Wikipedia:Peer review/Instant-runoff voting/archive1

Instant-runoff voting
This article seems to be almost ripe for a Featured Article nomination. Please let me know if there is any way in which it falls short of the criteria. We expect to be getting more images soon from FairVote, so that should help with criterion #3. Thanks, Captain Zyrain 06:52, 14 October 2007 (UTC)

Review by Jeff
Ok, a tough topic because I'm sure it gets disrupted by POV pushers. I vaguely remember this article from a while ago, I see that it's changed a lot since then. It's looking fairly good, not too many neutrality issues. Here are some specifics:
 * Of course, more refs please, and check them for accuracy/neutrality. I'm sure you already know this one.
 * The lead is pretty good, but I would say you need to add a sentence regarding the history. Perhaps something like "First proposed in 1870 by American architect William Robert Ware, Instant runoff voting..." The history section should probably say why he came up with it, and maybe flesh it out with information about the history of its adoption.
 * Turn some of those lists into prose, such as the pro and con section. They should be written in sentences organized into paragraphs. Make sure to point out that it is the adoption of such a voting system that is supported or opposed.
 * Evaluation by criteria - this section needs to be fleshed out. Use complete sentences and paragraphs to explain the concept. Give the reader sufficient background to understand the concept without going into unnecessary detail. (OK now I'm really starting to sound like a page out of the manual of style)
 * Explain (but don't elaborate upon) jargon at the first possible opportunity. Examples include First past the post, Majoritarianism (?), Condorcet, tactical voting, exhausted ballot etc. Explain the concept once the first time it is used, and only once. There is no need to explain it again every time it occurs.
 * I would cut down/simplify on the Similar systems section. It's OK to mention comparisons to the IRV system, but that comparison should focus IRV and how it differs from other systems, and how the outcomes differ, rather than bogging us down with procedural details of systems explained in other articles. Keep the focus on IRV. Reading through Similar systems, I felt confused.
 * "What happens in practice in Australia is a simplified count is sent through to..." yikes.
 * "The common way to list candidates on a ballot paper is alphabetically or by random lot, a process whereby the order of the candidates published on the ballot paper is determined by lottery. In some cases candidates may also be grouped by party." Simplify. How about "Candidates may be listed on the ballot by party affiliation, alphabetically, or randomly." Keep it crisp and concise.
 * "However, if this election were in Vermont for governor under proposed IRV legislation there, if there is no majority winner, the election would to the Assembly, to be decided among the top three by secret ballot [citation needed], so the fact that the election was IRV would have been moot." I don't understand how this is relevant.
 * "IRV produces representation very similar to those produced by the plurality system" here is where this article could use some expansion, try to cover how this voting system has worked out for the countries that use it. Is it pretty much the same? Is it substantially different? Advocates and Opponents give us long lists of how it is better or worse, but how has the system worked out in practice? Mention in the introduction!
 * Check for duplicate material that is explained/discussed several times. I didn't get the feeling like I was reading a well-organized article. The content seemed to jump around a lot. Consider writing out an outline from scratch, and then moving the text around to fit the outline. Over time, a good outline can become eroded by lots of little edits.
 * "Some of these arguments may may be false or otherwise misleading" Oh boy, this is just what we need on wikipedia...
 * References should follow after the punctuation. Esp in the pro and con section for a place that needs work.
 * "Scholars of electoral systems often compare them using mathematically- " them?
 * I know you have probably made a good effort to keep it balanced, and not too heavy on the United States, but the list of US IRV elections in the History and current use section needs work. If you choose to use an example, from the US or anywhere else, explain why it is important. Was it the first one in that country? Were there mass protests? Did it go smoothly? Was the outcome unexpected? A dry list of details of every place it was ever used is mega-boring.

OK that should get you started. Jeff Dahl 18:47, 14 October 2007 (UTC) I have to disagree that the article is anywhere near ready for Featured Article status. It was edited and maintained for a long time, and vigorously defended against change to make it more balanced, by a few sincere supporters of IRV, plus sock puppets and, in fact, the Executive Director of FairVote, see the history of the block placed on the article which is still there as of today. The anonymous and liberal use of Undo, more than violating 3RR, was from Rob Richie, and he acknowledged it. Unless, of course, someone is imitating him, which would be difficult. Lots of people have personal email from him with the same IP, that's how we identified him. But instead of heavily revamping the article, and in spite of complaints from Richie that now I would have a "free rein," I made only modest edits, pending something more important: the development of a template for election methods articles that applies evenly to all of them, so that the very structure of the article cannot so easily be manipulated to bias how readers will see it.
 * I'll just toss in one last piece of advice from FA director Raul654, which he appropriately calls Raul's Razor: "An article is neutral if, after reading it, you cannot tell where the author's sympathies lie. An article is not neutral if, after reading it, you can tell where the author's sympathies lie." Good luck with the article. Jeff Dahl 04:06, 16 October 2007 (UTC)

The comment about the arguments made by the reviewer is necessary if the arguments are to be presented at all. Here is the reason: many of these arguments are very carefully framed by activists such that they seem reasonable at first glance. If we present the arguments in a rigorously neutral and balanced way, they become quite wordy. By reporting the arguments raw, without comment, we keep it simple, and we frame it with a warning. I agree that it isn't the best way. However, it enabled us to start listing the arguments, otherwise we'd have to simply cut them all out, and I repeatedly gave that as an option. Captain Zyrain has started a stub to do that, but has gone ahead without understanding that what should be done -- as a user who objected to the stub being formed suggested, if I recall correctly -- was to discuss the Controversy. That is, I'd call the article Controversies around Instant Runoff Voting. And each claim would be examined in detail; that is, what might be, for example, a Pro argument would be immediately balanced with Con arguments that relate to it, explained. We really should be able to agree on this, *unless* some of us have an outside agenda and insist upon it. This *is* the situation we had, very clearly, and it remains, just no sock puppets as far as I can tell. When the block comes down, we'll see: one of the continuing socks, in my opinion, an SPA formed at the right time, as User:BenB4 was about to be blocked, has been using reverts on the Approval Voting article.

What are the issues remaining? Let's see: I attempted to make the reference to Robert's Rules alleged "recommendation" more accurate, it was undone and removed for the most part. I attempted to move it to the middle of the article, because having it in the introduction is precisely where an advocate of IRV would want it to be, Robert's Rules has cachet, but for the mention to not be misleading takes more words than appropriate in an introduction, so being in the middle is better, but that was blocked.

The very name "Instant Runoff Voting" was politically crafted, I believe by FairVote. It makes implications about the method that are, at the very least, controversial. Note that Robert's Rules does not mention "Instant Runoff Voting" by name. You'd never have known that when the article was purely as FairVote wanted. It's "preferential voting." Is it the *same* as "IRV"? Well, RR actually is referring to a class of methods, not just the one that it describes, and it actually expresses some reservations about the whole class, it's a compromise. It's actually *not* recommended, but it *is* possible. Then it describes one specific method. Is that IRV? Yes, it is *one* of the forms of IRV, that is, there are forms being advocated. In Vermont, the form advocated was, as near as I can tell, the *same* as the form in Robert's Rules. But that is not what was described in the article, it is different: if there is no true majority winner, it fails and the question is then resolved with further process. I try to put this in the article, and it is taken out. The article was being maintained by a cabal of editors. One of them is still active. Captain Zyrain is not one of them, but he is, from my point of view, naive about the political implications of the article and the arguments.

I'm *not* aiming for an anti-IRV article, contrary to hysterical claims from some. I'm aiming for an NPOV article, developed by a consensus of editors, on a topic which is hotly contested in the public arena. That's not an easy task, but I believe we can do it. However, my opinion is that we need to establish some principles. The whole question about how controversy is handled is tricky. The basic principles are clear: controversial statements are sourced. But sourcing isn't enough. How the arguments are presented must be balanced, and that is not so easily defined.

There are more issues, indeed, but I'm only one editor, working on many other projects. I've been attempting to attract experienced editors, familiar with election methods, to the article. User:Scott Ritchie has said he will help, though I don't know how much time he has. I've deliberately avoided recruiting opponents of IRV, but I have mentioned the existence of problems with the article on the Election Methods mailing list. I acted to make it clear that User:TBouricius was not part of the abuse, helping to get his block lifted, even though he is a FairVote consultant and co-author with Rob Richie. I believe that all points of view must be represented and the people holding them must be a part of the consensus that creates a truly great article.

One more comment: the reviewer referred to this: " 'However, if this election were in Vermont for governor under proposed IRV legislation there, if there is no majority winner, the election would to the Assembly, to be decided among the top three by secret ballot [citation needed], so the fact that the election was IRV would have been moot.' I don't understand how this is relevant."

I agree that this was incompletely explained. However, there is a critical issue here. IRV is being sold on the grounds that it allegedly guarantees a majority winner. If it is done as described in the article, it does, but through a trick. Imagine you have an election between 3 candidates. One of them gets 41% of the vote, one gets 39%, and the third gets 20%. None of them have a majority. Under Plurality the first one will win. Now, make this an IRV election where no voters choose to add second rank votes. Magic! The third candidate is eliminated and the first one still wins with a majority *of the remaining ballots.* In Vermont, no. There is no majority winner (which is what Robert's Rules would assert, clearly), and so the election, as provided in the Vermont constitution, goes to the legislature for resolution. And any of the top three can win. The devil is in the details. Is this some minor point with no effect in real elections? No. In San Francisco, a number of races have been won by candidates who had less than 40$ of the legal votes cast. Supposedly a big selling point for IRV was that it would eliminate runoffs. It did. And, at the same time, it eliminated an important element of the democratic process, which would be the ability of voters to reconsider their votes and actually make a new choice between the top two, when none enjoy a majority. Abd 00:44, 16 October 2007 (UTC)


 * Please see automated peer review suggestions here. Thanks, APR t 13:47, 16 October 2007 (UTC)