Wiktionary talk:Votes/pl-2009-08/Voting eligibility 2

Definition from Wiktionary, the free dictionary
Jump to: navigation, search

maybe we should add Concordance too? -- Prince Kassad 12:44, 17 August 2009 (UTC)

Perhaps. But there are only 50 pages in that namespace apparently, with very little activity, so it wouldn't be much of a loss. But I'll add it. --Ivan Štambuk 13:48, 17 August 2009 (UTC)

With the introduction of sub-votes, is one user entitled to vote in one or in both of them? The uſer hight Bogorm converſation 14:36, 17 August 2009 (UTC)

Both of them. The one which gains more support will pass. --Ivan Štambuk 17:47, 17 August 2009 (UTC)

I don't think we should specify the namespaces, but rather say something like, "contributions to content pages (entries, appendices, 'Rhymes' pages, etc.)". I say this firstly because our list is currently missing three content namespaces (Concordance, Index, and WikiSaurus); secondly because we do add new namespaces occasionally, and they should automatically get counted (or not, as appropriate); and thirdly because the Template: and Category: namespaces are split between content pages and non-content pages, where the former should count and the latter (under this rubric) should not.
Better yet, we could dispense with this whole restriction, and just say "contributions (not counting votes, comments, vandalism, and user-page edits)". After all, if our threshold is just 50, then when there's doubt about it, we can take a minute to look through the contributions.
RuakhTALK 22:59, 17 August 2009 (UTC)

But it's so much easier and clearer to count when the definition is clearly restricted to certain namespaces. For the record, I wouldn't count Index as a content namespace, btw (it's pretty much just a second Category namespace). -- Prince Kassad 23:46, 17 August 2009 (UTC)
I've added the phrasing "content namespaces" and listed those which were previously explicitly mentioned in the brackets. So if new content namespaces get introduced, we only need to agree that they're "content namespaces", and they get included. I've omitted the Index namespace because it's mostly another type of categorization, as Kassad says (tho I've personally wasted quite a few hours there), mostly maintained by the bots. --Ivan Štambuk 13:43, 18 August 2009 (UTC)


In my opinion the vote should be structured as follows:

  • Support Option 1
  • Support Option 2
  • Oppose (both)
  • Abstain

This makes the vote much clearer and more unambiguous. Any comments? -- Prince Kassad 23:47, 17 August 2009 (UTC)

Symbol support vote.svg SupportRuakhTALK 00:44, 18 August 2009 (UTC)
Done. --Ivan Štambuk 13:37, 18 August 2009 (UTC)

Approval voting inappropriate.[edit]

I should have thought of this sooner, but approval voting is actually the wrong approach here. We have a "more severe" option 1 and a "less severe" option 2, where it's impossible to prefer option 1 over the status quo without also preferring option 2 over the status quo. So, people should vote for the one they prefer, with the winner determined thusly:

  • if {option 1} has consensus (relative to {option 2} + {oppose both}), then option 1 wins.
  • otherwise, if {option 1} + {option 2} has consensus (relative to {oppose both}), then option 2 wins.
  • otherwise, neither wins, and the status quo remains.

In approval voting, the only way option 1 can win is if its supporters "game the system" by pretending they don't approve of option 2 — which is a risky approach, because it might end up causing option 1 and option 2 both to fail. (This is already happening, though I assume it's because people are accidentally gaming the system by not understanding how approval voting works.)

RuakhTALK 13:58, 19 August 2009 (UTC)

(To clarify, real approval voting wouldn't have this problem. The problem arises because we use a modified form of approval voting where we use "Oppose both", a default position that can only be overridden by consensus, rather than a normal "Status quo" option that competes on a level playing field with the other two.) —RuakhTALK 14:03, 19 August 2009 (UTC)
(I was trying yesterday to put my finger on what was wrong with the vote; thanks for unwittingly lending me your finger, Ruakh. However:) Because the vote was explicitly set up not the way Ruakh describes but rather so that the option with the largest number of votes wins if it also has a requisite percentage of votes, people should vote for only Option 1 if that's the one the want most, as, if they vote per approval (hence for Option 2 also) it's bound to win. (Unless we decide now — mid-vote — to change the decision-making process to the one Ruakh outlines. I don't mind doing that as long as we inform all prior voters (on their respective talkpages) and make sure they've seen it (viz, have edited since they were notified). Thoughts?)​—msh210 16:55, 19 August 2009 (UTC)

There should really be some example vote with the multiple-options structure (which are either partially or completely mutually exclusive/inclusive). --Ivan Štambuk 19:13, 19 August 2009 (UTC)

I have no idea how "approval voting" works. I voted for #1 because it is slightly better than nothing at all. #2 is no better than nothing and the status quo is actually better since nobody needs to try to count anyone’s edits. If approval voting works in some mysterious way, as Ruakh seems to intimate, I have no idea what it could be. As far as I’m concerned, it’s between #1 and nothing, and #2 is a red herring. —Stephen 21:08, 19 August 2009 (UTC)

Missing points[edit]

Apart from problems discussed above, this proposal is much better than the previous one, but it insists on one point: who whould be allowed to vote, and most important points are forgotten:

  • what is most important is not who votes, but arguments provided, and to try to reach a consensus after discussion. The proposal does not mention at all how the decision should be taken after the vote!
  • the project is not for contributors, but for users. In most votes, the users' opinion is more important than editors' opinion. Why not applying this proposal only to votes without any impact on users?

Lmaltier 20:39, 30 August 2009 (UTC)

This is a vote for voting eligibility, not for how to close a vote, step-by-step guideline how to reach a consensus, or how to interpret the relevance of voters opinions. Its primary aim is to filter out bad-faith votes cast by users out of sheer ignorance, malevolence or stupidity, which should ultimately prove to be a giant leap forward in reaching consensus in votes in which the abuse of the absence of voting threshold is likely to surface. This vote is essential for reaching consensus in any other relatively contentious vote.
This project is for both contributors and users (and who says that the contributors are not users?) In case you haven't noticed, lots of very prolific contributors here utilize Wiktionary as an abiding learning device. As for your remark on confining this vote only on votes without impact on users - I fail to see what would be achieved by this. Every single vote has an impact on users, and there is no way for us to replace the general wisdom of contributors with random sampling of relevant users opinion, as 99% of Wiktionary users are anonymous and unregistered and could not care less, and it would be impossible to engage the surfing minority which is relevant in the vote affecting them to express their opinions on the issue. The relevant contributors are the best relevant users. --Ivan Štambuk 21:32, 30 August 2009 (UTC)
Jeez. I thought it was a vote so that the Serbo-Croatian proposal could get ramrodded through by distorting all existing processes to that end. Thanks for the clarification.
What is your source for the 99% assertion above, BTW?
"The relevant users are the best users." So, only the "best" users should be voting? Well, if artistocracy was good enough for the Ancient Greeks most of the time, then why not for en.wikt? DCDuring TALK 22:58, 30 August 2009 (UTC)
What exactly is "distorted"? If we apply these criteria retroactively to all the votes that passed in the last year, how many of them would fail? None. Please don't spread FUD.
That percentage is not the point (it matters not whether the regular non-contributing viewers are registered or unregistered - they're out of our reach): the point is that there is no way that Lmaltier's scheme on non-contributors exclusively deciding on what contributing users should be doing can work. He is proposing what appears to be an alternative solution without giving any details how it should work in practice. Quite in sync with your late vote, if I might add.
Yes, only the "best" should vote. If you let anyone vote you get modern "democratic" system where a vote of an idiot is worth just as much as a vote of an educated, intelligent person who wisely makes his choices. That kind of electoral system is fundamentally flawed, because the first group will always statistically outnumber the second group. Plus, the majority is completely passive anyway, so various ardent PoV fundamentalists will always be overrepresented. We, however, satisfy for much less then blue blood: 50 edits hardly makes you a member of "Wiktionary aristocracy". Canvassed nationalists from Croatian Wikipedia have been ennobled anyway, making 50 edits and being gone till the SC vote reiterates, so you shouldn't be bothered. --Ivan Štambuk 23:33, 30 August 2009 (UTC)
I do believe you just demonstrated why this proposal won't actually do much good, and the costs will outweigh the benefits ("Canvassed nationalists from Croatian Wikipedia have been ennobled anyway, making 50 edits and being gone till the SC vote reiterates"). I agree wholeheartedly with Lmaltier's first point. The best way to combat meatpuppetry is to take a reasoned approach to the process as a whole, seeking to resolve more matters by discussion rather than counting heads, and de-emphasizing vote-counting in favor of weighing arguments. Dominic·t 00:09, 31 August 2009 (UTC)
What costs does this proposal have against non-contentious vote? What are you talking about?
"reasoned approach to the process as a whole, seeking to resolve more matters by discussion rather than counting heads" - this doesn't scale with contentious votes, as voters are not interested in discussing with arguments but spreading FUD and paranoidly defending their worldview. Lots of them even lie (see SC discussion on "languages have nothing to do with linguistics", or "Croatian and Serbian intellectuals independently chose the same dialect for their common language"). Why didn't you argument with obviously meatpuppeted users from Swahili and Croatian wikipedia, if it can all so easily be solved by talk? --Ivan Štambuk 06:57, 31 August 2009 (UTC)
The whole point of your proposal is to bar participation from an entire class of people. Don't act like there are no costs. I never said that we need to argue with people until we are blue in the face (that seems to be your particular talent). My point is that if we didn't have so much devotion to vote-counting, we wouldn't have so much a problem discounting the meatpuppets. You're saying it won't work because contentious voters are less interested in reasoned discussion, but that's my whole point. If we emphasized the weighing of arguments, then the voters who had no real arguments would have no real impact, whereas now we simply count it regardless. You appear to be happily enfranchising all the meatpuppets with an arbitrary number of edits by disenfranchising everyone, thoughtful or not, below that number. Dominic·t 08:31, 31 August 2009 (UTC)
No, not to "bar participation", but to bar from voting. No one is being barred from participating in the discussion in the relevant community discussion board which precedes every single vote. In the case of SC vote, of those voting for oppose, only 3 (Ullmann, Lmaltier and DCDuring) raised their arguments (whatever they be) in the BP discussion that preceded it. All the others simply kept silent and then waited for the vote. Votes aren't meant to discuss anything because it's too late to discuss anything once the vote has been started (although the discussed matter can be reutilized for the vote's further incarnations, if it failed). Anyone willing to comment on the matter as an outsider is free to do it in the period of several weeks before the votes start, and if his arguments are logically coherent it is likely that they will be taken into account by the regulars.
In practical terms terms the impact of this vote to all the non-contentious votes is none. Absolutely nothing would change in terms of passing/failing votes as regards on what should be included in ELE, who should be bestowed sysophood and similar things that are usually a matter of vote. That's because the usual set of voters (overwhelming majority, let's say 99%) are all regulars who already satisfy the trivial criteria of 50 edits. The only votes that would be affected would be the ones that would attract a significant amount of voters that do not pass the threshold, which we didn't have up until that SC vote, and probably won't have until it reiterates or some similar votes with possible implications for invalidating somebody's perception of political reality reappears. So "costs" would be infinitesimal in practical terms.
The point of the proposal is to simply set some basic barriers to bar evil-minded trolls of various kinds to significantly impact the voting process. Even if it doesn't solve the problem of meat-puppetry in its entirety (which it neither aims to solve) - it would be much easier to deal with 5 than with 40 of such votes. Every sister project has this kind of eligibility threshold for a very good reason. --Ivan Štambuk 11:58, 31 August 2009 (UTC)

The obvious questions...[edit]

  • Have you engaged in a widely-participated discussion of the issue in a public venue first?
  • Who determines who may vote in this vote?
  • Since voting is a fundamental decision-making concept, shouldn't voting rules be accepted strictly by 100% consensus? (That is, without consensus to follow a voting standard, there is always justification to ignore any result from a vote.)
  • Public voting is clearly divisive, thus constitutes community harm. Shouldn't a private ballot be used on those issues over which the community is unable to achieve any reasonable consensus?
  • Have you read Meta:Polls are evil?

-Amgine/talk 01:25, 1 September 2009 (UTC)

  • Who is "you"? And why "widely-participated"? Surely if someone chooses not to participate in a discussion, that's their own business?
  • The existing guidelines. (Dunno if they count as a "who", though.)
  • No.
  • No.
  • Who is "you"? I have; I can't speak for anyone else, though.
RuakhTALK 19:48, 4 September 2009 (UTC)
  • Actually, there was a discussion on BP. Widely-participated is the only possible measurement of community acceptance in online communities (see meatball:FairProcess and c2:ProcessPatterns.)
  • The point of this 'vote' is there is no guideline which is accepted policy.
  • Then you do accept that voting result in consensus, and will not follow the results of any vote. (iow: there is an underlying consensus that results of polls have at least some binding characteristics on the community.)
  • Excellent! You support, pardon the phrase, community balkanization.
  • A standard of English is to address the audience (readers) as "you". It's a rhetorical technique or device. Just an fyi. - Amgine/talk 20:19, 4 September 2009 (UTC)
  • Re: "Then you do accept that voting result in consensus, and will not follow the results of any vote.": Sorry, I don't follow. Could you please write in English?
  • Re: "Excellent! You support, pardon the phrase, community balkanization.": Nope.
RuakhTALK 20:27, 4 September 2009 (UTC)
  • Sorry, Re: "Then you do ^not accept that voting result in consensus..."
Voting assumes a consensus agreement to abide by the results of the poll. If there is not consensus on elements of a vote - such as who may vote, who counts it, what the rules for determining the outcome of the vote might be, etc. - then the vote will not likely result in the consensus to abide by the results of the poll.
  • You merely believe everyone should know who their opponents/targets are. - Amgine/talk 20:46, 4 September 2009 (UTC)
  • Re: first point: Consensus does not require unanimity. For example, there are editors who literally and professedly believe that Wiktionary should be an anarchy, with every editor being allowed to do whatever (s)he wants. Such editors are mistaken, and there exists consensus that no, Wiktionary should be a dictionary, with every editor being required to contribute in a way that improves the dictionary.
  • Re: second point: Not at all. Look, I believe that we should foster an environment where people can feel comfortable expressing their opinions openly — this is my big beef with [editors who shall remain nameless] at the moment, BTW, that they're openly accusing all opinionated editors of genocide or other forms of overt nationalism, without caring that their accusations may be wrong. But I also believe that if we're failing to foster that environment, then the solution isn't to say, "oh, well, that's O.K.: we can just tell people to shut up and cast secret votes."
RuakhTALK 21:03, 4 September 2009 (UTC)
I believe consensus, in decision-making process, does not require unanimity in support for the result, but does require unanimity that a decision has been made. Else the decision-making process continues. Witness this specific poll.
I think we're saying the same thing, but with completely different points of view. Perhaps my pov is cynical, that public votes "on those issues over which the community is unable to achieve any reasonable consensus" results in opposing camps/parties being defined and reinforced. My experiences suggest that when discussions do not progress toward compromises, votes or polls entrench enmity rather than resolve issues. - Amgine/talk 22:13, 4 September 2009 (UTC)

Counting the vote[edit]

Everyone who voted for more restrictive option 1 automatically supports less restrictive option 1. If that is not particularly clear, I suggest that all of those who already voted for option 1 be notified to express their possible non-support for the option 1, and that be taken into final account. --Ivan Štambuk 11:47, 3 September 2009 (UTC)

In this case, it will be necessary to change the end date of the vote. You can't change your vote after the end of the vote. --Rising Sun 13:43, 3 September 2009 (UTC)
No, it's not additional voting, but explicitly saying what is already implied, just apparently unclear to some. Vote is already expired and cannot be prolonged. --Ivan Štambuk 14:26, 3 September 2009 (UTC)
I thought a month was customary, anyway. At least that's what the header recommends (which is apparently now canon, or this vote wouldn't have been necessary). Generally, it is accepted that if people support more than one option, they should add their vote to both, rather than have us make assumptions about their opinions, since some people may be refraining from supporting both intentionally. In the future, when there are multiple options, a vote design more like Wiktionary:Votes/pl-2008-04/WMF jargon might work better. Dominic·t 14:17, 3 September 2009 (UTC)
There are no general rules on multiple-option voting. This vote was originally set so that you can vote independently for and against either of the options, but the format was revised per Ruakh's and Prince Kassad's suggestion, so that you can vote only for 2, for both, and be against both, which is quite natural since option 1 is really more restrictive variety of option 2, and it wouldn't make much sense to e.g. vote only for option 1 and be against option 2. --Ivan Štambuk 14:30, 3 September 2009 (UTC)
That vote on WMF jargon is not applicable here because it deals only with exclusive type of single choice without overlap in options. --Ivan Štambuk 14:33, 3 September 2009 (UTC)
Ivan, I don't understand what you are doing. You say that people who voted for option 1 can't vote against option 2, they can only vote for it. In this edit I question the reasoning. Why can't they vote against it? And I can't see a reason, why it would be "already implied" that they would also support option 2. This is the reason, why I think the vote should be prolonged (or I don't see why we can accept your late vote). --Rising Sun 14:48, 3 September 2009 (UTC)
Because there's only ==Oppose both== header. There is no ==Oppose option 1== or ==Oppose option 2== header, and if they already supported either, they cannot vote against the other one.
And I can't see a reason, why it would be "already implied" that they would also support option 2. - Because option 1 is a logical superset of option 2. You cannot vote for option 1, and simultaneously be against less restrictive form of it (well you could in theory, for some principle or sth, it's just it doesn't make much sense from the standpoint of logic). --Ivan Štambuk 14:57, 3 September 2009 (UTC)
Well, Ivan, I do not agree with the last ratiocination - Because option 1 is a logical superset of option 2. You cannot vote for option 1, and simultaneously be against less restrictive form of it. This is Stephen's position and it is clear-cut: If the superset is noted as B and the set (of voting option 2) is A you may want to see at any cost the set (B-A) implemented/promulgated, i. e. the 6 month restriction and without it for the person in quæstion the set A is not too different from the empty set (whichis currently valid) and therefore the support may well be only for set B. The uſer hight Bogorm converſation 15:18, 3 September 2009 (UTC)
Well, the alternative is no voting eligibility rules whatsoever, and I sincerely hope that pragmatical concerns will outweigh those coming out of principles. --Ivan Štambuk 15:25, 3 September 2009 (UTC)

Note also that e.g. I voted only for option 1 not because I'm against option 2 (I am in favor of any kind of restriction on voting threshold!), but because I wanted it to succeed to option 2. If all of those who voted for option 1 also explicitly voted for option 2, option 1 could never win as long as there is single vote only for option 2. Ideally (as I imagined this vote) people would vote only for one of the options, and that with had most votes would win. But, as it turned out, some folks voted equally for both of the options (from understandable reasons), which messed things a bit. The way this vote should be counted is:

  1. those people who voted for both of the options, or only for option 1 should only be counted supporting option 1
  2. all those people who voted for option 2 should be counted as supporting only option 2, unless they also voted also for option 1 (this is to compensate for those who have redundantly cast their votes for both of the options)
  3. the option that has most of the votes "wins". If the option 1 "wins", only those votes that were cast for it are counted as "winning". If option 2 "wins", all the votes cast for it are counted, plus the votes of those who voted for option 1 and who didn't (again, redundantly) cast their vote explicitly supporting also option 2.
  4. the opposing votes are counted against the total votes of the "winning" option.

This is necessary because options 1 and 2 overlap in choices, and are not mutually exclusive. Ideally one should vote only for one of the options, but this algorithm of counting compensates even if users voted for both of them by giving preference to that which is stricter. --Ivan Štambuk 19:00, 4 September 2009 (UTC)

Where the "rules" went wrong[edit]

here, where Ruakh changed my original formulation:

Users can vote for either of the proposed options, and the one which gains the must support will pass.


Users can vote for either or both of the proposed options; the one with more support will pass, provided it gets a majority.

Too bad I didn't catch this earlier. Can we agree that the current "rules" which allow users to vote for both of the options, but suggest counting them independently (i.e. user voting for the stricter option not being counted as being supportive of less strict option, when determining the total count of votes for the "winning" option, and users explicitly voting for both of the options not being counted as being actually supportive only of stricter option, when determining which of the option "wins") are logically invalid?

This change of "either" to "both" can be effectively compensated by the above algorithm of counting I provided. --Ivan Štambuk 02:11, 6 September 2009 (UTC)

No, I don't think that can be agreed upon. In fact, there has clearly already been disagreement to that way of counting, so I am not sure why you are just proposing it again. Dominic·t 07:55, 6 September 2009 (UTC)


I am forwarding a discussion on the accusation of RU and on canvassing from the vote page to the talk page:

Well, being uncivil isn't "prohibited", either, so if "not prohibited" means "acceptable", then I'm not sure what you're going on about. (Actually, more generally, I'm not sure what you're going on about. Your first comment in this thread — "If there is a formal complaint of sockpuppetry, there should be a request at WT:CU. If not, this is slander or libel, a form of harassment which is not acceptable in this community" — borders on nonsense. Do you seriously contend both (1) that meatpuppetry is acceptable and (2) that pointing-out of meatpuppetry is libel? Because those two contentions are mutually contradictory: if it's acceptable, then how can it be libelous to point it out? I swear, of all of Ivan's comments, I have no idea how you decided that this is the one that bothers you.) —Ruakh 17:33, 8 September 2009 (UTC)
But the possibility of libel does not refer to an accusation of meatpuppetry but rather of sockpuppetry. Meatpuppetry is undesirable, while sockpuppetry in voting is forbidden and is said to lead to a block. That RU is being accused of sockpuppetry can be estimated from "Wow Robert, you just couldn't resist could you?" and "If that is really Robert, it is grossly unethical and ... ", although that is rather equivocal. The "Wow Robert..." seems to be addressed to Rose Waswa by being posted under the vote of Rose Waswa, which implies that Rose Waswa is a sockpuppet, not meatpuppet, of RU. I have read "that" in "If that is really Robert..." as "Rose Waswa". In any case, this ambiguous communication that accuses someone of something without making it explicitly clear what is the accusation should better never have happened. It gives me the impression of a provocation that, as it turned out, was not left without a heated response. --Dan Polansky 18:08, 8 September 2009 (UTC)