Wikipedia talk:Non-admin closure/draft

Changes and suggested changes

• I've made a few changes, if someone disagrees feel free to exercise BRD.

I also feel that For example, does "10-2" in favor of keep count as "nearly unanimous"? - I feel that "nearly unanimous" is not a helpful marker to give - the restriction is on close calls, whereas nearly unanimous is a much stricter requirement, beyond that we should be limiting NACs to. Nosebagbear (talk) 22:11, 19 January 2020 (UTC)

Fully agreed. In general I think the pitfalls section is quite unnecessary if it became a user right. Perhaps changing the heading to advice for new non-admin closers or something because most of these items seem to be exclusively suggestions rather than rules. ‑‑Trialpears (talk) 22:37, 19 January 2020 (UTC)

WP:BADNAC

If AfD closures are given as a permission, why would we still need WP:BADNAC? --Enos733 (talk) 22:22, 19 January 2020 (UTC)

The current items there are involved closes, controversial and difficult closes and closes requiring admin tools. I still think all three of these are needed, ‑‑Trialpears (talk) 22:29, 19 January 2020 (UTC)

Involved closes are a separate issue, and can affect administrators and non-admins alike. Closes requiring admin tools are the only category that is complicated by adding permissions for closing XfD. My thought is that if the community trusts a non-admin to make close AfD, then they should be trusted to close. I would suggest that if a non-admin closes a discussion as delete (or other action that requires other administrative tools), that notice be placed on the WP:Administrators' noticeboard, to finish the close. --Enos733 (talk) 22:46, 19 January 2020 (UTC)

That is kind of what is done at TfD as can be read about on this very page. It is also done unofficially at CfD due to the immense backlog, but other venues have been reluctant to the idea. I think a non-admin closer could make it easier to get through though since that would largely remove the concerns of articles being deleted after a clearly bad close. This would have to be a future discussion though and it would not be appropriate to remove that section now. ‑‑Trialpears (talk) 22:52, 19 January 2020 (UTC)

I think now is the best time to have (a preliminary) discussion. --Enos733 (talk) 23:23, 19 January 2020 (UTC)

Allowing non-admins to close as delete at AFD would likely be considered an even bigger change than making non-admin closer a user right and whether non-admin closer is user right would massively affect that discussion. I don't disagree with your proposal, I just don't think now is the time. I've seen several discussions where too much is discussed at the same time and the overall result is no consensus because the discussion got to fractured. I really don't want that to happen here. ‑‑Trialpears (talk) 23:31, 19 January 2020 (UTC)

I understand where you are coming from, but the premise of giving a permission is a signal of trust from the community to the user. This goes to the question of the problem we are trying to solve. Is the problem that there are too many bad NACs, or is it that there needs to be more structure around who can close. If the problem is that there are too many bad NACs, we can make the criteria for a NAC close more stringent or prevent administrators to close. If we just want to limit closures to a certain set of editors, there should be a question of how much trust we have in that set of editors to close any discussion. We allow non-administrators to review (and speedily delete) new articles, accept or reject content on protected pages, and several other tools relating to removing content. So, I think that embedded in the question of establishing a form of permission is how much trust and latitude an editor (with permission) should have. --Enos733 (talk) 23:59, 19 January 2020 (UTC)

I'd be against allowing editors with this permission to delete at AfD (obviously it can make sense for other XfDs if they think that's logical). This userright's requirements are targeted at the current level - if we wanted them to be able to consider delete, I'd want to re-write them and I'd probably still be against it. The damage of an incorrect Keep (let alone relist) is far less than an incorrect delete judgement. Non-admins can't speedily delete new articles, only nominate them, so they're only half of that process. If we expected the processing admin for an AfD to do the same level of review, then the NACs wouldn't be saving any work. Nosebagbear (talk) 09:57, 20 January 2020 (UTC)

The thing is that by creating a permission to close AfD is that we are signaling trust in that user to make the correct judgment. Why would we spend time and energy to create a user right that only allows a closure (or relisting) of obvious discussions. If the current rules are followed, we don't need to create a new user right as everyone would follow the guidelines in WP:NAC. Since that is not the case - that as it is, current administrators are closing discussions that are not always obvious, the guidance question is one of trust - either as a community we trust administrators to close AfD discussions properly (even potentially controversial discussions) or we don't.

As an aside, I do note that even deleted articles are not permanently deleted from this project (except in special circumstances). Thus, the harm to the project by a incorrect delete judgment is minor and temporary. --Enos733 (talk) 04:46, 21 January 2020 (UTC)

@Enos733: - to your first paragraph: there are several reasons. Firstly is that trust has two components - in the character and in the ability. Trust also isn't an all or nothing exercise - we also have a userright for pending changes, which indicates some degree of trust, but not as much as a Bureaucrat flag. Handling non-controversial ones lets editors practice interpreting consensus, which includes both the end decision but also learning to pick out the reasoning, filtering out non-policy, and summarising it. Nosebagbear (talk)

harm to the project by a incorrect delete judgment is minor and temporary - this is a viewpoint I used to share. However, through OTRS, the Teahouse, and seeing a couple of arb cases, this is an issue of perspective. To you or I, having an article of ours deleted is annoying - but not harmful or permanent. To new editors, it's downright catastrophic, confusing, difficult to understand the process of disputing, and causes a major loss of trust in "the system". Even from good faith new editors, it frequently comes with thoughts of bullying. The consequences are in fact major, which behooves us to make sure that the decisions are taken as major events. Nosebagbear (talk) 10:23, 21 January 2020 (UTC)

To the first comment, I would quickly say that there is not much training on non-controversial keeps. If there is only the nomination delete and the rest of the comments are keep, it is obvious what the close should be. Similarly (and as it is a non-administrator cannot close these), if there are only delete votes, it is pretty easy to see how the close will go. Even a close of no-consensus is easy to spot - if there are policy arguments made on both sides and there is a roughly . The tough calls are those where there is a discussion of sources or in cases where there is not a clean policy argument made (especially if that side has a slight numerical majority). In addition, there is no requirement to summarize the rational behind the decision (although I do think there should be in controversial cases). Where we should want to see growth in an editor's ability is how they close the more difficult cases. --Enos733 (talk) 17:21, 21 January 2020 (UTC)

I completely agree that non-admins should be given the trust to perform non-obvious closes and that is partly why I support making it into a user right. By having a more experienced pool of non-admin closers it is less likely that bad closes occur and the general opinion of NACs will improve leading to people trusting non-admins to perform less obvious closes. There is already quite some discretion in where to put the line for what is a close call and I wouldn't be surprised if BADNAC could go from avoid close calls to avoid closing when there are several appropriate outcomes assuming there aren't much in the way of BADNACs at deletion review. Trying to give non-admins more leeway by letting them close as delete at all venues is not a good way to do it. It is uncertain whether it cuts down on backlogs at all since admin action is still needed, the harm done by a bad deletion is so much larger than a bad keep and it can easily backfire by stalling the rest of the discussion. It may be appropriate to have this discussion later but now is definitely not the time. ‑‑Trialpears (talk) 21:29, 21 January 2020 (UTC)

Spreading the Word

(Placed here in response to @Trialpears: so discussions don't get split across pages)

I think that we should carry on just with those editors interested from the NAC page until a rough consensus is found amongst us. Then drop a link and request for additions to the various specific groups - XfDs and PM in particular, and only then present a clear updated proposal through RfC/VPP/CENT Nosebagbear (talk) 11:28, 20 January 2020 (UTC)

Sounds like a plan. Good thing you brought up PM. I have some experience from mergers and don't think it's ready for any closing restrictions. Many, many mergers are performed boldly and that is currently the only way to be even close to keeping up with the backlog. Most mergers aren't closed at all and only a negligible proportion pass through PM. It should be left out from the proposal entirely. The same problems but worse are present at WP:PROPSPLIT. ‑‑Trialpears (talk) 12:21, 20 January 2020 (UTC)

Drafting RfC

I think we should be drafting the actual RfC as well. Here is my very rough draft of how the nomination could look, it still lacks most of the arguments for the proposal since that will be highly dependant on what the actual proposal will be and perhaps better presented as initial !votes so we can each present our own reasoning leaving the nomination to explain the proposal, its aims and how it will be implemented. Feel free to edit it freely. ‑‑Trialpears (talk) 21:52, 21 January 2020 (UTC)

Any opinions? ‑‑Trialpears (talk) 07:57, 7 February 2020 (UTC)

@Trialpears: - I'm going to drop a line to the various relevant groups, asking them to come give their suggested additions and any specific local thoughts. Nosebagbear (talk) 09:07, 7 February 2020 (UTC)

@Trialpears: - sorry for the inaction (you might have seen something in the news about the disruption that caused it). Before I notify the relevant groups, should I like them both to the draft page and also to here (the talk page), where the draft userright is, or should we add that in as a provisional section in the main draft document and amend as feedback is gathered? Nosebagbear (talk) 09:12, 4 May 2020 (UTC)

Nosebagbear Don't worry about inaction, we all have it though now and I haven't been particularly active either.

I think it would be best to link both the main draft and this talk page since there is some information about implementation since there is some information about implementation that wouldn't be in the final version. A temporary section is also possible, but I think it's simpler just linking both. ‑‑Trialpears (talk) 19:59, 5 May 2020 (UTC)

Draft

Should non-admin closer become a user right?

There is a long history of non-admin closures being a valuable tool in reducing admin backlogs and contributing to the smooth running of our discussion boards. There has also been a long history of inexperienced users closing discussions inappropriately leading to the community spending valuable time at avoidable discussion reviews, inappropriate content being kept and the community's voice not being heard. This proposal aims to both improve the quality of closes by our already competent group of experienced non-admin closers by formally showing our trust in their judgement and help them improve as well as preventing the large negatives of questionable closes by editors without the required experience.

While we believe this proposal has great potential to improve many discussion venues we recognize that a one size fits all system won't work with large differences in culture and technical requirements for different boards. Therefore this RfC has one section for each venue where details specific to the venue can be discussed, including opting out entirely and keeping the status quo. There will also be a main discussion section where general comments can be made, changes proposed to the NAC rewrite documenting the change or implementation discussed. The final section will be for voicing your opinion in the form of a support or oppose to help the closer to determine an overall consensus.

The proposal is to make non-admin closer a user right giving access to tools used for closing discussions such as WP:XFDcloser as well as allowing any user to revert closes by users without the non-admin closer user right, no questions asked. This system will still allow any user to close discussions as withdrawn or perform uncontroversial clerking such as closing as speedy delete in deletion discussions concerning already deleted pages or move discussions filed at the wrong venue.

The proposed criteria for granting are modeled after the page mover right which was intended to be used by RM closers with the same recommended tenure and edit count. The 3 other criteria are concerned with experience and knowledge of policies and guidelines reflecting our current expectations for a good non-admin closer. They are as follows:

1. The editor should be a registered Wikipedia user for at least 6 months.
2. The editor should have made at least 3,000 overall edits.
3. The editor should have voiced an opinion in at least 50 discussions.
4. The editor should have demonstrated knowledge of and competency in applying policies and guidelines relevant to the discussions they wish to close.
5. The editor should have read and understood guidelines related to closing discussions.

As with other user rights administrators can grant it at their own discretion and there will be a dedicated section at PERM. Granting access to scripts will work similarly to how it currently works for AWB with users being listed at a checkpage.

For the user right to be removed the user has to demonstrate a pattern of performing inappropriate closes or getting them reverted, fail to explain their closes when challenged, be inactive (no edits) for 12 months or lose the trust of the community for example by performing vandalism, engaging in sock puppetry, using the user right to gain an advantage in disputes etc. Proposed wording can be found at Wikipedia:Non-admin closure/draft.

If this proposal passes the sections Guidelines for granting and Criteria for revocation of the NAC rewrite will become policy, a new request for permissions page will be started, scripts used for closing discussions will be modified to require the new user right and administrators active at affected venues will be encouraged to add all active and competent non-admin closers at the venue to the check page to smoothly transition to the new system.

• I would strike the 3,000 overall edits. This is a high bar compared to other permissions and edit count does not necessarily lead to experience in understanding policy. The important one is that someone voices an opinion in at least 50 discussion. --Enos733 (talk) 17:46, 4 May 2020 (UTC)

  Enos733, the edit count is definitely not the most important factor and I mostly choose that number because of the page mover RfC. It may not be needed or could be lowered, but I think it is useful to include a figure since most people would expect an edit count in the guideline. It shouldn't be a deal killer either way though since admins have a large amount of discretion as can be seen at WP:PERM/TPE where many successful candidates don't actually fulfill all 5 points in the guidelines for granting section. I would imagine that admins would look at requests here in the same way weighing experience in discussions significantly more than edit count. ‑‑Trialpears (talk) 20:26, 5 May 2020 (UTC)

Reading through the page mover RFC, the closer recognized that the edit count did not reach consensus, but described it as a compromise. If you move forward as is, I would like to insert the language: "The above items are guidelines. An administrator may grant non-admin closer rights to users they otherwise deem competent and may deny the requests if they do not see a need for the tools or have other concerns." --Enos733 (talk) 20:57, 5 May 2020 (UTC)

Good point. I've added your text now. Thanks for the feedback! ‑‑Trialpears (talk) 21:04, 5 May 2020 (UTC)