPEAL IRV- KEEP VOTING SIMPLE
 http://repealirv.blogspot.com/

IRV promoters are desperate as FairVote and IRV lose credibility and market share. You can only fool SOME of the people, SOME of the time. After a few IRV elections, people get wise to the game and may not know exactly how the game is run, or why it feels slimy, but they know the odds are always in favor of the house. Check out IRV for Newbies.
\____________________

Instant Runoff Voting - Facts vs Fiction
 http://www.instantrunoffvoting.us/

The blog below is about the problems with Instant Runoff Voting in the US - the latest reports, news,and commentary. Educating and informing the public and government officials about IRV.

/ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ
 http://instantrunoff.blogspot.com/

There's an article and also blog post in the Wall Street Journal about instant runoff voting. (There was referendum on IRV in UK that recently failed). Rob Richie [well-funded IRV advocate] of Fair Vote argues against voters having all of the vote data, because seeing it might cause voters to lose confidence in the election results.


MAY 13, 2011, 8:54 PM ET
A Split Vote on Alternative Electoral Systems
....
Richie said he prefers that the full details not be released, in part so that candidates who lose don't parse all the results in search of an argument that they should have won.
"I've come to believe we shouldn't put all the rankings out there for people to mull over and second-guess the outcomes," he said.
An important part of elections, he added, "is for those who have lost to accept they lost and move on."
He also would rather not arm instant runoff's opponents with material for what he calls "post-election `math masturbation' that IRV opponents can do when they have full tallies to work with."
 link to blogs.wsj.com
\____________________

Since IRV has all these problems, and has caused so much political upheaval, I just don't understand how the Green Party continues to advocate for it. I have noted that it seems to have been repealed in over a dozen places. It simply does not work, and can easily be manipulated by organizations with powerful computer programs. Why do the Green Party people persist in supporting a voting method that destroys democracy, and third and forth parties as well?

I have advocated the infinitely simpler "padded score voting" (PSV). This method is very simple: you can give as many candidates as you please, say, 10, 11, or 12 votes, and give other candidates no votes at all. Then the number of votes given is simply added up. Hand counted paper ballot counting is very easy this way. The voter's strategy is very simple. Some people insist that voters should be denied recourse to strategy, and should vote "honestly" (really heroically). But if they can't use strategy, they can have no power.

I helped this election process and ran for Secretary of State in 2008 on an election reform platform for the Greens.

Unlike Rob Richie, we promote all details being made public about IRV elections. In this election, we took range, approval, Condorcet, and IRV counts and the order agreed in all cases, so the election method used was not important. We did this because we know that Arrow's Impossibility Theorem shows that no election method can satisfy all criteria and so details were kept well enough that we could compute any common criteria. IRV is especially useful because it is the election system yet devised most immune to strategic manipulation by special interests in a real election. That's a mathematical result. Some mathematicians dislike IRV for other reasons and promote election methods that meet other criteria. The problem with our elections isn't so much any of the criteria they are promoting, however, it is the spoiler effect. Since IRV is the most immune to the spoiler effect and strategic voting of all election systems available, it is supported by most third parties. We are however open to what the community wants to support as the "best fit" for Portland, and this may include forms of Proportional Representation like Single-Transferable Voting or other alternatives. Most of the criticism of IRV on the links provided are incoherent or mathematically wrong. I could spend all day explaining the math of election systems, but I have other things to do, so I will limit it to the specific one mentioned.

The election method mentioned as an alternative fails heavily due to manipulation. Adding candidates increases spoiling and dilutes final votes. It collapses to a form of approval voting when it is NOT being manipulated by strategists. At least use range voting which fails to match approval voting when it IS being manipulated. Every election system has artifacts because (here's a layman's crude explanation of Arrow's Impossibility Theorem) it relies on the idea that we can take an entire electorate and choose one candidate that will satisfy everybody. This is why we support direct and liquid democracy systems above representative democracy and one of the questions for our candidates was whether or not they support the work of the Charter Review Commission to expand our city charter to be more inclusive and to reform the election system in Portland that has single-candidate city-wide elections where only the rich and powerful typically end up winning.

The commenter about R&B is correct about the location. Brady was met at R&B just because it was ideally located next to St. Francis, but it was a small gathering. Political groups often meet in groups smaller than eight at random to local service centers with an open door and tables. I put about $40 to the tip jar, all the cash I had, as they seemed uncomfortable with Brady there and like what they are doing and enjoyed the discussion their staff created with Brady. If we were publishing it widely we would have asked permission in advance, but it was really just an attempt to have a discussion with a candidate at one table before hopping over to the Occupy office to show them around. The vote result was very close between Brady and Smith. The press release was updated at the request of Brady to change from unsolicited to unexpected, sorry for missing that detail. I should have caught that when I read it for review. In fact, Smith, Brady, and Hales all solicited the PBA endorsement very strongly.

Our first choice candidate, Cameron Whitten, spent his nights camping at Occupy when the camp was going. He talked his way through a police blockade that shut down City Hall to file for a run for mayor while protesting at City Hall. We should focus on our first choice candidate and note that >50% approval was met by the other two and people are free to interpret the results as information rather than as a directive to vote a certain way. Our goal is to inform people when it comes to endorsements. No party member is required to campaign for endorsed candidates, and in fact, Tre Arrow filed after our endorsement process. Some in this environment and no doubt some of our members will end up supporting his run.

In the end, Brady has taken the endorsement process in a positive direction and decided to give Cameron more prominence at debates to highlight the issues that mainstream candidates don't address. I was surprised by that, but IRV is known for creating positive environments in elections to mutually campaign. Please comment and provide whatever feedback you like, even if critical. PDX IMC is an excellent open forum.

Forget about the mathematical nonsense of Arrow and his "law" and his Nobel Prize (Obama got the Nobel Prize, as did the inventors of modern financial derivatives, Robert C. Merton and Myron S. Scholes). It's not just math; it's about Power.


The Irv Voting Method:

With IRV you have to "rank" the candidates, giving one candidate first rank, another second rank, and so on. If it's done on paper (almost an impossibility) it goes like this: All the ballots are transported to a single counting center. If someone has 50% +1 first-rank votes, they win. If nobody has that many, then the ballot of every voter who gave a first-rank vote to the one with the least first-rank votes is "amended" such that their "first loser" is eliminated, and all the other candidates that were ranked below their first loser are "moved up" a notch. In a real democracy, there might be at least 30 candidates, so there would be quite a job possibly amending millions of ballots.

/ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ
Check out:

Voters in District 5 will have 22 candidates to choose from when they go to the polls Nov. 2
Link to articles.sfgate.com
 http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2004/08/12/BAGJ086K4F1.DTL

And:

25 candidates file for Washington County judgeship
Link to www.startribune.com
 http://www.startribune.com/local/east/96110574.html
\____________________

Then, the first-rank votes are counted again, including the "new" ones that were "moved up" from second-rank. If somebody now gets 50% +1 "first-rank votes," they win. If not, the second poorest loser is "eliminated," the amendment process is repeated on perhaps 100 million ballots, and the old and new "first-rank votes" are counted again to find if someone gets 50% +1 "first-rank votes" (who may not be the one that initially had the most first-rank votes). And this entire process is repeated until someone gets 50% +1 "first-rank votes." Something is a little odd here? Okay, say we have an election with George W. Bush, Al Gore, Pat Buchanan, and Ralph Nader. In the first round Al Gore gets exactly 50% of the first-rank votes. He would have won with 50% +1, but some "honest" (or "sincere") voter (as most are wont to say — but "heroic voter" would be a more neutral term for this) gave first rank to Nader and second rank to Gore, so we must do a big shuffle (vote "transfer"). People have actually claimed that since since Gore already has 50%, he surely will get one more in the second shuffle, thus win! Not so fast! Due to the shuffle, many of the second-rank votes are now first-rank, so Gore no longer has his 50%! In fact, Bush, who had only 47% of the original first-rank votes, "inherits" many votes when Buchanan is eliminated, and winds up having 52%, and of course thus wins. But if that "honest" voter had only given first-rank to Gore instead of Nader, Gore would have won. This is actually the spoiler effect (or black hat effect) yet again! And this is why communities that try IRV end up right back with two-party duopoly. It is rather obvious that this method requires computer tabulation. I was one of the first people to state that computers have no place in elections, on BlackBoxVoting.org and BradBlog.com, in 1995. Virtually no one else was saying that. Now most of them agree completely. Computers cannot be made safe for voting. Using computers for that is like using a bulldozer to mow the lawn. So I played around with voting methods. Some, such as Condorcet voting are much safer than IRV, but are even (much) more complex.

This is not like a few people in a small meeting electing petty officers. This is the big leagues, where the contest is NOT between individual candidates. NO! This is between candidates supported by the Power Elite Oligarchs and supporters of the few who might actually serve the people and nation. It's an entirely different ball game.

The voting public must use strategy, not foolish heroic voting, to have any chance of gain. In a padded score voting (PSV) election, I would give Nader 12 votes (or 11 if I really trusted the counting processes, but I tend not too), Gore 10, and give Buchanan and Bush no votes at all. This would not hurt Nader, and would hurt Gore very little (it would be only a 20% hit). It is simple, and can be done with any number of candidates in a contest of any size with hand counted paper ballots.

This is not much like Warren D. Smith's range voting, with its confusing 0 to 99 scoring and pointless "percentage" system that requires computer tabulation.

We can argue forever about the "theorems" of complicated systems, but padded score voting (PSV) is easy to understand, and, foe all practical purposes, eliminates the spoiler effect if the public is aware enough to cast intelligent strategic, rather than falsely heroic or "honest", or "sincere" (really naive) votes.

This is the real way to achieve real spoiler free democracy.

It was late, when I said "In a padded score voting (PSV) election, I would give Nader 12 votes (or 11 if I really trusted the counting processes, but I tend not too), Gore 10, and give Buchanan and Bush no votes at all."

Actually, If I really greatly trusted the counting processes, I would would STILL give Nader 12 votes, but would give Gore 11, rather than 10 votes. This would improve the effectivity of my ballot. (The ability to give low numbers of votes to candidates (an unpadded method) would create an excessive spoiler effect in the election process, and make voter decisions and vote counting pointlessly complex.) I would know that I was giving Nader a (very nearly) 10% advantage over Gore; so I would clearly know the effect of my decision.

Clearly, padded score voting (PSV) is much simpler than other methods, and makes strategy much more effective for the intelligent strategic voter. People should vote strategically, not heroically, since the latter approach leads to the election of ogres (enemies of the 99%) such as Bush.

I feel there should be no need for further complex mathematical "game theory" arguments now. A small amount of extra complexity could make them go on for many pages!

blues, facts are important, ad hominem attacks are not.

I'm not sure where most of your assertions come from, but as a computer scientist my day job is to implement mathematical methods based on algorithmic research, so I'll try my best to clarify a few facts that contradict most of your assertions.

First, the total percent of the vote for a candidate doesn't ever go down. It can only go up as ballots without full rankings are exhausted and the total voter count decreases, leading to only monotonic increases in vote totals for all candidates equally, either by count, or by percent. In our election method we didn't reduce the total voter count when ballots were exhausted so that approval had to be from all cast ballots, not remaining unexhausted ballots, but that wouldn't matter anyways as adding or subtracting from the denominator affects all candidates equally.

Second, the transfer of ballots only happens if nobody gets a majority in all previous rounds. Only one candidate is eliminated at a time (unless you perform an optimization that is mathematically equivalent for IRV and STV elections, removing all candidates whose total votes is fewer than the Nth place candidate's first choice votes (where N is the number of seats)). You take the removed candidate (or candidates) and transfer only those ballots. So Gore's 50% means Bush can never get more than 50% of the vote, only equal it, Buchanan can't have enough first choice votes when he is eliminated to get Bush up to 50%, only 49.99..%. Technically Nader will be eliminated first, but it doesn't even matter what order between the two who are eliminated. What will happen is Nader's having only one first choice vote, in the second round Gore is ensured to win no matter what Bush and Buchanan's ballots said as at the end of that round Gore will have 50%+1.

This only required the transfer of one ballot to determine the winner.

Third, IRV is hand counted in many jurisdictions. I hand count IRV ballots all the time. It's the simplest task. Once you've piled up votes together you see if anybody got >50%, if so, done. If not, you take the smallest pile and move them to the other piles. Then you see if >50% was met, repeat if not. It doesn't even require central tabulation. You can wait for total counts before doing transfers at all decentralized stations or you can provide the central office the count of all unique ballot rankings and then let the central tabulator perform the rounds, either by hand or by computer, but that too can be done by hand very easily.

I'm a computer scientist and I oppose computerized elections. Thankfully IRV doesn't require computers at all and existing hardware is compatible with performing the elections.

Fourth, Range Voting also doesn't require computers, either, as it can be hand tabulated since it only requires sums, the same as your proposed method. I oppose Range Voting in general because it's manipulated by strategic voting, greatly. It reduces to approval voting when used strategically (only ranking with the highest and lowest scores). That is why I only use Range Voting for informational or straw polls where manipulation would be evident and correctable since they are non-binding.

Fifth, PSV as you call it and describe it, isn't much different from "Range Voting". Internet searches for the term only bring up your writing  http://www.google.com/search?q=%22padded+score+voting%22 , so it's clearly not an established voting method. For example, if you give them a range from 0 to 12 to vote with, is it any different at all from Range Voting? Range Voting is merely the sum of scores, just as yours is. So mathematically, it appears equivalent, thought maybe you neglected to mention more rules that would make it different. How do you determine how many votes are allowed? The problem of course is that it reduces to approval voting in the strategic case so your attempt to introduce ranges generally fails with intelligent voters.

Sixth, the idea that we should allow strategic voting is fundamentally an odd one. Social Choice Theory defines strategic voting as the use of insincere voting to maximize sincere outcomes. Any vote system designed to allow insincere voting violates the principle of one person one vote where everybody has an equal power. If voters have to vote against their sincere interests just to exercise power it introduces spoiler effects. Oddly enough, your PSV will lead to a spoiled election in the example you mention. If you vote 12 and 11 for Nader and Gore, that is one vote point less. Let's assume all Buchanan voters voted 12 for both Bush and Buchanan (since they are smarter than you, they know how to use strategy better and know Buchanan has no chance of winning). Let us also assume that no Bush voters voted for Gore (since, as I pointed out, they know strategy better than you do and know Gore has a chance of winning and must vote him 0 to ensure he doesn't get more votes than Bush). This means Bush has 50% and Gore now has 50%+11/12 (your vote). That looks like Gore wins, right? No, see, I didn't mention that no Gore voters also voted for Bush. If just a few Gore voters (12) ranked Bush with 1 vote, all of a sudden, Bush takes over Gore. So it is in their interest to vote Bush with 0 votes, but they were being sincere in saying 1 and now Gore loses, spoiling the election, all because they didn't pick the right strategy while others did.

In IRV, the sincere vote comes up with the correct and expected result.

See blues, math is important, it's fundamental to understanding election methods.

Since you used a two party system argument against IRV, it would be useful to point out that all single seat election systems, regardless of counting method, force a two party system upon the electorate. Only Proportional Representation systems prevent that. Approval or your method also create a two party system. Only a system where a significant minority is allowed to take a useful minority of seats will break the two party duopoly, and that cannot happen in any election method where the top two parties are most likely to win.

Based on your description of PSV in your first post, I misunderstood that it required separate vote amounts for each candidate and figured different properties from Range Voting. From your post today I figured out that you meant you could apply equal values, meaning it is more like Range Voting, and exactly like it, essentially. That is why my first reply gave a different technical critique, but that critique was slightly wrong (and I suggested Range Voting would have better properties). My posts today have the correct critique, the ones common to Range Voting.

I used to work with second order symbolic logic. Liked it because, unlike set theory, it required no (substantial at least) axiom system. Today I work with symbolic identitative cognition, which is a symbolic abstract theory of human (and other) thought. Oddly enough, it has nothing to do with computer science. It's probably much more abstract than, say, lambda calculus. So I know a little about the math.

Despite (or because of?) all my efforts to avoid gratuitous concision, you refuse to understand me.

1- Your point —› "First, the total percent of the vote for a candidate doesn't ever go down."

My response —› If enough voters (foolishly) refuse to choose a 1st-rank candidate, these "biggest losers" could get their 2nd-rank votes to suddenly appear as new 1st-rank votes, subtracting from the near-winner's 50% margin. But I suppose you could get around this by requiring the winner's 1st-rank votes to exceed the total number of ballots cast. A bit pointy-headed of me, here.

However, if it makes no difference whether you give a candidate 1st or 2nd rank... then that's weird! It must make some difference, so the voter is still saddled with the perceived viability/ preferability dilemma (the spoiler effect). Unless said voter is willing to allow a Bush to be elected!

Also, why are the "biggest losing voters" given first shot at deciding the election? Why not let all the voters who didn't vote for the one with the most 1st-rank votes also get their 2nd-rank votes transferred into 1st-rank? Ha! Maybe because there might be a new winner of most 1st-rank votes, and then all the ones who didn't vote for this new one with the most 1st-rank votes, would get theirs transferred up (mostly re-transferred!), and this see-sawing could go on almost forever. Which just goes to show how weird this method is.

Also, there could exist three parties whose members would vote for no other party members, so no one could ever get 50+1 percent!

2- Your point —› "You can wait for total counts before doing transfers at all decentralized stations or you can provide the central office the count of all unique ballot rankings and then let the central tabulator perform the rounds, either by hand or by computer, but that too can be done by hand very easily."

My response —› In a real state election with a dozen candidates and millions of voters, your paper-piling method would become absurd.

Also, what is the difference, in terms of keeping local control, between sending in all the ballots, and sending in the information they bear? Also, why would "doing transfers at all decentralized stations" and re-sending the results multiple times be simple, secure and observable?

I find it quite hard to believe that millions of ballots involving dozens of candidates will be counted by hand (using piles?).

3- Your point —› "For example, if you give them a range from 0 to 12 to vote with, is it [padded score voting] any different at all from Range Voting?"

My response —› PSV does not "give them a range from 0 to 12 to vote with." They do not get any "0" vote, they just leave the ballot blank regarding a candidate they reject. And they do not get to vote from 1 to 9 either. They can only give 10, 11, or 12 votes. This provides effectivity of preference that is lacking with the approval method, without confusing choices such as 3, 5, 6, etc. This also makes it possible to count quickly, allowing for hand counted paper ballots. (This is why it is called "padded.")

PSV is effectively a form of range voting. However, Warren D. Smith and Jan Kok, the champions of range voting, seem to emphasize some advantage of "Averaging versus Totaling," but if the votes are "averaged" rather than simply totaled, computers will surely be required. See:

 http://rangevoting.org/Averaging.html

4- Your point —› "Sixth, the idea that we should allow strategic voting is fundamentally an odd one. Social Choice Theory defines strategic voting as the use of insincere voting to maximize sincere outcomes. Any vote system designed to allow insincere voting violates the principle of one person one vote where everybody has an equal power. If voters have to vote against their sincere interests just to exercise power it introduces spoiler effects."

My response —› Social Choice Theory? You are kidding right? Really, the inability to use strategy is a RESULT of the spoiler effect. If you manage a football team and you bet for your opponents, that's "insincere." If manage own a grocery store and you try to low-ball your suppliers on wholesale prices, that is called "business." Politics determines whether we get to eat or not, so it's much more like grocery store management! If the power to utilize strategy is removed from you, you become known as a "slave."

Conclusion —› in real large-scale elections involving big money interests, IRV is strange, incomprehensible, and has a pronounced spoiler effect, but PSV is simple, understandable, easily machine free, and is optimally spoiler effect free.

It's really hard to write about election theory, because one must deal with some math in combination with very complex human relations factors.

You make a statement that is very peculiar and frankly, very incorrect. Perhaps because you deny the need for strategic voting? You just said:

/ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ
Since you used a two party system argument against IRV, it would be useful to point out that all single seat election systems, regardless of counting method, force a two party system upon the electorate. Only Proportional Representation systems prevent that. Approval or your method also create a two party system.
[...]
\____________________

It's simply a fact that all single seat election systems DO NOT "force a two party system upon the electorate."

Padded score voting (PSV) most certainly does not "force a two party system" at all! For example, If you give 12 votes to Nader, and 11 to Gore, if enough voter want Nader he will win, of course. At the same time, you are giving Gore 11/12 (that is 91.6%) as much support as you would have if you had given him 12 votes. So you are not helping Bush and Pat Buchanan (who you are giving no votes) much at all. This could easily break the the two party system. This seems quite clear to me.

I'm a rather formalist person that when I deal with math, and I work from axiomatic systems with well-defined concepts. I understand the attraction to loose language -- it doesn't require any work if your bar for precision doesn't exist. But since there is no precision, it's not very useful.

So when I see you malign axiomatic systems and promote looser systems as more "abstract" (they are not), I have an inner sigh. Being less useful doesn't make something a more distinction-promoting thought (abstraction), it actually is the opposite, it muddies concepts, and prevents abstraction that leads to clarity of thought.

Take your "theory" that has no definition:  http://www.google.com/search?q=%22symbolic+identitative+cognition%22

The only instance it has ever been mentioned is in your one post. Nobody can possibly know what you mean by it. It's not defined, anywhere.

The same was true of your new voting method. Each time you have introduced it you have been imprecise. Each time I have tried to provide a formal description of it you have changed the description to something else. The goal posts essentially move each time I try to compare it to other election systems when you give a different example rather than describe your election system with precision -- a precision that would be required to actually use it in the real world. Perhaps it's not worth it to discuss it until you provide a truly formal description.

As far as your other points, a second place vote with no first place vote is not really a valid IRV election -- as philosophers of science and logic have been pointing out for a hundred years: just because something could be expressed in syntax doesn't make it semantically meaningful -- and for an IRV election it would only make sense interpreted as a shifted vote where your second is first, your third is second, and so on. In fact, when candidates are eliminated, second choice votes become the top vote, and so on during each runoff iteration. You don't just run an election without counting the top vote getters in IRV just because you left out a "first choice vote" thinking it's some sort of special strategy. What the ballot captures is the order of preferences, and order doesn't mean anything other than that certain things are positioned relative to one another on a linear axis, either to one side or another side. The order is the same if two people are in positions one and two versus two and three, respectively.

My point about never going down thus stands. (I'm surprised I have to define such simple concepts as "order", but oh well, formalism has its benefits in that it can deeply explain such things.)

"Biggest losers" do not decide the election. Voters who expressed a first choice vote that wasn't popular get their vote recast for their next place choice. That's what it is doing. The elected candidate must get a majority of unexhausted ballots in order to win an IRV election. That is the criterion. It is true that somebody who isn't the least popular of first choice votes could be eliminated, for example the candidate with the least total votes cast for them with a none of the above ballot. The criterion for elimination is simple and convenient in normal IRV elections and follows the pattern of existing two round systems, but with more rounds, making it an easy transition. But more importantly eliminating the one with the least first choice votes makes it not risky for a more popular mainstream candidate to have competition from a similarly-minded radical candidate. This is the spoiler effect and this risk is exactly what we try to address when proposing a solution to the spoiler problem. IRV thus does the most appropriate action when eliminating candidates to prevent the spoiler effect.

In the case I provide, which is very, very common, under your system, there is an incentive for the mainstream candidate to be voted for by all radical voters. This ensures that the radical voters get no larger set of votes as any non-vote for the mainstream may end up electing the opposite mainstream candidate. This is _exactly_ the spoiler problem. By allowing a voter to actually temporarily withold their vote for the mainstream candidate until their preferred candidate gets eliminated, the better radical candidate simply has a better chance of getting elected versus yours. If no radical candidates can get elected, third parties are effectively shut out of your proposed voting method. This also happens to be another problem with approval/range type election methods as there is no strategy for them to get what they want.

Regarding strategy you are confusing the election concept of strategy with an informal definition of strategy. In my use of "strategy" we are talking about insincere voting to effect a better but not optimum position to avoid greater losses. This is the spoiler problem. That the strategy one must use to defend against the worse outcome never allows the optimum output a chance makes strategy in a non-spoiler-fixing system a very bad thing. You are talking about strategy as the most effective use of of a ballot to get what they want. I want to separate those two concepts, one is negative and the other is positive. Of course you would want to use strategy that gives you the most effect, but you do not want to have to use a strategy that leads to non-optimum outcomes for you.

Now that we've cleared up the mix of concepts of strategy, I hope you can see that it is a bad thing that one may be forced to use a strategy that doesn't optimize ones desired and sincere outcomes. IRV makes the perfectly sincere ballot the strategically best (in your view) in the general case. In your election system the general case requires that you vote in a way that makes all radical candidates never win with a candidate who plays to the center. It would lock in further the two party system that forces the two major parties together and toward the center.

As far as actually fixing the two party system, there is one single-seat election method that does eliminate the two party system because it doesn't actually take preference into account: random selection. But it does this by actually being a multi-seat selection method at the same time (the single and multi-seat form are identical). If preferences are taken into account such that a plurality or majority are required to win the election, a single-seat election system results in a primarily two party system as long as distributions of opinion are relatively homogenized throughout the election area. Only geographic differences in district elections are proportionally voiced. If no geographic differences were proportionally voiced in single-seat district election systems, the entire body would be one party, the one with the most support. That's not a proportional election system. If your vote method were used and all opinions were of equal mixes, the same type of candidate all belonging to the same party, would get elected in every district. That solidifies the two party system, and in fact promotes a one party system, which is honestly what we more or less have aside from geographic differences. Democrats win urban areas, Republicans win rural areas. Due to the geographic distribution, our one party system becomes two party because any party that wins urban areas can be differentiated from a different party representing rural areas (with different issues due to being in a different geography).

Even IRV won't solve that problem unless districts are eliminated and a multi-seat, proportional form such as STV is used. Your proposed method most definitely will not solve that issue, and is worse than IRV for solving spoiler issues, as I demonstrated above.

Given some candidate "X" give either no mention (no vote at all), or ten (10) or eleven (11) or twelve (12) votes. That is the "algorithm" of padded score voting (PSV). It's totally simple. You get to choose the strength of your vote versus the background noise of error or cheating. You can give or not give that to any candidate you choose.

So you can give Nader 12 and Gore 11, for example.

Try to do that with IRV and you could end up electing a Bush.

IRV is very complicated. PSV is simple.

People who don't understand Instant Runoff Voting think it's bad. For more irony: at a guess many of these people object to the two party system...and then tell you to vote for Ron Paul. As if that's going to help....

It would be good if someone could clarify the history of "Portland Green Party". The domain was registered in mid 2010; very little obvious activity since.

I've already explained some reasons why. Fairvote also mentions another reason why here:  http://archive.fairvote.org/irv/approval.htm Campaigns would promote bullet voting to try to ensure their candidates don't get outvoted by somebody working lukewarm positions to try to get backup votes, and it would devolve into plurality voting in that case, which reintroduces the spoiler effect. Your idea that minor changes from 10 to 12 would affect the outcome aside from the spoiler issues is nonsensical and unsupported. Range Voting is the extreme case of your limited case and runs into the same problem -- it devolves into approval voting, and then devolves into plurality voting when campaigns learn how to strategize against it. Fundamentally your proposed system doesn't actually solve any election problem we have.

This is yet another meaningless "argument" from the IRV machine mavens. Voters saddled with IRV attempt to get rid of it a fast as they possibly can, since it always leads to two-party domination. See the example of Australia:

 http://scorevoting.net/AustralianPol.html

Or see:

/ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ
 http://irvbad4nc.blogspot.com/2010/03/burlington-kiss-instant-runoff-voting.html

Tuesday, March 2, 2010
BURLINGTON - 'KISS' INSTANT RUNOFF VOTING GOODBY
Burlington voters voted to ditch IRV today. The measure to repeal instant runoff voting won by a healthy 4% margin. This is a case of a David verses Goliath, grassroots activism verses well funded out of state interests. Congratulations to the grassroots activists.
\____________________

/ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ
 http://instantrunoff.blogspot.com/2009/06/places-that-have-ditched-instant-runoff.html

Monday, June 8, 2009
Places that Have Ditched Instant Runoff Voting or are Moving to Ditch It

THESE JURISDICTIONS ARE MOVING TO DITCH INSTANT RUNOFF VOTING OR HAVE ALREADY DITCHED IT:

*UPDATED ON May 30, 2010*

MOVING TO DITCH, ASPEN COLORADO. [Aspen has now ditched it.]

DITCHED. *Update. Burlington rejects instant runoff voting March 3, 2010 ..."a system which, on paper, persuaded people to give it a try but in reality resulted in a very confusing and poor system." Repeal IRV reports that "2005/2010 Voter disapproval of IRV increased in every ward in the city"

DITCHED. *Update. ON NOV 3, 2009. PIERCE COUNTY WASHINGTON.Majority of Pierce County voters reject Instant Runoff Voting on Nov 3 Instant runoff voting was rejected by an overwhelming majority of Pierce County Washington Voters. 44,145 of 64,106 voters said yes to ditching instant runoff voting, also called ranked choice voting. That is 71.76% for eliminating IRV and 28.24% who wanted to keep IRV. Pierce voters ditch instant runoff voting - save $500K for taxpayers immediately
MOVING TO DITCH. PIERCE COUNTY WASHINGTON. The voters will get to reconsider this issue in November. 63% of Pierce County WA voters don't like Ranked Choice Voting. That is 56,751 out of 90,738 Pierce County voters who answered a questionnaire included with their ballots that asked, "Did you like this new Ranked Choice Voting method?" December 7, 2008 The News Tribute. The county could save $600,000 if they scrapped instant runoff voting now.

DITCHED. BRITISH COLUMBIA (2ND TIME) 61% of the voters gave a thumbs down for STV, Single Transferable Vote, a ranking method in British Columbia. May 12, 2009.

DITCHED. CARY NORTH CAROLINA Cary North Carolina rejected a second go at IRV, voted to keep current election method WRAL News Apr. 30 2009 Cary, N.C. — The Cary Town Council voted against a proposal Thursday to change the current election method. WRAL News and Protect NC Elections Stop IRV Blog . Also see Cary NC tries IRV, then says 'no more'

DITCHED. GEORGETOWN UNIVERSITY. February 21, 2009 Georgetown University ditches Instant Runoff Voting - cites problems The Hoya and No IRV in NC Blog

Instant runoff voting was invented in 1870 by American architect William Robert Ware yet has not been widely adopted. IRV has also been rejected by a few jurisdictions that used it. Perhaps the problem is that IRV is loaded with the potential for perverse outcomes and is difficult to count in a transparent fashion (since it it not additive and votes are redistributed).
\____________________

Here is how I would vote using PSV in a padded score voting election:

Buchanan, Patrick — No vote given
Bush, Jeb — No vote given
Clinton, Hillary — No vote given
Gore, Al — (10) Votes given
Kucinich, Dennis — (12) Votes given
McKinney, Cynthia — (12) Votes given
Nader, Ralph — (12) Votes given
Paul, Ron — No vote given
Romney, Mitt — No vote given
Ventura Jessie — (11) Votes given

These would simply be added up by hand at the polling station. (Note that I can give only no vote, 10 votes, 11 votes, or 12 votes.)

Note that I would give, for example, Ralph Nader, 12 votes. I would know exactly how much support I was giving to each candidate. With IRV, after all the shell game shuffling and vote transferring, I whould have no idea of the effect of my vote. By giving Nader "first rank," I could be helping to elect Jeb Bush!

Now if Ralph Nader were to ask me to "bullet vote" only for him, I would take it that he was asking me to preserve the two-party system, and he would get "No vote given" from me.

Padded score voting is similar in most ways to range voting, but emphasizes simplicity. By removing to "0 to 9" scores ("padding"), it:

1) Is far less confusing to the voters, regardless of how smart they are. (How does one decide to give someone 6 votes, when doing so affects the outcome of an actual election?)

2) The voters must deal with much simpler concepts.

3) It's four times easier to count three possible votes than to count twelve.

So it's just vastly better.

With padded score voting, the two-party system will shrivel and die!

However, the 1% oligarchy much prefers that the 99% remain powerless forever! So they pump vast sums of money into dead ends such as IRV. See their donation pages: