Meta talk:Rewriting/Stewards policy/Vote

Required percentage

An anonymous user (thanks for logging in next time) indicated that 80% support would be needed for this voting process. The idea behind that would be that that is as much as the position. Personally, I do not think that a policy and a position are directly linked through the required percentage. A position is about trust that a community has in the candidates (and it is not even a real election, it is only a nomination so that the Board of Trustees can appoint) and a person having that position should have much trust.

However, a policy is not about trust. A policy is about opinions, about guidelines. With a policy it is very well possible that a group of people think it is too harsh, and another group thinks it is too soft. If both groups would be 20%, that does not make it a bad policy, as the middle would still be that very policy. That is not the case with trust (as you can't trust a person too much). With opinions, it should just be clear that there is a majority for the proposal, and the 55% is already to indicate that there is 10% difference between the pro and con's. 55% seems thus a fair percentage to me, and not a very odd percentage for policies in general. Remember, this is not about trust now, but about opinions. Effeietsanders 21:37, 30 December 2007 (UTC)

A general policy with broad influence should require the support of a strong consensus. It is paradoxical in theory that, a group of users failing to muscle up the required 80% vote can, in principle, decide to change the policy by declaring an arbitrary number (be it 55% or 50.001, it matters not) is the threshold for policy change.

Put in another way, if your new policy cannot even get 56% support on Meta, it cannot be good.

Re :"it is very well possible that a group of people think it is too harsh, and another group thinks it is too soft" you have ignored the possibility that the editors, despite their differences, have enough good wills to come to a compromise. Your low threshold enourages takinig camps quickly and discourages dialogue.

Thank you for your comments (I really would think it shows good manners if you'd log in, thanks), but as you might except, I disagree. Again you mix up the policy that describes the behaviour and guidelines for stewards and the steward positions itself. The position is about trust, and that is why there is such a high threshold. However, note that this is totally different from a policy, which is about what a group of people thinks is best. If you are talking about what is best, which opinion should be leading, a normal majority is normal. As you might have seen, this proposal is being build upon already for a long time, actually already since last summer. I do not see where you see here a quickly taking camps. Also, it is not my threshold, but it is a general threshold for this kind of policy votings afaict. It is no strong requirement to have for everything a strong consensus, although it is desireble. But maybe you should make a difference between desireble and required for policy votings. I think that it would be desireble and wanted for a proposal to have broad support, and I think that you can assume easily that I will always try to gain that broad support, but that does not mean that we should tie ourselves up to a very high percentage just because of that. It means that we should attempt that, and I do not think that people who are active on meta need encouragement like a strong high threshold to keep trying to gain a broad support. But if that fails time after time, it should also be possible to continue with a lower support, which is the general basis for a percentage like 55%. However, please note that in this case I would not expect any trouble, I have seen barely any objections to the policy. However, I do think that it would be best to stick with the 55% (or 50, if you prefer that) of the pro/con votes as an absolute minimum threshold. Effeietsanders 00:14, 31 December 2007 (UTC)

"but it is a general threshold for this kind of policy votings afaict." What is "this kind of policy" and where is 55% a general threshold? Even the US Senators can filibuster with 40%. And you know how hard it is, for example, for the English wikipedia to adopt a new policy. A policy proposal which cannot even garner 56% support is to be fixed. Think of the 45% opponents with their multitute of reasons. Are you afraid to put the proposal to the real test that it deserves? If you have worked on it, you should be confident that it will pass the test. The basic reason why a general policy which has wide effects requires the test of a high threshold is that, when in the future there is a dispute on the mechanics on the policy, you can point to a strong support. —The preceding unsigned comment was added by 82.133.109.205 (talk • contribs) 07:35, 31 Dec 2007 (UTC)

I caught this one late yesterday and thought I would see where it was going today.

I can understand the idea of a simple majority (50% + 1 vote) as a basis for this. Equally a 2/3 or 3/4 majority are quite normal concepts. However I do find the idea of 55% rather strange & random. I will watch & learn I'm sure. (And logging in/signing post would be considered polite thanks) --Herby ^{talk thyme} 08:12, 31 December 2007 (UTC)

Yes, 50%+1 would be the the simple idea, and maybe it is better to stick with that, but there are actually reasons for 55% :) As you might see, there is a 10% barrier between 45% and 55%. So that's a nice number, and 55% is not totally ad random :P The idea is that with a 10% barrier, you create enough space between contra and pro voters to make sure that if you would do the vote again one wee later, the result would be the other way around. For instance, now we would get 50.001%, and next week 49.999%. With 55% it is less likely that next week we get 45% support. But actually that threshold is more relevant to projects with a lot of votings, and not really to meta. Effeietsanders 09:27, 31 December 2007 (UTC)

In "barrier" one can sense a "win-lose" mentality. There is no winner or loser in a policy discussion. The point is not to out-vote your opponent (where your model would be ideal), but to get the best policy out of the discussion, and one which reflects the community consensus. A policy affecting *everybody* should not be passed if it has almost 1 opponent to every supporteer. The reason is that wiki is above all a collaboration. Now, whether there is difference between *trust* in a person or *trust* in a policy is a point of view and is arguable. What remains is that the policy affects everyone. That includes users identified only by their IP-addresses. --82.133.109.205 10:23, 31 December 2007 (UTC)

If this were en wp we could initiate a discussion on a vote to decide on the approval threshold....:)

In practice these policies must be acceptable to the majority of those affected by them (in a sense presumably all stewards?). Without being extremely prescriptive there will always be "grey" areas and goodwill & trust will be needed for these.

One problem is that an exceptionally small number of those affected by these policies will be aware of the discussion (or indeed of Meta itself). That said stewards are elected mostly by that very small number as well. The fact that stewards is a sense "govern" with such a small mandate always has the potential for difficulties & misunderstandings.

Certainly the concepts of "winning" or "loosing" are unhelpful - however arguments about "consensus" frequently generate more heat than light too. Personally I do think that the acceptance threshold for steward policies should be higher than 55% given the extremely small involved population. --Herby ^{talk thyme} 10:39, 31 December 2007 (UTC)

• reset

Reflecting on this I am certainly of the opinion that a far higher figure for approval can be justified. If you consider the number likely to vote as a proportion of the possible voters then a very high figure will be needed for the approval to be statistically significant. Strictly statistically you would probably require approval of 100% but I imagine that will not be acceptable & we do live in the real world. 80% probably should be the minimum in my view --Herby ^{talk thyme} 07:50, 5 January 2008 (UTC)

Well, actually you could say the same of course for opposers, to have a statistically relevant opposition, it should also be very high :P And you could and should also take into consideration that generally the people who oppose will find their way to the vote more easily then the people who agree with it. At least, that is what history on Wikimedia Projects taught me. If you take that into account, you would come to a lower percentage. That is actually one of the reasons why i think that 55 (or 50) % is more suitable. Effeietsanders 11:54, 5 January 2008 (UTC)

Please look at real statistical significance. Equally to consciously introduce a bias based on a perception that voters who oppose will "find it more easily" seems frankly plain wrong. I would hate to think that was the basis a project that has any care for democracy would work on --Herby ^{talk thyme} 12:13, 5 January 2008 (UTC)

It is however how it generally works. People which disagree with a proposal, will be more alert for such proposals, are generally more willing to express their opposing publicly, but also are more willing to "do" something for it (create an account to vote etc), then people who find the proposal "ok" (especially when the new thing is not so much of a radical change). Of course, in an ideal world it would not work like that, but this is a bit how people think and set their priorities. People tend to shout much louder that they disagree then that they agree with something. Effeietsanders 14:55, 5 January 2008 (UTC)

Required Percentage part II

See this diff, an anon has changed the percentage requirement to 75%. While there did seem to be some disagreement about 55%, I'm not sure we were at consensus for 75% either. More discussion to clarify this might be helpful. Personally I think 75% is too high, I'd favour something more like 2/3 majority of those voting. ++Lar: t/c 20:08, 7 January 2008 (UTC)

In practice, as I have already said, the "sample" vote that will be here is so small as to be almost meaningless anyway. Statistically the support really should be close to 100% to be valid - I suppose 90% might be ok. I would also expect 100% of the stewards to support it - otherwise its validity will be questionable anyway --Herby ^{talk thyme} 09:51, 8 January 2008 (UTC)

I'm with Lar on this. As Effeietsanders said above, of the people who oppose, probably all will vote, of the people who agree, only a small percentage will vote. So setting the requirement to 75 or even 90 % will give the opposers much more power to let improvements fail than the supporters will have. In every "normal" voting, 50% (or maybe 55) is the default, considering that stewards policy is something very serious and important, it is probably better to set to 2/3, but that's the absolute maximum I would support. --Thogo (talk) 10:09, 8 January 2008 (UTC)

I confess I find the view that we need to find a percentage that works because we know the "no" voters will be more likely to vote than "yes" voters extremely worrying in a democratic sense. Stewards would have had a far harder time being elected if that had been the case - you do all really need to remember to trust your electorate - they decided to trust you --Herby ^{talk thyme} 10:42, 8 January 2008 (UTC)

You remember Aphaia's election? There you have seen that problem. --Thogo (talk) 10:44, 8 January 2008 (UTC)

I don't think that is relevant to policy discussions, it is about that fact that a specific community had issues of trust with a specific individual. Quite a number of you were elected/confirmed despite some "local" concerns in quite a few cases. I see no reason why policy discussion should be affected by any local issues (assuming you don't get on to licensing:)) --Herby ^{talk thyme} 10:49, 8 January 2008 (UTC)

Hm, I didn't mean local issues in the case of policy rewriting. I mean people who have certain ideas of how things should go which is not the opinion of the vast majority of users. But those "aberrant" people always vote whereas the mainstream users in most cases don't. So if you set the barrier to 90%, you are very likely to get a result that most people don't support, just imagine that 85% support the change, and 15% do not, than 15% get their opinion through which is not appropriate at all, and, that's not how democracy should be like... --Thogo (talk) 11:04, 8 January 2008 (UTC)

You (& Effeietsanders) are ignoring the issue I raised above - statistical significance. (equally "aberrant" data is what is usually disposed of because it affects statistical significance, much more convenient than allowing people their opinions) However I am beginning to think that this whole exercise is fairly futile given the approach to democracy that seems to exist among those democratically elected! --Herby ^{talk thyme} 11:14, 8 January 2008 (UTC)

Well 55% is too low and 85% is too high..I would prefer 2/3 or atleast 65% but also we have to make sure how many of the votes are null and void as well before making the final decision since most just vote based on personal grudge or socking by creating multiple accounts to oppose or support someone or try to make a point..--Cometstyles 11:19, 8 January 2008 (UTC)

Please "assume good faith" - the comment since most just vote based on personal grudge or socking by creating multiple accounts to oppose or support seems to lack that and not be based on fact. I am unaware that there were any sock accounts in the steward elections for example or equally on any RfAs since I have had CU rights. It is rare --Herby ^{talk thyme} 11:57, 8 January 2008 (UTC)

Yes, sockpuppets are not used on Meta normally (maybe in closing/new project discussions, I don't know). Anyway, is there any argument to go as high as 75 or even 90% instead of 2/3 (which already is more than in normal votings)? The default for a community decision should always be 50% in the sense of democracy (if you really want democracy here). This is of course too low for policy changing, but as in most states for changing the constitution (which is nothing else than changing a/the general policy) a 2/3 majority is sufficient, why not here? --Thogo (talk) 12:15, 8 January 2008 (UTC)

A bit late comment. Generally I agree with Herby including my own case if we limit the topic about voting rate. However I would point out there was apparently canvassing (Alex.S.H . later found it and gave me its URL) at that time. Meta has no rule about canvassing (as well as no avoiding "COI" rule). Perhaps we would like to put words about canvassing? There are some users, and I think it would be ja.users case, who have no idea why canvassing is not welcome. If you are involved into Nanking Rape issue on English Wikipedia, you may know a Japanese critic with a certain renown did a canvassing on his blog and failed to perceive why Wikipedia didn't welcome such activity. --Aphaia 08:26, 10 January 2008 (UTC)

Much talk about not much

A good policy will be supported by a large majority. If a policy cannot reach 75-80% of support (or 2/3, I don't really mind), then it's not a good policy, and it should not be accepted. Lower percentages around 51-55% are clearly not enough in my view, they give the impression that this percentage is chosen to be sure the policy is accepted. Either we adopt a policy with a large majority, or we don't adopt it, but lowering the required percentage to have it adopted seems a bad practice to me. My 0.02€. guillom 12:18, 8 January 2008 (UTC)

Could we go with 2/3 as a reasonable compromise? ++Lar: t/c 00:52, 10 January 2008 (UTC)

Yes Please..I believe that will be perfectly reasonable compromise :P ..--Cometstyles 00:55, 10 January 2008 (UTC)

2/3 is ok for me as stated above. --Thogo (talk) 07:54, 10 January 2008 (UTC)

I'm with guillom. If we cannot get the support of 75/80% of the voters then there is an indication that the policy is not as good as it should be. Equally dealing with any dispute would be harder without a high level of validity --Herby ^{talk thyme} 07:56, 10 January 2008 (UTC)

Let me raise some other issues here. To examine if some result from the voting may be accepted, we would like also to consider: If there are enough voters (votings for constitutions were mentioned: do most of parliaments have a requirement of the minimum number who vote for a particular bill?). I eventually found this talk and found therefore the voting is planning and opens soon. That said, I would miss the voting, while I knew the Eia's proposal on foundation-l. From my experience as an organizer of several online votes, it is always a challenge to attract eligible voters and get a fair amount of votes. If you are stick to the current schedule (from Jan 15 till the end of the month), you'd better to advertisem this voting including the date and the eligibility of voters more aggressively, I supppose. And this may affect all Wikimedia project, it wouldn't be enough to put a notice on meta and foundation-l. ---Aphaia 08:20, 10 January 2008 (UTC)

In many senses that is my problem

If you attract few voters from a large electorate then the level of acceptance must high to be at all valid

If it can be made more valid by attracting a substantial number of voters that would be good (while there may then be logistical issues) --Herby ^{talk thyme} 08:25, 10 January 2008 (UTC)

Starts tomorrow?

Hi, today is 14th. The planned schedule says it starts on 15th, hence tomorrow. If it counts in UTC, it means it will starts in 14 hours. The last edit on this page was on 10th January and this talk seems almost dead.

Will the vote really start be opened tomorrow? --Aphaia 04:17, 14 January 2008 (UTC)

As far as I am concerned, yes :) I will make a final announcement on foundation-l and meta-l today I think, with the request to spread the word. BR, Effeietsanders 08:18, 14 January 2008 (UTC)

I see. So it is a good time to refresh the sitenotice. I'm going to blank it ... --Aphaia 12:19, 14 January 2008 (UTC)

Removal of annual confirmation

It seems to me that this new policy removes annual confirmation of Stewards? If I am correct I shall certainly vote against this for that reason --Herby ^{talk thyme} 08:42, 14 January 2008 (UTC)

By no means. The confirmation would be performed whenever there are elections in progress. However, the policy states that it would be the stewards themselves that would vote in the confirmation process. --FiLiP ¤ 08:52, 14 January 2008 (UTC)

"Whenever" does not equate to "annual" though? The wording is confusing - stewards currently decide on the results of confirmation so is there still an annual community input? --Herby ^{talk thyme} 09:07, 14 January 2008 (UTC)

Diff

Could there please be a link or a page describing the actual differences between the old and the new policy to see what has changed? It would also help a lot, if those changes could be commented explaining the rationale behind each. Thanks. sebmol ^? 08:48, 14 January 2008 (UTC)

There is already such a link. Please take a further look. --Thogo (talk) 09:31, 14 January 2008 (UTC)

However I also think "rationale" would be helpful --Herby ^{talk thyme} 09:35, 14 January 2008 (UTC)

In the very most cases it's only a reformulation of the existing policy, due to make things more clear. The newly introduced sections are no new rules but add conclusions of the existing rules. For example the point about "Super-Admins" shall make it clear that we stewards can not overrule any local decisions, which already is policy but since we come across *many* requests of the kind "The local sysop/crat decided wrong, please check." it was necessary to make this point more clear. --Thogo (talk) 09:48, 14 January 2008 (UTC)

So the rationale behind removing annual confirmation is? --Herby ^{talk thyme} 09:51, 14 January 2008 (UTC)

There is no removing of the annual confirmation. --Thogo (talk) 09:52, 14 January 2008 (UTC)

It clearly says "at the time of the next election" now and makes it equally clear elsewhere on the page that there may well not be annual elections (removing the opportunity for input by the community on what they may want) --Herby ^{talk thyme} 09:58, 14 January 2008 (UTC)

Since there are elections every year, there will be confirmations every year. If there are no elections in a year, there is no need to confirm the other stewards, since removing steward access of some users would make it necessary to elect new ones which would mean an election anyway. And if a steward turns out to be inappropriate or inactive or whatever the steward access will be removed anyway. Community input can be given at any time. --Thogo (talk) 10:01, 14 January 2008 (UTC)

Hum - I think we have a circular argument here. You say Community input can be given at any time, however without a formal process it will merely be the odd lone voice which can easily be ignored. The new policy states (effectively) that there will not be annual elections of new stewards. Confirmation takes place at each election (per policy). We will never need new stewards unless we get a chance to confirm the old ones. Chicken - egg, badly thought out at best --Herby ^{talk thyme} 10:42, 14 January 2008 (UTC)

OK, this is not really clear indeed. I changed the sentence in "Steward access is granted until the next confirmations, when users will be invited to comment or ask for removal of their access. Confirmations will generally be held during steward elections and/or roughly annually." I think the idea is that elections are generally held annually, but with a different schedule, this could become complicated indeed. Hence I changed it in this. I think that the intention of having roughly annual confirmations is considered consensus, but for practical reasons it could be 14 or 10 months imho. Hence no strong wording to *have it annually*. Effeietsanders 11:33, 14 January 2008 (UTC)

I'll go back to what I said before then. While I still feel twice a year is a good idea, I am OK with once a year since I am in the minority on twice a year. But I strongly feel we should explicitly institutionalise having elections annually, not leave it as "whenever the stewards feel like it", and the policy should be firm and unambiguous about that. I think the overall community would give some latitude about exactly when they are held (not hold us to a specific day, as long as it was roughly the same, late Nov early Dec kind of timing) but we've held them close to the same time for at least two years in a row now, and as long as Pathoschild doesn't get archive happy :), the framework built from last time means the next ones will be far less work to organise, so it's not really an issue as far as organization goes. I think it sets a bad tone for the stewards if the stewards do not commit to having definite times for elections. ++Lar: t/c 17:24, 14 January 2008 (UTC)

The trouble is looking at this from the perspective of someone outside the "Steward enclosure" I see the fact that stewards directly or indirectly will control if/when there is another election. This in turn means that they determine the next confirmation. Equally they determine who goes and who stays. I regret saying this but if that does not appear to the "man in the street" as an attempt to control power I would be surprised.

As has been made clear we cannot have a decent level of support in the voting as "trouble makers" will turn up to oppose. It may well be that you are creating the trouble makers. I ask you to reflect on that. Equally to remember the community elected you - its views must be considered --Herby ^{talk thyme} 18:35, 14 January 2008 (UTC)

"Confirmations will generally be held during steward elections and/or roughly annually."

"Generally" and "roughly" seem very strange words to use in a policy proposal. It seems to me that yearly reconfirmation is something people have come to accept. There has also been some talk of steward elections happening more frequently. Tying the reconfirmation to the election raise the poblems that, if elections happen more frequently, stewards will face reconfirmation too frequently and, if elections become less frequent, stewards may not be seen as sufficiently accountable. I take Effeietsanders's point that for logistical reasons there may not be precisely 12 months between steward elections. How about the clear phrasing: "Confirmations are usually held annually, timed to coincide with the election of new stewards"? That way, if the next steward election is 13 or 14 months later it can be pointed out why the reconfirmation is being delayed, without given the appearance that reconfirmation happens at the stewards' convenience. Alternatively one could give a time span during with reconfirmation could happen (e.g. within 18 months, though usually annually). WjBscribe 19:24, 14 January 2008 (UTC)

I put some wording in pointing out that the last two elections were held at the same time and that there is a "reasonable expectation" they will be held the same time every year. Would that be a good compromise, I wonder? At least a step in the right direction? ++Lar: t/c 22:39, 14 January 2008 (UTC)

Where is the voting advertised?

keep the track!

• foundation-l: Effe? (with request to spread the word)
• Wikimediameta-l: Effe
• en.wb noticeboard (one of them)
• Commons VP
• ar.wikipedia VP

"Who can vote"

Surely it's not enough that the Meta user page links to the page for an active user on another project. What stops me from creating a new Meta account and writing "My Wikipedia account is w:User:Jimbo Wales" on the user page, then voting? The requirement must be that a link exists in the other direction (well, both ways, for simplicity.) — Kwi (talk) 21:06, 16 January 2008 (UTC)

yup exactly, or else they can verify it some other way for example through IRC or e-mail correspondence..--Cometstyles 21:12, 16 January 2008 (UTC)

Yes, a full crosslink really is what's needed. ++Lar: t/c 22:37, 16 January 2008 (UTC)

snowball

Rather better to close it now, while there are still three days according to the schedule, it may not happen the proposed draft get required 2/3 support of the total votes. Then we will be able to move on postmortem and seek for consensus on the points where the current stewards find difficulties or contradictions and feel urgent needs to reform? Apparently we are not going to have a reconfirmation voting in few weeks, and that is the point people feel bad mainly, it would be more realistic to focus on problems on day-by-day issues and need to be fixed as soon as possible. Or three days make no big difference or am I too early to call for WP:SNOW here? --Aphaia 08:22, 25 January 2008 (UTC)

Personally I requested it a while back so I would be happy. I am working on a new one anyway which I will publicise when it has rather more flesh on it. The last oppose pointing out that it really should be translated seems highly relevant too. Equally obvious is the fact that the community does want their involvement to be relevant and so it should be rather more open for ordinary users input rather than focussed on stewards views of their own role. To me this is important policy and needs to be dealt with in the best possible way & not quickly --Herby ^{talk thyme} 08:30, 25 January 2008 (UTC)

To me this is important policy and needs to be dealt with in the best possible way & not quickly. I agree. FloNight 18:10, 26 January 2008 (UTC)

Conclusion

The vote has concluded, with the result that the changes in the policy were not approved and the old policy is still in effect. I think it makes sense to put together a list of the concerns raised towards trying to put another rewrite in place that addresses the concerns. It's unfortunate that these were not raised during the rewrite process but at least now they're on the table. The place to do this is Meta talk:Rewriting/Stewards policy rather than here I suspect. ++Lar: t/c 15:55, 29 January 2008 (UTC)