Wikipedia:Requests for comment/Recall check-in

Recall check-in

English Wikipedia's recall process was largely based on German Wikipedia's recall process, but it has played out differently here than it did on German Wikipedia. Now that we've had 10 recall petitions it seems like a good time to examine the process.

Support 1 or more of the following:

1. Process is working well, no changes needed
2. There should be some way of enabling support for the admin during the petition phase
3. There should be fewer signatures needed
4. There should be more signatures needed
5. 30 days is too long, the petition process should be shorter
6. 30 days is too short, the petition process should be longer
7. Keep recall, but develop a different process than petition leading to a re-RFA
8. Keep recall, but do some other change to how re-RFA works
9. Keep recall, but do some other change to how the petition works
10. Recall should be abolished
11. Prohibit admins from !voting in RFCs to amend recall

When closing the closer is encouraged to think about overall support relative to participation in the RfA (e.g. if 5 people support Foo, 10 people support the opposite of Foo, and 30 people didn't support either but participate elsewhere, the consensus may be no change rather than opposite of Foo) and where a bartender's close may be appropriate. Barkeep49 (talk) 14:31, 21 October 2025 (UTC)

Recall check-in survey

• Support D, weak support for B, open to G. I am very opposed to abolishing a community recall, but I think we can do better than our current process. Right now we have changed a system where 15 highly scrutinized and widely endorsed editors make a decision about an editor's fitness for adminship (ARBCOM), to 25 self-selected people. I don't think that's a good change. This also stands in contrast to the 200+ people that can be counted on, in the current era, to show up to an RFA. We based our process on the German Wikipedia's process, but didn't adjust the number of supports needed despite being a much larger community. Perhaps for this reason, perhaps for other reasons, we also haven't seen most candidates go through with the re-RFA, which is a difference between the two. This can perhaps be explained away for good reasons (the "rightness" of the recalls to date) or perhaps ones that deserve scrutiny (differeing cultures of enwiki and dewiki). The idea of allowing editors to support may help admins who have support from the larger community realize that and not think they are starting off a re-rfa with 25 opposes or perhaps it's just better to raise the the threshold so borderline petitions without widespread community support fail at that stage, but I don't think we should continue where a 25 signature petition essentially is a desysop because I find it completely out of line with our other community practices, norms, and values. Best, Barkeep49 (talk) 14:36, 21 October 2025 (UTC)

  I support Mz7's analysis of the issue below. So one change I would be open to, in addition to what I've already supported, is actually dramatically lowering (to say 3 or 5) the number of signatures needed combined with requiring everyone involved to be extended confirmed. This would (hopefully) keep out the purely frivolous petitions (none of which we've had so far depending on how one views the inactivity ones) while also not starting the admin "down" 25 votes at a re-RFA. Best, Barkeep49 (talk) 16:11, 22 October 2025 (UTC)

• Support D and/or E, Support B assuming the details work, Very open to and interested in developing G - Very similar to Barkeep here, this is a net positive process that would benefit from tweaks. I think the easiest thing to do would be to reduce the duration of a petition down to 14 days - most petitions we've had have long-since succeeded or failed at that point. I like B in principle, but am wary of accidentally creating a mudslinging mess, in which case we might as well be at RFA. The biggest issue is that the RRFA at the end in practice hasn't happened, so the 25 signatures is a de facto desysop. I haven't seen a recall that went strongly off the rails yet, although Necrothesp's came close. Tazerdadog (talk) 14:52, 21 October 2025 (UTC)

  I very much like Mz7s proposal below, and think it would be a large improvement over what we have now. Tazerdadog (talk) 16:56, 21 October 2025 (UTC)

• Support A, but also remove the little clause that allows people (i.e. other admins) to remove a recall petition prematurely if other means have not been explored first or whatever it is the get-out clause says. (edit: "Other methods of dispute resolution should be attempted before a recall petition is initiated.") - oh and also "An editor may sign no more than five active petitions at a time." is sketchy af.  Tewdar  14:54, 21 October 2025 (UTC)

  Also there should be alerts for recalls like we get every time there's an RfA, and a very big alert for every user (like, email and Web notifications) any time anybody attempts to make even the slightest alteration to the recall process like what seems to be happening here. And support K, what a great idea. Oppose B, D, E, G, and J.  Tewdar  10:21, 23 October 2025 (UTC)

• Support what let's call I.a (1 signature per user per 90 days) per my comment below. Weak support D as second choice to I.a: I don't think it'd make a huge difference to go with a bigger number, and there's some argument that it could be even more discouraging of RRfA than it already is, which is why I would prefer I.a, but if we retain a system where anyone can support every recall (beyond the so-far inoperative restriction on 5 simultaneous signatures), then increasing the threshold is better than nothing. If we're going to do that, I don't think it'd make much sense to just tweak it a bit, though; something like 40 or 50 would be preferable. Neutral on G&H. Oppose all others, in particular B (that's called RfA!), C (obviously per !vote on D), and of course J. -- Tamzin^{[cetacean needed]} (they|xe|🤷) 14:55, 21 October 2025 (UTC)
  • I'd be all in favor of your I.a - even if you can't vote against another admin solely because their recall came up before your timeout period was over, nothing's stopping you from using the discussion section to try to convince other users to sign - except that, in practice, I think it'd be more likely to encourage sockpuppetry than active persuasion. Same reason I didn't seriously propose that the ambiguity closed by these edits be resolved the other way. —Cryptic 02:23, 24 October 2025 (UTC)
• Support A as first choice. After 10 recalls, no obvious issues with the system have become apparent, so there is no actual need to change things. Having said that, I would not be opposed to shortening to nothing less than 15 days, and am certainly opposed to making the process longer than the current 30 days. Lowering the number of signatures seems like a bad idea, raising it to more than, say, 40 seems like a bad idea as well (the 25 is "there are serious concerns with your adminship, please either stop being an admin or at the very least go through RfA again": if you raise the bar to 50, 60, ... signatures, then you are just saying "you are not fit to be an admin at the moment" and can simply abolish the whole RRFA thing). I see no reason to support BCFGHI at the moment, and no reason at all to support J. Fram (talk) 15:10, 21 October 2025 (UTC)
• Support A given this RfC is being held. This is not due to a particular strong feeling regarding many of these, but because that lack of feeling is due to many of these seeming potentially cosmetic tweaks that would not greatly change the underlying process. Oppose B though, on practical/procedural grounds. The point of the petition is to bulwark against stray fires. Allowing supports at that stage seems liable to essentially turn the petition into another (R)RfA. CMD (talk) 15:35, 21 October 2025 (UTC)
• Support E and/or D per Barkeep. 30 days is a long period. 14 or 15 days seems about right to make sure that the community is aware of a petition. Oppose others. --Enos733 (talk) 15:37, 21 October 2025 (UTC)
• Support A largely per Fram, oppose limits on supporting at this time. --SarekOfVulcan (talk) 15:39, 21 October 2025 (UTC)
• Support B and D Struck in favour of User:Mz7/How to fix administrator recall While I take Tamzin's point that B is what you get at RfA, I still think this would be a good idea, as it directly addresses my main concern with the recall procedure, namely that it only counts votes against the admin, completely disregarding the level of support they may have. So the relatively small number of 25 people (which IMO should be doubled, hence D) can force an RfA/election, even if 250 were perfectly content with the admin in question. And before someone says "but in that case they'll have no trouble getting the mop back"... well, maybe, if they choose to go for it. What if they feel so betrayed or snubbed or whatever that they don't even want to try? Then they'll never find out that their supporters outnumber the opponents by an order of magnitude. We don't grant the mop by counting only support votes, why potentially take it away by counting only opposes? --DoubleGrazing (talk) 15:43, 21 October 2025 (UTC)

  Only just seen Mz7's excellent User:Mz7/How to fix administrator recall, which I fully agree with. That would be my preferred approach going forward. I guess that then amounts to Support G/H/I, specifically on those lines. (Will strike the bolded part of my earlier opinion.) -- DoubleGrazing (talk) 12:30, 22 October 2025 (UTC)

• Support B Agree with DoubleGrazing – if we select admins via a vote, why shouldn't the process for deselecting them be the same? Basing recall purely on oppose votes would be like only considering support votes in an RfA (which I don't think anyone would consider a good idea?). Number 57 16:02, 21 October 2025 (UTC)
• Support B and D per DoubleGrazing, but perhaps more immediately we should require an actual consensus in favor of the recall (or at least a 50% threshold in favor of it), in order to avoid wasting resources on bureaucratic processes and potential system-gaming as an end-run around disputes. BD2412 T 16:06, 21 October 2025 (UTC)
• Support A seems to be working fine enough. The first part is not supposed to be a full debate, it is just a check to avoid wasting the subject and the communities time with frivolous requests - and that is working. — xaosflux ^Talk 16:26, 21 October 2025 (UTC)
• Support D and E, okay with B. If you can't find, in a month's time, 25 people an admin has irritated at some point or another, they either haven't been one long or are totally unwilling to make tough calls in contentious areas. That should not be what we want; there are a lot of cases where someone's going to hate your decision no matter what it is, and often those are the cases which most need to be dealt with. So, make it shorter and require more signatures, to reach a point of "This admin really has made an egregious bad call or a lot of them", not "This admin has made tough but defensible calls that were sure to annoy someone no matter what." B would make a reasonable alternative to that in principle, but then we basically have the possibility of two repeat RfAs, and the thought of one is bad enough. Seraphimblade ^{Talk to me} 16:33, 21 October 2025 (UTC)
• Support D I certainly don't think these need to be open longer as it seems like most of the recalls get 25 signatures long before running out of time. It might be good to provide some way to allow a defense of admins during the survey phase though. This seems like something missed. Simonm223 (talk) 16:36, 21 October 2025 (UTC)

  Is the best solution raising the number of signatures, or is it shortening the time to gather them (i.e. would option E be just as good?) Tazerdadog (talk) 16:44, 21 October 2025 (UTC)

  I would also support raising the number of signatures. I'm not sure shortening the time would really do much TBH. But I do think it's a good idea to allow some form of support statement for admins going through this. I'll be honest: the main thing that has prevented me from trying out for an admin position is the fact that they have to run a gauntlet just to pick up the mop and then, if people get upset about their comportment, we expect them to sit in silence and wait to see if they've upset enough people to then be told not to do this anymore. It's not a great way to treat some of our more-engaged volunteers TBH.

  I understand we need checks and balances for when admins screw up. I've even signed one or two petitions. But I think we need to give some way for the process to allow people who think the admin is doing good work to be involved in the decision. Simonm223 (talk) 14:11, 22 October 2025 (UTC)

• Support G. Broadly, I do agree now that a community-based desysop process like this is and has been desirable. (I did not always hold this view.) However, there are several key flaws with the recall system as it is currently formulated, and I have written a bit of a lengthy essay at User:Mz7/How to fix administrator recall with my thoughts.
  The high-level summary is this: a certified recall petition is a desysop. When we supported creating the recall process, 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 and discuss whether to desysop. The reality is that the petition itself is what functionally desysops administrators and forces them to reassert themselves as an RfA candidate just like any other RfA. The only functional difference between a regular desysop and a successful recall petititon is that the effect is delayed by 30 days, and if the recalled admin chooses to do an RfA within those 30 days, the support threshold is slightly lower. So what does this mean? It means that we now have a process that lets 25 self-selected editors desysop admins with no requirements for evidence, consensus, or prior dispute resolution.
  What do I think the solution is? I've detailed this a bit more at User:Mz7/How to fix administrator recall § Proposed solution, but at a high level, I think that the recall process should require consensus to desysop in order to desysop. A successful recall petition should lead to a request for de-adminship (RfDA) instead of a "re-request for adminship" (RRfA). This shifts the focus of the recall discussion to evidence that an administrator is actually causing some active harm (or would be likely to cause harm in the future), and it converts the petition system into being a legitimate filter against frivolous desysop requests, as opposed to being the desysop process. Mz7 (talk) 16:41, 21 October 2025 (UTC)

  I would support User:Mz7/How to fix administrator recall#Proposed solution, but I do think it is important to hash out where ADMINRECALL (or a similar process) fits into the structure of Wikipedia (and whether doing something different would cause issues). In my mind, a ADMINRECALL is a community-led L2 desysop for obvious long-term errors and I wonder if a RDfA raises the bar there or keeps it the same. Sohom (talk) 17:07, 21 October 2025 (UTC)

  I would support this proposal. JuxtaposedJacob (talk) | :) | he/him | 05:11, 25 October 2025 (UTC)

• Splitting my !vote into multiple parts.
  • Support D > A Raise the required number of signatures to 30. Defense of the admin can always be put in the discussion section or the talk page. This was especially prevalent in the first two petitions, which had bolded Opposes. Also, based on the discussion below of Cryptics query, we should raise the suffrage requirements. I don't know to what standard exactly, but we definitely need to raise them.
  • Strongest possible Oppose J While Novem and others make great points about removing admins at a time of needing admins (which is very true), sometimes we need to prioritize quality over quantity (and I have very low standards compared to most). The whole point of recall was to give people more reason to !vote support for candidates, because if the community made a mistake, we can always recall the admin. It's unfortunate that even with recall, some people still refuse to support decent admins, as evident at AELECT and a few RfAs.
  • Weak Support K I'm torn here. On one hand, I do think everyone should have at least some say on the process; on the other hand, this looks like a conflict of interest on the admin's part, as they are directly affected by the outcomes of this RfC, and future RfCs. As others have mentioned, this is most definitely a bad look that we do not want to have. I agree with both sides on this option, but unfortunately, I agree with supporters slightly more than opposers of K. fanfanboy (block_talk) 12:31, 27 October 2025 (UTC)
• Somewhere between G and J -- I share some of Mz7's concerns and remain unconvinced that recall is the necessary format or that existing measures wouldn't have been adequate to address the 10 examples we've had. I think in the current format both the time period and the number of supports are out of whack (too long, and too few). I reiterate the points I made in Night Gyr's case about the potential for misuse of recall to allow a smaller local consensus to bypass our project-wide policy on administrator inactivity. But addressing that is not one of the available options, so I'm stuck between "throw it out and start over" and "throw it out and move on", either of which I'd be fine with. ⇒SWATJester ^{Shoot Blues, Tell VileRat!} 19:38, 21 October 2025 (UTC)
• Support for B and/or D, weak support for G recognizing that it's very easy to say "come up with something else" but that doesn't really get us very far. _signed, Rosguill ^talk 19:43, 21 October 2025 (UTC)

  A lot of the discussion here has focused on the inactivity-based recalls, and those in favor of the status quo have expressed a dismissive attitude to hypothetical situations. While that's not unreasonable, I would flag that my primary concern here is how easy it would appear to be to abuse this process--a hypothetical, but hardly irrelevant, concern. With the rules set as they are now, I imagine that with a bit of legwork and sockpuppetry, I as a lone individual could get someone de-sysopped. I imagine it would take about a month of work to do this sloppily, six months to a year to do it semi-convincingly, and a few years to do it in a fashion that would be difficult to detect or dispute. That's my assessment of what an individual could do--a well-funded organization or state could accomplish it much more readily and convincingly. Of course, such a recall would likely raise suspicions and be called out at least part of the community, and I do have some faith that we'd survive the ensuing crisis of trust. But I think it would be better to modify our practices to avoid such a crisis entirely, and that support for the status quo is being naive with respect to what forces external to our community would be willing and able to do. _signed, Rosguill ^talk 14:50, 23 October 2025 (UTC)

• D (with 50 signatures) and E (down to a week or two) are the absolute minimum that is needed. —Kusma (talk) 19:46, 21 October 2025 (UTC)

  And I (there should be some actual issue stronger than "admin meets the letter of the activity requirements"). Lots of good stuff in Mz7's essay too. —Kusma (talk) 10:04, 22 October 2025 (UTC)

  And K (proposed by a single-purpose IP) shouldn't have been added to the list of proposals. A topic-ban for a thousand people (including me) should not even be debated without evidence that it is required. oppose, in case it isn't clear, but suggest striking. —Kusma (talk) 18:09, 25 October 2025 (UTC)

• Support for B Support for the admin needs to be factored in to the equation, and not simply await an RfA that the admin may have no stomach for..--Wehwalt (talk) 20:17, 21 October 2025 (UTC)

  I support also any alternative (obviously not A) which addresses my concern. Wehwalt (talk) 11:03, 22 October 2025 (UTC)

  And I agree with most of User:Mz7/How to fix administrator recall and consider the proposals stated therein an improvement on the status quo. Wehwalt (talk) 12:59, 22 October 2025 (UTC)

• I'm hesitant to provide any !votes on a process this open-ended, because the number of proposals more or less precludes specific consensus emerging, and I worry about what a bartender's close might produce. In the interest of aiding a temperature-taking exercise, I will note that I explicitly oppose A, C, and J. I am open to the other ideas but specifics are needed: there are distinct directions we could go that are mutually exclusive but would represent improvement.
  The rationale behind a two-step recall process is reasonable: the first step is supposed to be "light-weight", and simultaneously act as a filter against bad-faith requests; and an admin who was moved to the second stage but still retained community support ought to be able to pass RRFA as we have set it up.
  I still believe this reasoning is sound. And to the extent that we have used it with active administrators who made serious mis-steps, it has worked as intended. The difficulty with it has been in two specific areas, in my view; first, that given our large editor community, getting any 25 editors to sign something is perhaps too easy; and second, that when the subjects of recall are admins who are not very active, the perceived cost-benefit ratio of an RRFA is not favorable, and consequently administrators resign instead of running at RRFA.
  The first problem is easier to fix. Given the size of our community, 25 votes in 30 days is far too few. Most admins who work at SPI or AE or ANI would likely accrue that many. If we preserve a two-step recall process, the first step must set a higher bar than the number of editors an active admin taking good decisions might annoy. The second problem is difficult: when the justification for recall is losing the community's trust, we cannot effectively prohibit inactivity recalls. But I am hopeful that addressing the broader issues will help.
  That said, I am not certain that a two-step process is necessarily the best process long-term. I can see a genuine case for a one-step process: essentially, immediate RRFA. This has some major attractions: my gut tells me that shortening the process and allowing support will improve the effect on admin morale even if the eventual outcome is the same. The biggest difficulty with this is potential inconvenience to the admin in question, but this is a fixable problem. I rather think we need to settle on whether we want a two-step process, before refining the steps, at least for a trial period. A second RfC will be necessary. Vanamonde93 (talk) 20:54, 21 October 2025 (UTC)

  @Vanamonde93, what would trigger "immediate RFA" in the one-step process? -- asilvering (talk) 02:44, 23 October 2025 (UTC)

  @Asilvering: I could see it playing out in different ways: we could, for instance, allow it to be triggered by a good-faith request after an admin action was overturned via noticeboard discussion. Arguably the simplest way is what's being implied with the suggestion of allowing supports at the petition stage: if we're allowing support and opposition for retaining admin status, we have an RRFA triggered at the request of any EC editor. It's possible I didn't choose the best words above, though. I'm not advocating for dropping the petition stage; certainly, I think allowing an RRFA to be initiated at any time by most editors would be a mistake, which is why I'm not supporting B. But I do think we need to give serious thought to whether drawing out the process is the best way to balance admin morale and community recall. Vanamonde93 (talk) 02:57, 23 October 2025 (UTC)

  I want to note in addition that I oppose K as contrary to our ethos, and on practical grounds. Like it or not admins and former admins are the only ones who have first-hand evidence of the experience of adminship. Prohibiting us from commenting on recall is an absurdity. It's also based on a conception of admin solidarity as distinct from Wikipedian solidarity that I see no evidence for. Empirically, most admins who have been desysopped or otherwise held accountable for their actions have had their feet held to the fire by other admins. Vanamonde93 (talk) 17:12, 21 November 2025 (UTC)

• Support A; nobody has presented any convincing evidence of problems. And oppose B in strongest possible terms as a suggestion that completely misunderstands the purpose of the petition. The sole purpose of the petition is solely to prevent people from having to go through a rough, adversarial process in the case of completely frivolous recalls, which every indicator shows that it is doing. It's not supposed to be a stumbling block for any recall with even the slightest chance of success. Making the initial petition adversarial would defeat the entire purpose for it - in that case, why not just go straight to a single RRfA? The people who decline to continue to a RRfA after a successful petition are usually doing so because they want to avoid that adversarial process; if we made the petition adversarial, they'd just resign as soon as it started getting off the ground. And if it were adversarial, what would be the point of the RRfA afterwards? Basically, for the people supporting B, how do they envision the petition differing from the RRfA, and why do they think we should have a two-step process at all? (Also, I will point out that this RFC contains a bunch of complaints about the petition stage, but doesn't offer the obvious option of "remove the petition stage, have recalls consist of a single RRfA." If people are viewing the petition stage as a referendum on an admin rather than a quick-and-easy sanity check, perhaps the problem is that we should just have gone with a single-step process in the first place? We already have a solution to obviously frivolous requests, it's called WP:SNOW. Anyone can attempt to open an ANI or AE thread about a user seeking to get them banned at any time, no matter how frivolous it is, and we somehow deal with that; how are recalls different? For that matter, why is it so much easier to seek to get an administrator banned than it is to get them desysopped?) --Aquillion (talk) 20:58, 21 October 2025 (UTC)
• Support G. I agree with Mz7; it is a problem that an admin can be removed without even a simple majority consensus ever being achieved for removal. I think User:Mz7/How to fix administrator recall#Proposed solution makes sense, though I would make it a little easier to remove an admin; 50% should be enough (which is what Commons uses). In any case, 60% is still better than 40%, which is what the current policy calls for. -- King of ♥ ♦ ♣ ♠ 21:21, 21 October 2025 (UTC)
• Support User:Mz7/How to fix administrator recall. So support E (having this hang over your head for 30 days sounds like torture) and I suppose support G per Mz7. I'd support D, but really would prefer a successful recall to automatically transition into a "request for de-adminship" as Mz7 proposed. Guettarda (talk) 21:31, 21 October 2025 (UTC)
• Support A, as the process has repeatedly shown that it's working just fine and as intended. I'm fine with supporting D as well, just in the hopes that it would stop more people from whining about every recall petition that happens. Doubtful that would be the case, but one can dream. Silverseren^C 21:58, 21 October 2025 (UTC)
• Support B, I and much of User:Mz7/How to fix administrator recall. Specifically it is a problem that 25 people can effectively desysop an admin before they're even aware that there is a recall process. Petitions should not be open for signatures until the earlier of the admin concerned explicitly noting that they are aware of the petition or 24 hours after the editor's first edit after being notified of the petition on their talk page, and there should be a way for editors to actively oppose recalling the administrator. Thryduulf (talk) 22:01, 21 October 2025 (UTC)

  I'm very surprised to see so many comments along the lines that recall is working as intended - please could those editors show some evidence of this with reference to the discussions that resulted in recall. I've read those discussions and the reality does not match the intent of what most commenters there were supporting. Thryduulf (talk) 22:01, 21 October 2025 (UTC)

  Thryduulf, it seems very odd that you're only demanding people to "show evidence" that recall is working as intended, but not for the reverse. How are you deciding which comments are self evident enough that the "reality does not match the intent of what people were supporting"?

  It seems to me that you are decided that RECALL must only work the way you think it should, and everyone else must always explain themselves. Soni (talk) 00:58, 22 October 2025 (UTC)

  Okay, here's some proof: In the RFCs that led to RECALL's creation, nobody expected it to be used to desysop admins who were skating on the thin edge of INACTIVITY. There is not a single !vote for it that says anything remotely like "This will be good for INACTIVITY cases, too". The comments are about actively abusive admins, not about absentee ones. But a majority of the RECALL petitions so far have been about INACTIVITY. Therefore, RECALL is not working "as intended". WhatamIdoing (talk) 02:51, 22 October 2025 (UTC)

  @WhatamIdoing 3/10 have been about inactivity, my by count. GreenLipstickLesbian💌🦋 02:58, 22 October 2025 (UTC)

  I would argue that out of those three inactivity cases, two were of the editors were WP:GAMING to stay within the activity criteria, which falls squarely into "abusive behavior" that should have community review. Sohom (talk) 03:06, 22 October 2025 (UTC)

  Firstly there is absolutely no consensus that either those editors were actually gaming, or that gaming the inactivity requirements is "abusive behaviour". Secondly, a recall petition is not, in practice, a community review - it's a place where an editor has 30 days to see if 24 other people agree with them that an admin should be desysopped (for any reason or no reason). If you can find any comments in the discussions that resulted in recall intending that then please link to them here as I've not found any. Thryduulf (talk) 03:12, 22 October 2025 (UTC)

  The RECALL petitions and surrounding discussions shows there was enough agreement among editors that there was a case to be made for GAMING the criteria. The RECALL petition is the start of a process of community review which culminates in a re-RFA, the fact that the administrators decided to not engage with the community review process is their decision. That does not allow you to assert that there were no concerns of GAMING the inactivity criteria. I'm not arguing that the process is perfect, but I don't think those two RECALL petitions are necessarily standout examples of RECALL going wrong. You'll see that I have explicitly said out of those three inactivity cases, two were WP:GAMING because I see the third one as being a particular display of what is wrong with the process, in that you can convince 25 people to vote against me in 30 days, if you spin any set of events correctly. TLDR, I think INACTIVITY is the wrong hill to die on, but I do think recall has it's flaws Sohom (talk) 03:42, 22 October 2025 (UTC)

  But the fact that (some part of) 25 people signed a petition doesn't mean that the community as a whole endorses the view that carefully complying with the written rules is GAMING, either. WhatamIdoing (talk) 03:48, 22 October 2025 (UTC)

  Gaming is weird, tbh. Like, take this guy, who made about ten stupid edits that they self-reverted to get their account autoconfirmed, just so they could fix a template.[1] Indef blocked, they couldn't figure out how to format their appeal so they went away forever. Many wikipedia power users... really don't like anything that could be perceived as gaming, I guess. GreenLipstickLesbian💌🦋 06:47, 22 October 2025 (UTC)

  You're right, there are ten now. I count DG as an admin losing the bit partly due to inactivity. WhatamIdoing (talk) 03:43, 22 October 2025 (UTC)

  I can't speak for others but inactivity and gaming most certainly could lead a a user [to believe] that an administrator has lost the trust of the community, which is what I supported and believed the spirit of ADRC to be. I don't think anyone created an exhaustive list of situations they considered to warrant recall but I suspect, had there been one, inactivity would have made the list. Sincerely, Dilettante 11:40, 23 October 2025 (UTC)

• Support D/E/G - I think that a formalized recall process of some kind is necessary, and have long felt that many users are somewhat immune to traditional forms of accountability, especially when it relates to civility. That said, the current system seems to incentivize knee-jerk recalls, the signature requirement without a requirement for discussion is contrary to the basic principles of this community, and the lengthy time the petitions stay open makes hitting the requirement almost inevitable. At a minimum, I think that the signature requirement should be increased and the petition length should be decreased. ThadeusOfNazereth(he/him)^{Talk to Me!} 22:16, 21 October 2025 (UTC)
• Support A. Weak support for D; the number of signatures could be raised from 25 to 30, but I'm not opposed to keeping it at 25. Regarding B, if editors want to show their support for the admin, they can do so in the Discussion section or the admin's talk page. Oppose J. Some1 (talk) 22:57, 21 October 2025 (UTC)
• Support A per Aquillion. Tazerdadog cites Wikipedia:Administrator recall/Necrothesp as problematic, but I see the community's rapid advising there that incivility should first be addressed at WP:ANI as proof that this process is not resulting in mistaken/premature desysops, even if the pace has been higher than expected. ViridianPenguin🐧 (💬) 23:47, 21 October 2025 (UTC)
• Support J. At a time when we are having a lot of trouble finding new admins, making it easy to destroy existing ones should in my opinion not be the priority. I find the 3 recalls related to inactive admins to be especially egregious. Admin inactivity should be handled by increasing the inactivity requirements via an RFC at Wikipedia talk:Administrators, not via dragging individual inactive administrators through the mud at WP:RECALL. We often talk about how we should not WP:BITE the newcomers. How about not BITING the admins either? Keeping administrators motivated is in my opinion just as strategically important as keeping newcomers motivated. –Novem Linguae (talk) 23:54, 21 October 2025 (UTC)
• Support G, above all, with secondary support for B, D, and J. Here's what I mean. I accept that there have not been any egregious cases of admins being subjected to petitions unfairly. Overall, this has worked reasonably well. But what I see as a problem is that it really boils down to a vote, posing as a petition, in which only one "side" gets to vote. And that goes against everything that Wikipedia has enshrined in a culture of WP:CONSENSUS. It needs to be more of a discussion, and a discussion that lasts more than a day or so. Making it expeditious isn't a good enough reason, nor is the fact that nothing really bad has happened yet. I'd like to see a couple of days of discussion, before anyone other than the petition initiator can add their signatures to the petition. --Tryptofish (talk) 00:35, 22 October 2025 (UTC)
• Support J – This process does nothing to encourage administrative accountability; it simply encourages mob justice. Administrators will inevitably create enemies in the course of their daily job, and we all know this. Let's not be coy about the political nature of Wikipedia editing. After a few years wielding that vaunted dustman's mop, the rank-and-file administrator will almost certainly have twenty-five editors champing at the bit to enact a public humiliation when the time is right. Those who disagree with the petition are not even granted a voice in the matter. Why subject our administrators to this farce? Whither that well-known piece of Wikipedian wisdom, 'Wikipedia is not a democracy'? I agree in principle that administrators are dustmen, and that their tools should be readily removed when problems arise. I do not believe, however, that this process is the right way to do it. We have ArbCom, the elected representatives of the community, who at least have an obligation to honour the principle of due process. Abolish this black stain on Wikipedia history, and leave justice to those who are capable of serving it with care. Yours, &c. RGloucester — ☎ 01:51, 22 October 2025 (UTC)
• I is for WP:INACTIVITY, which should not be grounds for a recall petition because complying with the actual, written policy should not be grounds for being desysopped. In practice, I expect this to mean those who think the INACTIVITY policy should be stricter will have to find some other excuse first ("I asked him about a routine admin action from six years ago, and it took him three months to reply") or just be vague about it ("I personally don't trust this admin with the tools any longer"). WhatamIdoing (talk) 02:07, 22 October 2025 (UTC)
• Support D (50) and E (7 or 10d). Lukewarm on G, but if that happens to be supported, make it a Request to Revoke Adminship instead. From what I've been seeing, the 25-signature threshold has hardly, if ever, taken more than a few days to meet, and of all but three of them (one pending, one failed RRfA, and one withdrawn request) the responce to a successful petition has been to just say "Fuck this" and resign the tools without bothering with a process that literally everyone but the people who came up with Recall realises is pointlessly redundant, and in a few cases also leave Wikipedia for good measure since it's clear there's no trust. —Jéské Couriano v^_^v ^{threads critiques} 03:15, 22 October 2025 (UTC)

  I'll even up the ante and propose Position L: Abolish the post-recall RRfA in favor of a successful recall being a desysop, with any new RfA needing to wait until some period of time after the recall. If we are going to insist on them having a chance to regain the tools, then we should also insist on them not being forced to start a loaded RfA that is doomed to fail because it's required to happen during a time when emotions over the petition will be running highest. This would basically make it equivalent to any other situation where an administrator would have their tools revoked for cause (be it misconduct or 172 exits) where they can still potentially regain the tools at a later date. —Jéské Couriano v^_^v ^{threads critiques} 00:43, 18 November 2025 (UTC)

  That seems like a less thought-through version of User:Mz7/How to fix administrator recall mentioned by many others here. Anomie⚔ 02:31, 18 November 2025 (UTC)

  I view it as more akin to a community-led equivalent to admin-conduct ArbCom cases/deops. HtFAR still suggests a discussion after a successful petition, which still has the problems that (assuming the petition is successful) it starts in the hole against the admin and tensions/ill will are still going to be heightened. It's better to eliminate the RfA/RfDA entirely and just have the petition itself just be the deop method, since that's the current practice no matter how people want to claim it isn't. (And there's no need to split hairs about timeframe for the RRfA/RfDA if we do this.) —Jéské Couriano v^_^v ^{threads critiques} 04:06, 18 November 2025 (UTC)

• Support A - "Process is working well, no changes needed". I haven't seen miscarriages of justice so far. Counterintuitively, recall is probably helping add additional admins; it's easier to support a candidate if you know they can be recalled if they turn out to be problematic. --A. B. ^{(talk • contribs • global count)} 04:48, 22 October 2025 (UTC)
• J, or, barring that, B and D. I'd go into detail but it's late. DS (talk) 05:10, 22 October 2025 (UTC)
• Support A, Wikipedia has improved as a result of this process. AugusteBlanqui (talk) 07:12, 22 October 2025 (UTC)
• Support D and E set to 50 signatures and 7 days due to the petition being a de facto desysop rather than some sort of protection for admins. --Guerillero ^{Parlez Moi} 08:22, 22 October 2025 (UTC)
• Support A. Give it a year, then review. I see it working reasonably well as designed, through a long-term RfC. →StaniStani 08:25, 22 October 2025 (UTC)

  It's had just about a year at this point (Graham's recall is 1 year old in about 2 weeks). —Jéské Couriano v^_^v ^{threads critiques} 15:32, 22 October 2025 (UTC)

• I'm very much a support A kind of person. I note that (a) every single successful recall petition was supported by highly-regarded sysops; (b) every single successful recall petition was for a sysop who'd exhibited genuinely problematic behaviour (bearing in mind the community thinks sysops should be active in order to keep their knowledge of custom and practice current); (c) the only recall that should have failed, did fail and got withdrawn promptly; and (d) sysop behaviour has noticeably improved now that there are consequences. This process is working perfectly and it's one of the best things we've ever done.—S Marshall T/C 08:41, 22 October 2025 (UTC)

  And, oppose K. Of course admins should get a voice in this. Every single successful recall has been supported by admins whom the community respects, and if there isn't a respected admin who thinks we're over the bar for a recall !vote, we probably shouldn't be holding one.—S Marshall T/C 21:46, 24 October 2025 (UTC)

  I'm not sure if I read your comment correctly, but I'm pretty sure option K pertains only to admins participating in RfC's about amending recall, nothing about participation in the process (recall) itself. fanfanboy (block_talk) 22:13, 24 October 2025 (UTC)

  Oop, yes, I misread that! ... But now that I've re-read it I still oppose K. You've got to give people a voice in discussions that affect them. To deny them that voice is oppressive and I think ethically untenable.—S Marshall T/C 07:50, 25 October 2025 (UTC)

  @Fanfanboy: Speaking of which, you haven't voiced an opinion on K. Are you neutral on it, or would you like to !vote on it? I'd ping ever !voter about it, but there's a lot and I'm a bit busy at the moment. QuicoleJR (talk) 17:44, 25 October 2025 (UTC)

  I would say I am neutral on it, but leaning support (both sides make great arguments). I feel like K should be a separate proposal altogether, though, instead of being lumped with the rest of the options. And to be honest, I need to rewrite my !vote so I'll probably do that later. fanfanboy (block_talk) 17:52, 25 October 2025 (UTC)

• Support B, D, E, & I. Vote threshold should be increased to 50 and time limit reduced to 14 days. Sometimes, I'm a bit wishy-washy regarding an admin I barely know, but at the same time not wanting them to get recalled until I get all the facts. "I" for "inactivity" per WhatamIdoing. Oppose A & J at all costs. The process has so many flaws yet it prevents messy and long ArbCom cases. Mox Eden (talk) 09:53, 22 October 2025 (UTC)
• Support A/Open to E to 2 weeks/Oppose All Others I think the process is working relatively well. As someone who was in an extended dispute resolution, the this can be a stress in real life. The tensions is likely even higher here both because the admin obviously didn't initiate the effort and it's really a waiting game as the signatures get added. I'm open to shortening the window to two weeks but, based on how quickly the petitions have been signed, I'm not sure this would make much of a difference in practice. - RevelationDirect (talk) 10:47, 22 October 2025 (UTC)
• Support Change in Petition process. No I do not think the 25 deciders survey is working, see selection bias. To me this present process should be similar to a 'motion to call the question' or a 'discharge petition' with a majority or slight over majority to succeed. We should not silence one side in this process, when we are doing this to one person, or force one side to be bystanders (see bullying). And if we want an esteemed group of users to decide, we should create a board, not this. Alanscottwalker (talk) 11:26, 22 October 2025 (UTC)
• Support A and B - In general the process works well, although making things more of a debate would be healthy. Carrite (talk) 11:40, 22 October 2025 (UTC)
• Support A, otherwise not opposed to B on the condition that by allowing an Endorse option it doesn't negate or change the petitions requirements itself; it is only in an attempt to influence editors not to sign, if there is strong support for the admin being recalled, effectively grouping these supportive comments from the general discussion to it's own section below signatures. So regardless the petition still functions on a signature basis per A and not broken. As for more/less signatures and more/less time (C-F) I see no basis for changing this. At present all recalled admins have not succeeded in passing an RfA, therefore we don't know yet if we need to extend or shorten either requirement. In response to J, admins should do better at finding suitable candidates now we have streamline prcoesses for adminship via the elections. Abolishing recall, when it hasn't functioned incorrectly yet (bar one withdrawn petition) is simply too soon to consider, we'd need a documented pattern of problems to justify that. In reality for every recalled admin there is probably a dozen suitable candidates for election. CNC (talk) 12:54, 22 October 2025 (UTC)
• Support A. The current process seems fine to me. 88.97.192.42 (talk) 13:40, 22 October 2025 (UTC)
• Support D, E, G, and/or H. The process needs to be more deliberative and less subject to responses by a small faction of the community. - Donald Albury 13:57, 22 October 2025 (UTC)
• Support G & H, oppose A & J: Mz7 hits the nail on the head with regard to the balance of power in the process -- the petition is the desysop, and we should move that power to the RRfA.
  Separately, I think exploring a change to the petition structure would be worthwhile; maybe something like having support and oppose in the petition and requiring it to reach a net support number within the seven days (say 25). Many admins are resigning instead of running RRfA, and I think that's a failure of the petition: we should try to make the it a better barometer of whether an RRfA has a chance of passing or not, rather than just a formality. Giraffer (talk) 14:47, 22 October 2025 (UTC)
• Support B outright and D &/or E, and possiblly G I'm not altogether happy with any of the recall petitions we've seen so far; as with rather too many things here, it seems too much like a pile-on with not much incentive for discussion or deep thought. As Giraffer says above, the numbers of admins resigning is a failure of the process ~ Lindsay^Hello 15:25, 22 October 2025 (UTC)
• Support G & H, and in particular I support User:Mz7/How to fix administrator recall § Proposed solution. If this fails, support B by requiring that recall petitions remain open for the full petition period of 30 days (or however long the time period eventually is) and allowing signatures against recall to subtract from the 25 signatures needed to initiate a recall. Best, KevinL (aka L235 · t · c) 15:36, 22 October 2025 (UTC)
• Support D and E, oppose B, G, H I, strong oppose J. I agree with people supporting A that overall recall has so far worked very well in practice, no matter its flaws people are pointing in theory; getting rid of the process altogether would be unthinkable to me. I do think though that 30 days is way too long, 14 days would be much better, and that 25 signatures is not enough relative to the size of enwiki, where 50 would make more sense (I would like be fine with any amount higher than 25 in case of WP:BARTENDER close). Regarding formal supports, I agree with every opposer of B: completely counterproductive and against the very logic of starting with a petition before a WP:RRFA (informal expressions of support in other places to the admin concerned already happen and are obviously fine). Any other complication of the process at petition or RRFA stage to me is currently unwarranted and potentially harmful. Choucas0 🐦‍⬛⋅💬⋅📋 15:47, 22 October 2025 (UTC)
• Support D. E. G. H. I. – Strongly oppose B, that's not how the recall process works. But I agree it's flawed. FaviFake (talk) 16:32, 22 October 2025 (UTC)
• Support A I'm not aware of any problems and the OP doesn't list any. If it works, don't fix it. Andrew🐉(talk) 16:40, 22 October 2025 (UTC)
• Support A, open to D. I don't see a problem here. Writ Keeper ⚇♔ 16:44, 22 October 2025 (UTC)
• Support A as this relatively new process appears to be working ok. Personally I avoid discussions about individuals' conduct but this process appears to be working, judging from the table at Wikipedia:Administrator_recall/Lists#Closed_petitions. Were that to be showing churn around recalls and successful Re-RFAs I might take a different position, but it looks fine for the moment. AllyD (talk) 19:27, 22 October 2025 (UTC)

  You seem to be saying that everything is working fine because administrators are being recalled, without looking to see whether you agree those administrators should have been recalled, nor even attempting to engage with the lack of re-RFAs (and hence lack of successful ones) being one of the problems? This is only half a step up from "I haven't looked for any problems, therefore I can't see any problems, therefore there are no problems", which is obviously fallacious. Thryduulf (talk) 20:56, 22 October 2025 (UTC)

• Support J, also Support D - The purpose of a system is what it does. This process is not a recall system. It is a 25-supports-means-desysop system. If that is what people in this community actually want (and it is what this process does) then it is a different community to what I thought it was. Even if we are WP:NOTAVOTE, it's goes against plain logic to have a system that is always a one-and-done process where no opposition to the motion can be heard.

I fully intend to bring a closure motion at the next recall "petition" (it'e not a petition, it's a vote) that is as palpably pointless as the one brought against Night Gyr. I hope people will support it. FOARP (talk) 21:13, 22 October 2025 (UTC)

EDIT: I'm also OK with the adminship-revocation-request process discussed, and literally anything that would reduce the mob-justice nature of this system. FOARP (talk) 08:22, 23 October 2025 (UTC)

It is a 25-supports-means-desysop system.: Well, it's only a desysop if the person being recalled doesn't take up the RRfA. If they aren't going to pass the RfA even with the reduced threshold, recall has done its job.

I do agree with you that the inactivity-related recalls are unnecessary at best. Children Will Listen (🐄 talk, 🫘 contribs) 00:23, 23 October 2025 (UTC)

Why on earth would they take up the RRfA? —Jéské Couriano v^_^v ^{threads critiques} 05:14, 23 October 2025 (UTC)

Seems like a good example of someone failing RRfA for good reason - the community clearly lost faith in that admin's behavior. Katzrockso (talk) 01:17, 24 October 2025 (UTC)

No-one is taking the RRFA. And why on earth would they even try?

This system was imported from DE wiki without any real thinking about how it would work in a a system where RFA is an “are you spotless and perfect?” test and there’s a massive “no smoke without fire” mentality. FOARP (talk) 05:38, 23 October 2025 (UTC)

Maybe I should use "if you can't stand the heat, get out of the kitchen" instead, which as heartless as that may sound is the unfortunately consequence of being an admin, and more or less how admins responded when recalled so far. Other editors might start a petition against you, heck you might even have to go through an RRfA to retain the bit. I'm not really judging this as good nor bad, but rather the fate of having elevated privileges among a large community. It comes with heightened accountability so yes there will be collateral damage, but no this hasn't occurred yet. CNC (talk) 12:46, 24 October 2025 (UTC)

• Support E (~2 weeks), D if not E, and potentially G per Mz7's proposal or something like it. Petitions to date show that getting 25 signatures within a day is possible in some cases, but it can sometimes take much longer. A shorter window would expedite the process, which I think should be to everyone's benefit, while weeding out the more marginal cases. I favor Mz7's proposal overall, but consensus to implement something new like this has proved extremely elusive in the past. While we're at it, making RfA more tolerable might help address Administrators who could pass RRfA may choose not to try. Anybody else getting deja vu? —Rutebega (talk) 21:47, 22 October 2025 (UTC)
• Support A, This vote is purely from an anon user who mainly reads Wikipedia and looks at the behind-scenes stuff from afar so feel free to revert/disregard it. I support the current system as is and believe it finally gives regular users a means to rectify admins who have gone off the rails without having to go through endless threads on ANI or a torturous Arbcom process. Honestly, looking at the recall cases so far, the only questionable one I would say is Night Gyr and even then, for an admin who so little uses the tools, the aftermath is negligible either way. Cheers. 115.64.183.57 (talk) 00:22, 23 October 2025 (UTC)
• Splitting my !vote into multiple bullet points for clarity:
  • Support A Honestly, if no change were to result to recall, I think I would be fine with how this processes works, although any process can be improved. I don't believe recall is "flawed" in the negative sense, more like "imperfect" as in "nothing is perfect but that's ok" and I can get behind some of the proposed improvements to the process, as expressed in my !votes below.
  • Comments on B I think having editors expressing support should be allowed on a petition, but I strongly oppose having those votes be counted against those signing a petition or for there to be consensus for the petition to succeed. I see the petition as a way to raise an issue with an admin's conduct. If we are dealing with an admin who has popular support from the community, I am concerned that any users raising concerns of misconduct might get drowned out by those who preemptively support the admin without carefully analyzing the evidence that is cause for concern, leading to those concerns being buried. I think the following RRfA should really be were consensus is necessary for any desysop to occur. Now if the signature requirements were dramatically lower, as suggested by Barkeep, then I could somewhat see myself supporting B as getting a net of 5 signatures should be easier than 25. CNC's !vote about option B is more in line with what I expect for this option to take effect.
  • Weak Oppose C, Support D 25 signatures feels too low and can be gathered fairly quickly. I remember having an intention of signing a recent petition, only for it to reach the target before I could have a chance to get on my desktop and log in. I would support raising it no higher than 50 signatures, that way we not doing more than doubling the requirement, although I'd also something in between 25 and 50 like say 40 which would be my preferred target to have. I'd be fine with 30, but I'm not sure if the 5 extra signatures that are required will cause any impact.
  • Support E, Oppose F 30 days is also too long, petitions are ending far sooner than that. I think we should lower it to 14 days, roughly half the current time of the petition phase, and if its still too long we can consider 7 days, but not any less.
  • Open to G, Support H I would be ok with a RfDA process as laid out by Mz7 in place of a RRfA (with the caveat of 50% support instead of 60%), that way we really do need a consensus to desysop instead of the petition actually being a de-facto desysop which I don't think was the intent.
  • Strong Oppose I.a (as proposed by Tamzin) I really don't think limiting signatures to 90 day intervals is a good idea. This is far stricter than the 5 simultaneous signatures requirement, what I.a would mean is that if two petitions are ongoing (which as far as I can tell hasn't happened or isn't common, yet should still be planned for just in case), users would have to choose which of the two petitions to sign, and depending on the conduct involved, that comparison of the two situations might be too complex to decided and may lead to an over correction of discouraging participation of petitions. I could also see a scenario of someone signing a petition for what might be conduct that falls just below our standards, but then another petition within the 90 days starts for conduct that is far below expectations, and that could lead observers wondering "why would this user support a recall for conduct that frankly wasn't that egregious, and yet they wouldn't support a recall for this other unacceptable conduct?"
  • Strong Oppose J Having a community check on admins is important to ensure admins are responsible for their actions, and I don't think the current is failing at all, let alone at the level that warrants scrapping it altogether. Gramix13 (talk) 00:31, 23 October 2025 (UTC)

  Strong Oppose K To topic ban admins from recall RFCs should only be done if there's evidence of disruption, and I don't really think I've seen any strong evidence for this. I wouldn't even classify this RFC as evidence to assume good faith of all participants. This would also be a fairly wide net to cast for such a topic ban. We also have admins participating in the recall petitions themselves, and these can be very important voices to hear about the concerns raise about an admin's conduct to give validity to the complaint. So taking away their ability to give input on a process they participate in, to me, goes against the collaborative spirt we are trying to build here on Wikipedia by having a topic ban rooted in distrust over a wide category of users (the admins) who are all guaranteed to disagree about what's the best way to recall an admin for misconduct. Gramix13 (talk) 22:12, 25 October 2025 (UTC)

• Support J all the way. The extreme deontological attitude that the Wikipedia community has for some reason adopted has already cost us enough good admins who lost the bit for making a few missteps that were greatly outweighed by all of the hard work they've done on balance (such as Kudpung, RHaworth and Athaenara), and that was pre-recall. Suffice to say that it was already far too easy to get desysopped before this absurd policy was pushed through. And for the record I am not licking admins' boots; what I have said also goes for any other editor who has butted heads with the Wikipedia community at times but is mostly a hard worker whose contributions are a boon to the project. Passengerpigeon (talk) 01:03, 23 October 2025 (UTC)
• I strongly agree with Barkeep49's comments in the opening vote, notwithstanding my strong opposition to arbcom being the only route for desysop. (For various reasons, high among them that it takes a very high level of know-how for someone to successfully launch an arbcom case, combined with my belief that the people most likely to be wronged by admins are the ones least likely to have the knowledge of how to challenge them.) Therefore, to my great surprise, I strongly Support G. AN threads against admins being shut down prematurely has bothered me for some time, both before and after I became an admin myself, and I thought that the recall petition would be a useful counter to that problem. However, far from being a backstop against frivolous recalls, petitions have ended up becoming the de facto desysop action in itself. Consensus, or at least strong support, in an AN thread to begin a "request for desysop" seems to me a much better approach than the simple vote-based petition we have presently. However, I still worry about AN threads against admins being shut down prematurely, and so would recommend some limits on closing threads, eg, "if someone has started a subthread proposal with intent to desysop, the thread cannot be closed within 24 hours". Similarly, I strongly Support H, following the general gist of Mz7's "request for desysop". Oppose B as written; if we're doing supports during the petition phase, what we're doing is an RRFA - let's just go straight to RRFA then, and get it over with. Lukewarm on C, D, E, F. I don't really think that tinkering with these particulars will significantly change how this works, but I wouldn't oppose it. However, Support I, in the sense that if we retain the petition system, we ought to address the issues raised by Cryptic in the discussion below. Outright Oppose A, J. -- asilvering (talk) 02:35, 23 October 2025 (UTC)
• Support E, I, H, and maybe D. My preferred solutions are: 1. Recall petitions should last a week, or at most 14 days. The current period is too long. 2. A recall petition must stay open for the full duration regardless of how many signatures it has received, unless the recalled admin requests it be closed early after reaching the threshold. This would encourage greater engagement from the admin being recalled and encourage apologizing and making amends. 3. Previous discussions on the admin's user talk and at a noticeboard (AN or ANI) must be a requirement, and they must both stay open for a certain minimum duration (I'd prefer 7 days and 3 days each). We have a bad habit of closing noticeboard discussions against admins early, forcing people towards recall when that is not the best solution. 4. I think RRfAs should be required to reach consensus to desysop rather than no consensus to remain as sysop, per Mz7. This means a majority in favor of desysoping. 5. I don't think we need to raise the number of signatures required, but we need some kind of activity requirement for petition-signers per Cryptic below. 6. We may want to exclude activity alone as a reason for recall. Perhaps this is better phrased as "require evidence of misconduct" 7. We could look into shortening the immunity periods, also per Mz7, though as they have not been an issue yet I'm not sure if that's necessary. 8. I would also support GLL and co.'s idea below, that the lower RfA thresholds should be valid for a longer period (6 months?) after a recall petition, to encourage recalled admins to reflect, improve, and come back. tldr; per Mz7. Toadspike [Talk] 06:50, 23 October 2025 (UTC)
• J as first preference, followed by D, E, H and I.
  As others state, a recall is at present a de facto desysopping. It is very easy for someone who does admin work in contentious areas to attract 25 detractors. As such, the existence of recall as-is exercises a chilling effect pushing admins away from doing contentious work. We also don't need people bringing into recall their own personal inactivity standards as an end-run around gathering consensus to change our existing policy.
  Accordingly, my preference is to move away from recall completely as a failed experiment, but if we must keep it, then the following changes would be appropriate.
  1. The majority of petitions initiated to date have hit 25 within a day or two. The current structure promotes haste over thought, pile-on over discussion, numbers over consensus. Therefore, the process should change to separate the moving of a petition from gathering of support. I would be in favour of a three-day hold: discussion only from when the petition is commenced, and only after 72 hours can signatures be added. This would allow for rational thought and consideration, as well as the recallee being given a fair chance to be heard.
  2. A recall shouldn't be a first or second step, and proper engagement on talk page and AN should be a prerequisite.
  3. To make sure that there is real community support for the recall, raise the number of required signatures to 50.
  4. For similar reasons, RRFAs should require consensus to demote, not a lower consensus to retain the bit.
  5. The availability of an admin election is a lottery at present depending on when a recall starts. To remove this, the lower re-election % required should be offered to a recalled admin standing at admin election within the 6 months after a recall closes successfully. (The recallee would not retain sysop status during that time.)

  Also fine with the Mz7 proposal and similar. I would prefer change to no change, and would not want the good proposals to die for no consensus over which of them to enact. Stifle (talk) 09:44, 23 October 2025 (UTC)

• Oppose C, F, J. Given that there has never been a petition which failed due to not reaching the signature threshold in time, I don't see any particular reason to lower the signature threshold or increase the time needed. Only one petition lasted for longer than two weeks, six of ten closed within 24 hours and one more closed within 48 hours. I'd be open to seeing the threshold raised or the time limit lowered. I'm not opposed to some other process being developed, but without seeing what alternative process the community could agree on nor am I going to support G. I don't see the point of B: if editors want to encourage an administrator for whom a petition is currently open, they can do that on their talkpage; turning the petition into a !vote is just duplicating the work which would be done in the RRfA. I'd potentially be open to scrapping the two-stage process entirely and just having the RRfA. Caeciliusinhorto-public (talk) 09:48, 23 October 2025 (UTC)
• Support A Process has worked well. Fram has expressed my personal thoughts quite well. At most go a bit shorter in time but change nothing else. It's there to say "hey, enough people think there may be an issue we need to see if the rest of the community thinks you should retain the bit. spryde | talk 13:01, 23 October 2025 (UTC)
• Support A, mostly. I just wish people wouldn't jump on it quite so fast. The recall opens, people sign it quickly, recall closes. I'm sure the signers aren't just signing willy-nilly and have fully considered it, to be clear. But maybe people feel like, if they want to sign, they have to jump in quickly before 24 other people beat them to the punch. I don't know. But perhaps there would be some value in codifying a minimum time. Something like "The recall will be open for 48 hours or until the signature count reaches 25, whichever is longer, and will be closed as unsuccessful if 25 signatures are not reached in 30 days". People know they have seven days to comment on an RFA, so they don't feel rushed. Might be nice to get some of that "rushed" feeling out of WP:RECALL. Probably still a good idea to cap it at 25, though. Useight (talk) 13:16, 23 October 2025 (UTC)

  I suggested a little above that a hold, or discussion period, of 3 days should be in place between the creation of a recall petition and anyone other than the creator being able to sign it. What would you say that that? Stifle (talk) 13:21, 23 October 2025 (UTC)

  Doesn't the AN or ANI thread that begins the recall act as such a hold? Recalls can't be begun without evidence of prior dispute resolution. Children Will Listen (🐄 talk, 🫘 contribs) 13:42, 23 October 2025 (UTC)

  My understanding is that a community noticeboard requirement is not currently a hard requirement, and attempts at prior dispute resolution is a best practice. One of the things Mz7's proposal does is change this from a best practice to a requirement that there be a recent adverse finding against the admin at a community discussion. Tazerdadog (talk) 13:56, 23 October 2025 (UTC)

  Tazerdadog's understanding is correct. Any extended-confirmed editor can start a petition against any administrator for any reason or no reason, with or without prior discussion, as long as it is not within 12 months of a (re)RFA, admin election, them being elected to Arbcom, within 6 months of a previous petition and the user has signed fewer than 5 currently open petitions. The administrator does not even have to have recently edited. Thryduulf (talk) 14:20, 23 October 2025 (UTC)

  Hmm, I'm not sure about a discussion period. Might devolve into three days of people just ragging on the person and now (s)he has to dread all the imminent signatures that will land the second the discussion period ends. Might work if we do it right. Clearly we want a mechanism for de-adminning people, but I want it to be as tasteful as we can. A person has been volunteering their efforts for some amount of time, hopefully doing their best, so it'd be nice if we could make the exit graceful instead of demoralizing. Useight (talk) 22:45, 23 October 2025 (UTC)

• A - and the admins supporting J should probably be candidates for the next petition. nableezy - 14:46, 23 October 2025 (UTC)
• Support D, E, G, and/or H. There has to be a way to retain a community-driven recall while better supporting admins working in areas that are contentious or have policy ambiguity. My preference would something along the lines of Mz7's suggestion (see also Toadspike), where a key change seems to be "RRfAs should be required to reach consensus to desysop". I'd also like to see changes to the petition in this case (increase signature count, reduce time period). NicheSports (talk) 16:01, 23 October 2025 (UTC)
• Support B, D, G Neutral on A but with the caveat that I'm not really sure that we've actually had enough discussions yet to be doing more than guesstimating here... It could in fact be the case that nothing is broken even if improvements can be made. Horse Eye's Back (talk) 16:09, 23 October 2025 (UTC)
• Support A The process is working well and I see no need to change it. The existing process is nothing but a means to garner support for requiring a RRFA or ADE. Iggy pop goes the weasel (talk) 16:30, 23 October 2025 (UTC)
• Support H, and failing that D or J. In my view, 25 signatures is an inappropriately low bar for desysopping someone, and the recall system should not be retained in its current form. Rather, before removing any user rights from an admin it ought to be necessary to establish a regular community consensus for this, and no rights should be removed until it has been shown not only that there are 25 or more users who think that this should happen, but also that this is the consensus view. I would be in favour of replacing the recall petition with a discussion which follows our normal consensus-building norms, which runs as long as it needs to until a consensus can be established, and which involves participation equally from those in favour of the desysopping and those against it. Dionysodorus (talk) 17:17, 23 October 2025 (UTC)
• Neutral - I'm neither running for administrator (I'd never get enough support), nor seeking to recall an administrator. PS - Let me know, how this RFC turns out. GoodDay (talk) 18:56, 23 October 2025 (UTC)
  • Support A broadly, with some tepid support for E. I think you can argue the overall period is lengthy enough to be unnecessary based on the real-world recalls that have occurred. But otherwise, I'm not seeing the point of process tweaks. If people actually have case examples where they think there was a miscarriage of justice beyond the aforementioned, that would be instructive, but I don't think the cases described are examples of recall failures. Der Wohltemperierte Fuchs ^talk 18:58, 23 October 2025 (UTC)
• Support A process is working well. Traumnovelle (talk) 20:41, 23 October 2025 (UTC)
• Support A. I believe the process is working well and as intended. I'm disappointed in the number of admins who are arguing to neuter the process in one way or another. Accountability is a good thing. ~~ Jessintime (talk) 00:16, 24 October 2025 (UTC)
• Support A or E. I have yet to see an argument against the current system that I find convincing. It does well to remember that adminship is a privilege, not a right. If the community opposes someone being an admin, the deciding factor shouldn't be whether they already have the tools. If you are worried about the number of admins, then the solution is to recruit more editors and grow the editor base. That's a far more important conversation to have than recall, but editor-growth discussions aren't nearly as dramatic or exciting and don't get the same turnout. Thebiguglyalien (talk) 🛸 00:52, 24 October 2025 (UTC)
• A. There was no need for this RfC. ~ ToBeFree (talk) 01:04, 24 October 2025 (UTC)
• A. The hypothetical issues with this process are still hypothetical and there exists no good reason to believe they will occur. Katzrockso (talk) 01:33, 24 October 2025 (UTC)
• A: When reviewing the recall results so far, the system appears to be working as intended, and I have seen no indication that it will fail to keep working as intended in the future. The current requirements are reasonable for demonstrating adequate concern to require an RRfA, and the RRfA process is reasonable for demonstrating if the broader community still has confidence in an admin. I don't believe the addition of futher bureaucracy, in the form of prerequesites or otherwise, will benefit the process. Oppose any change. fifteen thousand two hundred twenty four (talk) 02:18, 24 October 2025 (UTC)
• I'd like to propose Option K: topic-ban administrators from initiating or !voting in any RFC to amend the recall process. This is the umpteenth discussion where the large majority of participants who support abolishing recall (or amendments that effectively accomplish the same) are administrators, with the majority of non-admin participants finding the process to be working well. I believe the administrators in this discussion are participating in good faith, but allowing a single user group - who represent less than 1% of active Wikipedia users - to overrule a community process meant to held them accountable is a terrible look. Admins need to recognize that they are WP:INVOLVED participants and recuse themselves from these RFCs and if they will not do that, the community should do it for them. 2601:540:200:1850:F43C:C448:7974:22BD (talk) 04:08, 24 October 2025 (UTC)

  If administrators are <1% of active users, then they'll easily be out!voted by the other 99%. But considering there were multiple admins who signed the most recent recall petition, I don't see why their input is any less valuable. Recall benefits admins too (one bad apple yadda yadda yadda). Nil🥝 04:39, 24 October 2025 (UTC)

  By the nature of their work and the project space pages they frequent, admins are more likely to be aware of and !vote in wiki process discussions than a typical active user, so they generally make up a disproportionate share of !votes in process-related RFCs. This is mostly fine - there is plenty of viewpoint diversity among the admin corps on other issues - but it becomes a problem when admins are disproportionately participating and favoring one side of a debate, across many different discussions. In the WP:BADLOOK RFC I linked above, 30 of the 89 RFC participants were administrators, who overwhelmingly (73%) !voted to make recall more difficult by reducing the petition period to 7-14 days (similar to option E above). 2601:540:200:1850:F43C:C448:7974:22BD (talk) 05:10, 24 October 2025 (UTC)

  To be clear - I do not mean to imply that admins are intentionally coordinating their !votes here or in other recall discussions. However, I do believe that admins, as individuals, are more likely to oppose the recall process and are more likely to support changes that make recall more difficult, which affects the !vote balance in these RFCs which may not represent a broader community view about the proposed changes. 2601:540:200:1850:F43C:C448:7974:22BD (talk) 05:21, 24 October 2025 (UTC)

  I'm an ex-admin. Would I be subject to your restrictions under Option K? —Jéské Couriano v^_^v ^{threads critiques} 04:22, 25 October 2025 (UTC)

  @Jéské Couriano I don't see this restriction as applying to you or to other ex-admins (except, possibly, those who temporarily resigned their permissions at WP:BN but could regain them at any time without going through an RFA.) 2601:540:200:1850:CC47:61C6:19C6:6028 (talk) 21:09, 25 October 2025 (UTC)

  Support Option K in addition to my vote above supporting Option A. I note that Option K allows admins to participate in recall petitions--it just prevents them from starting RFCs to gut or remove the process. AugusteBlanqui (talk) 07:49, 24 October 2025 (UTC)

• Support A With no prejudice to a shorter timescale and more signatures. I have never liked that there is no 'support/oppose', but note the process has not so far failed. Best Alexandermcnabb (talk) 05:49, 24 October 2025 (UTC)
• Support A. I'm not saying that there is no reform that might be worthwhile, but the evidence that there is any problem is very thin. A whopping 1% of admins have been recalled in the first year of the process. More than three times as many have had their admin access removed for inactivity in the same period. In neither case is this evidence of a problem with the process of removal. If recalled admins were routinely winning subsequent RRFAs or admin elections, then that could be an indication of a problem with the recall petitions. But that is not what is happening. --RL0919 (talk) 07:47, 24 October 2025 (UTC)
• Support A. Something tells me that certain people are feeling awfully miffed that despite all the hemming and hawing, the recall process has yet to produce a single clear miscarriage of justice. Now it's down to theory-crafting about the RRFAs that might have been, tsk tsk... Dr. Duh 🩺 (talk) 12:25, 24 October 2025 (UTC)
• Support G/H/I What we seem to have wound up with is a weird mash-up of the dewiki process and enwiki processes, which has lurched along without going off the rails yet (dispite the multiple discussions over WP:INACTIVITY it spawned) but I can't say it really seems to be working "well". I see multiple proposals in the discussion so far that I think worth exploring. User:Mz7/How to fix administrator recall is one: if we're going to have an enwiki-style discussion, let's do that from the start instead of having what's effectively an oppose-section-only pre-RFA. Or we could go the opposite direction by returning the petition to "We the undersigned believe it's worth asking the question whether this admin still has the trust of the community" with no discussion, only signatures, and have the actual discussion if the petition passes. And then probably still change re-RFA (and RFA itself, for that matter) to be less emotionally driven, but that's a long-standing intractable problem. Anomie⚔ 16:22, 24 October 2025 (UTC)
• I also support K and will add my comments tomorrow. NotAGenious (talk) 18:23, 24 October 2025 (UTC)
• E as first choice. I agree with others above that in our highly active community, 30 days is a long time to leave a petition open. My gut feeling is 7 days (the length of an RfA) is an appropriate timeline. 3 of our 10 recall petitions took longer than that to accumulate 25 signatures, though that includes the first two petitions when perhaps there was some hesitation to engage in the new process. I'd also support shortening the window between petition and Re-RfA to 7 days. All that said, long petitions are rare (7 of 10 closed within 48 hours) and re-RfAs rarer still (1 in the books; 1 on the way) so I'm also content if we stick with A and let the process continue to run. Ajpolino (talk) 02:18, 25 October 2025 (UTC)
• Support A, mostly - There has been no evidence that there is any problem, and as such no changes need to be considered at all. Interested in K - not necessarily ready to support this outright at this point, because I don't necessarily agree that admins should be removed from discussions about changing recall just by virtue of being admins. However, there is a big problem with how admins are coming and acting like their personal fear of recall is a problem - it's not. To be quite blunt, if you're scared of getting recalled for taking an action, then it's probably not an action you should be taking unilaterally and without prior discussion in the first place - or at a minimum it's an action you should immediately take to a noticeboard for wider evaluation and discussion. I furthermore highly, highly doubt any administrator is ever going to be recalled for one or even a few controversial/borderline decisions - especially decisions that there was clearly a reason to "shoot first, ask later" - as long as they appropriately seek others' opinions at a minimum after the fact, and abide by the consensus of that discussion. And if they are recalled anyway, they have a high chance of passing a RRFA if they still hold the trust of the community.
  I could see a reduction of the recall timeframe to 15 days (option E) - but on one condition - that there is some manner in which to extend that 15 days if it would be warranted. A couple cases in which I could see it being warranted is if the recall was started early on in a community discussion about an admin or an action they took, and that discussion is still going on at the end of the 15 day period. The community shouldn't be "penalized" in my opinion for starting a recall petition early in a discussion process, which is what would happen if it's shortened significantly, and is why I can't support anything shorter than 15 days. I'm also not sure how the extension would best work - I guess the easiest way would be that any user (whether they are supporting the recall or not) should be able to request the bureaucrats to determine if the recall should be extended and for how long, but I appreciate what others have said about adding more tasks to them that they may not be willing to do.
  Alternatively, what about allowing early closure in lieu of shortening all petitions by default? I am imagining this would be something like "if discussion of the administrator's actions on any noticeboard have died down, and there have been no new supports for the recall in the last 24 hours, the recall should be closed as unsuccessful". The problem I see with this (and why I'm not proposing an actual additional proposal yet) is that "died down" is vague. But that may be beneficial - since many discussions don't ever end up getting formally closed at all, so it would enable whoever gets this ability to early close to determine whether discussions have died down to the point that more time is not necessary. This, in my view, would help - it'd remove the "sword" hanging over the admin for 30 full days (barring withdrawal) even after the community has finished discussing them, while also not shortening from 30 for instances in which people are holding off supporting (or deciding not to support) pending further discussion in those discussions and/or further review of other actions.
  Almost done here, fine with option B - I can't really say I'm massively in favor/support for it, but I wouldn't mind it - so long as it's just a way for editors to show their support for the admin/opposition of the recall and it does not increase/change the number of votes to recall that are required to force the RRFA/election. Specifically allowing support for the admin in the recall vote phase would basically remove all arguments of "but the admins just resign cause they don't know how many people support them so they don't even bother RRFA". If an admin is recalled, gets the 25 votes, but also has 50 editors who've shown up expressing that they'd support an RRFA if it becomes necessary, then there's no valid argument that the admin didn't want to bother with it because they didn't know they had support. If the admin in this case still chooses to not run for RRFA, that's their right.
  Lastly, fine with option H so long as the changes made are to make it less of a "gauntlet" for the admin. Two changes I could see myself supporting are potentially a slight reduction in the minimum standard to keep the bit at RRFA/election, and allowing the administrator to postpone the RRFA/election for a decent length of time (or indefinitely) - in other words allow them to delay it but keep the lower standards for keeping the bit when they do eventually run - but they will lose the bit in the meantime. This would, imo, alleviate the concerns that admins feel "rushed" to decide and to form their statements/etc to run RRFA in the allowed timeframe. It would also further mitigate the issue of "admins form enemies by virtue of their duties", i.e. a slight reduction (ex: to 50% being sufficient and 40-50% being crat chat) would mean that people who have almost half of the !voters' support still have a chance to retain the bit if the crats feel that, after discounting vendetta-type !votes, they should retain the bits.
  Hopefully this all makes sense and gives some food for thought. I'll try to respond if anyone has questions for me but I'm very, busy IRL right now. If anyone feels that this would be better formatted/understood as bullet points or something please feel free to change this from paragraphs without checking with me first. Regards, -bɜ:ʳkənhɪmez | me | talk to me! 05:06, 25 October 2025 (UTC)
• Support B & E To an extent, it feels like admin recall has become a bit of a space where wedge issues with one person can remove an otherwise-acceptable admin. Having a space for support & shortening the time can remove some of the bias against admins. JuxtaposedJacob (talk) | :) | he/him | 05:08, 25 October 2025 (UTC)
• Support A and K. Surely if there was even the remotest truth in the claim RECALL has been used frivolously, maliciously, or has some structural flaw, we would have already seen one if not more of the successfully recalled Admins go immediately to reconfirmation and pass easily. It surely has to be because, in all cases, issues were identified that can only be rectified with a sustained period of editing as a mere editor. And the people recognizing that fact most clearly, are surely the recalled Administrators themselves. It certainly doesn't remotely hold water that they would just be too busy, too scared or too jaded to do such a thing. They can certainly claim they have too many enemies, but surely if they're not willing to even stand to see if their enemies can pull the wool over everone else's eyes in the forensic environment of RfA, we're entitled to view that as merely an excuse. And if they're refusing to do it just to teach Wikipedia a lesson in what they've lost, well, good riddance to bad apples, and no reason to think RECALL isnt working as deaigned there either. Gordon Maximo (talk) 10:30, 25 October 2025 (UTC)

  Struck !vote by blocked sockpuppet per WP:SOCKSTRIKE. QuicoleJR (talk) 20:58, 26 October 2025 (UTC)

• Support G per Mz7. In particular, I agree with his observations that desysop happens at the point of cetification; that a petition can be signed without argumentation; and that RfAs are prepared for, but RRfAs cannot be. Arcticocean ■ 13:03, 25 October 2025 (UTC)
• Support A because it seems to me like it's working as intended. If A fails, a distant second would be Mz7's proposal, although I'd rather have it at 14 days instead of 7. Also, weak oppose K since I don't think it's fair to remove the opinions of admins from the equation just because they are admins, and it IMO is against the spirit of Wikipedia. If there is any actual evidence that admin participation in these RFCs is causing problems, I'd be willing to reconsider. Also, I don't know if an option added so late in the RFC can get consensus here, and it would probably need a different RFC. QuicoleJR (talk) 16:55, 25 October 2025 (UTC)

  I believe that admins participating in these RFCs is causing problems because, unlike a regular RFC where policy-based !votes are weighted, an RFC to change Wikipedia policy is much closer to a straight-up headcount.

  By my count, there are 90 editors who have !voted in this RFC so far (33 admins and 56 non-admins). Option A is supported as a first choice by 34/56 non-admins (61%), either alone or in combination with other options. If this RFC were closed based on non-admin input it would be a consensus for A. However, administrators have overwhelmingly voted against Option A, with only 6/33 admins (18%) supporting. They account for the majority of !votes to weaken or abolish recall despite admins representing less than 1% of active Wikipedia editors. Their participation is probably enough to shift the close from a consensus for the status quo into at least 'no consensus' territory.

  I understand why people are saying that admins should have some voice in this discussion, but they should not be able to dictate the outcome by sheer force of numbers which is what is happening right now. The structure of the RFC process stacks the deck in favor of admins who want to amend or abolish recall and I suspect this is the main reason it took 15+ years for the community to adopt such a process in the first place. 2601:540:200:1850:CC47:61C6:19C6:6028 (talk) 22:23, 25 October 2025 (UTC)

• I support E (shortening the recall period to any period in the interval [7 days, 15 days], inclusive) and D (raising the number of signatures required; I am agnostic on the precise details and would like to be counted as support for whatever number here might gain consensus); I oppose A, C, F, and J. Independently, I support Mz7's proposal (whatever letter than stands for). --JBL (talk) 17:34, 25 October 2025 (UTC)

  Also oppose K. --JBL (talk) 18:07, 30 October 2025 (UTC)

• Support G My feelings are similar to the ones expressed by Barkeep and Mz7, I agree with Mz7's observations. I think RECALL should serve as a small barrier from frivolousness before a full community decision should be reached and not an already condemning vote. Additionally I think that every admin how has done a tough decision such as a large RFC close or the likings will have 25 people unhappy with their judgement and still admins can and should be making tough decisions and should also be able to feel comfortable doing so. I also support Mz7's proposal to change to an RfDA, if the outcome is "no consensus" an admin has not lost the trust he earned in my opinion. Squawk7700 (talk) 00:07, 26 October 2025 (UTC)

  Amendment: A loose idea just formed in my mind and I'm not sure at all that it could work but I want to share it anyways. As stated above in my opinion the RECALL should be the first step just making sure no frivolous stuff happens that forces an admin into a long painful but totally unnecessary process. Also as stated above, I think one will find 25 opposers for every active admin. So the idea was something along the lines of, someone could propose/nominate an admin for RfDA/RRFA and it automatically happens unless X people oppose the nomination. This would allow for Supports without having to have a kind of pre-rrfa. But of course I this could run into the very same issue; one could find X supporters for any admin which would be a lot worse and I have not yet found either the right number of voters or another way to prevent this. Squawk7700 (talk) 00:15, 26 October 2025 (UTC)

  I appreciate the attempt at brainstorming ideas on alternatives to our current petition system, but I don't believe this one will work out for the concerns you mentioned. I'd prefer always having 25 supports bringing an admin to RRfA where we as the community can judge the complaints being made than to have 25 opposes blocking any attempts at recalling an admin for misconduct. I'd much prefer having a direct RRfA over an opposition-based petition (i.e. what you are suggesting here), but that runs the risk of frivolous recalls, hence the need for support-based petitions (i.e. our current if imperfect system). Gramix13 (talk) 00:34, 26 October 2025 (UTC)

  Yea, I have to agree... Perhaps one day I'll have an enlightenment on how to make it work but I think I was way to early putting it here. Should I WP:MUTUAL? Squawk7700 (talk) 00:42, 26 October 2025 (UTC)

  I don't think that's necessary at all, I don't see any ill will from your proposal and I didn't intend to comment on it as entirely negative. Ideas that don't work in brainstorming may lead to ones that do, so keeping this thread here might be helpful for anyone looking back at this RFC in the future. WP:MUTUAL, in my reading, is for unproductive comments like incivility, and I think there's something productive that can still be gleamed from this even if it doesn't work out (yet).

  And in case my tone wasn't clear, I don't hold any grudge towards you, the opposite is more true, I think coming up with ideas, even flawed ones, can be helpful! If you find a novel way to make this all work, that could be useful to improving recall. I hope my comment didn't come off as harsh towards you and my apologies if it did. Gramix13 (talk) 00:58, 26 October 2025 (UTC)

  I was more thinking about preventing an unneeded debate but works for me too, Kind Regards Squawk7700 (talk) 01:27, 26 October 2025 (UTC)

• Support C D Generally I think the system is working well and I support it existing. However, I do think it can be improved and creating a higher number for recall would be a good idea so it requires more support from the community to recall and prevents it being a smallish number who decide on if to recall or not. I do also agree with  Tewdar  that alerts should be sent out to the communtiy for recall petitions like RFA do. I also would support removing the 5 sig limit(which I believe Tewdar also proposed.) All editors in good standing should be able to sign a recall petition not have a arbitrary limit.GothicGolem29 (talk) 16:33, 26 October 2025 (UTC)

  @GothicGolem29: Did you mean D, not C? 173.79.19.248 (talk) 21:17, 26 October 2025 (UTC)

  I did thanks I will correct it GothicGolem29 (talk) 23:35, 26 October 2025 (UTC)

• Support A We need a recall process, this one's working despite various misconceptions about it, and there's no case for making changes that would have selectively prevented any particular petitions so far from passing while letting others through. Strong oppose B: we already have a discussion section; turning petitions into votes would elevate them into RRFAs/AELECTS, meaning that any single ECR editor could by initiating a petition launch an immediate full-scale vote on the admin. (It may be instructive to note that the recall process for the UK legislature is similar to our current process, in that an election may be triggered by a petition, but petitioning is not a process of voting for and against.) Strong oppose C and F: so far there's no need to make it easier to recall. Mild aquiescence for D or E, depending on the number tbd, but not both, but as petitions generally reach 25 in far less than 30 days there's no indication that any difference would be made by anything but an extreme amendment, or that such a change would be beneficial. Oppose G: petitioning is an appropriate process for initiating recalls. Open to hearing specific proposals for H or I, though if we have a vexatious run of fruitless proposals, we might need to call a halt to them. Strong oppose J: the community should be able to recall admins as well as appoint them and it shouldn't be seen as a lifetime appointment. Strong oppose K: admins are members of the community, many of whom have already participated in recalls in just as diverse away as Wikipedians in general, and we should value their perceptions as much as every other editor's, not disenfranchise them on appointment. A pox on RFCs with so many disparate options as this one. NebY (talk) 19:21, 26 October 2025 (UTC)
• A. The limited set of certified petitions show that the process is effective at filtering for RRFAs where there is a probability of desysoping. The process isn't unfair to admins, the community's just been very disciplined in starting them (save for one that was withdrawn). Mach61 00:46, 27 October 2025 (UTC)
• Support A not a single improper recall so far. Mztourist (talk) 15:03, 27 October 2025 (UTC)
• H/I but A is acceptable. I would prefer some minor changes such as admins being explicitly allowed to waive their recall immunity at RRFAs if they think it will help their chances (which I brought up at the Graham87 RRFA and the abortive recall reworkshop [2]). I also think that the initial signature on a petition should not solely be based on activity unless it meets a very high test for gaming (e.g. useless userspace edits/admin actions or borderline useless edits) as I articulated at the previous thread at the Village Pump Idea Lab [3] based on the large amounts of ink spilled at Night Gyr recall until actual tool misuse was brought up. Otherwise, I think recall is fine as it is and don't think any of the successful cases were without merit. ---- Patar knight - ^chat/_{contributions} 17:38, 27 October 2025 (UTC)
• A: working as intended, better than expected. No bad petitions so far. None of the doom-and-gloom predictions have come to pass. Zero revenge petitions. No groups of socks block-signing petitions. No admin recalled for the mythical "correct but unpopular decision." No gang of disgruntled editors going about voting to recall everyone (it's been a very diverse pool of signatories; most signed 1/10 or 2/10 petitions, the highest is 4 petitions; no one signed half or a majority of petitions one person signed 6, one signed 5, everyone else signed less than half). No rush to judgment over one mistake, or a few mistakes. No frivolous or thoughtless petitions. No evidence-less petitions. The worst part of the recall process has been the behavior of certain editors (including, sadly, admins) who've engaged in bludgeoning and incivility, but even that's been tolerable. For each petition, the outcome would have been the same if it had been 20 signatures or 25 or 30. Or if it had been open 15 days or 30 or 60. Most of these are wrapped up in one day or a few days. The lack of RRFAs shows how well the system is working, and also how conservatively the community has been deploying this tool: so far, these have all been extremely clear cases, years in the making. None of the proposed tweaks will make a difference, and it's time to put recall RFCs on ice for a good five years; we've had enough. Levivich (talk) 15:55, 29 October 2025 (UTC)
  • The highest is six petitions (plus a redundant 26th signature on a seventh). —Cryptic 16:30, 29 October 2025 (UTC)
    • You're right, thanks. Fixed. Levivich (talk) 17:15, 29 October 2025 (UTC)

  Also support K, after reading others' comments about it. It's a conflict of interest, plain and simple: people with a user permission have a COI when it comes to deciding the rules for removal of that permission. Levivich (talk) 16:35, 30 November 2025 (UTC)

• Support A, Oppose B I think there has been nothing presented that a change needs to happen. So until that time I do not think there should be a change. In regard to option B, that goes against what was decided very deliberately during the lead up to the recall process. It is a petition not a vote, people that want to have an oppose vote have the RRFA. PackMecEng (talk) 22:29, 29 October 2025 (UTC)

  Looking over the discussion I would also like to Suppprt K as pretty obvious and sensible. There is a clear and obvious conflict of interest there. Even assuming good faith we should treat them like COI editors. PackMecEng (talk) 18:36, 1 December 2025 (UTC)

• Support A, open to D, K, & potentially B, strongly oppose J - The process so far has worked as intended & I've yet to see objections provide convincing evidence to the contrary. Every editor here has rules they have to follow, privileges earned through trust, & consequences when those rules are broken & that trust lost. Admins carry a lot of power, responsibility, & community trust that can easily be abused, so the potential consequences should match. If you can't pass the RRfA (which has a lower threshold then regular RfAs) then the community has shown they've lost trust in you to perform admin duties. Adminship isn't for everyone, but you can still constructively contribute as an editor. I'm undecided, but remain open to K as the inherent conflict of interest presents an issue. Obviously an admin has an incentive to argue against their checks & balances, so I'm not against some form of restrictions to remedy this. - Butterscotch Beluga (talk) 23:14, 29 October 2025 (UTC)
• Support A and D, oppose B. The recall process is Mostly Fine. I can see subtly upping the number of signatures required due to avoiding trigger-happy recall petitions that are "referendums on one unpopular action" that maybe could be staved off, but nothing too dramatic - 35 signatures maybe? Very opposed to B, which seems to be turning the referendum into a two-part RfA. RfAs are already stressful enough, don't threaten people with the possibility of going through it twice. SnowFire (talk) 03:38, 31 October 2025 (UTC)
• Support D, Oppose B: I was going to cite other people but SnowFire has pretty much summed it up, though I think we might need much more than 35 signatures. At the same time Mz7's proposal seems really good (Support G?) and I'd really like to see it trialed. Aaron Liu (talk) 16:42, 4 November 2025 (UTC)

  I agree with S Marshall that so far the petitions seem to be reaching the correct outcomes, but the processes of reaching these outcomes seem to be flawed, and the way we've seen them play out exposes possible problematic outcomes as detailed in Mz7's essay. Aaron Liu (talk) 16:46, 4 November 2025 (UTC)

• Support D and/or E, at least. Insofar as some editors have claimed that it's "worked as intended", there have been some recalls like Wikipedia:Administrator recall/Master Jay where the main complaint was essentially "the admin isn't using the tools enough" (something that is not against the rules at this point), and allegations of policy and guideline violations weren't lodged until later (if at all). Increasing the signature count and/or reducing the timespan for recalls would reduce the incentive for petitions motivated by factors other than wrongdoing. If an admin has actually engaged in recent egregious violations of guidelines and policies, the petition can accumulate the required number of signatures in a shorter time period, anyway. – Epicgenius (talk) 19:56, 4 November 2025 (UTC)

  The easiest way to reduce the incentive for petitions motivated by factors other than wrongdoing would be to just directly implement a requirement that complaints must start with a non-activity related complaint (with perhaps a well-defined carve out for the most egregious of gaming inactivity such as purely doing edits/admin actions in one's own userspace). -- Patar knight - ^chat/_{contributions} 21:43, 12 November 2025 (UTC)

  That seems fair, and would go a long way toward alleviating my concerns in that regard. Alternately, if we do want to make admins eligible for recall merely for inactivity, it may be better to just increase activity requirements to prevent that from happening. – Epicgenius (talk) 02:29, 29 November 2025 (UTC)

• Support D. I'm of the opinion that the current process is not sufficiently robust against hardcore sockpuppet abuse. MER-C 17:02, 11 November 2025 (UTC)
• Strong oppose B as this would functionally make it redundant as a consensus discussion, support H per User:Mz7/How to fix administrator recall, and (ironically) strong support K in order to keep recall as a check the community has on admins and avoid a "closing ranks" effect. Neutral on changing the duration or petition threshold one way or the other. Open to G depending on the alternative mechanism being proposed. Chaotic Enby (talk · contribs) 05:52, 13 November 2025 (UTC)
• H The process has two major problems: it is one-sided (only looks at those in favor of recall) and the period between recall and RRfA is too long and creates problems, particularly when the admin continues the behavior that led to recall. I’d rather see a quicker petition process that required fewer votes, had some protections against abuse (e.g., person could only sign one petition per year) and that immediately segued into a Request for Deadminship that would have set word limits and time for pro/anti arguments , then the deadminship would pass if >50% voted in favor. So most of the process would be focused on a pro/anti discussion rather than signature collection. — rsjaffe 🗣️ 06:17, 20 November 2025 (UTC)

  the period between recall and RRfA is too long The 30 days is maximum time possible between a petition and RRfA. We've so far only had one other petition go to RRfA, and that was 11 days after it was closed. This most recent petition is so far an exception and is currently the only where the time between a petition closing and RRfA starting (in this case it's AELECT) is longer than 30 days. fanfanboy (block_talk) 13:29, 20 November 2025 (UTC)

  But others have not desysopped until the 30 days were up. If a person should be desysopped, that's a problem. — rsjaffe 🗣️ 02:31, 29 November 2025 (UTC)

• Support J with conditions. It does not make sense to me that ArbCom can use its existing processes (e.g. motions, cases) to desysop, instead of having a separate page ("Wikipedia:Arbitration/Desysops") just for desysops, while we (the community) have to use a separate page to desysop instead of using existing processes we use for other conduct disputes (like WP:ANI). It makes intuitive sense to me for there to be no dedicated recall process, but instead to give ANI/AN the power to desysop. Any desysop requests on ANI/AN would be procedurally treated as a proposal to topic ban the admin from performing activities only available to administrators (excluding global rights actions in line with the global rights policy), enforced via removing the admin bit, and rescinded immediately and automatically upon successful re-granting of the bit (through RfA, AELECT or any future adminship process), but with higher participation requirements, keeping the "extended-confirmed editors only" restriction at the current recall process, a mandatory notification requirement (even if the recall discussion started from another discussion the admin in question was already notified about), and not logging the "ban" at WP:AEDR as having no point; this would also render RRfAs as redundant. (I also support this other proposal as an alternative.) Failing that, support E to 15 days; every successful recall other than Night Gyr finished in less time than this (often in only 1 day), and we don't need vexatious petitions to be hung over the admin's head for 30 days. OutsideNormality (talk) 05:24, 22 November 2025 (UTC)

A lengthy rationale...

• B, D, E; maybe I, maybe J; it did seem, to me at least, that 25 is a bit on the lower end (considering that there might be about 200 users in an RfA opining... upper end, I think 300, although in my time I think that was around 100), and it might be explained (not sure, though) by how it was adapted from dewiki (uncertain if the number was, too), and considering there's fewer people, on average, there... and, uh, not to unnecessarily highlight a subjective anecdote which is, of course, just that, but sometimes I happen to block up to 25 disruptive users daily, even, and I certainly do try to be as fair as possible... with, I think, more warnings than most usually send them, and I'm usually responsive etc.... and, I know this is supposed to be a possibly rational process, with evidence presented, and such (even if someone below mentioned that might not be necessary), I can't help but wonder if... well, any number of those blocked might seek to use the process (comments below indicate that a few did already), and it's (just for comparison reasons, with the 25 necessary in the process) about a thousand now, maybe more (haven't checked), in the last half a year, or such, so that's quite a lot of possible 25s (and some, if one can be explicit, can be quite vicious, with wishes of death etc... so, trying to use this process wouldn't, quite, be the worst thing they'd try to do, if they ever, of course, did try anything, and hopefully it's just the usual online commentary... I am just trying to highlight the reality, though, beyond the 'dry' discussions of various technicalities, here, as actually doing admin stuff goes beyond some politics, honestly, and it's a bit intimidating with what kind of responses some blocked disruptors have, albeit a few). I am, of course, not at all saying that there shouldn't be accountability, far from it; I try to be as lenient with disruptors, myself, in the possibility that they might revert to productive editing (if there was ever any), so accountability is something I appreciate, but...
• It could be 'useful', in a sense (even if seen from a utilitarian vantage point, i.e. the growth of the site, regardless of any individuals), to be... well, not sure if this is strange to say, but empathic? In the sense that every 'user' is, indeed, an individual, and even when dealing with vandals I usually approach it from a "is this necessary" kind of standpoint, and unless they're acting in clear bad faith, and don't seem likely to not disrupt going forward, I usually don't indef (and, even then, disabling the talk page is a last resort, or based on prior accounts, similarly), as they could, possibly, return to positive editing... analogically, perhaps, it could be useful/helpful (just a suggestion, of course), if people did try to approach it (this process) from a good/bad faith perspective, and possible likelihood of future change, because wouldn't that be beneficial for the site overall, too? So, this is the sense that... people make mistakes, everyone does, so why not try to be a bit more tolerant of them? I certainly can never claim to never will, in my monitoring of AIV, especially when it's at all times, not just when 'perfectly' refreshed (and, of course, no one is perfect), and... I'm not sure how aware people are of the traffic at that page, but considering that this is the biggest wiki, sometimes there's quite a flow (other times, as today, not so much; but, even then, it's only relative, i.e. e.g. I handled 10, and with the amount others handled, in comparison to smaller wikis where... I suppose one person, if as much online as possible, can manage it on their own, as long as there's others at different times)... uh, what I'm trying to say is; there is a mixture, here, occasionally there's multiple people monitoring for disruption, but other times, also occasionally, I wake up to some mass page-mover that has been going on for, maybe, half an hour... and, I know these are wikis, and everything's reversible, I suppose, but it's quite arduous to revert 100 page moves... and, I realize not everyone monitors this, or that page, but the more admins there are... even if they handle just the occasional report, I'm not sure what the problem is? (Sure, I admit I 'have no life', which is why I end up monitoring it any time I'm awake, but as long as technology doesn't exist that dispenses with the annoying need for sleep... I'm just saying, but also being facetious, as I realize I'm meandering, which I hope not many mind (I don't often go on at length like this, though, so if I can a couple of times)... point is, statistically the more admins there are the less mass page-moving vandals will go on at length, and some may find possibly vulnerable templates that transclude to many other pages, so tackling them earlier is crucial...)
• This is why, in my opinion, B could be useful, to highlight not just the mistakes (or otherwise), and try to get a whole picture, and E as, as it is, it's quite lengthier than the usual processes to get there, and making them parallel would make more sense, but one last thing I'd like to highlight (why I might go for J) is the sense of arbitrariness of it, as I previously commented (so I won't go into length about this, as I did already, before)... with regards to that kind of case, at least (other recalls seemed to have specific, individualized reasons for them, and, by all means, bring up incorrect blocks etc. I, myself, wouldn't want to block anyone by mistake (or misjudgment, perhaps), but it would be better if second chances are given)... with regards to inactivity (besides, as mentioned previously, from my point of view even the occasional action could be positive, as some others have lives out there (wonder what that's like), although that's just my opinion, and suggestion to consider)... it does end up seeming arbitrary if there's some amount of actions, or edits, beyond the pre-established consensus, as, if not, then is it not Kafkaesque, possibly? In the sense that, hypothetically (and this might be a strawman argument, I admit, but the logic might still apply), someone (as it's down to a subjective opinion) might still decide that even, say, ClueBot-like activity isn't enough, and then... what's the point of all this, honestly, on a volunteer encyclopedia, especially? Isn't the objective to expand it, which is done by having more volunteers, regardless if they're the most active editor, or the least? (And even the most can't not sleep.) Just... random thoughts, suggestions; feel free to ignore them, and I hope I haven't annoyed anyone by writing all these, at length (at least, it seems to be at around the end of this discussion, so the text isn't sitting in the middle, all stark; I'll try to hide it in a template, just in case, but last time it didn't function so well...) ~Lofty abyss 05:10, 25 November 2025 (UTC)

• Oppose Options C, F, and K. K, because it's essentially a pre-emptive topic ban on hundreds of users and a remarkable assumption of bad faith (the implication being that administrator !votes on suggested amendments can never be trusted to 'do right' by the process). Administrators have just as diverse a collection of views on the process as any other user group – as shown by their active participation in recall discussions, where there have been multiple administrator voices in favour of recalling some petition subjects. Yet apparently the same administrators who have !voted to recall – thus implicitly backing the process – wouldn't be welcome to !vote on proposed changes to the process...why? How do we know their input wouldn't lead to a beneficial outcome? Why assume an entire user group thinks along the same lines? --- C and F, because the petitions to date show that the current thresholds are sufficient. --- Indifferent/Don't Know on Options A and J. The process is too far along to be abolished, and while I have reservations about the process as it stands, I'm not opposed to it just carrying on as is. --- Broadly support all other options, especially Option B. What I'd prefer is the 'petition stage' be more of a deliberation, where !votes opposing making the subject go for RRfA can be weighed against those in favour. SuperMarioMan (talk) 02:05, 1 December 2025 (UTC)
• Support A but not necessarily opposed to D. A lot of editors (especially administrators) seem to be missing the point of how this works. An RRfA would be an onerous process, so the petitions is to make sure that we don't start one unless there is a good chance of it failing. The fact that no administrators have had a successful petition and then successfully run for RRfA is a feature, not a bug, it means the 25 signature threshhold is working at weeding out RRfAs that would've been an unnecessary burdern for an admin that would retain the bit anyway. Any changes such as allowing supports and opposes during the petition just turns it into a pre-RfA RfA, which defeats the purpose. --Ahecht (TALK
  PAGE) 18:51, 1 December 2025 (UTC)

  it means the 25 signature threshhold is working at weeding out RRfAs that would've been an unnecessary burdern for an admin that would retain the bit anyway We have absolutely no evidence to support this given that there have been exactly zero petitions that were open for 30 days and failed to reach 25 signatures, and only one admin who has stood for ReRFA. Thryduulf (talk) 22:06, 1 December 2025 (UTC)

• Support D. The amount of signatures should be at least 50. If an administrator is notable for being problematic, at least 50 editors will recognize this. Steel1943 (talk) 07:07, 2 December 2025 (UTC)

Recall check-in discussion

• To me the biggest issue so far seems to be that a relatively small number of people with particularly strict standards can make up a sizable portion of the 25. (And I say that as someone with standards that are stricter than the median.) I'm not sure if that's yet caused any undesirable recalls—it may have hastened them, but most recalls have been so fast that I think they would have gone through anyways—but it concerns me as something that could lead to problems over time. I wonder if a good Option J here would be something like "A user may not sign a recall petition (except as initiator) if they have signed one in the preceding 90 days". -- Tamzin^{[cetacean needed]} (they|xe|🤷) 14:42, 21 October 2025 (UTC)

  I have added an I "some other change to the petition" and moved "abolish" to J (since no one had used either of them). Best, Barkeep49 (talk) 14:46, 21 October 2025 (UTC)

  I'd endorse replacing the "sign no more than 5 concurrent petitions" clause with Tamzin's. That clause is a vestige from when we thought recall would look very different. Only concern is possible bookkeeping hell. Tazerdadog (talk) 14:57, 21 October 2025 (UTC)

  I don't love this proposal. It's perfectly plausible that two well-founded recall petitions could take place in the same three month period and it could well make sense to countersign both. I wouldn't want to have to engage in game theory about timings before presenting a petition.—S Marshall T/C 14:13, 22 October 2025 (UTC)

• Some recent discussions that suggest some dissatisfaction with how WP:RECALL has interacted with WP:INACTIVITY: Wikipedia:Village pump (policy)/Archive 203#Admin inactivity rules workshopping, Wikipedia:Village pump (idea lab)/Archive 69#Revisiting WP:INACTIVITY, Wikipedia:Village pump (policy)/Archive 204#Fixing the admin inactivity requirements. Anomie⚔ 14:45, 21 October 2025 (UTC)

  The tension these seemed to come down to, in my biased opinion as a participant, is that in principle it would probably make sense to exclude a purely inactivity-based rationale as grounds for recall, but the two inactivity-related recalls so far have involved gaming the system, and it would be a much slipperier slope to exclude gaming. And while I do think there was gaming in both of those cases, if we just banned purely inactivity-based rationales, surely anyone looking to recall for inactivity would just cry "gaming" even if there wasn't any. -- Tamzin^{[cetacean needed]} (they|xe|🤷) 14:59, 21 October 2025 (UTC)

  As a mostly non-participant, my personal (and probably also biased) opinion is that a lot of those discussions may have come down to some people thinking that WP:INACTIVITY should cover everything, rather that being a bright-line rule like WP:3RR is to WP:EDITWAR. Another major thread I noted was questioning the extent to which a minimally active admin (even a gaming one) is actually a problem, with the "not a problem" viewpoint holding that they're around enough to (hopefully) avoid "inactive account compromise" and they're not running into "taking action that is out of touch with current practices" because they're hardly taking any actions at all, while the "is a problem" viewpoint seemed to hold that "not using means no need" and "they might decide to be more active without getting back in touch with current practices" (and also some "anyone who became an admin 15+ years ago is suspect because they didn't pass under the current higher standards", sigh).

  Regardless, the discussions do reflect some concern that WP:RECALL as it exists is open to a small group of people forcing their viewpoint, in this case that people who want a high threshold for admin activity can force an RRFA, and if the low-activity admin decides to bother running the gauntlet at all the support may be too lukewarm for it to pass. Anomie⚔ 15:37, 21 October 2025 (UTC)

  Regarding INACTIVITY, I currently think the policy does not fully meet community expectations, and improving that interaction would remove some of the heat-and-sound around these RECALLS. It was my intent to run some scripts to collect data. And then propose superceding the current policy with just "Inactivity thresholds require 1 admin action within the last 12 months" +appropriate checks and balances.

  Unfortunately, I ran out of personal steam and IRL energy to focus on that, so didn't get back to that discussion. CC @Firefangledfeathers who had pinged me a few weeks earlier as well. Soni (talk) 01:11, 22 October 2025 (UTC)

• I find myself unclear as to what There should be some way of enabling supports during the petition phase means. Is "supports" meaning "support for the admin", which would effectively be "opposition to the recall"? That seems to be how voters so far have interpreted it. Anomie⚔ 14:58, 21 October 2025 (UTC)

  I went ahead and reworded (supports --> support for the admin). Please revert if that wasn't the understanding of any participant so far. Tazerdadog (talk) 15:12, 21 October 2025 (UTC)

  @Tazerdadog, "support for the admin" to me sounds like "help for the admin", not "votes in opposition to the recall". I may be in favour of help for the admin, but whether I want a pre-election is a different question. —Kusma (talk) 16:28, 21 October 2025 (UTC)

  I was trying to make the statement as unambiguous as I could with the smallest tweak I could. There's definitely a pretty fundamental issue inside of option B on how you allow both sides to express opinions without getting a Pre-RfA RfA. I think we probably have too many !votes in now to change the wording unless there's a serious issue - if you see that serious issue though, I'd encourage you to fix it sooner rather than later. Tazerdadog (talk) 16:39, 21 October 2025 (UTC)

• Why no option to increase the amount of time an admin has to re-RfA after a recall? Looking at the one petition that actually ended in a re-RFA, and I think more space between "really stupid admin action" and "RfA" would have been a positive - Graham could have pointed to improvements in the way he handled issues, and the community would have had time to sit back and judge whether or not that was sufficient. For the inactivity ones, it gives the inactive admin time to either raise their activity or, if they weren't active because they were sick or had irl responsibilities, address that. Have an automatic desysop after 30 days, two weeks, whatever - but is there any reason why the RfA with a reduced pass threshold can't happen 90 to 180 days later? (Arbcom would still be around for emergencies)
  Main worry with increased signatures is that, assuming for every signatory there's another person who won't sign, but believes the person shouldn't be an admin & would !vote so on an RfA (based on several private conversations, I believe that's about fair), then we'd start off any re-RfA with 80 to 100 oppose !votes (using Tamzin's 40 to 50 signatures requirement) and that's not.... that's not going to go well. GreenLipstickLesbian💌🦋 15:11, 21 October 2025 (UTC)

  As far as I know, the short RRFA period is so admins can show "the recall was wrong" as in "it didn't represent the actual consensus of the community". If you need to improve, change, ... and thus need a longer period, then there is no reason why your success percentage should be lower than e.g. for someone who failed RFA1, improved, and has a go at RfA2. Fram (talk) 15:14, 21 October 2025 (UTC)

  Fully agree with the first paragraph of this. If you want to use the election in 4 months, fine. If you want to demonstrate personal growth between an ugly petition and asking for the bit back, great! That's what I'd expect a good admin following best practices to do, it also removes the awkwardness where an election isn't a real option for an admin 4/5ths of the time. Tazerdadog (talk) 15:20, 21 October 2025 (UTC)

  +1, This seems to be one of the most impactful changes to RECALL imo. Having an RRFA happen a longer time period after the recall petition, as well as allowing WP:AELECT an option for all such admins, would significantly decrease the temperature of the process, but still allow admins clear avenues to return.

  GreenLipstickLesbian What would you think of something like "In 30 days, adminship is removed, but every RRFA within 3? 6? months of the petition would have reduced thresholds" Soni (talk) 01:03, 22 October 2025 (UTC)

  Yes, that's about the wording I was thinking! GreenLipstickLesbian💌🦋 02:00, 22 October 2025 (UTC)

  Interesting. Yeah one of the reasons I thought the RRFA option was worth a try was the 60% threshold. I'm not sure what I would've done if I'd had the option to take it later but still be desysopped within a maximum of 30 days (because I always could've resigned at any time). My particular situation was bizarre for many reasons, the easiest to sum up here being (a) the timing, which couldn't have been worse due to long-standing RL plans in December and (b) my importer rights, which at time of writing I share with only one other person ... I would never have thought in a million years of bundling them with history-merging abilities (which was one of the sysop features I was desperate not to lose with de-adminship) until that possibility was brought up in my re-RFA (also see the relevant entry in my personal Wikipedia timeline). Graham87 (talk) 08:23, 22 October 2025 (UTC)

• Regarding B, the solution to admins not opening RRfAs that doesn't turn the petition into an RRfA in itself (I remain surpised there is a strong feeling that it is desirable to create a system where any user can immediately start an RRfA on any admin) is to automatically open RRfAs. That does have its own downsides, which is why I presume it was not initially chosen, however some could be ameliorated. In the case that at admin might presumably not want to be forced into an immediate RRfA (we have an example now of wishing to join in the scheduled election, among other potential reasons), you could make RRfA opt-out rather than opt-in. That an admin might have to answer the RfA questions under unusual time pressure would perhaps be taken into account by voters if said answers were short or absent. CMD (talk) 15:55, 21 October 2025 (UTC)

  I think this idea, skipping the petition and going straight to the RRFA, is an interesting counterpoint to the people opposing option B for the possibility of it turning the recall petition into an RRFA-before-the-RRFA. Without going back to re-read the original RFC that created WP:RECALL, I wonder to what extent the idea behind the petition was that it would be a simple check on frivolous RRFAs rather than the "just the Oppose section of an RRFA-before-the-RRFA" it seems like it has turned out to be. Anomie⚔ 16:17, 21 October 2025 (UTC)

  Scrapping the petition stage really should be what option B is, makes it a bit clearer. If there is a petition stage, I'm not sure there is a practical difference between "simple check on frivolous RRFAs" and "just the Oppose section of an RRFA-before-the-RRFA" in terms of design, although in terms of appearance the more signatures required the more it will shift towards the latter. On the former, from the samples available 10% of the petitions have been stopped at this checking stage, and I'm pretty sure I recall at least one other that I am not sure would have made it out of that stage if further issues had not occurred while the petition was open. CMD (talk) 16:51, 21 October 2025 (UTC)

  I'm not sure there is a practical difference between "simple check on frivolous RRFAs" and "just the Oppose section of an RRFA-before-the-RRFA" in terms of design If we really want it to be a simple check that at least 25 people support there being an RRFA, we could forbid any discussion or evidence in the petition at all. Just signatures. Anomie⚔ 17:24, 21 October 2025 (UTC)

  Anomie That (Having no discussion during recall petitions) was precisely one of the key components of the Original Phase I proposal, but the community rejected it in Phase II. Now that we've seen a year of recalls, the community can gauge better what worked and not, but I do believe all proposals that improve RECALL must necessarily "reduce the temperature" of the process. I think a "no discussion" option should very well be discussed. Soni (talk) 00:54, 22 October 2025 (UTC)

  Soni can probably offer the most insight but I suspect most, if not all, of the early proponents of what would become WP:ADRC did believe it would be a simple check on frivolous RRFAs. There was definitely significant discussion as to whether the discussion would just serve as a preliminary RfA but we concluded it would be better than just going straight through RfA or allowing people to make claims with no section for rebuttals. Additionally, there was some discussion as to whether signing a petition indicates 'I would vote oppose' or 'I think this should go to RRfA, even if I might support'. In practice, I suspect very few signatories have used the latter interpretation—certainly most don't vote support. Sincerely, Dilettante 16:59, 21 October 2025 (UTC)

  Dilettante Memory is a bit shoddy, but the petition process was indeed intended to filter out all the frivolous RRFAs in a "lower stress" setting. Admins would go through RFA afterwards anyway, but introducing a lower temperature "This admin does have concerns raised for them" check before the RRFA itself was a better approach than Arbcom or such. Hence, the petitions process. Soni (talk) 00:54, 22 October 2025 (UTC)

• I've done an analysis of the users signing petitions at quarry:query/98343. Of particular note, there were eight users who signed at least four petitions; twelve users with under a hundred mainspace edits this year (two with none at all, including one who signed four petitions); and one user who is currently blocked for sockpuppetry. —Cryptic 16:06, 21 October 2025 (UTC)

  That seems worrying. Especially given how easy it sometimes is for people to run sock farms. Guettarda (talk) 21:36, 21 October 2025 (UTC)

  Of those eight users who signed 4+ petitions, how many of them actually had meaningful interactions with the administrator being recalled? —Jéské Couriano v^_^v ^{threads critiques} 03:27, 22 October 2025 (UTC)

  two with none at all, including one who signed four petitions good grief. I hate to suggest minimum activity rules, but maybe we should institute minimum activity rules. -- asilvering (talk) 10:58, 22 October 2025 (UTC)

  dewiki's process (on which ours was supposed to be based) does have minimum activity rules (de:WP:SB asks for 2 months, 200 article space edits, 50 of them in the last 12 months). The "50 mainspace edits in the past year" rule would disenfranchise some people who are extendedconfirmed. —Kusma (talk) 12:07, 22 October 2025 (UTC)

  It would be ironic to have higher edit requirements than those needed by an admin to avoid "inactive" desysopping, so I guess having at least 100 edits in the last 5 years would be a good activity requirement? Fram (talk) 13:06, 22 October 2025 (UTC)

  I do think voting rights and inactivity desysops are separate issues, but I don't really want to introduce a special class of voting rights only for admin recall. Note that dewiki uses the same criteria for all votes (and they use voting more than we do). The anti-metapedian "count only mainspace edits" is also a bit questionable given how many admins are fairly metapedian in their approach (but it would at least tell people that editing their user space is not enough). —Kusma (talk) 14:03, 22 October 2025 (UTC)

  I think complex suffrage rules are a bad idea because they are hard to check. We should have an easy-to-check rule such as "!voters must be extendedconfirmed", then call it a day. –Novem Linguae (talk) 14:52, 22 October 2025 (UTC)

  I agree that complex suffrage rules are hard to check, that voting and inactivity desysops are separate issues, that it's not great to introduce a special class of voting rights only for recall, and that it's ironic to have higher edit requirements than those needed by an admin to avoid an inactivity desysop. But I just can't square the idea that, for example, we expect prospective RFA candidates to have strong content understanding (GAs, often), but are happy to let people who have no interaction with mainspace at all decided who ought to be desysopped. I don't mean to say "someone who doesn't participate in mainspace editing shouldn't have a say in adminship", exactly, but this disconnect seems to me emblematic of the broader issue with the petitions - I just don't think they reflect how we make community decisions on English Wikipedia, and I think that's a problem. -- asilvering (talk) 02:24, 23 October 2025 (UTC)

• I've been pondering option B. The concern about turning the recall into an RRFA-before-the-real-RRFA is a valid one, but on the other hand having the recall be just the "Oppose" section of an RRFA isn't great either. 30 days of getting dumped on, then you have another 30 days to decide whether to open yourself up to potentially another week of it in the RRFA? Ugh. OTOH, glancing through a few of the past recall petitions, I see some people had used the discussion section to offer support. Documenting that option (so people doing it don't have to worry about getting shouted down with "save supports for the RRFA") may be all option B really needs to be. Anomie⚔ 16:08, 21 October 2025 (UTC)
• My personal advice to any closer is to avoid relying on the concept of a bartender's close as much as possible. I understand why this idea is attractive. I feel, though, that it provides a disincentive for participants to work out a best compromise through discussion, and it sets up the evaluation of consensus to be challenged on the basis of the closer imposing their own solution. Now it may be good idea to start moving towards a model where a conciliator can weigh the voices and find a middle ground solution that leads to the greatest overall satisfaction (even if that means the least amount of dissatisfaction). I think, though, in order to ease the transition from the current consensus-based decision-making traditions, the community should discuss this approach first and establish that there is support for a conciliator role. isaacl (talk) 16:51, 21 October 2025 (UTC)

The Sword of Damocles

• I'd just like to pull out the business about sysops living under a constant Sword of Damocles where they could be recalled at any time for any reason. It's most clearly articulated by Thryduulf when he says "It is a problem that 25 people can effectively desysop an admin before they're even aware that there is a recall process."

That's how it is, Thryduulf, for people who aren't sysops. Normal editors are under a Sword of Damocles 100% of the time. Except that it's worst for us: one person can siteblock me before I'm even aware that they think there's a problem with my edits. Welcome to Wikipedia.

In my case the one person was a sitting arbitrator, and they did it without warning. The first time they edited my talk page was when they put up a block notice. And mine isn't an unusual experience. It's actually quite normal for people who accidentally cross a sysop who's tired and not reading straight.

There was no consequence for the blocking sysop. There never is.

The Sword of Damocles over my head is on a far thinner thread than the one over yours. Nuff said.—S Marshall T/C 09:13, 22 October 2025 (UTC)

How does this mean that there is not a problem with recall? Thryduulf (talk) 10:49, 22 October 2025 (UTC)

What problem? Nobody has established that there is a problem with recall. If there's a problem with recall, then show me the person who got recalled but shouldn't have.

What Mz7 has established is that there is a problem with fear of recall, and I started this subheading to talk about that fear. I want to do so in the context of the basic asymmetry between how easy it is for you to block me and how hard it is for me to recall you.—S Marshall T/C 12:06, 22 October 2025 (UTC)

I think the part you miss is that you can ask for a review by slapping a unblock template, admins have no recourse if 25 people decide to petition to remove you (besides going through rRFA which is almost garuanteed to never pass). Sohom (talk) 12:55, 22 October 2025 (UTC)

I don't think it's almost guaranteed to never pass. I think that all the people who've been recalled, should have been recalled and all but Graham87 knew their RRFA was doomed from the start, so they mostly didn't bother. I think that if you or Thryduulf went through the same process, then your RRFAs would not be doomed from the start at all.—S Marshall T/C 13:31, 22 October 2025 (UTC)

I don't buy that for a single minute. Any RRfA would by default start "in the hole" at 0/25/0 (assuming all petitioners voted nay) which would make retaining the admin rights next to impossible given the fact that they not only need to have 70+ percent to pass, but also need to overcome a by-default 25-nay deficit to do so. This is easier said than done regardless of who the administrator is. If anything, Graham87 showed the community that RRfA is redundant with the actual recall petition - it passes, your RRfA is doomed, and so there is no point participating in a process that is guaranteed to fail. —Jéské Couriano v^_^v ^{threads critiques} 14:57, 22 October 2025 (UTC)

That's why it's not 70%.—S Marshall T/C 15:09, 22 October 2025 (UTC)

Doesn't change the fact they still start up to 25 nays in the hole. —Jéské Couriano v^_^v ^{threads critiques} 15:10, 22 October 2025 (UTC)

For good reason, based on the recalls that have gone through. AugusteBlanqui (talk) 15:23, 22 October 2025 (UTC)

Regardless of whether the reason is good, it still points to the actual RRfA - the point where an admin has to defend their actions - being completely redundant/unnecessary. Why bother with an RRfA when it's patently clear you're not going to pass, especially if emotions are running high because of the recall? This also isn't a matter which you can let sit; cunctate for too long and you've effectively 172'd yourself out of adminship regardless. —Jéské Couriano v^_^v ^{threads critiques} 15:29, 22 October 2025 (UTC)

"almost garuanteed[sic] to never pass"

Yeah, citation needed.

Last year's Arcbom elections, which are kinda, by their nature, looking at people who incredibly active in project space administration (where it's apparently very easy to make enemies), and where people are holding them to a much higher standard than regular admins, and who even may strategically vote against people they trust, and we see that every single one got over 50% approval - even Beeblebrox, who was controversial to say the least, and all but he passed with over 60% approval. The pattern hold in 2023, 2021, 2020, 2019, and 2018, which is when I stopped looking.

Well, you say, maybe Re-RFA would be different- then we need to look at WWT's and HF's as well as Graham87's - and yes, Sohom, I know you're aware of those.[4][5]. And, okay, yeah, I'll concede those are a little different - Graham87 had several adverse ANI threads closed against him in the previous months, had several opportunities to course correct and instead continued making the kinds of blocks that were getting him in trouble during the recall petition, while WTT and HF did not. But no, I don't think you can look at one RfA and conclude that the others would "never pass", not when 2 out of 3 re-Rfas that year did.

Similarly, while an unblock may give you a chance to appeal - in all but one of the cases[6] I listed in my Bbb23 evidence involved an unblock appeal, at least one other admin looked at the unblock appeal and left it in place.[7][8][9][10][11][12]. Unblock admins look all day at bad faith actors who lie to them, they are not a good place to overturn unfair blocks. S Marshall would be fine, but that's because lots of people know him and like him. Random account with no userpage?

Some day, somebody needs to write an essay about how the "what if it was me" mindset shapes these discussions - admins who make a lot of blocks or deletions think "well, what if I make a bad one and I get recalled?", ignoring the fact that, for all but one case, it hasn't just been one bad decision (seriously, y'all need more self confidence), and non-admins say "but what if it was me that was blocked", fully cognizant that, yeah, reversing an admin action is not always easy. (The fact that that doesn't apply to the average VPP participant, is, well, a matter for another day) GreenLipstickLesbian💌🦋 16:19, 22 October 2025 (UTC)

TL:DR; self-fulfilling prophecies, evidence suggesting some re-rfas would pass. GreenLipstickLesbian💌🦋 16:26, 22 October 2025 (UTC)

I think the biggest issue is that Graham87's RRfA going so deep underwater had a major knock-on effect on later recall petitions. Whether it's fair or not, and irrespective of the merits and facts of the case, the fact the first petition led to a doomed RRfA very likely influenced the other successfully recalled admins and convinced them that any RRfA attempt was going to meet the same result. —Jéské Couriano v^_^v ^{threads critiques} 03:40, 23 October 2025 (UTC)

Yes, I have noticed you think that, Jéské Couriano. [13][14][15][16][17] I do understand where you're coming from, and I do concede that fear is not rational, but I disagree - there were other factors that were more influential to the decision not to go through a re-RfA, such as a hisoty of sanctions, an an apparent lack of desire to talk about admin stuff or a realization that the petition hit the threshold so quickly that any RfA would not pass. (Just saying, getting Wikipedians to agree on something is like herding cats. If you get a twenty five of them to agree on anything in less than 24 hours, that would be a WP:SNOW decision in literally any other venue. Most admins are trained to respect blizzards above all else) GreenLipstickLesbian💌🦋 08:10, 23 October 2025 (UTC)

I cite Fastily's recall as a counterpoint, as that took 11 days. They still decided to both resign and, for good measure, fuck off afterward. —Jéské Couriano v^_^v ^{threads critiques} 15:47, 23 October 2025 (UTC)

A not unique reaction... And one that both sides here seem to interpret very different... One seems to view retiring as proof that the admin was only in it for the thrill of the tools not the good of the project and on the other side its seen as proof that this process is brutally driving editors in good standing away from the community permanently. Perhaps we could split the difference if we treated it less as a punishment and more as a democratic procedure which included thanking the admin for their service, highlighting their best contributions (their worst were likely highlighted in the recall), and welcoming them back into the realm of non-admin editors with a celebration. Horse Eye's Back (talk) 15:55, 23 October 2025 (UTC)

Here's another idea. It could be because the admin lost the spark and started seeing Wikipedia as more of an obligation. With the removal of the bit, the admin no longer sees any reason to continue working on something they don't enjoy working on anymore. fanfanboy (block_talk) 17:39, 24 October 2025 (UTC)

Thats a very fair point, I take it all within the context that ragequitting/retiring seems to be a rather temporary phenomenon most of the time... I would be easier to name all the retired editors who actually retired than the retired editors who either un-retired or still edit despite a prominent claim of retirement on their user page. Horse Eye's Back (talk) 18:31, 25 October 2025 (UTC)

I thought you might - but to do that, you have to ignore their previous losses of adminship & commons re-RfA. GreenLipstickLesbian💌🦋 16:57, 23 October 2025 (UTC)

Different process is different. More as this story develops. FOARP (talk) 07:15, 23 October 2025 (UTC)

I would go even further to say that anyone who is so attached to their admin bits that they consider the possibility of losing them to be a "Sword of Damocles" situation should become less attached to them, one way or another. I don't know to what extent the whole "janitor" thing has ever been true--it might well have always been a polite fiction--but we are certainly straying further and further from it, to our detriment. Writ Keeper ⚇♔ 13:30, 22 October 2025 (UTC)

S Marshall, I think any number of admins, including the former ones recalled for inactivity, have far less of a margin of mistake than you do. You are an active member of the community who does important and highly visible work which garners you a large amount of respect. That respect means that if some admin were to attempt to sanction you there would be pushback such that it seems likely that the sanction either would not pass (if put up for discussion) or would be reversed (if it was). However, I would think that a sanction would become more likely to pass or stick if the people whose voice who respect you and would defend you against such a sanction, carried no weight. Best, Barkeep49 (talk) 15:20, 22 October 2025 (UTC)

Thank you for those kind words. You're right to say that I'm one of the vested contributors that Wikipedia likes to pretend it doesn't have. I have the credibility that comes with a username that people know, and I'm accultured to Wikipedia so I know what tone to take with people. As you rightly say, it inoculates me against some of the behaviours that some sysops freely display with less accultured, less vested editors. Please could you clarify your final sentence? I'm afraid I haven't been able to follow it.—S Marshall T/C 15:46, 22 October 2025 (UTC)

Sorry for not being clear. What I was trying to say is that in the scenario where some admin is out to get you because of some honest mistake you made, I think the current processes make it unlikely that if proposed it would get consensus and if it were enacted without discussion that it would stick. However, I think that's true because the discussion would give appropriate weight to those who would defend you as it would to your critics. Right now, in the petition process, the critics are the only voices given any weight and admins have so far been less willing to go to the second phase of the process (2 out of 10 times) than in the model we copied from. Best, Barkeep49 (talk) 16:08, 22 October 2025 (UTC)

The recall petitions I've read have all been preceded by discussions on Talk Pages, ANI, and elsewhere--usually where the admin in question has enjoyed the support of other admins. So there has been plenty of defense before the recall. In fact, it was this circling of the wagons, the admin choruses of 'but they do valuable work so overlook the abuses,' that led to the recall process in the first place. AugusteBlanqui (talk) 16:21, 22 October 2025 (UTC)

I've certainly made honest mistakes while editing Wikipedia, and also stupid mistakes. But exactly 0% of the blocks in my block log are because of my mistakes. I got site-blocked once in 2008 by sysop error that she immediately reversed, and once in 2012 in these circumstances. I've also been blocked once in 2021 from one user talk page because Mjroots took it on himself to invent a novel and ill-conceived blocking rationale that I never bother to challenge because I was never going to edit that page anyway.

If those blocks had needed 25 signatures in advance before being enacted, they would never have got through. Not a single one of them.

This is why I reject the contention that 25 signatures is somehow a low threshold for sanctioning someone. It's an extraordinarily high threshold for what's intended as the start of a process. I mean, if I brought a case to Arbcom asking for a desysop, and it had twenty-five independent signatures on it, then Arbcom would definitely take that case. I guarantee it.

I'm not surprised at all that in the limited number of cases where we've got 25 signatures for recall, it's been justified and the sysop (having stood back and taken a look at the diffs) has thrown up their hands and admitted defeat. That's healthy.

What's missing here is any example of a person who got desysopped by this process and shouldn't have been. Produce one of those. In the absence of that, I think Mz7's suggestions are trying to solve a problem we don't have.—S Marshall T/C 16:50, 22 October 2025 (UTC)

What's missing here is any example of a person who got desysopped by this process and shouldn't have been is not the standard you laid out originally. The standard you laid out was The Sword of Damocles over my head is on a far thinner thread than the one over yours. Nuff said. This is what I rebutted. That said, I do not think 25 people should, without any chance of discussion or opposition, get to rewrite our inactivity rules. Best, Barkeep49 (talk) 02:03, 23 October 2025 (UTC)

They don't, and haven't. If people choose not to run for RRfA, that's their business. The consistent refrain is that people choose not to run for RRfA because their RRfAs would inevitably fail, but if they can't pass an RRfA, why should they be an admin? Writ Keeper ⚇♔ 02:24, 23 October 2025 (UTC)

The consistent refrain is that people choose not to run for RRfA because their RRfAs would inevitably fail I've never heard this from anyone who didn't actively support desysopping the admin in question. What I hear from everyone else is that people choose not to run for ReRFA because why would anyone voluntarily submit themselves to a week of criticism and abuse at RFA after they've just had days or weeks of criticism and abuse in the form of a recall petition? Thryduulf (talk) 02:47, 23 October 2025 (UTC)

Feel free to read the conversation in this very section, then. Writ Keeper ⚇♔ 02:56, 23 October 2025 (UTC)

I have, and I stand by my comment. Thryduulf (talk) 02:57, 23 October 2025 (UTC)

Look, if 25 independent people came to me in the same month to say, "Listen, S Marshall, we've lost confidence in you and we don't want you to close any more RfCs", I would definitely stop closing RfCs. Because of course I would! Completely step back, and leave all discussion closes alone until I was clear that I had the community's confidence again. 25 independent people in the same month is quite an extraordinary number of people considering the limited number of active Wikipedians nowadays. I can't understand the contrary view at all. When 25 independent Wikipedians say the same thing in the same month how can you possibly just brush that off?

If we were at block review, 25 independent "endorse" votes looks like a snow close for a site ban. It's a crystal clear signal that the community has totally had enough. If we were at Arbcom, 25 independent people certifying the need for a case is a slam dunk red flag that a case needs opening.

I mean, people do get annoyed with S Marshall. One person or two people having a go at me for one of my closes is just Tuesday. I've definitely annoyed more than 25 people with my closes, but they're sure as hell not all coming to me in the same month to say stop doing it and if they did, I'd totally step back.—S Marshall T/C 03:13, 23 October 2025 (UTC)

But: Would you? Should you? For any 25 editors? Even if the 25 editors are a couple of people who are mad at you because this afternoon because the closing statement you posted this morning is the opposite of what they voted for, and their loyal friends? Context matters, and I'd hope that you'd take the "sore loser" aspect into account when deciding whether the complaints warrant such a response. WhatamIdoing (talk) 03:58, 23 October 2025 (UTC)

Yes, I think I would and should.

I'd likely start a post mortem in some central venue to ask: what went wrong? Had I just royally screwed up? I mean, if the community largely thought I hadn't, and this was just a case of outraged redditors all registering accounts in order to brigade me, then I'd likely start back up again. But while that discussion was going on I definitely ought to stop, shouldn't I?—S Marshall T/C 04:35, 23 October 2025 (UTC)

I am not sure why you talk about 25 independent users. They are not independent. They read each other's comments, and it is not uncommon that they write smth like "I was not going to sign this, but the arguments of X who pointed out a major mistake in judgement of Y convinced me". And of course X pointed out smth which they perceive as a Y's major mistake in judgement but they were not interested or did not have time or whatever to go and see how characteristic for Y this mistake is, and what does actually the rest of contribution of Y consists of. Ymblanter (talk) 10:18, 23 October 2025 (UTC)

This was notable particularly in the Night Gyr fiasco, where the reason for desysopping morphed part way through from inactivity to that fact that Night Gyr hadn't responded ... to the petition ... FOARP (talk) 12:16, 23 October 2025 (UTC)

That's not at all my reading of that petition. It seems to me that the concern was not simply NightGyr's failure to respond to the petition, but that their nonresponsiveness to the petition was part of a decades-long disregard of WP:ADMINACCT and WP:ENGAGE. Caeciliusinhorto-public (talk) 14:17, 23 October 2025 (UTC)

"Given that they have edited since the launch of this petition and thus are or should be aware of it but have not replied, I think a revocation is justifiable" - This was an argument cited as convincing by a number of the participants. FOARP (talk) 14:36, 23 October 2025 (UTC)

Lack of responsiveness to criticism was one of the points made in the original petition: it wasn't a new issue brought up part way through. The specific argument by WellingtonBay about not having responded to the petition itself was cited as convincing by one signatory, who had already signed the petition two full weeks before WellingtonBay made that argument.

By contrast Levivich's argument about a pattern of ADMINACCT issues was cited as convincing by eight of the subsequent ten signatories (nine if you count QoH's "per EggRoll", which in turn cites Lev). The point that it was part of a pattern had already been alluded to by several prior participants: three specifically cite the fact that NG had not responded to a single post on their user talk page since 2007. Including the original petition.

You may or may not think that the concerns raised in the NightGyr petition were reasonable, or that they rose to the level of meritting desysopping – if I were god-emperor of Wikipedia I'm not sure I would have desysopped on the basis of the arguments made – but it's not true that lack of responsiveness was a new argument that came about partway through the petition, and it's not true that lack of responsiveness to the petition specifically was ever the primary justification for people to sign the petition. Yes, a couple of people did cite lack of responsiveness to the petition specifically, but both of those also cited either inactivity or more general concerns about responsiveness. Caeciliusinhorto-public (talk) 15:47, 23 October 2025 (UTC)

Lack of response to the petition being seen as a clinching argument (you'll note the petition stalled part way through, this was seemingly the argument that got it over the top) is necessarily an argument that wasn't made at the start of the petition. FOARP (talk) 19:37, 23 October 2025 (UTC)

a sysop who's tired and not reading straight Can confirm. In my case, I had only been around for a few months, I asked a regular (now vanished) to stop marking non-minor edits as minor, and they responded by reporting me to AIV. Gave me a heart attack, but I was unblocked by the blocking admin within minutes with the apology of "sorry, I was reading too many tabs" (paraphrased). Just goes to show how easily it is for most editors, especially newcomers, to be shown the door. QuicoleJR (talk) 18:17, 25 October 2025 (UTC)

seemingly got it over the line in what sense? If you removed the zero people swayed by Wellington Bay's vote into signing the petition, it still would have got over the line. If one vote can be credited with getting it over the line, it's Levivich's which was, as I already said, explicitly cited by eight subsequent signatories. Nor was Wellington Bay's signature the one which re-started people signing the petition after it stalled on 4 July; they were the seventh to sign the petition after signing resumed on the 11th. Even the one vote you cite – which in no way got the petition over the line except in the trivial sense that as the petition is closed once it reaches the threshold, removing any of the votes would drop it back below the threshold – does not cite non-responsiveness to the petition as the primary issue! Caeciliusinhorto (talk) 20:09, 23 October 2025 (UTC)

• A modest proposal - Petition to remove this system, 25 unreasoned votes should do it. FOARP (talk) 21:22, 22 October 2025 (UTC)

  @FOARP if by "this system" you mean WP:RECALL, then I would support removing it pending significant reforms along the lines of those I've proposed and those Mz7 has proposed (after further workshopping). Thryduulf (talk) 21:30, 22 October 2025 (UTC)

  Sure! Provided proponents are allowed to host an rfc (basically a vote) with a 60% support threshold to keep it. Sincerely, Dilettante 10:15, 24 October 2025 (UTC)

Evidence

The OP doesn't seem to provide any evidence or links. So, I went looking for a list of the cases to date and here it is. I'm still not seeing any problems. Andrew🐉(talk) 17:14, 22 October 2025 (UTC)

This is a recall system and it has not led to any recall votes. That would seem to be a flaw in its design. FOARP (talk) 21:25, 22 October 2025 (UTC)

It led to one recall vote. That was withdrawn as hopelessly underwater at the 11th hour. And happened to be the first recall case. Is it any wonder none of the others bothered with the vote? —Jéské Couriano v^_^v ^{threads critiques} 03:37, 23 October 2025 (UTC)

The OP didn't provide a link to the process in question so here it is: Wikipedia:Administrator recall. This could use more background but the point of the process seems to be as described by recall election: to "remove an elected official from office". The mechanics of petition, RRfA and the rest are just means to an end – the removal of dysfunctional admins. The process seems to be working as intended and no-one has produced a counterexample.

It seems obvious that, in these cases, most subjects did not proceed to an RRfA because they expected to lose and so chose to avoid further humiliation. This seems to be a good thing as, if an RRfA were started regardless, the resulting pile-on would generate unnecessary drama and bitterness. The current option softens the blow rather than insisting on such a walk of shame.

Andrew🐉(talk) 07:44, 23 October 2025 (UTC)

"The process seems to be working as intended" - Please show evidence that this was ever intended to simply remove admins with 25 votes (not !votes, votes). All of the discussions up until this system was brought in had the petition as simply a sense-check/filter before the main decision. Instead it's all decided there, every time. FOARP (talk) 11:37, 23 October 2025 (UTC)

It was intended to recall admins who would fail to pass an RfA, the fact no recalled admins have succeeded so far is not a sign of being broken but a sign of functionality. We haven't even started to have admins recalled who have passed an RRfA in order to revise the parameters better. The fact community consensus hasn't re-affirmed such a petition yet is not a sign there isn't consensus, only that it hasn't been tested yet. CNC (talk) 12:02, 23 October 2025 (UTC)

This is railroaded-prosecution system where the result is pre-decided by the time the convict gets to court, and so pleads guilty every time rather than go to a full trial. FOARP (talk) 12:23, 23 October 2025 (UTC)

It was intended to recall admins who would fail to pass an RfA, this is completely and fundamentally incorrect in every respect. The petition was intended to be a sense-check with the sole purpose of preventing frivolous ReRFAs where there would be no chance of failing. It was intended that every successful petition would result in a re-RFA that some administrators would pass and others would fail. Thryduulf (talk) 12:31, 23 October 2025 (UTC)

That's a blatant lie. When I proposed the petition process, I was very well aware that some admins will choose to not go through RRFA. That happens during Arbcom all the time. And in case of at least one recalled admin, it happened in the past before happening again in Recall.

Of course, in an ideal world, more admins will attempt to RRFA, and some of them will succeed. We should workshop ways to make the RRFA itself less stressful to make that happen more. But nobody in the making of recall, best i can recall, thought to force literally every admin through RRFA, kicking and screaming, against their will.

I understand the need to badger or stand behind your points. I just request you to not rewrite the past. Soni (talk) 14:28, 23 October 2025 (UTC)

" I was very well aware that some admins will choose to not go through RRFA" - We're you aware that all of them wouldn't? If so, could you let us know what next week's lotto numbers are? FOARP (talk) 14:39, 23 October 2025 (UTC)

That's all completely irrelevant to what I said, please try reading it again. What you were thinking when you proposed a petition process is something irrelevant to both what I said and to what the community approved.

The petition process, as approved by the community, was intended to prevent frivolous reRFAs - nothing more, nothing less. Of course nobody suggested forcing admins to stand again, and the desysopping if they don't was to there to ensure they couldn't just ignore the request to stand at RFA again. The RFA is what determines whether someone has the support of the community or not - the petition is explicitly and intentionally not a consensus process.

If you think this is an attempt to rewrite history then you are either failing at reading comprehension or not here in good faith. I don't know which I prefer. Thryduulf (talk) 14:40, 23 October 2025 (UTC)

It was intended that every successful petition would result in a re-RFA that some administrators would pass and others would fail.

This was still not true then, and I hope this is still not true now. The only way to guarantee it after a successful petition is by forcibly making the admin go through an RFA, whether they want it or not. Soni (talk) 14:50, 23 October 2025 (UTC)

Or, you know, not having this process because, when it isn't actually resulting in any RRFA's, then it isn't working as it was supposed to. FOARP (talk) 15:06, 23 October 2025 (UTC)

(edit conflict) You're right, I should have been more cautious to your reading bad faith into my contributions here. Accordingly, I'll rephrase: It was expected that there would be a mix of successful and unsuccessful petitions. It was intended that nearly-every successful petition would result in the admin standing for re-RFA. It was expected that some re-RFAs would be successful and some would be unsuccessful. It was not expected or intended that in less than 48 hours 25 editors could drive a fellow editor away from the project completely without any attempt to see whether those editors views matched the consensus of the community. Thryduulf (talk) 15:07, 23 October 2025 (UTC)

There's a lot of passive voice in your post so I'm not sure who had these expectations and intentions. I don't see them in the policies or the discussions that created them. AugusteBlanqui (talk) 15:14, 23 October 2025 (UTC)

I'm surprised you can tell the passive voice apart from the passive aggressive voice, lol. Although I can see why Thryduulf would be so sensitive about "It was intended to recall admins who would fail to pass an RfA" because they almost certainly fall within that category, almost no chance of passing a modern RfA. Horse Eye's Back (talk) 15:20, 23 October 2025 (UTC)

If you think that I don't have the consensus of the community to be an administrator, then go ahead and start dispute resolution proceedings against me, citing your evidence that I'm a net negative to the project as an administrator. Or you could just jump straight to a recall petition, nicely illustrating one of the flaws in the system as implemented. Thryduulf (talk) 15:31, 23 October 2025 (UTC)

I notice that you don't appear to disagree that you wouldn't be able to pass a modern RFA. I have no interest in there being less admins, I think we're understaffed as it is... But I think its important to point out that you have a clear conflict of interest when it comes to how to interpret this, and that your chosen interpretation is in line with your conflict of interest. What Thryduulf the Admin thinks the meaning and intent is also happens to be the meaning and intent which is most useful to Thryduulf the Admin. Do you see the issue here? Horse Eye's Back (talk) 15:41, 23 October 2025 (UTC)

RFAs are uncertain processes. I honestly have no idea if I would pass one, because I never went through one. I went through AELECT, where no-one asked any actual questions or expressed any actual concerns, yet somehow I still ended up with circa 100 oppose votes. FOARP (talk) 15:56, 23 October 2025 (UTC)

Thank you for that context. Horse Eye's Back (talk) 15:59, 23 October 2025 (UTC)

My WP:ORFA was very positive, but there are apparently 106 editors in good standing on Wikipedia who thought in October 2024 that I should not be an admin (admittedly versus 268 who thought otherwise), and I don't even know their reasons for voting the way they did, so who's to say that 25 editors could not be gotten together?

Ditto since I never passed RFA, who's to say I'd pass RRFA regardless of circumstances? FOARP (talk) 07:52, 24 October 2025 (UTC)

In Oct 2024 6 of 11 successful candidates had >100 oppose votes, then in Jul 2025 only 2 of 9 and the mean oppose dropped significantly. I really don't think that it's comparable to an equivalent RfA/RRfA which presents significantly different numbers for successful candidates. AELECT is clearly more difficult than RfA so don't be thrown off by anonymous votes that wouldn't sign a petition publicly, let alone chime in at ANI etc. CNC (talk) 10:26, 24 October 2025 (UTC)

In the October 2024 AELECT, we had a sizeable minority of editors voting against anyone whom they ran out of time to research. Instead of "I don't know, so I won't vote" for that person, it was "I didn't plan enough time to research this person, so I will assume they're bad and vote against them" WhatamIdoing (talk) 23:33, 24 October 2025 (UTC)

Not quite. There were 18% less votes overall, down from 659 to 541, so the abstentions and the oppose votes decreased proportionately. A couple of candidates dragged the mean support up, but that came from a notable decrease in oppose votes, not from abstentions which remained balanced. CNC (talk) 08:37, 25 October 2025 (UTC)

We literally had editors saying that they voted against anyone they didn't have/make time to research. WhatamIdoing (talk) 04:48, 10 November 2025 (UTC)

I have no idea whether I would pass a modern RFA, because modern RFA is an unpredictable hell-hole that bears little relation to the process as it was when I passed near-unanimously in 2005. There are a small number of non-admin editors in generally good standing who I would feel confident saying would not pass an RFA if they stood today. There are exactly zero editors (admins or otherwise) in any standing who I would feel confident predicting they would pass an RFA today. There are many that I hope would pass, but that's not the same thing.

If there wasn't such community opposition to voluntary reconfirmation RFAs then I would have stood for one at some point in the past few years. As it stands I have no interest in people opposing me because they object to the idea of admins verifying they still have the trust of the community when not being forced to (and make no mistake, some people will oppose candidates at reRFA simply because it is a reRFA) which is completely irrelevant to whether I do or do not hold the community's trust.

As for the conflict of interest, that's identical to the conflict that literally every administrator on the site has and exactly equivalent to the conflict of interest that every non-adminstrator has in the same process, making it utterly meaningless. It is also not a "chosen interpretation" it is the only interpreation that logically follows from the facts as far as I am aware of them. Thryduulf (talk) 16:14, 23 October 2025 (UTC)

I agree that modern RFA is not ideal and perhaps even broken... Which leaves us in something of a chicken and the egg situation, perhaps if we fixed RFA the recall process would be easier to figure out. On the other point "exactly equivalent to the conflict of interest that every non-adminstrator has in the same process" that simply isn't true because one has a position of power and prestige to lose and the other doesn't. Thats like saying that the politician and the voter have exactly equivalent conflicts of interest when it comes to whether or not that politician should stay in office. Horse Eye's Back (talk) 16:21, 23 October 2025 (UTC)

On your first point, it was hoped by at least some people that the existence of recall would be what fixed RFA on the basis that it would make adminship less of a big deal, and several others were less hopeful but felt it wouldn't hurt to try (I think I was in this camp). Unfortunately that hasn't proved to be the case, but I don't see that as being related to how recall has turned out in practice versus how it was intended to work. Thryduulf (talk) 16:36, 23 October 2025 (UTC)

If you'll pardon my tangent, Thryduulf, I know I've said that I think most re-RfAs would pass. I don't think you would, though, sorry - your last 500, nonminor mainspace edits (most of which are redirect or template gnoming related[18]) bring us back over a year, which isn't actually an issue, except for the fact that you've not built up a reputation as somebody who does a lot of encyclopedia editing. Instead, I'm hazard a guess most people know you from talk or projectspace discussions- which you habitually comment more than anybody else, and accuse anybody who says something you disagree with of not answering your question, not refuting your arguments, or being disconnected from reality.

Quick anti-aspersion defense, from the past few months:

• this conversation - you've made 18 comments, including saying that people who disagree w/ you were disconnected from reality[19] - to which I'm going to respond w. your other comment what is not acceptable is you claiming that evidence-backed views that differ from yours are contrary to reality which you made literally in response to somebody who called you out on your first reality comment.
• Talk:London attack#Requested move 26 September 2025 - as of writing, you've made 10 comment & have added over 1/3 of the pages's text[20] - accused one person of not answering your question because you didn't agree w. their answer?[21]
• WP:VPR#RFC: Adding featured and good content status to the tagline - you have made 43 comments, by my count over, quite literally, if we should tack on half a dozen words to a banner that regular people do not read - [22] + saying that people were responding to you but not refuting your arguments.[23] (which, using the same argument you've used here- how do we know they haven't been refuted until a closer judges that? Schrodinger's Wikipedia Consensus cuts both ways)
• edit summaries referencing other editors relationship to reality[24] (some false positives)
• Wikipedia:Village pump (proposals)/Archive 219#Creating an edit filter or automatic new page flag for likely LLM-generated material - in which you justified the amount you wrote, when called on it, by saying as I've "spilled all this ink" repeatedly explaining.[25]
• Night Gyr recall- you made 54 comments, well above anybody else who wasn't bludgeoning [26] and, when asked about it, asked if others would like to shut up w. the es no need to make multiple requests for people to answer the question if they just answered the questionb [27]

Sanctionable bludgeoning? No idea! But you can't do stuff like this & expect to build up good will. (I say this, not as a moral judgement- I talk too much myself at the best of times, :P ) GreenLipstickLesbian💌🦋 18:19, 23 October 2025 (UTC)

@AugusteBlanqui I read those expectations and intentions very clearly in the significant majority of the comments that resulted in RECALL. If you don't, then maybe you could start by showing some evidence that something else was intended. Thryduulf (talk) 15:33, 23 October 2025 (UTC)

You didn't expect admins would just fuck it off if 25 editors signed a petition? I kind of expected that mainly to be honest, not much else. CNC (talk) 15:21, 23 October 2025 (UTC)

Recalls also don't need to provide evidence of links :P Stifle (talk) 08:47, 23 October 2025 (UTC)

Looking at the first example of Wikipedia:Administrator recall/Graham87, it seems that the initiator provided lots of links to issues and previous discussions. Its closer likewise provided lots of links. I've not looked through the other cases to see if they followed this example but would expect any respectable petition to do something similar. Andrew🐉(talk) 09:58, 23 October 2025 (UTC)

Petitions also don't need to be respectable.

A petition can be opened against any admin, based on any reason, and the issue is decided once 25 votes are amassed. I'm only being semi-unserious when I say that "FOARP is a filthy deletionist who smells" would probably get to 15 votes at least, possibly 25, and the RRFA would be a whole bunch of editors !voting "Neutral - need to see more evidence about FOARP's bodily odour before I decide either way". FOARP (talk) 11:45, 23 October 2025 (UTC)

With all due respect, FOARP, you seem to be arguing against some hypothetical recall that might be called, instead of the actual cases and the examples at hand.

Plenty of editors believe that the recalls that have happened so far were correct decisions. You may disagree.

just don't think your over the top comments are taking those disagreements in good faith Soni (talk) 14:08, 23 October 2025 (UTC)

FOARP's comments are much closer to the reality of recall than many of the comments from those who think there is nothing wrong with the current situation. Especially with regards to the inactivity petitions which, intentionally or otherwise, are applying a much stricter activity requirement than the one that has community consensus. While their language is emotive in some cases, I see absolutely no evidence that they are here in anything other than good faith (and I say that as someone who has had distinct concerns this regard about some of FOARP's comments regarding other matters). Thryduulf (talk) 14:26, 23 October 2025 (UTC)

As someone who called out your misinformation literally a couple minutes ago up thread, I just don't take you for a neutral arbiter of reality, sorry.

I understand that both of you hold very strongly held views on RECALL, but either you are just refusing to accept that other people may also have valid perspectives. Or we are both looking at two very different Wikipedias.

I'm remembering an adjacent RFC closed by @Tamzin which has a very stark divide between admins and non-admins. I can't recall what RFC it was, but I still remember the close for how starkly different the two perspectives were. How are we both looking at the same process and come to this disparate a conclusion Soni (talk) 14:39, 23 October 2025 (UTC)

You have not called out any misinformation by me. You have simply claimed that my view is wrong because it doesn't match your preferred viewpoint. There are multiple different perspectives here, and that's understandable, but what is not acceptable is you claiming that evidence-backed views that differ from yours are contrary to reality. Please desist with the accusations of bad faith. Thryduulf (talk) 14:46, 23 October 2025 (UTC)

Soni is correct that the literal statement is incorrect: it wasn't assumed that every petition attaining the threshold number of supporters would result in a re-request for adminship, but that some admins might cede their privileges (just as some admins have said, if enough people raise concerns with their actions, they would resign as admins). It is also true, though, that the petition process was designed to be a precursor for substantial discussion to be held during the re-request. Current practice is that the community is proceeding straight into substantial discussion in the petition. isaacl (talk) 15:37, 23 October 2025 (UTC)

I want to note that while I think certain things are wrong, I am calling this a check-in for a reason. I probably should have linked to the page that listed the 10 recall petitions to date, but otherwise, I intentionally was trying to allow editors wide latitude about how it's going, including by offering clearly mutually exclusive options. Best, Barkeep49 (talk) 20:17, 23 October 2025 (UTC)

Another link that seems to be needed is WP:BARTENDER to explain what is meant by a "bartender's close". That's an essay, not policy, and OP's reference to this seems to be a blatant violation of WP:RFCNEUTRAL. The problem of resolving this multiple-choice morass arises because the RfC has been framed badly, contrary to WP:RFCBRIEF. Tsk. Andrew🐉(talk) 08:01, 23 October 2025 (UTC)

Yeah. This RFC probably should have been workshopped into like 10 individual RFCs that ask specific, actionable questions such as "should recall be repealed?", "how many days should a recall run?", "what are the requirements to vote in a recall?", etc. Once the workshop was finished, then the finished questions and options could have been presented for !voting. Ah well. –Novem Linguae (talk) 09:15, 23 October 2025 (UTC)

Note that this entire mess is the product of such a decision-making process, where a small group of editors, based on assumptions about how the system would work that have not proved correct, pushed this through. FOARP (talk) 11:39, 23 October 2025 (UTC)

These were also my thoughts. Even trying to remember the 9 different options is difficult enough for !voting, let alone trying to follow the discussion. It's unlikely there will be consensus for anything at this rate. CNC (talk) 11:48, 23 October 2025 (UTC)

I wonder if this is why the OP took the unusual step of providing instructions to the prospective closer in the OP itself? AugusteBlanqui (talk) 12:20, 23 October 2025 (UTC)

Note a followup request for comments discussion was workshopped starting last November and heading into this year. The election of theleekycauldron to the arbitration committee stalled the effort and it didn't proceed. (*) There's a hard balance to strike, because the number of people who will engage in workshopping ten RfC questions is typically considerably smaller than the number of people who hold opinions on them, making it hard to reach agreement on the RfC questions. So sometimes it may just be better to have a broader request for comments discussion that essentially workshops proposed changes. Unfortunately, it's not very easy to tell which approach is most suitable for a given situation.

(*) Of course, there is a reasonable argument that a review at the start of 2025 was a bit soon, and that having a review now, after a year, allows for more data to be taken into account. isaacl (talk) 15:23, 23 October 2025 (UTC)

• As a further bit of evidence, I was just checking my watchlist and noticed that Graham87, who was the subject of the first case, seems to be still editing quite amiably. That case seems to be a good example of the process working well as they went most of the distance and only withdrew and resigned when the outcome seemed settled. And now they seem to have taken that outcome in good heart and moved on with their other good work. This stoic, "no big deal" attitude should be encouraged and applauded. As another example, founder Jimmy Wales has had their privileges criticised and curtailed over the years but he's still staying the course, "One of my fundamental beliefs is that Wikipedia should always stand ready to accept criticism and change..." Andrew🐉(talk) 17:20, 24 October 2025 (UTC)

What was recall created for?

Above, many editors who are opposing recall completely or want to either turn it toothless or change it into some RRFA itself, seem to question what recall was intended for. So, let's check this.

Wikipedia:Requests for adminship/2024 review/Phase I#Proposal 16: Allow the community to initiate recall RfAs introduction: The intention of recall was "the community will be able to revoke the permissions of an administrator pending a new RfA", because "There is no means to remove permissions for administrators who have lost the trust of the community" except Arbcom. Recall was never intended to be some vague, non-binding discussion that some admin may need, perhaps, some check, no biggie. It was explicitly intended for admins who had (or seemed to had) lost the trust of the community. Only one has so far decided whether they really had lost that trust, the others realised they had no chance of passing (e.g. the inactivity ones) and/or had other reasons not to run: that's up to them, or to issues with RfA itself, not up to the recall process.

Wikipedia:Requests for adminship/2024 review/Phase II/Administrator recall closure: "adopting a community-based process for the involuntary removal of administrator rights, administrator recall."

I don't think there was much doubt in the mind of the proposers or closers of these proposals about what exactly was proposed. When 25 extended confirmed editors express their concerns whether an admin still has community support, the admin loses his toolset unless they go to RfA again. The "de-adminning" was a clear, major part of the proposals, not some unintended consequence we are now stuck with. Fram (talk) 13:21, 23 October 2025 (UTC)

I support this Iggy pop goes the weasel (talk) 16:46, 23 October 2025 (UTC)

I will pull out Wikipedia:Administrator recall/Night Gyr as the "unintended" case here. Back when I was following the discussion, I perceived two distinct waves, the first wave representing the first rationale that for all intents and purposes did not reach consensus and a second rationale that strung together a set of small mistakes that together with the old rationale found enough to de-admin a user. This makes me wonder whether if there was actually any "loss of trust" before the recall itself. Sohom (talk) 17:11, 23 October 2025 (UTC)

I think a lot of my disappointment with recall comes from it being originally advertised as "based on the dewiki process". That process (where recall petitions against every admin are open outside of certain grace periods, and recall votes seem more independent of each other) is a lot slower and calmer than what we have here. Our recall petitions look more like old-style user RfCs, just without the dissenting views and a right to reply. That admin recalls end in desysop is fine, but the level of drama is not. —Kusma (talk) 18:42, 23 October 2025 (UTC)

Are there clear reasons for the difference? Is the procedure significantly different? Is it a difference in temperament and culture? Or is it just happenstance, as we haven't had many cases yet?

And is the difference significant or just cosmetic? Your insights might help us here, please.

Andrew🐉(talk) 19:07, 23 October 2025 (UTC)

I am not very steeped in dewiki culture, but I am a native speaker and have observed their processes quite a bit (mostly from the outside). My gut feeling is that the difference is related to some deep wikicultural differences: dewiki did not have such a strong "polls are evil" anti-voting culture as enwiki from the start, and they still use voting in many places, not enwiki's weird blend of discussions and voting. Their RfAs and re-RfAs and recall petitions are closer to straight voting than ours, which takes away some of the drama and lowers the temperature. —Kusma (talk) 19:46, 23 October 2025 (UTC)

@Kusma I think a lot of my disappointment with recall comes from it being originally advertised as "based on the dewiki process". That's because it was originally closer to dewiki process, until it got blended with enwiki's priorities on "Discussion is important". Phase II RFC had discussed every single aspect of recall, and chose to explicitly allow discussion in the Petition process instead of being a straight voting.

I still believe that disallowing discussion in the petition process would reduce the drama of recall significantly. If that's not feasible, I think we would benefit from other ways to reduce temperature through the recall process. Soni (talk) 04:50, 24 October 2025 (UTC)

I don't think we can disallow any type of communication between editors, but we could force it off the recall pages. We won't be able to police off-wiki discussions.

For a fairly extreme version, imagine that the person initiating it is allowed to post a couple of diffs or links to logs, and that nobody else is allowed to post anything except their signature. Would we get more editors checking the admins edits/comments/actions themselves, instead of responding only to what someone else claimed said? That might be a good thing. What would we lose with such a system? WhatamIdoing (talk) 05:16, 24 October 2025 (UTC)

Reasoned debate, is what we'd lose; although experience suggests we wouldn't, and it would mainly just move to talk pages or somewhere in project space. But my main objection to this suggestion is the game theory of it. "Ooops, I screwed up. I'd best use one of my other accounts to start a recall petition against myself that's got shaky reasoning and really terrible evidence, so nobody else can present the diffs."—S Marshall T/C 10:00, 24 October 2025 (UTC)

Isn't discussion what the RRFA part of the process is supposed to be for? Under the original model, the petition part was just to make sure there were at least 25 editors who think the discussion should happen. Anomie⚔ 13:52, 24 October 2025 (UTC)

Is "reasoned debate" what we're getting under the current system? For example, do any declared opponents change their minds? WhatamIdoing (talk) 23:36, 24 October 2025 (UTC)

Clearly not! What we're getting is cloture when self-selected closers end the debate, often after less than a couple of days. Their motive tends to be a mercy killing; but by doing so they prevent the supporters from changing their minds and withdrawing their support.—S Marshall T/C 10:26, 25 October 2025 (UTC)

I agree on this – I would prefer recall petitions to have a short, fixed duration, rather than be closed as soon as they hit 25 signatures. That would allow for more self-reflection from the admin being recalled and perhaps more time for things to cool down and alternative resolutions to be found. Toadspike [Talk] 03:19, 26 October 2025 (UTC)

Every single day, we see new accounts gaming the system to become "extended confirmed" in order to obtain various tools and benefits. It is equally certain that there are accounts created for such purposes that are not so obvious in their efforts to reach that status. At the same time, The Signpost has reported on efforts by outside organizations to disrupt Wikipedia through various means, including doxxing of editors. We would be suicidally naive to think that entities intent on damaging Wikipedia would not exploit the vulnerabilities inherent in enabling a relative handful of manufactured accounts to undermine the trust and safety structures of this project through targeted recall efforts against the admins most likely to counter their efforts. BD2412 T 19:29, 23 October 2025 (UTC)

We don't have evidence that this is happening, though. One of the obvious targets would have been Wikipedia:Administrator recall/Bbb23, but that had many trusted Wikipedian signatories including multiple admins. —Kusma (talk) 19:54, 23 October 2025 (UTC)

This conspiracy theory works both ways. If malign parties wish to subvert Wikipedia then they would want to aim high by gaining admin powers. We have steadily lowered the bar and now have more open admin elections. To maintain effective checks and balances, we need methods of undoing such mishaps. Andrew🐉(talk) 20:25, 23 October 2025 (UTC)

See Lourdes for an example where that actually happened. QuicoleJR (talk) 17:55, 25 October 2025 (UTC)

I'm quite familiar with Lourdes as I was the first person to oppose their RfA. But I was almost alone as the final score in that RfA was 207/3/1. This demonstrates that the community is far too trusting and gullible at RfA and so provides little protection against bad actors. The recall option is therefore needed to help correct such mistakes. Andrew🐉(talk) 22:25, 25 October 2025 (UTC)

"Trust" and "gullibility" of course being swords that can cut both ways. FOARP (talk) 10:57, 27 October 2025 (UTC)

Indeed – "Live by the sword, die by the sword". Andrew🐉(talk) 23:00, 27 October 2025 (UTC)

If we're considering well-funded bad actors, then personally I think the more fruitful and direct path would be to build up a farm of editors making good edits, establishing their reputation. They could then gradually participate in discussions to make decisions that subtly favoured their employer. The real way to control content is to outnumber editors opposed to the employer's viewpoint; admins have no ability to override what is decided by the self-selected portion of the community that shows up to a discussion. isaacl (talk) 23:36, 23 October 2025 (UTC)

While I recognize that BD2412's point is valid, I would expect these admins to be likely enough to pass a RRFA. I've supported Admin Recall for a long time, and survived one attempt to recall me under my chosen criteria. I also almost lost my bits in an Arb case, resigned them anyway, failed RFA3, and then succeeded with RFA4. I still think A is the right choice at this point. SarekOfVulcan (talk) 19:55, 23 October 2025 (UTC)

"I would expect these admins to be likely enough to pass a RRFA - Since thus far no-one has passed RRFA, nor has even one full RRFA been held, what is this based on?

You're thinking an RRFA would just be the same as a normal RFA, instead it's more like when FRAM tried to get his bit back after getting desysop'd by the foundation: lots of people are going to say "no smoke without fire" and either oppose or stay neutral. FOARP (talk) 07:56, 24 October 2025 (UTC)

I was speaking particularly towards recalls initiated by coordinated bad actors. With Fram, there was an element of "well, it's the foundation" that wouldn't apply in this case. SarekOfVulcan (talk) 13:56, 24 October 2025 (UTC)

See because I agree with Fram that the intent of this process was, clearly, to have a mechanism to desysop admins is precisely why I think there should be a mechanism for people who think the admins are doing good work to argue against that desysoping. Right now what we're effectively saying is that admins need to avoid upsetting 25 people if they want to keep on being admins. I'm concerned this creates a set of perverse incentives. Simonm223 (talk) 14:02, 24 October 2025 (UTC)

There is a mechanism, and it's the RRFA, with a built-in lower pass level than for regular RFAs. SarekOfVulcan (talk) 14:07, 24 October 2025 (UTC)

But given that ReRFAs are not happening something is clearly wrong. Reasonable minds can (and evidently do) disagree on what that something is, but it is unarguable that the process is not working as intended. Thryduulf (talk) 14:26, 24 October 2025 (UTC)

But given that ReRFAs are not happening something is clearly wrong: What if they realize they are never going to pass for the reasons they were desysopped and give up? How is that wrong in any way? Wikipedia:Requests for adminship/Worm That Turned 2 shows that passing these is technically possible if you are in good standing. And mind you, WTT is an arb and has participated in several contentious cases. Children Will Listen (🐄 talk, 🫘 contribs) 14:37, 24 October 2025 (UTC)

Perhaps recall is not being deployed as widely as envisioned when designing a safeguard against frivolity, and the 25 editors who might feel WTT is not good admin material in a non-frivolous manner are not creating recalls for whatever reason. CMD (talk) 14:48, 24 October 2025 (UTC)

I don't think "unarguable" means what you think it means. Fram (talk) 14:43, 24 October 2025 (UTC)

Start a recall for me, and I commit now to running a full RRFA without withdrawing early. SarekOfVulcan (talk) 16:41, 24 October 2025 (UTC)

@Thryduulf Either is means something is wrong with RRfA, or it means something is right with the recall process and the only successful petitions have been against admins who know there's a good reason they wouldn't pass RRfA. --Ahecht (TALK
PAGE) 19:27, 27 October 2025 (UTC)

If there had only been one or two petitions and they had been the result of unambiguously and egregiously bad behaviour (not quite the right term, but it'll do) then you might have a point, but we've had only one ReRFA despite more than one of those being about things where there wasn't even a clear consensus among those commenting on the petition that the complained-about (in)action(s) justified a desysopping. That's not evidence of a process that is working correctly. Thryduulf (talk) 22:29, 27 October 2025 (UTC)

More Specific Proposals

The discussion above has done a good job at taking the temperature of the community on recall and generating ideas for proposals, but I don't think it is structured in a way that is good at actually measuring the consensus for/against each of them. I'd like to break out the proposals that seemed to gain traction, and give them a space for refinement, and eventually a space to see if we have consensus.

Note: Please treat this section as an idea lab/RFCbefore - I'm not looking for bolded !votes yet, I'm trying to make sure that the right ideas are put in front of the community properly in their best possible form.

I see 3 ideas that should probably get an up or down discussion:

• Mz7's proposal at User:Mz7/How to fix administrator recall
• Reducing the time a petition stays open from 30 days to 14 days (input helpful on number to poll)
• Raise the number of signatures needed for a petition to succeed from 25 to 40 (input helpful on specific number)

There are 3 other proposals that could be discussed, but got weaker receptions

• Allow an admin to start a RRFA or stand in an election anytime within 6 months of a petition closing. Administrator access would still be removed after 30 days.
• Limiting petitions signers to 1 petition every 6 months
• Having a slowdown at the start, either preventing signatures for the first day or so, or requiring all petitions to be open for the first couple days or so.

Any thoughts about which ideas should be polled, or the specifics of any idea, or the mechanics of this (can I just add a section on the bottom and ping everyone when we're ready to go?) would be welcome. Tazerdadog (talk) 17:19, 23 October 2025 (UTC)

I think that if proposals depend on each other, they should be put together as one proposal instead of having a case where one sub-proposal passes, but the other fails. My preference is multiple polls, each with the idea and a support/oppose-discussion, such that they are all independent proposals on their own merits.

Inactivity is one of the things a lot of editors have decently strong opinions on, I just don't know if it'll be appropriate handling them in same space of proposals. I had planned to flesh out "Inactivity thresholds require 1 admin action within the last 12 months" as a proposal to replace WP:INACTIVITY, as an ideaL. Soni (talk) 17:26, 23 October 2025 (UTC)

Yep, I was going to structure it as 3-6 sections with a support/oppose/discussion. Are any of these proposals dependent on each other/in need of a merge? Tazerdadog (talk) 17:33, 23 October 2025 (UTC)

I would think that the idea of having opponents of the petition play a role in the first round got sufficient support that perhaps something along those lines should be included. Wehwalt (talk) 19:55, 23 October 2025 (UTC)

No, because then you have to have the discussion twice, instead of a short vote on whether there needs to be a discussion followed by the actual full discussion. SarekOfVulcan (talk) 20:00, 23 October 2025 (UTC)

It did get a fair amount of support. How would you word a proposal in a way that puts the idea forward in it's strongest form? I think it needs to have a lot of specifics hammered out (which to be fair is exactly why this section exists). Tazerdadog (talk) 20:34, 23 October 2025 (UTC)

Probably something along the lines that if the admin can show at least as many supporters as the number that signed the petition, that the petition does not advance. If proponents of recall can't even get a majority of those expressing a preference, then probably a substantial number of people are going to support the admin in the second round. Obvious timeframe details to be worked out, I'm just speaking generally. Wehwalt (talk) 16:38, 24 October 2025 (UTC)

How about a proposal to leave the recall system untouched? AugusteBlanqui (talk) 18:12, 23 October 2025 (UTC)

"Extended confirmed" participation is not enough. We know that there are malicious accounts that game the system to achieve that status, undoubtedly already more than 25 or 40. We need a better barometer of account trustworthiness in this regard than having made a still relatively small number of edits in a relatively short period of time, which any reasonably clever bot could do. BD2412 T 19:33, 23 October 2025 (UTC)

I understand where you're coming from here, @BD2412; do you have any suggestions for what sort of benchmark that may be? Staraction (talk | contribs) 00:31, 26 October 2025 (UTC)

@Staraction: Some number of administrators in the mix would be good, perhaps a minimum of five. I know that this will draw the objection that admins will cover for their own, but a quick perusal of WP:ANI archives will demonstrate that this really does not happen, certainly not with any kind of uniformity. I would require all participating editors to have at least a year on Wikipedia (thirty days is basically a waiting period, not a useful measure of participation). An admin subject to the process (or any other editor) should be able to challenge the validity of participation of editors who do appear to have gamed the system to reach an EC number of edits. Finally, to prevent "revenge" voting, I would disqualify from participation any editor who had been subject to an unequivocally good block by the admin sought to be recalled. Admins need to be able to make clean blocks for edit warring and other actual misbehavior without concern that such activity alone will result in votes to recall that admin. BD2412 T 00:47, 26 October 2025 (UTC)

One thing I suggested in a previous discussion (I don't remember where) was requiring that all signatories (except the initiator?) be uninvolved with regards to the subject of a petition. Something like, if you were an arbitrator would you have to recuse from a case about the subject? If yes, they you cannot sign the petition. I don't recall the idea getting much discussion last time beyond a few knee-jerk opposes to the idea of excluding anyone from signing a petition. Thryduulf (talk) 12:06, 26 October 2025 (UTC)

Looking at statistics on previous recall-related RfCs, there is absolutely a phenomenon where admins tend to be much more opposed to recall than non-admins. This would turn recall from a (flawed) way for the community to keep admins accountable, to a way for admins to protect their own, while non-admins can't do anything about it. Plus, regarding the disqualification of editors subject to an "unequivocally good block", who decides what blocks count as good? There is a genuine risk that this might incentivize "revenge blocking" on technicalities to prevent one's opponents from having a clean block log. Chaotic Enby (talk · contribs) 05:36, 13 November 2025 (UTC)

I said this in my !vote earlier, but I think that something along the lines of "Having a slowdown at the start, either preventing signatures for the first day or so, or requiring all petitions to be open for the first couple days or so" would be a very good idea. Personally, the way I see it, having a period of a few days at the start, during which only the initiator of the petition could sign, but all editors can discuss the situation, before the signing would begin in earnest, would transform the process into a genuine discussion, as opposed to a petition which can start and end in less than a day. I think that would be all-upside, with no downside. Even those editors whose opinions have been Option A would not be "giving anything up", except maybe some ill-founded sense of urgency. --Tryptofish (talk) 21:46, 23 October 2025 (UTC)

• I wonder if there is any possibility of giving bureaucrats some role as a check on the democratic excesses of this process. For instance, bureaucrats could hold a 'cratchat' once the petition is closed to determine whether any extenuating circumstances require further consideration. Perhaps they could refer such cases to ArbCom or some other community process for further examination. Yours, &c. RGloucester — ☎ 00:31, 24 October 2025 (UTC)

  The crats didn't volunteer for that, so we should ask them before putting it in a formal proposal.—S Marshall T/C 07:33, 24 October 2025 (UTC)

  The process already involves the bureaucrats specifically and gives them some discretion. WP:RRfA includes:

  "The bureaucrats may remove adminship if an admin does not run in the allotted time; they may grant slight extensions on a case-by-case basis."

  "If the administrator receives between 50 and 60% support, the community's consensus will be determined by the bureaucrats."

  "Should the administrator fail to pass an RRfA or administrator election, bureaucrats may remove their privileges."

  Andrew🐉(talk) 07:51, 24 October 2025 (UTC)

The process that I am referring to is the petition, not RRfA, which might as well not exist. Bureaucrats could be given latitude to pause petitions or otherwise refer to them to ArbCom for review. Some kind of certification process to hold the 25 editors accountable is necessary, especially considering that there may be objections lodged by other parties that require further examination. There is no other situation where 25 votes are sufficient to override any and all other opposition. Yours, &c. RGloucester — ☎ 08:06, 24 October 2025 (UTC)

The bureaucrats provide the teeth in the current process. Notice that the key clauses use the word "may" not "must". If they thought that a petition was suspect or inadequate then they could just sit on their hands and decline to act. Per WP:NOTBURO, it's not a mechanical process. See also WP:IAR which happens to be under discussion elsewhere. Andrew🐉(talk) 08:55, 24 October 2025 (UTC)

What do you mean by it might as well not exist? GothicGolem29 (talk) 16:25, 26 October 2025 (UTC)

This would go against the concept of recall itself. Recall is supposed to be a check by the community on administrators, and giving a smaller group of administrators the power to decide whether the community should even be listened to goes against the whole idea. Chaotic Enby (talk · contribs) 05:29, 13 November 2025 (UTC)

• The risk with some proposals to address certain concerns is that they exacerbate other concerns. For example, there are views above that RRfAs are not started because it is felt 25 opposes is a significant hill to climb. Increasing signatures to 40 near doubles that, so it's trading off certainty of the petition with viability of the RRfA. However, the "anytime within 6 months" seems to not raise such a consideration. People will have different periods of reflection, and some may be longer than 30 days (especially if that 30 days includes not only reflection, but also preparation for the questions etc.). I trust bureaucrats to be able to notice keep track of any different passing percentage requirements. CMD (talk) 01:57, 24 October 2025 (UTC)
• I already shared by thoughts about the essay here. I'm not opposed, nor do I think it would change a great deal of the outcome by implementing such an obvious improvement. CNC (talk) 18:15, 24 October 2025 (UTC)
• Regarding Mz7's proposal, I agree with some aspects such as getting rid of recall immunity and converting the system to an automatic "request for de-adminship" (although I believe 50% of opposition is already enough to show a lack of community trust).
  However, the hard minimum requirement of evidence requirement seems unworkable, as it asks for a closed discussion where the consensus of that discussion was that the administrator engaged in conduct below the standards expected of administrators. The issue is that the vast majority of these discussions are either not formally closed, or closed with a non-committal decision such as "no consensus" or (ironically) "take it to RECALL", which would make them non-starters. Plus, requiring an explicit affirmative consensus prior to a petition even being started is a much higher threshold than what the petition itself is for (finding probable cause), and as high as the RRfA/RfDA's own threshold.
  In my opinion, a better option for the requirement would be an attempt at dispute resolution where multiple uninvolved participants agreed that there was a conduct issue, without requiring it to have been closed with affirmative consensus. Chaotic Enby (talk · contribs) 05:44, 13 November 2025 (UTC)