Wikipedia:Administrators/RfC for BARC - a community desysopping process

Abbreviations
RfC: Request for Comment
BARC: Bureaucrats' Administrator Review Committee

Introduction

[...] We know that for two years now, the number of people being made admins is too low. And yet we have valid concerns that admins are overstressed, and that they don't always live up to what we hope in terms of thoughtful, kind, and welcoming conduct. I think that solutions lie precisely in these directions: make it easier to become an admin so that more people can share the burden, and easier to lose the bit when there are problems.

Jimbo Wales (2012)

In August 2012, the Community de-adminship proof of concept concluded with a clear consensus for the motion:

The current methods for removal of administrator rights are not sufficient and an additional community-driven method should be implemented.

A second, less firm consensus suggested that some form of gatekeeper should be put in place to stop frivolous complaints. This will be included by sharing the task with a group of trusted users, who are generally regarded as level-headed (Bureaucrats).

Benefits

• The lack of an effective, dedicated procedure for addressing admin behaviour is one of the reasons for the perceived failure of WP:RfA to produce sufficient candidates of the right calibre. A community-based system should help to alleviate that perception.
• Devolving simpler cases to a more community-based system would reduce the workload on the Arbitration Committee.
• A transparent system would help strengthen accountability of administrators.

Current situation

Current methods to remove adminship
Process	Does not require fresh request for adminship	Requires fresh request for adminship
Voluntary removal	For most reasons that the administrator chooses to resign their userright	If the administrator chooses to remove the userright "under a cloud", i.e. whilst their actions are under scrutiny, perhaps during an arbitration case or elsewhere. If the administrator has agreed to removal due to a successful "recall"
Inactivity	If the administrator has been completely inactive for between 1 and 3 years	If the administrator has been completely inactive for over 3 years
For cause	N/A	If the administrator has the user-right removed. Arbcom can do this with a case, or without in certain circumstances.
Other	Technically, the stewards or bureaucrats have the ability to remove the tools in an emergency. Also, Jimbo Wales could theoretically also remove administrator access at his discretion. These situations would likely have subsequent repercussions.

Bureaucrats

1. Bureaucrats are empowered to judge consensus and add the sysop flag following a candidate's successful pass at WP:RfA.
2. Bureaucrats have the technical ability to remove the sysop flag but not the community mandate to do it except in situations outlined in the above table.
3. Bureaucrats have demonstrated a very high level of community trust, by passing an RfB, which requires an 85% level of support, despite the small number of additional tools.
4. There are 33 Bureaucrats (as of this proposal).

Sources

Further reading and previous proposals
Wikipedia:Removing administrator rights • Wikipedia:Arbitration Committee/Procedures#Removal of permissions • Wikipedia:Blocking policy • Wikipedia:Banning policy

Note: Please do not introduce other alternative methods of admin review. If you wish to suggest significant changes or to propose a significantly different system, start your own separate RfC

Proposal for lightweight desysop process (BARC)

Note: Full Wikipedia:Bureaucrats' Administrator Review Committee/Procedures will be debated in a second phase to this discussion if consensus is reached to accept this proposal in principle.

A new group, the Bureaucrats' Administrator Review Committee (BARC) will be created, comprising bureaucrats and other editors from the community. The purpose of the group will be to accept/decline and subsequently adjudicate requests from the community for the removal of an administrator's sysop user-right. The group's deliberations should be public and separate from general community discussion. The process should be considered a formal, binding process, driven by the community but not as open a forum as the "Administrator's noticeboard".

The committee should act on behalf of the community and will comprise any five bureaucrats from the pool of active bureaucrats, on a case-by-case basis, alongside five other static community members. The community members should be appointed annually by the bureaucrats, with community consultation. There is no requirement for the community members to be administrators.

The committee will vote to decide:

1. Whether to accept case (i.e. a legitimate case exists and if it does, whether it should be referred to the Arbitration Committee).
2. Whether a temporary injunction is required for the duration of the case.
3. Whether removal of the administrator user-right is appropriate.

Bureaucrats are authorised, on consensus being reached, to remove the administrator userright.

In the event of a hung committee, the case will be referred to the Arbitration Committee. If the case involves private or off-wiki information, it will be referred to the Arbitration Committee.

An open case will be referred to by a non-descriptive numerical value. The main case page will be open to committee members and those directly involved in the case. Outside views will be kept on the associated talk page. Once deliberation has begun, the deliberation page may be only edited by the committee. Complaints about patterns of inappropriate behaviour must be brought by at least two editors. All cases must be substantiated by diffs. Temporary injunctions are possible, i.e. the use of admin tools may be procedurally forbidden or physically removed while the case is being heard.

Timeline

↓Complaint

↓BARC decision to open case

↓BARC deliberations

↓Final decision

│

0

│

1

│

2

│

3

│

4

│

5

│

6

│

7

│

8

│

9

│

10

│

11

Timeline in days

• From the point the complaint is lodged, the administrator(s) in question are to be notified on their talk page and by email. They have 72 hours to respond. This period is designed to allow the situation to cool, therefore cases should never be opened in less than 24 hours. Emergency desysop procedures should still be handled by the Arbitration Committee.
• After 72 hours, the committee will decide whether the case should be heard or referred to the Arbitration Committee, with or without the administrator's comment. They also decide at this point if a temporary injunction is required.
• If a case is opened, it remains open for 7 days, where discussion by involved parties happens on the case page and non involved parties on the case talk page.
• At the end of the 7-day period, the case page will be closed and the committee will deliberate on a sub page for up to 24 hours. The outcome of the deliberations will be posted to the involved party's talk pages and at the Bureaucrat's and Administrator's noticeboards.

Appeals
Appeals will be heard by the Arbitration Committee as a Request for Arbitration. The Arbitration Committee may decline the request, issue a motion or hold a full case request, according to their own procedures.

Voting

Support

1. As joint-proposer. --Kudpung กุดผึ้ง (talk) 06:16, 24 July 2015 (UTC)
2. As joint-proposer. I've been advocating for a more community based desysop process for a while now, but my time on Arbcom only made my resolve stronger. Arbcom does deal with problematic administrators, but the process is cumbersome and daunting for all participants, often humiliating for all participants too. The thing I really liked about this process was that was lightweight and streamlined - designed to be as pain free as possible, for every participant, especially the subject. Clear timelines, reduced bureaucracy, a structured but transparent forum - these sorts of things are key to a process like this working. Worm^TT(talk) 06:30, 24 July 2015 (UTC)
3. I think I'm among the oldest admins still active, even if my efforts are sadly sporadic nowadays. I was a bit leery of this proposal at first, but having studied it, it seems to be taking a worthwhile step towards breaking down the "us and them" combative mentality which has arisen (going back to at least 2003). No doubt the committee will receive a number of baseless complaints, but I think that hassle will be worth it for the greater community trust it will hopefully engender. Manning (talk) 07:24, 24 July 2015 (UTC)
4. Yes, strongly. Recall is a dead loss and always has been, and should be done away with as openly misleading. Devolving stuff from arbcom is good and increasing bureaucrats' role in this way I think is worthwhile. Cas Liber (talk · contribs) 07:31, 24 July 2015 (UTC)
5. In priciple. BethNaught (talk) 07:34, 24 July 2015 (UTC)
6. Strong support. There is no real way for the community to take the mop away from an admin, and we do not need our admins to be appointed for life. The only real way now for an admin to be removed is ArbCom - you can't trust them to remain open to recall once the issue is brought up to them. Too many have backed away from it at that point, and regular editors need some way of holding admins accountable. GregJackP Boomer! 07:40, 24 July 2015 (UTC)
7. Strongly support. As noted by User:Ihardlythinkso "...the process (removing an administrator) is so overladen with bureaucracy... Not only is it slow, but tunnel-visioned, by its very nature. Jimbo Wales was right when he said the solution is to make adminship easier to get, and easier to lose. What we have today is a painfully slow and inept system that is overall impractical to address the problem appropriately, thus the problems fester and multiply and linger in view of the seldom-and-difficult-to-use process to cleanse the WP body of undesirable admins. They win. Jimbo was right. The structure is inadequate to deal with the reality. The structure needs to be changed." Memills (talk) 07:42, 24 July 2015 (UTC)
8. Support, per Worm above. Cavarrone 07:42, 24 July 2015 (UTC)
9. Strong support. Arbcom is the only current avenue for desysoping. As we all know, those proceedings can be grueling, overbearing, humiliating and highly stressful time sinks of a process for everyone involved. At the other extreme is recall, a widely varying process that most administrators don't partake in, and anyone who does can opt-out of at anytime they choose. While a good idea, recall is simply not, in practice, an effective system for desysoping. Another, easier means to desysop problematic administrators would definitely benefit the community and the project, but I don't think desysoping should be easy. I bristle at the very thought of a community-driven sideshow to desysop someone. I was very surprised at the striking simplicity and reasonability of this proposal. There doesn't seem to be a risk of major drama, nor arbitrary witch hunts. I have nothing against Arbcom, but there's no denying the resentment, distrust and controversy can surround the committee. Crats generally don't seem to have that problem as they are probably the most trusted, competent, and least controversial members of the community—I will point out that the Arbs generally don't even get the amount of support that a crat needs to get appointed. I think this concept is about as good as it's going to get as far as desysoping goes. It would be a lighter, faster, easier, less painful process than Arbcom, and with more qualified and trusted people placed in charge of it. I support this proposal unreservedly. Swarm ^{we ♥ our hive} 07:44, 24 July 2015 (UTC)
10. Support. Tony (talk) 07:45, 24 July 2015 (UTC)
11. Strong support in principle, but I would want to make sure that when the procedures are hardened out, they would make this process be fully fluid and truly community-based.--Jasper Deng (talk) 07:52, 24 July 2015 (UTC)
12. Strong support. Per Worm, and also: It's a necessary counterweight to admins being appointed without definite (renewable) terms. Ijon (talk) 07:55, 24 July 2015 (UTC)
13. I was just invited to comment. Had never heard of BARC before. I don't think it will help a great deal, but I don't see how it can hurt, and it looks worth a try. In principle, the community should resolve common issues, and ArbCom should be the last resort. I find it hard to believe that more Wikipedians will choose to run an RfA because there is a community desysopping procedure. It might cause people to be more relaxed about welcoming new admins. --SmokeyJoe (talk) 08:03, 24 July 2015 (UTC)
14. Support, worth trying it out. Almost anything would be better than the current situation. Graham87 08:09, 24 July 2015 (UTC)
15. Good admins have nothing to fear. Those learning their trade have nothing to fear, since admission of an honest mistake is a valid defence. Admins who are less than satisfactory and who transgress will get a wakeup call that may return them to the path of righteousness. Genuinely bad admins will be weeded out in a few months as this system, progresses. This has my unqualified support. I am not now, and do not ever intend to be, an admin. Fiddle Faddle 08:13, 24 July 2015 (UTC)
16. Support, sounds like a good idea of a quick process for simple cases. The complex cases should still be carried by Arbcom Alex Bakharev (talk) 08:22, 24 July 2015 (UTC)
17. We doubtless need something; ideally, while this committee will probably be heavily used at first, its format, structure, and remit will ensure that over a fairly short period it will settle into being a part of the community used occasionally (much as Arbcomm, which, if i recall correctly, seemed to have far more cases on at a time ten years ago). Thanks are due Kudpung and WTT for their thoughts and suggestion. Cheers, Lindsay^Hello 08:24, 24 July 2015 (UTC)
18. Support. This is a very well thought-out proposal; it has enough procedure to prevent its frivolous use, yet is not overly complicated. It also makes better use of our existing, highly-trusted bureaucrat group, whose role is currently limited to the menial task of promoting and demoting administrators and bots. In addition, I agree with just about all comments made before me. — This, that and the other (talk) 09:32, 24 July 2015 (UTC)
19. Support - I've authored my own proposals, helped on others and have been looking for a solution since I became an admin. Of all of them, this has the best blend of safety, equity, speed and flexibility. This will help increase the number of admin and reduce the impression that admin are guaranteed to be admin for life. Dennis Brown - 2¢ 09:59, 24 July 2015 (UTC)
20. In principle, although I'm somewhat sceptical of the ability of any panel of users to consistently decide cases in the timeframe proposed. I think it's worth a try though to see how it shakes out. Lankiveil ^{(speak to me)} 10:06, 24 July 2015 (UTC).
21. Support This is a well thought out and much needed proposal. Sam Walton (talk) 10:10, 24 July 2015 (UTC)
22. Support, nothing here to complain about, really. The proposers have done their due diligence on this, though I should note that there should be a means to remove and replace a member of the committee who steps out of bounds in some way (i.e. participating in a case they are involved in instead of recusing). This is especially needed if admins will serve on the committee, since it would make no sense for an admin who was dragged before the committee to remain on it if he was. —Jeremy v^_^v ^Bori! 10:16, 24 July 2015 (UTC)
23. Support, we need something like this. It would probably slightly reduce my likelihood of opposing some of the RFAs that are marginal in my mind. I share Jasper Deng's caveat that "I would want to make sure that when the procedures are hardened out, they would make this process be fully fluid and truly community-based," and his remarks in the Discussion section. I think we should give this a try. --Stfg (talk) 10:20, 24 July 2015 (UTC)
24. Support: a well thought out proposal by two users I have a great deal of respect for. Process seems very simple but not lacking any substantial details and I certainly agree with the principle of promoting more admins but making it easier to demote. Everyone seems to agree that RfA is incredibly flawed and perhaps it's because the !voters don't have enough information—no-one can actually work out how a user will act when they have the mop until they have it, and most !votes (whether support or oppose) seem to be quite superficial. The solution would be to just have an easier removal system, so there's more of a feeling that people are granted adminship temporarily, or until they start causing problems. — Bilorv_(talk)^(c)(e) 10:23, 24 July 2015 (UTC)
25. Support per noms. Widr (talk) 10:24, 24 July 2015 (UTC)
26. Support They say the devil is in the details, and I don't know if the detailed procedures will find consensus, but let's go forward and try -- the outline looks fine. Alanscottwalker (talk) 10:59, 24 July 2015 (UTC)
27. It's ludicrous that desysoppings have to go through Arbcom.—S Marshall T/C 11:26, 24 July 2015 (UTC)
28. Support Clearly a well thought out idea, that looks like it will work. Brustopher (talk) 11:37, 24 July 2015 (UTC)
29. Support Yes, there are some sub-standard decisions made by people not up to scratch, and with ANI worse than useless as a venue for coming to clear cut decisions once the peanut gallery clouds the issue, there has to be a better way. - SchroCat (talk) 11:46, 24 July 2015 (UTC)
30. I support, and the proposal is well thought-out and balanced. Though I have to comment, if it's lightweight... then that really says something about how cumbersome the current system is (=desysop by arbcom). Well, I knew that. As for recall, as an example, I've had a quite frustrating experience of trying to use recall against an admin who had made certain promises in their RFA, who would surely never have reached sufficient support without those promises, and who chose to renege on/wikilawyer them when they were invoked. Talk about a timesink, and they're still an admin. Yes, recall is a joke. Bishonen | talk 12:11, 24 July 2015 (UTC).
31. Support - It is long overdue that there should exist a community-based and community-initiated process for addressing long-term behavioral issues of administrators that "fly below the radar" of ARBCOM. The process should be relatively quick, it should be tempered with reasonable due process and protection from the ANI "pitchforks" mentality, the deliberative decision-making body should include non-administrators, and there should be an appeals process. Kudpung and Worm That Turned's proposal satisfies those criteria, and the details may be tweaked in the future, as necessary, based on actual experience. This is a necessary step forward. Dirtlawyer1 (talk) 12:16, 24 July 2015 (UTC)
32. Support, well thought out process, some additional details can be ironed out in the next phase, but the basis of this proposal is sound, and it is needed. --kelapstick^(bainuu) 12:18, 24 July 2015 (UTC)
33. Support: this process seems very well thought out - while a number of more specific details will need to be worked out, as a framework, this proposal seems effective. G S Palmer (talk • contribs) 12:58, 24 July 2015 (UTC)
34. Support conditionally, as I'd need to see detailed procedures before fully supporting. In particular, I would want to see community-oriented procedures to appoint the five non-bureaucrat members (as giving 33 bureaucrats full control over a de-sysop process by choosing who votes is not community oriented) and what thresholds of successful/unsuccessful complaint you plan to set. I'd also be very interested in a restriction that only admins can bring complaints. If a legitimate complaint exists, any editor acting in good faith could find a sysop that would be willing to open a case for procedural reasons if nothing else. If a legitimate complaint does not exist, this would help limit the drama. ~ Rob^Talk 13:15, 24 July 2015 (UTC)
35. Support Lord Sjones23 (talk - contributions) 13:31, 24 July 2015 (UTC)
36. Strong Support in principle; a system needs to be put in place, and this one seems well thought out and not overburdensome. This deserves serious consideration. Scr★pIron^IV 13:36, 24 July 2015 (UTC)
37. Support Largely per Dennis Brown. I do note some thoughtful concerns in the opposition of Nick-D, but I think the concern about frivolous complaints (which I share) can be fixed in the procedures phase, where I would recommend that multiple editors have to sign on to a complaint before it would be considered.--S Philbrick(Talk) 13:42, 24 July 2015 (UTC)
38. Enthusiastic Support. A community-driven desysopping process is sorely needed, and this is a great step in the right direction. The integrity of the en-wiki bureaucrats (when they don their bureacruat fezzes, anyway) is neigh unimpeachable. Yes, the proposed process is a little bit complicated, but not nearly as burdensome as Arbcom. If this actually results in a more "easy come, easy go" view of adminship on RFA, all the better. HiDrNick! 13:49, 24 July 2015 (UTC)
39. Support Long overdue. Arbcom can't deal with this effectively. Intothatdarkness 13:53, 24 July 2015 (UTC)
40. Strong support per Worm, Dennis Brown, SchroCat, and many others. We're lacking an effective forum to bring complaints about substandard admins (nether ANI nor ArbCom are effective for this) and this is looks to be a well-thought-out proposal to create one; it's worth a try. I am somewhat less concerned about the process than I am about what the criteria would be for demotion, however I trust that both will form part of this discussion and/or be decided by the community, so it is not a major concern for me. Ivanvector 🍁 (talk) 14:30, 24 July 2015 (UTC)
41. Support - Arbcom does a good job handling de-sysops once a case is accepted. However, Arbcom is a bit of a juggernaut and takes entirely too long to prosecute cases. The DangerousPanda case was one of the swifter cases, but still took 56 days. The arguments for crats serving on the committee are compelling. My only mild reservation is the provision for allowing crats to appoint five editors from the community at large. I think there should be an popular election process, and I think the number of admin seats should be proportional to the number of active admins in the active editor population.- MrX 14:56, 24 July 2015 (UTC)

    The DangerousPanda case was one of the swifter cases, but still took 56 days, only because DangerousPanda said that he was busy and could not participate in the case, so the deadlines were extended to accommodate his needs. Two other cases I drafted where no such problem was present were handled much more quickly: Wikipedia:Arbitration/Requests/Case/Kiefer.Wolfowitz and Ironholds, 1 month and 1 day; Wikipedia:Arbitration/Requests/Case/Nightscream, 1 month and 3 days. Salvio ^{Let's talk about it!} 15:07, 24 July 2015 (UTC)

42. (edit conflict) Support. There are technically possible ways to remove the tools as of now, and I have argued that due to the availability of such processes the community should not make RfA so difficult, but I understand that the community desires a process which, although orderly and reasonably protected from frivolous complaints, is still somewhat community-based, unlike ArbCom. I would only suggest though, that panel members have more time than 24 hours to deliberate a case. --Biblioworm 15:00, 24 July 2015 (UTC)
43. Support - Well thought out and supported, especially "lightweight, efficient yet robust community de-adminship process" per BethNaught in discussion. . Buster Seven Talk 15:07, 24 July 2015 (UTC)

Oppose

1. Yet another go-round the same loop. ArbCom will desysop if there is evidence of misuse of tools, this is simply an invitation to witch-hunts. We have seen many examples where offsite groups have mounted co-ordinated attempts to rid themselves of admins they don't like. The number of instances where this is likely to be invoked other than through grudges, seems to me to be extremely small. Guy (Help!) 08:34, 24 July 2015 (UTC)

   Surely this is better than the current situation? ArbCom cases are long and drawn out, here cases should last a maximum of 11 days or if there is little basis for a complaint only 3. I'd hope that we would elect committees with enough sense to spot a frivolous report when presented to them. Sam Walton (talk) 08:41, 24 July 2015 (UTC)

   Arbcom does desysop, after months of discussion. Many who have been desysopped by Arbcom go on to leave the project, so we lose an editor as well as an admin. Perhaps if some of those cases had gone through this instead, they wouldn't have felt the need to leave. Worm^TT(talk) 09:03, 24 July 2015 (UTC)

   This proposal is specifically designed with the very purpose in mind of avoiding the traditional witch hunts and kangaroo courts that ANI is, and RFC/U used to be until we got it deprecated. The community will be able to have its say on the case talk page, but there's no guarantee, any more than at Arbcom, that their rattliing of sabres will be heard. One must not make the error of assuming that this system will only treat cases of misuse of the tools; it will hopefully do something which Arbcom does 'not' do so very well: hear cases of long term patterns of behaviour unbecoming of admins that often manage to stay just under the radar but nevertheless does quite a bit of damage. --Kudpung กุดผึ้ง (talk) 09:23, 24 July 2015 (UTC)

2. Bureaucrats act based on community discussion. Not based on their own discussion (like Arbcom) - even taking into account the five community spots. This would basically just create another arbcom-lite. If you want to do that, just create a process for a cut-down desysop case with a panel of 3 active sitting arbiters with rigid submission rules. Much simpler to set up and administer - I or anyone else who works in refining process-flows (I believe at least one admin offered to do this recently) could get a draft in a day or two. Arbcom's case problem is that all types of cases are mostly handle d the same - when better up-front organisation would cut down the time needed massively. I would however support any proposal that allows crats to desysop based on clear community discussion - there is no functional reason why AN could not handle a discussion on if an administrator's actions deserve them losing their admin status. The recent 'block while involved' misuse of tools by Kww shows that. The only reason that case has gone to arbcom is that there is not a process in place for removing tools by community discussion. Give the crats the ability to desysop yes - but we already have one time-consuming and ineffective body, lets not turn the part of wikipedia's bureaucratic processess that actually works well into another. Only in death does duty end (talk) 10:03, 24 July 2015 (UTC)

   To be honest, I would prefer 10 user's selected at random like jury service (although of course, it would have to be opt-in) to be a better option... Only in death does duty end (talk) 10:12, 24 July 2015 (UTC)

   In this system, the Bureaucrats are not just gatekeepers, they are part of the deliberating committee. Knowing that what we see of Arbcom's work is only the tip of the iceberg, I felt that to respond to the community's perennial calls for a more efficient and more streamlined system to address sysop issues, devolving it from Arbcom but keeping it under the aegis and participation of our most respected members,would be the logical step. It's either that or increase the number of Arbcom members so that they can form a sub committee for the purpose. That would not however address the communiy's claims that Arbcom as a body is not proficient to resolve sysop problems. At the same time, the effort is to deliberately avoid am ANI type structure at all costs. It's even much harder to maintain decorum and common sense at ANI than it is at RfA.Kudpung กุดผึ้ง (talk) 10:27, 24 July 2015 (UTC)

   Well I would prefer streamlining arbcom and increasing members than involving the bcrats as part of the deliberating committee. I want them to flick the switch, not get stuck in a discussion about if it should be on or off. The majority of the community's problems with desysops are about the process required, not that it is arbcom doing it. Arbcom are proficient at desysopping, given the current process to work with. Give them a better process and most problems would disappear. Personally I *still* think a decision to remove admin rights needs more community involvement, not less - which is what happens when issues are escalated to arbcom or (as above) to a new arbcom-lite. Which is why I would also support giving the crats the ability to remove admin rights given a clear consensus to do so. (It isnt inconceivable to do both) Only in death does duty end (talk) 10:53, 24 July 2015 (UTC)

   It's certainly one suggestion for improvement and if you were to propose it, I'd give it significant thought. However, I'd probably oppose because I'd like to see Arbcom's powers devolved, having sat on it for 2 years, I am of the opinion that it is far too powerful a committee and handles a lot of stuff which would be better handled elsewhere. That's why I'm throwing my weight behind this suggestion. I will, however, challenge one point - "The majority of the community's problems with desysops are about the process required, not that it is arbcom doing it". Having spoken to a large number of people on this, there is a sentiment that Arbcom are not always up to the task due to its structure as well as it's processes. Many people have a low opinion of the committee generally, it's even mentioned in the archive of this proposal's talk page. On top of that, there is a concept that if you go into an Arbcom case, all parties generally come out worse off. The amount of work to raise an Arbcom is significant. There's all sorts of reasons that Arbcom is a less than optimal solution and a lightweight community alternative should be considered. Worm^TT(talk) 12:12, 24 July 2015 (UTC)

3. While well meaning, I think that this proposal contains three serious flaws: 1) While I have a very high regard for the bureaucrats, they weren't selected for this role on their ability to investigate complaints against admins, as this has historically been the role of ArbCom - where it is a major issue considered by participants in their annual elections. Empowering the current bureaucrats with a role which has historically only been granted to popularly elected arbs and which they don't necessarily have the skills or desire to perform seems unduly risky. 2) The notion of the bureaucrats then picking (with undefined "community consultations") "five other static community members" to join them in passing judgement is even more problematic. These editors won't have passed through any formal community processes to give them a mandate to judge the conduct of people who have passed through a RfA process, and this process runs counter to how Wikipedia has historically handled selecting editors for significant positions of responsibility. 3) There doesn't seem to be any kind of process to prevent or even discourage frivolous complaints (which, for all its many faults, the labour required to lodge a RfArb does). There's a real risk that this will turn into an offshoot of ANI where grumpy editors file complaints against admins, thus creating yet more drama for few gains. I'd also prefer establishing a sub-committee of ArbCom to handle this kind of thing if the current processes aren't working given that its members are elected to handle serious complaints against admins and doing so would avoid creating another governance body. Nick-D (talk) 12:25, 24 July 2015 (UTC)

   I believe that if anyone possesses the required skill set, our Bureaucrats have demonstrated, in so far as they are or have been active, more than sufficient level of tact, diplomacy, and wisdom. I am not sure that every one of our past and present Arbcom members can rise to that compliment or even have passed the stringencies of RfB.

   While it may not have been within the original remit for Bureaucrats to deal with or be involved in such matters, consensus can change, and over the years Wikipedians have demonstrated in many cases that they are able to rise to the challenge of successfully modernising and modifying our policies and procedures as the encyclopedia evolves. Most members of our pool of Bureaucrats have been with us for a very, very long time - indeed, many of them have fallen into quasi inactivity. If we wish to retain the very notion of Bureaucrats, like the anticipated increase in the flow of RfA perhps this current initiative will help build a larger team of Bureaucrats.Kudpung กุดผึ้ง (talk) 13:20, 24 July 2015 (UTC)

4. Preliminary stance. I'd like to see the procedures involved - particulary with the idea of strongly restricting frivolous complaints. Also #2 of what Nick-D said, while I don't distrust the current Bureaucrat roster at all that looks a bit too in-groupey to me. Jo-Jo Eumerus (talk, contributions) 13:32, 24 July 2015 (UTC)
5. Emphatic Oppose Another half-baked proposal. (1) We're to approve this and then figure out how it's supposed to work (see note here)? Oh hell no. Either put your cards on the table and tell us how this is going to work or don't propose the process for approval. (2) There's no outline for how "other editors of the community" will be appointed to this board except some vague process involving the community. (3) This proposal suffers from the fatal problem some other proposals have failed from: Bureaucrats have not been elected to review admin conduct, as that has been an ArbCom function...one for which we vet and elect ArbCom. Nobody has ever reviewed a bureaucrat for their suitability in reviewing admin conduct and conflict resolution. (4) This proposal puts the cart before the horse. We can keep throwing processes at the wall for years (and in fact we have). But, until we first come up with the problems we are trying to solve, every proposal is a total shot in the dark as is this one. (5) To call this a "lightweight" process is absurd. There's some vague procedure for how the five bureaucrats are to be appointed, there's what will no doubt be a nightmarish procedure for members of the community to be appointed to it, voting mechanisms which will no doubt involve clerks like we have at ArbCom, and some process for placing injunctions. To call all of this lightweight is akin to calling the RMS Queen Mary 2 a rowboat. (6) Before even considering whether this could possibly be a valid process one needs to ask this question: "What administrator who was not desysopped would have been desysopped by this process?" If you can't come up with a single example, then it becomes blatantly obvious this process has no purpose. In summary We already have a seriously broken process. We don't need another one. --Hammersoft (talk) 13:40, 24 July 2015 (UTC)

   No process this big would ever get through a Wikipedia RfC process fully baked and I think you're naive to think otherwise. What we've spent our time doing is finding the balance of putting enough flesh on the bones that people can picture what the final solution will look like, but allowing for change to allow the community to find the right solution. Once the general solution is in place, we can firm up the different areas. Finally, as to the word "lightweight", compare it to the current options, a multiple month RfArb case, including four phases, hundreds of words of evidence, a committee which needs 2 months of elections and Jimbo to appoint. That's pretty much it. This process is designed to last a specific short period, with a group that is fairly simple to create, that's why I call it lightweight. Worm^TT(talk) 14:13, 24 July 2015 (UTC)

   I'm sorry, but passing something and then figuring out the way to implement it isn't the way to do this. I'm not going to support anything for which I don't know how it's supposed to work. "Hey! I've got this great idea! I'll tell you what it is after you approve it!" Hell no. And while it might be lightweight compared to ArbCom, it is _far_ from being lightweight. This process is nothing more than ArbCom Lite. And, I ask again, "What administrator who was not desysopped would have been desysopped by this process?" --Hammersoft (talk) 14:43, 24 July 2015 (UTC)

   I'm sorry, but all major policy proposals begin with a framework first. That's the way truly collaborative projects evolve and not by dismissing them outright before fully understanding what is involved and what the stakeholders have asked to be debated by the community. This is in fact the third in this series of debates - the previous two having received the consensus on which this one is built. Kudpung กุดผึ้ง (talk) 14:58, 24 July 2015 (UTC).

6. How is it better to have a group of bureaucrats adjudicating desysops rather than arbitrators? I don't see how this would make any difference. Arbcom generally seems to do a good job deciding who should have the bit revoked in any case, so I don't see any point in setting up an additional committee. Gatoclass (talk) 13:51, 24 July 2015 (UTC)

   In my experience people disagree with arbcom when they get an answer they don't like. Under the new system if you can't make your case with evidence, you can always go to a venue that is a lot closer to a public vote. Chillum 14:52, 24 July 2015 (UTC)

   The answer to your question is in the proposal. It's been clearly outlined why this proposal is being made. Whilst it may be true that cases that are brought before Arbcom are generally successfully concluded by Arbcom, we need to look at the bigger picture, understand the hundreds of comments around the site over the years that claim that inappropriate behaviour by admins too often escapes sanction alltogether, and then take the initiative to offer the commuity a proposal that addresses their concerns. Kudpung กุดผึ้ง (talk) 14:49, 24 July 2015 (UTC)

7. Oppose Our current system is based on evidence, this system sounds like an invitation for a witch hunt every few days. This system looks more like an ANI discussion, well an ANI discussion can already block/ban etc an admin but it does not. This system seems vague and lacking in detail. Crats were never selected with the ability to make these choices in mind, they were selected for other reasons. I don't see what the current system offers that arbcom does not other than a reduced burden of evidence replaced with sentiment.

   I am amazed at how quickly the first 40 supports arrived here after this was opened to the public at large. May I ask what exactly you are all supporting because it seems to me that all of the meat of the proposal is yet to be established. What has been presented here is so lacking in detail that any number of systems could be spawned from it. This would establish crats as one of the most powerful authorities on Wikipedia and we have not even made it clear what form that authority will take other than leaving most of it up to them. Has anyone asked the crats if they even want this job? Chillum 14:40, 24 July 2015 (UTC)

8. Oppose. Definately not, we shouldn't be voting on anything until the proposal has been firmly thrashed out and discussed. Feels like votes for lynching all over again. If this proceeds. I think crats would all need to be reconfirmed to confirm that the community trusts them to desysop users. Spartaz ^Humbug! 15:04, 24 July 2015 (UTC)

   If we did that, then we would likely have another problem to deal with: we would not have any crats remaining, since all or almost all of them would likely be demoted. --Biblioworm 15:16, 24 July 2015 (UTC)

   Exactly, because this is not what they were selected for. The whole choice to use crats in my opinion is killing what otherwise would be a step in the right direction. Chillum 15:18, 24 July 2015 (UTC)

9. Oppose we really don't need a "ArbCom2" to have another level of bureaucracy. What's wrong with the present system? That has not been demonstrated. Why yet another governing body? Why does there have to be 5 bureaucrats, and what happens if there are not 5 standing that are worth voting for? The reason for a lack of administrators is that RfA is terribly broken, and this does nothing to fix this; in fact it could put people off trying, especially if there becomes a mob mentality after the latest block of Eric Corbett. Will encourage grudges and pushes for desysops for unpopular actions that may well need to be done. --Jules (Mrjulesd) 15:19, 24 July 2015 (UTC)
10. Hmm, being in this column puts me in some rather questionable company, doesn't it? Kudos to Kudpung for taking the time to come up with ideas, and I support the concept of some kind of misconduct committee separate from Arbcom—or a more general RFC committee with the powers to issue binding closures to RFCs, coupled with a revival of WP:RFC/U. I definitely don't support giving powers of this nature to the 'crats; this would be taking powers away from people who can be voted out, and giving them instead to people who are appointed for life. Wikipedia's governance structure is dysfunctional and corrupt as it is; the last thing it needs is its own House of Lords added to the mix. – iridescent 15:21, 24 July 2015 (UTC)

Discussion

• Assuming this process can be accepted in principle, there are a number of procedures that would be up for debate in a second phase of this RfC. In my head that would be things like final timescales, potential situations where this process could not be used, recusals and quora, what threshold would be a "legitimate" case, appointment processes and so on. The important thing here is to look at the big picture - a lightweight, transparent, community based desysop process. Worm^TT(talk) 06:49, 24 July 2015 (UTC)
• A concern I have is that the appointment of the community members of the committee by bureaucrats rather than community election would, in that sense, make it less of a community body than the Arbitration Committee. Also, I'm wondering exactly how, in each case, bureaucrats will be selected. I'm also concerned that the power of this committee is not well-defined: is it answerable to arbcom? What happens if the community is displeased with one or more committee members? On a somewhat related note, does the Arbitration Committee intend to consider this a "dispute resolution method of last resort" like it currently does for AN/I? If yes, then if an administrator has engaged in conduct that not only warrants desysopping but site blocks or bans, does this mean that the Arbitration Committee and BARC will have to run simultaneous cases? Finally, a bit of bikeshedding, but why exactly five each of bureaucrats and non-bureaucrats?--Jasper Deng (talk) 07:28, 24 July 2015 (UTC)

  I'd expect that it would effectively be a community election and we can debate how best to do that. My thoughts that something like secure poll or even formal election would lead to election fatigue. It's all about the balance. The rest of the questions are good, certainly. Worm^TT(talk) 07:44, 24 July 2015 (UTC)

  Answering in no particular order... The 5:5 was my idea. I wanted to ensure there was a proper community involvement, size is based on my experience on arbcom, also to give the community a strong say. It's up for debate in the next phase. As is how the bureaucrats are selected on a the case by case basis, I just imagined first come first served. Terms are 1 year, therefore displeasure should only last a finite time. I'd expect Arbcom to take this as a "dispute resolution" with respect to admins, but it's really up to them. Finally "answerable to arbcom"... I don't see why it would be - any more than say, MedCom is. If an Arbcom case is being opened, I'd expect that to be a reason that this is rejected. Worm^TT(talk) 07:53, 24 July 2015 (UTC)

• "Bureaucrats are authorised, upon consensus being reached, to physically enact the removal of tools." This sentence is a bit awkward ("physically"??) especially as Bureaucrats cannot actually remove an admin flag. They need to submit a request to stewards at meta. --99of9 (talk) 07:30, 24 July 2015 (UTC)
  • Yes they can remove the administrator user right. --ceradon (talk • contribs) 07:32, 24 July 2015 (UTC)

  Yes, we can. I have done many times for inactive admins. I thought I tidied the sentence, I'll drop the word "physically". Worm^TT(talk) 07:43, 24 July 2015 (UTC)

  Well I learn something every day around here! On Commons we cannot (I'm a crat there). Strange that we have different settings. Thanks for the grammar fix. --99of9 (talk) 08:26, 24 July 2015 (UTC)

• There is great value in having a lightweight, efficient yet robust community de-adminship process. In detailed discussions, I think we need to address the time scale (allowing lengthier deliberations would be better) and appointment process in particular, and make sure the process is wikilaywer-proof. However this is an excellent proposal; I would urge everyone not to nitpick too much at this stage since there will be a follow up RfC. BethNaught (talk) 07:34, 24 July 2015 (UTC)
• The committee should be at least 51% non-admin, non-bureaucrats. GregJackP Boomer! 07:41, 24 July 2015 (UTC)

  I tend to agree with your sentiment here, although I would only go so far as to say that the non-bureaucrat contingent should have, say, at least 2 seats reserved for non-admins. In any case I don't think this concern is fatal to the proposal (you see that I have supported it above) but I just think this point needs to be made. — This, that and the other (talk) 09:37, 24 July 2015 (UTC)

  Probably best for discussion in the next phase, but I am concerned with the difficulty in finding a significant number of non-admin, non-bureaucrat users with sufficient trust from the community to perform this role and who are also willing to do so. Worm^TT(talk) 11:57, 24 July 2015 (UTC)

• The proposal says that the committee should be "comprised of any 5 bureaucrats from the pool of active bureaucrats on a per case basis alongside 5 other static community members. The community members on the committee should be appointed annually by the bureaucrats, with community consultation." If I'm reading this right, and please correct me if this isn't what you meant, but does this mean that any 5 bureaucrats will be appointed to a case, but the same 5 community members will be part of every case? Sam Walton (talk) 07:43, 24 July 2015 (UTC)

  Yes, that's what I was aiming for. Worm^TT(talk) 07:54, 24 July 2015 (UTC)

• Will there also be a community Bureaucrats' Administrator Review Committee member removal process? —Xezbeth (talk) 07:44, 24 July 2015 (UTC)

Plerase see the introduction. Kudpung กุดผึ้ง (talk) 07:48, 24 July 2015 (UTC)

• Interested in knowing what would count as "consensus for removal" here - it looks a bit "free for all" to me right now. The possibility of any community-driven deadminship process being abused/misused for petty grievances or outright bad faith requests is a major reason why previous proposals have failed. Some concerns about whether assessing tool misuse is a radical new job that is at odds with the normal bureaucrat responsibilities. Jo-Jo Eumerus (talk, contributions) 08:37, 24 July 2015 (UTC)

  My thoughts were that it would effectively be a vote of those on the committee, but a vote with visible comments, so that they can effect other votes. Similar to how Arbcom does it's "proposed decisions". Worm^TT(talk) 09:08, 24 July 2015 (UTC)

• Regarding appeals to arbcom, what would be the grounds for the appeal to be successful? Are arbcom there to simply judge whether consensus was reached correctly, or a full case from scratch like it is now, or somewhere in between - "yeah a mistake (or two) were made and there's legitimate complaint, but we think a telling off would suffice, so we'll overturn a desysop decision"? -- KTC (talk) 08:45, 24 July 2015 (UTC)

  That's a good question. I would expect Arbcom to overturn based on additional information, possibly private. Perhaps if there was evidence that the case wasn't fair, for whatever reason. I'd be surprised if Arbcom would overturn a a case that was handled reasonably. Worm^TT(talk) 09:08, 24 July 2015 (UTC)

• "An open case will be referred to by a non-descriptive numerical value." Why? If each case concerns only one administrator, then surely it would make more sense to name it according to the same scheme as RFA, for example "ExampleAdmin" for a first case, or "ExampleAdmin 2" for a second case, rather than just a simple number. — This, that and the other (talk) 09:37, 24 July 2015 (UTC)

  It's one of the most stupid arguments that we seemed to have on almost every case on Arbcom. The non-descriptive numerical value stops any future accusations of "scarlet letters", or marks against persons real name, allows for better vanishing and privacy. The fact is, this is never going to be a pleasant process and little things like that will make it slightly less unpleasant. Worm^TT(talk) 10:03, 24 July 2015 (UTC)

• (edit conflict) We elect our bureaucrats on an open, transparent process where we subject them to a greater inquisition and hold them to a far higher level of integrity than the Arbitration Committee candidates. In actual theory, if not in practice and policy, a group of Bureaucrats should be able to trump the Arbcom, and not vice versa. Kudpung กุดผึ้ง (talk) 09:41, 24 July 2015 (UTC)

  Indeed. Only one arbitrator since 2009 has managed to top 85% support, and that wasn't me ;) Worm^TT(talk) 11:59, 24 July 2015 (UTC)

  My RfB in 2007 enjoyed 98% support, with only 3 people opposing. In the 2008 ArbCom elections, I had 67% support, with 127 people opposing, at the time I withdrew (i.e. only just enough to be appointed to the committee had that been the final result). It is regrettably a fallacy to think that the high percentage support required to pass RfB would be reflected were a bureaucrat to seek the community's endorsement to fulfil a different tole. WJBscribe (talk) 14:59, 24 July 2015 (UTC)

• so what do we do when this doesn't increase the numbers passing RFA?©Geni (talk) 10:52, 24 July 2015 (UTC)

  Increasing numbers passing at RfA isn't the sole purpose to this proposal, just one hopeful benefit. Whether or not the numbers are increased, future arguments that RfA should be difficult to pass as it's hard to remove the userright should be reduced. Worm^TT(talk) 11:53, 24 July 2015 (UTC)

  The main purpose of this proposal is to demonstrate that while adminship is for life (at least for the time being), its tenure in the face of a streamlined desysoping process is one that admins will hopefully treat with more respect. Once the new committee has shown that it does what it says on the tin, the community will introduce more homogenised and hopefully less stringent criteria on which they base their votes at RfA. At the moment we have RfA participants demanding FA, 20,000 edits, a raft of DYK, and 100s of AfD which along with the need for some of them to deliberatekty turn RfA into a drama is what is keeping otherwise perfectly qualified candiates away. That said, it's nevertheless unlikely that over the next five years RfA will be back to dozens per month rather than the two dozen per year at present. I started drafting this proposal three years ago, given a chance over the next three we can them come back to your question if necessary. Kudpung กุดผึ้ง (talk) 12:29, 24 July 2015 (UTC)

• Doing this to fix RFA is pointless. RFA needs to be solved at RFA not with desysoping rules. Thinking that this will work is crystal ball gazing.
  That said I'm not against this in principle. But there are a few problems with the proposal. One thing that needs to be clear is that there *are* times when a sysop needs to be stern and the process must outline that there's a difference between a sysop chewing out someone who deserves it and them abusing their "position" and being a jerk. On top of this a few thoughts on reading this:

1. Crats should be allowed only suspend sysops bits on consensus but must refer to ArbCom for approval of that decision (not for an RFAR). In all likelihood this will only happen in a few select circumstances so adding that layer of check and balance wont hurt anyone.
2. It has to be editors in "good standing" who can file this kind of complaint and moreover I'm not sure that two is an appropriate number, this is an overturn of community consensus, two people is very low for that. However I am conflicted about this, because one person should theoretically be enough if the evidence is strong enough. Also maybe there should be clear sanctions for clearly disruptive complaints (like there are at WP:AE)?
3. Like RFAR other avenues of dispute resolution must have been attempted and ignored by the sysop before going here. We can't have a situation where disgruntled sockmaster goes straight to BARC. I know no action would be taken but it's a waste of time. And as we all grow-up we don't have the infinite time that some people seem to have to dedicate to pointless disputes.

So in short can't support as currently worded but not against it in principle--Cailil ^talk 12:46, 24 July 2015 (UTC)

• This would be a big change in the bureaucrat role - something not envisaged when we originally were RfB candidates. If this proposal were to pass, I think I would feel the need to seek reconfirmation from the community. I don't know how other bureaucrats feel about this. WJBscribe (talk) 14:51, 24 July 2015 (UTC)

• So for clarity, you don't feel that a consensus at this RfC (which by all projections is going to be much better attended in the end then a typical RfB) is not, in-itself, a mandate that you (as Bureaucrats) have the trust of the community for this role. Put another way, if we're going to trust 'crats, why wouldn't you think that doesn't include you? Regards, Crazynas ^t 15:20, 24 July 2015 (UTC)

• Do we keep statistics on the percentage of oppose comments by admins, compared with the percentage of support comments by admins? We should.- MrX 15:00, 24 July 2015 (UTC)

• My question is: what standards will BARC use to determine whether to desysop an administrator? I ask because, in my opinion, the main problem with our current procedures is that they ony allow us to desysop administrators who have been misconducting themselves, while nothing can be done about admins who the community no longer trust to have the tools. If BARC is going to use the same standards used by ArbCom, then I don't really see any reason to have it, but various reasons not to: a. the community members are not elected by the community, unlike arbitrators; b. 'crats were never appointed to evaluate the conduct of administrators; c. in my opinion, this process gives far too little time to administrators to defend themselves and for the community/BARC to discuss the issue; and, d. having had experience with ArbCom cases, I can foretell that the deadlines indicated will turn out to be nothing but wishful thinking. Yes, ArbCom cases last far too long, but part of the reason they do is to give the opportunity to everyone, who wants to have a say, to express their opinion (not to mention that sifting through evidence to prepare a draft decision is a time-consuming effort) and discussion of difficult cases often takes days especially because editors live in different time zones. If, on the other hand, it's anticipated that BARC will be desysopping administrators who have lost the trust of the community, then I don't see why we can't have a community discussion closed by a 'crat or a panel of 'crats evaluating consensus (which would be my preference). Salvio ^{Let's talk about it!} 15:02, 24 July 2015 (UTC)

• It is part of our job to make decisions that are going to be unpopular with some people, if you do your job in a neutral fashion you will eventually annoy a lot of people. One of my big concerns with these sorts of proposals is that desysoping will move from being based on actual misconduct and instead be allowed to be achieved through sufficient sour grapes. Chillum 15:05, 24 July 2015 (UTC)

A few points

If we're turning bureacrats into the admin watchdogs, I can understand that. Given that the user-right has of late has seemed to change focus that somewhat makes sense. But no, please no, to a "committee of any kind. No "yearly durations" etc. Besides that, what's the point of having 5 community members input if they're appointed by the bureaucrats anyway? And what if one or more need to recuse for various reasons?

This should be done the same way we choose closers for contentious discussions. Ask for volunteers on a case-by-case basis.

Also, this should reflect the current processes done by arbcom. For example, bureaucrats should be able to advise or admonish as well - giving other options than just "mop or no mop" helps take this more into "preventative, not punitive".

All that aside, I'd rather see this as a "reverse RfA". If bureaucrats are ready to accept this responsibility (in the past when I polled them, they were dead set against it), then I would re-write WP:RRA to remove the arbcom section, and replace with a sort of policy defined "crat chat".

But the community - not representatives thereof - needs to be able to be directly involved. The bureaucrats are already our representatives, let's not add another layer as well, please. - jc37 13:27, 24 July 2015 (UTC)

• I would think that if the committee of Crats and Editors decided the problem was singular but didn't want to deysop, they could issue an admonishment in the final ruling. You don't need to formally declare that "right" in the policy, that is a matter of procedure. We already can admonish at ANI, for example. Your previous community desysop proposal only allowed for desysop, mine allowed for other actions, we learned that the community was really only interested in the desysop portion back during that period. And it would be ANY 5 Crats, so they could recuse as could the community members. That doesn't need codifying, as existing policy (and in particular WP:COMMONSENSE and the spirit behind WP:INVOLVED) applies. I would also note that complicated cases would still be pushed to Arb. This fills the space between "emergency" and "highly complicated". It simply adds a community driven option. Dennis Brown - 2¢ 14:21, 24 July 2015 (UTC)

This is so far away from what crats were selected for it makes me wonder why they were even chosen. It would make for more sense for arbitrary members of the community to be selected than it would to use them. This seems like you just picked an existing group of people and taped this task to the side. Chillum 15:09, 24 July 2015 (UTC)