Introducing Democracy Chronicles New Presidential Rank-Balloting Poll

Taroko Karting Winners

IMPORTANT: READ THE VOTING-INSTRUCTIONS BEFORE VOTING BELOW

This is a Presidential rank-balloting poll – a poll in which you rank the Presidential candidates in order of preference. 1st choice, 2nd choice, 3rd choice…etc.

There is another Presidential poll in progress too at Democracy Chronicles.  That poll is an Approval poll, which asks you to mark each candidate whom you like.  The candidate receiving the most marks wins.  If you haven’t voted in the Approval poll, then I invite you to vote in that one too, in addition to voting in this article’s rank-balloting poll.  Also, please see my previous article on Approval voting found here.

Why have this poll, in addition to the other one?  The Approval poll is intended to demonstrate the do-able, feasibly enactable voting-system reform that lets everyone fully support their favorite(s) in addition to any compromise(s) that they feel compelled to support too. Approval is the voting system that has that property, and also qualifies as a serious public proposal, due to its minimal simplicity. But the purpose of this  rank-balloting poll is different.

A ranked ballot takes into account more of your preferences, for the purpose of immediately finding the voter-median candidate, the best candidate whom we all can help to beat everyone who is worse. That’s the ideal of rank balloting, and when counted properly, that ideal is realizable.

Example of Approval Voting Presidential Rank-Balloting Poll

Example of Approval Voting

A rank-balloting poll, such as this one, can inform and guide our voting in a subsequent ordinary “vote-for-1″ official public election, such as the presidential election in November of this year.  That’s the purpose: to show what party, candidate, or candidate-category is the best one that we could help to win, in an ordinary “vote-for-1″ official public election.

Voting instructions will follow, in the next paragraph, but let me say that there is also another article whose purpose is to explain,in more detail, the reasons behind the voting instructions given on this page. That other article will discuss the following matters: The purposes, in more detail, of the Approval poll and this rank-balloting poll; what was learned by the results of the Approval poll.  You can find the other article here.

The other article also includes discussion and explanation of strategy considerations in voting, with a more complete explanation for why certain strategies are not needed in this poll–in other words, why you should rank the candidates completely sincerely in this poll. The other article will also describe the count-method that will be used, for the rankings in this poll, and why that count-method provides the strategy guarantees given here and discussed in the other article.

Voting instructions:

Rank the candidates in order of preference: 1st choice, 2nd choice, 3rd choice, etc.  Post your ballot as a comment.  Do that by writing in the “comment” space, below the candidate list where it says “Leave a Reply”.

How to write your ballot? Simply write the candidates in a vertical list. Your favorite candidate goes on the first line. Your 2nd favorite candidate goes on the 2nd line.  Your 3rd favorite candidate goes on the 3rd line…etc.

But you can rank two candidates equally if you want to, by writing their names on the same line. For instance, if you write two or three candidates on the first line, that means that they are your equally preferred 1st choices. Likewise for any other line. You can write several candidates on any line, such as 2nd or 3rd line, etc., if you want to.  If you rank two or more candidates on the same line, then I suggest separating their names with commas, for clarity. But that isn’t a rule, just a suggestion.

A Different Example of Ranked Voting or Condorcet

A Different Example of Ranked Voting or Condorcet

You don’t have to rank all of the candidates, but I urge you to do so.  Any candidate whom you don’t rank is regarded as ranked below everyone whom you ranked.  You could, but don’t have to, number each line with “1.” , or “2.”, etc.

Now, because sometimes people are inclined to use strategy when they vote, I’d like to describe 3 kinds of strategy that are not needed in this poll.  So, though, I suggest ranking the candidates in your sincere order of preference, I’m specifically telling you some reasons why you have no reason to do otherwise in this poll:

First, just as in the Approval poll, this poll gives you no reason to ever vote anyone else over your favorite. If you like candidate X at least as much as anyone else, or better than all, then there can’t be any reason to not write X’s name on the first line. There is no benefit or additional protection that can be given to a compromise-candidate, by ranking him/her over your favorite.

Additionally, If you’re part of a majority set of voters who all prefer several candidates over the others, it’s important to actually rank all of those several over the others whom you like less. That’s because, with this count rule, all of the candidates preferred by such a majority can lose, if their supporters don’t sincerely rank each other’s candidates over the ones they like less.  In other words, majority coalition support must be mutual–A faction can’t get help from other voters from whom they withhold their sincere support over less liked candidates.

Why do we use a count method with that property? Because, otherwise, with your faction and some other faction having, together,  a majority, you might support their candidate, by ranking him/her 2nd, and thereby giving him/her majority support to beat those you rank lower–but the members of that other faction could decline to help your candidate, to ensure that their candidate would win instead of yours. That means their faction has taken advantage of your faction’s co-operation. It creates a game of “chicken”, a co-operation/defection dilemma, in which the message is, “You help, you lose.”

That won’t work for them, by our count rule. That “defection” by that other faction can only worsen their own candidate’s chances, and elect someone they like less than your candidate.  Our count rule doesn’t have any of that problem. When majority support has to be mutual, the defection referred to above won’t happen.

And that means that you can help that other faction’s candidate, by ranking him/her 2nd, without any fear of defection on their part. There is no game of “chicken”. ”  And that means that you can freely rank the candidates sincerely, without any worry that you’re the supporters of your 2nd choice candidate will defect in the way described above.

Ranked Voting Example Presidential Rank-Balloting Poll

Ranked Voting Example http://upload.wikimedia.org/wikipedia/commons/1/15/Rankballotname2.gif

Lastly, our count method avoids the strategy-need described above. But the truth is that there is no such thing as a voting system or count rule that doesn’t give any strategy incentive. If this were an actual official public election, using our count rule, here is a strategy that you might want to use, even with our count rule.

Suppose there were one or more really unacceptable candidates who might win. Suppose that the merit difference among the “unacceptables” and among the “acceptables” were negligible in comparison to the merit difference “between” the acceptables and the unacceptables. In that case, your best strategy, in an official public election, would be to rank all of the acceptables together in 1st place, by writing their names on the 1st line.

But this isn’t an official public election. As i said, the purpose of this poll is to find the best candidate, party, or candidate-category whom we can help to beat those who are worse.To that end, you should sincerely rank all of the candidates in order of preference.

That’s the way that we recommend that you vote in this poll, because that’s how this poll will provide the information that will tell you something about the winnability of the candidates and parties–and which candidate is the best one whom we can make to beat those who are worse.


Presidential Rank-Ballot Poll:

Rank these candidates, as many as you want to rank, preferably all of them, in sincere order of preference, by writing their names in the comment-space directly below, as described above:

 

Roseanne Barr (Green Party of the United States)

Virgil Goode (Constitution Party)

Gary Johnson (Libertarian Party)

Barack Obama (Democratic Party)

Mitt Romney (Republican Party)

Dr. Jill Stein (Green Party of the United States)

Comments

  1. Adrian Tawfik says

    Here is my vote:

    1. Barack Obama
    2. Dr. Jill Stein
    3. Gary Johnson
    4. Mitt Romney
    5. Virgil Goode, Roseanne Barr

    • Michael Ossipoff says

      Here is smy vote:

      1. Dr. Jill Stein
      2. Roseanne Barr
      3. Barak Obama
      4. Gary Johnson
      5. Virgil Goode
      6. Mitt Romney

      Here are the votes of 2 additional members of my household:

      1. Roseanne Barr
      2. Dr. Jill Stein
      3. Mitt Romney
      4. Barak Obama

      And the other household-member’s vote:

      1. Dr. Jill Stein
      2. Roseanne Barr
      3. Gary Johnson
      4. Barak Obama
      5. Virgil Goode
      6. Mitt Romney

  2. says

    This poll is flawed.

    Michael has chosen to count the results using the “Improved-Condorcet-Top” method, yet this method is obscure and is not a good choice for counting 1-2-3 ballots.

    That method does not appear in the “Voting System” Wikipedia article — which lists the relevant methods supported by election-method experts — and that method is not one of the named methods in the “Declaration of Election-Method Reform Advocates.”

    Michael has a conflict of interest in making this choice. He wants his favorite method, Approval voting, to look good by comparison.

    An additional flaw with this poll is that ballots cannot be cast anonymously.

    • Adrian Tawfik says

      We regrettably could not find a program that could set up a poll using a ranked system using Mr. Ossipoff’s parameters. I can not address your critique of his counting method because I lack the necessary mathematical skill. But I do remind you that all of the information is public and the poll is primarily to demonstrate ranked polling as a political election methodology. Please cast your vote if you please, it is for everyone’s benefit to see the comments you post and, in this instance, your voting preferences. Everyone please see Mr. Fobes articles on election methodology found here or the interview he conducted with Democracy Chronicles here.

  3. Michael Ossipoff says

    A few comments on Richard’s post:

    When choosing ICT, popularity and the wikipedia weren’t a consideration. As I made clear in the articles, I chose ICT for its properties, which I covered in the articles.

    I’ll quote Richard’s post:

    Michael has chosen to count the results using the “Improved-Condorcet-Top” method, yet this method is obscure

    [endquote]

    ICT is new. Chris Benham, a voting-system expert and participant in EM proposed it only a few months ago.

    Richard continues:

    …and is not a good choice for counting 1-2-3 ballots.

    [endquote]

    Regrettably, Richard seems to have forgotten to share with a meri-related reason why he thinks that ICT is not a good choice.

    Richard continues:

    That method does not appear in the “Voting System” Wikipedia article — which lists the relevant methods supported by election-method experts — and that method is not one of the named methods in the “Declaration of Election-Method Reform Advocates.”

    [endquote]

    Copying or following authority means everything to Richard.

    As I said, I chose ICT for its properties. I don’t regard the wikipedia or EM opinion-polling as the way to find the best rank-count method.

    ICT has only first proposed a few months ago. The declaration Richard refers to was written long before that.

    Richard says:

    Michael has a conflict of interest in making this choice. He wants his favorite method, Approval voting, to look good by comparison.

    [endquote]

    I clearly stated why ICT is the best way to count the rankings. Speculation about motive is a sure sign of a complete lack of better arguments.

    There are ways in which Approval does “look good by comparison” to other methods, as I described in my Approval article. Rank-methods are contraptions. None of them can have the solid and elegant simplicity of Approval. But ICT, in addition to Approval’s freedom from favorite-burial incentive, brings an addiitional desirable property, defection-resistance.

    I wanted the rank-balloting poll’s results to not be distorted by strategy. ICT gets rid of the two main kinds of strategy incentive that could adversely affect the results.

    I’m sorry if that bothers Richard enough to make him post about it.

    Richard says:

    An additional flaw with this poll is that ballots cannot be cast anonymously.

    [endquote]

    If anyone wants to vote anonymously, maybe they could just write “Anonymous” in the “Name:” field. Or maybe “John Smith”, or something. Be creative.

    Michael Ossipoff

  4. Ursa Major says

    My vote:

    1. Dr. Jill Stein
    2. Pres. Barack Obama
    3. Roseanne Barr
    6. Gary Johnson
    6. Willard Mitt Romney
    6. Virgil Goode

  5. says

    The most important point about ranked ballots is that they can be counted in many ways. This article fails to even mention this very important point.

    Instead of only counting the results based on the newly proposed method that Michael favors, I propose that the final ballot preferences be posted on the Election-Method forum where the advocates of various counting methods can calculate the results using the other counting methods. Then those results can be compared alongside the new “ITC”-method results.

    The other counting methods include several that have been used successfully to elect officials in small organizations, and they have been carefully analyzed for their fairness characteristics, so they very much deserve to be used in this poll. In contrast, although the newly proposed “ICT” method deserves to be used here as well, it has not yet been used in real voting situations, and it has not yet been mathematically analyzed, so it is still experimental.

    Part of the reason I mentioned the “Declaration of Election-Method Reform Advocates” is that those of us who signed it, including Michael, have agreed to educate people about the multiple voting methods that are supported in that Declaration. In this article and his previous article, Michael has failed to call attention to those other ranked-ballot counting methods.

    I applaud Adrian, the co-editor of Democracy Chronicles, for his part in helping to educate citizens around the world about fairer voting methods. As he points out, there is so much to learn about voting methods that it is difficult for a non-expert to judge the choices Michael has made. This is why I am taking the time to clarify the important point that multiple counting methods are available for counting ranked ballots, which are also called “1-2-3 ballots”.

    As for this poll’s primitive cut-and-paste balloting method, it is inexcusable. There are lots of online resources for handling this kind of ballot. For example, Survey Monkey (dot com) offers a free service that (when I last used it for this purpose) is capable of handling this kind of preference information.

    As a better balloting alternative, I have set up a ballot collection process for this poll at VoteFair.org (one of my websites):

    http://www.VoteFair.org/cgi-bin/votefairgenballot.cgi/votingid=10881-53423-12381

    It demonstrates how easy it is to mark a ballot of this type, and that anonymity is easy to accommodate. It can be used for votes that are not submitted here using in the cut-and-paste method. I’ve already cast my vote there. At the end of this poll of course I will post (on the Election-Method forum) the ballots that are cast there.

    If desired, we can even embed this ballot (the non-Javascript version) into a Democracy Chronicles webpage, for easier access.

    I am in favor of conducting this Presidential poll. Yet I’m in favor of doing it right. A cut-and-paste ballot and only one, dictatorially chosen, calculation method is not appropriate, and it gives many false impressions about ranked ballots.

    This is unfortunate because so many election-method experts regard ranked ballots as the best choice for use in governmental elections.

    Richard Fobes

    • Michael Ossipoff says

      Richard says:

      The most important point about ranked ballots is that they can be counted in many ways. This article fails to even mention this very important point.

      [endquote]

      Actually, that fact is mentioned in the explanatory supporting article that was published with the ballot-introduction/voting-instructions article.

      Perhaps Richard needsd to be more csreful about what he says, or should read what he wants to reply to before he replies.

      As I said there, there are infinitely many ways to count ranked ballots. And that’s the problem, isn’t it. You’ve just been shown a sample of what it’s like, with each EM method-rivalry warrior, angrily be arguing the issue. I haven’t said that Richard shouldn’t use his “VoteFair” at his website. That’s his business, and not my concern.I couldn’t care less. But Richard, like a typical devoted voting-system infighter and flamewarrior, needed to come here to tell us that we shouldh’t be using the rank-count that we’re using. Without giving a reason

      I asked him what, in particular, he thinks is wrong with it. In today’s rant, he seems to have again forgotten to tell us.

      Richard says:

      Instead of only counting the results based on the newly proposed method that Michael favors, I propose that the final ballot preferences be posted on the Election-Method forum where the advocates of various counting methods can calculate the results using the other counting methods.

      [endquote]

      Nonsense. Richard wants me to say, “These rankings will be counted by many various methods; good luck in guessing,then, what your best strategy is, since different count-rules have different strategy.

      No, the only way I could make the strategy-guarantees that I made, would be by guaranteeing a count by a method for which those guarantees can be made. Which part of that does Richard have difficulty with?

      In the two articles accompanying the introduction of the poll, and also in my previous reply to Richard, I stated those guarantees, and explained why ICT was chosen, and why ICT offers those guarantees.

      No one is stopping Richard from copying the rankings from this comment space and posting them to EM. But hopefully, if he does, he won’t post the names of the voters to Em. Let’s hope that he’ll only copy the rankings, and not the names, because I don’t think the participants here want their names posted elsewhere without their permission.

      And then, having posted the rankings to EM, he and anyone else, can count them by any method that they want to. That will be _their_ count, and their results. It won’t be our poll results. Our poll result is the one given by the method whose properties we have guaranteed to our voters.

      Richard says:

      Then those results can be compared alongside the new “ITC”-method results.

      [endquote]

      Compare away, Richard. Maybe, after comparing the results, you’ll be more able than you now are, to tell us why you prefer some other count method to ours.

      Richard says:

      The other counting methods include several that have been used successfully to elect officials in small organizations, and they have been carefully analyzed for their fairness characteristics, so they very much deserve to be used in this poll.

      [endquote]

      Again nonsense. What deserves to be used in our poll is what offers the guarantees that this poll seeks to offer. IRV has been used in many public elections. So has Plurality. That doesn’t mean that we have to use them here, when we want properties that they don’t possess.

      I specifically stated certain strategies by which I didn’t want our result to be distorted. For example, I’ve observed favorite-burial, to protect a lesser-evil compromise, in a Condorcet presidential poll. That isn’t what I want in this poll.

      To anyone voting in the poll: We will count the rankings, as we said, by ICT, and the guarantees that I offered are in effect, due to the use of ICT. Neither of those guarantees would be there in an ordinary Condorcet count, but that doesn’t matter here, because this isn’t an ordinary Condorcet count.

      Richard says:

      In contrast, although the newly proposed “ICT” method deserves to be used here as well, it has not yet been used in real voting situations.

      [endquote]

      The instance of favorite-burial that I observed was in a real poll voting sistuation, like this one.

      I can and did guarantee that certain strategies are not needed in ICT, though they are needed, and have been observed in ordinary Condorcet.

      Richard says:

      , and it has not yet been mathematically analyzed

      [endquote]

      Ridiculous. Specifically, what maathematical analysis is it that Richard thinks that ICT needs, but hasn’t had?

      I wrote two long articles, mostly about ICT’s properties–the properties for which ICT was chosen. If Richard wants to say that mathematical analysis would disfavor ICT, then he needs to specify the mathematical analysis that he thinks would disfavor ICT. Vaguenes and sloppiness seem to be Richard’s style.

      Richard says:

      , so it is still experimental.

      [endquote]

      Certain method properties can be described without experiment. But, as I said, I’ve experimentally obsesrved favorite-burial in ordinary Condorcet.

      Richard says:

      Part of the reason I mentioned the “Declaration of Election-Method Reform Advocates” is that those of us who signed it, including Michael, have agreed to educate people about the multiple voting methods that are supported in that Declaration.

      [endquote]

      I agreed to not oppose them. No, I never agreed to teach and promote all of them. Richard, too, agreed to not oppose them. Perhaps Richard has forgotten that agreement. Condorcet has various versions. The Declaration didn’t specify a particular version. In the accepted usual usage, Condorcet refers to a broad family of methods. Improved Condorcet still comes under the broad umbrella of “Condorcet”.

      II certainly never promised that, if I conducted a poll, I’d describe and count the poll by every method mentioned in the Declaration.

      In any case, the declaration emphasized that it wasn’t saying that its methods weren’t the only good ones.

      By the way the poll took a not-so-objective partisan position about Approval. I wasn’t going to say anything about that, because I wanted to oo-operate and get along, and not fight–a concern not shared by Richard.

      Richard says:

      In this article and his previous article, Michael has failed to call attention to those other ranked-ballot counting methods.

      [endquote]

      Nor do I have any responsibity or obligation to do so. I offer, propose and recommend the simplest, most minimal change to the existing Plurality system–Approval, which nevertheless offers the guarantees described in my Approval article. Richard’s sort will be dithering and battling about which rank-count is best till the cows come home.

      As I said, this poll’s count rule was chosen for the guarantess tha it offers.

      Richard says:

      I applaud Adrian, the co-editor of Democracy Chronicles, for his part in helping to educate citizens around the world about fairer voting methods. As he points out, there is so much to learn about voting methods that it is difficult for a non-expert to judge the choices Michael has made.

      [endquote]

      Actually, I explained, in detail, the reasons for my choice of ICT.

      Richard says;

      This is why I am taking the time to clarify the important point that multiple counting methods are available for counting ranked ballots

      [endquote]

      I clarified that in my explanatory article, and told why I chose ICT.

      Richard says:

      As a better balloting alternative, I have set up a ballot collection process for this poll at VoteFair.org (one of my websites)

      [endquote]

      I wonder if it even occurs to Richard, the impropriety of asking that this poll be moved to his website. Richard, vote if you want to. Don’t vote if you don’t want to.

      Richard thinks that it’s too difficult to write “1.” in front of his 1st cchoice, and “2.” in front of his 2nd choice, etc.
      Richard says:

      It demonstrates how easy it is to mark a ballot of this type

      [endquote]

      See above.

      Richard says:

      , and that anonymity is easy to accommodate.

      [endquote]

      If anonymity isn’t achieved by writing “Anonymous”, or “John Smith” in the “Name:” field, then ask Adrian if you can send your ballot to him, to post for you.

      Richard says:

      I am in favor of conducting this Presidential poll. Yet I’m in favor of doing it right.

      A cut-and-paste ballot and only one, dictatorially chosen, calculation method is not appropriate

      [endquote]

      Richard seems to be saying that we should have first conducted a vote on what count rule to use. For one thing,people would be much less likely to participate in both polls–the one on voting systems, and then the one on presidential candidates Secondly, I doubt that anyone other than Richard would think that it’s dictatorial to want voters to not experience the strategy needs that ordinary Condorcet, or various other methods, would impose. I didn’t just want to conduct a poll. I wanted to conduct a poll whose result wouldn’t be distorted by the familiar, observed strategies that I spoke of. Hence, ICT.

      Richard says:

      , and it gives many false impressions about ranked ballots.

      [endquote]

      Such as?

      Richard says:

      This is unfortunate because so many election-method experts regard ranked ballots as the best choice for use in governmental elections.

      [endquote]

      I wanted to do an Approval poll and an ICT poll.

      A rank method can bring its own properties. Choose the rank method for the properties, the guarantees, tha you want.

      As for governmental elections, Richard is showing us why is why we shouldn’t exptect the enaction of a rank-method for national elections. He and those like him will battle forever, and nothing will happen.

      Michael Ossipoff

      Democracy Chronicles (http://s.tt/1iB8d)

  6. says

    Michael claims that I did not specify which mathematical criteria have not yet been established for the “ICT” counting method. They are listed in the “Voting system” article on Wikipedia, which I referenced. For convenience the relevant fairness criteria are listed here:

    * Majority criterion

    * Mutual majority criterion

    * Condorcet loser criterion

    * Independence of Smith-dominated alternatives

    * Local Independence of Irrelevant Alternatives

    * Independence of Clone Alternatives (Cloneproof)

    * Monotonicity criterion

    * Consistency criterion

    * Participation criterion

    * Reversal symmetry

    * Later-no-harm

    * Later-no-help

    Specifically, Michael has not indicated which of these fairness criteria the ICT counting method meets, and which criteria it fails.

    • Michael Ossipoff says

      Richard says:

      Michael claims that I did not specify which mathematical criteria have not yet been established for the “ICT” counting method. They are listed in the “Voting system” article on Wikipedia, which I referenced. For convenience the relevant fairness criteria are listed here:

      […]

      Specifically, Michael has not indicated which of these fairness criteria the ICT counting method meets, and which criteria it fails.

      [endquote]

      No, I haven’t. Very many criteria have been defined. We choose a voting system because it has certain particular properties that we want.

      To post a list of criteria, and complain that I haven’t evaluated by those criteria–That won’t do.

      I’ve told, in detail the properties for which ICT was chosen.

      If Richard wants to evaluate ICT by some particular criterion, then he needs to tell us why he thinks that his criterion important is more important than the properties that I’ve specified as the reasons for choosing ICT.

      As I’ve repeatedly said, I wanted to conduct a poll in which the voters wouldn’t have the strategy-needs that most voting systems (including ordinary Condorcet and Richard’s “VoteFair”) have. So I chose ICT, and I described its strategy-guarantees to the voters.

      If Richard prefers other properties instead, then I invite him to choose methods that have the properties that he likes, when he conducts a poll.

      If he doesn’t approve of the properties that were chosen for our count, then he’s free to not participate in our poll.

      Polls are conducted by many methods. Does Richard also write to complain to everyone who conducts a poll, to object that he doesn’t approve of their count rule?

      By the way, this behavior on the part of Richard is flagrantly contrary to the spirit and letter of the Declaration agreement.

      Michael Ossipoff

      • says

        Michael says “this behavior on the part of Richard is flagrantly contrary to the spirit and letter of the Declaration agreement.”

        Quite the contrary. My goal in the Democracy Chronicles articles and comments is to educate voters about alternatives to single-mark ballots, without soapboxing a specific method. The whole point of the Declaration is to stop soapboxing specific methods and educate voters about alternatives. (I was one of the two main writers of the Declaration so I certainly know what it says.)

        For clarification, the ballot link I provided provides access to a balloting interface that does not show any results, and after a ballot is cast the only choices are to invite other people to vote and to click a link that returns here to Democracy Chronicles. I now realize that Michael and Adrian wouldn’t know that without having used it to cast a ballot.

        I am not advocating a specific counting method. My goal is to prevent Michael from using his articles to keep blinders on the eyes of voters. We’ve already had too much of that.

  7. Michael Ossipoff says

    Richard says:

    Michael says “this behavior on the part of Richard is flagrantly contrary to the spirit and letter of the Declaration agreement.”

    Quite the contrary. My goal in the Democracy Chronicles articles and comments is to educate voters about alternatives to single-mark ballots, without soapboxing a specific method. The whole point of the Declaration is to stop soapboxing specific methods

    [endquote]

    Nonsense. Most polls are counted by only method. Is Richard saying that his Declaration forbids anyone from conducting a poll counted by one count-rule? Or from advising voters about the strategy-gurarantees of the method used, so tha they won’t feel compelled to unnecessarily use defensive strategies that other methods often compel or give incentive for?

    …or even from telling about some desirable properties for a voting system, and naming a voting system that has those properties?

    If so, then Richard’s notion of his Declaration is starting to sound a bit totalitarian.

    Richard says:

    and educate voters about alternatives.

    [endquote]

    Fine. Do so. No one is stopping you. I supported the Declaration’s position that Plurality should be replaced, and its advocacy of non-antagonism between advocates of different alternatives to Plurality.

    I didn’t promise that everything I say about voting systems will consist of a survey article describing all voting systems, or all proposed voting systems, or all of the ones that you chose to recommend in your Declaration.

    Richard says:

    (I was one of the two main writers of the Declaration so I certainly know what it says.)

    [endquote]

    Then you know that it advocates tolerance among advocates of different alternatives to Plurality, because replacing Plurality is more desirable than attacking eachother’s alternative proposals. One would have expected that you’d have understood that. You said it enough times.

    Richard says:

    For clarification, the ballot link I provided provides access to a balloting interface that does not show any results, and after a ballot is cast the only choices are to invite other people to vote and to click a link that returns here to Democracy Chronicles. I now realize that Michael and Adrian wouldn’t know that without having used it to cast a ballot.

    [endquote]

    If we decide to out-source our balloting to an outside polling-service, there are plenty of impartial polling services that we can choose from. I emphasize the word _impartial_.

    Richard’s conduct here has hardly been such as to assure any confidence in his impartiality or his qualification or suitability to receive and record the rankings for this poll, and then report the results.

    If we decide to outsource our balloting to an outside polling service, it will most certainly not be Riohard.

    Not having previously been conducting a presidential poll, and finding that we have begun a poll, Richard has decided to copy is, and conduct an idential one, with the same candidates. He probably doesn’t understand the impropriety of advertising it in our voting-space, or appointing himself, without permission, as our polling-service.

    In any case, writing a numbered ranking of candidates in this comment-space isn’t difficult, and it is far from certain that we need to out-source the balloting.

    Richard says:

    I am not advocating a specific counting method. My goal is to prevent Michael from using his articles to keep blinders on the eyes of voters. We’ve already had too much of that.

    [endquote]

    So, in Richard’s mind, to conduct a poll counted by some one particular count-rule, chosen for its strategy-guarantees,and to advise voters about those strategy-guarantees, so that they won’t feel compelled or have incentive to strategically mis-represent their preferences as most voting systems often require…would be to try to “keep blinders on the eyes of voters”. :-)

    No one is stopping Richard from writing survey-artiles, or conducting a poll counted by many different count-rules.

    Michael Ossipoff

  8. says

    Michael’s original claim was that ballot preferences needed to be entered as comments because balloting resources were not available online. Now he admits that there are “plenty of impartial polling services that we can choose from.” That was one of my two main points.

    My other main point was that Michael described the “ITC” counting method as if it were the only reasonable way to count ranked ballots. A few days ago Michael published another article (here in the Democracy Chronicles) in which he clarifies that there are other Condorcet-compliant counting methods. Bravo!

    Now I can finally, in the proper context, make the point I would like to have been able to introduce initially.

    In a real election, where there are thousands of voters, the different Condorcet counting methods are all likely to produce the same results.

    The “favorite burial” voting strategy that Michael has explained indeed is an important issue in existing elections that use single-mark ballots, and the Approval voting method indeed basically eliminates the the need to “bury” a favorite candidate. However, Condorcet-compliant counting methods all have about the same amount of resistance to the “favorite burial” strategy.  (Detail: That’s because the Condorcet-compliant methods differ only when there is “circular ambiguity,” and that seldom occurs in political situations.)

  9. Michael Ossipoff says

     Richard says:

    Michael’s original claim was that ballot preferences needed to be entered as comments because balloting resources were not available online.

    [endquote]

    Wrong. At no time did I say that. Adrian said, correctly, that this website’s polling facility doesn’t handle rankings,and so, to conduct a poll _at the website_, it is necessary to write the rankings in the comments space.

    Neither Adrian not I consider comment-space balloting to be a problem. Richard doesn’t like it, and he is free to not vote.

    At no time did I say that there were no online balloting services, and at no time did I oppose using one, if that were Adrian’s choice. But we agree that the comment-space balloting is fine.

    What I _did_ say was that Richard hasn’t shown the objectivity, neurtality and impartiality that is necessary for a balloting-service. That’s why Richard doesn’t qualify to receive, record, store, count, &/or report the rankings.

    By the way, anyone who votes in Richard’s copycat poll, at the URL that he posted in one of his comments here, is invited to also vote in our Democracy Chronicles poll. I emphasize that voting in Richard’s poll, at his website does NOT constitute voting in our poll. The rankings received and reporeted by Richard will not be counted in our poll. The two polls, Richards and ours, are entirely separate. So we invite you to vote in ours, even if you vote in Richard’s too.

    Richard says:

     Now he admits that there are “plenty of impartial polling services that we can choose from.”

    [endquote]

    No, I don’t “admit” that. To admit something is to reluctantly agree with it. There is no reluctance in my agreement that there are plenty of impartial voting services that we could choose from (if we wanted to use one).

    And, as I said, at no time did I say otherwise.

     

    Regarding those impartial balloting services, Richard’s is not one of them, for this poll.

    Richard says:

    My other main point was that Michael described the “ITC” counting method as if it were the only reasonable way to count ranked ballots.

    [endquote]

     

    No, I didn’t. What count method is “reasonable” for someone depends on what properties that person wants. I told why ICT is suitable for our poll.

    Richard can reasonably use one or more different coulnt methods, if he seeks different properties.

    Richard says:

    A few days ago Michael published another article (here in the Democracy Chronicles) in which he clarifies that there are other Condorcet-compliant counting methods. Bravo!

    [endquote]

    I was discussing the use of Condorcet’s Criterion (CC) with different meanings for the verb “beats”. And I discussed the ordinary unimproved Condorcet versions that use a different meaning of “beats” than does ICT.

    Richard says:

    Now I can finally, in the proper context, make the point I would like to have been able to introduce initially.

    In a real election, where there are thousands of voters, the different Condorcet counting methods are all likely to produce the same results.

    [endquote]

    Maybe if they’re all ordinary unimproved Condorcet versions. But if one of them is ICT, and another is an ordinary unimproved Condorcet version, and if there are many equal-top-rankings, then different results are likely.

    Richard says:

    The “favorite burial” voting strategy that Michael has explained indeed is an important issue in existing elections that use single-mark ballots, and the Approval voting method indeed basically eliminates the the need to “bury” a favorite candidate. However, Condorcet-compliant counting methods all have about the same amount of resistance to the “favorite burial” strategy.

    [endquote]

    Incorrect. ICT has absolutely no favorite-burial incentive or need. That is not true of unimproved Condorcet.

    And if Richard says that ICT isn’t “Condorcet compliant”, what he must really mean is that it doesn’t meet a CC that is defined around unimproved Condorcet–a use of CC with unimproved Condorcet’s definition of “beats”.

    And notice Richard’s bizarre reference to “resistance to favorite-burial strategy”–as if favorite-burial strategy were a sneaky way that some people were trying to cheat the system, against whom the system needs to protect itself.

    That reveals a an abyssmal misunderstanding of the favorite-burial incentive problem.

    The problem is that a voter shouldn’t have that need imposed on him/her.

    Richard says:

    (Detail: That’s because the Condorcet-compliant methods differ only when there is “circular ambiguity,” and that seldom occurs in political situations.)

    [endquote]

    Again, Richard is using “Condorcet-compliant” to refer to methods that meet Condorcet’s Criterion when we use the “beats” definition from the definition of unimproved Condorcet. It’s a method-specific criterion compliance. There’s no reason to limit the meaning of “beats” to the unimproved Condorcet definlition of “beats”.

  10. Michael Ossipoff says

    RANK POLL RESULTS SO FAR:

    So far, Dr. Jill Stein is winning in the rank-balloting poll.

    Stein is pairwise-beating Barr and Obama 4 to 1.

    The Approval ballot area, at the left margin of the home screen at this website, reports the current winner-so far, and so far the Approval winner is Obama.

     

     

     

     

     

  11. augustin says

    I recently realised that I had forgotten to cast my own ballot in this poll.

    Here it is:

    1. Roseanne Barr (Green Party of the United States)
    2. Dr. Jill Stein (Green Party of the United States)
    3. Barack Obama (Democratic Party)
    4. Gary Johnson (Libertarian Party)
    5. Virgil Goode (Constitution Party)
    6. Mitt Romney (Republican Party)

  12. Michael Ossipoff says

    Augustin–

    Thanks for your ballot. Roseanne Barr is my favorite too, due to her outspoken forthrightness.

    I report the winner once in a while, usually when a new ballot is cast.

    Stein is still the winner, but now she isn’t beating Barr by as much as she was before.

    Though my top favorite isn’t winning, it’s satisfactory that at least someone with significantly better-than-Democrat policy proposals is winning anyway.

    Michael Ossipoff

Leave a Reply

Your email address will not be published. Required fields are marked *

Current ye@r *