Wikipedia talk:Administrator recall/Reworkshop

Cleaning up the questions and merging them

As I've expressed in the Admin elections workshop phase, we should be careful about not repeating Recall's Phase 2 in effect. There will be lots of smaller levers we can pull, but it's important to not let them all turn completely independently - That risks the final product again be another hodgepodge of effects going in completely different directions.

I have very strong opinions on the overall number of independent questions be severely reduced for the workshop phase - Perhaps we don't relitigate things that already got asked in an RFC (recently), we likely merge multiple related questions together (Discussion-Discussion Location-Explanation of signatures, Closing-Withdrawing etc), and we reduce the overall number of questions anyway. Some of these questions have much larger implications than others; so I'd like someone to probably volunteer to decide which of these questions we will keep post Reworkshop. That way we don't risk having 30 questions again and allow policy fatigue.

Perhaps @Theleekycauldron since she's already helping format the reworkshop? Soni (talk) 16:38, 9 November 2024 (UTC)

The matter of withdrawing a petitionhas more or less reached a conclusion—that it's only allowed if one is the only signatory or all signatories have agreed—through normal editing. The simultaneous petition limit hasn't been brought up anywhere else nor are we likely we are to have >5 petitions open simultaneously in the near future. The word limit extent is, as is mentioned, not relevant yet. The discussion section, petition length and explanation of signatures are being discussed elsewhere.

I've ordered the remaining by roughly how important I think each sub-RFC is:

1. Number of editors
2. Requisite support of uninvolved editors
3. Opposition
4. Opening a petition
5. Word limits
6. Closing a petition
7. Can an administrator voluntarily opt into a RRFA?
8. Discussion elsewhere
9. Monitors

I understand it's a lot to ask, but I'd be interested in seeing other's rankings so that that someone (preferably uninvolved but not required) could have an idea of the five or six most important issues. Sincerely, Dilettante 17:36, 9 November 2024 (UTC)

I think that a lot of the questions don't change the status quo much, not saying they don't matter. I must admit that I am quite dizzy with the number of questions here and spent an hour just to get an idea what the main contentions are. We should pick four or five of the juiciest questions and run them through RfCs first. I have some concerns with relitigating minor details, such as 30 days timing, but they can be put into RfCs at later dates. Kenneth Kho (talk) 23:18, 22 November 2024 (UTC)

Ideas or proposals to execute

I'm thinking about addressing double jeopardy-like situations (i.e. discretionary scopes) and bureaucrats' decisions on petitions and scopes of them, but I can't figure out words to say precisely and accurately. Shall I say just the ideas without yet describing and providing options? George Ho (talk) 06:49, 14 November 2024 (UTC)

Instructions

I don't know it's appropriate to add this to the main workshop page; so, I'll add it here. I think that WP:RECALL#Petition probably needs some clear instructions regarding what to do when starting a petition and what to do when closing a petition. I understand that this whole process is new and there are still wrinkles to iron out, but perhaps like to format used in WP:RFC where there are different subsections devoted to "Starting a petition" and "Ending a petition" (with as much detail as deemed necessary) would help make the process a bit easier to understand. -- Marchjuly (talk) 00:22, 17 November 2024 (UTC)

History repeats itself

Way back in the early days of Wikipedia, we used to have a process called Requests for Comment/User conduct, which included a subsection on admin conduct. It was shut down in 2014 for many reasons, at least one of which was that anyone could start an RFC/U, and people who had any kind of concern about the user just piled on. What started as a single point of concern could rapidly devolve into a long series of negative commentaries. The difference was that there was no mandatory response to such pile-ons; sometimes people took it to heart, sometimes they just ignored it, and sometimes they just left. Only rarely did it ever result in an outcome positive to the encyclopedia. So, we stopped doing it. This very much feels like the same thing, except that it only takes 25 people to force removal of an admin's tools. An admin who has completed tens of thousands of useful administrative actions without complaint can wind up being de-adminned because of a minuscule error percentage or a few grumpy comments over the course of years.

I do think that the concept of admin recall from the community is useful and potentially workable. But adminship isn't workable if admins have a sword over their heads that could come swooping down as soon as a few people disagree with them. Twenty-five people isn't a lot of people. I have no doubt anyone who's been an admin for more than 5 years, or who has worked in difficult and/or controversial areas, could easily make that many enemies. Risker (talk) 01:55, 20 November 2024 (UTC)

Thats just factually incorrect. An admin who retains the trust of the community would certainly pass RRFA, even if 25 people recalled them. The more pertinent question is why would you want anyone who couldn't pass a RRFA to remain an admin? That is how community trust is asessed, anyone who right now couldn't pass one shouldn't be an admin. Magisch ^{talk to me} 10:37, 20 November 2024 (UTC)

Even in your worst case scenario, an admin who has ticked off enough people would have to stand for reconfirmation every year. That is in and of itself not a bad thing. Magisch ^{talk to me} 10:38, 20 November 2024 (UTC)

I can think of a way to never piss anyone off with my admin actions. Valereee (talk) 16:39, 21 November 2024 (UTC)

I rather think the optics of all this critique of the process are terrible. To an outsider this looks like established admins trying to hold a thin blue line. I get that's most likely not what any of y'all are going for but the institution of administrators is losing a lot of trust right now given that nearly all complaints about the process seem to come from established admins who are in effect arguing against a mechanism to hold them accountable for their usage of advanced permissions. Magisch ^{talk to me} 10:41, 20 November 2024 (UTC)

The reason is that many would decline going through an RFA. Because to go through an RFA requires a "Type A" personality and making an "I want this / I need this" assertion. North8000 (talk) 16:09, 20 November 2024 (UTC)

I very much agree with Risker, and disagree with Magischzwei. An admin who makes difficult calls is likely to attract 25 signatures over a month, regardless of whether there is really something that rises to the level of a desysop. And if that admin were to be able to then pass the re-RfA (as opposed to quitting Wikipedia rather than face the aggravation, as has just happened), it is downright abusive to make that admin go through the process. ArbCom is entirely able to deal with admins who really need to be desysopped. This process has been a fiasco. --Tryptofish (talk) 16:39, 20 November 2024 (UTC)

@Tryptofish I disagree that RFA is abusive, and I think it's not unreasonable to expect everyone who wants to retain admin permissions on a longer term basis to go through it regularly. It's important for the trust in administrators as an institution and the general good of the project to have only administrators who continually retain the trust of their electorate. It should not be a lifetime appointment. This is why I don't think a recall petition is "a big deal". If it passes, and you're a good admin, you will have no problems at RRFA and then nobody can start another one for a whole year. Magisch ^{talk to me} 18:36, 20 November 2024 (UTC)

We disagree with each other. I didn't say that RfA is abusive. I said that the recall petition is. If you think it's no big deal, then maybe we should open up a page where editors can post about things that they don't like about you. --Tryptofish (talk) 20:16, 20 November 2024 (UTC)

My talk page is in the usual place, feel free. I don't edit much but I wouldn't mind that much, as long as it remains within CIVIL boundaries. Also note I hold (to my knowledge) no advanced permissions, nor am I seeking to do so. Magisch ^{talk to me} 20:18, 20 November 2024 (UTC)

This from an editor who in over 5 years has made 79 edits, of which 54 have been to project space.--Bbb23 (talk) 18:49, 20 November 2024 (UTC)

The identity of the person making a claim doesn't change the fact that they have a point. Thebiguglyalien (talk) 19:18, 20 November 2024 (UTC)

Just because I don't edit much doesn't mean I don't read what's going on here, nor does it mean I don't support wikipedia in other ways. If there's a minimum edit requirement to give an opinion on this process that I haven't seen, feel free to redact my comments and receive my apology. Otherwise, I haven't adressed arguments based on any particular person making them, and I would ask for the same courtesy. Magisch ^{talk to me} 19:31, 20 November 2024 (UTC)

A recall petition is a big deal, just like RfA is. Those who have gone through it, or been close to someone who has, recognize the difficulty and stress of it. There's a reason a number of admins plan to just give up the tools if recalled instead of trying to fight to retain the tools and volunteer their time. There's a balance to be struck, and I certainly don't believe admins shouldn't be held accountable, but downplaying the awfulness of it all doesn't help anybody. Hey man im josh (talk) 19:22, 20 November 2024 (UTC)

That to me seems like a problem with the RFA process that should be adressed there. Magisch ^{talk to me} 19:32, 20 November 2024 (UTC)

Then tell us how to fix RfA. Hey man im josh (talk) 19:34, 20 November 2024 (UTC)

I don't have the insight to do that. But I don't think the bar for accountability should be higher for people seeking to gain admin tools then it should be for people who seek to retain them despite community concerns. Magisch ^{talk to me} 19:36, 20 November 2024 (UTC)

I passed with over 300 supports last September, but it would take only 25 signatures to force me to go through the process again (if I decided I even wanted to). Those concerns don't even necessarily need to be valid or serious to reach 25 signatures, something that's evidenced based on a number of opposes at various RfAs I've seen. We, routinely, see opposes that are clearly based on past interactions where one person was in the wrong, but it's not the person being opposed, or someone upset that they didn't get their way and consensus formed against them. I am absolutely for accountability, but the system, in its current state and with the low bar that needs to be cleared, has potential to force people through the process a second time when it may not be necessary or appropriate to do so. That's MY issue with it, as an admin who's not one of the old admins you've referred to. To be clear, I'm owed absolutely nothing and my past success at RfA says nothing about how I've been as an admin. But I mention it for context and that, even for those of us who had smooth RfAs, we don't want to go through it unnecessarily again. Hey man im josh (talk) 19:40, 20 November 2024 (UTC)

A recall petition is a very big deal. You may disagree that RfA is stressful, but many of those who've been through it recently disagree. It was quite stressful for me. I spent a solid week pretty much every waking hour watching it unfold, and there was some stuff said or implied that made me want to vomit. Valereee (talk) 16:43, 21 November 2024 (UTC)

I posted a reply to this in the discussion at Wikipedia:Village pump (policy)#We need to fix the admin recall process (comment link). I encourage anyone who has thoughts about the "thin blue line" theory to look at the statistics. Thebiguglyalien (talk) 17:56, 20 November 2024 (UTC)

The "thin blue line" argument has never had evidence to support it. It was an aspersion from the beginning, and it's definitionally a failure to assume good faith. ⇒SWATJester ^{Shoot Blues, Tell VileRat!} 19:00, 20 November 2024 (UTC)

I linked to evidence in the very same comment you're replying to. Thebiguglyalien (talk) 19:16, 20 November 2024 (UTC)

You linked to a partial analysis of data based on raw support count that did not even attempt to address the substantive reasons for supporting, and used that to jump to what you called an "obvious conclusion" apparently solely based on assumptions about what you think 48 administrators believed rather than what they actually expressed at the RRFA. I would not call that useful evidence, IMO. ⇒SWATJester ^{Shoot Blues, Tell VileRat!} 21:08, 20 November 2024 (UTC)

Well I really hope that the evidence indicates what it looks like to me. The alternative is that we have a hundred editors willing to write off things like User talk:Lazborn#Blocked and User talk:श्रीमान २००२/Archives/2020/September as minor issues that it's okay to overlook, and that a sixth or seventh chance is warranted. I cannot respect anyone who would be willing to vote support for that reason. Thebiguglyalien (talk) 21:55, 20 November 2024 (UTC)

I respect editors who evaluate evidence carefully and on the merits, as opposed to rushing to judgment. --Tryptofish (talk) 22:18, 20 November 2024 (UTC)

It's not a binary between opposing Graha87 and being willing to write off things like User talk:Lazborn#Blocked and User talk:श्रीमान २००२/Archives/2020/September as minor issues. My !vote, for example, explains another reason for supporting. If you choose not to respect me, so be it, but don't write a blanket statement about 100 support !voters. Sincerely, Dilettante 23:08, 20 November 2024 (UTC)

Perhaps admins are more aware of the scrutiny they face and the stress of RfA so are more empathetic to Graham87's plight. Sincerely, Dilettante 21:11, 20 November 2024 (UTC)

• Risker, do you feel that either Graham87 or Fastily (or both) were treated unfairly? It's striking that neither case looked remotely likely to attract 25 signatures, until, in both cases, previously uninvolved editors brought additional evidence to the community's attention, leading to a sudden rush of signatures. This seems right to me. Most Arbcom cases don't attract 25 participants, so it's clearly not the case that 25 people with a grudge will come for you because you blocked their buddy for typing rude words about Palestinians. In fact, looking at participation in Arbcom cases, 25 extended-confirmed editors is an extraordinarily high threshold.—S Marshall T/C 08:46, 21 November 2024 (UTC)

  25 extended-confirmed editors is an extraordinarily high threshold. Cmon now, it's fine if you don't agree with changing the threshold but this is a silly statement. You might as well propose 10 signatures being enough if you truly believe this. This threshold exists because that's what it was on a smaller wiki that uses recalls, not because it's a significant bar to clear. Arbcom is also not a fair comparison in this case. I simply never participate there, and I know that's the case for a lot of others, but that doesn't mean I don't support a case in a number of situations. It's significantly different to sign a petition than to participate in an arb case, which is why participation is higher at petitions. A more accurate comparison would be ANI, a place which we see people with grudges pile on all the time, and a place which certainly helps to make the case that 25 is too low. Hey man im josh (talk) 11:04, 21 November 2024 (UTC)

  This threshold exists because that's what it was on a smaller wiki that uses recalls, not because it's a significant bar to clear. That's not really true. The smaller wiki also had "6 months", both of which are much more likely to require a higher threshold. At this point, the specifics of EnWiki's RECALL are different enough from other projects that the same numbers cannot be projected across multiple project. We have discussion allowed, and 30 days for petitions. Both of these variables change the threshold calculation a lot.

  The main thing I am personally gauging this by is "Did the petitions so far look too easy to pass", and both of them kinda... didn't? They both steadily got too many signatures, but both looked likely to fail until a big burst of activity. That sounds more like recall at work than canvassing or frivolousness - Signatures happened when evidence was found, not purely out of pile on momentum.

  All I am saying is... We either compare all the factors with the other wikis. Or, we start treating our recall as it's own thing, now that we had two clear cut cases to read. Based on my read of things, the threshold was not as insignificant or as ill thought of as you suggest. Soni (talk) 11:46, 21 November 2024 (UTC)

  Yes: I've always thought the threshold should be about ten extended-confirmed editors. That kind of number alleging sysop misconduct would virtually guarantee an Arbcom case being opened.—S Marshall T/C 11:49, 21 November 2024 (UTC)

  As mentioned, statements at arbcom are drastically different than just signing your name to a petition. Ten signatures to start the process of re-rfa would cause so many people to say screw it and give up their tools instead of going through rfa again. Hey man im josh (talk) 12:11, 21 November 2024 (UTC)

  Look at it though. See the actual examples of this working in practice. The people certifying the recall give their reasons in full. Sometimes they do so in pre-recall discussions on AN or AN/I, but there's clearly a lot of thought and judgment going into them. The vindictive retaliatory certifications you're worried about have not happened at all.—S Marshall T/C 13:04, 21 November 2024 (UTC)

  @Soni: Perhaps if people allowed the recall petitions to stay open then we'd have a better idea of how many signatures they would have actually gotten. Unfortunately, some people wanted them closed sooner rather than allowing them to play out (I know you and I discussed this and this isn't directed at you specifically as a number of people apparently wanted them closed immediately). We really can't know because of this.

  We're in the early stages of recall, where people are trying for what they see as the easier cases, which, probably would have reached the 50 signature threshold as well.

  I'm looking towards the future. There will be people who are unfairly the victim of the recall process and we owe it to the health of Wikipedia and to the integrity of the process to be proactive instead of reactive. I've spoken to a lot of people about this, and many believe they could reach 25 signatures, but that doesn't make them a bad admin. Despite how much I work to take criticism, grow from it, and not make the same mistake twice, I still believe I'd have 25 signatures if a petition was started against me. My RfA was relatively smooth compared to others, but do I want to go through that again? Hell no.

  As more people become aware of recall the participation will grow, especially if advertised via watchlist like some people want. Take note of some of the ridiculous and vindictive oppose reasons in RfAs. I'm for accountability and some form of recall, but I'm deeply concerned about the potential for problems with our current iteration of the process. Hey man im josh (talk) 12:20, 21 November 2024 (UTC)

  You and me both. I knew from the very start that recall could be a massive can of worms, it just looked clearly a good idea worth pursuing regardless. I genuinely believe that the recalls we have so far were a good thing for enWiki, despite being not ideal in a number of ways.

  As for the petitions to stay open, I don't think there's much point re-litigating that. The original intent was always to close them sooner, else you risk having a pile-on of opposes which could be demoralising. But death of the author, yadda yadda, so I get that we're going to consensus our way through this. I think just "How many days did it take to reach threshold" is still valuable to judge recalls as they exist. It will eventually become clearer one way or another, I just hope it doesn't require as much bickering as every other aspect of this process. Soni (talk) 13:02, 21 November 2024 (UTC)

  I think you and I are mostly on the same page, and you do have a point that it's relevant how long it takes to reach said threshold. I disagree with aspects of why each person was recalled so far, but I do have difficulty defending either recall at this point. Hey man im josh (talk) 13:52, 21 November 2024 (UTC)

  Editors have been noting, accurately, that both of the petitions we've had so far followed a pattern of starting slowly, and then accelerating when new evidence was brought forth. We'll probably need more experience before we can really answer this, but I think one can interpret this observation in either of two opposing ways. One way is what other editors have already implied here: that the process worked because the right kind of evidence came forth as a result of keeping the petition open long enough. But an alternative way to look at it is that the petitions were started with evidence that was unconvincing to most of the community, but if you leave something open long enough, something will eventually shake out. I find the second possibility concerning. --Tryptofish (talk) 17:35, 21 November 2024 (UTC)

  Generally agree with Tryptofish here. One thing I'd add is that for any highly active admin, there are always likely to be mistakes and errors of judgement; we need to be asking things like: is the proportion of mistakes concerning, are the individual mistakes serious, is the admin learning from mistakes/correction. I feel the current system prioritises finding mistakes but addresses the real questions poorly. Espresso Addict (talk) 17:58, 21 November 2024 (UTC)

  If every active admin had a skeleton in their closet that once revealed would make them fail a RRFA, I think we would need stronger medicine than this process (like having to be re-elected every 5 years). I suspect that's not the case though and we're just seeing some pent up demand. RevelationDirect (talk) 05:35, 22 November 2024 (UTC)

  I still believe I'd have 25 signatures if a petition was started against me. No way. I'd eat my hat. Unless you've made dozens of horrible, bite-y blocks we don't know about, just no way. You don't have the predicate ANIs/whatever, which is an almost-conclusive indication you haven't done anything that would get 25 people to sign a recall petition.

  I think you're vastly underestimating how serious people take signing a petition. They're not going to sign it because you did something they disagree with. They're going to sign it only in the face of serious, longstanding problems that have been raised before without sufficient improvement. So far, the two petitions we've had confirm this. Levivich (talk) 19:07, 22 November 2024 (UTC)

  Perhaps I am more fearful than I should be, and perhaps I am underestimating how serious people take signing a petition. In my mind, I'm thinking of the frivolous oppose reasons that we sometimes see at RfA when considering the possibility of signatures. I'm also hypothesizing that we'll start to get more views at these petitions the longer we hold them, leading to more willing signatories, which is part of what's adding to my fear about it and why I'm hopeful to have the process tweaked before then. Hey man im josh (talk) 19:12, 22 November 2024 (UTC)

  S Marshall I have received your ping but am AFK all day and will not be able to respond until late this evening. But I do view this current iteration as a serious systemic failure. Risker (talk) 13:50, 21 November 2024 (UTC)

  Responding to the initial question asked of me by S Marshall: To me, the fundamental issue is that our system was misused to bring in a policy under the radar of most editors including the administrators who are directly affected. I have made it clear that I think there is probably a place for community-based admin recall. I would be in favour of a broad community discussion on this specific topic. But I go back to the initial RFC: it was about Requests for adminship. It was NOT about admin recall. The two have as much in common as RFA and the blocking policy, or RFA and the deletion policy. They're all commonly the subject of questions at RFA, but that's where the similarity ends. I've gone back and reviewed the central notices, the watchlist notices, the posts to AN, the (single) post to WT:ADMIN, the Administrators Newsletters during the course of the RFC(s). A few of these mention there's a recall proposal, but it's pretty much buried in the middle of a list of other proposals that logically would be seen in an RFC on RFA. In many cases, it's not mentioned at all. Most people would not expect to see a proposal about what they'd probably consider an unrelated topic in an RFC about any given subject. I'm not a great one for analogies, but I think it might be like seeing a proposal about notability in the middle of an RFC on blocks; it just wouldn't compute. I'm not assuming bad faith on anyone's part; the editor who initiated the proposal was no doubt acting in good faith, the people organizing and coordinating the RFC were probably hesitant to remove any proposal that could even vaguely be considered related to RFA, especially one that came from a non-admin, and it seems that there wasn't really all that much discussion there.

  Given that my thinking is that this policy was not created within our project's normal standards, I don't think that subjecting admins to it without actually having a community discussion about this specific topic was fair to anyone. The discussion at VPP that led to this becoming policy was one that probably should have been closed as "too soon to get consensus, back when people finish modifying it" because even as the VPP discussion was taking place, modifications were being made to the proposed policy. We don't normally pass policies that way. So in that sense, the process has been unfair to both the community and the administrators who were the subject of recalls.

  There are a lot of things that we still need to nail down. Generally speaking, our community has decided that only in rare and exceptional circumstances would we consider evidence that is more than two years old. There was lots of older evidence in both the recall petitions. We need to have a discussion on why this should be an exceptional circumstance, and whether those same diffs and that same old evidence will be allowed at a subsequent recall petition if the first one does not succeed. We had lots of votes about deleted edits from people who can't actually see deleted edits, but there is no way to discount those votes. We allowed extensive discussion, and some of those discussions illustrated that some of the votes were based on a mistaken interpretation or assessment of a situation, but those votes still counted. Given the really low threshold for success, these are all things that need to be properly considered. So in that sense, I think the recall process was inherently unfair to the two admins who were subjected to it, and it will continue to be unfair to any other admin who might be subject to it before we clean up this policy. I do think that there were some really valid criticisms made in both of the recall petitions.

  And, speaking strictly as an individual and completely unrelated to any recall or any particular person, I really do worry about people who take on unusually heavy workloads for an extended period. Having been here for almost 20 years, I can honestly say that the vast majority of problematic behaviour seen in both editors and administrators is related to spending way too much time on Wikipedia and seeing oneself as The Key Person for any role. We need to find ways of saying quietly to people "maybe you should take a bit of a break, or work on a different area, or train someone else to do Task X" before things get to the point where we are blocking people or desysopping them or driving them from the project. Risker (talk) 03:44, 22 November 2024 (UTC)

  The long-term behavior by an admin toward multiple new editors that came out during the RRFA was honestly shocking to me. That's because my experience with admins has been very positive and admins deserve a fair process.
  
  I hope we have enough concern for fairness to also include consideration for those new editors who were improperly blocked. RevelationDirect (talk) 05:23, 22 November 2024 (UTC)

  I think of them as part of the community. I agree that they deserve fairness. I suspect, though, that it is difficult for the average editor who hasn't worked in areas where there is a really stunning amount of paid/COI editing or LTA/chronic socking to grasp how serious both of those problems are, and why in some cases the best thing to do is refuse to give them the attention they seek. There are editors involved in this discussion who've seen some of it, but across the project it's hundreds of accounts a week. Hence my mention of "don't do too much of one thing" and "teach someone else how to do it too", because it is really easy to become jaded and to start seeing all new editors as suspect, or all articles about organizations or people as probable paid editing. In an ideal world, admins would be more of a team that worked together to spell each other off on tasks like that, but this is not a perfect world. Risker (talk) 06:15, 22 November 2024 (UTC)

  On the one hand, it's dissapointing that--even when discussing non-admins--most empathy is reserved for fellow admins. On the other hand, I appreciate such a candid answer. RevelationDirect (talk) 23:31, 22 November 2024 (UTC)

  Depends on which part of my comment is disappointing, RevelationDirect. If it's that people shouldn't do too much of one thing, that applies equally to non-admins as it does to admins. Too much use of AWB or Twinkle, too much categorization, too much focus on editing in a contentious topic area: they're all just as prone to the development of biases and jaded behaviour as anything admins see or do; and I'll note that most admins spend much more time in the non-admin sphere than wearing their admin hats. If it is that there aren't a lot of non-admins who have intense experiences with LTAs/chronic socking or paid/COI editing (or that the average admin who blocks socks and looks at COI sees more of it), I think I'm more likely to envy their experience than to empathize with them. ("Geez, you don't have a lot of socking in your preferred area of editing? What a shame!" said nobody ever...) I'm hardly the most active admin, and even I am likely to see a few COI/paid editing issues a week, not including what I see as a CheckUser. Risker (talk) 05:28, 23 November 2024 (UTC)

  Yeah if I'd done my recent watchlist purge (mentioned in my RRFA) five or six years ago, I don't think I would have become anywhere near as jaded, thus avoiding my bad blocks, and I'd still be an admin now. But we also wouldn't have had my expansions of the articles about James West (a microphone pioneer) or "Cunnamulla Fella" (an Australian country song) ... so, take your pick. Graham87 (talk) 09:17, 22 November 2024 (UTC)

  Risker's analysis of the ways the "consensus" process went wrong during the process of adopting recall exactly matches the way that I experienced it in real time. This is really important. Either we have a legitimate procedure for adopting policies that affect real live people, or we have some very serious problems. I remain very frustrated by how I repeatedly raised this issue, and was blown off.

  I also agree strongly with the importance of evaluating work in the context of the total volume of that work. That's not in conflict with the importance of treating new editors with patience and respect. The petition process does a truly awful job of accomplishing that. If you look carefully at Fastily's petition, you will see me and some other editors trying to go carefully through every diff of evidence, and raising legitimate questions about whether all of the evidence really was what it purported to be, and then other editors signing the petition, arguably without really having examined that analysis on their own. --Tryptofish (talk) 14:48, 22 November 2024 (UTC)

  This idea that the consensus process "went wrong" seems entirely misconceived to me.

  Wikipedians routinely make decisions without consulting those who they affect. We might decide to delete an article about your company, or your product, or about you, without involving you and without giving you the chance to make representations. We might decide what the article says without involving you. We might decide to keep it without involving you. If your marketing department makes an account to discuss these matters with us, we will block it because of rules that we understand and you don't. If you personally make an account to discuss these matters with us, we will ask you aggressively about "conflict of interest", as if you owed Wikipedia a professional duty of care. If you involve your legal representatives, or warn us that you might, then we will block you. We do not consult people affected by our decisions. I think it's wrong in principle to treat power users differently.—S Marshall T/C 20:49, 22 November 2024 (UTC)

  That's not unlike what I was told when I tried to raise concerns about the consensus process while it was happening. At this point, I am sick of repeating what I've already said, and I despair of this issue ever getting an appropriately thoughtful discussion. But I'll hand it to you: your argument that a user who became an admin by running the gauntlet of RfA and obtaining the trust of the community, and who does very large amounts of difficult work on the community's behalf, should be treated the same as a drive-by vandal, because, hey, we treat all editors equally – that's very creative. --Tryptofish (talk) 00:11, 23 November 2024 (UTC)

  Risker, when did "our community has decided that only in rare and exceptional circumstances would we consider evidence that is more than two years old"? At the recent RFA election I was called upon to explain events more than twelve years ago! Hawkeye7 (discuss) 19:17, 22 November 2024 (UTC)

  Hawkeye7, I think it would be reasonable for people to ask about previous RFAs and desysops and the situations that led to them when the candidate is running for RFA; it hasn't been a written-down exception, but I think most people would see that as a reason to cross that 2-year line. At the same time, I don't think questions about editing 12 years ago, or an admin action you made when you held adminship, would have been reasonable. I hope you found this most recent RFA less stressful than the last one (and probably the first one, too). Risker (talk) 05:45, 23 November 2024 (UTC)

• Okay. Because my position is that neither Graham87 nor Fastily were treated unfairly. The problems you and others foresee have yet to appear. The process is working well in practice at the moment.

We as a community need oversight and direction of how admin tools are used. Sysops can't be unremovable, because they're elected by humans, and humans make mistakes sometimes. In most communities that have elections, the elected people serve for a defined period and then there's another election. Our experiment with people elected once, forever, was a grave mistake and one that I'm delighted we've moved on from.

Arbcom must not be the only route to hold sysops to account. Arbcom has ongoing problems with job abandonment by committee members. This process is reaching good decisions in less time than the current committee takes to decide whether to accept or decline a case. We desperately need to reform Arbcom and lighten its workload so it can focus more on private-evidence stuff.

I do not think of recall as policy. This is not its final shape. There will be tweaks and fixes as problems arise.

I, too, remember RFC/U and I loathed it from afar (having never used it or been dragged to it myself). I agree that this must not become that. It should be faster and less proctological than an Arbcom case which is the current alternative.

I do not think sysops should enjoy amnesty for misdeeds more than 750 days old. Use of the sysop tools is held to a higher standard and misuse should be remediable for far longer.—S Marshall T/C 08:28, 22 November 2024 (UTC)

I don't know what to say when you express the opinion that I do not think of recall as policy.. The discussion at VPP was closed as it now being policy. It has been publicly announced as policy. It has been incorporated into the Administrators policy as policy. It has been used as the policy by which we have had two recall petitions. Unlike several other outcomes of the RFC on RFA, it has not been framed in any way as an experiment or a trial. I agree with you that the current policy needs to be reformulated (This is not its final shape.). That's what we're doing here. Risker (talk) 05:28, 23 November 2024 (UTC)

I don't think it's policy, and the community doesn't think it's policy. It's a consensus which is, as you rightly say, being tweaked.—S Marshall T/C 10:53, 23 November 2024 (UTC)

• Tbh, I disagree with the premise. Firstly, because by the same reasoning, any editor always has a sword over their heads because any other editor can take them to WP:AN and request a site ban. Which is absurd, of course, and doesn't happen. And there are editors who do far more contentious work than the average admin. e.g. editing in hotbed topic areas. And some of these do get dragged to noticeboards, of course, and yet they still edit. So we shouldn't pretend that admins suddenly become scared to do work because they can be held to account, same as any editor can. Plus, WP:BOOMERANG exists. Secondly, because alongside Graham's RRFA we saw Worm's second RFA, and he's a multi-time arbitrator (ie: the definition of getting involved in contentious issues) who flew through his second RFA.
  More generally, I think I have no doubt anyone who's been an admin for more than 5 years, or who has worked in difficult and/or controversial areas, could easily make that many enemies., despite being repeated a lot, is almost completely untrue. You would fly through an RFA, and I'm sure you've done some controversial work. SFR has been manning AE for a while, dealing with lots of contentious problems and probably making enemies along the way, and I bet he would fly through RFA again (and we'll see this partially tested with the ACE results).
  I say "almost completely untrue" because I think the admins most adversely affected are those doing high-volume work. Even if their false positive rate is the same as any other admin (or even better), because of sheer volume there will be more instances of errors, and the community tends to care more about volume than proportion. It's not a good benchmark IMO and defies the basics of how we do statistics anywhere else in the world, but that's the standard we have, and it's the same standard ArbCom uses afaik. Otherwise, doing contentious work competently doesn't cause one to get desysopped, and nor should it. Doing it incompetently should, however. ProcrastinatingReader (talk) 09:22, 22 November 2024 (UTC)

Flattered that you think that I'd fly through RFA, ProcrastinatingReader. My RFA was absolutely by far the most stressful week of my English Wikipedia career, and I know for a fact that a very large percentage of the admin corps feels the same way. Risker (talk) 05:28, 23 November 2024 (UTC)

• The problem isn't that admins would be scared that they'd be held accountable, the problem, from my POV, is the threshold to force a potentially unnecessary re-RfA. I also don't think we should be focused on the ideal that the first two worked, so we won't have an unnecessary petition pass at some point. Being proactive to address what some see as flaws in a process is important. Hey man im josh (talk) 13:30, 22 November 2024 (UTC)

  Nobody forced Fastily to resign and retire. I want to be clear that this was entirely his own decision. But nobody should conclude that the process was fair unless they can link to his re-RfA and show that it was unsuccessful. --Tryptofish (talk) 14:38, 22 November 2024 (UTC)

  I agree, it was Fastily's choice to do so. We really don't know how it would have played out if they waited 30 days, expressed remorse, outlined how they'd do better, and ran. Hey man im josh (talk) 14:40, 22 November 2024 (UTC)

  We can't know for certain of course. But, if we look at the 1 case that did go to an RRFA and the stated reasons for votes, I think it would have at least gone into the discretionary range if not for the block during the petition. RevelationDirect (talk) 15:26, 22 November 2024 (UTC)

  focused on the ideal that the first two worked well, whether they 'worked' depends on how we define 'worked'. For instance, I think it's unfortunate Fastily left or was even -sysop, and I'd like to think he'd have passed RRFA. But the claim being made in the OP (AFAICT), also commonly made in past recall discussions, is some fear that an admin doing [presumably good] admin work will "make enemies" and get desysopped by presumably these enemies and their friends. This is a much lower bar than the process working.

  so we won't have an unnecessary petition pass at some point e.g. let's look at Graham's petition. Although I didn't support it and personally I hold Graham in high regard, there was at least "reasonable cause" given in the petition (ie: it wasn't frivolous or vexatious), and still people were initially in strong disapproval of it. It had basically lost all momentum by 29 October (signature #11), despite the attention it had. It blew up on the 5th, after several days of radio silence, due to Graham's block during the petition. If that didn't happen the petition looked to be failing. Obviously people were trying to Graham a chance. Again, this is in a petition where at least a semblence of a case was made. Just looking at that, there's absolutely no way an admin with overall community support is going to get recalled based on a frivolous case, much less actually fail RRFA, just because they 'made some enemies' doing their good admin work. We don't even see RFARs against such admins. So I must ask, where's the evidence here?

  More generally, I think what we've seen in recent reforms is that some long-term hypothetical fears came not to fruition. Look at AELECT and the immediately preceding fears of tons of admins getting elected without scrutiny (eg here). Probably a book worth of text was expended on these hypotheticals, yet they didn't happen. I could name many more examples, to the point where I think it's easier to list instances where 'hypothetical fears' to change were actually vindicated after the fact. I think this stuff has historically been seriously detrimental to community development; it just paralyses everyone with fear to do anything. I do think the process needs tweaking (unsurprisingly), but let's not pretend there's anything wrong or unfair happening here. It's great we're finally able to move past aforementioned barriers, and I hope now we can spend time discussing how to tinker with the solution to make it better, rather than going back to discussing the age-old, and hopefully now discredited, tropes. ProcrastinatingReader (talk) 19:03, 22 November 2024 (UTC)

• Quite a few people have commented, here and elsewhere, on how many admins make loads of 'enemies' and detractors who'll jump on the opportunity to start or sign a petition. If that's the case, why haven't any of the prominent sock-hunting admins had a petition yet? or the mainstays at AE or ANI? or really more than just 0.25% of admins? The fact of the matter is that the process, whether or not you agree with the merits of either petition, has not yet been abused for the simple reason that any malicious actor would be placed under further scrutiny and risk a boomerang or find themselves subject to a SPI for votestacking. There is a risk of abuse, yes, but the same is true of ANI, XRV, or any other venue that allows criticism of conduct or actions. Sincerely, Dilettante 17:08, 22 November 2024 (UTC)

  This was a major, recurring theme in Fastily's recall discussion among the admins that have been around for longer. I don't think they would agree with any version of recall. It's probably similar to the way law enforcement often become desensitized, AGF just becomes more difficult when 90% of your day is dealing with the underbelly of the Internet. OXYLYPSE (talk) 17:59, 22 November 2024 (UTC)

  @OXYLYPSE: Can we stop with bad faith assumptions? I've seen so many admins who have been around for a long time who support a recall process in general. Being critical of the current iteration is far from the same thing of being against a recall process in general. Hey man im josh (talk) 18:03, 22 November 2024 (UTC)

  I'd suggest you expand on your comment or withdraw it. OXYLYPSE (talk) 18:10, 22 November 2024 (UTC)

  Sure. The text of "... among the admins that have been around for longer. I don't think they would agree with any version of recall." is a bad faith assumption and unfair characterization of a large group of experienced editors. Many of which 'DO support a recall process. Hey man im josh (talk) 18:12, 22 November 2024 (UTC)

  My comment was based on my experience at Fastily's recall where there were a number of older admins discussing recall and "enemies" they've made. My take on it was that they believed recall can never work in any form. This is the specific group of editors I was speaking about. The characterisation is fair and accurate given their comments at the petition.

  You've trimmed my comment to make it look like I was speaking about all admins that have been around longer, not just that pool at the petition. OXYLYPSE (talk) 18:44, 22 November 2024 (UTC)

  My take on it was that they believed recall can never work in any form. – Which is an unfair inference to make. The characterisation is fair and accurate given their comments at the petition. – And I maintain this is a bad faith assumption to be making, so let's stop with those type of preconceived notions. You've trimmed my comment to make it look like I was speaking about all admins that have been around longer, not just that pool at the petition. – Trimmed or not, and with no intention of attempting to frame things differently than what you wrote, I stand by stating it's a bad faith assumption that shouldn't be made. Hey man im josh (talk) 18:51, 22 November 2024 (UTC)

  OK Josh, you win. We'll pretend the admins comparing recalls to a guillotine don't exist. Please leave me alone. OXYLYPSE (talk) 19:02, 22 November 2024 (UTC)

  None of this is about "winning", it's about trying to have constructive conversations that lead to improvements. That's why it's important to not frame groups involved in negative ways just because of their tenure and because they have criticisms. For what it's worth, an admin elected in 2005 also supported Fastily's recall. Hey man im josh (talk) 19:06, 22 November 2024 (UTC)

• Our problem, as Risker notes, is that more and more of the administrative burden is falling on fewer and fewer administrators. The reason for this is that our RFA process is not processing enough new administrators to compensate for separations. One response has been to implement this recall process, with the aim of lowering the pressure at RFA by reassuring voters that admins can be recalled. Twenty-five signatures is a significant bar, and does not in itself trigger the loss of admin status - it just creates an RRFA. Now if admins would not go through an RFA process then that brings us back to my second point. Hawkeye7 (discuss) 20:49, 22 November 2024 (UTC)

  Maybe it's time to lower the threshold for new RFAs to 60% so that we don't have one standard for new admins and another for existing ones, especially since we have more accountability now with recalls. RevelationDirect (talk) 23:35, 22 November 2024 (UTC)

  I don't think that will solve the issues, RevelationDirect. To be honest, of the experiments that have been undertaken as a result of the RFC on RFAs, the one that has been most successful is the WP:AELECT one, and there was no variation in the pass level. Once it is tweaked, I think it might be a reasonable idea to make it the standard approach to RFA. It became the standard on a couple of other projects where adminship and RFA became contentious for various reasons; that's been a chronic problem on this project since about 2007.Risker (talk) 05:43, 23 November 2024 (UTC)

Watchlist notification

Can we get a watchlist notification when a recall is under discussion, similar to all RfAs? — — Benison (Beni · talk) 07:00, 20 November 2024 (UTC)

given that petitions don't currently have an oppose mechanism, a watchlist notice would make almost every petition certain to pass. You can watchlist the relevant subpage, though. theleekycauldron (talk • she/her) 10:50, 20 November 2024 (UTC)

This RfC needs to be shortened

Right now, we're up to 15+ questions. That's too long for the average community member, and we're going to end up with trainwreck that doesn't fix what needs to be fixed. This RfC is only for problems with recall that need to be fixed in the next month. Does anyone have suggestions on which questions do not need immediate answers? theleekycauldron (talk • she/her) 12:21, 20 November 2024 (UTC)

I liked the section above where Dilettante just listed all questions in order of priority, that way we can just cut the questions we agree are not urgent.

So my preference of questions is to just ask

• Number of editors
• Explanation of signatures + Word Limit + Opposition (merged section as 2 questions)
• Recall discussion
• Opening a petition Q1
• Opening a petition : Time delay Q2-Q3 (merged)

I think the below questions should not be asked. But the other questions by my preference are -

• No consensus at Re-RFA
• Closing a petition Q1
• General process
• Minimum time before petition following resysopping at BN
• Administrator inactivity
• Percentage threshold to pass
• Closing a petition Q2-Q3

Soni (talk) 12:57, 20 November 2024 (UTC)

I very strongly disagree that this RFC is only for problems with recall that need to be fixed in the next month. The RFC is for problems that need an RFC to fix, regardless of how many questions that is or when they need fixing. It's important to avoid a second RFC if we can help it. That said my preference is (in no particular order):
Definitely ask:

• Opening a petition
• Number of editors
• Opposition
• Recall discussion (maybe combine with opposition?)
• Word limits (ditto)
• Closing a discussion Q1
• Minimum time before petition following resysopping at BN
• Administrator inactivity
• No consensus at Re-RFA

More discussion needed first to determine if an RFC is necessary

• Closing a discussion Q2 and 3
• Withdrawn petitions and subsequent petitions

Don't ask:

• Explanation of signatures
• Percentage threshold to pass. Thryduulf (talk) 14:32, 20 November 2024 (UTC)

The RFC is for problems that need an RFC to fix, regardless of how many questions that is or when they need fixing. My perspective is that the last time we had a large RFC (Recall Phase II) there were too many changable aspects and that confused too many people. I'd rather focus on our most important 5-6 questions, let recall run with those changes, and then if we still have problems to fix or consensus to have, solve them then. That way, even if it's a constantly changing policy, people know what to expect from from the process at large. Soni (talk) 14:39, 20 November 2024 (UTC)

Writing as an admin who has a pattern of annual hibernation, I would definitely include

• Opening a petition
• Number of editors
• Opposition/Recall discussion
• Administrator inactivity
• No consensus at Re-RFA

From my perspective, the process needs to both be & appear fair, and also needs not to generate an atmosphere in which admins feel unable to go about their necessary business. Espresso Addict (talk) 15:41, 20 November 2024 (UTC)

Having so many questions makes it too complicated to develop with an RFC. Better to develop one proposal here that has pretty wide consensus and them make an RFC on adopting it. North8000 (talk) 16:12, 20 November 2024 (UTC)

The longer this stays open, the more questions we get that only loosely, if at all, address the key causes of the toxicity. My view is that as soon as the timespan RfC is closed (since !votes to several questions are likely contingent on this), we should start the reworkshop RfC. Sincerely, Dilettante 16:49, 20 November 2024 (UTC)

Since the discussions have changed so much, I'll reorder them.

Vital

• Number of editors
• Opening a petition
• Opposition
• Word limits
• No consensus at Re-RFA
• Closing a petition

Nice-to-have

• Administrator inactivity
• Minimum time before petition following resysopping at BN—Can't think of any reason this should matter. The 24 hour hold is in no way equivalent to even 1/7th of an RfA.
• Withdrawn petitions and subsequent petitions—We'll cross that bridge when we come to it if we don't now

Not needed

• Recall discussion—Redundant to Discussion Elsewhere; see arguments there
• Percentage threshold to pass—not seeing significant opposition to the current thresholds, even after Graham87's RRfA
• Nominations at Re-RFA—CREEP, precedents including Graham87, Floquenbeam, and Fram

Sincerely, Dilettante 21:07, 20 November 2024 (UTC)

Moved from Wikipedia:Administrator recall/Reworkshop § Deciding which questions to ask

We need to start moving towards deciding which questions we are going to ask. We shouldn't restrict it to an arbitratary number (as was suggsted on my talk page) but we do need to ensure that we ask all the important questions without including ones that aren't. I've laid out my brief first thoughts below, which are based on my opinions as informed by the discussion that has happened regarding them. They are presented in the order they appear on the page, which order they are asked in the RFC is the step after we agree which questions to ask. I don't think there are any additional potentially important questions that haven't been raised on this page, but it's likely that unless they are added in the next couple of days they won't become apparent before a new petition is initiated. Thryduulf (talk) 14:05, 20 November 2024 (UTC)

• Opening a petion (3 questions including time delay)
  • All three questions are important, but more workshopping is needed before it's ready for an RFC. Thryduulf (talk) 14:05, 20 November 2024 (UTC)
• Number of editors
  • There is probably appetite to ask this question, but it is dependent on the RFC on duration concluding first. Thryduulf (talk) 14:05, 20 November 2024 (UTC)
• Explanation of signatures
  • Not needed IMO, but what things other than a signature may be included with it and what should be in discussion may be needed, depending on the outcome of other questions. Maybe this could be folded in to the recall discussion question? Thryduulf (talk) 14:05, 20 November 2024 (UTC)
• Oppostion
  • This is definitely needed, but it needs more workshopping. Thryduulf (talk) 14:05, 20 November 2024 (UTC)
• Recall discussion
  • This is needed I think. Thryduulf (talk) 14:05, 20 November 2024 (UTC)
• Word limits
  • Another important question. Thryduulf (talk) 14:05, 20 November 2024 (UTC)
• Closing a petition (3 questions)
  • Question 1 clearly needs to be inclued, questions 2 and 3 need answers but maybe not at the RFC (more discussion on that is needed). Thryduulf (talk) 14:05, 20 November 2024 (UTC)
• Minimum time before petition following resysopping at BN
  • Important and simple. Possibly good to go as is? Thryduulf (talk) 14:05, 20 November 2024 (UTC)
• Adminstrator inactivity (2 questions)
  • No clear consensus on whether this is needed, but I favour asking. Thryduulf (talk) 14:05, 20 November 2024 (UTC)

  This question largely seems to become relevant if the community answers question 1 (Under what circumstances may a petition be opened?) in a manner that does not limit when a petition could be brought forward. If the community believes that we should limit petitions after a discussion on an administrator noticeboard, or similar, then it is very unlikely an inactive administrator will have a petition filed against them. So, I think this question raises some theoretical questions, it becomes more complex as we would need to also try to define what is meant by inactivity. - Enos733 (talk) 07:01, 22 November 2024 (UTC)

• Percentage threshold to pass
  • This might need to be asked later, but not in this RFC. Thryduulf (talk) 14:05, 20 November 2024 (UTC)
• No consensus at Re-RFA
  • This is important to get clarity on and it needs to be an RFC-level consensus. Thryduulf (talk) 14:05, 20 November 2024 (UTC)
• Withdrawn petitions and subsequent petitions
  • Needs answering, but needs more input regarding whether it needs an RFC or not. Thryduulf (talk) 14:05, 20 November 2024 (UTC)

@Thryduulf: could you please take this to talk, where we're already discussing it? theleekycauldron (talk • she/her) 14:08, 20 November 2024 (UTC)

Didn't see that. Thryduulf (talk) 14:32, 20 November 2024 (UTC)

Moved discussion ends here. Sincerely, Dilettante 16:58, 20 November 2024 (UTC)

I wholeheartedly disagree both that this RfC is getting too long, and that it's "only for problems that need to be fixed in the next month." Where are people pulling these ideas from? ⇒SWATJester ^{Shoot Blues, Tell VileRat!} 19:02, 20 November 2024 (UTC)

From my experience wrangling Recall Phase 2, I found the number of different tweakable settings and the depth of options in each of those questions to be the part that caused confusion for editors the most. No single person (not even myself) was fully clear what the final structure of RECALL would look like, even vaguely, because "25 signatures in 30 days, discussions allowed" is vastly different from "50 signatures in 6 months, no discussion allowed" and so on.

The participation suffered a bit due to it, because this lack of overall clarity meant editors who want to engage had to dissect every aspect of it we were tweaking around. Now that we have a basic structure for RECALL, it makes much more sense to restrict questions to the most important most changing factors. And once those are set, to tweak around the dependent questions at a later date. An RFC with 5 primary questions is inherently much easier to grasp than with 15; and I'd rather reduce the number of people clamoring for "This process is a mess, burn it all" when we have a real shot at a healthily working RECALL process Soni (talk) 19:14, 20 November 2024 (UTC)

As I said about the discussion for the admin elections RfC, I appreciate there's a tension between trying to ensure that everyone feels the concerns most important to them are being worked on, while also trying to make a discussion effective by focusing on a narrower set of questions in order to maximize engagement. I disagree that we should put everything into one RfC at all costs just to avoid another one. I think it's reasonable to try to group questions into areas of highest concern, or with interrelated dependencies, and then to have separate RfC phases. Each should be a manageable size with which most potential participants will be willing to follow and comment on for a significant portion of time. I realize, though, that this is tricky balance to achieve. isaacl (talk) 19:51, 20 November 2024 (UTC)

I appreciate that context, and I think there's probably a middle ground solution that avoids an "everything and the kitchen sink RfC" that would also address the concerns about having sufficient opportunity to voice criticisms and identify deficiencies of the process. For instance, I'd be less concerned about getting any individual question in if the final product made an up-front commitment to a second workshopping at some point in the future to deal with those issues that got left off. That'd allow for prioritization without disenfranchising people from a process that's already been criticized as having been insufficiently participated-in and not surfaced as well as it could have been. It would also mitigate my concern that we're going to hear "well why didn't you include that in the reworkshop RFC" in the future for anything that doesn't make it, if we know for sure that there will in fact be a future opportunity to address it. Additionally, ensuring that the proposals for an RfC questions are non-variable options (such as Question #1, with three distinct, unambiguous choices) as opposed to those that require further specification (such as Option 3 of "Recall Discussion" which requires you to specify a location) or allow for multiple selections (such as "Closing a Petition" where the ability to overlap multiple options complicates things). ⇒SWATJester ^{Shoot Blues, Tell VileRat!} 23:32, 20 November 2024 (UTC)

I would be happy with a future RFC or discussion designation for anything that hasn't been confirmed here.

I don't think we need an RFC for every specific thing we'll hash out. Nearly everybody who has weighed on "Closing a discussion", for example, agrees. So I believe that we can just straight up use regular consensus to parse that.

In my opinion handled by a future RFC/discussion" should be left for broader dependent discussion, without necessarily committing ourselves to only an RFC, even if it's not necessary. Soni (talk) 02:50, 21 November 2024 (UTC)

Grouping It wold both simplify voting and make the outputs of the reworkshop clearer if we grouped these by theme: proposals to raise the bar to ensure recalls are used judiciously, proposals to clean up the process to ensure recalls run smoothly, etc.

There's been a mostly unpspoken consenus in the reworkshop that recalls should be reigned in (which I respectfully disagree with). I'm totally fine letting the community decide on that, but we shouldn't bury the lede and not make the intent behind some of the proposals clear. - RevelationDirect (talk) 13:32, 21 November 2024 (UTC)

The ordering of the questions is something that we should think about, but not yet. We need to agree what questions we want to ask first, and only then will discussing ordering be usefully spent time. Thryduulf (talk) 14:24, 21 November 2024 (UTC)

• Noting that I have added "General Process" on my listing. I do not believe that question should be asked. Others should add it at the appropriate place in their list, else we'll go through "add it/no?" again when the final sections are being chosen.

As an aside, I'd quite like to close some sections that seem to show no support for being asked right now. We have too many sections (and more are being added every day). A bikeshed benefits nobody. Soni (talk) 05:14, 23 November 2024 (UTC)

Edit notice

We seem to be getting more and more people mistaking this for an RFC. Would there be any objection to setting up an edit notice to remind them that it isn't? Thryduulf (talk) 01:20, 21 November 2024 (UTC)

Realistically, how much can editors be expected to distinguish between "here's an important question" and "here's a change I agree with"? We have several proposals here to raise the bar to ensure that recalls are used judiciously that, until I prodded, was a mostly an unspoken consenus because it wasn't clear that could be a controversial position.

Certainly, we can't turn this into a proxy RFA where every choice is hashed out. But I also don't think it's inappropriate to say "I personally like this change so let's ask it" instead of hinting that there are questions just floating in the ether that we're transcribing. RevelationDirect (talk) 13:44, 21 November 2024 (UTC)

But that doesn't mean you object to an edit notice, correct? Valereee (talk) 12:16, 22 November 2024 (UTC)

Not at all. I just want to make sure that doesn't get unintentionally used to limit the range of perspectives. RevelationDirect (talk) 12:48, 22 November 2024 (UTC)

I don't see any reason why we shouldn't put an edit notice on. May not help, but won't hurt. Valereee (talk) 12:17, 22 November 2024 (UTC)

After another couple of instances, I've created the edit notice. Please improve. Thryduulf (talk) 21:46, 22 November 2024 (UTC)

Waiving the 12-month immunity period after a successful Re-RFA?

Moved from WT:RECALL § Waiving the 12-month immunity period after a successful Re-RFA?

 – Levivich (talk) 18:24, 22 November 2024 (UTC)

At our first Re-RFA (Wikipedia:Requests for adminship/Graham87 2) one of the points brought up by opposers that if the Re-RFA is successful, then there is a 12-month period in which another recall is not available as a means to desysop. Should it be possible for admins to offer to waive this right against new petitions to try and convince the community that this accountability mechanism will be available if new circumstances arise that might warrant one? That might make it easier for Re-RFAs to succeed, though it might also erode the right if it becomes expected at Re-RFA. Thoughts? -- Patar knight - ^chat/_{contributions} 02:08, 19 November 2024 (UTC)

One of the problems RECALL was proposed to fix, was that potential admins were often being pressured into giving up to an unassorted set of demands just to get/keep their bits. These were unenforceable, hence the problems with voluntary recall; but I also believe it broadly suffered from being "extracted" by one or two editors as opposed to any reasonable community consensus.

Hence I broadly oppose any process that attempts to informally waive away admin rights, even if we consider this enforceable. Admins can always choose to voluntary stand for recall for any reason they prefer. I see no reason to let people strip away the "right against new petitions" on top of that. Soni (talk) 02:29, 19 November 2024 (UTC)

I agree with Soni. We should not be encouraging hounding of admins, even if done with good intentions, and allowing admins to waive their right to 12 months without a petition could encourage editors to hound admins into doing so. Thryduulf (talk) 03:15, 19 November 2024 (UTC)

I don't think Thryduulf/Soni's logic will actually work - the community seems to want the ability to make binding promises of one's future self at Graham87's RRFA even though there's currently no way of making them, so we may as well give them their wish. See my lengthy comment at Special:Diff/1258337787. * Pppery * _{it has begun...} 04:09, 19 November 2024 (UTC)

And people wonder why RfA is so unbelievably toxic and hasn't been fixed despite a myriad of tries that the damn debate should be at WP:PEREN. —Jéské Couriano v^_^v ^{threads critiques} 17:28, 19 November 2024 (UTC)

Not sure how "without a petition" would be relevant if the recall petition has already met the threshold and the admin is in their RRFA. If there's a concern of hounding, waiving the immunity could be limited to admins with an active RRFA. No one would be stripping away other's rights, it would be a decision for the admin undergoing a RRFA. -- Patar knight - ^chat/_{contributions} 16:10, 19 November 2024 (UTC)

This all refers to the fact that an admin who retains the tools at a re-RFA cannot be subject to a second petition within the next 12 months. So it applies to the period after the Re-RFA has concluded, not during it. Thryduulf (talk) 16:17, 19 November 2024 (UTC)

Yes, that's how the immunity period works currently works. I'm not sure how that's relevant to my reply, which simply notes that if an admin is currently in an RRFA, then by definition they are not in a situation where they are being hounded to do anything "without a petition" since an RRFA requires a successful petition in the first place. Many opposers at the first RRFA have explicitly cited the lack of recall, even if they otherwise might have supported. Admins in an RRFA should be allowed to try and win over those people by addressing their concerns, but they currently cannot even if they wanted to. -- Patar knight - ^chat/_{contributions} 21:01, 19 November 2024 (UTC)

All that twelve months of 'immunity' means in practice is that for that period the admin can only be desysopped via ArbCom, which is the same situation all admins have been in for the twenty one years before last month. – Joe (talk) 16:30, 19 November 2024 (UTC)

I think Graham's RFA shows that's not true. There's subset of RRFA participants who would support if they had the assurance that a recall petition could be started again within 12 months if there are new issues, but since that's not the case, they feel compelled to oppose. Not allowing admins to waive this right means admins whose behaviour falls between a clear consensus either way are more likely fail their RRFAs even if they would have been willing to waive the right and would otherwise have been net positives and on a tight leash to avoid the problematic behaviours. -- Patar knight - ^chat/_{contributions} 20:52, 19 November 2024 (UTC)

I agree with Soni that nothing prevents admins from pledging to follow their own personal recall process. The community has to weigh how much they trust the admin to follow that process. isaacl (talk) 22:37, 19 November 2024 (UTC)

If an admin creates a personal recall process that says anyone can start a recall petition at any time, they would be barred from following that pledge if one of the immunity periods applies. -- Patar knight - ^chat/_{contributions} 15:46, 20 November 2024 (UTC)

They wouldn't per NOTBUREAU, IAR, and general precedent that we allow leeway towards self-imposed recall processes. Sincerely, Dilettante 16:43, 20 November 2024 (UTC)

I agree. This is (in part) why I asked (a while ago) the question whether an administrator could voluntarily opt into this recall process (including the different threshold for retaining administrative privileges). I think as this process was created, any administrator can voluntarily opt in. - Enos733 (talk) 17:03, 20 November 2024 (UTC)

Sure, if thi is the case, then we should just make it clear that it is available instead of misleading people that it is not (e.g. "Administrators can choose to waive this immunity period."). As the Graham87 RRFA shows, there is a subset of !voters who do care about this and we should be telling admins at RRFAs that this is an option to try and convert participants. -- Patar knight - ^chat/_{contributions} 17:22, 20 November 2024 (UTC)

I mean, they wouldn't really be barred from it? An admin can request (or even start) an RFC on their conduct, promising to step down if the results are negative; and can then choose to step down, at any time, for any reason. It would just be completely nonbinding up until they actually stepped down (after that point I think any reasonable interpretation would be that they'd resigned under a cloud and couldn't request the bit back without a new RFA.) Anyway, as far as whether they can agree to binding things, I'm inclined to say no, because we don't want elaborate webs of binding agreements extracted by a single editor during high-pressure processes; but I don't think that editors should be forbidden from requesting non-binding agreements or commitments, and I do think that ArbCom or a later legitimate recall could, as part of a larger process, take into consideration whether an admin has adhered to such non-binding promises. It is a grey area (and perhaps it should be) - I think that if the promises were manifestly reasonable, and there wasn't any real rationale to break them, then breaking them even if they were non-binding is potentially unbecoming conduct and could tip the scales in close cases later on. This means that while such promises are not binding, they are not completely weightless, either (which I think is as it should be - we don't want elaborate webs of binding promises; but an admin candidate shouldn't glibly promise whatever they think will get people to agree with no intention of following up, either.) --Aquillion (talk) 17:32, 20 November 2024 (UTC)

The bar to opening a recall petition needs to be much, MUCH higher

Moved from WT:RECALL § The bar to opening a recall petition needs to be much, MUCH higher

 – Levivich (talk) 18:28, 22 November 2024 (UTC)

What an absolute disaster the 2nd-ever recall attempt has turned into. There are a few stark differences between Fastily's case and the prior one:

1. Fastily had exactly one recent ANI case opened about them [1]. All other accusations of behavioral issues were from 2012 or earlier.
2. A mere 12 hours elapsed between the opening of this ANI thread, and the opening of the recall petition
3. Despite some rightful concerns, the ANI thread seemed nowhere near conclusive or widely damning of Fastily's behavior.
4. Even still, Fastily accepted a rebuke, and promised to adjust their behavior.

Now the petition was closed as successful, and Fastily has decided that RRFA isn't worth their time, and handed in the bit.

I'm a fairly new user around here, and I was not following the ANI case or recall closely. But does anyone else think that this was an absolutely absurd outcome? Isn't one of the main points of ANI to accomplish a shift in behavior? Why, then, did we need to go through recall if Fastily was never given a chance to shift their behavior? At least in the case of Graham87, there had been multiple recent ANI thread, all with seemingly stronger consensus against the behavior, and repeated re-occurrence of similar behavior.

IMO the bar for admin recall needs to be tied in some way to failed ANI threads. If several people bring up concerns at ANI, then several editors agree at ANI these are concerns, and these concerns are not addressed several times - THEN a recall petition could be allowed.

(as a minor point - the bar of 25 EC signatures needed to confirm a petition also seems too low. But perhaps if the bar for opening the petition were raised, this bar to successfully close a petition wouldn't be such a concern). PhotogenicScientist (talk) 17:31, 20 November 2024 (UTC)

My two biggest issues with the process, in its current form, are the number of signatures required (should be 50, at least, based on our user base) and the lack of opportunity for users to improve on the behaviour that's been deemed problematic. The first problem can be solved, but I'm not sure how to address the second one. Hey man im josh (talk) 17:34, 20 November 2024 (UTC)

Part of the problem requiring X users to sign onto something here is the ever-changing size of the active user base, in any given topic area. It's an incredibly non-robust way to build a process - but at the current time, 25 signatures seems an okay level to me, based on things like ANI/AE participation levels.

As for solving the second problem, what do you think of hinging this somehow on failed ANI threads? PhotogenicScientist (talk) 17:39, 20 November 2024 (UTC)

In both of these cases, the petition's threshold succeeded in its intended purpose, namely only allowing RRFAs to be initiated if there is a strong chance they actually fail. In Graham87's case, an RRfA actually was run, and was set for failure (<50%) before he withdrew. Mach61 17:36, 20 November 2024 (UTC)

if there is a strong chance they actually fail. Hard disagree here. Graham's RFA accumulated 275 votes (in <6 days). 25 signatures represents 9.7% of all votes in that RFA - hardly what I would call showing a strong chance of failure, right out the gate.

Sure, more opposes came in based on the same rationale as those 25. But kicking off an RRFA because you have 15% of the votes needed to desysop? (25 in hand, 165/275 needed for 60%) PhotogenicScientist (talk) 17:45, 20 November 2024 (UTC)

@PhotogenicScientist It is entirely possible to determine an admin has a significant chance to fail their RRFA using a petition threshold significantly below the amount of people that would actually need to oppose them in it. RRFA's are advertised on a greater level (CENT, watchlist notice) than petitions are, and users have greater motivation to vote on RRFA (both to support and oppose), since there's an immediate threat with a definite conclusion. The underlying principle isn't that different from how polling firms can use representative samples of a few hundred to get information about millions of people. Whether 25 signatures over 30 days specifically is ideal can be determined at a future date. Mach61 18:39, 20 November 2024 (UTC)

The issue with your representative sampling example is that this petition is not similar enough to RFA to be a proper sample. Namely, people can't provide Supports. So a recall petition with 100 positive comments and 25 negative comments would still clear the bar for sending us to RRFA, for a case that would in all likelihood result in a pass. PhotogenicScientist (talk) 20:45, 20 November 2024 (UTC)

@PhotogenicScientist I agree that petition signatories aren't a representative sample, which is why I merely said the underlying principles between sampling and the petition process are similar. Anyhow, I would argue, from the limited data we have, that it is unlikely a petition with a very low % of support from the community would get the attention necessary to get 25 signatures from the small % of editors that do support it (whether or not 25 supporters/30 days makes that scenario unlikely enough is again an issue that can be discussed later). Also, the ratio of signatories to people who left negative comments isn't a great heuristic for how things will go down; formal opposition and counterarguments in favor of the admin are really meant for the RRFA, and anyway, Graham87's petition, which lead to a RRFA that was clearly headed towards a desysoping, probably had more unique editors leaving negative comments than the 27 signatories. That was because it got extra attention by being the very first petition, and since most of the negative comments were made before Graham shot himself in the foot, but it's easy to imagine a scenario in the future where an admin has a strong core support base who swarm the petition page, but who ends up being losing an RRFA once their issues are brought to the wider community. Mach61 23:52, 20 November 2024 (UTC)

This should be discussed at WP:REWORK. Levivich (talk) 17:41, 20 November 2024 (UTC)

I'd rather gather some input from others a la WP:RFCBEFORE before going that route. PhotogenicScientist (talk) 17:48, 20 November 2024 (UTC)

After the edit conflict, RFCBEFORE is less applicable... PhotogenicScientist (talk) 17:51, 20 November 2024 (UTC)

(edit conflict × 3) See Wikipedia:Administrator recall/Reworkshop#Opening a petition for ongoing discussion about requirements for opening a petition, and Wikipedia:Administrator recall/Reworkshop#Number of editors for discussion of how many signatures are needed. Thryduulf (talk) 17:41, 20 November 2024 (UTC)

I think the number should be higher, and the time the petition stays open should be shorter. I think these things are going to be RfC'ed. Then again, I think this entire process should be repealed. --Tryptofish (talk) 20:32, 20 November 2024 (UTC)

@Tryptofish do you think it'd be beneficial at all to consider the number or outcomes of recent ANI threads (or threads at similar noticeboards)? PhotogenicScientist (talk) 20:46, 20 November 2024 (UTC)

Maybe I misunderstand your question (sorry if I do), but for me the most relevant metric is whether there are requests for arbitration that fail to lead to someone being desysopped, when they should have been. I guess there is also a question of whether editors are reluctant to go to ArbCom when they really should go there, but it would have to be accompanied by evidence that, if only they had contacted ArbCom, ArbCom would have found legitimate cause. --Tryptofish (talk) 21:18, 20 November 2024 (UTC)

I think we may be both hitting on a similar theme - recall should only be allowed after other attempts to address admin behavior have failed.

One such avenue that can be tried is ArbCom - though I personally think ANI is a better place to start with any admin issue. If the issues repeatedly persist, or if one issue is reprehensible enough on its own, going to ArbCom makes a lot more sense to me. But if those efforts end with no sanctions and/or no changes in behavior, only then may recall be initiated. PhotogenicScientist (talk) 22:02, 20 November 2024 (UTC)

See Wikipedia:Administrator recall/Reworkshop#Opening a petition for existing discussion around this idea. Thryduulf (talk) 22:10, 20 November 2024 (UTC)

Thanks for the clarification. I was actually trying to explain that I think evidence is lacking, that we need recall in addition to ArbCom. --Tryptofish (talk) 22:23, 20 November 2024 (UTC)

Perhaps. Though I was hoping to use this thread to discuss ways to improve one particular aspect of recall, rather than ask for its repeal.PhotogenicScientist (talk) 00:09, 21 November 2024 (UTC)

Going to ArbCom should be a last resort for anyone. One of the reasons for the sudden popularity of WP:RECALL is no doubt that for many editors it is preferable to going to ArbCom. And I don't blame them. Hawkeye7 (discuss) 23:05, 20 November 2024 (UTC)

I didn't have an issue with Graham87's recall. I have a massive problem with Fastily's. Meanwhile, I've just noticed that the RfC that created RECALL was closed by User:Maddy from Celeste [2], and over at WPO the user claiming to be Maddy from Celeste posted (when Graham87's recall started) We have a first petition, 26 hours after my close. The purge has begun? with a dancing banana GIF emoji ... I'm glad it's fun for some people. Black Kite (talk) 23:03, 20 November 2024 (UTC)

What was the massive problem that you had? Was it with the process? Hawkeye7 (discuss) 23:08, 20 November 2024 (UTC)

Pretty much what the OP here articulated. I don't see a major misuse of tools, which is what I would expect from a recall. I also note that the editor who started the recent ANI on Fastily also bludgeoned the opposes on Graham87's recall, but I couldn't possibly assume bad faith, could I? Black Kite (talk) 23:18, 20 November 2024 (UTC)

I concur. WP:RECALL as written is a trainwreck. Full disclosure: I'm not against holding admins accountable. I've been open to recall for many years. See my user page. And while not something I enjoyed, I did take another admin to ARBCOM resulting in their being desyssoped. The bar here is far too low and Fastily is strong evidence to that end. -Ad Orientem (talk) 16:09, 22 November 2024 (UTC)

@Ad Orientem Simple question: Do you think that, if Fastily did not resign, their RRFA would have a support level well above the 60% passing requirement? Because from my vantage point, that seems rather unlikely, which indicates that the bar for petitions is high enough to fuflfil their intended purpose of filtering out weak cases. Mach61 20:00, 22 November 2024 (UTC)

Yes. -Ad Orientem (talk) 22:34, 22 November 2024 (UTC)

And with popcorn. @Maddy from Celeste, is it fair to say you generally you consider the opening of a recall petition an occasion for happy dances and popcorn? Valereee (talk) 16:57, 22 November 2024 (UTC)

In hindsight, I absolutely agree that was a bit of a daft thing to post. Now the way I intended it was as a sarcastic remark on the situation I found myself in, having gone from "interesting policy rfc at WP:CR; let's close it" to "there's a big drama and people are blaming me". It all felt a bit absurd at the time, a bit like a dancing banana or a plot to purge the admin corps. Now it's clear Poe's law and the general lack of context in my post got the better of me, and ironically enough that has caused more people to blame me. -- Maddy from Celeste (WAVEDASH) 17:30, 22 November 2024 (UTC)

I'm too lazy to go figure out where I made a flippant remark in a serious discussion recently myself and lived to regret it when I was called on it. Lesson learned: serious business is best treated like serious business. Thank you for the clarification. Valereee (talk) 17:34, 22 November 2024 (UTC)

• All other accusations of behavioral issues were from 2012 or earlier. This is just a blatant lie, PhotogenicScientist. Fastily's recall petition specifically closed as passed because of the evidence presented by A Smart Kitten, which was 10 cases of Fastily misusing their admin tools with newbies from this year alone. From the past 6 months alone, even. Silverseren^C 23:54, 20 November 2024 (UTC)

  I suppose I should clarify - all over accusations of behavioral issues prior to recall being initiated.

  My main point on this still stands - do we not want admins to have a chance to defend themselves at a discussion board? Or to have a chance to amend their behavior after even a single discussion board thread closes with consensus found for a rebuke?

  Fasitly got neither of these, considering how soon after the ANI thread the recall petition was started.

  (also, "misunderstanding" sounds a lot nicer than "blatant lie"). PhotogenicScientist (talk) 00:03, 21 November 2024 (UTC)

  Fastily commented plenty in the ANI thread and the recall petition was started because of editors deciding the response did not meet what they expected of admin conduct. And recall petitions also have a Response section entirely for the admin to respond to anything brought up by the petitioners. Fastily just declined to defend any of their actions brought up by the editors in the petition. And they seemingly declined to defend their actions after that point either, despite what was brought up as months, if not years-long, harmful inappropriate admin conduct going far beyond the initial ANI thread topic of their interaction with Clovermoss. Silverseren^C 00:10, 21 November 2024 (UTC)

  the recall petition was started because of editors deciding the response did not meet what they expected of admin conduct. A mere 12 hours later, yes. No real chance was given to let our existing process of ANI work - to garner more community input, and to come to consensus.

  The recall petition is suited by design to being just that - a petition. It is not a good platform for discussion. No process where people can just log votes and log off is. PhotogenicScientist (talk) 00:16, 21 November 2024 (UTC)

  And? 12 hours was enough time for responses from Fastily and the initiator determined that the response wasn't good enough in light of their actions and conduct. And it turns out they were correct in that belief, particularly when it led to a much larger amount of evidence of misconduct being revealed. Furthermore, GreenLipstickLesbian brought up a very relevant point in the ANI discussion after starting the petition after the response there to it, ie that admins in the ANI discussion had determined that "no misconduct had happened" and it was that very circling the wagons around an abusive admin that even more so necessitated the petition happening. Both petitions that have occurred thus far have massively revealed a wealth of abuse and misconduct on the part of the admins in question, showcasing that recall was the right call in the first place. Silverseren^C 00:29, 21 November 2024 (UTC)

  It seems widely acknowledged (most recently here) that bringing Admins to ANI never amounts to anything (at ANI).

  The adminship process also allows you to log votes and log off, I assume you'll be petitioning to change that too? OXYLYPSE (talk) 22:12, 21 November 2024 (UTC)

  ANI can amount to plenty considering it resulted in the petition against Fastily. Hey man im josh (talk) 22:42, 21 November 2024 (UTC)

  I strongly disagree, but am struggling to word this in a nice way, so I'll stop. OXYLYPSE (talk) 07:08, 22 November 2024 (UTC)

  The adminship process also allows you to log votes and log off What a shining example of a functional and pleasant process you've proferred. I suppose as long a RECALL is just like RFA, everyone will absolutely love it. PhotogenicScientist (talk) 23:11, 21 November 2024 (UTC)

  Fastly had a chance to defend themselves in the RRfA. They declined to do so. If your goal is to come up with a process that would have somehow saved Fastly then I have to strenuously disagree. The entire purpose of a WP:RECALL process - and this is something that has always been controversial, but which was ultimately settled when the process was made - was that WP:CCC applies to admins as much as anything else and that an admin who no longer has the support of the community should no longer hold the bit. The purpose of the petition state is solely to filter out recalls that have no chance of success, not to force a "cooling off" period or to give admins "time to change" or anything like that; if an admin wants those things, they can make that argument in the RRfA, allowing the community to weight it on a case-by-case basis as appropriate. -- Aquillion (talk) 04:42, 21 November 2024 (UTC)

  There is an argument for "time to change" and a "cooling off period", but in the form of a minimum gap between a noticeboard thread and a petition (see WP:REWORK#Opening a petition and the time delay subsection). There is also an optional "cooling off" period in that the admin has up to 30 days after the closure of a petition that meets the threshold to stand (or choose not to stand) for re-RFA. Graham87 waited 11 days to start their re-RFA. Fastily resigned 4 days after their petition closed. Thryduulf (talk) 05:12, 21 November 2024 (UTC)

  Graham87 didn't have the luxury to wait. It was brought up during the petition that he didn't have much time to do an RRfA in December: "For what it's worth if an RRFA does end up being required, I would very much prefer it to start as soon as possible; December will be a relatively busy month for me IRL." + "I'm aware of that. I'd want the RFA to be held as close to 27 November as possible."

  The only modification that I can suggest for RRfAs, with the exception of banning outside links, would be to slightly extend the RRfA period to 40-45 days. That way there is a slightly decreased rush for an admin involved while not keeping it pending for too long. --Super Goku V (talk) 06:54, 21 November 2024 (UTC)

  That wouldn't have helped me. I needed the albatross of my admin status off my neck by early December, at the very latest, preferably late November. Graham87 (talk) 11:38, 21 November 2024 (UTC)

  Thanks for sharing your perspective. (Not sure I could read through these comments and cooly provide clarifications.) RevelationDirect (talk) 00:09, 23 November 2024 (UTC)

  If your goal is to come up with a process that would have somehow saved Fastly then I have to strenuously disagree I'm not familiar with Fastily's work in the slightest, so I can assure you this isn't my goal. It just feels like with this recent petition, the admin under scrutiny wasn't given a fair shake.

  and this is something that has always been controversial, but which was ultimately settled when the process was made Settled as it may have been, you said yourself - WP:CCC. We should always be open to improving the process. ESPECIALLY since the previous consensus was formed before this procedure was ever actually used. PhotogenicScientist (talk) 14:43, 21 November 2024 (UTC)

The ANI was about recent (within past month) deletions and WP:ADMINACCT. Before the recall petition was started, 20+ editors posted comments, writing 7,000+ words. One of them was Fastily, who made five comments in the first six hours, starting with "of course I'm going to respond the way I did...Are you really surprised you got the response you did?...the last thing I need is abuse from my colleagues" and "Having fun twisting the narrative there? ... I've been both patient and cordial with you, yet you have exclusively responded with vitriol and hostility. Well cool, I'll do the same ..." and "I wasn't the one who went around looking to pick a fight in the first place."

The recall petition was open for 11 days. 84 editors, 82 page watchers [3], over 20,000 words. 25 editors signed it. Top 10 editors by edits: 3/10 signers, 7/10 opposed, leaned opposed, or criticized the process. Top 10 by added text: 7/10 signers, 3/10 opposers.

10 blocks, and over 30 diffs, log entries, or discussion threads, all from 2024, were posted in that discussion, and a greater number from before 2024, including in every year for the last five years, and going back to before 2012.

The petition has an entire section for the admin to respond called "Response". Fastily made 19 edits during the 11 days the petition was open [4], 4 of which were comments to the petition: 1, 2, 3, 4.

Now we can disagree about whether Fastily should be an admin, whether the recall process is good, whether this particular recall petition should have been started when it was or not yet, but nobody can claim that there wasn't thorough discussion, that not a lot of people participated in it, that Fastily didn't have an opportunity to respond, that the signers didn't explain themselves, or that they bludgeoned it, or that opposers weren't heard, or that there weren't a lot of specific, recent, and ongoing examples raised. We are entitled to our own opinions, we are not entitled to our own facts. Levivich (talk) 04:20, 21 November 2024 (UTC)

• I've been thinking about suggesting a "three strikes"-type rule where if N actions by an admin are not endorsed at XRV within a given time period, that admin has to start an RRFA (XRV rather than ANI because the latter is focused on how to respond to an allegedly bad action rather than reaching a consensus about whether the action was bad). I imagine this would be in addition to the petition-based trigger, though. – Joe (talk) 07:48, 21 November 2024 (UTC)

  DRV should be counted as equal to XRV in this regard, which can't review deletions. —Alalch E. 09:10, 21 November 2024 (UTC)

  Good point, probably also MRV, maybe also unblock requests? But it can be hard to distinguish successful appeals from overturned bad blocks. – Joe (talk) 09:20, 21 November 2024 (UTC)

  There are all kinds of reasons why that's a bad idea. It's not going to raise the bar on recall attempts - people will make note of overturnable actions, then formally appeal them in quick succession to get around the time restriction. It'll skew the decisions made at those venues - some people will endorse solely so an admin they like won't get driven off the project, some will overturn solely so an admin they dislike will. It'll disproportionately affect the admins most active in these areas - if you block a thousand spammers in a month and one percent are wrong, you get driven off the project; if you're one of five hundred admins who block two each and you get half wrong, you're a-ok. And it'll further discourage anyone from closing difficult discussions that would've gone to review either way. —Cryptic 09:56, 21 November 2024 (UTC)

  I don't think we can or should make policy based on the assumption that people will try to abuse it. There's no such thing as a gaming-proof process. And remember, the idea here is to trigger an RfA, not a desysop; that is where questions of e.g. acceptable error rate would be addressed. – Joe (talk) 10:02, 21 November 2024 (UTC)

  I'm sorry, but it is effectively the same thing. No-one whose recall petition reaches 25 votes is going to pass an RRFA, even at 60%, let's be realistic here. Black Kite (talk) 10:13, 21 November 2024 (UTC)

  We're not talking about the petition here. – Joe (talk) 10:21, 21 November 2024 (UTC)

  It applies equally to the alternative method (though I don't think the mostly moribund XRV is the way to go either). Black Kite (talk) 11:32, 21 November 2024 (UTC)

  XRV is a noticeboard I wasn't even aware of. As far as I knew, if someone has a problem with an admin's actions (and realistically, also their conduct after performing those actions) they started a thread at ANI.

  And I agree that the process seems pretty unfair at the moment. But I disagree that necessarily no-one whose recall petition reaches 25 votes is going to pass an RRFA, even at 60%. As I laid out in Graham's case above, those 25 votes represented just 9.7% of all votes in that RFA, and just 15% of the votes that were required to desysop. I get that these 2 petitions have attracted massive discussion, and have acted kind of like mini-RRFAs already, but they're set up in a way that is a bit antithetical to that. Petitions can absolutely pass against admins who have very many supporters in their corner.

  @Joe Roe: see also, WP:BEANS. PhotogenicScientist (talk) 15:20, 21 November 2024 (UTC)

  I don't favour using a count of administrative action review discussions as a hard threshold for starting a re-request for adminship discussion. The administrative action review process was designed to be a no-fault examination of what is the most appropriate action for a given scenario. The discussion outcomes can be used in a separate discussion about what consequences should ensue, if any. isaacl (talk) 22:55, 21 November 2024 (UTC)

  I'd be concerned that a simple numeric threshold would snag up the most active admins. (A percentage might hypothetically work but would be too complex to manage.) RevelationDirect (talk) 00:15, 23 November 2024 (UTC)

You're telling me the first 12 hours of an ANI thread, between the filers and the accused party, did not start off with any hearts and minds being changed? Frankly, I'm shocked.

Still, after 41 hours of discussion, with 21 users contributing, Fastily accepted a rebuke, and said they'd do better. Of course, they were never given the chance to prove themselves - the recall petition had already been opened 29 hours prior.

And I stand by my point above - Recall in its current format functions more like a petition, and less like a discussion. That discussion happens there doesn't really change the fundamental way it's set up. If 25 users come by and register votes, it passes. No discussion required.

Contrast that to ANI, where discussion and consensus-forming is absolutely required. As an uninvolved editor who hadn't been following either admin's cases, but still cares to participate in processes like RFA and RECALL to improve the encyclopedia, two things help me make a choice in this process: 1) What is the severity of their bad behavior, and 2) Have they been given chances to rectify it already. One ANI thread with a particularly nasty abuse of power may satisfy the first condition on its own - but typically, we should be giving people more than a single chance to respond to community criticisms. PhotogenicScientist (talk) 15:08, 21 November 2024 (UTC)

Yes, recall is specifically a petition and it should be, that is by design. Then you go to the RRfA, assuming it gets confirmed, where votes and discussion happen like a normal RfA but with a much lower bar to jump over. But also looking at the last two petitions there was loads of discussion there. Honestly everything you describe seem to be working as intended and as it should be. Even the 25 people seem about right. PackMecEng (talk) 01:23, 22 November 2024 (UTC)

• Comment - I think the issue for a lot of people is the process for passing RRFA petitions was something that just went on and on and on and on and most people just lost interest in it and assumed nothing would ever happen, in spite of notifications which I think a lot of people just ignored since they had seen them so often, and then it just plopped out in its present form and now a lot of people are unhappy. I've got to be honest, I also don't like the look of what happened in the Fastily case where it looks like what should have been dealt with at ANI just turned in to straight desysopping, with no real discussion either.

I think we really need to look twice at these super-long processes that serve to exhaust the community. I don't like, for example, the insistence that any re-examining of RRFA should necessarily happen through yet another exhaustive process with re-workshopping. FOARP (talk) 20:52, 21 November 2024 (UTC)

I very much agree with what FOARP said. Also, I was constantly frustrated with what seemed to me to be many knowledgeable editors who simply did not bother to raise objections when there was an opportunity, and only seem to have taken an interest when it was already too late. Combine with that, there were others who felt that we simply need to have a community-based system, just because, and who were willing to push things through without really making the effort to get a solid base of community consensus, and who now argue (with some validity) that the time for second-guessing has already passed. When I see comments like some that were posted today, to the effect that since both petitions resulted in the admins no longer being admins, that means the process is working the way that it's supposed to, I feel like I'm in some sort of bizzarro world. --Tryptofish (talk) 21:06, 21 November 2024 (UTC)

Yeah, I felt that way myself, exhausted by it and unaware of specific developments. It really is exhausting trying to keep up with some stuff while also doing work and contributing to Wikipedia outside of what you feel you need to be informed on. Not that I assumed nothing would happen, but man it dragged on. Hey man im josh (talk) 21:08, 21 November 2024 (UTC)

I completely understand the difficulty in maintaining sufficient focus to participate in lengthy discussions (I believe the longest delay for the recall phase 2 discussion was waiting for someone to evaluate the outcome). The key challenge is that there are many interested editors who really want to be able to weigh in (typically dipping in and out based on their availability or degree of current focus), and the number of potential interactions go up as the square of the number of participants. The catch-22 is that English Wikipedia's current free-for-all discussion-based decision-making process can only be changed by... going through that same process. If the community wants to keep making decisions this way, then discussions about major changes will continue to take up a considerable amount of time, and the outcomes more strongly influenced by the much smaller number of people who can afford to engage throughout. isaacl (talk) 23:11, 21 November 2024 (UTC)

Indeed, I was hibernating for much of the period, and when I saw all the confusing notifications about it (RfCs to decide whether the RfC had decided...), thought it must had already been decided, so there was no point in spending a couple of days wading through the walls of text to participate. Espresso Addict (talk) 00:27, 22 November 2024 (UTC)

You're probably going to think it's ironic for a supporter of NSPORTS2022 like me to be complaining about a long process that ends up plopping out a result they didn't expect, but here I am I guess ¯\_(ツ)_/¯ FOARP (talk) 11:03, 22 November 2024 (UTC)

I can't help thinking that the recall and RRFA should be combined. Allow opposes to the recall and then if the supports have a majority of some kind the admin just loses the bit. One step rather than two, less process and paperwork. If the editor then still wishes to stand again they can use the normal RFA or election process. -- LCU ActivelyDisinterested «@» °∆t° 12:01, 22 November 2024 (UTC)

That would mean that literally anyone could start a RRfA, which would obviously be way more problematic. Chaotic Enby (talk · contribs) 12:53, 22 November 2024 (UTC)

So, it sounds like the bar to starting a weeks-long process that consumes so much community time and asks people to weigh in by Supporting or Opposing should be much higher than "literally anyone"... PhotogenicScientist (talk) 13:03, 22 November 2024 (UTC)

Any extended confirmed user can launch a petition but it takes 25 extended confirmed users to launch the support/oppose phase. PackMecEng (talk) 15:22, 22 November 2024 (UTC)

Is that right? Because I see 20+ bolded opposing votes at this petition along with 132,000 bytes of discussion, and quite a bit of opposition logged at this petition in the 166,000 bytes of discussion.

Both petitions thus far have functioned nearly exactly like mini-RRFAs.

Which in my estimation is exactly why Fastily handed in the bit after their petition closed, rather than initiate an RRFA. They'd already been subjected to 11 days of extensive discussion over their fitness for adminship. That sounds just like an RFA, except it lasted 4 more days, and Oppose votes held no sway over the outcome. PhotogenicScientist (talk) 15:35, 22 November 2024 (UTC)

@PhotogenicScientist If you think excessive discussion makes the petition phase worse, you're free to express that opinion in #RfC: Should we add text prescribing just signatures, no discussion? above :) Mach61 15:39, 22 November 2024 (UTC)

Beat me to it. There is a discussion already to address that. But I will note that during the petition phase anything short of a support holds no weight. Also a lot of the opposes were focused on opposing the process. So the opposes dont really matter. PackMecEng (talk) 15:41, 22 November 2024 (UTC)

The talk page of the project page for the process is a perfectly acceptable place to discuss the process. Rather than telling me I've submitted my form in the wrong location, it'd be more helpful for you to engage with the merits of the issue. PhotogenicScientist (talk) 15:43, 22 November 2024 (UTC)

Nah, its discussed elsewhere. You are just splitting that discussion for no reason which is confusing the issue. PackMecEng (talk) 15:45, 22 November 2024 (UTC)

@PhotogenicScientist if you want to engage with the merits of the issue then please contribute to the existing discussion where the merits of the issue are being engaged with. Thryduulf (talk) 15:46, 22 November 2024 (UTC)

I don't see the validity in saying anyone could start an RRFA, it is the outcome not the process that counts. It takes many EC editors to finalise the process, whether it was one stage or two. Also it seems that one recall petitions has effectively served as an RRFA just requiring 25 EC editors, while the other has been a timesink by running two processes where one could do the job. -- LCU ActivelyDisinterested «@» °∆t° 16:54, 22 November 2024 (UTC)

I am not really an ends justify the means kind of person. So the process is important as wwll. Also, while that was the effect of one that was also their choice and not the processes fault. PackMecEng (talk) 17:49, 22 November 2024 (UTC)

Due process could still be done without having to run separate processes, doing so just creates bureaucracy. As to choice this came out of a discussion specifically created to detoxify the RFA process, so saying editors may not want to go through what is widely regarded as a toxic process is hardly contentious. -- LCU ActivelyDisinterested «@» °∆t° 20:12, 22 November 2024 (UTC)

Abolish Immunity Period & Require 100 Signatures

Moved from WT:RECALL § Abolish Immunity Period + Require 100 Signatures

 – Levivich (talk) 18:30, 22 November 2024 (UTC)

A lot of participants in Graham's RRfA were risk averse, even though the very purpose of the recall process is to lower the bar for RfA. A lot of participants were willing to support Graham, if the topic ban can be enforced. If we really want to protect admins against rash desysopping, and truly make adminship no big deal, we must offer less severe alternatives.

It should be no big deal to give someone the mop, RfA or RRfA, under a cloud or not. It should also be no big deal to sign a petition, currently the bar of 25 is so low that people are equally risk averse to signing it.

Requiring 100 signatures mean people are less risk averse to sign, less risk averse to remove immunity period, and meeting only half the threshold would serve a good admonish function as there is no immunity.

In case of ambiguity, I am in favor of abolishing all immunity, immunity always creates risk aversion. Kenneth Kho (talk) 16:26, 22 November 2024 (UTC)

@Kenneth Kho: Recall won't lower the bar for RfA because RfA as it is is already a Hell Week and nothing about that will change short of figuring out a way to tone RfA down proper, rather than introduce an exploitable process that duplicates Hell Week and makes people think twice about going thru it, especially if they work in a hyperpartisan topic area. —Jéské Couriano v^_^v ^{threads critiques} 17:05, 22 November 2024 (UTC)

I have never done RFA and can only imagine the stress, but I have done Arbcom and that was way longer than a week. - RevelationDirect (talk) 23:37, 22 November 2024 (UTC)

See WP:REWORK. However, the type of petition you're suggesting failed to gain consensus only a few months ago, and it'd be unfair to have another pass at it so soon after. Sincerely, Dilettante 17:09, 22 November 2024 (UTC)

Are we supposed to rework in this page?

Levivich moved three sections from WT:RECALL to WT:REWORK. I moved it to WP:REWORK. Josh reverted it. My sense is that WP:REWORK is for the actual proposals, while WT:REWORK is for organizing how to process the proposals. Kenneth Kho (talk) 19:03, 22 November 2024 (UTC)

Theleekycauldron since you created the page. Hey man im josh since you're mentioned in the post.

My assumption is that the WP page is for concrete RfC questions and this talk page is for vague points of discussion. Sincerely, Dilettante 19:47, 22 November 2024 (UTC)

That's my understanding as well: the WP page for RFC questions, and any pre-RFC workshopping (with the goal being to formulate the question, not to answer the question) would be done here. Levivich (talk) 19:53, 22 November 2024 (UTC)

The workshop page is already the pre-RfC workshop page, and has been used extensively already to do workshopping. I don't think it's a good idea to start splitting the workshopping onto this talk page. isaacl (talk) 22:27, 22 November 2024 (UTC)

Just saw this thread here. So, what is the point of the other talk page, exactly? PhotogenicScientist (talk) 20:16, 22 November 2024 (UTC)

@PhotogenicScientist, it's a pretty common way of handling policy talk pages: when a particular policy discussion becomes so very large that it dominates that policy's talk page, it often gets moved to its own talk for that discussion so that other discussions about that policy can be done on the policy talk. It helps keep talk pages navigable. I know it can be a bit confusing to find the new locations. Valereee (talk) 20:58, 22 November 2024 (UTC)

Should admins that are subject of recall petitions be allowed to use the Election RFA?

I'll preface this by saying that the recent trial of RFA by election worked out better than I imagined it would. We got a bunch of quality admins, the stress levels were much lower for everyone involved, and there were no issues identified that can't be addressed with some tweaking of the process. The majority of comments poste-vote, from both candidates and voters, was very positive. I believe that there are already plans to run another trial with some process modifications.

Would people object if an admin whose recall petition result requires them to start a RRFA elected to use the RFA by election method, rather than the standard RFA? Would they be allowed to wait until the next scheduled RFA by election, if it is more than 30 days?

I'm putting this out there now, because I wouldn't be surprised if RFA by election becomes either the standard, or a regularly scheduled option, within the year. Risker (talk) 05:55, 23 November 2024 (UTC)

This is explicitly allowed and mentioned in WP:RECALL. If an election exists and is less than 30 days away, they can use WP:AELECT. Else they mostly cant* (crat discretion means there's some leeway).

There's also a pre-RFC workshop phase at Wikipedia:Administrator_elections/October_2024/RFC_workshop for AELECT that is hashing out some of the details. There's a couple recall related questions in there. Soni (talk) 06:00, 23 November 2024 (UTC)

Hmmm. That creates an issue. I'm fine if it's yes, or if it is no, but there's a fairness issue when it comes to "only if you get recalled within 30 days of an election and there's an open slot". Perhaps the solution is to campaign for monthly elections when the AELECT rework RFC comes up. Risker (talk) 06:31, 23 November 2024 (UTC)

What if an Admin choosing an election agrees to completely suspend their use of the tools pending the election outcome? IOW, if an Admin were petitioned for recall 110 days before an election they could choose RfA with no effect on their tool access and an immediate resolution, or election with the understanding they would cease all tool use in the interim? Chetsford (talk) 07:28, 23 November 2024 (UTC)

The German admin recall systems have what they call Temporary Desysoping. That may be worth considering and applying to a wait period for an upcoming AELECT. Kudpung กุดผึ้ง (talk) 10:30, 23 November 2024 (UTC)