Wikipedia talk:Removal of adminship

This is a proposal for a de-adminship procedure based in part on my own observations and ideas, Wikipedia:Requests for comment/Wikipedia:Requests for adminship (specifically Friday's section) and some of the earlier proposals for this. I'm purposely leaving many details vague or to be filled in later because I know that filling in specifics before the discussion will lead to too many complaints about the specifics and going back to change them. Anyone is welcome to make reasonable edits that don't significantly change the proposal (so that all the discussion is still applicable). Major changes should be discussed first. Specifics should be filled in as they are determined in the discussion. Please discuss on the talk page. If you can think of a better name that is not already taken, please suggest it. Mr.Z-man 01:07, 7 October 2007 (UTC)

Bureaucratic

I think the process should be just like a new RfA, just as if the person were trying to become an administrator for the first time. a.z. 01:20, 7 October 2007 (UTC)

That's essentially what Step 3 is, obviously the questions would have to be different though. The first 2 parts are essentially for "abuse resistance." Mr.Z-man 01:29, 7 October 2007 (UTC)

I think that step 1 is unnecessary. If the request makes no sense, and few people think it's a good idea, it will just fail. I think any established editor should be able to be the one to review the requests. I am afraid to make it too complicated to remove the tools. a.z. 01:37, 7 October 2007 (UTC)

The point of Step 1 is to prevent a group of people that have a grudge against an admin from just starting one on the admin they don't like and then just having someone else in the group start the discussion, forcing the admin to go through the unnecessary process. How is this more complicated than an WP:RFAR? Mr.Z-man 01:42, 7 October 2007 (UTC)

It is more complicated than an RfA. I think WP:SNOW could perhaps apply to this process, in order to solve the problem. a.z. 01:56, 7 October 2007 (UTC)

Unneeded

Currently unneeded. Current methods are more than adequate. Good job writing this, but unneeded. Regards, Mercury 01:27, 7 October 2007 (UTC)

Additionally, good admins are not necessarily the popular ones. I Think this is game able. I vehemently oppose this. Mercury 01:33, 7 October 2007 (UTC)

Could you please elaborate on how this can be gamed? I tried to close off as many holes as possible without making users jump through a week's worth of hoops but I might have missed something obvious. Mr.Z-man 01:46, 7 October 2007 (UTC)

Ditto. You said it all here: For an admin to be desysopped in an Arbitration case, there must be serious, documented, long-term abuse. That is all what is needed at this point. We do not a bash-the-admin-you-don't-like process: post a user RfC if you need to. ≈ jossi ≈ (talk) 01:34, 7 October 2007 (UTC)

• I agree with Jossi. When serious problems arise with admins, they are dealt with through ArbCom. In cases of emergencies, we deal with the situation in appropriate ways. There is no need for this and it would likely degenerate into a bash-the-admin venue.--Alabamaboy 02:23, 7 October 2007 (UTC)

Agree, is there a particular problem this is addressing? Are there bad admins running around doing terrible things? Doesn't seem like it. ArbCom already de-admins people in extreme situations now, that system seems to work. This system seems like it would be good for removing adminship from unpopular admins though, which would not be an improvement in most circumstances - cohesion 02:36, 7 October 2007 (UTC)

There has been cases of people being emergency desysopped by a bureaucrat which then goes to Arbcom for affirmation/reconsideration, which I agree with in the cases for which it is needed. Also agree with Cohesion in full - there isn't so many bad admins running around that Arbcom can't deal with things. Orderinchaos 16:30, 7 October 2007 (UTC)

There needs to be some better way of removing adminship that doesn't involve arbcom. Arbcom should be the last resort, to occur only when consensus cannot be reached by the community, and the consensus should be a community consensus decision. There probably are bad admins out there, and they would never get desyooped, as 98% of problem cases never reach arbcom. Major cases of obvious administrator abuse shouldn't waste the time of the arbcom when the community can clearly handle and resolve the case easily through desyooping. Yahel Guhan 05:49, 7 October 2007 (UTC)

I don't see the point in creating a whole new bureaucracy for an problem that doesn't verifiably exist, especially when it will likely create problems of its own. - cohesion 06:06, 7 October 2007 (UTC)

The fact that several almost identical proposals have been outright rejected in the past indicated this is unlikely to go anywhere. ((1 == 2) ? (('Stop') : ('Go')) 16:28, 7 October 2007 (UTC)

The process is urgently needed, and ArbCom and user conduct RfCs are not adequate. Why? Because there are so many admins who don't do anything bad enough to get desysopped under the current system, but are habitually incivil and uncommunicative in their exercise of admin tools. When user conduct RfCs are filed against such people, they end up being criticised for not being "focused" enough, or for presenting a broad range of minor grievances rather than one big problem. But lots of minor issues are just as bad as a few instances of serious abuse, IMO (indeed more so, since it results in editors being driven away). Walton^One 21:43, 7 October 2007 (UTC)

Reinventing the wheel

The fundamental features of this proposal are the fundamental features of arbitration. What exactly is wrong with the way that arbitration presently handles this type of thing? --bainer (talk) 02:26, 7 October 2007 (UTC)

That it is not up to the community to decide. a.z. 02:40, 7 October 2007 (UTC)

huh? What do you mean? Isn't everything up to the community to decide? --Ybbor^Talk 02:41, 7 October 2007 (UTC)

No. The Arbitration Committee, also called ArbCom, decides some things, including taking admin tools from people. a.z. 03:25, 7 October 2007 (UTC)

No, in every ArbCom case there's a workshop where community members who have nothing to do with the case can voice thier opinions. ArbCom is supposed to follow those opinions loosely. And if the community decides to dissolve ArbCom or replce its members then it'll happen--Phoenix 15 10:28, 7 October 2007 (UTC)

See my comments to Seraphimblade below. Generally, to get desysopped by ArbCom or by a steward, an admin has to do something very bad. A lot of admins get away with borderline abuse - being uncommunicative and incivil on a daily basis, or using their tools out-of-process - and manage not to get desysopped. Yes, this system would be prone to abusive requests, but with a community discussion in place, it is highly unlikely that someone with a personal vendetta against an admin would actually be able to get that admin desysopped. Also, I think there is a real need for admins to be accountable to the community as a whole. Walton^One 10:51, 7 October 2007 (UTC)

And I strongly agree. dmins are typically not accountable to anyone unless, as walton said, they fuck things up spectacularly. Viridae^Talk 11:19, 7 October 2007 (UTC)

Are there actual examples of arbcom refusing to de-admin someone the community strongly believes should be? ((1 == 2) ? (('Stop') : ('Go')) 16:53, 7 October 2007 (UTC)

No, because there is no forum for that on Wikipedia. The community has essentially no place to discuss the withdrawal of its trust. — Dorftrottel⁠ 18:00, 7 October 2007 (UTC)

No, I think people would mention it anyways. I can never get people to even give an example of where arbcom has failed to desysop someone the community did not want an admin. I don't think this problem is as big as people think or people would be able to rattle off examples of this problem with ease. ((1 == 2) ? (('Stop') : ('Go')) 21:08, 7 October 2007 (UTC)

That was my basic view of it. Considering you would have to go through RFC and such to accomplish this end, and since it involves "trusted" members of the community anyway, you might as well just take it through the existing arbcom process rather than coming up with a new one that creates more bureaucracy. SchuminWeb (Talk) 23:15, 7 October 2007 (UTC)

Extent of problem

Can someone provide some numbers so that we can see how big of a problem we have. How many admins have been desysoped in the last 12 months? How many appear to need such attention? Vegaswikian 02:27, 7 October 2007 (UTC)

I can't answer the later, bus as for the former, since October 2006, 12 adminsitrators were desysopped by ad-hoc decsions making, many for compromised accounts. Of those, 6 had their adminships restored. 7 were desysopped by arbitration committee ruling. 1 was desysopped for failing to respond to an Arbitration Committee ruling. source. --Ybbor^Talk 02:35, 7 October 2007 (UTC)

In other words, we have a solution looking for a problem. Unless someone can show an overwhelming need, lets drop this now. Vegaswikian 02:59, 7 October 2007 (UTC)

The problem is rather about the possible existence of those who ought to be desysopped, but where the circumstances are less than needed for a full RfA. The questions to be asked a/ are there many of these--which would not be easy to answer--I guess one approach is the number of rfcs filed relevant to the matter and b/ whether anything less than arbCom's procedure should be allowed. I'm not saying I am the least sure of the answers, but I think those are the key basic questions. DGG (talk) 04:11, 7 October 2007 (UTC)

"possible existence"?? Who? Where are all these admins that need desysoping but arbcom won't do it? I don't see them. ((1 == 2) ? (('Stop') : ('Go')) 17:05, 7 October 2007 (UTC)

Don't ask for other people to violate WP:NPA in order to make a point. — Dorftrottel⁠ 18:01, 7 October 2007 (UTC)

If we don't see the problem or we can't talk about and quantify it, then it does not exist so lets end the discussion. Vegaswikian 18:21, 7 October 2007 (UTC)

I believe it's safe to assume that some current admins would not succeed at RfA again for good reasons that nevertheless may never lead to their de-sysopping through ArbCom. That's proof enough and doesn't require anyone to violate WP:NPA. — Dorftrottel⁠ 18:48, 7 October 2007 (UTC)

(unindent) I could certainly name one or two specific admins who should not be admins. The trouble is that, if I name them, I will be criticised for violating NPA, or at least the letter thereof. And the trouble that they cause isn't bad enough to file an RfC, or to lead to their desysopping by ArbCom; it's the sort of low-level, day-to-day incivility and aggressive attitude that slowly harms the project by driving good editors away. Indeed, whenever anyone tries to file an RfC against such people, it's usually criticised for being "a miscellaneous collection of small incidents" rather than a single large grievance; that shows the problem with the system, IMO. Walton^One 19:57, 7 October 2007 (UTC)

Reasonable criticism of on-wiki behavior has never been a violation of NPA, if it can be backed up with evidence. ((1 == 2) ? (('Stop') : ('Go')) 21:09, 7 October 2007 (UTC)

That doesn't stop people from declaring that it is, unfortunately. It's happened before. -Amarkov moo! 21:17, 7 October 2007 (UTC)

Well, I'm not going to "name and shame" now, because I'm logging off (it's getting late here) and I can't be bothered to dig up the relevant collection of evidence on several admins. Walton^One 21:38, 7 October 2007 (UTC)

I too cannot see a need for this process. If I understand correctly, 8 admins have been desysopped through Arbcom in the last 12 months but reading the discussions, it seems there are admins who should be desysopped but no evidence can be provided due to the WP:NPA policy. However, should the first port of call for these admins not be RFC? This would be the community's way of informing them of the activities that are causing problems and give the admin a chance to rectify their ways. How many RFCs are there for abuse of admin tools? Can we gather any stats on this? → AA ^(talk) — 21:28, 18 October 2007 (UTC)

Comments for clarification and development

I rather like this idea, because it provides not only a reliable way of desysopping rogue admins, but could guarantee the kind of due process that could protect admins from unfair desysopping. I think that the process would best be arbitrated by the existing bureaucrat system, rather than creating a new one. I think one of the key worries with implementing a system like this is warding off trivial and excessive attempts at desysopping by disgruntled persons. This might best be achieved by having the same beginning process as ArbCom; that is, first a vote to accept or reject a proposal by the presiding bureaucrats. VanTucky ^Talk 04:25, 7 October 2007 (UTC)

I concur, further, to assure current admins there should be an amnesty on any gripes or complaints before a measure like this is enacted. I'm sure that some admins have fans just waiting for something like this. (I get the feeling that admins who might support the notion in principle wouldn't support it in reality due to old grudges they fear coming back to haunt them.) Anynobody 04:56, 7 October 2007 (UTC)

We already have "the kind of due process that could protect admins from unfair desysopping", it is called arbcom. ((1 == 2) ? (('Stop') : ('Go')) 17:06, 7 October 2007 (UTC)

It's not the ArbCom's job to determine the community's ongoing trust with the tools. — Dorftrottel⁠ 18:53, 7 October 2007 (UTC)

Simplified process

The process here seems a little too complicated and strict. My suggestion is for a simplifed process, essentially the inverse of a request for adminship:

• Stage 1: Prior discussion must take place, in the form of an admin RFC, a discussion on the Administrators' Noticeboard, or elsewhere. The admin must have been informed of the discussion, and must have had a chance to respond.

  Reasoning: An RfDA is a method to solve disputes that cannot be resolved in any other way, similar to an RfAr. As such, it should only be undertaken if good-faith attempts to find another solution have failed.

• Stage 2: A request for de-adminship is created. The reasoning for de-adminship is laid out. The request is submitted to Wikipedia:Bureaucrats' noticeboard for approval.

  Requiring approval is a way to keep frivolous requests from wasting the community's time, and to filter out abusive requests. If someone feels their case was unfairly rejected, they can always file an ArbCom case.

• Stage 3: If the request is approved by a bureaucrat, discussion is opened. At the end of a week, a bureaucrat decides if there is consensus to de-sysop the admin in question.

  Many of our best administrators could not pass a second RfA. Because of this, voting is for removal of sysop priviliges, rather than for retaining them.

I don't see a need for specifying penalties for abusive requests, or for requring endorsements, or requring a certain amount of time between requests. If someone's trying to abuse the process, Wikipedia's existing disruption policy seems adequate. --Carnildo 04:56, 7 October 2007 (UTC)

I agree with most of that, except the need to seek approval from the bureaucrats. The point of this is that it allows the community, not an élite committee, to hold administrators directly to account. Just as bureaucrat approval isn't required to submit an RfA nomination (even a completely frivolous one), it shouldn't be required for an RfDA either. The community will decide if the request is frivolous, and requests started by someone with a vendetta against an admin will almost certainly fail. Walton^One 10:31, 7 October 2007 (UTC)

The reason I added the bureaucrat approval is to stop frivolous discussions started by groups with a grudge from starting in the first place and creating undue stress on the admin and wasting the community's time with pointless discussions. Mr.Z-man 16:28, 7 October 2007 (UTC)

What need is this meant to fill?

I guess the question I have here is, why is this necessary? Generally, by the time an admin has a long enough pattern of misbehavior that a significant portion of the community is complaining, ArbCom will accept the case, especially if there's already been an RfC which raised significant concerns. ArbCom is much better equipped to handle complex cases, and has available to them more nuanced remedies (probation, prohibition of certain actions, "do this again and you're done") in addition to outright desysopping. On the other hand, when an admin account is clearly compromised (vandalizing the main page, blocking Jimbo and everyone else they can hit the button on, etc.), a steward will do an emergency desysop with no need for procedure at all. What is the need this is filling? Seraphimblade ^{Talk to me} 06:15, 7 October 2007 (UTC)

As an addendum to above, the other reason I like ArbCom handling these issues is that they examine the behavior of all involved parties. Even if an admin has done wrong, they may not be the only one; one arbitration case can untangle all of those problems. I also fear that, despite the safeguards proposed, this system would be highly prone to abuse. Seraphimblade ^{Talk to me} 06:18, 7 October 2007 (UTC)

I think this is filling a specific need. Generally, to get desysopped by ArbCom or by a steward, an admin has to do something very bad. A lot of admins get away with borderline abuse - being uncommunicative and incivil on a daily basis, or using their tools out-of-process - and manage not to get desysopped. Yes, this system would be prone to abusive requests, but with a community discussion in place, it is highly unlikely that someone with a personal vendetta against an admin would actually be able to get that admin desysopped. Also, I think there is a real need for admins to be accountable to the community as a whole. Walton^One 10:28, 7 October 2007 (UTC)

One could argue we have users who get away with being uncommunicative and incivil on a daily basis who manage not to get blocked. Why is there such a strong push against our community's volunteers? Yes, they need to be accountable, but no, the People's Court (playing judge, jury and executioner) is not the place for it. Orderinchaos 16:37, 7 October 2007 (UTC)

Ah, but an admin who is incivil and uncommunicative causes far more damage than a regular user. Admins are trusted with a position which, whether we like it or not, carries a certain amount of de facto authority as well as technical abilities. An uncivil admin, or one who is too block-happy, will drive potentially good editors away from Wikipedia (which ties in with my essay Wikipedia:Editors matter). As to your second point, this doesn't need to be a "court"-like proceeding at all. The judicial role on Wikipedia is filled by ArbCom, and I'm happy for that to continue. Rather, this is about giving the community the power to select and remove its own office-holders at will, which they have a right to do. Plus, consider WP:CSN (a process which many people dislike, but we'll leave that aside for a minute); at CSN, we already give the community the power to ban users, a far more dramatic sanction than removal of sysop privileges. We need to trust the community: we already trust ourselves (as a community) to select admins, to determine policy and to ban users, so why can we not remove admins at will? Walton^One 19:54, 7 October 2007 (UTC)

The "People's Court" - you mean community consensus? The idea that most of Wikipedia is based around? If adminship is no big deal, why does it take the highest authority on Wikipedia to remove it? Mr.Z-man 14:45, 8 October 2007 (UTC)

It's been tried before

Wikipedia:Requests for de-adminship

This remains an attempt to vote unpopular admins out and as such is very unlikely to fly - David Gerard 10:39, 7 October 2007 (UTC)

WP:CCC - just because it's been tried before doesn't mean it won't work now. In terms of an "attempt to vote unpopular admins out", that's a very simplistic way of looking at it; also, I believe that all admins should be accountable to the community, and should not be admins unless the community has confidence in them. Yes, a user with a vendetta could start a frivolous request; however, with a full community discussion process in place, a frivolous RfDA would be unlikely to succeed. Walton^One 10:50, 7 October 2007 (UTC)

Have you seen how quickly community discussion have been enacted without long discussion and without many contributers. This is not a good system and it is gameable. No, not going to work, and not workable. The current system works fine, its not broke, don't fix it. I fyou want someone desysopped, WP:RFAR is a couple of doors down. Mercury 18:24, 7 October 2007 (UTC)

The ArbCom doesn't determine community trust, the community does. If you think some kind of community-consensus- driven process to determine and express ongoing trust (or a lack thereof) is not a good idea, then how do you justify RfA? — Dorftrottel⁠ 18:58, 7 October 2007 (UTC)

Exactly. Admins should hold office only while they enjoy the confidence of the community. If the community decides, collectively, to withdraw that confidence, then they should lose office. Walton^One 10:12, 9 October 2007 (UTC)

Even if they act per policy and take unpopular actions? Vegaswikian 19:14, 9 October 2007 (UTC)

Isn't that concern solved by having the bureaucrats as gatekeepers to the process? Abuse of the process is dealt with the same way that POINT-y abuse is handled everywhere else. Thoughts? --Bfigura ^(talk) 06:25, 11 October 2007 (UTC)

Hmm.. after reading everything below, and the resulting links, it would seem this is largely an academic question. Feel free to ignore the above. Best, --Bfigura ^(talk) 06:33, 11 October 2007 (UTC)

Problems

Trusted users? This is building a hierarchy. A better idea would be to simply place every admin on Category:Wikipedia administrators open to recall. This process would also take about a week to de-admin someone--Phoenix 15 10:40, 7 October 2007 (UTC)

I'm not trying to create a new class of users. We already have the users in what I called the "trusted users" category. Bureaucrats and arbitrators already have the community's trust and deal with similar situations as this. If you disagree with the wording, feel free to change it to something less hierarchal. However, with the potential for abuse in a de-adminship process, letting any user start these discussions would lead to way too much abuse. Mr.Z-man 16:33, 7 October 2007 (UTC)

Adamantly oppose "simply placing every admin on CAT:AOTR" ... the voluntary recall process works, and works well, precisely because it is voluntary. I encourage all admins to decide for themselves whether to join it but completely understand that not all admins will choose to. What of just having all admins get reconfirmed by the community every year like many other wikis do? ++Lar: t/c 20:47, 9 October 2007 (UTC)

There is no established need. No numbers to support a problem. So I'll say again, this is a solution in search of a problem. Vegaswikian 21:39, 9 October 2007 (UTC)

Would you wait until you get in a car accident to wear a seatbelt? Why do we have to wait until we have an urgent need for something before we can even consider it? Mr.Z-man 14:22, 11 October 2007 (UTC)

Failure to wear a seat belt can result in death, a serious problem. Also, there are numbers to back up the risk of not wearing a seat belt. The problem here is not as serious and the extent of it is not clearly know but it may be very small and the need to change is not clear. Lets not compare apples to oranges. Vegaswikian 00:41, 12 October 2007 (UTC)

There may well not be a need to change, and you're certainly correct in assuming that it's not your responsibility to prove a lack of need. But there might be such a need, so unless you think this process would be worse than what we have, why oppose it? -Amarkov moo! 00:44, 12 October 2007 (UTC)

Yes, the fix 'in search of a problem' is worst then the status quo. Vegaswikian 00:57, 12 October 2007 (UTC)

No it isn't. When you have something you want to work well continually, you regularly maintain it. You don't insist on the status quo until something fails. -Amarkov moo! 20:11, 13 October 2007 (UTC)

This is admins recall

This is just admin recall. Replace step 3 with an actual RFA (which it is equivalent to), and it is just a way of forcing people to go through RFA again. That's not automatically a bad thing, but if you're proposing a system of recall, then call it recall, don't invent a new process identical to RFA just with "support" and "oppose" swapped around. --Tango 12:25, 7 October 2007 (UTC)

My opinion is that RfA needs to be reformed so that instead of a "yes" or "no", bureaucrats have recourse to a third option in controversial cases to give them the tools and see how they go, and they are subject to review after three months and another RfA if the community fails to reconfirm them. This is for those borderline 70%ers and the like. Orderinchaos 16:39, 7 October 2007 (UTC)

I agree that "Request for re-requesting adminship" would be more to-the-point. — Dorftrottel⁠ 18:13, 7 October 2007 (UTC)

Who makes the call?

Have the authors of this proposal actually spoken with any Arbs or 'Crats about their proposed involvement in this process?

Are there any bureaucrats that want this responsibility? It will come loaded with substantially more grief, bickering, and abuse than their current role. The firestorm that results from promotion of borderline or controversial candidates will pale into relative insignificance when compared with a bureaucrat call on Request for Desysopping.

Bureaucrats that currently exist were not approved by the community with the expectation that they would exercise this extraordinary new authority. I would not be surprised if there was push-back from the community on that basis. On the flip side of that problem, I note that otherwise-superb bureaucrat candidates see their requests for 'cratship sunk by arcane discussions surrounding the minutiae of 'discretionary ranges' and absolute thresholds for promotion of adminship candidates. Will we ever again successfully promote a 'crat when the debates and discussions expand to include handling of desysopping as well?

Arbitrators already have an established, formal place in Wikipedia process. They are expected to weigh and evaluate evidence in the context of Wikipedia policy, guidelines, and practce. They aren't supposed to count !votes in a referendum or recall, and they don't work under deadlines. If you want an Arb to consider a desysopping, then file for Arbitration. TenOfAllTrades(talk) 17:16, 7 October 2007 (UTC)

Well, if it were up to me, we'd have a straight vote, meaning that no discretion would be needed in closing it. Sadly, due to the inbuilt cultural revulsion to voting around here, it probably won't work that way. Walton^One 21:40, 7 October 2007 (UTC)

First off, 'crats give admin bits, they are not given the power to remove them based on an intentional division of power. I do not see it as reasonable for them to have the power to "accept" or "close" de-sysopping proposals. I don't think they would even want the grief. Arbcom will deal with these issues, but they already have a system in place and I doubt they would like to move from the controlled environment of arbcom to something considerably less structured.

A straight vote is right out, we are not a democracy. This really is at the heart of the problem here. ((1 == 2) ? (('Stop') : ('Go')) 21:42, 7 October 2007 (UTC)

It isn't just at the heart of the problem, it is the problem. This principle of "Wikipedia is not a democracy", which no one has ever satisfactorily justified, has done more harm to the encyclopedia than Willy on Wheels. Because everyone screams, rants or cowers when they hear the dreaded word "vote", we endure processes that are alternately authoritarian, bureaucratic, confusing, inconsistent or downright unfair. I realise I'm in a minority of one on the English Wikipedia in suggesting that voting might not be so bad, but if people actually thought about it rather than just screaming "Voting is evil!", it might occur to them that there's a reason why voting is the accepted method of decision-making in the real world. Walton^One 18:25, 8 October 2007 (UTC)

Likewise, I think there's just as much (if not more) a lot of problem with people not thinking before !voting as not thinking before decrying a vote. :) EVula // talk // ☯ // 21:36, 8 October 2007 (UTC)

Perhaps you can go to WT:NOT and suggest the removal of the idea that we are not a Democracy, but I doubt the community is ready to give it up. ((1 == 2) ? (('Stop') : ('Go')) 21:32, 8 October 2007 (UTC)

I know the community won't give it up, so we should probably end this thread. However, I long for the day when the community sees sense, and asserts its right to govern Wikipedia and elect its own office-holders. That's all I have to say on the matter. Walton^One 22:35, 8 October 2007 (UTC)

Not bad but...

It looks familiar, this is essentially just doing part of ArbCom's job. There is one important advantage and that is the community discussion. GDonato (talk) 18:48, 7 October 2007 (UTC)

Exactly. — Dorftrottel⁠ 18:59, 7 October 2007 (UTC)

Incidentally, are there examples for point 5. of WP:ARBPOL#Scope? How would such a discussion be started? It would certainly require some central location to grant that community consensus actually exists. — Dorftrottel⁠ 19:05, 7 October 2007 (UTC)

Suggestion

Let's leave this as a proposal for the time being. Then we can test the process to see if it enjoys the confidence of the community; next time an administrator needs to be desysopped, we can start a discussion here. If the process actually works in practice, we can adopt it as policy. Walton^One 20:00, 7 October 2007 (UTC)

No, we'll mark it historical, like the last 20 times it's been proposed. It's clear that the consensus is the same as it has always been - that is, that this is bad idea. Raul654 20:07, 7 October 2007 (UTC)

I don't see any consensus. I firmly believe that testing this proposal is the only way to determine consensus. Walton^One 20:34, 7 October 2007 (UTC)

Uh, no, we don't put half-baked proposals into action, and then figure out if there's consensus for them. The onus is on the proposal to gather consensus before we try it, and if you think it does, then you were not looking very hard Raul654 22:27, 7 October 2007 (UTC)

Addendum; read my arguments in the various sections above. Then it might be helpful if you try to counter them, rather than just saying "this is a bad idea, because we've always thought so, and consensus can never change". Walton^One 20:41, 7 October 2007 (UTC)

First off, obviously there is not a consensus to make this historic. Second, ever heard of the phrase consensus can change Raul? With the slow but steady increase in the number of admins and desysoppings, a function such as this one is more needed every day. VanTucky ^Talk 20:49, 7 October 2007 (UTC)

Yes, I have heard that consensus can change, but just claiming it can happen does not mean it has happened. Raul654 22:22, 7 October 2007 (UTC)

There may well be consensus that there is no problem to fix, so a solution is not needed. Vegaswikian 21:28, 7 October 2007 (UTC)

Yes. There may also be evil pink aliens that want to eat us all. Without evidence that those things are actually the case, it's rather silly to state that they could be true. -Amarkov moo! 21:31, 7 October 2007 (UTC)

There is a huge problem with sysops who are incivil and uncommunicative, but don't do anything bad enough to get desysopped under the current system. Hence why RfA standards are far too high at present, because people are reluctant to promote anyone they don't have 100% confidence in, knowing how hard it is to take the tools away. We need an "easy sysop, easy desysop" culture, in which adminship would actually be "no big deal". Admins are not judges, and don't need to hold office quamdiu se bene gesserint. They should hold office only for as long as the community has confidence in them. Walton^One 10:08, 9 October 2007 (UTC)

Misses the point

The idea behind a different method of desysopping is that people don't have to go through numerous hoops to even get a discussion on desysopping. Whatever group we decide to let oversee deadminship requests is likely to require just as much as Arbcom would. If adminship is no big deal, then why do we insist on creating so many hoops to jump through before removing adminship may even be discussed? -Amarkov moo! 20:57, 7 October 2007 (UTC)

Because since it currently is so hard to have adminship removed or to get adminship, it has become too big of a deal. As de-adminship discussions get easier to start, the potential for abuse increases. While obviously frivolous discussions would be shot down after beginning, I don't think making the discussions easier to start is worth the wasted community time or extra stress on admins who already work in controversial areas (where the potential for abuse by people with a grudge would be greatest). If however, we can make adminship no big deal again, this process could be simplified. Mr.Z-man 21:03, 7 October 2007 (UTC)

Arbcom requires that you seek dispute resolution, and provide clear evidence before bringing a case. I don't think of either of those things as hoops, I would oppose removing such requirements. ((1 == 2) ? (('Stop') : ('Go')) 21:44, 7 October 2007 (UTC)

Right, these aren't "hoops", they are possible solutions. Solutions to the problem of behavior, editing history etc. The problem is never that someone has admin rights, the problem may be that they are misusing them. This method assumes the solution going in. I would rather have a method, like the current dispute resolution system, that remains open in terms of solutions. Again, this seems like a "solution in search of a problem" as Vegaswikian says above. - cohesion 21:56, 7 October 2007 (UTC)

• Clear evidence is reasonable to expect, and while I think that requests with no evidence would just be dismissed as clearly frivolous, I'd be fine with requiring evidence. But dispute resolution is often just a pointless, annoying, and painful obstacle. If someone is doing bad things, and deliberately continues after criticism, that's all the useful dispute resolution there will be. Why must I repeat that many times in different venues before I'm allowed to pursue sanction? -Amarkov moo! 22:30, 7 October 2007 (UTC)

If dispute resolution fails then take it to arbcom, but assuming it is pointless before trying is not AGF. I would not approve any system that resorts to brute force without trying dispute resolution. ((1 == 2) ? (('Stop') : ('Go')) 22:58, 7 October 2007 (UTC)

You're really very much opposed to any of this, aren't you? — Dorftrottel⁠ 23:16, 7 October 2007 (UTC)

I though I had made that much clear, yes I am opposed to any of this. ((1 == 2) ? (('Stop') : ('Go')) 05:00, 8 October 2007 (UTC)

I support this idea, I think it should be roughly as simple to remove admin privileges as it is to give them. Tim Vickers 22:39, 8 October 2007 (UTC)

Limited time appointment

Giving people power for life is often a problem. I propose that users get admin powers for a year, after which they have to have another RfA. Arrow740 21:44, 7 October 2007 (UTC)

• While not a perfect solution, a way to test out a person as an admin could be an improvement. My concerns with this would be someone could be on their best behavior during probation and then go bad after their probation. On the other hand, this type of approach could let in many good admins who are not able to get past the current screening process. Maybe you need to move this suggest over to RFA talk where it would be more appropriate. I would guess this has been suggested and discussed before. Vegaswikian 22:18, 7 October 2007 (UTC)
• Redoing RfA isn't going to work. In the course of admin duties, an admin will tick off quite a few people, no matter how right the admin is (which is usually the case, no matter who you argue it). We should not punish admins for enforcing policy. --71.141.117.207 22:28, 7 October 2007 (UTC)

So you don't think the people can be trusted to choose their leaders. To clarify, I'm proposing that every adminship expire after a year and would have to be renewed each time. Arrow740 22:45, 7 October 2007 (UTC)

I strongly oppose removing admin tools from people who have not done anything wrong automatically like this. Admins are not "leaders", they are people who have special access to tools. ((1 == 2) ? (('Stop') : ('Go')) 22:47, 7 October 2007 (UTC)

Unfortunately, they are. They block users and lock articles all the time, thus controlling wikipedia. Arrow740 22:51, 7 October 2007 (UTC)

Blocking users and locking articles is not control. Most of the time admins have actually been requested by non-admins to do such. That's not control. --71.141.117.207 22:53, 7 October 2007 (UTC)

I'm not sure about that "most of the time." A problem arises when an admin believes that a user is being "incivil" and blocks him for an extended period of time. This happened recently with User:Matt57. In the aftermath, some admins have said that admin decisions should not be critiqued so as to maintain "harmony" and the appearance that wikipedia is being run well. Read about it here. This betrays a worrying attitude. The main issue is not the clearcut enforcement of wikipedia guidelines. The problem is making decisions restricting other users' access in more arbitrary situations. Arrow740 23:01, 7 October 2007 (UTC)

I think I used the wrong word: "redoing" means "going through it again." I hope that clarifies, because I didn't quite get your statement. --71.141.117.207 22:50, 7 October 2007 (UTC)

I prefer any user being able to start a request for confirmation which would work just like an RfA at any time. However, I would support this proposal, with the reservation that I think three or four months with the tools before having to go through another RfA is more appropriate than one year. a.z. 22:54, 7 October 2007 (UTC)

Has been proposed. The problem is that this would immensely increase the workload concentrated on RfA. Three or four months couldn't be possibly handled using the current RfA process. — Dorftrottel⁠ 23:14, 7 October 2007 (UTC)

I don't believe this, but there would be no problem with making it a three-day process, since we would just have to decide whether that person should have the tools for three more months, and that's not such a big deal that needs a long and complicated process to be done. a.z. 00:20, 8 October 2007 (UTC)

My preference would be for a reconfirmation after two years. Every year, and it would become a meaningless rubber-stamp session.--Nydas^(Talk) 17:24, 17 October 2007 (UTC)

There is still not strong reason for every admin to need reconfirmation. There is a process for admins who by their actions should have their access to the trusted tools removed. The problem is getting needed new admins through the review process. If they can be placed in the position for a period of time and then under go a one time review asking should they be removed, we would be able to approve more admins without significantly increasing the workload in the review process. Vegaswikian 19:53, 17 October 2007 (UTC)

Any 'bad' admin will simply wait until after the review before misbehaving. In my experience it's the established admins which are the most likely to be problematic. Your review would just give them another badge to hide behind.--Nydas^(Talk) 21:04, 18 October 2007 (UTC)

Change to step 3

I propose that step 3 be changed such that instead of a normal RfA (that one must get a consensus to support the candidate for adminship), it should be the other way around (one must have a consensus to oppose the candidate in order to get desysopped). A vocal minority should not be able to overturn a majority of the community, but must rather be able to convince the community that grave transgressions of Wikipedia policy have occurred; therefore, a consensus to oppose the furtherance of the user's admin privileges will display the community's dissatisfaction with the user's admin actions, etc. --71.141.117.207 22:38, 7 October 2007 (UTC)

Also, don't hound me for using "majority" as I have used in the above post; the dictionary definition of the word was intended, not the Wikipedia definition of it. --71.141.117.207 22:40, 7 October 2007 (UTC)

Doomed proposal

Unfortunately, due to the resistance of a vocal minority, this will go nowhere as usual. A shame however, since in my opinion desysopping simply for a lack of ongoing community trust is perfectly in order. Failing at RfA doesn't always involve proof of any serious misconduct, a lack of community consensus is all it takes for an RfA to fail. So why shouldn't adminship be as easy to withdraw in cases where community trust has expired due to continued minor things? Ach, whatever. — Dorftrottel⁠ 23:22, 7 October 2007 (UTC)

"Vocal minority"? While I admit we might not the majority, but insinuating a nonexistent "silent majority" on your side is doing everyone a disservice. --71.141.117.207 23:53, 7 October 2007 (UTC)

Yeah? How? — Dorftrottel⁠ 00:17, 8 October 2007 (UTC)

Because what you are implying simply does not exist. --71.141.117.207 06:54, 8 October 2007 (UTC)

Tell that to a religious person. — Dorftrottel⁠ 15:28, 8 October 2007 (UTC)

I am religious myself. Faith has nothing to do with the perception of majority opinion. But we digress. --71.141.117.207 17:40, 8 October 2007 (UTC)

Simplification

I think that this whole proposal could be simplified to:

• Since we already have the Arbitration committee in place, Allow them to "hear" cases at Requests for administrator review, in addition to Requests for arbitration. The criteria for placing the request is the same as placing a RfAr, that is, that other forms of dispute resolution have already been tried.

Everything else is essentially the same, the only difference is what the expected outcome is: Whether the admin in question should or should not be desysopped. - jc37 03:47, 8 October 2007 (UTC)

The same as this proposal or the same as an RFAR? Mr.Z-man 04:27, 8 October 2007 (UTC)

The general goal or outcome of a request for arbitation is to determine wrongdoing (if any) as a result of an event or set of events.

The general goal of a requests for administrator review would be to determine if the administrator in question is abusing the trust the community entrusted them with. (Not necessarily a single event or set of events.) This would also allow for a broader selection of possible results than just de-sysopping.

As for format, the current Arbitration system would seem to work fine (main page, workshop, evidence, proposed decision)

AFAICT nothing in what I said above is contrary to what's on the proposal, except that it's the Arb committee, rather than a single user who makes the final determination, and it expands it to take advantage of all the tools we already have in place (such as the clerks system, etc).

I hope this helps clarify. - jc37 04:44, 8 October 2007 (UTC)

Hypotheitical Example

Let me give an example from the recent past: User:Jayjg was accused by several to have possibly abused his checkuser priviledges during an RfA discussion. There was quite a bit of discussion about it in various places, resulting in This RfAr (I'm intentionally not researching it further, since I'm really only using it as a hypothetical.)

So let's presume that there was an WP:AN (or some sub-board) discussion about it, which then resulted in an RfC about it. Mediation would likely be inappropriate in this case, since this isn't about a content dispute, or having editors "get along". So the next step, depending on the outcome of the RfC would be a request for administrative review. The arbitrators decide whether to accept or not. If accepted, it follows the same format of a request for arbitration, but with it roughly limited to the question at hand (though the question could obviously be expanded if new evidence comes to light). Once that's done, the arbitrators make their final decision. Note that this doesn't preclude a simultaneous arbcomm case if there were also events related to this which warrant it (if, for example, someone was disruptively uncivil, or whatever).

There are more benefits that I can think of, but I think that this is good for now. - jc37 05:03, 8 October 2007 (UTC)

Looking at the history of that and other cases, i do not see how it would have helped--the matter would have gone just the same, and no simpler. Actually, I think the main thing needed is to provide a better t=ace for extended discussions, rather than a new procedure. We could simply branch discussions from ANI to a subpage, without the formality of RfC. After that, either the opinion of people as expressed there would convince someone to change his ways, or there would be reason for a RfC or RfA. Don't downrate informal methods--lots of people here--admins and other editors too--have improved their work after hearing what people have said about it, on or off wiki. The last hing this place needs is a further elaboration of formal procedure. People are always complaining that its too complicated and inefficient, and the proposals for reform tend to create yet more of the same. We need freer and more general expression of opinion, and we can do that ourselves. DGG (talk) 09:19, 8 October 2007 (UTC)

Actually, if you look over that specific RfAr page and its subpages, it's essentially what I was proposing. The focus was on whether the tools were abused. (Though it meandered a bit.) That aside, I was only using the event as a hypothetical, I wasn't attempting to go into the process that was used at the time.

As for "ourselves". last I checked, that included arbcomm and every other Wikipedian. This isn't and shouldn't be about "us and them", it should be about all of us : ) - jc37 14:01, 8 October 2007 (UTC)

I think it will go better if it does not directly feed into formal process. All admin actions are visible in summary--just go to an admin's page, and look at the log. And by ourselves, I meant each individual one of us, starting here and now. Though if arb com is going to eventually rule on things, its members probably should not join the informal discussions, just as an admin should not join a AfD discussion he intended to close. DGG (talk) 15:00, 8 October 2007 (UTC)

ArbCom isn't really community-consensus driven, which as best I understand, was the whole point of this proposal. shoy 14:06, 8 October 2007 (UTC)

I agree with you about that--but no matter what is done about this process, they will have the last word in the end. DGG (talk) 22:54, 8 October 2007 (UTC)

User subpage

Personally, I've always thought that discussions focused on a user's conduct, or on a user's use of admin tools, should be a subpage of that user's page, and linked from the user's page. That way things are obvious and transparent. Currently, if you click on "what links here" for a user's page, you get links to all that user's signed comments. Which is unhelpful if you are looking to see if the user has past history of poor judgment. You could look through the contributions, or look through the history of the talk page, but it is easy to miss things. The same goes for RfA. But it has always seemed silly to me that RfA is the only point where an editor's contributions are (theoretically) scrutinised thoroughly. There is editor review and Arbcom and RfC, but people should check other people's contributions more if they think there is a problem. The trouble is, most people I know (including me), find it difficult enough to keep track of their own contributions. Carcharoth 13:46, 8 October 2007 (UTC)

I agree that it would be great if you could filter a user's signature out of "Whatlinkshere". The subpage would be a problem because based on current rules of speedy criteria, they could request it immediately deleted. I also think it would be nice if we could compact a contributions page to list only their last edits to a particular page (and that includes my own). But I suppose those are proposals for another place : ) - jc37 14:01, 8 October 2007 (UTC)

Finding a concensus

I really have no idea how all this policy fenangling works. It looks to me like a handful of users are trying to say that there's no consensus support for a policy like this one, going so far as to mark this proposed policy rejected, but over on the RFA RFC running right now there are over 50 users who support a view advocating a community-based desysoping process. Do those users just not count over here, or is this not the process we were all advocating? I wanted to post over here to explicitly say that I support this proposal, or any proposal like it that gains enough steam to actually be implemented. Adminship is no big deal, and if taking it away weren't such a big deal people might not be so eager to treat it like a big deal in the first place. ➪HiDrNick! 19:41, 9 October 2007 (UTC)

That rfc tells us little, it has a place for a nice numbered list of "Support" votes, yet the format does not provide a similar location for opposition. Despite this in-balance in the format, there has been significant opposition on that view of the rfc. Looking at this talk page, there certainly does seem to be a lack of the consensus needed for such a major policy change. ((1 == 2) ? (('Stop') : ('Go')) 20:09, 9 October 2007 (UTC)

For the record: there appears to be no consensus and you appear to be seeking to prevent that consensus. — Dorftrottel⁠ 23:04, 9 October 2007 (UTC)

How about we have a poll on it? (No doubt several people will announce that it's not binding, but a large majority in support would certainly be persuasive. In fact...) Walton^One 21:24, 10 October 2007 (UTC)

I would support conducting a straw poll. VanTucky ^Talk 21:29, 10 October 2007 (UTC)

Yes, why not. — Dorftrottel⁠ 22:13, 10 October 2007 (UTC)

Straw poll

(Poll moved to Wikipedia talk:Requests for adminship.) Walton^One 21:48, 10 October 2007 (UTC)

Admin Oversight Board

There's been a proposal on the Wikipedia:WikiProject Council/Proposals page regarding this idea languishing for some months now, the Proposed admin oversight board gone into detail here. It actually has several people indicating support, although the proposer stopped editing back in January or so. If any of that proposal could be integrated in to this one, some of the individuals who expressed interest in it might be willing to act in whatever capacity is called for. Also, for my own strictly selfish purposes, I could then remove that seemingly dead proposal from the proposals page as well. John Carter 22:04, 10 October 2007 (UTC)

In reading over the discussion, I don't see much support. As for your "problem", why not just create an archive for them, leaving a note on the main page that anyone can revive an archived proposal (though not the supporters or discussion due to too much time passing). - jc37 22:17, 10 October 2007 (UTC)

Probation or mentorship

The only way that admins can really make difficult decisions that are correct is if they can reasonably expect that they will not need to defend every decision like this in some type of review board. It is bad enough the number of user and talk pages of admins that get vandalized probably upset by correct decisions that have been unpopular. If an admin is really bad, then it should be noticed and action taken, but adding a term or a regular review by every editor is unreasonable.

As a possible solution, lets free up the bottleneck for new admins and maybe at the same time improve the quality of their work. So lets consider two possible changes for new admins instead of the proposals that affect every admin and will actually reduce the time that admins have to do admin activities.

The first could be a probation period for new administrators. After a set period of time, their admin actions could be reviewed and if they show that they understand the policies, there responsibility and how to deal with other editors, they can be made permanent administrators. Or they could be given extra probation time. Or if they are really bad, dropped as an administrator. This only needs to be a very short review and discuss inappropriate actions taken. Mistakes can happen, so the review would look at these, how the admin adjusted and how they learned from those mistakes.

Another option would be to assign them a mentor for a period of time to help them acclimate to becoming an admin. This has problems in that it would require frequent monitoring and communication by an experienced admin. Vegaswikian 20:46, 13 October 2007 (UTC)

Admins should be prepared to defend every difficult decision of theirs. If it's a difficult decision, then it needs discussion, not just one admin acting how they think is right. -Amarkov moo! 20:48, 13 October 2007 (UTC)

However there is a trade off. How much time needs to be spent in those discussions? If the amount of time for the administrator is too high, then it is better to not make a decision. I don't think that is the direction most editors what to be pushing administrators. Why should an administrator have to defend every difficult but correct and proper decision? We should be worried about the bad decisions or actions. Vegaswikian 21:17, 13 October 2007 (UTC)

That's the point of discussion, to determine if a decision was correct and proper. I don't advocate discussing things everyone knows were correct. -Amarkov moo! 21:24, 13 October 2007 (UTC)

And that is the problem. Just because someone perceives a decision to be incorrect does not mean that it is. I'll hazard a guess here. For every decision/action any administrator takes, someone will think it is wrong. Using your logic, that means every decision then needs to be discussed. Sorry, but that is unrealistic. Vegaswikian 22:59, 13 October 2007 (UTC)

Problem is, the alternative would be demanding that some actions not be discussed. And that's no less unrealistic. -Amarkov moo! 23:18, 13 October 2007 (UTC)

I find that time spent explaining actions before or as they are taken is a valuable investment. While it won't necessarily discourage those with an axe to grind from grinding it, it will make defending each other much more practical. --bainer (talk) 22:57, 13 October 2007 (UTC)

Modification

Suppose

• Anyone can call a RfDA
• It needs a quorum to be put to discussion: I.E. those that think there is a discussion to be had, perhaps 3 or 5 admins?
• At that point it follows the normal discussion rules: no socks, looking for consensus, generally time limited, closed by an admin, implemented by any-one with the appropriate power.

Obviously all the normal WP caveats apply - the proposer or proposed, or others involved won't close the discussion. NPA. AGF. etc. etc.

Seems easy, can't really be gamed (any more than is inherent in the software). Abuse of process won't get past the first hurdle without abusive admins - even the worst are not going to make more than one or two frivolous RfDA's quorate. Unsuccessful RfDAs might still act as a wake-up call to problem admins. If the process is simple, we don't invest too much time in it, if, as some maintain, it will never be needed.

Rich Farmbrough, 13:07 15 October 2007 (GMT).

If it requires admins to remove admins, the community is not likely to be happy with it. --Tango 21:11, 18 October 2007 (UTC)

Essay

Since this is an essay by definition, both in the manner it is written and the fact that it has not been circulated amongst the wider community, it needs to marked as such which I've taken the liberty of doing. Odd nature 23:57, 22 October 2007 (UTC)

perhaps it is an essay, but it is also a proposal, one which has been rejected as is clear by this talk page. Perhaps the essays tag should be returned, I don't think so, but I do think it is evident that this is a proposal which has indeed been rejected. 1 != 2 18:20, 7 November 2007 (UTC)