Wikipedia talk:Administrator recall

Other discussions

• Wikipedia:Requests for de-adminship – index of previous unsuccessful proposals
• Wikipedia:Administrator recall/RfCs - index of RFCs and unassorted discussions related to this process

Questions

• The bureaucrats are responsible for ensuring that an RRfA is started within a reasonable time frame. If this does not happen, they may remove the administrator privileges at their discretion. Should the administrator fail to pass an RRfA or administrator election, bureaucrats may remove their privileges. - should we instead say The bureaucrats are responsible for ensuring that an RRfA is started within a reasonable time frame. If this does not happen, they should remove the administrator privileges. Should the administrator fail to pass an RRfA or administrator election, bureaucrats will remove their privileges.
• Does it include a power to revoke crat/CU/OS right? Otherwise an admin that is recalled will keep these rights - and what about a standing ArbCom member be recalled (which can happen one year after election)?

GZWDer (talk) 16:48, 6 June 2025 (UTC)

I can't remember in which of the many discussions it came up, but the consensus was very clearly for "may remove" not "should remove" - the crats have discretion to act in the best interests of the project. Normally that will be removing admin privs from someone who has not passed a re-RFA or election but exceptional circumstances can happen. One example given was that if they have signed up to participate in an admin election that happens shortly after the 30 days is over.

Checkusers and Oversighters serve at the discretion of the Arbitration Committee, only they can remove those privs. If someone with those rights is desysopped then it is almost certain that the committee will review their suitability for the position(s) and remove them if they feel it is appropriate.

If a sitting arbcom member is recalled and fails to pass a re-RFA then they will continue as an arbitrator unless they resign or the provisions at WP:ARBCOND are applied. Thryduulf (talk) 18:02, 6 June 2025 (UTC)

As per Wikipedia:Bureaucrat § Removal of permissions, administrators whose admin privileges are removed through the recall process also have their bureaucrat privileges removed.

Bureaucrats are trusted to understand the background of the process and are given discretion around the timing of the removal of administrative privileges, with the understanding that this does not give them discretion to disregard community consensus in a recall petition or re-request for adminiship. isaacl (talk) 18:15, 6 June 2025 (UTC)

administrators whose admin privileges are removed through the recall process also have their bureaucrat privileges removed: I wish that was the case, but I read "If an administrator has failed to pass a required recall" as applying only to the +sysop right, particularly given this discussion. Extraordinary Writ (talk) 20:08, 6 June 2025 (UTC)

My apologies; I accidentally misinterpreted the heading "Removal of permissions" and skimmed the section too quickly. (I was also misremembering the details of Wikipedia:Village pump (policy)/Archive 181#Alternate proposal (Procedural community desysop).) isaacl (talk) 21:48, 6 June 2025 (UTC)

No. The process is not looking for a consensus (certainly not from the community), it is looking for a factional vote, and is decided by a factional vote, based on any or no rationale -- Alanscottwalker (talk) 10:40, 13 June 2025 (UTC)

My understanding is that Arbcom members, CU and OS can only be removed by their colleagues and the Arbcom respectively. CU and OS folks probably shouldn't be necessarily be directly answerable to the community primarily cause their work by definition includes non-public information. Sohom (talk) 21:46, 6 June 2025 (UTC)

Speaking as an Oversighter, I'm answerable to the community for the actions I take as an editor and as an administrator. For my actions relating to suppression, I'm answerable only to my colleagues, the Arbitration Committee and the m:Ombuds commission because they are the only ones able to view all the relevant evidence. Thryduulf (talk) 23:17, 6 June 2025 (UTC)

My understanding with committee members would be that they would just lose the admin rights, but retain their membership. Granted, I would presume that it would never get that far. An incident where the community felt the need to recall a committee member probably would end up resolved by action from the group. Without going into excessive details, the committee has disciplined its own members in the past and I don't see a situation where this would be an issue in the future. --Super Goku V (talk) 04:51, 7 June 2025 (UTC)

Regarding the first question above, I would support GZWDer's proposal to change "may remove their privileges" to "will remove their privileges" if the administrator "fails to pass an RRfA or administrator election". The case that Thryduulf is referring to, when the RRfA or election has not yet been held, is different. But if the RRfA or election has been held and the admin failed to pass, the expectation is that bureaucrats will remove the privileges. Tim Smith (talk) 04:10, 8 June 2025 (UTC)

I have come here directly from AN (related to the Bbb23 admin recall case). I think there should be some clause where an admin can automatically have their rights revoked, without a) the 30-day time period (if the matter is of a serious nature - I would leave it to the community to discuss where that line should be drawn) and b) without the case going to the Arbitration Committee, as cases can take a while there, and the process is "expensive" (that essay is on SPI, but it is easy to draw parallels). My concern is that, within that 30-day time period, there is enough time for an administrator who has passed recall to continue carrying out administrative actions that may have been previously condemned by the community (or Arbcom, where that applies). Patient Zero^talk 22:16, 8 June 2025 (UTC)

@Patient Zero, any other admin can block an admin if their behavior is a serious problem. Valereee (talk) 22:25, 8 June 2025 (UTC)

That’s fair Valereee, but I do suppose that in cases where the issues are purely caused by the administrative tools, immediate revocation of said tools is arguably better than a block, as a block could be argued to be punitive. It would have to be dealt with on a case-by-case basis but what could be proposed is a sort of flowchart system to this procedure (which could be diverted from, of course, per IAR if felt necessary). Patient Zero^talk 22:43, 8 June 2025 (UTC)

A block for misuse of tools isn't generally considered punitive but preventative. I'd do it without hesitation if I thought there was a crisis. There are hundreds of active admins, only a handful of bureaucrats. Valereee (talk) 22:58, 8 June 2025 (UTC)

No, you're absolutely right there Valereee. I feel reassured now that there are adequate procedures in place for if this were to ever happen - if anything, my coming here was more over fear of a potential loophole being created, but you've done well to reassure me. I appreciate it. Patient Zero^talk 23:17, 8 June 2025 (UTC)

The Arbitation Committee has the power to remove the admin bit if the circumstances require it. Donald Albury 23:27, 8 June 2025 (UTC)

They absolutely do Donald Albury, but Arbcom cases can take a while to get going at times, and rely on a system where arbs have to support/oppose/recuse on points of fact, and that of course influences the final decision made. My proposal would hopefully "cut out the middle man" in that regard, but once again, I am now confident (owing to Valereee's response) that current procedure surrounding administrators willing to carry out difficult blocks, who would be willing to step in when required, is sufficient. Patient Zero^talk 23:32, 8 June 2025 (UTC)

Level 1 desysops are typically pretty fast, Level 2 desysops are slower, but are also a option as well in these kinds of cases where time is of essence. Obviously if the admin account is compromised, (or intentionally causing harm) they can always just be blocked or locked. Sohom (talk) 23:42, 8 June 2025 (UTC)

ArbCom can desysop by motion, and have done so in very short time when necessary. There is no need for any new process. Donald Albury 23:42, 8 June 2025 (UTC)

Of course - you're absolutely right, and I feel rather foolish for forgetting this is a possibility. This is also reassuring - thank you Donald Albury. Patient Zero^talk 23:49, 8 June 2025 (UTC)

Just for data, the first level 1 I could find in my inbox happened in 14 minutes. They can happen very quickly (and obviously complement a block and/or lock). Best, Barkeep49 (talk) 21:48, 16 June 2025 (UTC)

Interesting observation

In past discussions, there has been pushback against editors (including me) who expressed the opinion that the current version of the recall process moves too quickly, leaving too little time for discussion and reflection until a potential subsequent re-RfA. The pushback has rested on the arguments that the process has worked well, as intended, and that nobody has presented any evidence of harm resulting from recalls.

Today, I made this comment at Wikipedia:Arbitration/Requests/Case: [1]. In response, a member of ArbCom, Liz, observed this: [2]. I suppose that, strictly speaking, this is a matter of overall observation rather than hard evidence, and the (accurate, I think) characterization as a "pendulum swing" may indicate some beneficial aspects, but I regard it as significant in terms of a potential unintended impact of recall that may be causing some harm. --Tryptofish (talk) 20:21, 16 June 2025 (UTC)

Which I guess is not really surprising. Ymblanter (talk) 21:11, 16 June 2025 (UTC)

For better or worse, establishing consensus requires patience. Among editors who like to discuss Wikipedia process, there is a sizeable number who highly value using a consensus-like, large-scale group conversation to make decisions. Often a shepherd would be helpful to guide discussion, but that requires co-operation from the participants, which is generally hard to sustain on English Wikipedia.

It's not clear to me that Liz's supposition on admin recall is a significant factor. (For instance, I don't think any of the admins in the arbitration enforcement discussion thread in question were concerned about a recall petition.) Rather, I have seen a trend for many decisions to get appealed. This makes the task of evaluating consensus even more thankless, as those who disagree with your evaluation start complaining about your reasoning and lack of detail in your closing statement. It wasn't so long ago where closing statements weren't expected to provide a summary of everything, but now evaluators are getting excoriated for not mentioning every single viewpoint. In this environment, it's to be expected for multiple people to cross-check their evaluation of a situation, looking for what generates the most overall agreement. That can be a good thing for getting a good read on community desires, but is inevitably slower and so less time-efficient. isaacl (talk) 21:14, 16 June 2025 (UTC)

Yes, I can agree that this isn't a matter of "all-good" or "all-bad". But, just as the community likes discussion and consensus in dealing with various disputes, I think the community should value discussion and consensus in determining whether or not an admin should be required to have a new RfA in order to retain the permissions. I'm in favor of the move over the last decade, to expecting less shoot-from-the-hip attitudes from admins, but I also think we should seek to calibrate this, so it does not go too far in the direction of inaction in the face of obvious problems. We may be starting to see a trend towards admins being reluctant to take timely action when it is needed, and, if I'm right about that, then we might want to make the petition process less tilted towards such rapid conclusions. --Tryptofish (talk) 21:31, 16 June 2025 (UTC)

I found Liz's observation an interesting one. To the extent that it's true I think there would be a natural second part: not only would admins be less willing to be BOLD and to see more consensus ahead of acting but they might, on the margin, be less willing to get involved at all. In general it takes a special kind of editor who is willing to wade into the hardest areas knowing the cost of doing so and increased threat of community criticism could change that equation. Fewer bad admin actions is certainly one of the goals of recall so this all of this could still be a net positive, as liz's comments itself notes. But if there's a sense that it is true, it might lead to some ways to think about how we might lessen the negative side of that net positive calculation. Best, Barkeep49 (talk) 21:44, 16 June 2025 (UTC)

In general, I think the increasing willingness of the community to escalate decisions up through multiple levels of consensus discussion (there's a base process, then a review process, then a general appeal to the community process; the recall process is in essence a review process for the RfA process) is creating more and more expectations on the steps admins should include in their workflows. I don't know how to reverse course, though, without moving away from trying to decide everything by large-scale, unmoderated group discussions. isaacl (talk) 01:58, 17 June 2025 (UTC)

The pushback has rested on the fact that recalls been preceded by discussion, both in the days prior and going back years, and that your opinion is not based on the facts. Thebiguglyalien (talk) 🛸 21:41, 16 June 2025 (UTC)

I don't think it is particularly likely that any of the admins in that discussion were worried about recall. -- asilvering (talk) 21:45, 16 June 2025 (UTC)

Well, if my "opinion is not based on the facts", then neither was the claim that previous discussions demonstrated that the time the petitions were open was an optimal amount of time. And for that matter, we don't have a factual basis to say that the AE admins weren't thinking about recall. Remember, the case that I had commented on had a strong element of WP:UNBLOCKABLES. --Tryptofish (talk) 21:50, 16 June 2025 (UTC)

@Tryptofish, I assume you meant this as a reply to TBUA? -- asilvering (talk) 22:09, 16 June 2025 (UTC)

It was mostly a reply to them, but I also took into account your assertion that the AE admins were not thinking about recall. --Tryptofish (talk) 22:26, 16 June 2025 (UTC)

I certainly wasn't (I can't even be recalled for another few months). But I don't think that necesarily diminishes the observation which is that there are a number of factors, including recall, which push admin to act collectively and slowly rather than unilaterally and quickly and RECALL may be one such reason. This could be of net positive to the community while remaining an interesting observation about the possible impact of RECALL. Best, Barkeep49 (talk) 21:51, 16 June 2025 (UTC)

I have been twice taken to ArbCom for a single admin action. In the past, I was succeptible to the calls to close a particularly difficult RfC or discussion, and did it more than a few times, sometimes alone, sometimes a part of the group. I knew that even if I am taken to ArbCom, I would lose a lot of time, but they would at least listen to my arguments (which indeed happened both times). Now I know that if instead I am taken to a recall discussion, and we know that so far all recall discussions have been successful, and at this stage I will get 25 votes does not matter what - well, I will think several times before closing anything contentious. Mind you, I am in the top 50 most active admins of all times, unless I am missing something. Ymblanter (talk) 22:05, 16 June 2025 (UTC)

Hmm. Building on this comment and my comment, and in the sprit of no bad ideas when brainstorming (because this might be a bad idea), I wonder if the community would explicitly allow voluntarily re-RFAs once a year. This would give the user feedback on how they're doing (a good thing having sought some myself) and could give them six months of "peace of mind" knowing that they could only be removed by ArbCom; that still would leave six months of "vulnerability" but perhaps the peace of mind time could be extended to 9 or 12 months. Best, Barkeep49 (talk) 22:12, 16 June 2025 (UTC)

In the spirit of "the more RFAs the better, since that helps take some of the 'main character' heat off the people running", that sounds like an easy positive. But I have to admit I'm surprised to hear it suggested by someone who took issue with WTT's recent run. -- asilvering (talk) 22:20, 16 June 2025 (UTC)

If I recall correctly, there was some concerns raised during the recall discussions about admins running re-requests for adminship in order to gain an exemption from recall petitions. It was a limited conversation, though, so I don't know what the community overall would think. Historically, it has tolerated the occasional re-request, but with indications that it didn't really want re-requests to be a regular occurrence, due to the time demands. isaacl (talk) 01:39, 17 June 2025 (UTC)

While I can't remember the name of the rerunning admin, if I recall correctly the most recent couple of Re-RfAs received opposes for wasting community time. CMD (talk) 01:46, 17 June 2025 (UTC)

The last one I recall was Wikipedia:Requests for adminship/Worm That Turned 2 Hawkeye7 (discuss) 02:11, 17 June 2025 (UTC)

In December 2024/January 2025 (after WTT's RFA) there was an RFC at Wikipedia:Village pump (policy)/Archive 200#RfC: Voluntary RfA after resignation that came to a clear consensus that the status quo of "former administrators who would be eligible to request restoration via BN may instead request restoration of their tools via a voluntary request for adminship (RfA)." should be maintained. Thryduulf (talk) 02:13, 17 June 2025 (UTC)

Well the Re-RfAs did pass, just over the minority. Thanks Hawkeye7 for the link, inactivity was the main topic of discussion there. I think the community would be unwilling to reject a "may" option, the trouble is always the parameters of that "may". CMD (talk) 02:57, 17 June 2025 (UTC)

Wikipedia:Requests for adminship/Hog Farm 2 was the one I remembered. User:Hog Farm is a great contributor and a long time MilHist Coordinator, but IMHO he was treated at RFA2 as if he was asking the community a favor. After voluntarily surrendering his tools because of RL issues, Hog Farm could have chosen to ask at BN but he chose to run as a re-RFA. I was fairly shocked at the negative response this run got, especially since he's one of our finest wikipedians. IMHO the response had a chilling effect on other admins who were under the impression that re-RFAs were inside anyone's reasonable discretion. I'm glad the community kept the SQ. BusterD (talk) 17:36, 17 June 2025 (UTC)

Since I've been pinged - what frustrates me is the obnoxious environment we've created for formerly active admins who expect to have some more time and want to return. If you redo RFA, then a significant minority argues that you're a diva, but if you take the BN route and you're prediction of having more time turns out to be wrong, then you might get hauled to Recall like Gimmetrow was. My choice to re-RFA wasn't based on the threat of recall (more along the lines of "how can I call myself an uncontroversial candidate when I'm pledging to stay away from a whole wikiproject?") but I think some of the recent recalls have highlighted the Catch-22 here. Hog Farm _Talk 00:48, 18 June 2025 (UTC)

Gimmetrow had at least two opportunities to voluntarily give up their tools again once it became clear they didn't have the expected time. For whatever reason they didn't do so and so we ended up where we ended up. Nil Einne (talk) 07:26, 26 June 2025 (UTC)

I wonder if the community would explicitly allow voluntarily re-RFAs once a year. I think others have shared the exact links as well, but there already is consensus to explicitly allow voluntary Re-RFAs anytime the admin chooses. It just isn't necessarily something on most admins' minds. Soni (talk) 04:17, 17 June 2025 (UTC)

Threatening an admin with recall petition

I came across something today (twice) where an admin was directly threatened (by two users) with recall. In the case I saw, there was no history of bad admin actions. It was a difference of opinion about policy and common practice. What we had was an admin with almost 20 years of service being threatened because of others' disagreement with their interpretation of policy (and how the admin phrased it). A philosophical difference.

Admins are allowed to make mistakes; it's literally written into every arbcom ruling on the subject. Similarly admins are allowed to vary in their good faith understanding with our policies and guidelines. None of this is controversial. We have systems for dealing with good faith disputes.

Even saying "I'm going to take you to a noticeboard" or "I'm going to take you to arbcom" are threats, but I believe both are in-bounds because the resolution at those boards takes time and is generally well-attended. Merely saying "I'm going to start a recall petition" or "if a recall petition was on the table I'd vote for it" is quite a different approach. In this situation recall looks more like a guillotine and less like discussion (so far).

In the abstract, how would other admins feel about such threats made against them (compared to threats of ANI or ARBCOM)? BusterD (talk) 18:02, 17 June 2025 (UTC)

I wrote this already in the above thread. If I am taken to Arbcom, I am confident they would listen to me. If I am taken to recall, I am sure I will get 25 votes pretty soon (I guess I can even name 25 editors who would sign), and then I would need to decide whether I want to go through the RfA (likely not). On the other hand, a threat or not a threat, there is nothing I can do about it. Anybody can take me to recall any time. Ymblanter (talk) 18:48, 17 June 2025 (UTC)

Having been taken to ArbCom, I feel that your confidence is misplaced. Hawkeye7 (discuss) 19:25, 17 June 2025 (UTC)

Having watched (with horror) Hawkeye7 being taken to ArbCom, he knows something about it most of us don't yet know. Frankly, I'm glad Hawkeye7 isn't burdened with admin responsibilities. He's the only lead coordinator NOT to have held the privs during their terms, and he takes advantage of the available time in a way the others couldn't. He creates and reviews pagespace. Wikipedia is better for it, with all due respect to Hawkeye7's fine abilities and vast experience. BusterD (talk) 23:55, 17 June 2025 (UTC)

For me "I will take you to ANI" is less of a threat than "I will take you to ArbCom", but the ArbCom threat is roughly equivalent to "I will start a recall petition against you" for me. In some circumstances such any of the three statements would to me be blockable (frivolous noticeboard threatening could be a sign of NOTHERE, to give an example Buster declared in bounds), while in others they would not. And I say that agreeing that they are a kind of threat in all three cases. The reason I don't agree with recall being different is that it was meant to be a barrier to the discussion (the discussion being the Re-RFA) and to provide a mechanism so the admin has some control over when the Re-RFA happens. Now I get your argument it isn't working that way in practice and I think this step could probably use some reform. But I do think it important to note that the one re-RFA we've had did have meaningful discussion even if it resulted in the admin giving up the tools. Best, Barkeep49 (talk) 19:39, 17 June 2025 (UTC)

You could argue that, but in what IMO was the one "bad" recall (Fastily) that we've had so far, no re-RFA took place, because the admin said GFY and hasn't edited since - and I don't blame them. Black Kite (talk) 20:29, 17 June 2025 (UTC)

I agree never getting to a discussion phase because admins don't go to re-RFA is an issue and is why I suggest some reform. But that is the answer to a different question than whether "I am going to start a recall against you" is a sanctionable threat which is the reasonable question Buster has posed. Best, Barkeep49 (talk) 20:48, 17 June 2025 (UTC)

I mean, I blame Fastily for repeatedly and consistently abusing editors both new and old and using their admin tools to actively harm other editors in a blatant violation of WP:ADMINACCT spanning years. But I know we don't share that opinion. Silverseren^C 21:26, 17 June 2025 (UTC)

No, we don't agree and that's fine, but my bigger issue with this one was the 11th signature (from a_smart_kitten) was long and quite damning, and at least half a dozen other editors signed the petition "per a_smart_kitten" - only for the whacking long list of supposed misdeeds then to be found not to be anywhere as damning as had been made out. Black Kite (talk) 10:50, 18 June 2025 (UTC)

Out of curiosity, where were the discussions where an admin was threatened with recall? PackMecEng (talk) 21:48, 17 June 2025 (UTC)

Because I'm asking in the abstract, I'm going to decline that query for now. Not to obscure anything, but to ask the question independent of any association. Bbb23's blocking issues went remarkably undiscussed prior to the recall itself (and now it's a community emergency we rehabilitate those accounts). I'd have expected multiple chances to make my case (and demonstrate improvements) in noticeboard threads first if I were to be summarily dismissed after 20 years of loyal dedicated service. I'd hope editors like the ones in this thread would tell me I'd done another editor violence, so to speak. If I didn't make an ideal hypothetical, I'm open to reasonable chiding from Barkeep49 and all others. I don't have the answers myself, and I'm not sure I should know, but when I saw the two threats today it had me reading a boatload of associated pages. I honestly want to know what others think of the mere threat of petition (since so far every started petition has 100% cost us a mop-swinger), distinct from the other ways cases often go. IMHO the recall process has no demonstrated protections against the appearance of impulsive, frivolous, involved, coordinated, or otherwise biased petitions against any of the subjects so far. BusterD (talk) 22:39, 17 June 2025 (UTC)

Further, I don't think we'd be allowed to treat an employee in this manner. How is it we can treat a volunteer in such an extraordinary way? BusterD (talk) 22:45, 17 June 2025 (UTC)

Bbb23 was warned many times over the years, and the conduct had been brought up repeatedly for discussion, but in the past it was too difficult to address admin misconduct and nothing ever came from it. Thebiguglyalien (talk) 🛸 23:04, 17 June 2025 (UTC)

As coincidence has it we have our first petition which didn't lead to dwadminahip (presuming that's what happens with Bbb) now. Best, Barkeep49 (talk) 23:25, 17 June 2025 (UTC)

Now I don't need to be coy. This was the situation I saw unfolding this morning. (Threat one by User:Mdewman6, threat two by User:SilverLocust) BusterD (talk) 23:35, 17 June 2025 (UTC)

My edit here at 18:02 preceded the new (now withdrawn) petition by two hours. BusterD (talk) 23:46, 17 June 2025 (UTC)

I should add now that they've both signed the petition, I'm relieved at least these weren't mere threats. It sounds wrong, but I'm glad they filed the petition (regardless of the outcome). I would have thought less of both editors if they wrote these words as threats but didn't act on them. It might have been better if they'd confronted User:Necrothesp on a public board first, but here we are. BusterD (talk) 00:09, 18 June 2025 (UTC)

A recalled admin whose attitude is GFY or who otherwise refuses to explain their actions or be accountable let alone make a commitment to improve is confirming that they shouldn't be am admin just as an editor who responds that way to a block would be seen by admins as showing that they shouldn't be an editor. Wellington Bay (talk) 21:34, 17 June 2025 (UTC)

I don't think that the most important issue here is whether or not such a threat should be sanctionable. The important issue, for me, is that it is possible to make such a threat and start a petition for insubstantial reasons, and still get enough signatures to force a re-RfA. (And the second most important issue is that there are too many members of the community who choose to wave this concern away.) --Tryptofish (talk) 21:53, 17 June 2025 (UTC)

I agree. The point of the petition process is to ensure that recall RFAs cannot be triggered frivolously or needlessly. That a petition can start and end in a matter of hours, with no controls against any signatories being INVOLVED or otherwise biased against the admin, indicate to me that it is not performing this function. The petition stage should be a surmountable hurdle, not a trip hazard. Thryduulf (talk) 22:30, 17 June 2025 (UTC)

Possible "fixes" if that problem is agreed to are: increasing suffrage requirement for signing the petition, increasing the number of participant required. I doubt a "minimum time" is going to help, as it only helps if the current participants are convinced to withdraw their support for the petition. — xaosflux ^Talk 22:40, 17 June 2025 (UTC)

Agree with each comment above. It's not the time period; it's the "ratchet" itself (to utilize Tamzin's descriptive noun). We have not calibrated this new tool properly, IMHO. I have made a similar (and similarly undefined) assertion before. BusterD (talk) 22:50, 17 June 2025 (UTC)

I don't think that's a complete list of the possible fixes, but that's another discussion. --Tryptofish (talk) 22:54, 17 June 2025 (UTC)

I have been mooting the idea of a 24-72 hour "waiting period" from opening of petition to being eligible to sign. I think the very reasons a discussion period didn't work at RfA (everyone signaled what they might do anyone) it might be useful here. Best, Barkeep49 (talk) 23:26, 17 June 2025 (UTC)

What do you expect would happen in that "waiting period" (other than hounding the petitioner to withdraw it)? — xaosflux ^Talk 13:07, 18 June 2025 (UTC)

I would expect discussion to occur. And that discussion might impact the petition itself - as it did yesterday. More importantly to me it might encourage admins to go to re-RFA rather than feeling like the deck is stacked against them so what's the point. Best, Barkeep49 (talk) 14:38, 18 June 2025 (UTC)

I wouldn't mind some sort of

1. (initiate) which is defacto the originator becoming signer #1
2. 24-hour hold for additional signatories, discussion is open
3. Petitioners may begin signing

process, provided that there is a expectation that hounding the initiator is considered unwelcome. — xaosflux ^Talk 14:52, 18 June 2025 (UTC)

Yes. That matches what I was thinking about. Best, Barkeep49 (talk) 14:54, 18 June 2025 (UTC)

Hounding is pointless, because the petitioner cannot withdraw the petition. There is no provision in Wikipedia:Administrator recall for that, and it is not allowed for that very reason. The only way a petition gets closed is after 30 days, at the subject's request, or when it reaches the 25 signature threshold. Hawkeye7 (discuss) 19:50, 18 June 2025 (UTC)

The most recent petition was withdrawn. See Wikipedia:Administrator recall/Necrothesp. Wellington Bay (talk) 20:45, 18 June 2025 (UTC)

There was some discussion last year on an initiator being able to withdraw a petition, though a question on it didn't make it into the RfC being drafted. As I said in November 2024, allowing an initiator to withdraw a petition that has no other supporters is probably the easiest way to quickly resolve the petition. isaacl (talk) 22:43, 18 June 2025 (UTC)

It seems like if a petition with NO support can be closed early is being debated some, but I can't see any reason the petitioner can't withdraw their own personal support for the petition. I certainly agree that a petition with any supporters on it would not be able to be closed early. And a petition where the only author is the initial petitioner could potentially be CSD'd. — xaosflux ^Talk 13:05, 19 June 2025 (UTC)

IIRC it was proposed that a petition could be withdrawn if it were eligible for G7 speedy deletion (and maybe G6 too?) even if it wasn't actually deleted. Thryduulf (talk) 13:11, 19 June 2025 (UTC)

The biggest issue I see with early closing/deleting/etc is: does this grant the six-month recall immunity? That protection is a good reason to not allow early closures except in odd situations where the initial petitioner admits to being confused about the process (as opposed to just rethinking it). — xaosflux ^Talk 13:55, 19 June 2025 (UTC)

I do not recall any discussion based on the speedy deletion criteria, just the one to which I linked that you started. I agree with Xaosflux that it would be good to avoid providing an incentive for spurious petitions being started deliberately. That being said, an arbitration request can always be made at any time. isaacl (talk) 16:38, 19 June 2025 (UTC)

Some of the petitions have gone fairly quickly, but I think the community has pretty much shown the restraint we'd expect from them. Fastily's petition was stuck in neutral for days and might not have made it up to 25 if they hadn't gone and done exactly the thing they were accused of yet again. Lots of signatories on the Bbb petition waited until after he had a chance to make a statement, and when he did – and said statement ruined any hope of him actually taking responsibility for his actions – that was what sealed the deal. Necrothesp's petition received immediate pushback and failed. So far, in my opinion, all six recall petitions reached the result they should have (i didn't vote on Graham's RRfA and I'm still neutral, but the reasons for the petition were valid and I'm glad they still went on to RRfA; for what it's worth, i have high hopes on them returning). theleekycauldron (talk • she/her) 12:46, 18 June 2025 (UTC)

I mean, the sole purpose of the signature requirement, as I understand it, is to serve as a barrier to clearly frivolous attempts to force admins through a RRfAs. The RRfA is meant to be the actual consensus-based / determinitive part. Have we, in fact, had any frivolous recall petitions get past the petition stage? So far, every petition that has passed that stage has led to de-adminship or a functional resignation by declining RRfA. To me, the main sign that the petition process needs to be more of a barrier would be if too many recalls were passing that step and then failing in the RRfA stage; yet unless I'm missing something this has never happened even once. Regarding the BBB case - while people might reasonably disagree whether an RRfA should have or would have resulted in them being retained as an admin or not, I don't see how anyone can credibly argue that the case against them was frivolous. The fact that they passed the petition stage near-instantly was the system working as intended; I don't see how that's an argument for more delays or a higher threshold. --Aquillion (talk) 08:04, 18 June 2025 (UTC)

Indeed, the only actual recall so far has been Graham. -- asilvering (talk) 16:22, 18 June 2025 (UTC)

If there's ever a recall re-RfA that passes, I will eat my hat. Reaching 25 is effectively a desysop, regardless of how logical those 25 are. Black Kite (talk) 18:05, 18 June 2025 (UTC)

But if that's true - if everyone who reaches 25 signatures does not succeed at RRFA - then the petitions are working as intended. -- asilvering (talk) 02:12, 19 June 2025 (UTC)

I have to wonder, though, if there will ever be a petition that reaches the threshold based on flimsy evidence or reasoning, where the recalled administrator could pass RRFA if they wanted to, but because such an experience might understandably cause one to lose enjoyment in participating here, they decline the RRFA and retire from Wikipedia instead. If only we had a process for removing administrators that is specifically focused on ensuring there is high-quality evidence and where that evidence is reviewed by a group of trusted administrators that are annually elected to serve in that role... Mz7 (talk) 06:01, 19 June 2025 (UTC)

The Fastily one would have been interesting, had it happened, given that a number of the 25 signees gave a criteria as "per another editor" and later on that editor's evidence was found to be not the smoking gun it was presented as. As I said above though, I absolutely don't blame Fastily for retiring from the project in the circumstances. Black Kite (talk) 09:35, 19 June 2025 (UTC)

Yes, but Graham's RRfA was impacted by outside links to the RRfA. If another user goes to RRfA, I am currently expecting a similar incident. In the beginning, I thought the 25 signatures was an appropriate barrier, but having seen the process I feel it might be too low a barrier. (It does seem like the number of petitions is increasing again, so maybe that might change my mind or confirm it in a few months.) --Super Goku V (talk) 05:57, 19 June 2025 (UTC)

• Comment I certainly did not intend my post to Necrothesp's talk page as a threat, but understand how it could be taken as such, and apologize for that. I was merely following the advice currently given at WP:RECALL: Before initiating a recall petition, it is best practice to attempt to contact the administrator on their talk page. This notice should indicate that recall is being considered, and give the administrator a genuine opportunity to address the concerns before a recall petition is filed. which is exactly what I attempted to do. If this case is an example of how the community does not think this should play out, then the phrasing needs some reform. Mdewman6 (talk) 02:07, 19 June 2025 (UTC)

Current guidance for starting a petition vs. reality

Given the rapid feedback I received against initiating Wikipedia:Administrator_recall/Necrothesp, the statement Any extended confirmed editor may start a petition for an administrator to make a re-request for adminship if they believe that the administrator has lost the trust of the community probably needs some additional context going foward, in addition to my concern above. Mdewman6 (talk) 03:38, 19 June 2025 (UTC)

How did you interpret "if they believe that the administrator has lost the trust of the community"? To me, that seems quite a strong bar which brings its own context. CMD (talk) 04:22, 19 June 2025 (UTC)

That is the third step listed in the recall process. The feedback you recieved at Wikipedia:Administrator_recall/Necrothesp was that some other way to resolve the issue should have been tried first: this is already the second step (Other methods of dispute resolution should be attempted before a recall petition is initiated). Caeciliusinhorto-public (talk) 11:35, 19 June 2025 (UTC)

A minor correction: the second step you quote was recently put there (by me). It was already present in the introduction, but I figured it needed a restatement. Firefangledfeathers (talk / contribs) 12:11, 19 June 2025 (UTC)

For what it's worth, it has seemed to me throughout the process that Mdewman6 was acting in good faith, and I don't want that fact to get overlooked. I felt negatively about the lack of prior discussion before the petition began, and said so, but I see the problem as residing in how the project pages for the process are currently written. It's a process that needs to be significantly revised. --Tryptofish (talk) 20:37, 19 June 2025 (UTC)

Ah, I missed that. Personally I think this is now sufficiently clear, but I don't see an issue with the suggestion below that we should specifically recommend dispute resolution at a community forum. Caeciliusinhorto-public (talk) 08:21, 20 June 2025 (UTC)

I of course agree other methods of dispute resolution should be attempted before a recall petition, but an essential step in dispute resolution is a discussion on the user's talk page, and I read this in parallel with the first bullet point. I see now the next step (as in any dispute resolution) was to go to ANI when Necrothesp ended the usertalk discussion. But at the time, I saw a long-term WP:ADMINCOND issue and I didn't see starting a recall petition as the nuclear option, I wasn't building the gallows or guillotine, or however one may put it: I thought I was merely starting a poll, and that discussion on the merits could occur there just as it could at ANI. That said, some additional guidance along the lines of "in most cases, a discussion at a community forum such as WP:ANI should occur before a recall petition is initiated" is probably justified. Mdewman6 (talk) 21:50, 19 June 2025 (UTC)

I support adding the additional guidance you propose in your last sentence. Mz7 (talk) 22:12, 19 June 2025 (UTC)

I support that, too. --Tryptofish (talk) 22:22, 19 June 2025 (UTC)

I support that as well. And that discussion should be open long enough that the admin concerned has the opportunity to, at minimum, meaningfully respond. Thryduulf (talk) 23:54, 19 June 2025 (UTC)

Withdrawal of signatures or petitions

Although it's clear from practice that an editor who signs a petition may withdraw his or her signature while the petition is pending, as far as I can tell, the page does not actually say so. Should we add a sentence along the lines of: An editor who has signed a petition may withdraw their signature by striking it out at any time until the petition is successful (having obtained 25 signatures) or is closed as unsuccessful? Relatedly, should we address the question (raised by the recent abortive petition) of when a petition itself may be withdrawn? Perhaps: The editor who opens a petition may withdraw the petition as long as no other editors have signed it, or if all the editors who have signed withdraw their signatures or consent to the withdrawal? Regards, Newyorkbrad (talk) 14:27, 24 June 2025 (UTC)

Personally, I don't think additional instructions are needed regarding someone removing their signature, since this follows normal collaborative practice on English Wikipedia. Regarding withdrawing a petition: the key aspect that I think should be clarified is whether or not the six-month respite period can be avoided once someone starts a petition. I think having conditions when a petition can be withdrawn and the respite period not take effect would be desirable. isaacl (talk) 17:22, 24 June 2025 (UTC)

+1 to all of that. Firefangledfeathers (talk / contribs) 17:23, 24 June 2025 (UTC)

I'm a bit concerned that allowing a withdrawal (and avoidance of the six-month respite period) when all supporters remove their signatures would provide them incentive to always do this when a petition seems unlikely to pass. The consequences of a failed petition in theory should act as a brake to slow quick decisions to support a petition (though I'm not sure how well that works in practice). I think I'm more inclined to support a withdrawal only when no one other than the initiator has signed the petition yet. isaacl (talk) 17:34, 24 June 2025 (UTC)

I agree with Isaacl. I think a relatively simple and fair approach would be:

• A petition that is withdrawn before any editor other than the initiator has commented should be treated as if it was never opened.
• A petition that has attracted comments (possibly excluding purely procedural ones?) but no supporters may be withdrawn but the six month waiting period still applies.
• A petition that has attracted support votes may not be withdrawn, even if all the support votes are.
• The nominator may withdraw their vote, but the petition will remain open and will pass if it reaches 25 non-withdrawn supports.

Thryduulf (talk) 17:42, 24 June 2025 (UTC)

I think a withdrawn petition is withdrawn and so it should not count for the six month period. A withdrawn petition can be a single signer or multiple signers all withdawing support. The process will simply not have played out to the extent that the six month period makes sense. In my mind that all triggers for something that is brought to completion whether through 30 days or through a re-RfA. Best, Barkeep49 (talk) 17:50, 24 June 2025 (UTC)

There is also incentive for a supporter of the administrator to sign a petition, in support of allowing the process to complete even if everyone else has removed their signatures, in order to trigger the respite. I think it would be better for the process to always run to completion to avoid these incentives, with the sole exception being the initiator withdrawing with no other supporters (to deal with spurious cases). isaacl (talk) 18:01, 24 June 2025 (UTC)

I also thought about that but decided not to raise in the spirit of assuming good faith that people wouldn't do that. Best, Barkeep49 (talk) 18:03, 24 June 2025 (UTC)

I think that editors can, in good faith, sign a petition in support of a re-request for adminship being made, with a desire to see the admin gain positive support from the community, either through a failure of the petition to pass, or through a successful re-request. isaacl (talk) 18:10, 24 June 2025 (UTC)

I think the second item will result in rapid comments in support of the administrator in question. There isn't much advantage in having a formal withdrawal procedure if the six-month respite period is going to always be triggered. isaacl (talk) 17:54, 24 June 2025 (UTC)

(edit conflict) @Barkeep49 I strongly disagree - withdrawing a petition should not be without consequence. We want people to be certain that opening a petition is the right thing to do before they open it, and to thing carefully about their proposal before they make it, and making it hard to withdraw a petition is part of that. If the admin is doing something seriously wrong they can still be taken to arbcom. Thryduulf (talk) 17:57, 24 June 2025 (UTC)

I think your proposal misaligns incentives. It makes withdrawing of huge importance such that it disincentivizes anyone to withdraw. So if someone opens something they shouldn't have - which I agree we don't want - your method then suggests "well don't even think of withdrawing lest you want to close this process off for others". So we live with a bad outcome. And we know this is likely to happen based on our experience with people filing ArbCom requests despite not at all understanding the reasons and consequences for doing so. Best, Barkeep49 (talk) 18:00, 24 June 2025 (UTC)

If I understand Thryduulf's proposal correctly, an initiator withdrawing their petition won't trigger the respite period, so it doesn't close the process. Someone else is free to start a new petition if they wish. isaacl (talk) 18:21, 24 June 2025 (UTC)

It does but I don't think Wikipedia is worse off that the Necro petition was withdrawn by two people and instead there was a thoughtful discussion at ANI. But I also don't think that immunizes him from future recalls. Best, Barkeep49 (talk) 18:26, 24 June 2025 (UTC)

I agree Wikipedia is not worse of that the Necro petition was withdrawn, but it is presently undefined whether the withdrawal makes them immune for 6 months (although hopefully this is moot). My view is that it should make them immune, because this reduces the likelihood of people starting petitions without thinking it through, or withdrawing without thinking it through. Thryduulf (talk) 18:32, 24 June 2025 (UTC)

Hmm, part of your initial proposal was: A petition that has attracted support votes may not be withdrawn, even if all the support votes are. My interpretation of that is that it would have prevented Necrothesp's petition from being closed early. Just to be clear, is your view on that any different now?

Regarding whether Necrothesp's withdrawal makes them immune for 6 months, it does not. Under the current rules, a petition needs to be open for 30 days and fail to reach 25 signatures in order to grant the 6 month immunity. I also agree with Barkeep49 that it shouldn't grant immunity either. It is already really difficult to withdraw a petition that has multiple signatories. We got lucky with Necrothesp in that there were only two editors and they were both quite reasonable about it. If we have learned anything about this recall process so far, it is that people are quick to pile on: the last four petitions were all closed within 24 hours. The reality is that we will get petitions that are poorly-thought-out in the future, and we should be making it easier for those to be nipped in the bud, not harder. Mz7 (talk) 05:04, 26 June 2025 (UTC)

While I agree that allowing a withdrawal without triggering the respite period was suitable for this situation, I'm still uneasy about giving the signatories control over the triggering of the respite period. I'd rather take it out of their hands. isaacl (talk) 18:33, 24 June 2025 (UTC)

I think the simplest way to prevent Gaming the system is to make it so that any petition that is open for X hours is to be considered to trigger the immunity from recall period. What to define X as? Something like 6 hours, 12 hours, or 24 hours would work, I believe. (This would also prevent silliness. For example, a user could post a malformed petition, withdraw it to fix it, and re-posting it 15 minutes later, leading to a messy debate on if the user is immune because it was withdrawn or if the malformation was at attempt at gaming or some other consideration. Given how many malformed requests are posted to Wikipedia, I expect this to happen sooner or later.) Super Goku V (talk) 06:56, 25 June 2025 (UTC)

I like the "any petition that is open for at X hours is considered to trigger the immunity from recall period" idea. I'd oppose anything long than 24 hours or anything shorter than 2 hours. Thryduulf (talk) 10:59, 25 June 2025 (UTC)

I think there should be a little time for discussion with the initiator (for example, their mentor or some other experienced editor might be advising them on the best path forward). I appreciate, though, that it is desirable for the initiator not to be hounded into an action. Perhaps a 48-hour period might be more suitable with this approach. isaacl (talk) 16:41, 25 June 2025 (UTC)

Separately from withdrawn petitions, should withdrawn votes count towards the five active petitions you are allowed to support? My first thought is yes, because we do not want to enable people to change their mind about supporting petition X just so they can support petition Y. Thryduulf (talk) 18:00, 24 June 2025 (UTC)

Or repeal that entire rule since the way we are using recall now makes it basically a nullity. * Pppery * _{it has begun...} 18:07, 24 June 2025 (UTC)

Personally, if there's going to be a limit, I think it's reasonable for an editor to change how they want to allocate their signatures. isaacl (talk) 18:14, 24 June 2025 (UTC)

+1 to Pppery. The 5 active petitions rule was mainly added to protect against clearly spurious gaming, because people were seriously worried RECALL will become an axe for everyone to sharpen on. If the community believes RECALL will not be used that way anymore; or that it can self moderate clearly bad faith cases another way, it makes sense to remove that restriction. Soni (talk) 06:52, 25 June 2025 (UTC)

I'm the guy that initially proposed that and fought for it. When I cooked it up I thought this process might look very, very different than how it turned out. I'm fine with removing it now. We do want to make sure that we have some way to stop recall petitions from being spammed. We can try to do this socially with disruptive editing and IAR, but I'd prefer to modify the rule so that you cannot initiate 2 petitions at the same time. What I don't want to see is someone starting a petition against every member of arbcom after they do something controversial, or recalling every admin whose names begin with a or b because they think that all admins should be re-examined once a year. The 5 simultaneous signature method is a terrible way to do it though based on the way it shook out. Tazerdadog (talk) 07:34, 25 June 2025 (UTC)

Replacing the "maximum 5 simultaneous signatures" with "cannot initiate more than one concurrent petition" makes sense. It would also make sense to prevent someone opening a petition against admin A but seeing it get no support so withdraw it and immediately start a petition against admin B, or starting a petition against admin C immediately after the petition against admin A reaches the threshold, but I don't know whether we need an explicit rule (7 days after a petition is closed or withdrawn before you can initiate a petition against a different administrator?) or just treat it as disruptive editing? Thryduulf (talk) 11:09, 25 June 2025 (UTC)

There's another way to read that sentence, without even changing it; it's synonymous with "An editor can sign no more than five petitions. Ever." So if, hypothetically, your only edits in a year outside your own userspace were to sign five recall petitions, you've at least got to get a new sock extendedconfirmed. —Cryptic 11:21, 25 June 2025 (UTC)

Confused. The line says An editor can sign no more than five active petitions. (Emphasis mine) Since you can only sign them when they are active, I don't see how there can be a reading where you must only ever sign a maximum of five petitions in your lifetime. --Super Goku V (talk) 12:54, 26 June 2025 (UTC)

I think the individual circumstances would have to be examined for situations of an editor starting petitions one after the other. Thus at this point in time I don't think an explicit rule is the best approach. isaacl (talk) 16:30, 25 June 2025 (UTC)