User talk:Mz7/How to fix administrator recall

Bullet Points 1 and 2

In your first bullet point, you state: "there was the impression that the petition would just be a lightweight filter for frivolous requests, and the re-RFA would be where we actually examine the evidence". But in your second bullet point, you state: "No hard requirement for evidence, consensus, or prior dispute resolution. All it takes is for 25 editors to sign—regardless of the strength of the evidence". Admittedly, I haven't fully read the entirety of the essay yet, but this is it right at the beginning and it lacks clarity if you think the recall should just be a breezy 25 signature process with the evidence presented and discussed later or if the 25 signers should come with evidence in tow. I would suggest reworking the top bullet points so they don't (appear to) fly in each other's face. Useight (talk) 18:34, 21 October 2025 (UTC)[reply]

Thanks for the feedback. I will try to rework the lead to clarify this, but my point was that the current process is essentially a 1-step process to reach a desysop: instead of being a breezy, lightweight filter for a more evidence-based process, the 25-signature process is the entire desysop process. We wanted to make it a 2-step process where the re-RFA would be where we analyze evidence, but the reality is that the petition alone is what desysops admins. We should make it a true 2-step process by having the second step after the petition be an automatic, evidence-focused consensus discussion rather than a re-RFA. Mz7 (talk) 23:13, 21 October 2025 (UTC), edited 23:17, 21 October 2025 (UTC)[reply]

Shorter petition time + removal of immunity

I like much of what's here (and will eventually go say so over at the VPP RFC), but I'm concerned you're being overly optimistic about the interaction of these two points. If there's no immunity period, and the petition's sitting at 20ish endorsers, what's to stop them from just starting another petition immediately after the first one times out? Any admin who tried to "deal with [them] as usual for editors who refuse to get the point" would so surely immediately face a recall of their own that I doubt any would take that step. —Cryptic 00:06, 22 October 2025 (UTC)[reply]

I agree there needs to be some immunity period and I'm not sure that the current one is too long - if new evidence comes to light that can't wait until after the immunity period then take it to arbcom. If it's not serious enough to take it to arbcom then it actually can wait until after the immunity period. Thryduulf (talk) 00:39, 22 October 2025 (UTC)[reply]

I also think that we need some type of immunity period. I think it's inevitable, actually: Either we have an immunity period from the start, or we add one after some WP:AE admin gets harassed repeatedly by a series of socks. We could start the betting pool now for which CTOP subject would be the first, and whether the admin will quit before we identify and solve the problem. WhatamIdoing (talk) 05:38, 27 October 2025 (UTC)[reply]

slightly

For someone doing their recall via AELCT, I think it is disingenuous to say that that 55% vs 70% is "slightly" lower. That is significantly different. — xaosflux ^Talk 15:59, 22 October 2025 (UTC)[reply]

Some thoughts

I appreciate the thought that has gone into this. Asilvering showed this to me a few days ago. There's a lot I agree with here, including concerns about length and the inactivity-related use. My strongest feeling about the whole process is that the threshold the petitions need to clear is too low given the size of our community: you can get 25 editors to sign approximately anything in a 30 day period. There is a clear need to tighten that.

I'm less certain about requiring consensus to desysop. The analogy to STATUSQUO only goes so far. We do give some weight to the status quo ante, but also the default state for an editor is to not be an admin. But at the moment this is a bit academic, because we've had exactly one RRFA, and it was withdrawn early when its trajectory became clear. About thresholds more generally, I find it difficult to advocate for requiring more than 50% opposition to retaining administrator status.

The portion of this proposal that merits most discussion in my view is its biggest innovation, requiring evidence for a petition. This is an excellent idea in principle, and is how we handle most other discussions involving a loss of status; but it is rife with difficulties as to who would evaluate evidence, and how, because noticeboard threads about admin conduct are not so easily categorized. If there's an interest in implementing this proposal wholesale, rather than in tweaking the numbers around petitions and RRFAs, we need to address this. Vanamonde93 (talk) 03:37, 23 October 2025 (UTC)[reply]

Also thoughts

Interesting reading, good food for thought. I'm not particularly opposed to this idea, despite supporting the status quo with recall for now (mainly on the basis of retaining it rather than abolishing it). It doesn't appear known how many recalled admins would resign during the proposed 48 hour holding time, rather than face an RRfA that would be slightly more in their favour. But overall I get the impression it'd be similar, as 55% for and 60% against is not an enormous difference here, even if one requires slightly better consensus than the other. Overall there is nothing wrong with the theory of requiring consensus for desysop, that I can support in principle (per not a huge difference to RRfA), even if I still don't see it as necessary quite yet and don't believe it would change the outcome of participation. If there were responses from the 8 recalled admins that didn't do an RRfA supporting such a change, I'd be more inclined to support however. To me it still seems like a solution to a problem that doesn't exist yet, even if overall could well be an improvement to how the process functions. So to sound like a broken record still, we should at least wait until it's broken before fixing it, even if we already think we know what the solution is. CNC (talk) 15:18, 23 October 2025 (UTC)[reply]

My thoughts on your points on RFA

I disagree very strongly that RFA sucks. Is it a perfect process no but no process we have would be perfect but as a process it allows for good feedback to be given and the recent RFA's I have seen have gone well with alot of valid points being provided. Also, if by RFA you mean just the RFA not ELECT I would note that is another option as well as RFA for the recalled admins depending on the timing. GothicGolem29 (talk) 16:02, 26 October 2025 (UTC)[reply]

I thought that WP:AELECT was not authorized for recalled admins, and it had to be the old-style WP:RFA. But I was wrong. WhatamIdoing (talk) 05:41, 27 October 2025 (UTC)[reply]

Does make sense that both are allowed though I was wrong on some of the specifics of it so I am not surprised someone didn't realise WP:ALECT could be used. GothicGolem29 (talk) 16:29, 27 October 2025 (UTC)[reply]

Similarity to 2005 proposed desysop process

This is very similar to what is proposed 20 years ago. GZWDer (talk) 16:14, 3 November 2025 (UTC)[reply]

Question period?

I believe we can add a "question period" between petition and vote period. The following is adapted from pre-2021 RfDA process in Chinese Wikipedia:

Discussion period: If a user (not necessary confirmed user) alleged an admin (or crat/CU/OS) is misusing their permission), they can open a discussion on an appropriate discussion forum (e.g. ANI in English Wikipedia).
Start of petition: A "user eligible to vote" (can be extended confirmed user) can create a RFDA at Wikipedia:Requests for de-adminship/USERNAME, if there is a prior discussion and the admin misuse happened at least 48 hours ago (but within one year). That user then becomes the first user to sign the petition.
Petition period: Petition period lasts for 7 days or when there are 7 eligible signers, whichever is earlier. If there are less than 7 signers, the RFDA fails.
Question period: The admin then have 5 days to (optionally) comment on issues raised by community.
Vote period: A vote starts 5 days after the petition period ends and lasts for 14 days. The users to sign the petition automatically becomes supporters (though can change the vote in the vote period). Admin will be desysoped if there are >=25 total votes and the support rate (support/(support+oppose))>50%.

--GZWDer (talk) 16:32, 3 November 2025 (UTC)[reply]