Jump to content

Talk:Electoral system/Archive 2

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1Archive 2Archive 3Archive 4Archive 5

Impossible

Surely to God the simplest imaginable system in the world has at least been considered before: everyone votes into one big pool and parties are given a percentage of seats as precisely equal as possible to their popularity. That's it. No need for fancy calculations, wierd voting sheets or anything. Check a box and toss it in, and the resulting percentage is equivalent to what people get in Parliament/whatever. It's beyond me why this isn't discussed here, let alone implemented anywhere in the world. As for choosing what individual gets the seat, that's another story and not entirely relevant to the actual way the vote takes place, other than whether people vote for a party or a face. Party lists, perhaps; that's not the point, though. The point is that I can't imagine a simpler system (not necessarily better, but simpler) and yet it's not discussed, not even a tiny blurb. It MUST have occured to someone - countless someones, even. Anybody?

Perhaps you are talking about Proportional representation? It is mentioned in the article, you must have missed it. Paladinwannabe2 20:32, 22 October 2007 (UTC)

The path toward a featured article

We certainly have collected the sheer amount of information, by now, that it would take to make this article into a featured article. I'd like to fix the presentation of that information so that FA status is within reach.

I've just made a major revision of the article, which I've been working on over the last couple of weeks. The biggest changes involved turning lists into paragraphs wherever appropriate, rearranging sections, and providing a gentler introduction for non-theorists. For example, the article did not begin with an explanation of why we study this crazy morass of voting systems, but it did begin with a detailed discussion of write-in ballots, "none of the above" options, and proportional representation. In my edit, the first two are briefly mentioned with links to their own articles (which already existed), and proportional representation is now under the "Multiple-winner methods" section where it belongs.

In turning lists into paragraphs, I needed to briefly characterize every method mentioned on the page. This involved describing methods I'm not very familiar with, like highest averages methods and cumulative voting. Please check that my descriptions are accurate - I'm not trying to add misinformation to the article or anything. Be sure to keep descriptions brief if you modify them; after all, we're listing a lot of methods, and they all have their own page where they can be described more thoroughly.

There are more steps before we can propose this as a featured article. To begin with, there are more revisions to the article that I don't yet have the time or knowledge to write, so I would (of course) appreciate other people's contributions:

  • Though I rewrote the "famous voting theoreticians" section slightly, I think what we really need is a "History of voting theory" section that would engulf that section, describing major works in voting theory (and their creators) in chronological order. This section should be in paragraphs of prose, not a list.
  • We should have a table for multiple-winner criteria. It should leave out criteria that don't apply (like Condorcet and Majority, which assume a single winner), and include proportional representation (using a suitably strict definition of it).
  • The "External links" section has grown too large. We should find a way to pare it down. For one thing, links that are about a specific aspect of voting theory, instead of covering voting theory as a whole, can be moved to other articles.

And then there are the actual requirements for featured article status:

  1. We should find an appropriate, free (as opposed to fair-use) picture to go at the top of the article, and perhaps more for later sections of the article.
  2. We should make a list of references for statements made in the article. This may be the hardest part; I imagine that most of us are writing based on received knowledge from discussions, classes, or mailing lists, and now we have to track down the original sources.
  3. Then, we submit the article to Peer review and make revisions that they suggest.
  4. Finally, we submit the article to Wikipedia:Featured article candidates.

Who wants to help? RSpeer 18:44, 3 October 2005 (UTC)

I can probably find a picture that's in the public domain. Unfortunately, from the source I'm thinking of (Spencer's American Ballot) the picture would likely be of an old ballot design (there's one with the socialist party, the repubs, dems and such that is particularly neat). Let me know if this would be appropriate. Joseph Lorenzo Hall 20:02, 3 October 2005 (UTC)

That sounds like a good picture - I don't see why it's unfortunate. Sure, it doesn't indicate anything about the mathematics going on, but I can't imagine any comprehensible picture would. And it's somewhat more specific to voting theory than the generic picture of people lining up at a voting booth that's on articles like Vote. RSpeer 21:43, 3 October 2005 (UTC)

OK. You're right, I can't think of a good image for this page and a neat ballot might do the trick in terms of making it more inviting. I'll scan it in... it might take me a week or so (I'm about to go on a short business trip). Joseph Lorenzo Hall 22:30, 4 October 2005 (UTC)

Have you got the picture yet? RSpeer 15:33, 14 October 2005 (UTC)

Restructuring Voting System Articles

I'm an avid reader of the voting system articles, and I decided to join this discussion because I felt the voting system pages could be restructured for greater simplicity and clarity. First, I feel the pages as they are currently structured require too much information to be duplicated throughout. In particular, the comparisons between Voting System A and Voting System B need to appear in both article on A and the article on B. A three-way comparison of voting systems should appear on all three pages. This duplication means a lot of work for the maintainers of the pages, and inevitably results in asymmetric information because the duplication is hard to do perfectly. Also, the reader who would like to read about comparisons all single-winner voting systems, for instance, needs to visit every single-winner voting system page to get the full story. Lastly, the fact that the pages on the voting systems include assessments of them makes them susceptible to religious wars between voting system advocates and invite the "neutrality questioned" stamp.

I would like to see all the articles on individual voting systems be trimmed down to include the undisputed facts of the system: basically just who invented it, it's mechanics, and where it's used. Then the individual voting system articles can point to a separate article that offers the various arguments and assessments of all the voting systems. So for instance, the FPTP article, the Approval Voting article, an IRV article, etc, would no longer include assessments of their individual systems; instead they would all link to a single article called something like "Single Winner Reform" where all the comparisons, assessments, and arguments reside.

First, this would reduce the maintenance load of Wikipedia writers and maintainers and free them of the need to duplicate all the voting system comparisons and assessments. Second, it gives those interested in comparison voting systems a one-stop-shop for information. Third, it unclutters the individual voting systems pages for those looking for just a simple explanation of how the system works. Lastly, it takes the individual pages out of the contentious realm of assessment and gives us a central location for all the debates to go down.

I apologize if this seems presumptuous of me to not contribute and then all of a sudden suggest a substantial revision, but I do care about the quality of these Wikipedia pages. My desire to revise these articles is what prompted me to create a Wikipedia login. Boot1780 01:57, 25 October 2005 (UTC)

I agree with you in many ways; most voting system articles are horrible messes of POV arguments, and badly need to be fixed.
I think each voting system article should include more than what you've listed, though; they should at least list criteria passed and failed, because criteria are the closest thing we have to objective comparisons of voting methods. I also think that the major arguments for and against the system should be included, because the article would be incomplete without them. But what should be left out is direct comparisons with particular other methods. So the IRV article should say why IRV is promoted and why it is criticized, but not, for instance, why Approval is better.
I don't think moving assessments of the systems to a merged page would help. It might prevent POV wars on the systems' own pages, but the merged page would consistently be an unreadable mess of POV. That's just moving the problem.
So, basically, I would propose a less drastic change, which is to standardize the discussion of criteria passed and failed for each method, and remove unnecessary comparisons to other methods. RSpeer 19:14, 25 October 2005 (UTC)

Good comments, thanks. I am still doubtful of the ability to just include "why system X is promoted" and "why X is criticized" arguments for each article on voting system X without inviting endless comparisons to other systems. The reason these articles are cluttered with the "horrible messes of POV arguments" is that the arguments over voting systems don't fit nicely into "pro" and "con" categories. Other political topics are mostly binary choices, like abortion (pro-choice vs pro-life), war (pro-war vs anti-war), etc; for these it makes sense to include the pro- and con- arguments in the article itself. But with voting systems, there are many, many alternatives, and it is basically impossible to say why a particular system is good without saying why it's better than some alternative.

Just including which criteria are passed and failed is plausible. But then again, a few of the criteria themselves are debatable. Also, they are just mathematical criteria and don't address certain important social/political aspects. So including them runs the risk of implying that these are THE criteria. What if we moved the voting criteria matrix from this article to the voting system criterion article and point this article and all the voting system articles to the voting system criterion page.

Sure, it gets difficult to stick to just pro- and con- arguments, but what's the alternative? A page that does the comparisons that evaluate each system with regard to the others would be unmanageable.
Also, we kind of do need to decide on THE criteria, or else articles can be made POV by selectively including criteria. See the discussion at /Included methods and criteria, which unfortunately has had less effect on the actual content of voting system articles than I would have liked.
The idea of moving criteria and the table from voting system to voting system criterion (which is currently kind of redundant and neglected) does sound reasonable, though, and it would help cut down the length of this page. What do other people think? RSpeer 04:15, 26 October 2005 (UTC)

You're right, we should include the pro- and con- arguments for each system, but try not to make them comparative. As you agreed with, I would move the criteria matrix in [[voting system] to voting system criterion, and I would move (or copy) the introductory text to the matrix as well. Then I would link each individual voting system page to voting system criterion. However, I would have each link state that these are mathematical criteria. The voting system criterion page itself should especially make clear that they are criteria of quantifiable features of the voting system when it is treated purely as a mathematical algorithm. They criteria can't, for instance, measure the inherent simplicity, intuitiveness, or administrative cost of any system. Boot1780 04:48, 28 October 2005 (UTC)

Voter Satisfaction

Explaining my revert: I'm not a voting theorist, but am a mathematician, and it is my understanding that the definition of the term "voting system" has nothing to do with voter satisfaction. The definition of what constitutes a voting system is given in the first sentence of the second paragraph of this fine article:

Specifically, a voting system is a well-defined method (an algorithm) that determines a winning result given a set of votes.

This makes no restrictions on voter satisfaction, although I admit that real-world voting systems always try to satisfy voters. For instance, "Minority rule," where the candidate with the fewest number of votes wins, is a voting system. It is a voting system because it is an algorithm for compiling votes into a decision. Most people will intuitively judge that it is "not fair", and we can discuss its failure to satisfy many important criteria, but it's still a voting system. Please correct me if I'm wrong. --Staecker 21:48, 26 October 2005 (UTC)

Actually, you're right. What I was trying to do was ensure that the article doesn't come across as just some inscrutable mathematical topic, but as something that relates to real voting and group decisions, to someone reading the lead of the article. But I suppose this needs to be done some other way, not in the definition. RSpeer 19:09, 28 October 2005 (UTC)

Finally

After a good month and a half, I've made a pass through rewriting the article, including writing the new history section to replace the list of voting theorists. Any comments on style or content? Are we ready to send this article to peer review?

rspeer 23:48, 9 November 2005 (UTC)

We have too many links. I'm taking a hatchet to the external links section. Links that are too specific or too general for the topic of voting theory are going, and this includes advocacy articles with no particular significance. Here are the links I'm removing:

Sorry if I've offended any contributors, but this needs to be done. rspeer 01:36, 10 November 2005 (UTC)

Peer review

Okay, it looks like the article's stable and ready for peer review. Here goes! rspeer 22:42, 11 November 2005 (UTC)

Tie-breaking votes

2.2 Voting Power: "Voting power can also be distributed unequally for other reasons, such as increasing the voting power of higher-ranked members of an organization. A special case of this is a tie-breaking vote, a privilege given to one voter to resolve what would otherwise be a tie."

How is it that a tie-breaking vote increases the voting power of a single member? A member who only votes to break a tie actually has less voting power than any other member as he does not have the power to make a tie by adding one vote to the negative side, causing a motion to fail (since it didn't achieve a majority). This is why Robert's Rules of Order specifies that a society's chair (if he is a member) may vote to either make or break a tie, giving him a vote equal to that of every other member. In other words, he can vote whenever his vote will affect the outcome.

Adam Konner 01:21, 15 December 2005 (UTC)

I didn't say that the tie-breaking vote had more power, actually. But I guess I'll go clarify that. rspeer 01:28, 15 December 2005 (UTC)

Voting Power

First, in response to the previous comment, giving someone a tiebreaking vote does increase their voting power if they also have a normal vote.

In the voting power section, it only refers to weighted voting systems, not the actual concept of voting power. Essentially, voting power refers to the probability that one can have a decisive vote. That is, the probability that candidate A will win if one votes for candidate A minus the probability that candidate A will win if one votes for candidate B. May I suggest looking at the article "The Mathematics and Statistics of Voting Power" by Gelman, Katz, and Tuerlinckx for further analysis?

Also, overall, fantastic job with the article.

Hmm, you're right - the term I picked does clash with that existing use of "voting power", which incidentally should have a Wikipedia article if it doesn't already.
Is there a standard term for what I'm describing? Does "voting weight" work? rspeer / ɹəədsɹ 05:12, 31 December 2005 (UTC)

Yes, voting weight would work fine.

Venetians

I'll leave it to someone more knowledgeable than I to integrate the information in this passage from Julian Norwich's book into this discussion:

On the day appointed for the election, the youngest member of the Signoria was to pray in St. Mark's; then, on leaving the Basilica, he was to stop the first boy he met and take him to the Doges' Palace, where the Great Council, minus those of its members who were under thirty, was to be in full session. This boy, known as the ballotino, would have the duty of picking the slips of paper from the urn during the drawing of lots. By the first of such lots, the Council chose thirty of their own number. The second was used to reduce the thirty to nine, and the nine would then vote for forty, each of whom was to receive at least seven nominations. The forty would then be reduced, again by lot, to twelve, whose task was to vote for twenty-five, of whom each this time required nine votes. The twenty-five were in turn reduced to another nine; the nine voted for forty-five, with a minimum of seven votes each, and from these the ballotino picked out the names of eleven. The eleven then voted for forty-one -- nine or more votes each -- and it was these forty-one who were to elect the Doge. They first attended Mass, and individually swore an oath that they would act honestly and uprightly, for the good of the Republic. They were then locked in secret conclave in the Palace, cut off from all contact or communication with the outside world and guarded by a special force of sailors, day and night, until their work was done.

So much for the preliminaries; now the election itself could begin. . . .

Norwich, A History of Venice at p. 166

--Italo Svevo

German election (2005)

The number of seats of the CDU is wrong (CDU is the first german party). 81.51.105.128 15:46, 6 January 2006 (UTC)


  • Is the German election an example of proportional representation? I understand the system to be a half and half. Half FPTP and half PR.
  • You are right, it is half and half, plus there is the problem with the status of the csu. I suggest to replace the image with one of a country not only with proportional voting but also without election threshold, provided that there are countries that still do not have one. If there is none, I would suggest an image of one of the Reichstag elections in the Weimar Republic.

Image Names

I suggest that the name below the checked box be changed to "Pat Buchanan." Jburt1 19:03, 6 January 2006 (UTC)

Fred Rubble?!

That is much too close to being a Flintstones referance for comfort.

Nice job

Nice job, folks, this is a great article. --Doradus 21:32, 6 January 2006 (UTC)

The single-winner revival ??

Why was it then that only and mostly the anglo-american nations continued with maximum unproportional FPTP?? Seen from the PR and coalition-government point of view the late 1800s was when the "single-winner" system was abandoned(except for the presidential institution, although that too was suggested as 2-party, president and vice-president, plus tribunal, etc systems??)

Skipping the presidential institution (run-offs), the number of "single-winner" systems seem to diminish every year, these days?? Even UK is struggling to conform with EU?? Most (european) medieval systems also had something like 3-4 parties??

(single-winner) president, governing majority (or minority), representation, coalitions

I believe these actual outcomes of, the (constitutional) balance between them, of elections should be sorted out before (or at least mentioned when) one starts describing voting systems?? Single or double-houses, presidents or prime-ministers, veto-powers and the whole shabang.. (Note, the US and UK systems are anyway very "extreme" in the world??)

No, because voting systems are not necessarily used for any such purpose. The article already gives political elections as an example of their use. KVenzke 04:12, 9 January 2006 (UTC)

Voting system criteria

The New Zealand Royal Commission on the Electoral System identified 10 criteria for judging an electoral system. I feel that they should be included in this article, as they are pretty good and have actually had relevance in changing an electoral system. They are

  1. Fairness between political parties
  2. Effective Representation on minority of special interest groups
  3. Effective Maori representation (note, this could be changed into indigenous or regional representation, whatever needs to be highlighted in that country)
  4. Political integration
  5. Effective representation of constituents
  6. Effective voter participation
  7. Effective government
  8. Effective Parliament
  9. Effective parties
  10. Legitimacy

I can expand on each of them as I have a copy of the Report, and they are a lot more self explanatory anyway. I want to know what others think of this though before I edit away though. --LeftyG 03:52, 24 January 2006 (UTC)

It's incorrect to call these things voting system criteria, at least as that term is used in the literature and on Wikipedia, since the things the commission presents aren't precisely defined. That presents inclusion with a potential POV problem - the royal commission's view on what constitutes "legitimacy" or "fairness", for instance, could be quite different from other notions. The current article does well at summarizing the objective mathematical voting system criteria (for which we have an article, and on each particular article the points of view about its relative importance are addressed), however mention is also given to less precisely defined things, which probably closely approximate what the New Zealand commission is getting at. Scott Ritchie 05:25, 24 January 2006 (UTC)

The Royal Commission refers to them as "Criteria for Judging Electoral Systems". While your criticism is fair, I would like to reaffirm that these were the criteria used to judge New Zealand's change of electoral system, and thus the POV problem is lessened/removed (as they have actually been used). Prehaps after/instead of/part of

In addition to the above criteria, voting systems are also judged with criteria that are not mathematically precise but are still important, such as simplicity, speed of vote-counting, the potential for fraud or disputed results, the opportunity for tactical voting or strategic nomination, and, for multiple-winner methods, the degree of proportionality produced.

Would that be better? --LeftyG 06:40, 24 January 2006 (UTC)

That was more or less what I had in mind. Something like "The New Zealand Royal Commission on the Electoral System, for instance, listed ten criteria for judging potential voting systems, including (a few examples from the list.)" Scott Ritchie 07:33, 24 January 2006 (UTC)

I'll work on it tomorrow. --LeftyG 08:25, 24 January 2006 (UTC)

If you want to write more than a sentence about electoral system criteria (as opposed to voting system criteria) you could create Electoral system, which currently just redirects to Election. Judging from the number of inbound links it could become a useful article. Pm67nz 11:31, 24 January 2006 (UTC)
I concur. Write that article. ;) —Nightstallion (?) 12:16, 24 January 2006 (UTC)

Wouldn't it actually make more sense for Electoral system to redirect to Voting system than Election in the first place? I'll add a sentence here, as currently I'm trying to bulk up the Royal Commission on the Electoral System page. But maybe in the future... --LeftyG 22:15, 24 January 2006 (UTC)

Re redirecting here, please no, that is precisely my point, that "electoral system" is not synonymous with "voting system". An electoral system is a description of how an election is to be conducted, and is generally implemented as a set of laws, while a voting system is an algorithm and thus implementable as a computer program. The choice of voting system(s) to be used is only one aspect of an electoral system. Most of the links to electoral system are from "Elections in X" articles. In that context messy real world election issues like electorate size, party list thresholds and parallel voting are of more direct interest than the subject matter of this page, which (rightly) includes some mathematically elegant methods which never been used in any election. Pm67nz 09:32, 25 January 2006 (UTC)


Social Choice Function also goes straight here, which is strange since this is an example of one but not the only possible one, and there isn't a single mention of what one is in the article.

Adding tactics to criteria table

I would like to humbly suggest that the by-system analysis on Tactical voting:Tactical voting in particular systems is far enough along and NPOV enough to be summarized and linked in the criteria-satisfaction table in an UNCOLORED column. I understand that it's not, mathematically-speaking, a criterion, but in the context of this article I think it qualifies.--Homunq 14:24, 13 August 2006 (UTC)

Also, I understand that neither consistency nor participation strictly implies the other (this took me a while - it's because the "participation" voters don't count as a "consistency" group because they may be considering non-"participation" over something other than their first choice), but still, they are very similar concepts, and in this particular table the columns match down the line. I suggest they be merged. There are plenty of distinct well-defined criteria out there, let's avoid any suggestion of "stacking the deck". --Homunq 20:46, 13 August 2006 (UTC)

I just did the second suggestion, since nobody has responded above and it seems reasonable to me. Anyone who disagrees, feel free to revert, but please at least respond to both suggestions here with reasons. --Homunq 18:49, 16 August 2006 (UTC)

It's a great article, but it only begins to explain why any of these alternate voting systems should even be considered for use.

I am restating the last (reverted) edit. Some towns and other small governments are introducing IRV and, even with that, people ask routinely, "why bother?" In the Intro to the article there needs to have some conceptual justification as to why, in plain language, a voting system like IPV or Condorcet would be desirable and that language is not there. It's a great article (I first noticed it when it was featured) and it appears to me to be very scholarly and well researched. But, for an encyclopedia, it misses the point if anyone reading it sees it as just a bunch of academic fancy. There are important practical and idealistic reasons for governments to be considering different alternative voting systems. And this article says nothing clear about those reasons. r b-j 17:43, 6 February 2006 (UTC)

It sounds like you intend to modify this article so that it advocates electoral reform. Don't do that; it's a POV. We might all like electoral reform, but we can't let that show through in the article.
So I'm reverting your edit now. I don't think there's much to "give a try" there; the section you added interferes with the flow of the article, and it's US-centric, somewhat off-topic, and verging on POV. It definitely doesn't belong in a featured article. You could add something similar as a "United States" section under Electoral reform instead. Let this page just describe the systems, and not advocate any. rspeer / ɹəədsɹ 21:47, 6 February 2006 (UTC)
Or a section on Electoral reform in the United States might be an idea to create. --GeLuxe 22:50, 6 February 2006 (UTC)

You guys don't get it. If there was election reform somewhere (and making election reform is a whole different topic), can you give your readers a hint as to why they would consider "reforming" their election rules to use IRV or Condorset (also known as "True Majority" in some other articles)? What do you say to the person who says (ignorantly), "We don't need any reform, we have a fair system, we have majority rule, the candidate with the most votes wins, what does this 'new fangled' system do other than confuse us?" The only statement made in motivation or justification to any of these voting systems is "The goal of most voting systems is to give a sufficiently fair way to choose the winner in such a situation. Different voting systems arise from different approaches to this goal and their behaviour can vary wildly." That is practically a tautology. It says practically nothing. You offer very little, in accessable language, to clue in a reader why there is even a problem with the basic "majority rule" systems in use nearly everywhere.

In terms of "POV", forgive the US-centricism but I am not familiar with elections in other countries where it was as clear of what a 3rd candidate can do to swing an election. (I'm aware that some European countries have had many competive parties and sometime had trouble forming a government as a result.) But the historical cases of Ross Perot in 1992 and Ralph Nader in 2000 are two well known examples, affecting different major parties adversely (that should even out the POV a little), for what can happen when a clone candidate gets in, doesn't get any farther than 3rd place, but manages to change the outcome of the close election from the candidate more like the clone (who was ahead without the clone) and toward the candidate that would have otherwise come in 2nd. That is the whole motivation for why these guys like Condorcet and Shultze are thinking about these vote counting algorithms in the first place. r b-j 01:01, 7 February 2006 (UTC)

"can you give your readers a hint as to why they would consider "reforming" their election rules to use IRV or Condorset[?]" No, that is putting across a point of view that that system is "better". Having rational arguments for and against each system is acceptable, as long it is balanced. And this should occur at the systems page, not on the general page. —The preceding unsigned comment was added by Gregstephens (talk • contribs) .
It is putting forth a plausible reason for why anyone is spending time thinking about and designing such voting systems. If there is no context where something else might be "better", it is inexplicable as to why bother with this "academic crap" (not really my words) in the first place. r b-j 04:43, 7 February 2006 (UTC)
rbj: Your use of "majority rule" is strange; you're using it to mean "plurality". Are you confusing the terms, or are you attributing it to the average ignorant person? It may be reasonable to express more specifically that "majority rule" does not imply the plurality system. And I concur with Gregstephens: there should not be an argument against plurality on this page. rspeer / ɹəədsɹ 03:35, 7 February 2006 (UTC)
Yes, I am attributing that to what might be the "average ignorant person". You wouldn't believe people's opposition to considering these solutions. It's almost as if these people think that forcing voters into tactical voting is a good thing. In the U.S., both of the major parties can look to examples where their own candidate had votes "stolen" by another candidate farther out on the fringe of the dominant axis of the political spectrum than their more centrist candidate. Forcing voters to consider voting for their 2nd favorite choice instead of the candidate they most support, out of fear of giving the election to their least favored candidate is the whole reason that systems like IRV and True Majority (what the article calls "Condorcet") have something to offer. That basic and salient fact is barely touched in the article. You guys go into all this nice detail (it reallly is a good article) about these methods without ever clearly telling your reader why bother thinking about such methods. That gap I want to see filled, even if it isn't with my words.
You barely touch it with the paragraph beginning "If every election had only two choices,...", but you don't make it clear. For an encyclopedia that is not desireable. Perhaps for a graduate school text, but not an encyclopedia. I realize that I was elaborating on the last two of the "criteria that are accepted and considered to be desirable by many voting theorists" that are listed and then put across a table for the different methods. Now, do you honestly believe that for the "average ignorant" reader, that they understand Independence of irrelevant alternatives and Independence of clone candidates to be the reasons they were thinking of switching their vote from Ralph Nader to Al Gore? That impetus, the desirability "by many voting theorists" of these criteria is not at all clear to the "average ignorant" reader. You need to state clearly what possible problem might be dealt with in the promise of IRV or Condorcet or Shultz or whomever/whatever method. r b-j 04:43, 7 February 2006 (UTC)
I'll agree to clarifying the difference between majority and plurality, as you said, and I suppose also to clarifying that IIA and ICC deal with the "spoiler effect". This could even be done with the Gore/Nader/Bush example. But it is not the job of this article to dispel opposition to electoral reform or to discuss the "promise" of IRV or Condorcet. That's a POV, and that you feel strongly that it's the right POV doesn't make it acceptable. rspeer / ɹəədsɹ 05:19, 7 February 2006 (UTC)
It is not the issue to clarify the difference between majority and plurality. It is not POV. You keep saying that without any justification. It is explanation. You fail to clue your clueless reader in as to why are we even talking about all these complicated voting methods when our ostensible "majority rule" system is the fair system. Some people do not realize that they are not getting their majority rule, and this article should give them a hint as to what circumstances have "deprived" them of that. Take a look at this section: Instant-runoff_voting#Adoption_in_the_United_States. That has some explanation as to what motivates IRV. What is your objection to including a similar explanation to what motivates the consideration of different voting systems? It is incomprehesible to me what this objection is about. I didn't write any of Instant-runoff_voting but the article would certainly be less explanatory if that historical example regarding the 2000 presidential election was left out. All's they're doing is explaining why anybody would bother considering IRV. They're not plugging IRV, just saying what has happened in some cases when IRV was not available (that voters were forced to consider insincere selections for fear of electing someone even worse than their 2nd choice). This article needs an explanation of that clarity. r b-j 05:35, 7 February 2006 (UTC)
And that's a great place for such a statement to be. It's relevant to compare IRV to Plurality there because "adoption of IRV in the United States" is about IRV versus plurality. This article, though, should not stop to point out why each method avoids pitfalls of plurality. rspeer / ɹəədsɹ 23:48, 7 February 2006 (UTC)
But it should finish the thought in that paragraph starting with "If every election had only two choices,..." and it does not. Not clearly, anyway. It should point out, whether this is good or bad, that election systems that are common (either plurality or majority without any ranking) what consequences are possible without ranking, which is what motivates researching systems using ranking. That is clear statement of motivation is not there and it needs to be plainly there. Otherwise there is an assumption of motivation that was not yet established with the reader. r b-j 04:07, 8 February 2006 (UTC)
I just edited the page to try to address this without stepping on any toes. I added the sentence "Thus, the choice of voting system(s) is one important component of a democratic government." to the paragraph in question. This is a conservative edit, I'd prefer to go farther but this is a featured page. Please comment. --Homunq 14:13, 13 August 2006 (UTC)

no, no Scott. what got truncated was far worse [4]. the extra "see" line was due to Rspeer's last revert. r b-j 05:49, 7 February 2006 (UTC)

Yeah, I noticed this. Not sure when it happened in the edit history though, but I think I fixed it with my last edit (which for some reason doesn't seem to be showing up in history) Scott Ritchie 08:22, 7 February 2006 (UTC)

No mention of the Web

This article makes no mention of voting systems as implemented on the World Wide Web. The Web offers unique opportunities. For example, the vote options can be suggested by the community rather than be imposed by the elite. Once everyone is happy with the various options, they can vote. There are many examples around the world of referenda skewed to avoid a particular outcome. By involving the voters in canvassing the vote options, that level of manipulation disappears. --JG Estiot 23:43, 11 February 2006 (UTC)

  • Interesting, but that's part of the larger election, not the voting system. A voting system doesn't specify where the ballot options come from. It may be something that could be said at the Voting or Election article. Do you have a source showing that voting online has notably different effects than voting offline? rspeer / ɹəədsɹ 15:07, 24 February 2006 (UTC)

Without time no morality

Will add my analysis and suggestion of the Borda fixed point method. Colignatus 16:19, 5 March 2006 (UTC)

  • Moved to the bottom. And I think that article should be deleted, as it is clearly your original research. rspeer / ɹəədsɹ 16:38, 5 March 2006 (UTC)
  • Thanks for helping to find the right place for this. Well, it is my research, but it has been published, first as a working paper on EconWPA, then in a book at Dutch University Press, and now also at Project Gutenberg, so it is not 'original' in the somewhat unfortunate use of terms at the page that you referred to. Please note that I have hesitated to include my own work, since I feel that I should not advertise my work (though I make it available on my website), but I was encouraged by the fact that another user of wikipedia originated the page on the Borda fixed point (referring to my work). Colignatus 16:59, 5 March 2006 (UTC)
  • OK, what will happen now ? I noted that someone has suggested to delete the page on Borda fixed point. I removed that again since I suspect that it is you while I consider your reaction premature. Would it be possible for you (a) to read the text, (b) study the references when in doubt, (c) ask me questions if still in doubt, (d) only take any action after this, (e) and can you explain to me what is going in the mean time now that someone put that message there ? Please note that I wrote a book on voting theory so that I know what you have done, writing these pages. It is just that it is unsettling for a scientist like me to be suddenly confronted with 'deletion signs' that are wholly irrational. Colignatus 17:23, 5 March 2006 (UTC)
Sorry, I was hasty about that. I hadn't checked your credentials; the fact that the article referred to you by your Wikipedia username was a red flag for me, as it sounded like the usual case of some random person thinking their random ideas deserve an article.
The current version of Borda fixed point shows the problem with writing about your own work on Wikipedia - it is not at all written from the neutral point of view, as it simply advocates the method. I'll discuss more on Talk:Borda fixed point. rspeer / ɹəədsɹ 16:06, 6 March 2006 (UTC)
That the article only advocates the method is not true. See the talk page, where the criticism that Rspeer gave is shown to be false. Now, it may be that someone else would reformulate the explanation in a different way, such that such criticism and its refutation is included in the article. That is fine. Go ahead. I don't mind such a re-edit. But the fact that you don't understand the method does not mean that you can imply that the text is a mere advocacy.
Also, Rspeer included a request for mediation, see http://wiki.riteme.site/wiki/Wikipedia:Requests_for_mediation#Parties.27_agreement_to_mediate. See my explanation why this is too early.
Now, I will restore the paragraph on the list of candidates, the budget set, and the need for temporal consistency. For Rspeer's piece of mind I will not mention the solution of the Borda Fixed Point, and leave it up to you to figure out whether that method should be mentioned and, if so, where. As I said: Have fun. If you are interested in voting systems, you should have that. Colignatus 02:34, 8 March 2006 (UTC)

Request for comment

I've made an RfC about this issue, at the request of other admins. See Wikipedia:Requests for comment/Colignatus, and add your view if you want. rspeer / ɹəədsɹ 05:18, 11 March 2006 (UTC)

Autobiographical edits

I think that one moral to derive from the Colignatus conflict above is that you shouldn't add information about yourself or your research interests to Wikipedia. This is a corollary of the WP:AUTO policy. The problem is that when the edits are about you, it is easy to overreact by construing a criticism of that edit as a criticism of you. (Of course, this is not nearly the only thing that Colignatus did wrong, but this is where I believe the whole ugly issue started.)

I've just reverted a link that Craig Carey posted to his own advocacy site. Craig, looking at your contributions, it looks like you've been a helpful Wikipedia editor, even discovering some libel posted in Wikipedia and fixing it. Your contributions are very much appreciated. However, I have to revert your link here, both because it doesn't meet the standards of notability I think links on this article need to meet, and because your adding of it creates an autobiographical conflict of interest.

I do this so that everyone doesn't just link to a site expressing their own view -- and there are just about as many views in election methods as there are people.

I feel that I'm inviting more comments that I'm "taking ownership" of this article. It's a tough balance, because I care about this article and want to keep it in the best state possible, while recognizing that some contributions may start out looking bad but evolve to become a useful part of the article. So I always welcome feedback on my edits to this article.

As long as that feedback doesn't consist of incoherent ranting threats, that is.

rspeer / ɹəədsɹ 05:55, 13 March 2006 (UTC)

disambig

I just added an Otheruses tage and created a disambiguation page. This is per a discussion I had with Rspeer on his talk page [User talk:Rspeer#Voting Machine vs. Voting System]. Do let me know if you've got issues with it or other ideas. -- Joebeone (Talk) 00:20, 29 March 2006 (UTC)

Talk:Social Choice and Individual Values

Announcement: The above is the discussion tab for a new article Social Choice and Individual Values. Input is welcome through the article, the Talk page, or to me. The plan is to gather comment, corrections, or suggestions for probably at least a couple of weeks, make final changes, then go from there. Links to related articles (indluding the present one) would come after revision. Thanks for your help.

Thomasmeeks 23:04, 27 May 2006 (UTC)

A wikipedia voting system

Using scale and the "voting only on well-known people" principle

(context: on Wikipedia we have Elections for the Board of Trustees)

Approval voting is a is a voting system used for elections, where a "list of candidates" must be pre-defined. If the list have "democratic origins", the election by approval voting will show a democratic result. It is a perfect system for wikipedia community...

But what is a "candidate" in a direct democracy? People that wants power? Marketing? Volunteers? Whant help others? It was good if I, or you, or your friends, be a candidate? (And how many candidates on the list to be a democratic list?)

We have on Wikipedia community a exceptional opportunity to do a "more truth" democracy... the goal of this community is to colaborate on articles, and through articles we know people. Only on this "known people" we can vote (!). Not names or promises on a "Elections list of self-candidates".

We haven't time to investigate another people, only ours "known people", and this is a good principle: colaborators voting on a colaborator from (and only from) the articles where he/she was colaborate. From this first ease voting we can produce a very surprised and democratic list of potenctial candidates... and invit the "best approval" of then to become candidates.


Operationally it is possible:

  • Principles:
    1. Truth democracy need also to generate, all time, new and truth candidates.
    2. Only about your "known (local) people" we can do a truth vote.
  • Possible wiki process to generate renewed and democratic list of candidates:
    1. Each colaborator can votate on N (1, 2 or 3) colaborators from (and only from) M (2*N or 3*N not very more) articles where he/she was colaborate.
    2. We will have a "per article", "per language" or "per etc." (level) elected ones. They will be invited to become "per level" candidates.
    3. The confirmed candidates will go to compose the "list of candidates".

-- User:krauss (please sorry my english if necessary)

Criteria merge

See Talk:Voting system criterion — Bitt 15:24, 17 September 2006 (UTC)

Range voting research paper

The last link in the "Research papers" section, the paper by Warren D. Smith, is intriguing. On the one hand, it does all kinds of simulations that were sorely needed in voting theory, giving quantitative answers to things that people could previously only make guesses at. That's a really impressive accomplishment. And it's even backed up by the code to his computer program, making his results reproducible. On the other hand, it doesn't look like this paper has been formally peer-reviewed, and the author has an unabashed agenda of promoting range voting.

Now, he's entitled to promote range voting if his results are really that incontrovertibly in favor of range voting. But these questions remain:

  • Are his theorems accurate? Many of his results are based on his conclusion that a large number of ranked systems reduce to the same strategy. If there is a flaw in this conclusion, it casts doubt on most of his results.
  • Are his models objective? We've seen many instances of people making models based on the assumptions behind a particular voting system, and then -- lo and behold -- the voting system that matches the assumptions of the model turns out to be the best. Does range voting only come out on top because he makes assumptions that benefit range voting?
  • Suppose this paper has no such holes. Then, the fact that it's an excellent source of information is pitted against the fact that it's not peer-reviewed (except where the peers are hobbyists who discuss voting theory online). Should Wikipedia incorporate that information into articles? Link to it only (the status quo)? Or remove the link, to avoid setting a precedent that unreviewed papers are welcome in the links section?

rspeer / ɹəədsɹ 00:25, 15 December 2006 (UTC)

I'm the one who initially included this reference, and for the record, I have no link to Smith, nor am I generally a particular advocate for Range Voting (though I do favor voting reform). I read the paper and would like to speak to these questions insofar as I can. (Though I'd like to "slide by", I must also point out that the article does more than link to this data; the second-to-last paragraph of the "criteria" section is essentially mine and essentially based on this paper).
  • Are the theorems accurate?... It is true that the paper gives a fair amount of weight to the argument that a large number of ranked systems reduce to the same strategy. I myself am very skeptical of this claim in general (though he does solidly prove it for Borda). However, if you carefully read the numerical results, discounting the "strategic something-I-don't-like" as a excessively-worst-case, it is clear that there are still conclusions to be drawn, just not ones as strong as his. The language I used in the article already reflects this analysis.
  • The models... this is difficult to assess, and certainly it is possible to imagine and argue for the importance of features that he does not include in his models. Especially in the case of "strategy", as his models essentially arbitrarily assign the frontrunners independently of their quality (though this only matters for "strategic xxx"). However, on the plus side, he has done a good job of varying his models and his results are substantially robust under such variation. I'd say that's a strong point in their favor.
  • Is this OR?... Well, my arguments here may actually weaken the case in that regard, as they show that my inclusion of the article here departs significantly from the tone of the article, and that could be seen to constitute OR on MY part. However, I would argue for a different interpretation: by critically reviewing the data, I have made it less OR. I'd also try to argue that this paper has been "almost peer-reviewed" through its presence on the internet, its presentation within the verbal register of mathematics, and the process of feedback and response within that register that Smith has obviously participated in... I know this is a stretch, so I'm not going to give citations of the latter unless this argument appears to matter. Or how about this one: the article itself mentions that significant developments in the field have come from "hobbyists" - many of whom have academic credentials - working on the internet, and that ONLY FOR THIS REASON the article's presence INCLUDING DEMONSTRATED RELATED GIVE-AND-TAKE SHOWING AN OPEN MIND, AND INCLUDING SOURCE CODE, constitutes a rudimentary peer review. (hopefully, this standard would not cast the net around too much crackpottery, either here or elsewhere on WP).
Basically, I agree that it's somewhat shaky ground, but it's worth it because the data is good and there is no other source of comparable data. Which is why I included this in the first place.--Homunq 23:10, 15 December 2006 (UTC)
Are you counting on an author to also post any critique? Where would one look for any review or critique of such an internet paper? I tried to view the simulation source code using the URL in the paper, and got a server-not-found error. I don't object to Warren Smith's paper and simulations being mentioned somewhere in Wikipedia, even though I think they have some significant flaws, weaknesses, and limitations.
The simulations presume there is some sort of voting system standard or criterion about maximizing social utility. If there really is a recognized voting system standard or criterion based on social utility, it ought to be discussed, even have its own article. If there is not such a recognized standard or criterion, it is questionable whether the paper should be discussed here.
At a minimum, the current content in this article about the paper needs improvement in the areas of accuracy, verifiability, POV, and balance. DCary 07:11, 3 March 2007 (UTC)
You raise some good points about why we shouldn't give a self-published paper like this the benefit of the doubt regarding verifiability. For one thing, with a real publication you know that its appendix of code isn't going to turn into a "server not found" error. I also recognize that critical evaluations of the paper will be difficult or impossible to find (if they exist at all). So I'm going to harden up my stance -- I only support the paper being an external link, not a source for statements in the article. rspeer / ɹəədsɹ 09:19, 3 March 2007 (UTC)


I don't know where you were looking, but I clicked right through to [5] the first time.
Obviously this is covered under WP:ATT. I'd say this material falls under "primary sources" (the results/output of the computer program) "Examples of primary sources include... written or recorded notes of laboratory and field experiments or observations... Edits that rely on primary sources should only make descriptive claims that can be checked by anyone without specialist knowledge. Any interpretation of primary source material requires a secondary source." The source code is available, and Smith is a faculty member of a respected institution whose job is on the line if he falsifies his actual data - I don't think anyone doubts his actual data. I agree, any interpretation he (or I) makes of this data, beyond what "can be checked by anyone without specialist knowledge", or that "introduces an analysis, synthesis, explanation, or interpretation", is inherently an unreliable source.
So, we can mention the possibility of simulating an election, talk about the limits of this methodology, and link to this study, mentioning that it was run by an advocate for Range Voting. We could even pull out data points and represent them graphically, as long as we were careful not to format them in a way that was biased towards a given interpretation, in any way that goes beyond the biases inherent in the data itself - an almost-impossible "if". Sadly, my simple piece of interpretation - that the data supports calling system A "better" than system B in this regard if and only if A's worst (full strategy) result is better than B's best (strategy-free) result - is just as wp:unreliable as Smith's tendentious claim that Range is "always the best system" (wrongly comparing best-to-best and worst-to-worst, not worst-to-best).
I still say that this is worth citing in the body of the article. Among Smith's references, there are published papers which apply the concepts of simulation and "regret" to voting, so we're on solid ground to mention that. And Smith's data, while only reliable as a primary source, is still the best data in this regard. I've edited the article to not make any claims about specific voting systems. Now, the article really just shows the lack (or opportunity) for a reliable source to deal with this issue.
And if anyone is listening who is interested in this opportunity and has the possibility of publishing in a reliable source... if you mention Smith's data, please mention the fact that it's only valid for best-to-worst comparisons (since that in removes any unwarranted assumptions about how much strategy will be used in different systems).--Homunq 05:48, 4 March 2007 (UTC)

Requested move

Voting system → Electoral system — The article currently entitled Voting system is about methods for converting votes into a final outcome. In my experience researching and working on electronic voting, the term voting system is consistently used to refer to the mechanisms that produce, handle, mark, collect, and count ballots. "Voting" also has this meaning in the term electronic voting. On the other hand, the Citizens' Assemblies in British Columbia and Ontario consistently use the terms electoral reform and electoral system to refer to the method of converting votes into elected representatives; this is what the Voting system article is about. The titles of the books cited at the end of the article use the phrase "electoral systems", not "voting systems". Finally, if we use voting system to mean "translating votes into seats", then there is no other term but voting system to mean "casting and counting votes" as well — confusingly, both would then be called "voting systems", and indeed this confusion currently exists (see Voting system (disambiguation)). A much clearer way to distinguish the two is to name them electoral systems and voting systems respectively. For all these reasons, I support the renaming of the current Voting system article to Electoral system. Ka-Ping Yee 06:38, 20 February 2007 (UTC)

Survey

Add  # '''Support'''  or  # '''Oppose'''  on a new line in the appropriate section followed by a brief explanation, then sign your opinion using ~~~~. Please remember that this survey is not a vote, and please provide an explanation for your recommendation.

Survey - in support of the move

  1. Support - original nominator. But see the table below for details; more discussion may be needed. Ka-Ping Yee 20:14, 21 February 2007 (UTC)
  2. Support - as per nomination. Jimp 06:48, 20 February 2007 (UTC)
  3. Support - for reasons listed and I think the move will add clarity for the general Wikipedia audience.--Electiontechnology 07:06, 20 February 2007 (UTC)

Survey - in opposition to the move

  1. oppose - for the reason I gave above under #Voting_system_criteria Pm67nz 21:42, 20 February 2007 (UTC)
  2. oppose - Voting systems are more general than just elections. Tom Ruen 22:06, 20 February 2007 (UTC)
    Oppose. If anything, "electoral system" should be a separate article about the components that go into a political election -- the voting system, the method of districting, the party structure... It's a different topic than this article. rspeer / ɹəədsɹ 22:50, 20 February 2007 (UTC)
  3. Oppose per Rspeer: "electoral systems" should cover the nonalgorithmic parts of elections. CRGreathouse (t | c) 03:47, 22 February 2007 (UTC)
  4. Oppose Right problem, wrong solution. Voting is just one part of an election. Save Electoral System for those things that create the broader context for voting. Algorithms are not systems. The topic is less about voting and more about what happens after the voting is complete. Election Method is the better page name and terminology. DCary 23:31, 28 February 2007 (UTC)

Discussion

  • I would like to point out that there are currently 295 entries that link to "Electoral system," 7 that link to "Voting method," 15 to "Electoral systems," 16 to "Voting theory," 4 to "Social choice function," 3 to "Voting methods," and 48 to "Voting systems." I think that is a clear need for some clarity
  • Is there a particular reason that this article uses "Voting system" versus "Electoral system?" In my experience, the term Electoral system seems more fitting and much more common. Even the template uses the term "electoral."--Electiontechnology 05:50, 19 February 2007 (UTC)
  • On a related note, what do folks think about Vote counting system? Should it be renamed Voting system? Ka-Ping Yee 09:03, 20 February 2007 (UTC)
    • Why do you say it should? It seems to have the right name now, given that it's about how you actually count the votes. I feel like these moves are being proposed just for the sake of moving things. rspeer / ɹəədsɹ 22:52, 20 February 2007 (UTC)
      • It couldn't be further from the truth. Nothing is being moved for the sake of moving. These suggested moves are for clarity. The vote counting system article probably should be part of a greater voting system article as in technology terms a "voting system" includes everything in the preparation, counting, and tabulating process. It's not quite as simple as it might seem. It is important to discuss vote counting in the larger context of a voting system (technology definition).--Electiontechnology 23:09, 20 February 2007 (UTC)
      • I don't want to move articles just for the sake of moving; the titles of the various articles seem a bit confusing at the moment and could be organized better. There are several different concepts here, and the goal should be to find distinct and reasonably clear titles for each in-depth Wikipedia article necessary to cover the area. Let's keep working this out here until we have a decent solution -- see below. --Ka-Ping Yee 02:21, 21 February 2007 (UTC)

Closing

I tried to close this move request as a "no consensus", but I was reverted by User:Electiontechnology with the summary, "discussion is ongoing, it doesn't make sense to close the debate". That's fine, and I hope the discussion arrives at an agreeable conclusion, as it looks like there's a bit of stuff to sort out. Meanwhile, this request had slipped into the backlog at requested moves, so I wanted to do something with it.

I see two options. We could close this request as a "no consensus", and let you guys sort out which article needs to be where, and then come back and request it when you're got a consensus. Alternatively, we could just bump the listing up to the top of WP:RM, and then I or another admin will check back in with you in five days. Which would people prefer? -GTBacchus(talk) 10:17, 26 February 2007 (UTC)

I guess maybe the request to move was made prematurely to the admins. It just seems like there is a lot of good discussion going that could lead to some much needed organization and I wouldn't want to stymie that debate. Is anyone against continuing the discussion and then coming back to admins? Can we just withdraw the original request until the discussion progresses some? I think labeling it as "no consensus" is somewhat counterproductive. Thoughts? --Electiontechnology 18:11, 26 February 2007 (UTC)
That's fair - I didn't mean to imply any sort of finality by calling it "no consensus". Shall we consider it withdrawn until a suitable consensus is reached? I think we can do that without any formal procedure: the request is already delisted at WP:RM, and I can take the move request banner off this talk page (or not?), but leave the discussion as is. Then, when you're ready to request some specific move(s), just start a new listing at RM and we're in business. Does that seem ok? -GTBacchus(talk) 21:48, 26 February 2007 (UTC)

Organizing articles about elections and voting

Let's talk about how to organize the articles in this area. Here's what I've surveyed so far. (Please help collect information about the current state of Wikipedia articles in this area and feel free to add it to this table.)

Concept Common names Currently used terms and articles in Wikipedia
#1. How elections are conducted: regulations, processes, political and administrative bodies. Includes the specification of things such as #2 and #5. Election, Election administration, Electoral system Main article: Election. Related article: Electoral reform. Stub-like articles: Election administration and Election law. Many country-specific articles such as Australian electoral system, Canadian electoral system, etc.
#2. How votes are cast, collected, and counted: artifacts, procedures, and machinery for voter registration, polling, and tabulation. The implementation of #1. Voting system, Voting, Polls, Election administration, Election management Stub-like articles: Vote counting system, Election administration. Related content at Ballot.
#3. Devices used to cast votes: equipment used to implement the vote-casting part of #2. Voting machine Voting machine currently describes both vote-casting machines and vote tabulation, though it calls these "Voting system recording technologies" and "Voting system tabulation technologies". Some of this is also described in Ballot.
#4. How votes are translated into an outcome: algorithms for aggregating many expressed preferences into an overall selection. Voting method, Election method, Voting system, Social choice function, Voting theory Main article: Voting system talks about both of these things. The navigation template with the list of algorithms is entitled Commonly used electoral systems. Throughout the article the individual algorithms are called methods or voting methods.
#5. How seats are allocated in a legislature: the application of #4 to choosing candidates, with additional issues about political parties and representation. Electoral system
  • Some thoughts and concerns I have:
  • My current thinking on a possible solution:
    • #1: Main article: Election. As now.
    • #2: Main article: Election administration. Merge in Vote counting system; more content needed here.
    • #3: Main article: Voting machine. Mostly as now, but focused on Voting_machine#Voting_systems_recording_technologies.
    • #4: Main article: Voting method. Just about algorithms and mathematical criteria. Voting theory might also deserve its own article.
    • #5: Main article: Electoral system. Application of #4 to the election of representatives (e.g. PR goes here, not in Voting method). Social and political criteria go here. Add references to electoral reform and various national electoral system reform efforts. Top of the article links to Election for the more general concept and to Voting method for the more specific concept.
    • Election method goes to a disambiguation page listing Voting method (more theoretical) and Electoral system (more practical).
    • Voting system goes to a disambiguation page listing Election administration (most general), Electoral system, and Voting method (most specific).
  • What do you think? -- Ka-Ping Yee 02:21, 21 February 2007 (UTC)
  • "Electoral system reform" can cover a broad range of topics, including nomination procedures / requirements, districting, campaign finance, voter eligibility/registration, logistics of casting ballots, etc. None of which are included in #5. I see #5 more as a subcomponent of #4. For some election methods, a full result is not determined until the allocation is done. DCary 00:13, 1 March 2007 (UTC)

Please keep in mind that the voting method article should encompass non-election voting situations. Specifically, not all voting situations involve the election of a candidate. Two examples are the voting done as a part of the American Idol TV show, and the ranking of choices that are not candidates (as done at www.FullRanking.com). VoteFair 18:00, 26 February 2007 (UTC)

I think that any "practical" article (#5) should attempt to summarize the "mathematical" fitness of applied algorithms (#4). This may be contentious. If we commit to acheiving it nonetheless, I'd support this proposal; otherwise, not. --Homunq 03:14, 1 March 2007 (UTC)

Two infoboxes?

I understand the relative virtues of the infoboxes as they stand, but really we only need one. --Homunq 04:01, 28 February 2007 (UTC)

Voting criteria compliance boxes

I see that someone has merged participation and consistency. I do not think this is a good idea as one criteria does not necessarily imply the other. For example, range voting complies with participation and consistency, but median ratings satisfies participation and fails consistency. I opine those criteria boxes should be separated again. --Fahrenheit451 22:43, 8 March 2007 (UTC)

I added that data as a footnote. The article makes it clear - most importantly, by naming them as separate criteria - that these are not the same. However, since they are conceptually related, AND all the given systems pass or fail them both together, to have separate columns would be to give undue weight to this conceptual area. --Homunq 16:40, 9 March 2007 (UTC)

Majority criterion

The previous wording of the majority criterion was inappropriate, as it referred to ambiguous and contested terms "preference" and "sincere" rather than being strictly worded in terms of ballots and outcomes, as is described in the article on voting system criterion.

The new wording (taken from http://fc.antioch.edu/~james_green-armytage/vm/define.htm#mc which was listed on majority criterion) makes it clear that approval voting trivially satisfies the criterion (it is equivalent to stating that if a majority of the voters approve only a single candidate, that candidate must win, which is always true of approval voting).

I have made similar changes to the article on majority criterion, notably removing the example of approval voting and replacing it with one for range voting, which does indeed fail mc. I'd be happy to discuss or debate these changes, and to review various alternative wordings of the majority criterion as it applies to non-rank voting systems.

--Sapphic 20:23, 2 April 2007 (UTC)

This is the perennial "votes-only vs. hidden preferences" debate, and I don't think you can resolve it so easily. Although I prefer votes-only criteria in most cases, there are two problems with using votes-only exclusively:
  • Most of the established literature refers to hidden preferences, making statements about votes-only criteria difficult to source. Votes-only criteria are a relatively new development, fueled by voting system hobbyists who tend not to publish their results in reliable sources. (Note, for example, that while James Green-Armytage tends to write useful and informative things about voting theory, he's not a reliable source.)
  • Plurality passes nearly all the votes-only criteria in vacuous ways. Why didn't you change the Condorcet and Condorcet Loser entries for Plurality to "yes"? Not that I think you should do this -- for me, it's a reductio ad absurdum, as Plurality is clearly not a Condorcet method.
It also concerns me more generally that entries in this table flip to their opposites so often, while the underlying mathematical facts are of course not changing. We really should aim to have a footnoted reference for each entry.
rspeer / ɹəədsɹ 22:15, 2 April 2007 (UTC)
I wasn't aware that this was a long-standing debate topic, and do not plan to engage in any edit-warring should the "opposition" decide to revert my changes. However, in this case, the intent of the criterion seems perfectly compatible with a votes-only phrasing, so I'd argue that we should keep the current form. That means that approval would indeed pass, whereas cardinal ratings would fail. I agree that examples should be cited for each failure of a system to meet a particular criterion (and I added an example of cardinal ratings failing majority criterion on that page, which I suppose could be linked here as a footnote), but in this case the flip was because the phrasing (and thus interpretation) of the criterion changed, and not any facts pertaining to the system itself (approval voting). I didn't change any other entries (yet) because I only came across these articles today.
Also, I'd like to say that I am not totally opposed to a hidden preferences phrasing of voting system criteria, but more that I don't like the fact that these terms are basically undefined. If the preferences were described in more explicit mathematical terms, and in a way that they could unambiguously be extended to non-rank methods, I'd think they were fine. Similarly for explaining what "sincere" means with respect to expressing those preferences. I don't think it's necessarily correct to treat AB and A as if they were both equivalently sincere expressions of the preference A:3,B:2,C:1 and since the examples purporting to demonstrate that approval violates MC all depend on such interpretations, this is a crucial point.
Mostly it seems to me that the problem is that none of these criteria were really developed with non-rank systems in mind, and so applying them to such methods is problematic, at best. Prior to my edits to majority criterion there was some text in the intro implying that the criterion simply did not apply to non-rank methods, but I deleted it as being inconsistent with the rest of the article (particularly with the examples, which originally included approval and now includes cardinal ratings). If we wanted to sidestep this whole issue of votes-only versus hidden-preferences and rank vs. non-rank methods by explicitly restricting the criteria to rank methods, that might work too.

--Sapphic 22:45, 2 April 2007 (UTC)

Universality criteria

When I added the Kemeny-Young method to the comparison table I considered adding the universality criteria as:

  • Universality: Are all the candidates ranked (in a consistent way) from most popular to least popular?

I did not add this criteria -- which only the Kemeny-Young method achieves -- because the table is limited to single-winner voting criteria (and widening the table disturbs the nice formatting).

Is there any interest in accomodating ranking-based voting systems as a part of this voting-system comparison? The widespread availablity of computers now makes ranking methods practical, and surely ranking-based voting methods fit within the topic of "voting systems". VoteFair 06:16, 14 April 2007 (UTC)

I'd say that this would introduce confusion with PR methods - a ranking sounds as if it should be usable for multiple seats, but it isn't proportional. --Homunq 23:43, 16 April 2007 (UTC)

Have all the voting systems been validated?

From QA/QC perspectives, if a method/system is developed then it has to be validated, otherwise biases will be all over the places where the methods are applied. It should be OK if a method / system is used for research scales but a risky practice when used in production scales. As can been seen in Table of voting systems by nation, all these systems are in production scales which may generate inharmonious A/Effects. —The preceding unsigned comment was added by 72.52.81.34 (talk) 02:11, 12 May 2007 (UTC).

Vote-sizing?

Hi, I just reverted edits from User:Thapadushy in an effort to include something called "Vote-sizing" into the article.[6] I've never heard of this and haven't seen it in any verifiable sources. Anyone else know of this? If it can be cited/sourced, should it be included? -- Joebeone (Talk) 21:30, 29 May 2007 (UTC)

Computing power

Dear VoteFair, you wrote: "The increased availability of computer processing has increased the use of the Kemeny-Young and Schulze methods that fully rank all the choices from most popular to least popular." Well, I am not aware that the Kemeny-Young method is being used somewhere. Markus Schulze 13:00, 6 October 2007 (UTC)

To Markus Schulze: Your lack of knowledge about the use of the Kemeny-Young method does not constitute a lack of use. For your edification the www.VoteFair.org and www.FullRanking.com websites provide (free) access to VoteFair ranking, which includes Kemeny-Young calculations as the main part of their results. These websites promise anonymity to those who use them, yet I can tell you that they have been used privately for officer elections (in small organizations), choosing logo designs, ranking political parties (in Ireland), and more. Each season's public VoteFair American Idol polls (involving thousands of voters) also use these calculations. Usage continues to increase as more people learn about these two websites. VoteFair 06:27, 8 October 2007 (UTC)
Dear VoteFair, Wikipedia demands reliability. It is not sufficient that you claim that some private organizations were using the Kemeny-Young method. You must also provide examples and references. Markus Schulze 13:37, 8 October 2007 (UTC)
Please don't use your own website as examples. See Wikipedia:no original research. Scott Ritchie 22:33, 9 October 2007 (UTC)

To Schulze: I provided "examples and references" above. Apparently you did not bother to follow the links I provided. So, to make things more convenient for you, here are three web pages at VoteFair.org that document many cases of using VoteFair ranking in public polls: [[7]], [[8]], and [[9]]. Also, at the bottom of the [[10]] home page is a link titled "Sample survey" and it leads to a public survey that has been filled out by more than a dozen people who have used the FullRanking.com service for surveys, polls, and elections. (While you are there, notice that the survey itself uses VoteFair ranking for the first question.) There are other people who have used the VoteFair ranking service without completing the usage survey, but as I said before I can't compromise their anonymity. I haven't counted the number of uses, but it is significant. Also note that some of the surveys are in foreign languages, and some use placeholder names, so I don't always know what they are about. As further evidence, a Google search reveals yet another example at [[11]], which is the Irish-party-ranking example I mentioned above. Surely this is evidence of my "claim" that VoteFair ranking--and therefore the Kemeny-Young method--is indeed used "somewhere."

As also requested by User:Mccready, if you need additional proof that the Kemeny-Young method is indeed used in real situations, please discuss this issue here--instead of believing something isn't used just because you don't want to know about its use. VoteFair 06:58, 9 October 2007 (UTC)

Dear Richard, everybody can run a private Internet poll on whichever question he considers interesting. However, it is not the same whether American Idol uses a given method or whether someone runs a private Internet poll with a given method about who should be the winner of this contest. It is not the same whether the president of the USA is elected by a given method or whether someone runs a private Internet poll with a given method about who should be the next president.
You wrote that the Kemeny-Young method was used for "officer elections in small organizations"; but you didn't provide any examples or references. Markus Schulze 11:58, 9 October 2007 (UTC)

Remember that this "voting systems" article is not just about voting methods used in government elections (which would disqualify the Schulze method), and not just about voting methods that are publicly and officially adopted by organizations (who tend to implement heavily promoted voting methods such as the Schulze method and Ritchie's preferred IRV method), but also about voting methods used in surveys and polls. Although the Kemeny-Young method is increasingly used in organizations I cannot name because of promises of anonymity, I've edited the article's wording to accommodate your current disbelief. With this edit I'm also adding the ranked pairs method, which I excluded originally because I didn't know if it was "used" anywhere. (I now see it's officially used by one organization.) It's ironic that Schulze's attempts to reduce the visibility of the Kemeny-Young method attracts attention to it, and gives it credibility as a very viable alternative to "his" self-named Schulze method. VoteFair 06:20, 11 October 2007 (UTC)

Why is the table so poor?

It's got a ton of ambiguous entries where there shouldn't be. Approval and Range voting can both fail to elect the Condorcet winner, and they (very rarely) elect the Condorcet loser. Both pass independence of clones (if they are true clones, they should be ranked the same in both systems). Also, it's ambiguous whether or not Approval passes the Majority criterion. Have there been been huge edit wars on this page or something? There's no ambiguity mentioned when discussing these criteria anywhere else on wikipedia. Paladinwannabe2 20:04, 22 October 2007 (UTC)

After thinking about it, I realized that Approval voting is ambiguous on independence of clones for the same reason that it's ambiguous whether it passes the majority criterion- the definition of clones assumes ranked methods. Range voting clearly passes, however. Paladinwannabe2 20:20, 22 October 2007 (UTC)

Hello, this is James Green-Armytage. There is a dispute over the following external link: [12] I believe that this link should be included, and Fahrenheit451 believes that it should not be included. I've reviewed wikipedia's guideline on external links, WP:EL, and since the web site in question is mine, I find that I should not unilaterally add the link myself, but should rather open an entry on the talk page, as I'm doing here. I note also that Matthewedwards added a link to the site on September 3rd (which Fahrenheit451 has just deleted), so I am not the only one who considers the link to be valid. Also, I believe that this article (in addition to a few other voting theory articles) has linked to the site pretty consistently for over two years, so I'm not quite sure why Fahrenheit451 chose this particular moment to remove it. It bears mentioning as well that Fahrenheit451's edit may fall under the category of "grudge edits". He and I have been in contention with each other over a number of years (points of contention have included the merits of the Borda count and our differing views on the standards for external links), and I did nominate an article of his for deletion, over a year ago. (The vote went in favor of deletion).

Below I will make a case for this being a good link; I hope that discussion will ensue, that most editors who have contributed to this page will come to the same conclusion, and that someone else will add the link again.

I suggest that the link in question contains a great deal of useful information for readers interested in voting systems. In particular, I'd like to highlight the voting methods survey, [13], which describes about 44 distinct election methods on a single page. In many cases, numerical examples are given to help the reader learn how each method is calculated. (Each example was devised to highlight the differences between related methods.) Thus, as a learning tool, it might help readers who are having trouble grasping the material as presented in wikipedia. Please note also that while wikipedia now covers much of what is on that page, this was not the case at all when the page was first launched in the spring of 2004. (Indeed, it was the first entry returned by a google search for "voting methods" for several months (possibly even a year), before this article overtook it.) I'll admit that it could use an update here and there, but I'd like to submit the opinion that it's kind of a golden oldie at this point, that it has been used as an information source by people interested in voting theory, and thus that it might have helped form the basis for some wikipedia content. Other pages of interest are the criteria page, and the delegable proxy page, the latter of which was the first google result for "delegable proxy" for over a year and until very recently (possibly up until the period when the site was unavailable due to a server transfer of the fc.antioch domain).

Fahrenheit451 may argue that the site should not be linked because it is not a peer-reviewed article. However, a careful reading of WP:EL will show that this standard is not germane; many good links on wikipedia, and in this article, are indeed not peer-reviewed, nor should they need to be. Fahrenheit451 may also argue that the site falls under the category of personal web page, which is listed in WP:EL under the heading "links normally to be avoided". While this label may be fitting (though I must emphasize that my pages on voting theory are kept quite separate from pages on other topics, being connected only by one link), I ask you to judge the site on its merits, as suggested by WP:EL. I do feel that I can be described as a "knowledgeable source" on the subject of alternative election methods (as point 4 of WP:EL#Links_to_be_considered recommends). I'm a PhD student in economics at UCSB, with a focus on voting theory. My dissertation advisors include Nicolaus Tideman and Theodore Bergstrom, whose own dissertation advisor was Kenneth Arrow. I've been active in the voting methods community since 2001, when I independently invented a Condorcet method and proposed it to a number of IRV organizations. I contributed frequently to the Election-methods mailing list from 2003-2005, and I served as an intern for the Center for Voting and Democracy in 2003. I have also, of course, contributed to many of the voting methods articles on wikipedia.

Overall, I believe that the site is an accurate, helpful, and relevant source. I put a lot of work into it in the hope that I would be able to raise awareness of alternative voting methods. I believe that the site is up to par with other similar sites that I admire, e.g. Condorcet.org, Ranked ballot voting methods, Accurate democracy, Voting and election reform, Eric Gorr's calculator, and even the defunct electionmethods.org, which was polemical but trailblazing. Many of these can be described as personal web pages, but it must be acknowledged that they contributed significantly to the field and helped drive interest in alternative voting methods to the point that allowed these wikipedia articles to reach their current level of quality. --Hermitage (talk) 09:58, 27 November 2007 (UTC)

I, for one, think your site is useful and should remain as an external link on many pages, Voting system and Borda Count at the least. If you're concerned about WP:COI I'd be happy to add the links myself. Your web site is more helpful than most I've visited; I'm surprised to see the links removed while others remain. Surely it's better (sorry!) than Cretney's site or the National Brainstorming project? CRGreathouse (t | c) 13:47, 27 November 2007 (UTC)
I see nothing in WP:EL that discourages these links, and a couple of things that encourage them. What should be linked contains "Sites that contain neutral and accurate material that cannot be integrated into the Wikipedia article due to copyright issues, amount of detail... or other reasons". Links to be considered contains "Sites which fail to meet criteria for reliable sources yet still contain information about the subject of the article from knowledgeable sources." The links belong there. rspeer / ɹəədsɹ 15:10, 27 November 2007 (UTC)
Thanks to both of you for reading my long-winded topic, and for the kind words. --Hermitage (talk) 15:21, 27 November 2007 (UTC)

Update: Fahrenheit451's response was to delete the link again, delete the discussion of the issue from his talk page without reply, and ignore the discussion here, although I clearly linked it to him on his talk page right after posting here. I have to conclude that he's editing in bad faith here, and not respecting the elements of discussion and collaboration that wikipedia requires. Quite possibly, he's doing this just to screw with me. In other words, I don't know what to do with this guy. --Hermitage (talk) 21:53, 27 November 2007 (UTC)

James, I really don't appreciate your violation of WP:AGF with me as is evident in your comments above. It appears that you have a WP:COI issue here. I did inform you that I would be willing to mediate on putting the link to your personal website in the links section. You wildly speculate about my motives, for example, the deletion of the stub article I wrote, I believe binary independence, and that you theorize I bear a grudge. Frankly, I would have deleted that article myself today as it was not notable enough for a seperate article. I realize that you are still a student and have insufficient experience in real life adult social situations, but what you are dreaming is going on here is far from factual. I find your personal website link adds nothing new to the body of knowledge of voting theory. While I am no longer an advocate of Borda count, I find your section on that method to be highly POV. There are critical papers on Borda count that are much more scholarly and NPOV. I hope you can take a mature point of view, cease your wild speculations, self-importance, and work with editors with other views on an adult to adult basis. --Fahrenheit451 (talk) 23:43, 27 November 2007 (UTC)
I'm not familiar with the mediation process, but honestly I don't see the need. Once again it seems clear from reading WP:EL that discussion among editors (i.e. on this page) is the customary way to resolve the issue of whether a link should or should not be included. What is it about this particular case that makes you feel that such discussion is inadequate? Does anyone else feel that mediation is necessary to resolve this?
I apologize, as my comments regarding your faith (good versus bad) were irrelevant. What I meant to say is that you were editing without regard for what others had said on the topic; it seemed that you lost patience with rational discussion and decided to pursue your agenda regardless. You insisted repeatedly that the WP:EL page mandates a published/peer-reviewed standard for external links. When I actually read the page and pointed out that it clearly does not do so, you ignored my comments, ignored the comments of others, deleted my comments from your talk page, and continued to delete the link, once again citing WP:EL. Anyway, I'll try to refrain from speculating as to your motives. I certainly don't understand them, but I suppose that I don't need to. Honestly, I'm glad that you're not upset about that article I nominated for deletion anymore (binary independence sounds right). I take no pleasure in deleting content, and felt bad about that at the time, but as you say, it didn't seem to merit a full article. I will also try to avoid conflict of interest, but I feel that there is a difference between unilaterally adding an unwanted link and reverting unilateral edits against a link when other editors have expressed an interest in keeping it. (I'd argue that the latter is not a conflict of interest.)
As for the Borda article, I agree with you that the tone is more polemical (I mean less neutral) than it would be if I was writing it now. It would be nice to revise it at some point, when I have more time. However, this is just a question of style, isn't it? I believe that the criticisms I raise in the piece are still mathematically correct. Of course, if you can find a mistake, I will change it. --Hermitage (talk) 01:54, 28 November 2007 (UTC)
This is an unfortunate dispute. I have seen Fahrenheit451 make many constructive contributions, but when I looked more carefully at Special:Contributions/Fahrenheit451, trying to figure out what had happened here, I also was dismayed to see many instances of the sometimes not so subtle violation of Assume Good Faith involved in demanding others Assume Good Faith or the incivility of chiding others for incivility. Fahrenheit451 is an experienced editor with many, many edits as well as extensive involvement in content disputes; therefore I'd expect him to show better behavior than the somewhat less experienced Hermitage. Yes, Hermitage should not be involved in putting in or retaining links to his own website, other than through Talk; that is well-established and proper. However, that web site is also quite a useful one, and in removing links to it, Fahrenheit451 has himself been quite a bit less than civil.[1]
This issue goes quite a way back. I see that the link was originally added to this article, probably by Hermitage, as an IP editor, and it was immediately removed as a "vanity link" by Fahrenheit451.[2] If the link was inappropriate, it isn't necessary to call it "vanity"! (This kind of incivility is all too common, biting the newcomers. Of course, Fahrenheit451 was himself a newcomer at the time.
Since it appeared that there was some communication between these two users around then, I looked more and found this, placed by Fahrenheit451: User_talk:Hermitage#Hermitage_is_operating_here_in_bad_faith.
This set of comments was from May 13, 2005:

Vanity links

Hi, James.

I should point out that it's bad form to link to your own web page all over Wikipedia. I left in one link (from voting system) because I think it's at least a better page than some of the other things that are linked to, but the links on other articles were excessive.

In general, it's better for external links to be added by someone who is not the author of the page.

RSpeer 16:56, May 13, 2005 (UTC)

Actually, the references are supposed to be published and peer reviewed. Even an academic home page does not qualify whether James does it himself or has an agent do it for him.--Fahrenheit451 21:25, 13 May 2005 (UTC)

Note the assumption that it would be an "agent" of the author who places the link. Not simply an editor who thinks the link useful! External links are not the same as references, which indeed must be specially qualified. Now this is old, and I'm only mentioning it here to note that the conflict between these editors is an old one.
However, ultimately, the link to the page was restored by ... Fahrenheit451! [14]
He made the comment on User talk:Hermitage: "James, I added the link to your home page to the voting methods page. It would be appropriate there I guess, but no where else in Wikipedia.--Fahrenheit451 14:34, 14 May 2005 (UTC)" So it is particularly odd that he deleted it a few days ago.
WP:EL does not prohibit linking to a "personal website" of the kind maintained by Hermitage, which is clearly set up as a public service to provide detailed information on election methods. The question of whether or not such a link is appropriate is a matter of consensus of the editors as to balance and the value of the site to readers; it is not true that external links must be "reliable sources," peer-reviewed, though they should not be deceptive or known to be carelessly compiled. Links even to partisan sites are routinely allowed *if* they are labeled as such so that the reader does not expect them to be unbiased -- and if the links are, as well, balanced, so that various sides to an issue are represented.
In the field of election methods, we have seen repeated situations where experts have chosen to develop and "publish" their work on mailing lists or web sites instead of the far more cumbersome and time-consuming process of submitting work to peer-reviewed journals. The paradox is that on a mailing list, there is *extensive* peer review, often more than involved with formal print publications, though, on the other hand, there is typically no clear *decision* made that would indicate approval of a "paper" so "published." I'm attempting to rectify this by starting the Election Methods Interest Group[3] as a Free Association using Delegable Proxy, about which Hermitage knows a thing or two; eventually we may see, from it, a kind of peer-reviewed publication that is more consistent with the hot medium of the Web. That is, decentralized, free (how many times I've been researching for an article here and the necessary publications would cost $30 just for a glance!), *but* still truly peer-reviewed.
My intention is to review the deletions or restorations of links to Hermitage's site for appropriateness. I think he knows now that his original placement of many links wasn't appropriate. I suggest to him that he refrain from editing those links, though I certainly do not mind that he has called our attention to the deletions, and I encourage him to hew closely to civility. It is not necessary to impugn an editor's motives to complain about an editor's actions. There is plenty else to do that can use his expertise! His site already gets prominent placement on Google for many topics....
As to Fahrenheit451, I assume of him good faith, but suggest that he be careful, as well, about civility and the seeking of consensus among editors.
Our common goal here is properly the creation of the best articles possible on the subjects involved, consistent with Wikipedia policy; to do this we need to use the resources available; sometimes the best is the enemy of the good, i.e., if insisting on the best means that the good is removed, without being replaced by anything better. Facts in articles are often asserted without sourcing, initially; indeed, this often happens when an article is originated by an expert! Ultimately, the possible Original Research is removed and asserted facts are properly sourced. The standards for external links, though, are quite a bit looser. External links are not required to satisfy WP:RS. Ultimately, the question is whether or not they are useful to the reader and not misleading or biased; and even misleading and biased sites can be linked *if* the apparent bias is labelled: a common example is linking to advocacy organizations and preferably to opposing ones if such exist.
--Abd (talk) 04:53, 30 November 2007 (UTC)
Hi Abd, and thanks for commenting. You must have looked at this pretty carefully, to find those old comments from my talk page! I generally agree with what you've said. I'll try to avoid getting personal, and I'll rely on editor consensus about where links to my site ought to be, rather than placing them unilaterally. --Hermitage (talk) 07:29, 30 November 2007 (UTC)
There are 14 articles with links to Hermitage's site: Condorcet loser criterion, Condorcet method, Instant-runoff voting, Instant-runoff voting controversies, Liquid democracy, Majority criterion, Minimax Condorcet, Mutual majority criterion, Participation criterion, Proportional representation, Ranked Pairs, Schulze method, Tactical voting, Voting system. That's too much. IronMan3000 (talk) 17:52, 3 January 2008 (UTC)

"Voting system" vs. "Voting method"

Is there agreement that there is a difference between a voting system and a voting method? Air Fortress (talk) 18:53, 25 March 2008 (UTC)

Comment about Bucklin removed from article.

I removed a comment about Bucklin voting taken from the article. This is the text involved:

In the United States in the early 20th century, various municipalities began to use Bucklin voting, but the results were not satisfying to voters. Bucklin is no longer used in any government elections, and has even been declared unconstitutional in Minnesota.

The source given is Tony Anderson Solgård and Paul Landskroener. "Municipal Voting System Reform: Overcoming the Legal Obstacles". Bench & Bar of Minnesota. Retrieved November 16, 2005.

Solgard and Landskroener appear to be activists promoting Instant runoff voting; the opinions they give in the article have not been confirmed by other legal sources, no citations are given for the claim that "the results were not satisfying to voters," and the case cited, Brown v. Smallwood (the complete text can be found with google), indicates quite the opposite: the court notes that its decision was unpopular, that the bulk of legal opinion was contrary to the court's decision, and so forth. Bucklin was used in at least 55 towns across the U.S., plus it was used for primary elections; in one case this continued for a long time. There is very little to be found on the net, from my searches, about what happened and why Bucklin systems were rescinded; but the same phenomenon occurred with what is now called "instant runoff voting," though the latter was never as extensive in use, in the United States, as was American Preferential Voting (Bucklin). I've come to suspect that the over-enthusiastic claims that it would be possible to find majorities without runoff elections may have been responsible. Sometimes Bucklin and IRV find true majorities, sometimes not. Only two Bucklin implementations are known to have been found unconstitutional: the Duluth, Minnesota, usage, in a decision which appears, from independent legal examination, to have been based on the very concept of preferential voting; this decision is at variance with what was found in other states. In Oklahoma, as well, a Bucklin system was tossed out by the state Supreme Court, but the basis for that was not the voting system but a rule that had been included that voided ballots if the voter did not rank at least one additional candidate in a three-candidate election. --Abd (talk) 21:32, 1 January 2009 (UTC)

Electoral fairness with respect to geography, regionalism and multi-ethnic federations and confederations

None of these articles seem to address issues of electoral fairness with respect to Electoral fairness with respect to geography, regionalism and multi-ethnic federations and confederations. There are many countries in the word where that are not very homogenous for some or all or even beyond the above issues. Some countries have "solved" issues and other countries have issues that could be resolved with an approach that moves beyond one man one vote or proportional representation. The US Senate is one example and the ethnic division of powers in Lebanon is another.

A discussion of the advanatges, disaavantages and the different approaches would be great. —Preceding unsigned comment added by 67.170.141.119 (talk) 05:27, 2 April 2009 (UTC)

Range voting satisfies IIA?

An editor recently changed the approval / IIA entry to "ambiguous", because "When it is ambiguous whether Approval meets Majority then it must also be ambiguous whether it meets IIA." This seems to be based on the logic of Arrow's theorem. But the way that Approval and Range get around Arrow's theorem is that they are not "universal". If there is a Condorcet winner (including majority winners), the Nash equilibrium is that that person wins. But if there is no Condorcet winner, then the Nash equilibrium is a mixed strategy - that is, perfectly strategic perfect information voters will determine their vote probabilistically, and the winner will be probabilistic. In this case, IIA is still satisfied for any given set of ballots, even though it is not satisfied for the whole game. I'd say a eeffee (light green) qualified "yes" is more appropriate for this situation than a white "ambiguous". Homunq (talk) 21:54, 14 June 2009 (UTC)

The majority criterion and IIA are incompatible. So when you argue that it is ambiguous whether approval voting satisfies the majority criterion, then you cannot claim anymore that it is clear that approval voting satisfies IIA. Markus Schulze 22:24, 14 June 2009 (UTC)

I'm not trying to claim it's clear. I'm trying to claim that a "qualified yes" is the most appropriate answer.

The basic question here is, are we talking about just the information that's on the ballots, or are we talking about what's in the voters' heads? If it's just ballots, then it's simple: Approval passes both MC and IIA (but doesn't violate Arrow because it is not "universal", that is, depends on a cutoff decision not contained in the preference orders Arrow considers). The Condorcet criterion would technically reduce to the Majority criterion, because if all that exist are two levels of preference, "approve" and "don't approve", then the pairwise margins are the same as the full approval vote margins. But really any talk of Condorcet is bogus because Condorcet ties are impossible. Of course, MC is especially trivial in this case - a "majority" candidate must not only be preferred, they must be the only candidate not completely unacceptable to a majority.

If you're talking about what's in voter's heads instead of just ballots, then there is an inevitable choice involved - how do I map the that onto numbers? Whether there are too few choices, as in Approval Voting, or too many, as in Range voting, it's not just a matter of voting honestly. Objectively, how do I feel about Barack Obama, on a scale of 0 to 100? Do I approve of him, or not? The questions are underdefined. The only way, at least as far as I know, to prove anything rigorous about that choice is Nash-style game theory: assume that it's made with perfect information and perfect strategic rationality. (Note that now, if there's a Condorcet tie, the correct strategy is often a "mixed" - that is, probabilistic - strategy; so with the same set of voters, there is an X% chance that A wins and a Y% chance that B wins, etc. This violates "universality" and puts it of the realm of Arrow's theorem, but it still happens to violate IIA.) In this case with 100% strategic voters, Approval and Range are identical: they meet the Majority, Condorcet, and local IIA criteria, but do not meet IIA.

So, how do we include this on the chart? Personally, I'd say that we should color by the answer for a ballot viewpoint (which I'll note below as +/- 1), but tint by the answer for strategy (which I'll note below as +/- .5). This is a fully consistent way of coloring them. We'd need different footnotes for every one distinct color there, and one footnote at the start of both rows briefly explaining the overall logic. Here's a draft, without all the footnotes:


Majority Monotone Consistency & Participation Condorcet Condorcet loser IIA Clone independence Reversal symmetry
Approval by ballot
by voter with strategy[4] 		Yes
yes[5]		 Yes Yes N.A.
Yes [6]		 N.A.
No		 Yes[6]
No(see local IIA note)		 Yes
Yes		 Yes
Range voting by ballot
by voter with strategy[4] 		No
Yes[6]		 Yes Yes No
Yes[6]		 No Yes[6]
No (see local IIA note)		 Yes
Yes		 Yes
  1. ^ [1]
  2. ^ [2]
  3. ^ [3]
  4. ^ a b Several of these criteria were first stated with regard to a full preference ordering for each voter, which is not necessarily present on Approval or Range ballots. There are at least two ways of extending the criteria to these systems: to consider only the information on the ballot, or to consider the preferences in the head of perfectly strategic voters with perfect information on each others' preferences. When these two interpretations disagree, or are the same but for different reasons, both answers are given.
  5. ^ see note
  6. ^ a b c d e NoIn Approval and Range, if all voters have perfect information about each other's true preferences and use rational strategy, any Condorcet or Majority winner will win in the Nash equilibrium. Laslier, J.-F. (2006) "Strategic approval voting in a large electorate," IDEP Working Papers No. 405 (Marseille, France: Institut D'Economie Publique)

One of these colorations could be questionable is Approval/MC. Some might argue that Approval can fail a version of the MC where the mathematician is omniscient but the voters are not. I'd say that is a strained definition and an even more strained scenario. If a majority prefer candidate X, in any realistic scenario, enough of X's supporters would bullet-vote to give her the win. It takes a lot less than perfect information or 100% strategy to achieve that. Anyway, I've left a white patch in the table to allow for this issue.

I'm not going to make these edits without some further discussion here. What do others think of my proposal?Homunq (talk) 03:01, 15 June 2009 (UTC)

What if someone insists that the Plurality/Condorcet box should be changed to "Yes", since there are scientists who claim that plurality always finds the Condorcet winner when the voters vote in a sophisticated manner? Markus Schulze 17:19, 15 June 2009 (UTC)
I don't have an RS for this, but such a claim about Plurality as Schulze suggests would not be borne out by a Nash analysis. The two-parties-everyone-hates Nash equilibrium for plurality is stable in all circumstances, in IRV it is stable in limited circumstances, whereas in Approval it exists but is always unstable. (The cited source does prove that the only stable Nash equilibrium for Approval/Range is the Condorcet winner if that exists, but says nothing about Plurality or IRV; that part is my OR, but it's not hard to check for Plurality at least.) Also, the standard criteria definitions and Arrow's theorem results work perfectly well for Plurality, and we all agree they require at least some extension/interpretation to apply to Approval and Range. Homunq (talk) 17:54, 15 June 2009 (UTC)
Response to Homunq: Strategic voting requires both accurate information about the preferences of other voters and an understanding of the best strategy for strategic voting. Typically both of these are limited. The distinction between what's in a voter's head and what's written on the ballot is not addressed by the various fairness criteria, and for good reasons -- including the fact that they bypass the rigorous mathematical precision with which they are defined. Either a voting method passes or fails each clearly defined criteria. If it makes sense to define new fairness criteria, that's a separate discussion. VoteFair (talk) 17:28, 15 June 2009 (UTC)
In that case, Approval passes both Majority and IIA, as any preference ordering which would make it fail those criteria is in the voter's head and not on the ballot. Yes, that leads to the idea that if you eliminate all but two approved or two disapproved candidates, a voter will always uselessly vote them as identical. This is not realistic but it is the only strict mathematical interpretation of what you're saying. If you want to include any doubt that approval passes both, then you are arguing for a strategic view of Approval and Range. The only way to make that rigorous is to go to perfect information and perfect rationality, which, I agree, is not plausible either. Reality presumably lies between these extremes, and both extremes are rigorously defined mathematically, so I'd argue for including both, with a smaller font size and tighter bounding colored rectangle for the strategic interpretation. Homunq (talk) 17:45, 15 June 2009 (UTC)
So, I'd be willing to accept the current article status for the "range" row, and the following colors for the "approval" row. I hope we can agree on this because the white squares are really just a copout. I think that as a whole the line below gives an inaccurately favorable impression of Approval, and so have made the concession of not putting the simple (and correct) answers of fully-shaded "yes" in the MC, CC, and IIA boxes. Any further concessions would be, I feel, not accurate - unless you expanded the row as above. Homunq (talk) 18:20, 15 June 2009 (UTC)
Majority Monotone Consistency & Participation Condorcet Condorcet loser IIA Clone independence Reversal symmetry
Approval [1] Yes[2] Yes Yes N.A.[3] No Yes[3] Yes Yes
I don't agree. The majority criterion and IIA are incompatible. A "Yes" for the majority criterion implies a "No" for IIA. Markus Schulze 18:22, 15 June 2009 (UTC)
Not true. I suspect you are thinking of a scenario like that posed on the IIA page. Such a scenario is simply meaningless on approval ballots, where "A>B>C" has no meaning (the ballot only says A>B=C or A=B>C or the perfectly valid A=B=C). Yes, that leads to silly A=B or B=C votes if you eliminate one alternative, but we are being mathematical, not realistic. A>B>C is meaningful if you include what's going on in the voter's head, but then, as I said, you can say nothing rigorous without considering rational strategy.
The ballot-only view does not violate Arrow's theorem because it's not a universal system by Arrow's definition. That is, Arrow's definition requires voter preferences, but in approval a given set of voter preference orderings can lead to different results, depending where voters set their approval cutoffs.
I could make the same exact argument about the Condorcet criterion, but it's even sillier, so I think a pink-shaded "N.A." is a good enough approximation. Homunq (talk) 19:45, 15 June 2009 (UTC)
The Condorcet criterion and the participation criterion are incompatible. So you cannot say that approval voting satisfies the Condorcet criterion and simultaneously that approval voting satisfies the participation criterion. Markus Schulze 00:01, 16 June 2009 (UTC)
I just did say that. Please make an argument as to why you believe this, instead of just repeating it. Arrow's theorem does not apply here, because it requires universality, and I've explained several times now how that's broken. The Condorcet criterion based on Approval ballots is just the win condition. "Pairwise beat any other candidate" means "get more approvals than any other candidate", since A>B>C is impossible on Range ballots.
Yes, these criteria are silly when defined down to just what's visible on range ballots. So there's another way to apply them: assume the voter has a preference list in their head and let them use perfect information and perfect rationality and the Nash equilibrium strategy. In that case, Arrow's theorem applies again.
I would be happy with either of the tables I've proposed so far. I would also be happy to listen to any reasonable argument. But if you just keep restating Arrow and related results (G-S, etc.) we're not going to get anywhere, because many of these results do NOT APPLY to range ballots. Homunq (talk) 01:44, 16 June 2009 (UTC)
Actually, I don't see any meaningfull arguments in your comments. Markus Schulze 08:35, 16 June 2009 (UTC)
I don't see any arguments at all in your comments, just statements. You can argue that I'm wrong, argue that you're right, or argue that I'm incomprehensible/not even wrong, but you have to argue something.
My "ballots only" view of Approval is that a voter is nothing but a function from a set of candidates to an approval ballot (subset of that set). Each voter has a subset V of all possible valid candidates which they "approve", and their ballot for any set of actual candidates E is simply the intersection of V and E. Approval ties are broken deterministically - say by a comparing a hash function of concat(candidate unique name,binary sequence of votes for that candidate sorted by unique voter name). Here are my proofs:
Majority criterion for Approval ballots. If a majority (subset W of electorate) of ballots say A beats B pairwise for any B, then those ballots must be approving A and not approving B. Thus, A's approval score is over 50% (measure of W + any voters in ~W who approve both) and B's is under 50% (measure of ~W - any voters in ~W who disapprove both) for any B. Thus A wins. QED
Condorcet criterion for Approval ballots. If A beats B pairwise for any B, then A has more approvals than B for any B. This is the definition of winning on Approval ballots. QED.
Participation criterion for approval ballots. If A is winning and an extra ballot causes B to win, that ballot must approve B and not A. Thus that ballot prefers B to A. QED.
IIA for approval ballots. If A wins an election among candidate set E and electorate X, then each candidate Ci has ci approvals in X, and the winner is candidate n. ((cn > cm) or (cn = cm and hash(Cn,votes) > hash(Cm,votes)). Assume IIA does not hold for extra candidates F. Then another candidate M in E wins the election (E+F,X). So cm > cn or (cn=cm and hash(Cm, votes) > hash(Cn, votes). Contradiction and QED.
Arrow's universality for approval systems. The statement is that for any set of voter preference orderings over a set of candidates, the system gives a unique answer. However, take the two-voter two-candidate election where voter X prefers A>B and voter Y prefers B>A. Voter X's possible approval votes are (1,1), (1,0), and (0,0); voter Y's possible votes are (1,1), (0,1), and (0,0). But ((1,0),(0,0)) gives winner A and ((0,0),(0,1)) gives winner B. Thus we have a counterexample, and approval is NOT an Arrow-universal system. QED.
Homunq (talk) 14:15, 16 June 2009 (UTC)
With the same argumentation you could say that plurality satisfies the Condorcet criterion.
The fact, that the voters cannot cast preferences, doesn't mean that they don't have preferences. Markus Schulze 15:23, 16 June 2009 (UTC)
You cannot construct a plurality voter model like the range voter model above, able to vote consistently in any election, without an internal preference ordering. And conversely, and more tellingly, you cannot construct an approval voter able to vote consistently in every election with just preferences, you must add some way of setting their approval cutoff or they will vote indeterminately.
So, which is more natural: a model like the one I argued, in which they have an intrinsic cutoff; a model in which they have some arbitrary cutoff rule (approve of half of the candidates rounded down, or some such); or a model in which they have perfect information and rationality and choose a Nash cutoff rule? In the first model, Approval satisfies ALL of the given criteria; in the second, at worst NONE of them (as Saari argues, approval can, for most voter preference sets, elect any given candidate if voter cutoffs are arbitrary); and in the third, it fails only Condorcet Loser (probabilistically in the worst case, and with asymptotically 0 probability as voter number grows above candidate number, but still a failure) and IIA (in the same circumstances and with the same qualifications as Condorcet Loser).
My contention is that we must either choose one of these models (Occam's razor would choose the first) and consistently score Approval on that basis; give separate results for at least two, perhaps all three, models; or leave the whole line blank and explain our dilemma. I would be happy to accept any of these solutions, though I think the third is a cop-out.190.149.24.202 (talk) 08:39, 17 June 2009 (UTC)
I agree with Schulze that Approval voting cannot satisfy both the majority and IIA criteria. As Schulze points out, Approval voting does not allow a voter to express their real preferences. In fact this is supported by your/Homunq's statement above: "any preference ordering which would make it fail [Majority and IIA] is in the voter's head and not on the ballot." This reinforces our point that a voter cannot really express their full preferences. A related perspective is that -- unless the voter really regards two (or more) choices as equally acceptable, or there are only 2 choices -- Approval voting always requires strategic voting, which means that sincere voting is impossible . VoteFair (talk) 17:05, 16 June 2009 (UTC)
I agree with the above, although I would point out that on the other hand, Warren Smith has proven that Approval and, more generally, Range systems are the ONLY systems (except possibly some unknown non-ordinal systems) where perfect strategy is always semi-sincere for 3 candidates, that is, where it is never strategic to rank a lesser preference above a greater preference (as opposed to simply even with it). ((He has further demonstrated a case where even Approval has a non-semi-sincere perfect strategy for 6 candidates; as you might suspect, it is quite a contrived example, and it also requires either an expectation by one voter of perfect collusion for all other voters, or exponentially perfectly-balanced utility distinctions for all voters, as voter number goes up.))
Anyway, none of that helps us score it on the criteria. Choose a single consistent model (I'd accept all-green or full-strategy, not all-red); give results for 2 or 3 models, as in my first proposal above (I'd accept adding a third subrow which was all-red); or leave the entire row white? 190.149.24.202 (talk) 08:39, 17 June 2009 (UTC)

I put the fix in on Approval_voting#Compliance_with_voting_system_criteria instead of in this article. 186.1.16.35 (talk) 14:10, 19 June 2009 (UTC)

Own method

I have developed my own electoral system that can be used in presidential elections. It is very is similar to the system used in Burlington mayoral elections. The system consists of a mix of prefential voting, IRV and a normal run off system. I would like to to create a wikipedia page on my system but am unsure if I would b able to because there would be no citations or references (because the system is developed by myself). Anyone know if i would be able to do this? —Preceding unsigned comment added by Supun47 (talk • contribs) 09:45, 26 July 2009 (UTC)

Probably not but, you could do something with the Burlington mayoral elections. What exactly is this system anyway? Mike92591 (talk) 20:13, 26 July 2009 (UTC)

The system works where voters give a 1st and 2nd preference vote. To be elected, a candidate must recieve a majority of the vote. If nobody recieves a majority, the candidate with the least amount of votes is elminated and their second prefernce votes are distibuted. If after distribution no candidate has a majority then the candidate with the least number of votes is once again distbuted and so on until a candidate reaches a majority. Now under this system, some votes may exhaust. So it is possible that even after all perferences are distributed that no candidate has a majority. If this occurs, then the two leading candidates AFTER PREFERENCES go to a runoff in a seperate election. This in my view is the fairest election system though im sure a number will disagree. Here is an example of the system...

Theres five candidates, Albert (a liberal), Michelle (a conservative), Chris (a conservative), Steve (a moderate) and Sarah (a conservative). In the presidential election 100 votes were cast with 51 votes needed for the presidency.

Results of election

Candidates 1st Round 2nd Round 3rd Round 4th Round
Candidate Ideology Primary total Votes distributed Progressive total Votes distributed Progressive total Votes distributed Final total
Albert Liberal 37 0 37 3 40 2 42
Michelle Conservative 34 1 35 2 37 10 47
Chris Conservative 21 0 21 1 22
Steve Moderate 7 0 7
Sarah Conservative 1
EXHAUSTED PILE 0 0 0 1 1 9 21
TOTALS 100 1 100 7 100 21

EXPLANATION: In this election, virtually all liberal voters voted for Albert, who is ahead of all candidates but doesnt have a majority. The conservative vote has been split between the three conservative candidates. Once all votes are counted, Sarah is elminated and her one vote is distributed to Michelle, a conservative. Next to be eliminated is Steve and his votes are distributed. Still, no candidate has a majority. Next and last to to be eliminated is Chris. Most of his 21 votes are distributed to Michelle and she is able to go ahead of Albert. She however, does not have a majority. Two votes from Chris go to Albert. The other nine of his votes exhaust, because they preferenced candidates who have been eliminated. Most likly, those votes went to Sarah who is a conservative. The result is Michelle on 47 votes and Albert on 42. Becuase neither candidate recived a majority, the election will go onto a runoff between Albert and Michelle. Note, that a runoff occurs when no candidates recieve a majority AFTER ALL PREFERENCES, and that the candidates in a runoff will be the ones with the highest votes AFTER PREFERENCES. So for example, it would have been possible for Albert to not reach a runoff even though he finsihed 1st on primary vote. Ill add that under the system, it would be very rare for runoffs to occur. Runoffs would only be caused by major vote splitting and/or a large number of candidates. I chose to do this because vote splitting can still pose problems in a normal runoff system. A good example is the French_presidential_election,_2002. In the election, the left vote was split amongst several left wing parties, resulting in neither party reaching the run off. Under my system, a runoff would probably have occured, between the major centre right party vs major centre left party.

I would like to hear any comments or feedback on the system from fellow wikipedia users.


I would call this method "2-slot IRV with a runoff". This method satisfies the majority criterion and the Condorcet loser criterion. Markus Schulze 14:28, 1 August 2009 (UTC)
Archive 1Archive 2Archive 3Archive 4Archive 5
  1. ^ Several of these criteria were first stated with regard to a full preference ordering for each voter, which is not necessarily present on Approval or Range ballots. There are at least two ways of extending the criteria to these systems: to consider only the information on the ballot, or to consider the preferences in the head of perfectly strategic voters with perfect information on each others' preferences.
  2. ^ see note
  3. ^ a b Cite error: The named reference approvalnash was invoked but never defined (see the help page).
Retrieved from "https://wiki.riteme.site/w/index.php?title=Talk:Electoral_system/Archive_2&oldid=767639925"