Wikipedia talk:Arbitration/Requests/Case/RHaworth/Proposed decision

Comments by Alanscottwalker
Maybe minor but I think "administrator status" in that first principle is, shall we say, infelicitous. It is true that they have been entrusted, but they have not been entrusted with status, they have been entrusted with tools and co-commitment norms/expectations of behavior. I suggest change "status" to "permissions". Alanscottwalker (talk) 15:23, 28 January 2020 (UTC)
 * , for the sake of context, this principle is lifted from Arbitration/Requests/Case/GiantSnowman. Myself, I'm fairly indifferent to "status" vs "permissions"; another common description is administrator "rights".  Maxim (talk)  16:34, 28 January 2020 (UTC)
 * Well, never too late to improve. :) "Permissions" is better than "rights", although granted they have the same origin, because permission is granted and denied by others, whereas a person's rights may exist even though other's attempt to deny their right.  Alanscottwalker (talk) 16:56, 28 January 2020 (UTC)
 * I think "user rights" works best, in the context of Special:ListGroupRights. –xenotalk 17:00, 28 January 2020 (UTC)
 * Ehk . . . but truly thanks for the responses. Alanscottwalker (talk) 17:21, 28 January 2020 (UTC)
 * I agree - I think it should be termed as privilege(s), as distinct from rights. When I first started to contribute to WP, I allowed myself to be bullied between 2014-2015 as I thought, naively, foolishly, that autopatrolled, rollbacker, reviewer represented an early stage recognition of rank, a lowerarchy that could lead to higher things. I found by chance that these were simply tools (loose quote) "dished out to anyone who requests them", intended to make repetitious actions easier. When I ascertained this, I then investigated the perpetrator who'd been bullying me - linked to this, I'm not here by chance, and I'm pleased to see that there are now arbcom members 'new' to my past experience (five admins, four being arbcom in the one sequence).--Rocknrollmancer (talk) 20:54, 28 January 2020 (UTC)

,, : A few questions. Here is Arbitration/Requests/Case/RHaworth/Evidence. 1. What do you find reassuring in that submission? 2. Are there other RHaworth statements you are referring to? Thank you for your time. -- Alanscottwalker (talk) 14:35, 29 January 2020 (UTC)
 * Personally, I appreciate the specificity of RHaworth's statements. My chief concern was his speedy deletions which did not meet the strict criteria - I felt his comments led to a reasonable commitment on that. Even so, I still want to remove him from the worst hit areas, A7 and G11. However, I will say that I do not see an admonishment as sufficient, and if neither 2a nor 2b pass (which is currently looking likely) - I will be seriously considering switching my vote on the desysop. WormTT(talk) 14:47, 29 January 2020 (UTC)
 * I find that RHaworth's evidence does a good job of not only acknowledging the issues raised, but suggesting reasonable steps for improvement. The admonishment is written to provide an avenue for desysop-by-motion if the issues persist.  Maxim (talk)  16:52, 29 January 2020 (UTC)
 * If that's the case, why could MSGJ find so many ANI threads involving him? Ritchie333 (talk) (cont)  17:02, 29 January 2020 (UTC)

I may propose an additional remedy or some clarifying language as the case continues to unfold. Meanwhile, I will need to be offline for a few hours today, and may respond to this or other threads later. Newyorkbrad (talk) 14:50, 29 January 2020 (UTC)

Comments by MSGJ
Thanks for the prompt proposed decision &mdash; Martin (MSGJ · talk) 15:30, 28 January 2020 (UTC)

Checkuser
I understand the committee may wish to take the opportunity to remind admins not to undo checkuser blocks, but this really wasn't the focus of this case and no such pattern of bheaviour was identified. I would prefer that these findings be removed from the final decision to avoid distracting from the central issues, which were deletion and responses. &mdash; Martin (MSGJ · talk) 15:29, 28 January 2020 (UTC)
 * I think part of the point of arbitration decisions is to serve as a restatement of principles and policy. Not every single finding is intended to be grounds for sanction, and there's plenty of reason to restate for everyone in the back that it's not a good thing to do, evidence of a pattern or not. Der Wohltemperierte Fuchs  talk 16:08, 28 January 2020 (UTC)
 * The undoing of the checkuser block was the precipitating event for the case, to not address it in any way seems peculiar. –xenotalk 16:57, 28 January 2020 (UTC)
 * Fair enough but I doubt you would have accepted the case if that had been the only reason &mdash; Martin (MSGJ · talk) 21:32, 28 January 2020 (UTC)

Desysopped
I appreciate that this is the standard term on Wikipedia now, but I really wish findings and remedies were written in English rather than jargon. "Adminstrator rights are removed from RHaworth" is slightly longer but perfectly clear. &mdash; Martin (MSGJ · talk) 15:29, 28 January 2020 (UTC)
 * How's Special:Diff/938016264? –xeno<sup style="color:#000">talk 16:55, 28 January 2020 (UTC)
 * Much better! &mdash; Martin (MSGJ · talk) 21:32, 28 January 2020 (UTC)

Comments from Ritchie333

 * "RHaworth interactions with users" - any particular reason I'm mentioned twice? <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk) <sup style="color:#7F007F">(cont)  17:21, 28 January 2020 (UTC)
 * , just because you're that special..? :p  Maxim (talk)  17:27, 28 January 2020 (UTC)
 * Remedies - Unfortunately the only one that appeals to me is a desysop. The problem with RHaworth is not just deletion - it's that and brusque replies to newbies and not understanding that he did anything wrong - this includes incidents where I agree that the deletion was correct and within policy, but winced at the response to the newcomer. So the most effective sanction short of a desysop is "RHaworth is banned from being incivil to newbies", and civility restrictions always cause way more hassle than they prevent (eg: cases involving Eric Corbett or The Rambling Man). A ban from A7 / G11 takes out most of the problem, but it only solves half the issue. As I said earlier, I had a lengthy chat with RHaworth at the last London meetup and he explained in exasperated fashion exactly why "Kindly wait until somebody with no COI thinks your [x] is notable and writes about it here" was, in his view, polite, and I understood what he meant. (For example, "Kindly" here is used as a synonym for "Please" ie: "kind") However, if I have to struggle and jump through hoops to realise why he's not purposefully being malicious or incivil, what hope has anyone else got? I concluded that there was little hope of improvement in this area because it's fundamentally part of who he is. I think the best option is for him to forget about adminship completely and go and do something like expand and source Beanos or any number of other Croydon-based landmarks, so he might discover the positive aspect of Wikipedia editing. (TLDR : What Thryduulf said) <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk) <sup style="color:#7F007F">(cont)  17:33, 28 January 2020 (UTC)
 * I'm going to be direct - if the case closes with an admonishment, and I still think RHaworth is being bitey, can I civility block him? Or would I be considered WP:INVOLVED? <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk) <sup style="color:#7F007F">(cont)  18:24, 28 January 2020 (UTC)
 * It might be best to first warn them about and highlight to them the continued undesirable behviour before blocking - aside from that, the civility policy may of course be enforced. Is the only reason are concerned you might be considered INVOLVED because of your submissions to this case? –<b style="font-family:verdana;color:#000">xeno</b><sup style="color:#000">talk  23:10, 28 January 2020 (UTC)
 * Well, in my case, I've already started two ANI threads about RHaworth and given him a clear and unambiguous warning (as linked in the Evidence), so I would think any block from me would come across as having it in for him, regardless of how correct and aligned with policy it might have been. <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk) <sup style="color:#7F007F">(cont)  10:53, 29 January 2020 (UTC)
 * Well you clearly felt uninvolved enough in 2019 to issue the warning... If your prior involvement has been purely administrative, this would not be considered capital-I involvement. Ultimately this is a personal decision, whether you feel that you become involved to the point (thread creation, personal interaction with RHaworth, etc.) that you can no longer make a dispassionate judgment on the issue. [speaking personally] –<b style="font-family:verdana;color:#000">xeno</b><sup style="color:#000">talk 14:28, 29 January 2020 (UTC)

I just noticed the lead page of http://rhaworth.net says "I also get my kicks from helping to delete the endless dross that flows into Wikipedia along with the good stuff." Administrators should not "get a kick" out of using their tools; they are to aid in maintenance of the encyclopedia. <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk) <sup style="color:#7F007F">(cont)  15:33, 30 January 2020 (UTC)

Comments by Thryduulf
In the RHaworth restricted remedy (2b) the G11 speedy deletion criterion is described as "Unambiguous spam or promotion", but it is described at WP:CSD as "Unambiguous advertising or promotion". The distinction is important as the criterion only covers material that is unambiguously promotional, which is just one form of spam with other forms typically being different criteria (e.g. nonsense is G1). The remedy also does not make clear what will happen should he breach it - blocks? amendment request asking for a desysop? something else?

Regarding Remedy 3 (RHaworth admonished). I think this remedy might be better phrased as a warning that if the assurances do not materialise that the committee will desysop rather than an unspecified amendment. Given that the same assurances have been given many times after ANI threads, I'm not confident this time will be different. Thryduulf (talk) 17:43, 28 January 2020 (UTC)


 * I've matched the G11 text to the link, thank you. –<b style="font-family:verdana;color:#000">xeno</b><sup style="color:#000">talk 22:25, 28 January 2020 (UTC)

I'm in firm agreement with Liz and Boing! said Zebedee's comments elsewhere on this page. It is very concerning that there are arbitrators who are willing to take at face value promises of behaviour change from admins (RHaworth and BHG, I'm not following the Kudpung case closely enough to comment on that) when there has been repeated promises of exactly the same that have not held water (RHaworth) and a repeated failure to demonstrate an understanding of why their conduct is problematic (both). "Stop or we'll tell you to stop again" is not an effective deterrent. If (I hope not when, but that's probably forlorn) RHaworth returns to his old ways, what is the next step? There is nothing AE can meaningfully do, so do we need to go through another four weeks of RHaworth2? Do we just file a request at ARCA? Can we be confident that will lead to a desysopping or will there be another final, final, final, final, we-really-mean-final-this-time final warning? What is Arbcom's plan to deal with the good faith editors driven away by incorrect deletions and/or incivility between now and then? Thryduulf (talk) 13:54, 29 January 2020 (UTC)

Comments by Liz
I'm a little concerned that ArbCom is dealing with three admins where incivility is an issue (although in lesser or more degrees) for long-standing, valued admins and there are such different outcomes, at least at this early stage of voting. I realize they are different cases with different evidence presented but it's hard not to see some comparisons regarding how the admin treats others and what the penalties are for mistreating others.

My experience viewing RHaworth's many cases at AN is that there is a promise to do better and that lasts for a short while before he returns to very familiar behavior, both with treating new editors & IPs poorly and the quick page deletes. These proposed remedies don't hold out any penalty if he isn't "shook" by being part of an arbitration cases. There is an assumption that being part of these proceedings will lead to longstanding changes of behavior but there is no remedy presented right now if this doesn't prove to be the case (and if his pattern is true to past behavior). I'm not necessarily advocating desysoping but restriction from deletions in certain problematic categories (and I'm surprised that U5s weren't presented as well) would be a remedy that would have some consequences if violated.

Right now, these proposed decisions seem like a stern talking-to but are ultimately rather toothless, if this early voting is any indication.

I had also hoped there would be a general statement to our other admins who have page deletions in the hundreds of thousands of pages to be more careful in evaluating EACH tagged page but that's just a wish from me after seeing what seems to me to be almost a competition to have the most deletions. You can say I'm assuming bad faith but I see a lot of deletions that seem premature and only the sheer volume of deletions that are performed daily keep me from reevaluating some of them that seem questionable. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 02:26, 29 January 2020 (UTC)


 * I am equally concerned. That the arbs are unwilling to sanction admins does not bode well for the community, or for the other cases before ArbCom.  If any one is "on trial" here, it is the arbitration committee.  How they handle these cases will demonstrate if there are two standards:  one for regular editors, and another for well-connected administrators. The Arbs are the community's last hope when admins treat regular editors poorly, and it is not appearing that they are willing to take this on board. I refrained from submitting evidence in the Kudpung arbcase for concern that this very issue would occur; the arbs may not understand how these issues affect regular editors.  Sandy Georgia  (Talk)  02:38, 29 January 2020 (UTC)


 * I understand your concern that should only remedies 1 and 3 pass, it looks like RHaworth is only being censured and not remedied further. A civility restriction wasn't proposed because in the past civility restrictions have caused more problems than they solved. Also most would likely consider it untenable for an administrator to be under a civility restriction (i.e. the civility policy should be enough to control their behaviour, failing that-they're not suitable). And I share 's belief that deletion is really one of the core sets, to restrict deletion risks someone choosing to use a hammer to clean a floor. Should RHaworth retain administrative privileges beyond this case, they are still subject to policies and guidelines, can be blocked for incivility, or brought back before the committee if deletion problems persist... –<b style="font-family:verdana;color:#000">xeno</b><sup style="color:#000">talk 14:00, 29 January 2020 (UTC)
 * I have bad memories of the Arb Committee failing to take meaningful action in one sockpuppet case. I won't mention that case, but those with institutional memory know of what I speak.  With the direction this is going, I hope the committee will at least put in place a follow-up mechanism, to help avoid a potential further case.  The notion of sysops blocking a fellow sysop doesn't usually end in good places.  If the arbs (the community's only fallback) are unwilling to remove the tools, I hope they will at least come up with a remedy that won't allow problems (should they re-surface) to fester as long as those in the sockmaster case did.  In that case, we went over and over and over the same things again and again, and had more swift action been taken earlier on, it would have saved a lot of time and agida for everyone. Sandy Georgia  (Talk)  14:10, 29 January 2020 (UTC)
 * if there is not, or should not be, a double standard, and we want to retain community governance, then administrators need to be far more willing to warn or block other administrators who are behaving poorly. –<b style="font-family:verdana;color:#000">xeno</b><sup style="color:#000">talk 14:22, 29 January 2020 (UTC)
 * I wrote that before the case turned the right direction. Sandy Georgia  (Talk)  12:46, 31 January 2020 (UTC)

Comment in response to Cryptic's statements:

I wish there was a middle ground between admonishment & desysoping. It's often a choice between not going far enough or going too far. In the old days, I think there was "probation" but I guess they had their reasons for doing away with that option. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 20:22, 29 January 2020 (UTC)
 * Article and behavioural probation generally evolved into topic bans, interaction bans and discretionary sanctions. Admin probation hasn't been applied in a very long time, but I vaguely recollect it not being particularly effective and is poorly compatible with the current philosophy of people being either trusted with all of blocking, protecting and deleting on their own or not trusted with any of the admin toolset, with limited exceptions for specific topics or users. Some sort of halfway house would be good, but it will need to be something that hasn't been tried before. In this case the various rate limit and mandatory explanations for each deletion suggestions in the workshop came close, but no version suggested would have worked without further tweaking. Thryduulf (talk) 12:31, 30 January 2020 (UTC)

Comments by Mz7
I like to call deletion, blocking, and page protection the three "core" admin rights. Although the administrator toolset comes with many different technical abilities, these are the three that sit at the center of the toolset, and we should (and do) expect administrators and prospective administrators at RfA to be proficient in how and when to use all three. They are interconnected; oftentimes administrators must choose between some combination of these three tools in responding to an administrative situation on Wikipedia.

This is why I'm not a fan of proposed remedies 2a and 2b, which would restrict RHaworth from deletions or certain kinds of deletions. If it were some niche aspect of the admin toolset at issue, like granting user rights or mass-message sending, I would be fine with restricting an administrator from those parts of the toolset. However, in my view, we should be able to trust an administrator to use all three core admin rights proficiently, or we should not trust them to be an administrator at all. Mz7 (talk) 02:58, 29 January 2020 (UTC)


 * When I submitted my comment above, the voting looked like this. I'm glad that since then, my thinking has aligned well with that of the committee. Mz7 (talk) 23:02, 30 January 2020 (UTC)

Comments by Swarm
I'm troubled that the vote is trending towards a "slap on the wrist", i.e. an admonishment and no desysop, in an apparent paradoxical betrayal of your own finding of fact that RHaworth's "responsiveness" has not corresponded with any meaningful resolutions, and when RHaworth has already been admonished by the community. The evidence clearly shows a wanton and willful disregard for our policies and guidelines, continuing in spite of many complaints that RHaworth has disingenuously "accepted feedback" from, and you appear to be acknowledging this problem as a "finding of fact", yet, insanely, you're bowing to RHaworth's fairly mediocre and inadequate defense in the face of overwhelming evidence, because you want to "take him at face value"? Really? I will actually defend RHaworth's CU unblock here, because the CU block appeared to be out of line with policy. If he was merely standing up against a bad CU block, I would support him. In fact, it's a bit concerning that this case has glossed over the dubious CU block. If procedurally incorrect, I'd defend RHaworth for that action alone, as there was no convincing rationale nor admin accountability relating to the block. It's actually ridiculous that you're going to chastise an admin over what was apparently a justified CU overturn. But it's additionally ridiculous that you're going to to let a chronic problem user get a free pass in response to a clearly documented pattern of abuse. There's no actual logical reason to defend this blatant rejection of the communal rules and norms that govern us, and, while I hate to say it, Arbcom's current move to do so is unjust and corrupt. I had great faith in this iteration of Arbcom, but how quick you have forgotten the lessons of WP:FRAM. Established users defending misbehavior without principle or cause is what led to the issue, and now we have Arbcom propagating the issue well after it has been established and even ratified as a legitimate concern. This is just sad. Grow a spine and get real about behavioral issues, or the WMF is going to continue to usurp power from the community. ~Swarm~ {sting} 03:36, 29 January 2020 (UTC)
 * And with another sysop case right behind this one.     Sandy Georgia  (Talk)  03:49, 29 January 2020 (UTC)
 * I have not made my own decisions here as yet, but I would note that only 4 out of 14 active arbs have voted on most of it. Beeblebrox (talk) 07:20, 29 January 2020 (UTC)
 * I'm not sure their actions (assuming no switch of opinions) in any way approach "corrupt", by any normal meaning of the word. Certainly an argument could be made for unjust (and an argument against), and potentially other complaints, but I believe that on a Civility dispute, it's important that those of us who comment on it also may sure our own statements are evidence based. Nosebagbear (talk) 10:44, 29 January 2020 (UTC)

Please see my comment in the next section. Beyond that, I accept that some people will disagree where the line is to be drawn between "truly final warning" and desysop in a case involving an administrator who has shown a long-term commitment to the project but made errors of judgment and decorum. Commenting on the proposed decision as it emerges is the very purpose of this page; criticizing arbitrators' votes one disagrees with and even seeking to change them is part of that, but calling us or our decision-making "corrupt" is merely offensive. Newyorkbrad (talk) 11:52, 29 January 2020 (UTC)

, you make a valid point, but I wonder if would be made more effectively without the allegations of corruption, especially at this early stage of voting. – bradv  🍁  16:02, 29 January 2020 (UTC)
 * I do not wish for a perceived personal attack against Arbcom members to distract from my actual argument and it may not have been the best word to invoke. Apologies, I will strike that. I really don't mean it as a personal insult, though, more of an honest criticism, not of the people, but of the system. While I obviously don't think the arbs are individually "corrupt" (I genuinely respect and look up to every member), we all know that there is a problem with our system that unreasonably and blindly defends misbehaving "power users", even at the expense of the greater community's wellbeing, even when no logical reasons to defend them exist. This unwritten, unofficial system is corrupt, meaning that it directly contravenes our cultural norms and community standards. If one institution is supposed to be a failsafe against this issue, it's Arbcom, and after Framgate, both the community and the Foundation looked to this institution to step up its game and work on this issue so that it doesn't have to happen again. The problem we're addressing here is fairly extreme. It's not simply a "lapse in judgment", a "human error", or a failure of "decorum". It's intentional admin abuse, spanning years. The evidence is overwhelming, the defense is practically nonexistent. A pattern of disingenuous voluntary responsiveness is well documented. Community warnings, condemnations and admonishments have been issued, over and over again. It's not like we got here lightly. We are here, as a last resort, after all else failed. Securing the obvious and straightforward remedy of desysopping an abusive admin at this point is the whole purpose of all the time and effort everyone spent filing the case and submitting evidence. Admonishments were useless, that's the whole point of the case. I get it, RHaworth is a dedicated editor, and no one's asking for them to be kicked out. They're just abusing their community-granted authority, and we can't do anything about that, so we're straightforwardly asking you to exercise a desysop in what should be an uncontentious case. Your own findings of fact agree with all of this. So the act of betraying all of that and only delivering yet another admonishment is, if not "corrupt", illogical and wrong, and that's what I mean by calling it "corrupt". I do apologize if anyone felt personally attacked. ~Swarm~  {sting} 01:20, 30 January 2020 (UTC)

Comment from Espresso Addict
I have to say that I am genuinely surprised and confused as to why the case about BrownHairedGirl appears to be proceeding towards severe sanctions where there is limited evidence of misuse of tools, while here there is no question that the tools have been misused and civility towards newcomers has been lacking over a very long time period. Both are long-term admins with a lot of positive work for the encyclopedia.

If I encounter A7 or G11 deletions by RHaworth that I consider unjustified in future, what am I advised to do? Espresso Addict (talk) 06:31, 29 January 2020 (UTC)
 * I feel the same. RHaworth has exhibited chronic misuse of admin tools and incivility towards newcomers and IPs for years, has been admonished multiple times by the community, but still doesn't seem to fully understand the issues. Yet so far it appears he's only getting a finger wagged in his face. BHG was severely uncivil and attacking, but only towards one admin and only over a much shorter period, and has accepted her lack of self-control was unacceptable and she won't continue it. And she has not misused admin tools. But she's heading for a desysop. I know there are significant differences in the cases, but it seems like uneven treatment to me. Why won't ArbCom accept BHG's promises at face value just like they appear to be accepting RHaworth's? Boing! said Zebedee (talk) 08:43, 29 January 2020 (UTC)

For what it's worth, by coincidence, the four arbitrators (including me) who have voted so far on the remedies in this case all opposed the desysop remedy against BHG, with reasons given there. This may be relevant in making comparisons at this stage. Newyorkbrad (talk) 11:46, 29 January 2020 (UTC)
 * I hadn't spotted that, thanks. Boing! said Zebedee (talk) 12:17, 29 January 2020 (UTC)

Comments by Martinp
The Committee is facing multiple cases of productive, long-serving sysops who develop some unfortunate behaviour patterns and/or whose judgment becomes clouded in certain areas; who aren't responsive enough to community feedback (in such venues as exist for it); and so end up before Arbcom. I'm pretty sure more such cases will come.

While circumstances vary, given the absence of other mechanisms, I wonder if we can find middle ground between admonishment and removal of the bit, where the sysop shows insight into the problem and wants to reform. We all know the path back to redemption via an RFA is a long and thorny one. One can take the position -- in principle or in individual cases -- that words are cheap / it's too late / trust can't be conditional. But if not, I put out for discussion the following:


 * 1) The admin must have an ongoing history of overwhelmingly positive contributions, in which the poor behaviour is an aberrant pattern
 * 2) The admin must show insight into where, how, and why they have violated expectations (what ideally they would have had without an Arbcom case, but if that's what it took to get there, fine!)
 * 3) The admin must commit to specific changes in their behaviour acceptable to the committee. Ideally, the precise specifics would be up to the admin, something they feel will keep them away from danger. The committee would have to concur, and would solicit feedback from the community (during the workshop and/or proposed decision phase?). Example specifics might include the admin avoiding certain actions, certain areas, mentorship from another admin, cooling off period, etc. -- depends on the problem, the context, and the admin.
 * 4) After this commitment is accepted, the admin retains the bit, but Arbcom "retains jurisdiction" for some time (facetiously: "suspended sentence", though importantly the bit itself is retained, not suspended). This permits a speedy, low-controversy desysop if the admin can't stick to the commitment or it proves insufficient to contain the problem. I'm not sure of the right formalism here: case is suspended but the admin is desysopped by motion if there continue to be problems? Or some sort of discretionary sanctions/arb enforcement mechanism as a remedy? Or just "strongly admonished with a warning that the committee may summarily desysop"?
 * 5) Importantly, there is a time limit. There will of course be extra scrutiny on the admin, and so they will need to shy away from controversial areas. We can't have that hanging around for ever. If the admin successfully reforms for x months, we assume (and need to have had that confidence when crafting the conditions!) the extra oversight expires and they are subject to the same rules and oversight as all other admins. After all, they should be following the same norms of admin behaviour anyway. We declare victory; a problem solved and an admin retained.
 * 6) I don't think we want the workload or chilling factor of a formal review at the end of the "suspended sentence" period. But an alternative to this idea would send the hopefully-reformed admin for a reconfirmation RFA to explicitly demonstrate they have fully regained community trust. However, this would be happening after several months of reformed, trustworthy behaviour with the bit, and so would hopefully have a more positive tone than an RFA even several months after forced bit removal by ArbCom.

Thoughts? In general? And whether this would be a good case to try something like this out?


 * In the dim and distant past (eg: Requests for arbitration/Pedophilia userbox wheel war), Arbcom did issue the temporary suspension of administrative privileges as a remedy. I'm not sure why this fell out of favour, possibly people thought it was punishment or possibly thought it would be fairer to let them undertake community scrutiny at RfA again. <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk) <sup style="color:#7F007F">(cont)  17:19, 29 January 2020 (UTC)
 * I personally would be open to such suggestions; I do not formally propose them because I know the overwhelming number of arbs are opposed.  DGG ( talk ) 21:33, 29 January 2020 (UTC)
 * See also Wikipedia talk:Arbitration/Requests/Case/Portals/Proposed decision and User:Thryduulf/What happened after a desysop. At the first link I wrote Thryduulf (talk) 12:37, 30 January 2020 (UTC)

Thanks for the comments,. I am aware this may be controversial. And I won't attempt to parse details of interpretation of Arbcom policy/terms of reference with Thryduulf. However, I would zoom back the lens and note that the community expects Arbcom to find the best solution to otherwise intractable conduct disputes. Admin conduct ends up in its lap not only since no-one else is empowered to yank the bit as a remedy, but because community conflict resolution mechanisms (e.g. WP:AN) are poorly equipped in dealing with conduct patterns, especially on- and off- or borderline ones by long-term users. Over the years, Arbcom has invested lots of effort to develop a spectrum between the book ends of admonishment and long blocks/bans for problematic users. It would be helpful to consider what could be an analog for admins. One can of course take a hard line that Arbcom can merely sadly note that trust in an admin was broken, and leave to the community when (if) it can be re-established. But if (and whether or not that is the case here, it will eventually happen...) an admin genuinely gains enlightenment during the often-maligned RfAr process, the more formal review helping them step out of tunnel vision and/or battleground mode in WP:AN etc., it's a shame if we can't keep their services if there is reasonable confidence the problem is solved. As to the concern that Arbcom is not empowered to judge "whether someone has regained [...] trust", I would consider it more as judging "under what conditions are we reasonably sure the trust will continue?"

Finally, an analogy: If an admin's account is compromised, arbcom procedurally removes their bit. When they (credibly) regain control of their account, I believe we return their bit following their shamefaced commitment to practice better security hygiene. We trust that the shock means their security behaviour will indeed change; rather than saying they lost our trust by not taking previous (ample) guidance on security seriously enough, and therefore back to RFA to demonstrate they have community trust. (There are of course limits to this and any analogy, but food for thought...). Martinp (talk) 17:14, 30 January 2020 (UTC)
 * Some form of "halfway house" as someone else put it (SandyGeorgia or Liz possibly?) between admonishment and desysop would be good. However, certainly for issues involving blocks, deletion or protections, I think it would need explicit community support for the principle before the current arbitrators would be prepared to support such a remedy. That discussion would need to happen independently of an arbitration case. On your final point, the bit can also be removed when it is unclear whether the account is compromised or not and, if determined that it isn't, it is returned with no repercussions. I know this happened some years ago to user:Chase me ladies, I'm the Cavalry when he was an arbitrator. Thryduulf (talk) 18:58, 30 January 2020 (UTC)
 * Perhaps something like
 * For (reasons), (user)'s administrator user rights are removed. As the committee recognizes (user)'s assurances to (moderate their behaviour), this remedy will be suspended upon the closure of this case. Should concerns persist, the remedy may be reinstated by the committee. The remedy will be reviewed in (six/nine/twelve) months to determine if the remedy remains necessary.
 * –<b style="font-family:verdana;color:#000">xeno</b><sup style="color:#000">talk 19:08, 30 January 2020 (UTC)
 * I wrote that some sort of mechanism was needed to avoid a repeat before the !votes to desysop turned in the right direction, and after I had stated that I had bad memories of a decade-old case where the committee failed to act decisively and a problem was allowed to fester in spite of significant history. The arbs had a "plan".  Putting in place a "plan" with a user or admin who already has demonstrated history isn't likely to work and is likely to burden the community.  When an admin has a history of problems, the community expects the arbs to act.  We elected you to do that, and I will even more strongly endorse arbcom candidates who do just that next year as well.  Because things need to improve in here if we are to be able to write content, which is supposed to be why we are here.  Separately, when an admin does NOT have a history of problems, and the arbs act in the wrong—and most unfortunate—direction, we should be looking at ways to solve the underlying dysfunction in the community that leads to these festering problems, for which a very good admin has paid the price. A few of my ideas are a) even stronger insistence on better content contributions before passing RFA (BHG understood content development; that ANI let that problem fester is absurd and a reflection that we need more admins who understand and are willing to enforce the fundamental and core policies of why we are here, as BHG did), b) less reliance on RFCs to address what is clear content disruption by small groups of users with an agenda, and c) a return to the historical process of RFC/U to deal with problems that are actually coming from individual or small groups of users rather than broader content issues.  It is somewhat shocking how often clear content issues come to ANI where an individual or small group is violating core policies and admins seem to have no clue how to address them. We end up sending people away to write RFCs to address a content issue that is really a user issue, and the users should have been shut down at ANI.  IMO, the portal case need not have come to Arbcom if ANI had dealt with it correctly. We have really run enough RFCs as if they were broad content disputes, when it is really one user, or one group, or one admin who is ignoring core policies and enacting an agenda.  The portal thing should have been handled in two sentences at ANI:  1) BHG don't call people liars even if you can demonstrate that they are, and if you do it again, we'll send you to arbcom, then close that chapter because people enforcing core content policies can snap; 2) people creating portals, read WP:NOT.  We aren't a webhost or a place for you to push an agenda. Go create a website somewhere to do that.  Instead, BHG paid the price because she snapped, with no history of same.  No, I would not welcome more Arbcom "plans" or halfway measures or mechanisms or suggested RFCs. Yes, we expect Arbcom to take a strong stance on admins with a festering history, and yes, I will continue to support that. But throwing content issues back to community RFC to support a small group of disruptive editors should stop at ANI and ArbCom, and we should find a way to reinstate RFC/user conduct to deal with the real problems. And in spite of one huge and glaring mistake, yes, the community expects you to act when an admin has a history of problems. Sandy Georgia  (Talk)  13:32, 31 January 2020 (UTC)
 * Yes, I would welcome the return of an RFC/U-similar process. –<b style="font-family:verdana;color:#000">xeno</b><sup style="color:#000">talk 13:36, 31 January 2020 (UTC)

Comments by SilkTork
Adding my concern that, given evidence was provided in the case that "RHaworth is unwilling or unable to change", the appropriate action of removing admin privileges is being voted down based on a view that RHaworth is now willing and able to change. There is a credibility gap here in that RHaworth's word appears to count for more than the evidence. If the desysop remedy is to be voted down, there needs to be evidence and rationale presented in the findings. SilkTork (talk) 17:42, 29 January 2020 (UTC)
 * I have stricken the above, as the situation has changed since I looked at it about half an hour ago. SilkTork (talk) 17:52, 29 January 2020 (UTC)

Comments by isaacl
This is a very minor point: although I'm personally indifferent to the use of the term "status" (it can simply denote a state, without implying a hierarchical class relationship), given how principles can be re-used in future cases, I suggest using "administrative privileges". isaacl (talk) 18:23, 29 January 2020 (UTC)
 * , Forgive me for posting here, but this seems so discrete an issue, if someone wants to me move, please move. You can see my similar point in my section up-top.  But I am not a fan of "privileges" either (I think, "permissions" is good), I wonder why you think "privileges" is good given its hierarchy connotations (and keeping in mind that the nomenclature all apparently comes from User access levels) ? -- Alanscottwalker (talk) 19:11, 29 January 2020 (UTC)
 * Alanscottwalker: for me, the context of "administrative privileges" connotes a hierarchy of available actions, rather than one of class hierarchy. Generally, I feel that "privileges" reinforces the idea that it is not a right earned, but a privilege afforded by the community. However for this specific principle, it may not be too important to underline this point of view. isaacl (talk) 19:35, 29 January 2020 (UTC)

Comments by Cryptic
Clerks: In FOF2, there needs to be spaces inserted into "G11(" and into "A10 ([22])and". Such things make me twitch. —Cryptic 19:47, 29 January 2020 (UTC)

Also in FOF2, I don't think your logic holds. I don't think what we're seeing is an average-to-low error rate multiplied by a huge speed rate (I used to; Administrators' noticeboard/IncidentArchive1021 and Deletion review/Log/2019 August 31 were what finally convinced me otherwise), but if it were, then "the obvious conclusion is that he is going too fast" isn't obvious in the least. The same number of pages are going to be deleted no matter who deletes them, at the same overall rate - RHaworth pulls his deletions from CAT:CSD and hasn't done any deletions in over a month now, the category hasn't overflowed, and I don't think it's been added to any slower - and if RHaworth's error rate were below average then by removing him we'll actually be seeing more errors. —Cryptic 19:47, 29 January 2020 (UTC)
 * I didn't explain my thinking there very well. I was trying to say that the rate doesn't matter at all. Admins should be aiming for a very low absolute number of errors. If you make a lot of mistakes and say well, that's because I do a lot, my response would be do less, until you can get nearly everything right. –&#8239;Joe (talk) 05:57, 30 January 2020 (UTC)

Arbs: I want to echo what User:Martinp says above, in spirit if not necessarily in specifics. Nonadmins at arbcom don't ever face sanctions more severe than a block that's appeallable (frequently successfully) after a year, and if they change their checkuser-visible data and can keep away from their old haunts and habits, they can come back right away anyway and there's not a thing we can do about it. There's also a host of lesser sanctions that have been used successfully.

But against administrators, every recent case I can remember except GiantSnowman's has resulted in a desysop, and that's permanent. No admin desysopped for cause by arbcom is going to pass RFA these days: you know it, I know it, and the watchlist-summoned folks at RFA who overwhelmingly base their votes on the nomination and questions and previous votes and don't do their own investigations know it. I don't necessarily disagree with desysopping here - in particular, I haven't seen remedies suggested for his treatment of new users that I think would be effective, and haven't been able to come up with any - but if you're stripping the bit from nearly every admin that comes before you, then either you're not hearing remotely enough cases, or you really desperately need to come up with lesser sanctions at least as options. —Cryptic 19:47, 29 January 2020 (UTC)
 * as I see it, we are finally resolved to deal adequately with the problem, which in previous years we mostly evaded. With ordinary users, there's no way the community can reverse or over-ride our decision except by electing different arbitrators. But here, if what we doing does not meet with the agreement of the community, they can resysop. We're doing this in a way that lets them make the final decision--not just by whom they elect as arbs, but in each individual case of this sort. And even if they do not resysop the first time, they will have other chances to do so later.   Personally, I think this is an ideal way to do things.  (there have been cases where the prior approval of arbcom is necessary to allow a run for arb, but those deal only with the grossest misuses or very important confidential evidence).   DGG ( talk ) 21:30, 29 January 2020 (UTC)
 * There was a recent discussion about resysoppings at Wikipedia talk:Arbitration/Requests/Case/Portals/Proposed decision. I looked at all the data I could find and there is insufficient data to say that nobody would be pass an RFA after being desysopped at Arbcom these days. Since 2015 only three people in that position have even stood at RFA, Rich Farmbrough in 2015 failed because he was still behaving in the way that lead to the desysop in the first place, Hawkeye7 in 2016 and 2019 failed for things generally unrelated to his being desysopped and Fram in 2019 ran immediately after the case closed when the drama was still running high and there had been no opportunity for him to demonstrate that anything had changed (and it was not exactly a "normal" arbcom desysop anyway, so I'm doubtful that anything useful can be extrapolated from it). There is no predictive value in anecdotes. For my own part, if RHaworth is desysopped and comes back to RFA after 6-12 months demonstrating consistently good interactions with new editors and an understanding of the importance of (speedy) deletions being correct then I would support. Thryduulf (talk) 12:21, 30 January 2020 (UTC)

General reminders
With their vote on the General reminder (remedy 4) Beeblebox says which seems like it would be a good thing to make standard practice for all remedies addressed to significant sections of the editing community, (e.g. "editors reminded", "community encouraged", etc). Thryduulf (talk) 11:39, 31 January 2020 (UTC)
 * There also seem to be 15 votes in that section, but there are only 14 arbs active in the case.-- P-K3 (talk) 21:25, 31 January 2020 (UTC)
 * The votes numbered 7 and 8 appear to be part of the same vote. Phil Bridger (talk) 21:33, 31 January 2020 (UTC)
 * I've removed the duplicate vote and left a note for Beeblebrox in case he wants to put his comments together. – bradv  🍁  21:46, 31 January 2020 (UTC)