Wikipedia talk:Supermajority

Template deletion threshold

The reason I ended up creating this page is because I asked what the deletion threshold for a template should be^(archived) after seeing one deleted with less than 66% opposition. However, I asked in perhaps the worst place to ask that question, and didn't get any actual answers to the question.

Therefore, I am repeating my original proposal here:

A template should be deleted if it is opposed by at least:

• 67% of voters when two or more editors are using the template;
• 60% if just one editor is using it; and
• 50% if it isn't being used by anyone.

Comments? --James S. 11:35, 15 January 2006 (UTC)

• This proposal is pretty ridiculous. Can't each situation just be judged individually on the merits of the rationale? For TFD rationale, If 50 people say "it's ugly" (something that is very subjective, but can easily be fixed through further discussion on the template talk page), and 5 people say it is useful and should be kept, I would keep it in such a situation. — ₀₉₁₈^BRIAN • 2006-01-15 14:00

By all means, I agree with your specific example decision. However, the situation which caused me to ask this question had to do with a userbox which advocated the fair use of small copyrighted images in userboxes. The only reasons against it were that it was expressing an opinion about a position which may or may not be contrary to copyright law in some cases but not others. There was a consensus that nobody, not even lawyers, could honestly predict the legality of the position being advocated by the userbox. There was also a consensus that the userbox was merely an opinion. It was deleted with less than 65% opposing it. I for one don't care about the userbox and would never put it on my user page, as I prefer that fair use images stay far away to keep Wikipedia free. But the deletion process seemed really weird, especially when a few other userboxes passed with more than 50% opposition, so I figured the two-thirds AfD tradition was in effect.

I have certainly learned a lot from asking the question, and I hope you approve of the way I've followed your advice. However, I continue to believe that there should be a two-thirds requirement for templates when they are in widespread use. I don't think I would want to try to create any templates if there wasn't. --James S. 14:16, 15 January 2006 (UTC)

You're going to be constently disappointed if you believe setting arbitrary numbers is better than having someone read the comments and parse out a course of action. Some "votes" are weak, and some are strong. Also, over the course of 5-7 days new information comes to light that the early voters may not be aware of. -- Netoholic @ 14:17, 15 January 2006 (UTC)

I am not proposing that. I understand the value of the tabulator's discretion. I'm only saying that given the controversy surronding userbox templates, which seems to have been heated for a couple weeks now, there should be some kind of a standard for the tabulator to make close calls without being forced to make an arbitrary subjective decision. --James S. 14:22, 15 January 2006 (UTC)

Well, I don't understand the proposal. You agree that votes have differing weights depending on who's voting and their rationale, but advocate having numeric guidelines to follow? When, then, would these numbers ever be used? If anything, I think that the use of numbers like this would only increase the number of inappropriate decisions. Perhaps what we should be discussing is what an admin should do when closing TfDs is too close to call. --Qirex 14:50, 15 January 2006 (UTC)

I agree, but in the short time I've seen probably several dozen AfD, TfD, and other votes, I can say that the number of both Strong and Weak votes on each side have been remarkably similar, which is to be expected when the votes are normally distributed, as almost all human votes tend to be. (The same goes for the slightly less common Strongest Possible Swift and Speedy... votes.) I do consider this a discussion about what to do when the votes are too close to easily call, and if I have not sufficiently specified the proposal to reflect that, then I will do so now. Note that the administrator guidelines and policy for deletion currently specify relisting when the outcome is unclear. I agree with that, but after a few times, it makes much more sense to draw the line than to keep wasting people's time with a deadlock such as the diacritics problem below. --James S. 14:58, 15 January 2006 (UTC)

In such a case as you outline, I'd believe that the admin would take a decision that would roughly correspond to your numbers. Any admin would and should be willing to explain his rationale for closing the votes, whichever way he does. However, formalizing these numbers is troublesome, and imo, instruction creep. --Gurubrahma 15:03, 15 January 2006 (UTC)

This is a proposed guideline, not a policy. I hope to give the community something to help reassure them that arbitrary descisions aren't being made, just as much as I hope to keep admins from derision by giving them something to justify their close call, to keep from being accused of favoritism, among other things. And as the text of the proposal makes clear, I fully support ignoring it for any good reason. --James S. 15:34, 15 January 2006 (UTC)

Getting confirmation as to whether an admin has improperly closed a deletion debate is wat Wikipedia:Deletion review is for. Can we just redirect this page to Wikipedia:Consensus and acknowledge that it's not needed? -- Netoholic @ 18:24, 15 January 2006 (UTC)

Isn't a deletion review a whole lot more hard work than being able to point to an objective cut-off in close calls? I'm trying to help here. I understand that challenging long-held viewpoints which worked fairly well when there were 1/10th as many users is potentially considered rocking the boat. There is plenty of evidence that some of the current, well-respected, long-serving closing procedures aren't scaling too well. --James S. 22:53, 15 January 2006 (UTC)

Diacritics in article names

Here is a great example of the need for objective ways to judge rough consensus in the absence of actual consensus, from WP talk:Consensus:

Let's discuss a practical example. The Wikipedia community is split on the use of diacritics in article names, but pages still have to be named. How does one build a consensus on such an issue as the number of people involved in the debate runs to more than 100 and the number who have changed their opinion on the issue could be counted on one hand? The problem is not so much new pages, but the moving of pages that already exist to fit in with one guideline or another because no wording which would unify the guidelines can be agreed upon. So we end up with debate after debate, see the Wikipedia:Naming conventions (Norse mythology) for a good example of the consequences of this. If you can suggest how such a contentious issue can be sorted out without discussion of the number of participants which go to make up a rough consensus, I would be most interested to read it --Philip Baird Shearer 13:40, 15 January 2006 (UTC)

Mediation has been confirmed as a better way to proceed under those kind of deadlock conditions than repeatedly voting. --James S. 02:26, 16 January 2006 (UTC)

Simply put

This page is not, will never be, and should never be given any level of credence. Wikipedia is not a democracy, and there are no hard and fast numerical thresholds - closing admins and bueracrats are always and have always been empowered to weigh the merits of arguments as well as the numbers of people making them. This is a fundamental aspect of how Wikipedia works. Phil Sandifer 20:22, 15 January 2006 (UTC)

There is no question that several dozen administrative votes of one kind or another occur each day here. Ultimately I defer to those who aren't new here like I am. That does not mean that I believe more than 60% of them agree with each other on this issue, though. I will refrain from calling for a vote for the time being. --James S. 22:55, 15 January 2006 (UTC)

Merge tag

The reason this page was created a few days ago was to get the supermajority numbers out of the Wikipedia:Consensus page, by request on Wikipedia talk:Consensus. Therefore, I'm removing the merge tag. --James S. 09:31, 17 January 2006 (UTC)

• Yes, but ironically that request was not based upon consensus. 1) guidelines are DEScriptive. 2) those numbers are widely in use on e.g. RFA/AFD. 3) RFA/AFD work on consensus, not supermajority. Hence, 4) the consensus page should describe the status quo. Radiant_>|< 11:39, 17 January 2006 (UTC)

Reciprocity

Whatever majority is decided to put into effect for AfD, the same majority has to be used for DRV. User:Zoe|^(talk) 00:09, 18 January 2006 (UTC)

I disagree. The numbers for DRV were considered rather carefully when it was crested out of VfU. DRV is not supposed to be, and usually is not, a re-run of XfD. DRV is supposed to consider primarily process errors. If a simple majority of those expressing an opnion at DRV feel that the process was in error that is a good reason to re-run the process. Only if a large supermajority (currently set at three-quarters) belives that both process and outcomne were wrong is DRV supposed to simply undo the closure without relisting in the appropriate forum. DES ^(talk) 22:28, 30 January 2006 (UTC)

Have this page redirect to Wikipedia:Consensus

It seems that all the information on this page is in Wikipedia:Consensus except for the specific percentages that are being used for supermajority. I can think of two good reasons to have this page redirect to Wikipedia:Consensus. First, having the specific percentages stated on a centralized wikipedia page makes it seem that there is broad agreement in these numbers. I don't think this is the case. Second and more important in my view, is that any discussion of Supermajority should always be framed with discussion about how it is just a tool to help reach rough consensus and not a process in and of itself. We don't want to send the message that Supermajority by itself is the method by which decisions are made.

I know that this page was created to try and help reach a consensus at Wikipedia:Consensus but since we now include the range of percentages on that page, is this one needed? -- Samuel Wantman 21:33, 29 January 2006 (UTC)

Issues are not decided by supermajority, they are decided by consensus. *sigh* Kim Bruning 11:08, 30 January 2006 (UTC)

• The logic behind this page seems to be thus... (1) People on Wikipedia:Consensus are opposed to writing down numbers. (2) Those numbers are in fact heavily in use all throughout the Wiki (even if most sensible people knows that they're not strict but only rules-of-thumb). Therefore, it logically follows that (3) those processes that heavily use those numbers are not in fact working on consensus, but on something else. Hence this page. Of course, the flaw in the reasoning is point (1), and once more it goes to show that if a guideline avoids explaining the status quo, people will create their own explanations. Radiant_>|< 11:48, 30 January 2006 (UTC)

This Means Well, But...

Perhaps instead, we should focus on what each "vote" is worth in a standardized value, both qualitative and quantitative values. (sociocracies such as Wikipedian project space require both.) Karmafist 06:52, 19 February 2006 (UTC)

Interesting proposal. I'm still pretty new here, but most of the votes I've observed come in three basic varieties: self-identified normal, "strong" and "weak". In the usual case, those votes are spread on the same distribution on each side of the outcome. So, in the usual case, just the raw numbers ought to do the trick. In a case where the proportion of "strong" or "weak" votes are different on each side of the issue, then the person determining the results can weigh them to "push" a close vote to the appropriate side.

Since supermajority is propsosed as a mere guideline, that seems like it would do the trick. However, (and this is somewhat ironic) since the whole point of the supermajory guideline is to explicitly quantify what has previously been implicitly quantified, then the best way to set the "exchange rate" of "strong" normal and "weak" votes would be a statistical review of prior outcomes. That's too much work for me, so let me suggest another metric: every self-identified "strong" vote counts as equal to two "weak" votes on the other side. Otherwise, all votes are counted equally (that is, left-over "strongs" and "weaks" count the same as a normal vote). If you ant to make it a little more complicated, you could use a 3:4 ratio of "strong" to normal, and a 2:3 ratio of normal to "weak" as equivalence ratios.

There are other ways to weigh votes, such as early votes vs. late votes, or based on the length of time someone has been an editor on wikipedia, or the number of edits they've made. In the past, I've observed only one method of weighing such distinctions: an arbitrary threshold is set, and votes from those editors who meet the criteria are weighted in full and equally to one another, and those who do not make the threshold are given a weight of zero. I don't see any reason to change this system. Thesmothete 03:57, 11 March 2006 (UTC)

Let's speedily get rid of this NONSENSE proposal

This article states the equivalent of "A Supermajority poll is a process for attempting to ascertain if a consensus exists when it is clear that a consensus does not exist." This violates the basic rules of logic. This proposed policy is PATENT NONSENSE. pat8722 22:37, 15 April 2006 (UTC)

Any objection to redirecting to Wikipedia:Consensus ?

Objections? - Stephanie Daugherty (Triona) - Talk - Comment - 12:25, 12 September 2006 (UTC)

I think it's better off as an explicitly rejected page. That way people wo do link to it can be shown the errors in their ways if they think Consensus = Supermajority voting. -- nae'blis 21:16, 12 September 2006 (UTC)