Wikipedia talk:Removing administrator rights/Proposal

One small step

I propose this as an intentionally conservative option, in hopes of this having wider support. Crat discretion is still an essential part of this. The crats specifically don't want to get involved in cases like this without an explicit mandate, so here's my attempt to show that such a mandate exists. Friday (talk) 17:55, 2 September 2008 (UTC)

I certainly support the idea. Tombomp (talk/contribs) 18:11, 2 September 2008 (UTC)

How will it be shown someone's confidence has been lost? Majorly ^talk 18:14, 2 September 2008 (UTC)

My intention was that the crats make that call. Maybe the phrasing needs work. Friday (talk) 18:17, 2 September 2008 (UTC)

Oppose

That is not what 'crats are for, that was not the criteria upon which we selected them. The community can demonstrate it's grievances with arbcom. Besides, a 'crat can already ask a steward to desysop, it just is not likely to get them anywhere(if anything it should be reworded to clarify if the 'crat is advising or commanding the steward to act). Chillum 18:29, 2 September 2008 (UTC)

The crats have said they will not do so, without an explicit community mandate. Friday (talk) 18:31, 2 September 2008 (UTC)

That is very wise of them. Chillum 18:43, 2 September 2008 (UTC)

So far you just sound like you're objecting because this is new. Do you have any more substantial objection? Would this be harmful in some way? I see this only as a slight addition to traditional crat powers. They were chosen to gauge consensus on who should be an admin. This proposal is just a logical conclusion from that- crats judge consensus on adminship. Consensus can change. Doing the math is not hard. Friday (talk) 18:46, 2 September 2008 (UTC)

I said "That is not what 'crats are for, that was not the criteria upon which we selected them". That was my reason for objecting, not because it is new. Chillum 21:08, 2 September 2008 (UTC)

• Strong Oppose. As per the many other times something like this has been floated. Administrators who have been active for a significant amount of time will almost have undoubtedly had to make difficult decisions that upset the individuals or subgroups against whom the decision was rendered. Furthermore, we prize administrators who continue to act as editors. Most editors will be involved, at one time or another, in tendentious and controversial topics. Putting both of these aspects together, it is distinctly possible that a sizable enough contingent of upset editors can cause repeated claims against sysops. Sysops need to be able to act in the best interests of wikipedia without a fear of their appropriate and proper actions being held hostage against them. It is the Arbitration Committee who is tasked with the final decision as to the appropriateness of any editor's actions—including sysops' actions. It is solely their (together with Jimbo for accepted historical and benevolent dictator reasons) determination to permanently remove the sysop bit (stewards may remove the bit, temporarily, in emergencies, and need to be very certain about their actions). The community is empowered to choose its sysops, and we see how fragmented that discussion can be at times. To allow it to be the arbiter of bit-removal would, in my opinion, cause too much paralysis and bureaucracy for anything to be done. For the greater goodof the project, the decision to remove the sysop bit needs to remain in ArbCom's hands, although, as I have proposed many times before, I believe it is incumbent on ArbCom to create a streamlined process, which may need to include a subcommittee of editors authorized by ArbCom to act as a tribunal and apprise ArbCom of their analysis, which ArbCom may choose to rubber stamp. -- Avi (talk) 18:47, 2 September 2008 (UTC)

  I'm inclined to agree with Avi; while tossing the ability to desysop to the masses (for lack of a better, less elitist phrase) could be crippling, it also opens the door for attacks against perfectly fine administrators who have dealt with contentious situations in the past; I know that there are probably several people with axes to grind against me, much less a more active administrator. A streamlined process that is dedicated to sifting through evidence strikes me as a better option than a RfA-shadowing process. EVula // talk // ☯ // 19:02, 2 September 2008 (UTC)

Err.. you guys did notice the bit about "the bureaucrats can, at their discretion, ask a steward to remove the admin permission", right? This is a an intentionally small change, in response to particular statements by the crats that they won't touch this without an explicit mandate. It's explicitly not about giving this power to the community. I thought it was viable for that very reason- I can't see any way to argue that such a policy is harmful. The worst that could happen is that the crats would not choose to use it. This is about untying their hands and letting them use their discretion. Friday (talk) 19:18, 2 September 2008 (UTC)

Friday, bureaucrats may ask a steward now, like any editor. If the steward removes the bit in a non-emergency situation, the steward will be held to task for overstepping their bounds. Furthermore, from where will this mandate arise, if not the community? So the community will be demanding that bureaucrats demand of stewards to remove the bit? The community may already approach ArbCom via an RfAr, so what is gained? What we need is that ArbCom respond more quickly and more efficiently, not to be sidestepped, in my opinion. -- Avi (talk) 19:28, 2 September 2008 (UTC)

That would be nice also. However, I find that when trying to solve problems, you can often save time by trying several approaches at once. I'm not remotely suggesting this is the only viable solution. I just thought it was a small change that would do no harm, and thus it may be worth pursuing. Friday (talk) 19:34, 2 September 2008 (UTC)

I understand, Friday, and you are correct in that the process as it stands now has distinct flaws. I personally believe, however, that the change you suggest will be misused by others to cause more harm than good, and that a more efficient implementation of the current procedure is what is required. -- Avi (talk) 19:58, 2 September 2008 (UTC)

This is unwise. The arbs were voted on largely for their ability to exercise this kind of sober good judgment. Chillum correctly points out that this was not part of the criteria by which the 'crats were chosen. Further, having an ad hoc collection of interested people recommend desysopping is a polite way of describing a really poor decision-making process. The collection of interested people, and the nature of the process, will inevitably make it about the man personally. This type of community-based approach has historically enforced a homogeneity that would be incompatible with our project, which needs a healthy mix of views. It would have a chilling effect on admins enforcing blp and copyright. Leave it with arbcom. In spite of its imperfections, that's more likely to keep the focus on what people do instead of who they are. Tom Harrison ^Talk 13:27, 3 September 2008 (UTC)

"The arbs were voted on largely for their ability to exercise this kind of sober good judgment." So why is it a problem giving them this kind of decision making ability? I honestly don't understand these objections I'm seeing. Do people think this is just-another let's-vote-people-out-of-adminship proposal? Friday (talk) 13:53, 3 September 2008 (UTC)

The arbitrators were voted on largely for their ability to exercise this kind of sober good judgment. The bureaucrats were not. It may not be what those drafting it intend, but I do think this will become a !vote for desysop. Saying explicitly 'this is not a vote' won't help any more than it helps anywhere else we !vote. Tom Harrison ^Talk 14:02, 3 September 2008 (UTC)

Bah. I read "arbs" as "crats". Here I was thinking "oh no, someone misread this" and I completely missed the word you actually wrote. I thought crats were chosen to exercise good judgement in deciding who becomes an admin. But, maybe I'm one of the few who thinks this. Other people seem to want them to merely count votes. The main reason I think this is a good idea is the (real or perceived) lack of effectiveness of the arbs at solving these kinds of problems. But, it does only work if we have good crats and people trust them to be reasonable. Friday (talk) 14:15, 3 September 2008 (UTC)

(ec) You said it yourself—the 'crats were chosen because they were expected to exercise good judgement in deciding who should become an admin, on the basis of advice from the community and within a fairly well-established – albeit slowly evolving – set of standards.

The 'crats were not elected on the basis of how they would choose to remove the sysop bit. None was asked about it during the RfB process, and no member of the community (as far as I know) evaluated any 'crat candidate for their suitability in such a role. Aside from Jimbo, there's no member of the Wikipedia community entrusted with the power to decide to revoke adminship singlehandedly. (I don't include in this temporary desysoppings which take place when it is suspected that an account is compromised.) Even the ArbCom has to hold a vote to pull someone's bit on a non-temporary basis.

Members of the ArbCom are elected by the community and thorougly vetted over a month-long process. They are required to reveal their identities to the Foundation and have access to private information that even 'crats can't see. Participation in ArbCom elections is even greater than that seen for RfBs (the average number of participants in this year's successful RfBs is 132; no successful 2007 ArbCom election candidate saw fewer than 200 voters, and Newyorkbrad received more than 550 endorsements). Even then, the ArbCom regularly faces heated criticism (torches, pitchforks, small improvised explosive devices, singing telegrams) for their decisions to desysop (or not desysop) admins who have done something controversial. The community doesn't have uniform standards or expectations for desysopping admins. (In a way, this is a good sign. We generally are happy with the performance of our admins, so desysopping is an exceptional occurance.) Expecting 'crats to deal with the breathtakingly wide range of circumstances under which a desysopping might be contemplated and the shifting goalposts of a fickle community's expectations just isn't reasonable.

I would demand the recall of any 'crat who attempted to exercise such authority, as it would be prima facie evidence of poor judgement. TenOfAllTrades(talk) 15:33, 3 September 2008 (UTC)

This is not meant to be the only way. If arbcom is the right place for a given case, send it to arbcom. This is just an additional way to deal with the problem. I figured that by promoting candidates, and seeing the results of this, crats have come to have pretty good judgement on who should or should not be an admin. This is an attempt to leverage that expertise, which I assume already exists. Some cases are obvious and don't require months of arbcom activity to sort them out. I figured the crats were in a good position to do something useful with those cases. Friday (talk) 15:40, 3 September 2008 (UTC)

To be fair, everything in my experience points to the community expecting sober, good judgement from its bureaucrats. EVula // talk // ☯ // 15:02, 3 September 2008 (UTC)

Bureaucrats deciding who should and shouldn't be an admin based on their interpretation of a community discussion? Doesn't seem like so much of a stretch. As long as the crats are willing to determine consensus in RfDesysopping discussions, let 'em. If a crat doesn't want to close a discussion for fear of heated criticism, retribution, or whatever, let another do it. If none are willing, it's probably a good candidate for ArbCom. The crat wouldn't have to guess the community's expectations, community would state its expectations as arguments for or against desysopping, in idea 2 at least. WODUP 21:53, 3 September 2008 (UTC)

• Strong oppose This is just one more reincarnation of a very bad idea. The balloon has huge holes in it, you can not inflate it, stop trying. Factual errors exist in the assumptions presented. Bureaucrats do not determine consensus in RfA... they just count the little magic numbers and do some simple math. It's all one moron, one vote. So the rest of the recommendation from there falls apart like the sweater with the pulled string. Jerry delusional ¤ kangaroo 03:42, 13 September 2008 (UTC)
  • You reject the proposal because of your incorrect opinion as to the role of the 'crat? Giggy (talk) 03:47, 13 September 2008 (UTC)

How is my opinion incorrect? Crats just count and find percentage of support. They absolutely do not determine who makes wiser arguments and weigh the votes differently. It is truly one vote per participant, and drive-by "*oppose. signature" votes get as much weight as *oppose because of this diff where the candidate said this, and then that and because of this policy and that discussion, et cetera". Jerry delusional ¤ kangaroo 13:18, 13 September 2008 (UTC)

(Ignoring IP rant below.) Mmhmm. I don't know where on Earth you get the idea that they'll give an oppose with no commentary the same weight as an oppose with evidence. Care to provide an example? Giggy (talk) 04:05, 14 September 2008 (UTC)

I just want to let the community know how Jerry does that counting he is talking about ("Bureaucrats... just count the little magic numbers and do some simple math"). Really? Please see this last month deletion: [http://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/NimbleX] and count the fact that 2 votes opposed the deletion and no vote for deletion, but Jerry simply proceded with deletion even if the 2:0 was clear vote to keep that article. Based on this consideration, I suppose Jerry vote here doesn't really count... --- 24.87.105.114 (talk) 06:20, 13 September 2008 (UTC)

You seem to be a little confused again. I was talking about bureaucrats determining the outcome of voting (note NO exclamation point in front of "voting") at RFA's, and you are discussing how Administrators determine consensus (sometimes called "!voting") at XfD's. The situation you mentioned here has absolutely nothing whatsover to do with this discussion at all. In the future, you might want to read the page that you are commenting on, and ask for help if you can not understand it. Jerry delusional ¤ kangaroo 13:18, 13 September 2008 (UTC)

Reworded

Nearly everyone who has commented has not taken this the way I meant it to be understood. I've redone the wording in hopes of addressing this. Friday (talk) 13:56, 3 September 2008 (UTC)

I believe my concerns as described above still stand. Thanks. -- Avi (talk) 14:18, 3 September 2008 (UTC)

My concerns as described above also still stand. I don't like how it just says a 'crat can decide this without defining criteria. I would prefer a system where evidence is applied to policies to make a decision about removing an admin's bit. Also a system where multiple people decide instead of just one thereby providing a sanity check. But we already have that, it is arbcom. If a 'crat wants somebody desysoped, they can just file for arbcom. Chillum 14:49, 3 September 2008 (UTC)

I will add a process to the proposal. A lot of editors don't like dealing with ArbCom, and would prefer a community based process. Jehochman ^Talk 14:56, 3 September 2008 (UTC)

Arbcom is community based. The community makes the complaint, the community presents evidence, the community presents arguments. The only thing the community does not do is decide it with a vote at the end, which is a good thing. Chillum 14:32, 4 September 2008 (UTC)

Hmm, this is departing wildly from what I imagined. I'll try presenting more than one option, I guess. Friday (talk) 15:10, 3 September 2008 (UTC)

I was trying to go with something so simple and basic that nobody could object to it. Obviously, I've been failing spectacularly so far, but I didn't want to go so far as telling crats how to make these decisions. I was trying to solve one simple problem that can actually be observed: the crats won't touch desysopping, because they say there's no clear mandate for this to be part of their job. That's the only problem I was trying to solve here. Friday (talk) 15:29, 3 September 2008 (UTC)

No big deal

Removing administrator access is not a very serious action. It is less serious than blocking somebody, and can easily be reversed in the event of error. Jehochman ^Talk 14:56, 3 September 2008 (UTC)

Maybe you're saying what I've been unable to articulate. With rollback, we have a nice simple easy-come easy-go way of adding or removing the permission. It's just plain weird that when it comes to adminship, the people who add the permission can't remove it. To me, it's just simpler and more streamlined to remove this special case and make it work like the rest of the wiki almost always works. But, I suspect huge portions of the community (including most admins, go fig) do think removing the admin bit is a very serious action. This fundamental difference in viewpoint may account for this not being the no-brainer kind of proposal I imagined it was. I think it should be easier than it is now, to depose a bad admin. I don't think it should be too easy, mind- but this is why I tried to introduce the crats as an extra sanity-check layer between the "rabble" and the decision-making. Friday (talk) 15:08, 3 September 2008 (UTC)

The difference is that it's easy for an admin to piss people off for doing the right thing (do I need to link to my list of insults? Nah, didn't think so...). Rolling back non-vandalism edits? Pretty clear-cut evidence of rollback abuse. Laying the smackdown on borderline abusive accounts? Not as clear-cut. We need to cushion our admins from axe-grinding, which is the primary reason a de-sysopping system has never been put in place (in my opinion). EVula // talk // ☯ // 15:28, 3 September 2008 (UTC)

I would support bureaucrats' disregarding or de-weighting comments from those who are sharpening tools. I think you all are smart enough to do that. Jehochman ^Talk 15:40, 3 September 2008 (UTC)

But if we make it like rollback, where any 'crat can just up and remove the bit, it will (in theory) turn the 'crats into a bit of a power-granting cabal, given our vastly smaller numbers (when compared with admins and rollback). I much prefer the idea of us being executors of community will, and allowing the community to have a say in whether someone should be desysopped or not.
I do agree that we're smart enough to spot axe-grinding, but I'm not unbiased on that front. ;) EVula // talk // ☯ // 16:11, 3 September 2008 (UTC)

Becoming an admin is no big deal, removing an admin bit is a big deal. We don't just take away an admin bit for the hell of it, we need a dam good reason, so yes it is a big deal when it is done. Chillum 14:34, 4 September 2008 (UTC)

So you're concerned about the risk. I understand this. But - we have crats. We count on their good judgement. Once we're already counting on their judgement, it adds no additional risk to count on their judgement again. If they would be so foolish as to remove a bit for no good reason, this is a problem we already have- it would also mean they're promoting or not promoting for no good reason. Friday (talk) 14:38, 4 September 2008 (UTC)

Crat's aren't perfect, and they have never established the communities trust to deadmin. Good judgment in one area does not mean good judgment in another. The additional risk is putting too much power in one person's hands. The fact is that admins are also chosen for good judgment(just like crats), and they are often doing what they think is correct when they abuse their tools(just as a crat may believe they are doing right when they deadmin someone). I think crat's will have the same problem as admins in that their judgment will not always be perfect. Unilateral decision making is not the answer to admins making unilateral decision making. Besides, why bother convincing a crat when you could just convince arbcom. Chillum 13:52, 13 September 2008 (UTC)

Bureaucrats

It might be helpful to clarify the term bureaucrats - does this mean any one bureaucrat? Or if it means a collective decision of bureaucrats, how many are needed and must they be unanimous? WJBscribe (talk) 15:27, 3 September 2008 (UTC)

I was willing to leave that question as vague as it already is with other crat activities. However, I did assume that, this being a potentially controversial thing, they would probably choose to seek consensus amongst themselves. I'm treating the crats as a black-box here- I'm attempting to not care about, or mandate, their internal decision-making process. Friday (talk) 15:35, 3 September 2008 (UTC)

(ec) I think your input would be most helpful. Perhaps you want to have a panel of three bureaucrats review each situation, or one, or the whole group. Feel to propose whatever you think would work best. Normally you folks get together when there is a controversial situation and do a group huddle. That might be appropriate here. Jehochman ^Talk 15:38, 3 September 2008 (UTC)

Requiring all of us, I think, is an excellent way of making sure that we go nowhere fast. A single 'crat, however, places too much power in the hands of those that shouldn't have it (not my opinion, but I can easily see this argument being made). I think a trio of bureaucrats would make for a good balance of power. EVula // talk // ☯ // 16:03, 3 September 2008 (UTC)

Okay, then. "After a period of two weeks, a panel of three bureaucrats will review the discussion and determine the consensus." Jehochman ^Talk 16:59, 3 September 2008 (UTC)

I would be comfortable with that, if only because of the "discussion" part that makes it clear it would be up to the community to make the arguments, leaving the bureaucrats only the task of determining consensus, just like with RfAs. As long as we (bureaucrats) aren't being made to be the end-all/be-all deciders, my concerns about too much power being concentrated in our hands remain addressed. EVula // talk // ☯ // 17:50, 3 September 2008 (UTC)

(<-)I would much rather have ArbCom set up a subcommittee, as I have said a few times before. If ArbCom feels that an appropriate subcommittee for them to rely upon and take their (the committee's) recommendation is one composed of 5 'crats, that's great. But the process should still be a form of RfAr, even if it is brought before a sysop tribunal. -- Avi (talk) 18:31, 3 September 2008 (UTC)

How stewards work, from meta

See meta page, which says: "To request the removal of another user's status, you must gain consensus on the local wiki first. All discussion must be kept on your local wiki. When there is community consensus that the user's access should be removed, a trusted person from that wiki should provide a link here to the discussion, a very brief explanation of the reason for the request, and summarize the results of discussion."

The stewards already allow what I'm suggesting. They've been doing it for years. I was merely trying to give our local crats the explicit mandate to act as that "trusted person". I was honestly a bit surprised that our crats weren't already willing to do this job, since it seems obviously in line with their explicitly-stated duties. Friday (talk) 15:49, 3 September 2008 (UTC)

PS. It's even been suggested that we skip the crats, and appeal directly to the stewards. I'd rather include the crats, as it gives more legitimacy to the whole thing. Crats are trusted members of the community, and they explicitly deal with the question of who should get the admin bit. We already depend on them to separate legitimate reasoning from the crap. This tiny little tweak of their duties introduces no new dependencies. Friday (talk) 15:56, 3 September 2008 (UTC)

This system simply can't work without someone on enwiki making the call, and then putting the request into the stewards. Creating a new position doesn't make sense, and bureaucrats are close enough to the proposed position that it makes sense to expand their duties. (this is wordy way of saying "I agree") EVula // talk // ☯ // 16:06, 3 September 2008 (UTC)

Woohoo! Someone drank the kool-aid. I was hoping it wouldn't be just me the whole time :) Friday (talk) 16:12, 3 September 2008 (UTC)

Let's not go crazy here; I'm not saying you're always right, just in this one instance. :P EVula // talk // ☯ // 17:28, 3 September 2008 (UTC)

The nature of the local discussion is important. I don't think our RfCs, structured as they are, are quite what the stewards are expecting a request to be made based upon. I would be perfectly happy requesting the de-adminiship of someone based on a discussion that looked like this or this (examples from Commons), but I would be hesitant to make a request based on something like this. If RfCs are to be the basis for de-adminship, I think that at the very least the issue of de-adminship needs to be raised and discussed as a discrete point. WJBscribe (talk) 16:17, 3 September 2008 (UTC)

Maybe we should have requests for de-adminship like commons? :) Yeah, like people would agree with that... Majorly ^talk 16:25, 3 September 2008 (UTC)

It would be the domain of bureaucrats to specify the input parameters and formats they would require in order to make a good decision. I feel confident that these details could be worked out. What I like about the RfC format is that there is a certification step that filters out some frivolous complaints, and we look at opinions, rather than votes. Yes, in order to desysop, the opinion should clearly state that as the desired result, and the opinions requesting desysop should represent a consensus. If however bureaucrats prefer the commons voting format, I could also support that. Jehochman ^Talk 17:01, 3 September 2008 (UTC)

Has there ever been a consensus that anyone should be desysopped? By "consensus" here, I mean going by the numbers. Even our most obviously incompetent admins have sometimes gone down still sporting large crowds of supporters. I have no faith in numbers, personally- but I'm willing to trust crat discretion. Friday (talk) 17:08, 3 September 2008 (UTC)

The job of bureaucrats - and where that discretion lies - is in assessing consensus. Are you actually suggesting that bureaucrats take over from ArbCom in determining whether people should be desysopped based on their own judgment? I guess the question is whether, under your system, bureaucrats would be evaluating the evidence or the discussion... WJBscribe (talk) 17:11, 3 September 2008 (UTC)

Seeing as the RfAs to request removal of the bit from dormant accounts failed, I suspect any request to desysop an active admin will be even more controversial. I'd recommend that we only ask the crats to evaluate the discussion (the consensus of it), instead of placing them in an investigative role of judging the evidence brought forward. If the evidence is unclear enough to require a judgment call, that is more the domain of Arbcom, as opposed to a consensus interpretation, which is the crat domain. MBisanz ^talk 18:06, 3 September 2008 (UTC)

I like the Commons format better too. One thing I like about this proposal is taking some of the workload off of arbcom. — Rlevse • Talk • 21:30, 5 October 2008 (UTC)

Putting the cart before the horse

Before trying to grant desysopping authority to new individuals or bodies, perhaps there needs to be a discussion about exactly what circumstances call for desysopping.

Recent Arb cases, the Everyking RfAs, and countless RfCs and AN/I discussions seem to suggest there is no clear community agreement on all the types of conduct that should justify (or require) desysopping. Existing processes (ArbCom, Jimbo, and emergency requests to Stewards) generally work effectively in cases of clearcut misconduct – self-unblocking, compromised admin accounts run amok, blocking opponents and protecting pages in a content dispute, etc. – but there is a great deal of uncertainty otherwise.

The (oh-so-nebulous) community needs to get its house in order. Figure out what types of conduct warrant desysopping. Make reference to specific cases wherever possible. Show where existing processes have failed. (Is 'I was too lazy to put together an ArbCom case' to be an acceptable reason for 'failure'?) Demonstrate broad input and broad support. Once you have clear standards, consider the best way(s) to enforce those standards. That's the stage where you might start asking for extra 'crat rights. TenOfAllTrades(talk) 16:08, 3 September 2008 (UTC)

It's good to have a dialogue on those issues. However.. this has all been tried before. No grand scheme has yet fixed this problem. This is why I made a tiny-little-suggestion instead of the great-big one you're making here. I have confidence that reasonable people (the crats) can figure out details as they go. I have far less confidence that we (the community at large) can make a large, comprehensive plan for dealing with tricky issues. Friday (talk) 16:56, 3 September 2008 (UTC)

I'm deeply uncomfortable about taking a 'throw-everything-at-the-crats-and-see-what-sticks' approach here. If the community can't clearly articulate the circumstances under which desysopping would be appropriate, it's not appropriate to spill the whole mess into the laps of the 'crats instead, and expect them to 'legislate from the bench'. Worse, we would be expecting them to make these decisions while under intense pressure, in the heat of a desysopping campaign.

I also wonder if what I suggest has been tried before. How about just a piece of it, then? I ask anyone here to show me specific cases where community consensus exists that an admin should be desysopped and has not been. I will offer bonus points if it can be shown existing processes (ArbCom, pretty much) have failed to desysop the admin in question. Give the readers of this talk page some details to sink their teeth into. What specifically are the proponents of this approach hoping to see happen? It's no wonder that previous proposals have failed if no one is prepared to explain which problems are being solved.

What you've proposed is only a 'tiny little suggestion' in terms of the number of words it contains. The implications are quite enormous. TenOfAllTrades(talk) 18:31, 3 September 2008 (UTC)

If the community were to entrust this system to the bureaucrats, I'd be (somewhat) comfortable with letting the bureaucrats come up with the exact criteria. Some of it is pretty obvious, though; abusive/punitive blocks, protecting articles while in an editing dispute, etc. (pretty much what Friday said, but I don't want to give him an ego by agreeing with him twice in one day) Perhaps we should look at some of the existing recall criteria to get a feel for what people self-identify as abuse (for example, I have a clear outline of what I consider eligible evidence on User:EVula/opining/admin recall). I don't see this ambiguity as a major stumbling block. EVula // talk // ☯ // 17:44, 3 September 2008 (UTC)

The thing is, the existing processes (mostly ArbCom) already seem to work for the clear-cut cases I identified. My impression is that people are hoping to have a process that can be used for edge cases—not necessarily abuse of the tools, but admins who are (for lack of a better term) jerks, who reflect badly on the project as a whole and probably shouldn't be in positions of responsibility. I could be mistaken, though. There may be (at least) two groups here. One seeks an 'express' desysopping process for clear-cut misconduct cases, the other seeks a way to deal with cases that aren't readily covered by existing policy and which the ArbCom is reluctant to handle. If that is the case, then it's no wonder no one can agree on anything; those two types of cases should probably be handled in very different ways. TenOfAllTrades(talk) 18:31, 3 September 2008 (UTC)

I was thinking about cases Arbcom is reluctant to handle, which often involve admins who are jerks. Emergencies are already taken care of, and clear-cut abuse of tools is often already dealt with (eventually) by arbcom. To your earlier points, I can think of historical cases of bad admins that I don't think were handled well by existing processes. I'm not sure it's useful to name them, because I hope to not argue over specific cases here. As to whether community consensus existed to desysop in those cases, that's a hard one. If we use the "look at percentages of votes" version of consensus, probably not. If we use the "weigh the strength of the arguments" version, then, in some cases, probably yes.

Is it best to sit around ahead of time and think about what constitutes desysop-able behavior? Maybe. Or maybe we could just have the crats jump in and start doing it and build the list of desysop-able behavior as we go. My vote would be a little of each. Friday (talk) 00:43, 4 September 2008 (UTC)

I like ya, Friday, but I'm going to need a bit more than that. I'm not prepared to write a blank cheque to the 'crats – or anyone else – to solve a problem that nobody is willing to describe with clear examples. I won't ask for an exhaustive list of 'desysoppable' conduct in advance — I'm all too familiar with how that type of policy is prone to insufferable wikilawyering. But I do need to see a clear indication of what problems you're hoping to fix. If we had some specific examples to work with, it might even be possible to avoid creating a new 'crat-monitored trial framework altogether. (Avoid bureaucracy creep!)

Currently, WP:ADMIN (the existing policy governing administrators) only states that "For severe or acute abuse of administrator privileges, an administrator's powers can be removed. Administrators may be removed either at the request of Jimmy Wales or by a ruling of the Arbitration Committee." ArbCom is generally reluctant to 'legislate from the bench', and has usually gotten an angry earfull (deserved or not) from the community when they do write their own rules. Given the wording of the existing policy, it is strongly implied that desysopping is only appropriate where there has been an abuse of the buttons. What you need is a modification of that policy – with clear community support – that opens the door for ArbCom to desysop under broader circumstances. (That may mean carefully structured RfCs preceding the first few Arbitrations, so that the Arbs have some idea of what the community thinks the ground rules ought to be.) TenOfAllTrades(talk) 02:36, 4 September 2008 (UTC)

Fundamental flaw

There is an unsurmountable flaw in the entire proposal. Admins have tenure for a very good reason: doing the job properly requires getting on people's bad side, because by definition any administrative act is one to act against someone's wishes: whether deleting, blocking, or protecting.

Just like academic tenure, adminship should be immune from disagreement outside of grievous misbehavior; and the Arbitration Committee is exactly the body made to examine misbehaving administrators. The fact that an admin is highly unpopular (whether because of his actions, or expressed opinions, or even militant behavior) should not be a reason to strip the bit away. Ever. — Coren ^(talk) 17:17, 3 September 2008 (UTC)

Admins are put in their position because they have community trust; similarly, they should be removed from their position when that trust no longer exists. Several people have argued this very same fact (that we can't have it be a popularity or unpopularity contest), myself included. EVula // talk // ☯ // 17:34, 3 September 2008 (UTC)

It might be worth considering some kind of tenure for established editors. Tom Harrison ^Talk 17:35, 3 September 2008 (UTC)

Er, how exactly do you propose that? Once someone has started editing, they're always an editor (unless they get indefinitely blocked, or retire). The concept of tenure isn't as applicable. EVula // talk // ☯ // 18:17, 3 September 2008 (UTC)

It's as applicable as we want to make it. It could be something like Contributors with more than X edits over Y months may not be blocked except by arbcom ruling unless it's an emergency, or they habitually begin their replies with "er" or "um". Make X and Y whatever is appropriate. Tom Harrison ^Talk 19:10, 3 September 2008 (UTC)

Erm... well, if you'd like to push that, by all means, but it should be totally separate from this proposal (and is likely to fail pretty badly). EVula // talk // ☯ // 19:18, 3 September 2008 (UTC)

If you're highly unpopular, you're almost certainly doing something wrong. Wikipedia is an inherently collaborative project. Working reasonably well with others is not optional for an admin. Friday (talk) 00:06, 4 September 2008 (UTC)

Your comment seems to reinforce Coren's point, and my own. Tom Harrison ^Talk 00:12, 4 September 2008 (UTC)

Sadly, Friday, I have to agree with Coren and Tom. Not to mention that one can be highly unpopular with one part of the highly multi-faceted group of people that make up wikipedia, and popular with others. Back to what has been said many times over the past many months, admins, by the nature of their job, must make decisions that are bound to be unpopular with some of those involved in the decision (AfD's, blocks, article lockings, etc.) They need to know that they can do so, for the benefit of the project, without fear of immediate and constant repercussions. It is during the RfA that the community needs to ask itself about the candidate's judgment and maturity, not popularity, and whether or not they trust that judgment to be exercised properly even if they disagree with it at times. -- Avi (talk) 00:39, 4 September 2008 (UTC)

Hmm. Maybe by "unpopular" I really meant "unpopular among reasonable people". Sure, vandals don't like admins reverting vandalism, and trolls don't like admins blocking trolls. Being unpopular among vandals and trolls isn't a bad thing at all. But, if you're unpopular among reasonable editors, you probably are doing something wrong. This doesn't mean you should lose the bit over it, but it does mean you should improve your person skills or maybe work in different areas. Friday (talk) 00:49, 4 September 2008 (UTC) PS Maybe "unpopular" is a bad word for me to use here. What I really mean is, if lots of reasonable editors find you difficult or unpleasant to work with, you're doing something wrong. Friday (talk) 00:53, 4 September 2008 (UTC)

I dish out lots of block, but don't seem to be unpopular, and I think I'd pass RFA easily if I went back there. For every troll or sock puppeteer I've blocked, there are ten thankful editors who are glad for my help. Who are these admins doing great work who are not respected by the community? Show them to me. I think they are an urban legend. Jehochman ^Talk 01:21, 4 September 2008 (UTC)

Coren has it correct. The fact is that every admin action is designed to stop someone from doing something they want, be it block protect or delete. An admin doing their job correctly will certainly gather a long list of annoyed people unless they stick to only the least controversial subjects. Well, these controversial subjects need admins too, fair use images need to be deleted from userspace, BLP needs to be enforced, disrupters need to be blocked. All this adds up to a pile of pissed off people, but if there is no abuse of tools then there should not be a desysoping. Popular opinion is not likely to make that distinction.

This is not needed, and will be a liability. Arbcom is doing fine, if they fail to deadmin somebody it is usually because it should not be done. This whole alternate method people are looking for seems to be because they don't like presenting evidence and would rather just have it happen. Chillum 14:38, 4 September 2008 (UTC)

This is a problem we already have. We very regularly see pissed-off people making complaints of admin abuse, when the admin in question was simply doing the right thing. We're already good at separating valid complaints from the crap. The crats in particular are expected to know how to do this.

This line of objection reminds me of a story I heard once. A little town had a stoplight, at their busiest intersection. It worked as well as stoplights tend to work. There was talk of adding another stoplight at their second-busiest intersection. One guy in the back stood up and said "I object! It's a bad idea. Sometimes people don't obey stoplights!" He's right- people don't always obey stoplights. But, this is an objection to the idea of stoplights- it carries little weight when used as an objection to one particular stoplight. Here, we've already chosen to have stoplights. They already work reasonably well. We're just talking about adding a second one. It's really a very small change from where we already are. Friday (talk) 14:48, 4 September 2008 (UTC)

Your analogy is incorrect, in my opinion. We have stoplights all over town, it's called an RfAr. You want to add a new procedure, not expand the existing one. It would be more like "Since we have stoplights, lets add manned police traffic stops" -- Avi (talk) 14:56, 4 September 2008 (UTC)

My stop light represented bureaucrat discretion. RFAr is our squad cars. They're already either diligently patrolling, or sitting in the donut shop, depending on your perspective.  :) Friday (talk) 15:00, 4 September 2008 (UTC)

I believe you may be under a misconception. There is no 'crat discretion, per se. Rather, 'crats are supposed to identify the community's consensus and implement it. Any discretion, if it exists, is solely as how to weight and interpret the arguments and statements of the various participants. Our only stoplights is the RfAr, all the patrol cars have been repossessed to pay for the mayor's latest limo ;) -- Avi (talk) 15:04, 4 September 2008 (UTC)

The discretion is filtering signal from noise. If a group of meat puppets or grudge bearers are trying to game the system to pass or fail an RFA, we rely on the bureaucrats to stop that nonsense. With a process for removing sysop access, we would expect the 'crats to help the same way. Jehochman ^Talk 15:33, 4 September 2008 (UTC)

(<-)And there we return to the primary concern with this; to have the removal of the sysop bit in any other hands but ArbCom (excluding Jimbo) will, in my and many others' opinions, cause significantly more harm then good. ArbCom needs to make teh final decision. And even should ArbCom authorize a subcommittee, tribunal, or other group to expedite the investigation and decisions on sysop misconduct, as I have been lobbying for for a while now, I still think that the actual request for bit removal needs to come from ArbCom, even if all they are doing is upholding the findings of the subcommittee. -- Avi (talk) 15:50, 4 September 2008 (UTC)

Does crat discretion already exist?

I think I'm beginning to understand the disagreement now. If you see crats as simple vote-counters, this is a radical change. If you believe we already count on their discretion, this is a minor little tweak. I can't imagine why the standards for crats would be so high, if we could replace them at any time with a bot. So, yes, I've assumed all along that we're already counting on crat discretion. I think crats are expected not to make original arguments themselves, but to weigh and evaluate the arguments presented by the participants in the discussion. I know we call this "consensus" for traditional reasons, but it's really a misuse of that word. It's more like what a judge does. I see the crats as already adding an additional sanity-check layer between the voters and the results, so that people can't so easily game the system.

For those who deny that crat discretion has been significant previously.. well, do you object to it becoming significant now? We picked these people for a reason, why not have them use their good judgement to help us out? Friday (talk) 15:29, 4 September 2008 (UTC)

My personal opinions as to the responsibilities of 'crats has been amply discussed in my two (failed) RfB's: 1, 2 (Perhaps this is why the community didn't want me just then :D ). In a nutshell, we want out 'crats to use their judgment to derive the consensus opinion in a close RfA. Not to use their discretion about a candidate, but to use their discretion in most accurately understanding and implementing what the community wants with regard to giving a candidate the sysop bit. At this point, the community is not empowered to cause the removal of the sysop bit. They may apply pressure to the sysop, ala the recent Majorly issue, or they may file an rfAr. But as per Wikipedia:Administrators#Removal of adminship, the community may only granteth, and may not taketh away (which is why I beg and plead that people think about the judgment and trustowrthiness of the candidate) and your suggestion is one that is basically asking for a complete overhaul of the system; one whose benefits (enhanced response to sysop abuse issues) is outweighed by its detriments (sysop paralysis for fear of bit removal misuse), in my opinion. -- Avi (talk) 15:57, 4 September 2008 (UTC)

Gotta agree with Avi; my three RfBs have given me the impression that bureaucrat discretion is only "in play" when it comes to the discussion and not the candidates themselves. I think that's where a lot of people are stumbling with this; they're thinking that, suddenly, bureaucrats are making judgement calls about administrators, whereas I'm seeing this as bureaucrats "presiding" (for lack of a better phrase) over a discussion, with the community making the judgement call and the 'crats merely executing (similar to RfXs) and being the point-of-contact person for stewards. EVula // talk // ☯ // 16:03, 4 September 2008 (UTC)

I agree with that- they weigh the arguments that were presented- they don't consider arguments that they believe should have been made but were not. It's like what a judge does, as I said. The judge does not introduce his own evidence. But, I don't see how this distinction makes this idea any less viable. Friday (talk) 16:07, 4 September 2008 (UTC)

Because you are asking that either the community, or the 'crats based on the community, be directly empowered to remove the sysop bit; that is a major change. -- Avi (talk) 16:15, 4 September 2008 (UTC)

(after some edit conflicts) I would hesitate to put it in such simple terms. It's a big leap from 'bureaucrats are mindless automata who do nothing but count votes' to 'the discretion of bureaucrats should be trusted in all things' — too big. Right now, we expect 'crats to exercise a limited amount of discretion under a specific set of fairly well-defined circumstances. Let's be honest here; a 'crat is free to employ his judgement in an RfA that sits between about 70% and 75% support, after votes from painfully obvious sockpuppets are discounted. Any 'crat who tries to use his 'discretion' to evaluate the quality of support or oppose votes on an RfA that's more than a point or two outside that window would get (and has gotten) piled on by the community. Whether or not that's the way all of us would like it to be, and whether or not that's the way the rules are written, are moot; that's the way that it is.

In other words, the fact that we've granted 'crats limited authority to exercise their discretion in one set of narrow circumstances does not mean that we would necessarily be pleased with the results if we leave matters to their discretion in others. At the time of their selection, 'crats were vetted by the community to perform certain very specific tasks, and I don't think it is fair to assume that they would be universally trusted in new responsibilities. As written, this proposal offers bureaucrats a dramatic increase in their authority. By being so completely open-ended, a bureaucrat will have the 'discretion' to decide the following:

• What fraction of 'votes' counts as a 'consensus'? Is it the same as for an RfA? More? Less?
• Is a quorum required? What is it?
• What are legitimate grounds for desysopping? Which policies must an admin adhere to religiously, and which ones have some flexibility?
• Which votes are for the 'wrong' reasons, and can be discounted? Does a vote that contains 2 'wrong' reasons and 1 'right' reason still count? Does it just count as a third of a vote?
• Does weight of numbers matter for anything at all?

The hapless 'crats would be lawmaker, judge, jury, and executioner. Saying that we'll leave all that to the bureaucrat's discretion is unfair to them and unfair to the community. TenOfAllTrades(talk) 16:25, 4 September 2008 (UTC)

Crats have already answered most of those questions- they apply equally to the RFA work they've already been doing. But I guess you're saying that their answers are only known to them, and have not typically been explicitly spelled out? It hadn't occurred to me that it was necessary for them to show their work, but sure, maybe it's a useful thing.

The third one is the new one, and it's a good question. That's what's interesting to me. The biggest determining factor I can think of is how an admin responds to corrective feedback from the community. Making mistakes is OK. Making the same mistakes over and over without modifying your behavior is what's bad. Friday (talk) 16:45, 4 September 2008 (UTC)

Making the same mistakes over and over tends to lead to an arbitration request, doesn't it? Most of the situations in which people clamor furiously for an admin's head and nothing comes of it are serious but isolated incidents, which arbcom generally regards as unactionable because they don't desysop over a single bad choice (most of the time). Wouldn't bureaucrats feel the same way (I would, if I were one), and wouldn't that lead to the same discontent among those who feel wrongerd? Chick Bowen 02:29, 6 September 2008 (UTC)

It's entirely possible that under this proposal, crats wouldn't do anything that arbcom isn't already doing. I hope that the crats prove more effective at dealing with a certain type of bad admins, but this all remains to be seen. What I do know is that, right or wrong, many editors have little faith in arbcom. Also, many editors think that since the community grants adminship, they ought to have some way of taking it away. In my view, these are plausible arguments. I don't know if this proposal will help or not, but I think it's worth a shot. Since nobody has presented any reasonable arguments why this proposal would be harmful, I don't see any reason not to give it a shot and see what happens. Friday (talk) 18:24, 8 September 2008 (UTC)

Support

I never understood why people had the perception that current admins would fail an RfA. Yeah, some admins do unpopular stuff. But do we really think the community is so short sighted that we'll get stacks of opposition along the lines of "Oppose - he deleted ma images!!"? Surely we can have faith that at least experienced members of our community will think to themselves "yeah, he deleted my images, but he did so for a reason and he uses the tools well" and will support?

Heh. Am I assuming too much good faith? I hope not.

If a bunch of trolls turn up in the reconfirmation RfA and say "Oppose - Dave's fat", which, even if Dave is fat, is not a valid oppose, surely we trust our 'crats to ignore them. Right?

On the other hand, if a significant part of the community thinks an admin is doing a bad job, hiding behind the moniker of "yeah, well, it's controversial!" shouldn't cut it. Do we not trust our 'crats to work out, based on an RfA, when someone does or does not have community trust? Heck, that's why we elect them, right?

I'm feeling a bit confused by the opposition arguments. That's why ^. Can someone make me less confused, or else tell me that I'm completely correct? I'd appreciate either. Giggy (talk) 12:08, 9 September 2008 (UTC)

This happens regularly at RfA. Why wouldn't it be the same at a reconfirmation? Synergy 12:13, 9 September 2008 (UTC)

Sorry, clarity is a big issue in these discussions. What happens regularly at RfA? People oppose because Dave is fat, or something else? Giggy (talk) 12:23, 9 September 2008 (UTC)

Something else and Dave is fat. Then the per above happens, and its all downhill from there. What I mean, is that the same "things" apply no matter what and are just as likely to come up. The only difference is that we know more about the candidates activity as an admin in a reconfirmation, and have no need to speculate. Do you think an admin goes about his or her normal activities expecting all of his edits would some day be called into question? Or that he or she would have to stand up for another RfA? My guess is no. Synergy 12:31, 9 September 2008 (UTC)

If this is true, this is a problem we already have, and it already needs to be fixed. This could only mean we're promoting incompetent crats- and I'm not convinced this is true. If it is true- run out there right now and find a few new people willing to stand for crat, and make it clear you expect them to use their judgement. The other thing we could do is try to structure RFAs like a proper discussion and not like a vote.. But I think that's been suggested or even tried before and people weren't confident that it was helpful. Friday (talk) 13:11, 9 September 2008 (UTC)

Wait. I wasn't insinuating that there was a problem with the crats. Yet its a clear indication that oppose !votes can cite anything and have it be their reason. How are we supposed to know how the crat closes? We can't, unless they tell us. We can only ask that they give a clear reason for closing a particular RfA. I wouldn't speculate which !vote a crat recognizes and which they do not (although there are often calls for crats to discount this !vote or that !vote). Whats fair to me, is not fair to you, and likewise. Any reformation that RfA undergoes will be epic in nature. We can't even agree on one set of particular standards. Until this happens, we will have the same chaos that normally happens (oppose per this, no i find that a good quality so I support, etc). Synergy 13:25, 9 September 2008 (UTC)

Part of the problem is that it is not arguments such as "Dave is fat" that are provided and then snowballed upon (per user:Example, Dave is fat), but rather controvertial views that we are reluctant to allow our Bureaucrats to ignore: to call out Friday simply because (s?)he is commenting here, his ageism views are seen by some as not worth counting, and by others as very much worth counting. (I state no opinion here; I simply call this up as an example.) What do we do then? How do the bureaucrats chose to ignore a comment based on community consensus as to whether that view is "worth counting," when the community itself does not even know? There are a fair few problems; I don't think bureaucrat incompetence is one of them, but rather a lack of firm, hard and fast views on a wide range of things that really need hard and fast opinion to go by. Anthøny ✉ 19:38, 9 September 2008 (UTC)

I couldn't agree more. Although we will have to save it for another time, and another place. This proposal is only for de-adminship after all. :) Synergy 19:42, 9 September 2008 (UTC)

I'm cautiously supportive — limiting this to bureaucrats means that the usual problems with frivolous or vexatious requests don't apply. Stifle (talk) 13:28, 9 September 2008 (UTC)

In reply to Giggy's original comment, I would be in the club of people who would comfortably fail an RFA if I attempted it today, and it would be due to a perception of being concerned more with the deletion of articles than with their creation. Stifle (talk) 13:39, 9 September 2008 (UTC)

For the record: Wikipedia:Requests for adminship/Stifle. Those objections came up then, and you passed. Have you done uber naughty deletionist things in your time as an admin? *shrugs* Giggy (talk) 13:46, 9 September 2008 (UTC)

Yes, I have :) More seriously, I think the RFA process takes content-creation far more seriously now than 2½ years ago. Stifle (talk) 14:25, 9 September 2008 (UTC)

• Yes, support this proposal, in theory, at least: I'd like to see an example model and specific arrangements for how a "deadminship" process would operate on enwiki before offering my support for this to start work. The proposal simply makes sense: we trust our bureaucrats to implement positive consensuses for RfA's and confirmation RfA's; why not vice versa? Anthøny ✉ 19:33, 9 September 2008 (UTC)
• Strongly support this - adminship should be no big deal, but would favour an increased level of detail that gives at least some guidelines to the bureaucrats. Brilliantine (talk) 04:02, 10 September 2008 (UTC)
• Support - this would be a vast improvement. PhilKnight (talk) 19:32, 12 September 2008 (UTC)

Centralized discussion

I've listed this on Template:Cent to get proper exposure to the proposal. Stifle (talk) 13:27, 9 September 2008 (UTC)

• Wise idea. Seems fair to me. Anthøny ✉ 19:39, 9 September 2008 (UTC)

Confused

I'm a little bit confused. The crats have said they will do this if a mandate is demonstrated, yes? I think we therefore already have an answer. If consensus exists for the crats to desysop an admin, you have the mandate. So get the consensus and then tell the crats. So far I don't think we have ever had one instance of an admin who the community consensually wanted de-adminned, so it is all pretty much pie in the sky. Hiding T 13:35, 9 September 2008 (UTC)

See Wikipedia:Bureaucrats'_noticeboard#What_happens_when_consensus_changes.3F - I went ahead and asked the crats to do this. The ones who replied said, "No, I will not do this job without an explicit mandate from the community". That is, they wanted it to be clear that the community is generally OK with crats handling this side of the job, before they will consider any specific request to do it in a particular case. Friday (talk) 14:02, 9 September 2008 (UTC)

I would say that this talk page has demonstrated that the community does not want 'crats to be handling desysopings. At least it is not an unanswered question anymore. Chillum 14:09, 9 September 2008 (UTC)

I too believe that it is safe to say that there is no consensus to have the desysopping procedure moved outside of ArbCom/Jimbo to either 'crats or the community at this time. Then again, it has only been one week since this proposal was made, so more time should be allowed for others to comment. -- Avi (talk) 15:51, 9 September 2008 (UTC)

I don't see the point in ending discussion because a half dozen editors have filibustered in the first week (in agreement with the second part of Avi's comment). I think there actually is fairly strong support for a variation of this idea, and it's been supported in the past (such as at WP:RFCRFA) but most of its proponents are exhausted advocating for it since it's typically shot down by a vocal minority, often with the simple technique of asking for consensus to be demonstrated over and over, at every step. I think many of the opponents see the proposition anchored to a few specific cases (in the case of admins often their own; otherwise a close ally's) and don't look at the big picture. Others seem to sincerely believe that there are admins who are wildly unpopular for doing the right thing. I agree that this is an urban legend. Admins who have made difficult, but correct, decisions enjoy incredible support from the community. As this is a baby step proposal we would be unwise not to give it a bit of a hearing. --JayHenry (talk) 01:02, 11 September 2008 (UTC)

• I've asked them before as well Friday. Wikipedia:Bureaucrats'_noticeboard/Archive_4#Query. The ones that responded said they would do it if we had a process. Hiding T 17:39, 9 September 2008 (UTC)

Renaming

The current subpage title "Proposal" is quite uninformative. Perhaps "Bureacurat proposal"? —Anonymous Dissident^Talk 06:49, 11 September 2008 (UTC)

Adminship and Tenure.

I've seen above, and elsewhere, some Admin propagating the idea that Admin have tenure on Wikipedia.

Admins do not have tenure.

Let's repeat this again. ADMINS DO NOT HAVE TENURE.

There has been a strange current in some people to assume that Admins have tenure, they do not and never did. Tenure is something that is only earned in the academic community by many years of hard work and general support from the community granting tenure. Adminiship is granted here much much easier than that.

I see no reason at all to grant Admins any kind of tenure, and I support any proposal that makes it easier to remove the admin bit from disruptive or abuse administrators.

Admin are community appointed caretakers, no more no less. Being an administrator on Wikipedia is not even something a wise person would list on their CV. Let's stop treating it like it's a divine-power. --Barberio (talk) 12:05, 11 September 2008 (UTC)

They effectively have had something very much like tenure. Arbcom has traditionally only removed the bit in the most extreme cases. I agree that we should not think of it as an appointment for life, but many people do see it exactly that way. In the well-known Archtransit case, we had an admin who was obviously incompetent and unfit for the tools. Anyone could have seen it. Yet, the bit didn't bet removed until sockpuppetry was discovered. To me, this case demonstrates the need for us to take admin misconduct a bit more seriously than we previously have. Friday (talk) 13:57, 11 September 2008 (UTC)

I think the community will be investigating committee members willingness to remove admin bits, and I doubt anyone who thinks admins have tenure will have a place in the committee for much longer. Admin misconduct will be the big topic in December's elections, as well as introducing the new changes to Arbitration Policy which should introduce a better system of getting redress towards committee members who don't act appropriately. --Barberio (talk) 14:10, 11 September 2008 (UTC)

Exactly what problem is this proposal supposed to fix?

This seems like yet another round of oh noes how do we get rid of abusive admins. Well, I'm not actually seeing a major problem of that that needs fixing right now. Its far too easy as an admin to get on the wrong side of people who harbour grudges. Genuinely abusive admins get hauled up in front or arb com and they are not shy of dysysopping them for their sins. So, please can someone tell me who this is aimed at specifically and if no-one can come up with a proper list of admins that need dusysopping then perhaps this is a solution in search of a problem. Spartaz ^Humbug! 06:36, 13 September 2008 (UTC)

Please point out one of these great admins who is suffering extreme unpopularity due to their good actions. This seems to be an urban legend that admins become unpopular by doing the right things. That has not been my experience. The community has the power to grant and remove sysophood. We don't want admins to feel that they have tenure. Jehochman ^Talk 06:53, 13 September 2008 (UTC)

Since I'm not the one advocating change the onus is not on me to prove the point. Spartaz ^Humbug! 13:02, 13 September 2008 (UTC)

Wikipedia:Requests_for_arbitration/C68-FM-SV/Proposed_decision#Contentiousness_and_demoralization, this proposal is an additional check and balance against us getting into this situation again. There is now a demonstrated tendency by some administrators to assume that they have tenure, and may act however they wish as 'Protectors of the Wiki'. We shouldn't have to go through a lengthy five month arbitration case every time there's a problem. Nor should there stomping of feet about claims to tenure, or people plain ignoring any problems raised.

I think it's grossly unwise to continue putting people in power without any clear way to remove them from that power if they prove to be a poor choice. --Barberio (talk) 13:47, 13 September 2008 (UTC)

I read the case a couple of weeks ago and its clear that the arbcom is completely deadlocked so I can't see how we can expect the community to be any clearer about it. Exactly who are you pointing this proposal at here? Unless you can clearly articulate exactly what problem you are going to fix here and cite clear examples of who/how it will affect I really don't see what this is going to achieve. Spartaz ^Humbug! 13:57, 13 September 2008 (UTC)

At the moment, an RfC is next to useless on an Administrator who subscribes to the 'Administrators have Tenure' view, or just plain ignore all criticism. This would give the RfC much more importance, as it could result in action.

The status quo requires going all the way to ArbCom, who won't accept the case unless an RfC has been taken; and which has a tendency to deadlock on any proposals to take action, instead opting for "Stop or we will say Stop again" remedies.

The new policy would put action, if any, being taken after the RfC. Making the ArbCom's position in handling administrator rights that of reversing removal on appeal rather than deciding if they should be removed, which I think would much better suit them. --Barberio (talk) 14:08, 13 September 2008 (UTC)

yes, I actually read through the thing and I understand how the proposal is designed to work. I'm not entirely stupid although I will confess to the occasional senior moment now and again. The point is that this is a solution to a theoretical problem. Please can you explain exactly who this proposal is aimed at? If there is a problem of abusive admins that need to be removed that process does not allow then surely there must be a list of the individuals somewhere with references to the RFCs overwhelmingly endorsing their defrocking. AS far as I'm aware this list does not exist so I really see this as a solution to a non-existent problem. I really would find it easier to understand this proposal if someone would articulate how/who this is going to affect? Your saying that there is no effective way of getting rid of bad admins. I'm asking you to tell me who these bad admins are and what they have done to need their bit removing. Spartaz ^Humbug! 15:19, 13 September 2008 (UTC)

The proposal is aimed at administrators who feel above the law. It is aimed at those who think they can befriend a few arbitrators and then do what they please without consequences. Jehochman ^Talk 15:22, 13 September 2008 (UTC)

Can you name names and point to the community discussion confirming this or is this just innuendo, smoke and mirrors? Spartaz ^Humbug! 15:26, 13 September 2008 (UTC)

The problem this might fix: Arbcom has sometimes failed to do something useful about obviously unfit admins. The reason I say it might fix is, we don't know yet. We haven't tried it. The other problem it might fix is one of perception: people see that this nebulous "community" picks admins, and has no way to get rid of them when mistakes are discovered. The question I'd ask to the naysayers is, is there any chance this proposal could hurt anything? So far nobody's given any credible reasons why it would. Friday (talk) 15:50, 13 September 2008 (UTC)

Who? What admins are you referring to. Without concrete proposals and naming names how can we possibly evaluate if this proposal meets its purpose? Right now you all tell me there is a problem with rogue admins who need to be sacked but this is the 5, 6, possibly 7th time (i'm losing count now) I have asked for a straight answer on who you aiming this proposal at? It seems you have targets in mind but are unwilling to admit who they are. Sorry, but if you want to persuade people to your point of view you need to be more open otherwise this is just process for processes sake and that's dangerous. I'm not letting this go. Please just tell me who you are aiming this at. Spartaz ^Humbug! 16:05, 13 September 2008 (UTC)

Well, every time people mention specifics, the conversation degenerates into a debate about the merits of those specific examples. This is the way to drag your feet, not the way to get things done. But, to name a few examples- Majorly demonstrated childish behavior unfit for an admin over a period of months or years. Nothing was done. Archtransit was obviously incompetent to anyone who looked, yet nothing was done until there was sockpuppetty. If you go back a few years, the Kelly/Tony show wasted lots of community time, and nothing was done for a long time. There are plenty of examples, but I don't think it's useful to discuss the merits of specific cases here. Friday (talk) 16:15, 13 September 2008 (UTC)

PS if you're holding out for RFC-like discussions where the community concluded by majority that these were unfit admins, don't hold your breath. Who would attempt such a thing, when there's no useful end result that can come from it? Also, this requires crat bravery and judgement- even our most obviously incompetent admins still often had crowds cheering them on, so if people want this to be about majority vote, it's hopeless. People whine that crats just count votes. Well, here's one way to make it clear that we expect them to do more than that. Friday (talk) 16:22, 13 September 2008 (UTC)

Well, that's your problem right there. RFA is a measurement of community support for potential adminship. You are saying that the community is unable to form a firm opinion on the withdrawal of that trust but nevertheless want to create a system to do this. Therefore, either the community will be unable to make its mind up or the system will only work by allowing a subsection of the community to veto something that the community as a whole agreed with. Sorry but I'm not buying that. I appreciate your responding with historical admins who were removed. I'm curious, are any current admins in need of defrocking and if so, whom? Spartaz ^Humbug! 16:38, 13 September 2008 (UTC)

I didn't see you create an RfC for me, or a RfAr. Why didn't you, if it was such a problem? It's all very well saying ArbCom don't do anything, but it takes people in the community to bring things to their attention. The only reason nothing was done is because no one did anything. Try using the procedures provided, instead of creating unnecessary new onesMajorly ^talk 16:25, 13 September 2008 (UTC)

This is why I think it's of questionable usefulness to discuss specific cases. But, you've pointed out another problem- Arbcom waits for people to bring things to them, usually. And, people are reluctant to do this because it's seen as an adversarial process. An RFA is a discussion about one editor. A request to pull the bit would be about one editor. An RFAr is seen as somebody versus somebody else. Arbcom is our big hammer here. And, they want a smoking gun. Saying "so-and-so acts like a child" isn't good enough for them, but it's good enough for any reasonable editor to conclude that adminship is inappropriate for the childish editor in question. This is like a prod, to Arbcom's AFD. It's lighter-weight. We don't always need a big hammer. Friday (talk) 16:31, 13 September 2008 (UTC)

If you look at my self-RfAr, it was going to be accepted before I resigned. If only someone had bothered to bring it to them earlier... and I would have been desysopped no doubt. This proposal is unneeded. Problematic admins will get removed in the end. Majorly ^talk 16:36, 13 September 2008 (UTC)

AFD removes bad articles in the end. Does this mean prod is not needed? I don't understand these objections, yet they seem common. I read them as "We have a way to do this! Stop trying to come up with a different way to do this!". Why? In real life there's almost always more than one way to get something done. Friday (talk) 16:41, 13 September 2008 (UTC)

We already have a confusing multitude of conflicting policies and guidelines. Why on earth should we come up with another one unless there is a clear current problem that the thing is supposed to address? How can we be sure that the proposal addresses its objectives if we can only discuss it in historical terms? It seems to me that unless you name names we are simply dealing with process for processes sake. Spartaz ^Humbug! 16:44, 13 September 2008 (UTC)

Wait, you want more examples? I named names. If you're willing to take the proposal at face value and object to what's actually written here, we can talk. If you're going to insist that this is about some secret plan and object to things that you yourself made up, how can I respond usefully to that?!? Friday (talk) 16:48, 13 September 2008 (UTC)

You mentioned my name, but failed to mention that no one even tried to use the current processes to deal with it. Until you can honestly prove that the current method is bad, I object to having any different. Majorly ^talk 16:53, 13 September 2008 (UTC)

I want current examples for the reasons I explained. Is that wrong? Spartaz ^Humbug! 16:57, 13 September 2008 (UTC)

This isn't a prod otherwise one objection will derail the whole thing. If its supposed to measure the loss of community trust then the whole community needs to be consulted. I'd say that AFD is more like it. What process will we create for the inevitable appeals? This proposal is clearly being written with certain current admins in mind but there seems some reluctance to name names and say why they should be defrocked. How can any of us know if we are being considered for the chop if this goes through? It may be difficult to name names but refusing to do so only makes you look like you don't want to be fully open about your intentions.

Examples from right now

I have no specific examples from right now. Other editors may or may not have current examples in mind. But, I hope this doesn't result in a response of "Well, then, this proposal has no useful purpose". There have been examples. It surely seems reasonable to assume there will sometimes be more. We don't need to wait until it's raining right now to decide it's useful to have some shingles and nails laying around. We can assume the roof will sometimes leak, for the foreseeable future. Friday (talk) 17:14, 13 September 2008 (UTC)

Have you tried using the current processes to deal with problematic admins? Majorly ^talk 17:21, 13 September 2008 (UTC)

Of course. I'm not a complete fool. I've asked the crats to step in and do the whole job instead of just half of it. They don't want to, without a specific mandate. I've participated (albeit minorly) in bad-admin arbcom cases when I've felt I had something useful to contribute. I'm a big fan of using all available avenues to solve problems. I see a usefulness in opening up a new avenue. I'm hoping to get others to see it also. Friday (talk) 17:26, 13 September 2008 (UTC)

Didn't Majorly resign his bit after an RFC? That suggests that existing processes are not completely broken. Frankly if there isn't anyone right that needs defrocking that the current systems can't deal with then I really do fail to see the point of the discussion. Consensus can change so what's the point wasting valuable editing time when this is a theoretical problem. Spartaz ^Humbug! 17:44, 13 September 2008 (UTC)

The current system is horrendously time consuming, and seems to have additional flaws, for example an admin can lose the community's trust by behaving poorly, however unless there is a clear cut abuse of the admin buttons, then ArbCom will probably refuse the case. I don't consider this system to be process creep - my hope is that it can reduce the amount of time spent on bureaucratic process. PhilKnight (talk) 18:17, 13 September 2008 (UTC)

Exactly how many admins do we have that need defrockingP? Friday doesn't think there are any so what is the point of this. Slow process is not an excuse to fling it out. Where is the real damage being done by the these rogue admins? Its process for process sake unless there is actually a substantial reason for this. Spartaz ^Humbug! 18:28, 13 September 2008 (UTC)

Spartaz, if you believe that at the moment community trust in admins is high, then I think you're out of touch. PhilKnight (talk) 19:17, 13 September 2008 (UTC)

Which admins don't you trust? Which of them have you expressed your concerns to, or tried to get a requests for comment on them? Majorly ^talk 19:19, 13 September 2008 (UTC)

I don't think its any worse then its been in the past. You need to cite specifics rather then casting vague aspersions around. Spartaz ^Humbug! 19:21, 13 September 2008 (UTC)

Don't agree - it's possible for the community to have less trust in admins generally. PhilKnight (talk) 20:01, 13 September 2008 (UTC)

can you prove that or is this an opinion? Spartaz ^Humbug! 20:05, 13 September 2008 (UTC)

I think a poll or rfc would help to clarify this. PhilKnight (talk) 20:10, 13 September 2008 (UTC)

My question or the proposal? Spartaz ^Humbug! 20:15, 13 September 2008 (UTC)

A poll or rfc could help to clarify to what extent the community trusts admins. PhilKnight (talk) 20:23, 13 September 2008 (UTC)

Don't you think that would be an unnecessary drama magnet? Spartaz ^Humbug! 20:25, 13 September 2008 (UTC)

If you think an RfC on the question "Does the community have trust in its admins?" would cause unnecessary drama, then I suggest that answers the question in it's self.

Of course, as I already pointed out, we have a majority supported arbitration finding of fact that the community *has* lost trust in admins because of these kind of issues. --Barberio (talk) 21:21, 13 September 2008 (UTC)

Would this be harmful?

There's a few people who've commented who think this is probably not useful. However, is there anyone who thinks it would be harmful? To my way of thinking, this can't hurt anything, and it might help, thus it might be worth trying. Friday (talk) 19:26, 13 September 2008 (UTC)

It'll hurt at least one person: the editor being desysopped. We aren't a lynch mob. RfAs are bad enough with people piling on with petty objections. I can just imagine it: a reconfirmation RfA in the form of an RfC. Brilliant idea... if the people who are allowed to swarm RfAs with their silly objections are allowed to comment/vote in it, no way should an RfC be binding and for a bureaucrat to close. Majorly ^talk 19:55, 13 September 2008 (UTC)

It would be a two stage process. The RfC could result in a mandatory reconfirmation, or else the administrator could choose to resign. RfC's tend to be open for quite a while, so there would be time for heat to dissipate. Then the reconfirmation could be done under calmer circumstances. Contrast this to the hellish process in Wikipedia:Requests for arbitration/Durova, or Wikipedia:Requests for arbitration/Matthew Hoffman. Those cases were handled much worse than what we'd get with this proposal. ArbCom isn't gentle either. By the way, I am very impressed with the way you handled yourself. Jehochman ^Talk 19:59, 13 September 2008 (UTC)

Still looks like a solution in search of a problem. Spartaz ^Humbug! 20:06, 13 September 2008 (UTC)

We already know you oppose the proposal. Repeating the same mantra does not advance the discussion. Jehochman ^Talk 20:27, 13 September 2008 (UTC)

I dunno, I still feel you need to think long and hard about what you are seeking to achieve. Spartaz ^Humbug! 09:15, 14 September 2008 (UTC)

I see no lasting harm to any editor desysoped by this process. It does not meaningfully restrict their ability to edit wikipedia as a normal user. And if they feel that they really still do warrant access to the special tool box, they can appeal to ArbCom or the Community by a new RfA.--Barberio (talk) 21:23, 13 September 2008 (UTC)

Yes, here's the harm: it would affect the way admins act. To some extent this would be good. Admins might raise their standards of civility etc. in order to avoid making enemies who would vote against them. (And I'm very much in favour of raising standards of civility.) But it might not be worth it, because admins might also be reluctant to carry out certain actions. There might be editors with lots of friends who watch RfCs and RfAs a lot, and they could get away with otherwise unacceptable behaviour because admins are reluctant to block them. Overall, there might be worse standards of civility on the project as a result, and a general atmosphere of power structures, coercion, threats and implied threats. I'm sure that admins would experience implied threats of desysopping; it could make for an uncomfortable environment. Many admins experience a lot of abuse already and are under stress. What are things actually like on projects that have such a system (e.g. meta)? ☺ Coppertwig (talk) 00:28, 14 September 2008 (UTC)

This doesn't hold up to analysis. Under the current system, your hypothetical scenario 'editors with lots of friends who watch RfCs and RfAs a lot' would almost certainly be Administrators themselves, having used those friends to get voted in. --Barberio (talk) 01:01, 14 September 2008 (UTC)

Aditionaly, the 'harm' you mention in this hypothetical scenario, is a 'popular vote' unjustly removing someone's administrator privileges following an RfC.

Again, this fails analysis for the following reasons,

• Assumes the bureaucrat will not notice any gaming of the system, well poisoning or other dirty tricks.
• Fails to recognise that the Administrator may re-apply by RfA.
• Fails to recognise that the Administrator may appeal to the Arbitration Committee.

I think that both the hypothetical 'harm' and the proposed method it might occur, have been over inflated. --Barberio (talk) 01:09, 14 September 2008 (UTC)

Ahem, no disrespect to our 'Crats but they all tend to follow the numbers very closely so if you are looking for someone to close the discussions who will exercise discretion and weigh arguments then the 'crats are not the body you want for that. Secondly, defrocked admins find it almost impossible to pass RFA a second time. Even good admins who have given up the tools volunterily have come unstuck when they tried again. Thirdly I honestly can't believe that having any kind of community consensus to remove rights should be vetoed by the arbcom. I disagree completely and until someone can actually give me a clear case of an abusive admin whose tool need removing that this process will deal with then this remains a solution searching for a problem and a waste of everyones time. Spartaz ^Humbug! 06:19, 14 September 2008 (UTC)

We did already, but you've ignored the examples given.

Added to that, your recent silly point making disruption edit.

I think I'm going to have to say that yours is minority view point poorly argued in a poor. Your contribution to this discussion has been noted, but I don't think anyone wants to keep listening to you say the same thing again and again. --Barberio (talk) 18:33, 14 September 2008 (UTC)

No I haven't ignored the examples provided. I asked about current abusive admins that need removing. You don't get to pick and choose which arguments you listen to on an wiki. Consensus requires that all points are considered. I keep asking the same question because no one answers it. Just a thought but you might like to be aware that you have slipped into your usual habit of disregarding comments that you don't like and being condesending to editors that espouse them. That's one of the reasons why your hobby horses fail to gain traction you know. The point about the 'crats is valid and needs addressing. There is clear evidence of them mostly being number counteres. If we rely on them to assess arguments how on earth can we expect them to act differently to their actions at RFA? You do realise that expressing a willingness to use their own judgement when closing AFDs is almost certain to fail a RFB nomination? What we select 'crats for is not what we are looking for here and that seems to be a major weakness in the proposal to my mind. Spartaz ^Humbug! 19:00, 14 September 2008 (UTC)

I disagree with your analysis, Barberio. The type of person I'm thinking of would likely not be an admin: they would be too uncivil and have too many enemies to become an admin. Remember that a strong supermajority is needed to become an admin, so just having lots of friends is not enough. If the desysopping also requires an equally strong supermajority for desysopping: that is, if desysopping will not occur if the results are 50/50 but only if about 75% or so vote for desysopping, then you might be largely right: most admins could ignore such factions as they'd be unlikely to be desysopped anyway. Anyway, if people such as I hypothesised were admins, what of it? Other admins might still be afraid to apply ordinary admin actions to the person (deleting articles written by the person; blocking the person for 3RR etc.). For that, whether they're an admin is irrelevant.

The popular vote would only be unjust in some peoples' opinion: or it might not even be considered unjust by anyone, and could still cause the harm I mentioned, i.e. admins being afraid to do things for fear of losing votes, admins receiving implied threats of desysopping, and admins being under stress.

I don't think bureaucrats would discount votes simply because they were from people who often vote in such discussions and who have friends who have interacted with the admin. I'm not talking about canvassing; I'm talking about people who regularly watch the RfAs and vote when they see an interesting name come up.

Re-applying by RfA may be difficult for the reasons stated. I don't think someone would be desysopped if, say, 30% of those in the discussion vote to desysop them. I assume desysopping would require a stronger consensus than that. (I don't know what level.)

I agree with Spartaz that Arbcom would be very unlikely to overturn the decision, if there was no evidence of canvassing or other obvious problem.

One more harm: we need admins. I think there are often (usually?) lots of backlogged admin tasks. Desysopping some could be harmful to the project. ☺ Coppertwig (talk) 19:06, 14 September 2008 (UTC)

I'll address each argument,

1. Hypothetical gangs gaming the system: The current proposal is to mimic the RfA process, but in reverse. Explicitly allowing 'crats to take action to remove administrators. But with the addition that the admin may appeal either by new RfA or directly to ArbCom. Hypothetical gangs of people could crop up to remove unpopular administrators, but likewise hypothetical gangs of people can crop up to put in place 'friends' without propper vetting. To abridge Churchill, "Democracy is the worst form of government except for all those others".
2. No one will get through a second RfA after being demoted: I think this instead reflects that there are a rump government of admin who were granted the status back in the 'Adminship is no big deal' era, and who would not have passed an RfA of today's standard. It's those administrators who would, rightly, not be able to get their admin bit restored. But borderline case administrators may well get through a second RfA, especially if they take some time out, and come back to the RfA acknowledging the points that were raised and saying how they would address them. And I do expect, and would urge, the ArbCom to overturn any removal of admin privileges due to abuse, improper process or some extraordinary mitigating factor.
3. We need more admin: No. We have plenty of people working on the project now, manpower is not an issue. However, we don't have productive manpower. I suggest that the problem is that a lot of people who have been elected to adminship aren't helping out clear the backlogs, or doing so in a way that ends up creating more work for other people. There's an awful lot of admins who seem to spend a lot of time arguing about the colour to paint the bike shed and intent on excessing their 'authority' in content disputes while letting the janitorial work slide. I suggest that may be because once you're in the Admin club, there's very little chance of being kicked out. I've used the baseball team analogy before, and this is becoming a bit overused... But a team is based on cohesive quality of all it's members, not having a large amount of members, or trying to retain 'super athletes' who don't play well with others. --Barberio (talk) 21:37, 14 September 2008 (UTC)

Let me get this straight: you think if we cut down the number of admins, then the remaining ones will work harder, and the backlogs of admin tasks will diminish? Remember, we're talking about volunteers here, not paid employees.

Another harm: Discussions like RfAs can be emotionally stressful for the candidate. This desysop proposal would not only increase the number of such discussions people would have to experience, but it would also cause them to occur at times not chosen by the subject: often inconvenient times. ☺ Coppertwig (talk) 02:10, 15 September 2008 (UTC)

Some of these objections amount to "It's harmful if it ever results in removal of someone's sysop access", which of course, misses the point completely. If you think sysop access should be the default for all editors, good luck, but this isn't the place to make that argument. Go over to Wikipedia talk:Requests for adminship and suggest the whole thing be marked historical. I was looking for objections of substance. Friday (talk) 14:30, 17 September 2008 (UTC)

"User is under 18 so I can't support" misses the point completely as well on an RfA, but you're still allowed to say that? That has less substance than any of these concerns here. Majorly ^talk 14:40, 17 September 2008 (UTC)

Possible harm?

1. False positives. The proponents of this proposal argue that the process of desysopping admins through ArbCom is too time-consuming, and that it results in too few desysoppings. The community is harmed if a lynch-mob driven high-speed process desysops a salvageable admin who made an innocent mistake or two in a high-profile area. Creating a process that is faster and more streamlined discourages sober second thoughts and time for reflection.

   This process would also be forced to make decisions based on incomplete evidence and hearsay. Members of ArbCom are freely granted access to Checkuser results. They are bound by strict privacy measures, and are both trusted and empowered to handle things like personal emails and IRC logs confidentially. The bulk of persons contributing to a RfDesysopping will not have direct access to this sort of information, or be able to evaluate it. Even the closing 'crats may not be privy to all evidence.

   The lone saving grace of a proposal that risks these sorts of rapid, flawed decisions is that it's unlikely to ever generate a useful consensus anyway (see below).

2. Drama. Stake out your position. Cast your votes. Canvass your friends. Look through a year of contribs to find the juciest dirt. Rehash old personal disputes. Nail the guy who voted against your RfA. Support desysopping, per concerns expressed by nom. After it's all over, everyone has all kinds of freshly bleeding wounds, no one wants to work together, and it will all have been a waste of time (next point).
3. Waste of time. How often will there be a clear consensus for desysopping of an admin who wouldn't have been defrocked by ArbCom anyway?

   Look at the previously-cited Archtransit case. He was desysopped (by a special action of the ArbCom) on 19 February 2008. It's been a while since I've looked at the Archtransit RfC, and it makes for interesting reading. Ultimately, RfC participants appeared to narrowly support a mentorship program with voluntary suspension of tool use, with 14 supporting and 11 opposed; the last comment on that proposal was dated 11 February 2008.

   In other words, barely eight days before Archtransit was caught socking and (quickly and efficiently) nailed by ArbCom, the community was unable to come to anything even close to a consensus to desysop. (Heck, we couldn't even come up with a simple majority opinion to desysop in lieu of mentorship.) Is it possible that some of those who supported mentorship might instead have supported immediate desysopping if an RfC (or this process here) had allowed it? Perhaps, but I'm not convinced.

4. No candidate cases. Nobody wants to explicitly describe how this process would have helped in past cases, and the most vociferous proponent on this page (Friday) has stated that he doesn't have any current specific examples. The past examples mentioned (Archtransit, Majorly, and Kelly/Tony) don't seem to demonstrate the utility of this process. I've already discussed Archtransit above. Majorly's most recent RfA was running 143/26 (84.6% support) when it was withdrawn four days in, which leaves me wondering if a process like this would have managed a consensus to desysop at the time of his resignation a month earlier. Perhaps a more detailed examination of events around Kelly/Tony (Kelly Martin and Tony Sidaway...?) would be illuminating, but I have doubts.

   The argument has been made elsewhere on this page that one shouldn't wait for rain before looking for shingles and nails, and I don't object to prudent preparation for foreseeable – albeit infrequent – future problems. However, we don't want to be building a ski lift in the Sahara, either.

5. Makes recruiting 'crats more difficult. I've said before on this page – and some proposal proponents disagree – that this represents a significant expansion of bureaucrat authority and discretion. The current 'crats weren't vetted for this type of call when appointed, and future 'crats will be grilled mercilessly on this topic. It may make an already difficult-to-pass RfB process nigh-on impossible.

So the harm is that we would be building an unnecessary new process that is likely to be unpleasant for its participants, divisive to the community, and ultimately ineffective at accomplishing its stated goals. I'm still hoping to see a clearly stated example – past or present, with at least a few supporting details – where this process would have been (or could be) applied successfully. Until then, these potential harms outweigh the very dubious potential benefits. TenOfAllTrades(talk) 16:18, 18 September 2008 (UTC)

A simplified process

This is really becoming far more complicated than it should be. If crats are able to determine consensus to add the bit, they obviously should be able to determine consesus to request its removal. What is needed is a reasonable process to get there. It should really be as simple as this:

• Every administrator is open to the recall process. Upon the request of 10 independent auto-confirmed editors, a Petition for Administrator Recall shall be opened by a Bureaucrat. The PfAR should work just like RfA in reverse. Alternately, any admin may direct a Bureaucrat to certify a petition against them regardless of the number of requests.
• Petitions to desysop should require the same level of consensus as the percentage used to grant adminship - 80% or more to recall, 70% to retain adminship, in between is up to the closing crat's discretion with the default being to retain adminship. If that percentage is met, and the administrator receives a greater number of recall votes than they recevied support for adminship votes in their last RfA, the admin shall be desysopped and permitted to re-seek adminship through RfA again at any time subject to standard RfA practices.

If the Admin passed 80-10-10, then they need 81 votes to recall and 80% of all votes in favor of recall. It really should be made that simple. This should be enough of a requirement to avoid frivolous petitions. It also protects those who supported the Admin in the first place by creating a higher barrier for removal than was used to grant. The process should not prevent a recalled admin from being renominated and going through the process again. If the actions rise to that level, then ArbCom should be involved. Comments welcome. Jim Miller ^{See me | Touch me} 21:54, 14 September 2008 (UTC)

I support this proposal in general, but not as outlined above - that bar seems too high to me. Why not make it a straight 50%? That would be much simpler: any admin who gets a majority of voting users supporting their recall, and when that majority is larger than the number who originally supported them for adminship, will be recalled. In the example given above, if an 80/10/10 admin was put up for recall and received votes of 100/50/0 (supporting/opposing recall respectively), they would not be recalled, which seems bizarre - how could anyone with so much opposition from the community rightly remain an admin, when they would obviously massively fail a reconfirmation RFA? I would argue that to pass a 'request for recall', an admin must earn the support of at least 50% of the voting community - in the proposal above they could 'pass' with as little as 25-30% support.Terraxos (talk) 18:06, 16 September 2008 (UTC)

I think it should be harder to remove an admin than it is to get promoted in the first place. There are too many controversal areas where simple numbers of opposes could be attracted by correct admin actions. It's probably my background in parliamentary procedure where it usually takes a majority to act, and two-thirds to reverse any action taken by the majority. Jim Miller ^{See me | Touch me} 16:33, 17 September 2008 (UTC)

Oppose, too open to frivolous and vexatious recall petitions. Stifle (talk) 14:25, 17 September 2008 (UTC)

While I admit that I see deep flaws in the entire proposed process, I should probably warn you that directly basing the recall percentage on a candidates percent support in RfA definitely isn't going to work. Tooting my own horn for a moment, my own RfA passed 58/2/0 in 2005. Even if I deleted the Main Page, unblocked Willy on Wheels, moved United States to Lower Elbonia, and set fire to Jimbo at the next Wikimania, I could probably still pull 3.4% support from sympathy votes, not-yet-banned trolls, and fans of performance art. TenOfAllTrades(talk) 16:31, 18 September 2008 (UTC)

Comment by FT2

I'm broadly in favor of devolving a number of matters from the Arbitration Committee to the community over time, where there's no good reason not to, subject to a few precautions. Eg, 1/ prevention of abuse/politicization, 2/ adequate appeal, 3/ protection of privacy-related issues, and 4/ what might be described as communal maturity sufficient to handle the matter well, without it becoming a point of excessive divisiveness or a "virtual" bloodbath, or discouraging editorial/administrative participation. I think over time the community can probably handle many "straightforward" desysoppings, as it has many bans.

The broad principle on desysopping is that once appointed, admins should be trusted unless seriously or repeatedly problematic. Nobody will make decisions that please everyone, and very unpopular decisions will not infrequently be needed. "Holding to ransom" shouldn't become a fear for administrators, which is why desysopping is currently not a communal decision.

I have one deep concern with this proposal (which some others have alluded to). Some users on the wiki do strongly politicize things, or "flock" in recognizable cliques/groups. As a result, some reasonable actions by admins, or some actions by admins that touched upon populist issues, have been instantly escalated by "mob appeal", into calls for instant desysopping.

I would trust our 'crats to handle those; my deep concern is I see the possibility that our crats themselves become targets, if they happen to disagree with the mob.

One sort of role that Arbcom has, is to protect the project in seriously disputed areas that the community cannot, or is unable to reach agreement, or where there are deep concerns. I have no problem with retaining the role of reviewing forcible desysop requests, if necessary. There's almost a quid pro quo needed here. The vocal minority and those who routinely hang round in cliques (however called), need to move to a less confrontational style, or mature in their view (if you like), and recognize that collaboration and respect with all -- not just their clique -- is usually needed. Not just "attack/undermine those who make decisions we don't like". It's possible to differ greatly and yet be respectful. The community needs to make clear that working with all others and avoiding excessive demands for desysopping for one unpopular or poorly judged block is needed. If that issue were reduced, then the community would probably be better able to handle it.

As it is I have grave concerns whether the community's crats should be put in a position where every few months they become targets for those seeking someone's head, and attacked for not providing the heads concerned. Arbcom volunteer for tough roles where abuse may arise. It's not desirable and it's wrong, but its "known". Crats volunteer to judge consensus, not to take abuse, and I would not wish to see them put in that position, without consensus by the 'crat community itself that it's okay.

This is similar to Avi's comment:

"It is distinctly possible that a sizable enough contingent of upset editors can cause repeated claims against sysops. Sysops need to be able to act in the best interests of wikipedia without a fear of their appropriate and proper actions being held hostage against them."

To which I would add, "... And sysops have very different views, so diversity will arise and needs to be handled respectfully and with good faith by all, when it does."

That said, I would be willing to use crats to judge consensus in many more matters, though. I think that's a useful direction.

Some quick thoughts. FT2 ^{(Talk | email)} 15:18, 17 September 2008 (UTC)

Thanks, FT2, for the comments. I am pretty good at understanding mobs, as that's one way to describe my professional work. The key to disbanding an angry mob is to give everyone a chance to have their say, and then to allow time for rational people to respond. (Restricting or discouraging comments is a good way to intensify mob anger.) Given time, heat will dissipate, and then good decisions can be made. The proposed process will require an RFC to determine if reconfirmation RFA is necessary. This process can be held open by 'crats as long as necessary to get broad community input. Any "single issue" cliques will eventually be overshadowed by the broader community. Once heat has turned to light, an RFA can be held if there are broad concerns and a factual basis. The process does not need to result in desysopping to provide a benefit. Dissipation of heat, and satisfaction of the mob are also benefits to the project. Bureacrats would have the ability to send a controversy directly to ArbCom if they felt undue pressure upon their persons. Were that to happen, I trust ArbCom would immediately accept such cases, and handle them per current practice. Jehochman ^Talk 16:11, 17 September 2008 (UTC)

But they don't, or aren't, at present, on other matters, so I'm not sure this theory is translating well into practise on similar matters. Allowing time helps a lot, but take a look at similar issues in the last 6 months. A lot of them were forced into squabbling, eventually not leading to any satisfactory resolution. A number ended up with people who tried to prevent ordinary poor conduct, smeared for it... others ended up with people who engaged in poor conduct feeling they had a license to continue it. Admins were demanded to be beheaded for cases that frankly, were highly questionable (one unpopular decision as an outside admin). I would want to see more level headedness and what I can only call "maturity" in how all parties act in such cases, no wild calling for heads, no justifying of bad conduct (by admin or non-admin)... fair balanced hearing. I don't yet have confidence we'd get that, and I do have as said, the profound concern that those trying to ensure fair hearings on such cases will themselves get attacked for their views and decisions. If the behaviors involved there, changed, then yes, probably. But for now... its a concern. FT2 ^{(Talk | email)} 16:54, 17 September 2008 (UTC)

As the parent of three small children, I have learned that one way to get mature behavior is to give responsibility and let the kids suffer the consequences of poor decision making (under supervision, with the prevention of serious harm). Protecting them too much does not work. We don't need a "nanny state" where we rely on ArbCom to keep us safe and secure. Community processes should be tried first, and only when they fail, should we resort to closed fora such as ArbCom. Jehochman ^Talk 17:43, 17 September 2008 (UTC)

On wiki, and for serious matters, its the other way round. Take bans for example. Initially only Jimbo could ban anyone. He delegated that power to a committee he formed, which would hear cases he asked them to. That was opened up to be cases the community asked, and now a lot of bans are community decided. But that took time, and it didn't start with community ban, but only went to that stage slowly and over time.

Desysopping is a serious measure, for very serious matters only. Right now there is a deep concern that the community doesn't have the maturity (in the sense of calm dispassionate measured consensus judgement) to handle desysop decisions. Developing that would be good, but that is a communal matter. Right now it would be likely used politically, for retaliation, to discourage or coerce, and at times to attack crats who don't endorse the biggest fist. That is where we are now and the stance all too often reflected in debates. Unfortunately in this issue, the people who suffer if it's "poor decision-making" (per your example) are not going to be the people who make those poor decisions, but the ones they wish to target at the time, the ones who fall out of favor with this clique or wish to impose site norms on that popular person, and hence ultimately all admins, all norms, and the project itself. What's more needed is the basics - ie, finding ways to develop gradual improvement in the community so we don't get such factional issues, or heated use and endorsement of excessive demands as we often see. Less high profile, less "glamorous"... but far more to the point.

Right now all to often, the only protection a neutral admin has against offending this or that clique, is the fact that 1/ their adminship is usually safe if they don't egregiously breach site norms. And 2/ if they do, and someone complains, then their conduct is as open to examination and criticism as anyone elses, but without heed to this or that "side"s view.

FT2 ^{(Talk | email)} 18:13, 17 September 2008 (UTC)

Cliques are clearly harmful; editors who behave in cliquish ways need to shape up or go away. Anything that helps expose cliques is thus a good thing in the long term. Friday (talk) 18:46, 17 September 2008 (UTC)

I don't buy the argument that ArbCom is smarter or more mature than the community. You (ArbCom) were elected by us (Community). Your power comes from us. We ask you to help occasionally when we cannot come to an agreement. If we decided that an admin should be desysopped, we have the right to bypass you. Even Jimbo has no power except that the community supports him, mostly. In the event of a serious disconnect, we can vote with our feet, and then Wikipedia is no more. Jehochman ^Talk 19:36, 17 September 2008 (UTC)

Err.... no. Wikipedia runs on a set of careful checks and balances, and Arbcom is voted by the community, but actually not a community appointment. It's formally appointed from outside the community, to handle matters that historically were always handled outside the community in the years before its creation and have never been "community matters" or "taken away" in the past from the community, but were always handled outside the normal communal process.

I would also express disagreement that the community always makes better decisions than those it votes annually as being some of its most trusted and insightful users. By its nature the community will contain almost entirely users who have not been voted to be the most trusted and insightful of its users, and also, on average its decisions may probably be more susceptible to "mob emotion" and the like. Blunt but blindingly obvious and needs saying.

Example of the moment:- Cato was appointed to checkusership via a community, supported for unban multiple times by a community -- and was seen through (repeatedly, and every time despite mass attempts at gaming) by an arbitration committee.

FT2 ^{(Talk | email)} 23:08, 17 September 2008 (UTC)

Response to FT2

I would like to respond to the comments of FT2, rather than specifically on the merits of the proposal. But let me state, first of all, that FT2 is an arbitrator whose work on Wikipedia I greatly respect.

It is indeed a common view, especially among administrators, that increasing the chances of losing administrator privileges is a "dangerous" direction to go down, increasing the likelihood of populist or cliquish influence. I would like to humbly suggest that this has become a virtual dogma that is worthy of at least questioning. First of all, is it really the case that unless administrators are "protected" from the mob, their time on Wikipedia will be all too traumatic? In my opinion, some kind of sensible mechanism that means abuse of administrator privileges will be more likely to lead to loss of those privileges will have little or no effect on most administrators. I believe it would be a healthy addition to the Wikipedia culture if administrators took it as a matter of course that abuse of privileges may well lead to a loss of those privileges. In short, I think the dogma that administrators need protection from the mob itself has the potentially dangerous consequence that administrators will be over-protected, and thus more likely to be cavalier with the administrator tools.

I myself do not believe this is very common, however, it does lead to my second point in relation to FT2: it is not only the case that administrators may be pursued or attacked by cliques; it is also the case that administrators may belong to those cliques. I do not say this is common, but I do say it does occur. And in my observation, those administrators who seem to me to be affiliated with cliques become viewed as leaders or figureheads of those cliques. And their clique membership leads to a reinforcement of the view that it is legitimate to abuse their admin privileges, because the way in which this abuse occurs is inevitably applauded by the other clique members. I believe this is the case for a current administrator who looks like they are about to have their privileges removed by ArbCom. Perhaps this is an extremely rare case, in which case perhaps no change is necessary. But perhaps as well, if there was a better mechanism for removing admin privileges, they could have been removed with less disruption in the first place.

As to what mechanism is appropriate, I have no firm view. I understand the position that bureaucrats should perhaps not be put in the middle of conflictual positions. On the other hand, judgment of consensus always inherently has that potential to be conflictual. Therefore I would suggest that the difference is less that one set of judgments is conflictual and the other not, nor even that one is necessarily more likely to be conflictual. I would argue that it is the current perception, based on the current Wikipedia culture, that removing admin privileges is such a "serious matter" that it will inevitably be highly conflictual. I tend to agree with the proposer that, just as it is possible to judge consensus in a proper way for things like article deletion, so too it is possible to objectively judge whether privileges have been abused, and hence whether the right to possess them has been forfeited. Peel away the rhetoric and hyperbole, and it is entirely possible to review a body of evidence and ask questions such as: did the admin block in a conflict in which they were involved? did they use their privileges to gain advantage? etc etc. It is my belief that were such a mechanism implemented, with good people in place to act as adjudicators (like I said, I am not saying who that ought to be; perhaps in the end it has to be those in a strong position of trust such as arbitrators), the community will discover that it is a way of decreasing rather than increasing melodrama.

In summary, I am not of the belief that Wikipedia will stand or fall based on the question of what mechanism ought to lead to loss of admin privileges. But I am of the belief that there are potential benefits to fostering a greater sense among administrators that abuse of privileges forfeits the right to possessing those privileges. And I am also of the belief that the arguments against doing so are grounded in an exaggerated fear of what the mob will do if it feels it can strip privileges away from those it doesn't like. Admins are not themselves above joining the mob, and it may be in precisely such circumstances that it is important that administrators feel that they are not "protected" from the community at large (after all, "cliques" sometimes see themselves in a populist way as representing the community, but they also sometimes see themselves as the defenders of what Wikipedia ought to be against the enemy forces of some other mob which they associate with the community). I believe that it is possible to have greater trust that some properly thought out mechanism for removing admin privileges will function successfully, that it will not lead to an escalation of melodrama and trauma, and that in the long run it will be beneficial for everybody if there is an understanding that the admin tools are a privilege, and hence that their granting is always conditional upon being properly handled. BCST2001 (talk) 22:19, 17 September 2008 (UTC)

Sadly yes. We have admins who belong to those cliques (on all sides of all debates), and if anything I would be fine seeing us come down much harder on these situations, if they use their adminship with different biases depending upon who they like/don't, which "sides" are involved, or "wiki-politics".

What I suppose I'm after and my point above is, this is a problem that strikes both ways. Some admins are in cliques and use wiki-social connections to further a stance. Other admins are attacked by those in cliques and their adminship is unfairly impugned (often viciously) for actually doing work and expressing opinions that are well within reason or non-egregiousness.

I started by saying I'm broadly in favor of devolution of Arbcom matters over time, where there's no good reason not to, and subject to a few precautions. But the schism on this in 2008 is too polarized. We need a general agreement by the "silent majority", the people who want to edit in peace and not have drama spirals, the users who don't flare up at others... there needs to be a re-asserting that the politicization needs to end, on all sides, and a collaborative collegial norm which we aspire to, needs to be pushed for by every user with actual actions and de-escalation (when a dispute arises) not just fancy words.

When we are less divided what an admin is and isn't to do, then we can probably handle this question better. Right now, neutral admins get harsh attacks from both sides, and Arbcom's hand on the desysop trigger is all that reassures people on all sides, that they will not be capriciously desysopped without genuine good cause.

FT2 ^{(Talk | email)} 23:24, 17 September 2008 (UTC)

Fair enough. Perhaps what I am saying amounts to nothing more, then, than that ArbCom should not be quite as concerned as it appears to be about the purported dangers of removing admin privileges, and thus that it should be slightly more willing to pull that trigger when it is in the interests of Wikipedia to do so. I certainly believe that this is not a decision that should be a free and easy outcome for groups of vocal editors to effect simply because they are conducting a vendetta, and hence my comment that it may need to be editors in positions of trust such as arbitrators who retain the sole right to make such decisions.

I completely agree that the overarching goal at the moment should be to find ways of lessening the melodrama and "politics" of Wikipedia. I guess I just tend to believe that a slightly stronger hand played against abusive admins might in fact be one way toward that goal. I certainly believe that most of the chatter about the purportedly terrible things happening on Wikipedia is just nonsense propagated by drama-mongers. Unlike many of the loud voices, it seems to me that ArbCom is in general pretty good at measuring what level of force is necessary to achieve the best outcomes; I just think that it is perhaps slightly held back by anxieties which in my view are not as well-founded as people tend to think. I have faith that ArbCom can make perfectly valid judgments about when the tools are being abused; I am just slightly concerned that unfounded anxieties about the effects of removing those tools may hold them back from constructive decisions to do so.

In short, I believe that establishing a clear atmosphere about what will not be tolerated from admins would be worthwhile (but I also tend to think, although I may be mistaken, that it is mainly administrators appointed in the early days of Wikipedia who are less willing to acknowledge the necessity of behaving cautiously and correctly with the tools). Which reminds me: I think there is currently a remedy being proposed in an ArbCom case that you haven't voted for (or against) yet...BCST2001 (talk) 23:38, 17 September 2008 (UTC)

Ahh... thats a bit different, and that I can mostly agree on. I would be happy to see Arbcom taking more action on substandard admin conduct when it happens, or supporting the expectations most users have of more visibly neutral and "fact-centered" (rather than "social-connections-based") handling by admins in disputes. I'd be fine with a more common use of temporary suspension or the like ("no big deal") as a means to make clear it is expected and sought, and change is needed, more than we do. Right now desysopping needs something horrible to have happened; I'd be fine seeing admins nudged towards better role model and to see the "no big deal" aspect used within reason as you suggest, as a part of helping encourage it.

Thoughts? (poorly worded, apologies, I'm between tasks here) FT2 ^{(Talk | email)} 01:53, 18 September 2008 (UTC)

Since I tend to think admins and arbitrators are too reluctant to embrace desysopping as a remedy to poor admin behaviour, it behooves me to ask myself where that reluctance stems from. The answer I tell myself is based on asking who it is that tend to become administrators: at the simplest level, those who become administrators are those who first of all must want to become administrators. This already sets this group apart from the general body of editors, most of whom have no interest in becoming administrators. So why do some people want to become administrators? My answer to this question is not at all intended as a criticism of administrators, even though it might sound like one: people who want to become admins must at some level like, want, or be interested in power or control. By this I do not mean that they are power-hungry or power-mad, but just that there has to be a level at which they enjoy the pleasures of possessing tools which are those of power and control. So that I am not misunderstood, let me immediately add that this does not mean I think the tools convey immense power to those bearing them: they don't. Yet the difference between editors with the tools and those without is essentially a matter of power and control. And if, as is sometimes said, loss of the admin tools is a "trauma" for the admin who loses them, then this is because at some level they want what those tools convey.

So where am I going? My feeling is that there is a very subtle level at which all admins have a common interest and to some extent act according to that common interest: the common interest in being admins. Again so that I am not misunderstood, this is not at all some kind of conspiracy theory, nor is it at all a criticism of the judgment of administrators. But it is my contention that in any administrative or bureaucratic culture, that is, any hierarchical culture, no matter how enlightened, it is almost impossible to avoid a situation in which those in the upper position want to hold onto the benefits of that position. So my feeling is that administrators are in general going to be more predisposed to accept arguments about the costs of increasing the chances of losing the tools. And, in my opinion, this means more predisposed to accept these arguments than they really need to be.

So what does that mean we need to do? I tend to believe that creating a new mechanism is likely to cause more problems than it solves. So my feeling is that the best hope is probably that arbitrators will come to realise that it would be a positive contribution to set the standard more clearly for what admin behaviour will not be tolerated. And as you say, I think a commitment among arbitrators to think of desysopping as "no big deal" would be a very good thing, even though it is of course entirely predictable that there will be loud voices each time this occurs. But from what I have seen, arbitrators also know that often the best thing is to just not respond at all to those voices, however loud they shout. Admittedly, seeing the solution as coming from the upper upper part of the hierarchy (that is, from arbitrators insofar as they are "above" admins) depends on a continuing faith that the right people will be appointed arbitrators, but it seems to me that faith is fairly justified.

So my recommendation, however insubstantial it may sound, is nothing more than that current arbitrators (and administrators) ask themselves the question about how solidly based their objections to desysopping are, and how much this reluctance might be based in common interests or common personality traits, at a very subtle level. And thus that they might shift their ground very slightly toward taking a stronger hand against those who misuse the privileges granted them. But I would like to conclude by saying that I am not actually thinking of any specific cases of admins whom I believe have forfeited that right to possess the tools, apart from one who is just about to lose them at the hands of the arbitrators anyhow. So I am definitely not prescribing any major changes, just a slight shift in outlook which will in the long term serve Wikipedia well. BCST2001 (talk) 03:40, 18 September 2008 (UTC)

It's a good point and well expressed. I'd make one correction - many admins are not power seekers except in the sense they wish to help others or wish to get involved in areas where this level of trust is needed (the wish to help can be altruistic or not). But with that point, your argument makes sense that admins and some non-admins may have a common unspoken undiscussed interest in keeping it slightly more difficult to be desysopped (except when calling for others to be!). That's putting it at its worst, and is not the universal case. As for Arbcom, I'd myself willingly use our standing to enforce higher standards or to use suspension of sysop standards as a stick/carrot to better encourage "is a role model" and not just "uses tools properly". If the community made clear it wanted us to help that way, then especially.

That would be a change in how we used that right. Bear in mind it would not be without some difficulty and cost, as a number of people would complain loudly, since unfortunately some established admins/non-admins who have strongly formed cliques, would naturally object to this curtailment of admins' current scope to act up. Also it's important to keep a sense of perspective -- most admins probably do use their access peacefully and well; abuse and desysopping is not a majority issue (though aiming for higher standards is always going to be a good thing). As said, we also really need a parallel sea-change in the community that some things are not acceptable or will be held to a higher standard (including desysop calls) otherwise it risks being a recipe for divisions.

A possible solution might be to see if a consensus can be formed to agree some areas of admin action that clearly are not okay, and also, some grounds for desysop that are frivolous or unhelpful, and try to cut down the extent to which there is a difference in these areas. That would help it a lot.

Thoughts? FT2 ^{(Talk | email)} 10:34, 18 September 2008 (UTC)

In response to the problem that a strongly formed clique of admin may act up and cause trouble in response to this...

If they wish to take the length of rope, build the scaffold, and offer their own heads into a noose, so be it. --Barberio (talk) 13:35, 18 September 2008 (UTC)

Response to FT2: I certainly did not mean to imply that admins are "power seekers" in any way that rules out positive motivations. I think the argument is the same for politicians: some are motivated by niceness and some by rottenness, but at some level they must all have an interest in power. Which is not in itself a bad thing. I agree with you that there are difficulties in delineating exactly what conduct warrants desysopping, and that if the community could make this a little clearer to itself it might help. For example, one might say that clearly admins should not contentiously block people when they are involved in a clear dispute with those people. On the other hand, there might be some occasions where this isn't really such a big deal, since the block was going to be done by somebody anyway (although even so, there isn't ever really an excuse not to get an uninvolved admin to do the blocking). But I would suggest that, even in the absence of the community clarifying what is and what is not clearly acceptable, it is still possible for the arbitrators to take a slightly firmer hand: many of their decisions rely on good judgment anyhow, that is, on interpretations of policy and understandings of what makes good sense and what doesn't. I think arbitrators would be capable of the judgment required if they decided that it was a remedy they wanted to use more commonly than they currently do, even though the situations they will be analysing will inevitably be complex. All that said, however, I agree that most admins act without problem, and that the scale of admin abuse is probably quite small. The purpose of thinking about such things is partly in order to maximise the chances of keeping it that way. BCST2001 (talk) 13:44, 18 September 2008 (UTC)

Strong Support-- and I'll tell you why!

So, this proposal, if enacted, would do a number of things that are good-- and some of them aren't obvious at first, so I'll just rattle them off.

First and foremost, it would allow us to desysop people who abuse their tools, without having to go through the time (and drama, and delay) of Arbcom. Removing problematic admins is a good thing, if for no other reason than it stops the abuse. But actually, of all the positive benefits of this proposal, actual admin removal is just scratching the surface.

Secondly, this proposal would remove some of the burden on Arbcom. Arbcom is a major bottleneck, burnout rates are high, case delays of more than four months. We need to do as much as possible to farm out tasks to the community. The community will not be able to reach a consensus on many allegations of problematic admins-- but in some cases, the community will have a conclusive opinion, and it's unnecessary to have arbcom weigh-in if the community has made up its mind. Imagine if every RFA had to go through arbcom! This is almost the same thing.

Thirdly, this proposal will drastically reduce the incidence of admin abuse, over and above any reduction caused by actual desysopping. It's just a basic fact of human nature-- the less oversight you have, the more people are going to play fast and loose with the rules. Right now, it's so difficult for an admin to get desysopped that they have substantial latitude to behave badly. Bending the rules might be very tempting for an admin who knows the only possible consequence could come from an arbcom case, usually months in the future. On the other hand, if the tempted admin knows consequences are likely to be swift, it will be all the easier for said admin to not even give in to temptation in the first place.

Fourth, this proposal will drastically increase the number of admins. This may seem shocking-- how can a proposal designed to desysop increase the total number of admins?!? Because, as things currently stand, the RFA process has to turn away many of its candidates for fear that the candidates ultimately won't act as good admins. Great caution has to be exercised when bestowing the bit, because the people at RFA realize that once they give someone the bit, it's damn near impossible to take it away again. Under this proposal, however, an admin could be returned to editor status if they don't appear to be working out. The people at RFA could then afford to be a little more relaxed, because they know that if there's trouble down the line, the project would have a simple way to hit the undo button.

I'm actually vaguely surprised a proposal such as this one hasn't already been enacted. It's bound to, sooner or later. The number of admins is growing all the time, while the number of cases arbcom can actually handle is remaining constant. Sooner or later, the sheer force of numbers is going to require a proposal such as this one to be adopted. --Alecmconroy (talk) 15:00, 27 September 2008 (UTC)

Postscript. Fifth, this proposal would be a less-traumatic path to removing the bit. There are going to be people who are great editors, great human beings, great wikipedians, who pass RFA with flying colors, but in the end, they just don't have the knack for being an admin. Currently, the only way to fix things is to have a huge Arbcom trial, where an indictment is drawn up, piles of evidence are collected, arbs render a judgement, and a sentence is meted out. The result feels like beingdrummed out in disgrace. But under this proposal, if consensus existed, we could solve the problem as quickly and as painlessly as possible. Yes, some demoralization is inevitable, but losing your bit this way would feel more like a failed RFA than an full-blown Arbcom-court-martial. --Alecmconroy (talk) 05:26, 5 October 2008 (UTC) (be aware that many endorsements were added before this postscript was written.)

• Exactly right. Jehochman ^Talk 15:28, 27 September 2008 (UTC)
• I 'endorse and support this statement. --Barberio (talk) 15:30, 27 September 2008 (UTC)
• Yes, I agree. Tombomp (talk/contribs) 15:30, 27 September 2008 (UTC)
• Strong support. 100,000 welcomes to this crystal clear view. I had this discussion page on my watchlist for weeks as I hoped for insight like this. Totally endorse. — Athaenara ✉ 22:14, 27 September 2008 (UTC)
• +1 Giggy (talk) 01:01, 28 September 2008 (UTC)
• Endorse. Synergy 06:55, 28 September 2008 (UTC)
• Support Hear, hear. Makes great sense. Colonel Warden (talk) 11:47, 28 September 2008 (UTC)
• Complete agreement. What the editoriat giveth, the editoriat ought to be able to take away. the skomorokh 19:57, 3 October 2008 (UTC)
• This certainly puts things very clearly. -- how do you turn this on 14:14, 4 October 2008 (UTC)
• Long, long, overdue. I'm Mailer Diablo and I approve this message! - 18:18, 4 October 2008 (UTC)
• Comment / Question "The number of admins is growing all the time"? Yes, but the number of active admins is not. I'm concerned that this assertion implies that there are more and more administrators actually administrating (and doing wrong so going to Arbcom and so forth), when this is demonstrably false. I'm therefore interested as to what effect that has on that element of the rationale. Pedro :  Chat  20:07, 4 October 2008 (UTC)

Pedro-- great point. As we add admins, yes, we will lose some to inactivity. Good eye-- that totally flew past me.

My hope is that we'll continue to gain people faster than we lose them. But I could be wrong. It's possible number of active admins could level off at some point in the future.

Even so, I feel that there's still room for improvement. In arbcom, we're asking a tiny group of Wikipedia's Finest to shoulder too great a burden-- the result is four month case delays and massive burnout. That's a problem, right here and now, waiting to be fixed. Friday's proposal here may not be the whole solution, but I do feel like it's a step in the right direction. --Alecmconroy (talk)

• Strong support per Alec. Everyme 21:18, 4 October 2008 (UTC)
• Support The near-impossiblity of taking away admin tools unquestionably adds to the drama at RfA. Granting the admin bit can't be not-a-big-deal unless taking it away is also not-a-big-deal. Jclemens (talk) 02:45, 5 October 2008 (UTC)
• Oppose Good faith or not, good proposal or not, this is the surest method available to empty out Category:Administrators willing to work on contentious articles edited by true believers. Even if there is almost no chance that an administrator doing their job on (say) Kosovo, Liancourt Rocks or Barack Obama will lose the mop, we now have an enormous counterincentive for this already thankless and difficult job. Protonk (talk) 05:26, 5 October 2008 (UTC)

• Followup from talk page. I'm less worried about the process being hijacked as much as I am worried that a large motivator for working in WP:LAME territory is the knowledge that you won't lose the bit unless you actually misuse the tools. I'm not sure as to how the proposal could be changed to prevent this without limiting it in such a way that it isn't a substantive change. We can't just immunize admins in difficult areas from criticism because some of the judgments they make there are wrong. But we also have to accept that the "tenure" part of adminship helps incent them to work there. So I don't know. I'm open to solutions, but I don't have any. Protonk (talk) 06:10, 5 October 2008 (UTC)

• Oppose - Replacing the elected arbs with the collection of interested parties who would inevitably drive these new processes is a change for the worse. It's not just or useful. It's unjust because we shouldn't be going after people, but trying to resolve disputes. It's not useful because it will have a chilling effect on blp and copyright enforcement, and will enforce homogeneity of viewpoints when we need just the opposite. Tom Harrison ^Talk 18:43, 5 October 2008 (UTC)

• Like it or not, resolving disputes frequently and correctly involves "going after people". Also, I don't see the connection you're making with regards to blp and copyright enforcement. Everyme 19:05, 5 October 2008 (UTC)

• Strong support - agree this is long overdue. Jennavecia ^(Talk) 14:31, 6 October 2008 (UTC)
• Oppose As stated multiple times elsewhere :) -- Avi (talk) 14:33, 6 October 2008 (UTC)

Re-re-re-announcement

I was just over at WP:AN and noticed that this proposal has been re-announced there: diff.

Links to this proposal – or to ones like it – appear with some regularity at the Village Pump or other noticeboards. Usually there's a brief flurry of activity, then the proposal fades into the background again. What was the purpose of the re-announcement this time? Is this a 'final form' to be voted on, a near-final draft that might be tweaked slightly, or just another invitation for everyone to restate their positions? TenOfAllTrades(talk) 21:54, 4 October 2008 (UTC)

Do you agree with Alec's points above? Everyme 22:14, 4 October 2008 (UTC)

Are you answering my question, or are you just curious? TenOfAllTrades(talk) 22:28, 4 October 2008 (UTC)

Sorry for not making myself clearer: I think you wouldn't ask if you agreed with Alec's points. So I wonder inhowfar you don't agree with his points. Everyme 22:37, 4 October 2008 (UTC)

Rather than repeating myself ad nauseam, I'll just encourage you to review my existing remarks on this talk page. And it still doesn't explain to me the purpose of Jehochman's announcement. TenOfAllTrades(talk) 22:50, 4 October 2008 (UTC)

Well, I for one happen to think that Alec's post alone is worth a notification. Also, AN is the only of the potentially relevant noticeboards I have watchlisted and I wouldn't have found this proposal at all if not for Jonathan's announcement. Everyme 23:09, 4 October 2008 (UTC)

The purpose of my announcement was to get more comments. This proposal is not something that should be decided by a handful of editors. Jehochman ^Talk 15:30, 5 October 2008 (UTC)

A very brief review: the crux as I see it

I saw this at AN, so looked this over and will identify what I see as the dichotomy in the intent and the wording; 'Crats are to decide if the case is proven to desysop (as they are to confer adminship), but the process of arguing the case for desysop is not apparent.

The 'Crats role is to weigh the arguments given and pronounce a verdict, and it appears that they are prepared to make decisions on desysopping an admin when provided with the mechanism to provide a consensus - but that no proper thought has been given to the mechanism of that process. Is it to be a reverse RfA, or the present RfC, or an ArbCom style presentation of evidence, findings of fact, and Remedies which the 'Crat (historically an individual, but are we happy to stay with that?) will consider and decide upon? Which is it to be? Or is there a process that I missed while skimming through?

I will not be !voting on this matter, as I am too obviously closely identified with the "problem", but I would share my ruminations on the processes that are currently proposed; the reverse RfA and the now available RfC are far too open for short term considerations and prejudgements, forged in the heat of the matter which brought the case to a head, and will be largely populated by those who have already formed an opinion. There does not seem to be a facility for a longer term review of the actions and judgement of the relevant admin, outside of that which is drawn by and supports the viewpoint of the participant(s). The drawback of the ArbCom style of review, consideration, proposal and determination is so obvious that I am addressing it separately immediately below...

...Why have a second/alternative ArbCom to decide upon desysopping? This is already within the remit of the current ArbCom - it simply has not often been approached in this way; the unsuitability generally of an admin to retain those tools. Any new "desysop ArbCom" will need to be drawn from the community, as is the present one, and demonstrate the ability to not only weigh the evidence presented but to conduct its own investigation and to draw conclusions from both the present data but also from historical records - ditto the present mob. I see only a duplication of an aspect of the current ArbCom, and nothing that a brief but radical rethink of the process of requesting (and being granted) an ArbCom upon the viability of removing the bit from a current admin.

My alternative, in keeping with the processes already available

There should be an expansion of Requests for arbitration with regard to removing the bit from abusive/bad sysops, not relating to particular disputes (in the manner by which Felonious Monk lost their flag recently) but how Can't sleep, clown will eat me (talk · contribs · blocks · protections · deletions · page moves · rights · RfA) was eventually desysopped. Editors should be encouraged to file RfAR's against individual admins under those terms rather than the dispute that they had some dealings in, and thus the entirety of the sysop actions will be the focus rather than needing a specific case under which to expand such a review. For example, a case against me would be requested as "WP:RfAR/Desysop of LessHeard vanU" rather than trying to find one matter with an example of my disputed action/behaviour, from which a general review might happen if the case is accepted. Editors should be encouraged to make such requests, and ArbCom should be encouraged to accept such cases (since if the complaint is found to be unpersuasive it can be quickly discounted as such) with greater facility than currently. If this proves to increase the Arb workload to a point where the current membership cannot cope then the obvious answer is a recruitment drive - which will likely include those who would have formed the quasi-ArbCom previously proposed.

Anyhow, this is my off the cuff conclusions upon reviewing this and the project page (any long term consideration would likely not change the basis, but there would be a major increase of verbosity - yup, I'm afraid that that is possible!!) Consider and/or ignore it as you will. LessHeard vanU (talk) 23:17, 4 October 2008 (UTC)

Reponses

Less-- I definitely wouldn;t support what you see in the current policy either. A "new 'desysop ArbCom' " or "shadow arbcom" is definitely not what I would want to see.

I would like to see something very analogous to RFA. After all, letting community consensus decide whether or not someone should person have the bit is nothing new-- we do it all the time. In essence, we've already decided that the community, buttressed by the 'Crats, is up to making this kind of a decision.

But-- a final safety-value measure, of course the closings could always be appealed to Arbcom. The point isn't to superseding Arbcom, do a run-around Arbcom, or overthrow Arbcom. They're the best we have, and they're overworked as is-- we should use them as sparingly as possible.

Lots of times, the community can't reach consensus, and we have to turn to them. That's what they're there for.

But sometimes the community will reach consensus. After talking it over, it will be clear that, for example, the community speaks loud and clear than an admin is doing a good job and should stay in his current post. Having reached a true community consensus-- should we really ask that arbcom take the case on its own, rehash the evidence (and the associated time, effort, and mudsling)???

Maybe sometimes Arbcom will have to review things-- but sometimes they won't. Unfortunately, due to the lack of a RFdA process-- every single solitary desysop proposal 'must go through the narrow, high-pressure bottleneck that is ArbCom. We can do better. --Alecmconroy (talk)

While I trust the community to confer adminship, I do not trust the community (or that part which would participate) to remove it - either a failing in me or in good faith generally, but those who would be involved in the latter have likely far more invested than in the former. I prefer uninvolved individuals to review, and that is the role of ArbCom. I would support the easing of the bottleneck of the current ArbCom, by the expedient of adopting new members to the existing/retiring group (and would suggest that up to a third of active members be allowed to withdraw by reason of previous commitments to any accepted request). I prefer the calm(ish) considered response by an otherwise uninvolved group than the urgent(ish) clammerings of those previously involved.

I would suggest that, should a non-ArbCom body be formed to review and decide upon such matters, that a large majority would be referred to ArbCom by the "losing" faction in any case; so why not cut out the middle man> LessHeard vanU (talk) 00:44, 5 October 2008 (UTC)

I hear you, and I think you have a totally valid opinion (particularly about the possible Arbcom Policy changes, where the ball is getting rolling on those too).

I sorta feel like-- if the community can have a true consensus on _anything_, and arbcom doesn't want to overrule it, and the board doesn't want to overrule it, and the state of florida doesn't want to overrule it-- then that's what should happen. That de-admining can't be initiated by a community consensus is unpleasant. But I do realize there's the potential, like any community-driven part of the project, that the community just plain will get it wrong. It's a risk with everything we do here. I think the project will be better off if we take that risk, just as we entrust the community with almost everything else around here. But I do get you if you think the risk of a bad-desysopping due to community consensus is too great to risk. --Alecmconroy (talk) 01:07, 5 October 2008 (UTC)

The community can take it to ArbCom, and have had that ability for some time. They historically haven't, or presented it in terms of specific articles/areas of dispute, but there is no impediment to start to do so. Once accepted (a hurdle I do think needs working on) then the community can provide the rationale and evidence in support of the request - same as now. Simply, I think we need test the existing structure to more easily desysop admins than create a new one - with the inherent birthing pains that seem likely. LessHeard vanU (talk) 01:34, 5 October 2008 (UTC)

This will not work - let me tell you why

On first glance this, or some version of it, allowing the community to desysop, seems to make sense to many of us. That's because we largely agree on two things 1) That there are some (a small minority) of people who should be desysopped who aren't. 2) That it is too hard to desysop.

However, at that point the consensus ends. You see, although there's a consensus that "some" people need desysopping, I suspect there is no consensus as to who those "some" are, and on what grounds (enough of us have different people in mind). So we actually have a false consensus. The idea that giving a crat the power to order a desysop where there is a "consensus for removing administrator access" - won't work, because there only occasions where there will be such a consensus is going to be in the type of open-and-shut abuse case which would easily have resulted in an arbcom or emergency steward desysopping anyway.

So, my prediction is that if you adopt this, you'll get lots of drama as people ask for the desysopping of JzG, SlimVirgin, CydeWise or whoever is politically unpopular this week, but you'll get no fewer desysoppings than under the present system. In short, lots of drama and few, if any, results. It will simply become a weapon in the various factional wikiwars.

I know. The immediate reaction is "but I know people who need desysopped". Well, yes, I do too. But ask again: would you ever get a consensus of wikipedians to agree with you? --Scott MacDonald (talk) 23:59, 4 October 2008 (UTC)

Totally concur with the thinking. To the other people who support this proposal-- if you are supporting this with an eye to immediately requesting the desysop of some of the big names Scott suggests, you might as well forget it. I can tell you right now, consensus doesn't exist to desysop them, so if that's what anyone's driving at with this, they should forget it.

If you think those people shouldn't be admins, you're going to have to take it to arbcom. This proposal would, I guarantee yield no-consensus on all the above.

I support it with an eye towards the future. Looking back over history, considering admins where there was a real consensus to desysop, ultimately each either resigned, was recalled, or was desysopped by Arbcom.

It's not that the current system doesn't produce the right results-- whenever consensus exists, the system HAS produced the right results. It just sometimes takes forever and the desysoppings take place weeks and months after the abuse occurred-- and during those months, there's mud slinging in every direction. Really a simple RFA-like process could have quickly and painlessly determined that, yes, the admin has lost the confidence of the community, or no-- the community still has faith in the admin.

This is a function for admins who have completely lost the confidence of the community.-- NOT for admins over whom the community is bitterly divided. That's going to have to stay ArbCom's job. --Alecmconroy (talk) 00:39, 5 October 2008 (UTC)

It may be true (although I'm less sure) that a process like this would have caused some desysoppings more quickly. However, the benefits of doing that, weighed up against the heat generated by the doubtlessly manifold unsuccessful attempts to use this process means that this is detrimental. Can anyone not imaging that, if this process existed, it would not have been initiated countless times against unpopular people, only to result in acrimonious "no consensus" results. If the only people who will be desysopped by this would have been desysopped anyway, then it really is pointless to create a process that's going to be used for political point scoring and inter-admin feuding.--Scott MacDonald (talk) 01:18, 5 October 2008 (UTC)

Feuding? Right under the 'crats eyes? I don't think so. As Alec correctly pointed out, the process by its very nature would lead to greater acceptance of individual responsibility and more self-regulation, especially among admins. Everyme 01:37, 5 October 2008 (UTC)

Given that admins feud on arb pages, and under Jimbo's nostrils, why would crat eyes stop it? Of course it will be used in feuds. Are you really telling me there would not have been requests for desysoppings of SlimVirgin, JzG, Cyde, jayig, and I could go on. Read what I wrote above, this process will desysop no-one who does not currently get desysopped, but it will cause many unsuccessful and dramatic attempts to do so. Some will be trolling and easily stopped by crats, but others will have some measure of good faith involved. People who earnestly believe that certain people should not be sysops. There are simply not a bunch of admins out there, currently un-desysopped, where there would be a consensus to desysop. We all think some people should be desysopped - but we'd never agree who, unless it is so flipp'n obvious, that they would get desysopped under the current system anyway.--Scott MacDonald (talk) 01:47, 5 October 2008 (UTC)

Worst case scenario, it just devolves into an RFC-- and we allow those. This would just say that if, by some chance, the RFC _does_ generate a strong community consensus, we might be able to settle the dispute there without automatically dragging Arbcom into it. --Alecmconroy (talk) 01:51, 5 October 2008 (UTC)

In most case scenarios it is a drama magnet, in search of causing drama with no hope of doing good.--Scott MacDonald (talk) 02:02, 5 October 2008 (UTC)

It depends on your opinion, I guess. If you think one or more of the admins you mentioned (or others) should and will at some point probably be desysopped e.g. by ArbCom anyway, then why not let a specific process handle that in a much more timely manner? The difference with ArbCom is that this process would have just one, very specific, verdict. Everyme 01:56, 5 October 2008 (UTC)

I don't think they should be desysopped. But that's not the point. The point is that the process CAN'T handle it, since in none of those cases, nor in the cases I think should be desysopped, nor in the ones you think should be desysopped, will there ever be a consensus to desysop. You'll only get a consensus where it is so obvious that desysopping would have happened anyway. This process is quite useless, and will not result in any more desysoppings, but will result in more drama and bad feeling.--Scott MacDonald (talk) 02:02, 5 October 2008 (UTC)

Doesn't it seem right that there should be _some_ way from the community to assess their confidence in whether an editor should have the bit? Right now, we have one and only one such process, RFA-- and its decision are irrevocable. Look at this as CONSENSUS CAN CHANGE applied to the RFA process. If arbcom appointed the admins directly, I'd understand only letting arbcom de-appoint them. But we appoint admins based on consensus. And we should acknowledge that consensus can change, and the candidate who we thought might be a great admin turned out to be better as an editor. No harm, no foul-- sometimes it's not going to work out. We should be able to acknowledge that, and sometimes use consensus to ask people to go back to focusing on editing, without the drama of subjecting them to the trauma of an Arbcom "Courtmartial". --Alecmconroy (talk) 05:14, 5 October 2008 (UTC)

Maybe it doesn't seem right. But wikipedia is pragmatic. This proposal will do nothing useful, and solve nothing useful. Thus we don't do it.--Scott MacDonald (talk) 09:11, 5 October 2008 (UTC)

Oppose

I oppose this proposal unless all current bureaucrats resign and get re-appointed with the clear understanding that they would have this new "power". They did not become crats because of this ability, and to give them a new authority that they didn't "campaign" on is changing the rules in mid-game. I know I don't have a lot of time here as a registered account, but I've read Wikipedia and used it as a customer for a long time now. Little Red Riding Hoodtalk 00:04, 5 October 2008 (UTC)

A definite point, NurseryRhyme. There is something a little naughty about electing someone to and giving them new powers midway though.

It's justifiable, to an extent, because ultimately 'Crats are empowered to assess community consensus about who should and should not be an admin. So, from a certain point of view, we already gave the, power to make this kind of call.

But, it's a definite tweak in the role of the 'crat, so I totally would understand requesting that each crats ask permission from the community before they being close RFdAs (or whatever). That shouldn't be a problem at all-- the 'Crats are some of our most trusted users ever. I can see this proposal passing or failing-- but if it passed, I can't imagine the community wouldn't trust the 13 active 'Crats to administer it. --Alecmconroy (talk) 00:52, 5 October 2008 (UTC)

• The point is entirely moot since new tools for specific user rights groups have been introduced before, e.g. admins' ability to grant rollback rights. Everyme 01:59, 5 October 2008 (UTC)
  • An opinion is never moot. Little Red Riding Hoodtalk 03:36, 5 October 2008 (UTC)

Amen. NurseryRhyme-- what would you think about this policy if we added a proviso stating that the current 'Crats, the ones who elected before this proposal, would have to get some sort of community endorsement before they could close out any RFdA (or whatever)? --Alecmconroy (talk) 05:00, 5 October 2008 (UTC)

To be honest, I don't see what problem this solution is trying to resolve. Little Red Riding Hoodtalk 19:57, 5 October 2008 (UTC)

Thoughts

I am not opposed to this proposal, but I'd just like to list a few concerns, some of which have been brought up above - I'm going to use the phrase "PFD" for "Proposal for Desysop" to make it more readable.

• Necessity - TenofAllTrades asked for recent examples where the PFD process might have been useful, and would've reduced the burden on ArbCom. Looking at the recent list of desysoppings, all I see is the following situations
  • Breaches so obvious that they wouldn't have reached PFD - examples: (User:Chet B Long & User:PeterSymonds, User:Archtransit, User:Robdurbar, etc.
  • Desysoppings that emerged from related ArbCom cases that covered far more than the actual conduct of that user in particular - examples: User:FeloniousMonk, User:Vanished User.
  • Ones that may have benefited from PFD - and there are very few - possibly User:Tango (very controversial anyway), User:Alkivar (this is probably the best one, but already we're almost a year into the past).
• This makes me wonder how much drama and controversy would be generated for the sake of a very few cases.
• Disincentives - for (a) admins to make any decisions that may be seen as controversial, especially in notable areas of dispute, and (b) for admins to do any work in certain areas - nationalist battlegrounds springs to mind, as does images - image work makes you no friends at all, only enemies - hell will freeze over before I see a posting on an admin talkpage that says "Thankyou for making this article more encyclopedic by removing all those spurious fair-use images".
• Checks and balances leading on from the previous point, admins that do work in such areas could be subject to repeated PFDs, every time they upset another group of editors working in such areas. Is there to be some sort of system to prevent repeated spurious PFDs? Just having something like "10 autoconfirmed users proposing" isn't going to prevent abuse here.
• We don't like that admin - there have been a number of RFAs that have involved quite a bit of dubious behaviour, involving (but not limited to) canvassing, attempting to show people in the worst possible light, and opposes that appear reasonable but are merely covers for "This user doesn't agree with my POV". How many times are such admins going to be PFD'd? This also leads onto
• Problems for crats - 'crats are going to have to be really careful when determining consensus in difficult cases. Another problem (and this assumes bad faith, but meh) is that I can also see possibilities in certain areas where groups of editors create lots of autoconfirmed sleeper socks purely for the reason of "voting" in a PFD. Are we going to run CU on all PFD voters? Black Kite 12:29, 5 October 2008 (UTC)

Having said all this, there is the kernel of a good idea here - I am merely wondering about the huge amount of drama that may be created for little valuable result. Black Kite 12:29, 5 October 2008 (UTC)

That's what I was trying to say above, but you've said it better. This proposal looks good in theory, but when you you ask "what will it cause to happen, that doesn't happen already?", the pragmatic response is a raspberry.--Scott MacDonald (talk) 12:41, 5 October 2008 (UTC)

Yes. I support Black Kite's view completely. Spartaz ^Humbug! 13:01, 5 October 2008 (UTC)

Some excellent points, Black Kite. In particular, I think we need to give more consideration to we avoid "bad-faith PFDs"-- I don't really think we have that where it should be. The same thing is true of the concern that admins will be targeted for doing their job well, but earning enemies along the way just for doing what an admin should do. Those are two inter-related objections that I do feel like haven't been really answered yet. I do have confidence that between the wisdom of the community, the 'Crats, arbcom, and the board, we won't see any actual desysoppings for bad-faith reasons-- any one of those groups can put the brakes on the process.

But, as it currently stands, the procedure could still be used to annoy or otherwise be abused. That everything would come out right in the end isn't sufficient-- we need to come up with some sort of abuse prevention.

The other objections don't concern me as much. Yes, PFDs would be dramatic, but the alternative is an automatic arbcom anyway, which the nuclear bomb of drama. Plus, just as an RFC is a reasonable step to take in trying to resolve disputes, it seems like getting a sense of just how much confidence the community has in an admin would be a valuable step to take before proceeding to arbcom.

Looking back over the past year, you say a PFD probably wouldn't have resulted in any desysoppings. Let's assume that true-- is it possible that a PFD might have averted any arbcom case-- that an admin's behavior would have been sufficiently endorsed by the community that a full arbcom case would have been unnecessary?

For that matter-- how many disputes might have never even arisen, either because an earlier PFD was "too close for comfort" and a wayward admin decided to mend his ways, or else because a PFD showed "such clear support" that a complaintant realized his dispute without merit?

And then, looking back over the past year, imagine instead the RFA people had known there would be an easy way to "undo" their decision? How many more admins might we have been able to recruit?

At the end of the day, we generally recommend that people do a user-conduct RFC to try to resolve disputes before hand. Whatever drama there is in a PFD already exists under a different name. The only difference between a PFD and a user-conduct-RFC is that, _IF_ everyone basically agrees on a course of action, the 'Crats can recognize that and act on the consensus.

That said-- we do need to come up with some way to prevent abuse of process, and I don't have the answer to that. As a start-- how would we deal with someone who kept filing RFCs or Arbcom cases against a user? Is there some check in place to prevent that, or just the discretion of arbcom to decline cases?

More thought definitely needed here. I feel like this is slowly stumbling towards a better project, but the concerns, about abuse in particular, definitely have merit. --Alecmconroy (talk) 13:28, 5 October 2008 (UTC)

Response to doubts

"This will lead to drama"

We already have plenty of daily drama at WP:ANI and also occasional drama at WP:RFC when editors are unhappy about an administrator. Unfortunately, these current processes for complaining do not have any definite endpoint, leading to endless drahmaz, such as SV-Cla68-FM-JzG-Voldemort-Harry Potter-et al. The new process would allow us to point the complainer to a venue where the could most likely get an up or down answer to their complaint.

"This won't lead to many more desysops"

No, it probably won't. But it will lead to a fair number of clear decisions. Complaints will be filed, and sometimes no further discussion will be needed, because it will be shown that the complainers have not attempted to resolve the dispute in good faith. Other times a complaint will be deemed worthy of discussion, and the community will support the administrator in question. A bureaucrat will come along and pronounce that the admin still retains the trust of the community. That has the power to eliminate drama. "The complaint was heard and evaluated, now stop complaining and get back to work."

I was the one who flushed out User:Archtransit. It took a month to get them desysopped via ArbCom. The proposed process would have gotten the job done faster, and would have saved a lot of wasteful discussions and unnecessary disruption.

"ArbCom will still have to decide most of the cases"

Probably true, but they won't be bothered with trivial cases, and the cases that do come to them will come with a nice set of evidence and plenty of feedback from the community. This may shorten the cases, and help ArbCom stay in sync with the community.

"This will scare sysops from working in hot zones"

If a sysop behaves neutrally with respect to content, the good faith editors on both sides will appreciate their involvement. Ask the homeopathy editors which "side" I am on, and both sides will probably say "ours". Ask the Sri Lanka editors who I favor, and you'll get the same sort of answers. Actually, I am on no side, except Wikipedia's. A pack of nationalistic editors might number a dozen. They will not be able to overpower a community consensus, and our bureaucrats are smart enough to see and discount that sort of voting pattern.

"This is scary"

New things are sometimes scary. I notice that virtually all of the non-sysop editors who have commented here support this proposal. The opposition seems to consist entirely of sysops. Dear colleagues, do not be afraid of something new. Jehochman ^Talk 15:27, 5 October 2008 (UTC)

I haven't a doubt about this. I've a fundamental objection along the lines of "sounds good, but when you examine it, it is both unnecessary and practically useless." The reasons for change are simply not addressed by the solution. It 1) will not decrease drama, 2) it will not remove any more poor admins, 3) it will not take significant load of arbcom 4) it will not make admins better. In short it is another process-driven non-solution to a variety of supposed problems, with no practical benefits. It is a distraction from the work in hand. To make a major change, you need to show that the change will make a significant difference/improvement to the status-quo, for a whole load of reasons, this will not.--Scott MacDonald (talk) 15:50, 5 October 2008 (UTC)

I concur with Scott, and I still have concerns about the effects on "hotzones." Jehochman, you may unfortunately be more the exception than the rule. -- Avi (talk) 15:56, 5 October 2008 (UTC)

Agree with Scott, this proposal would not be helpful. The conflict-junkies might like it, as it would give them a new level of bureaucracy to play with, but for the vast majority of other editors and admins, it would just cause a lot of wasted time and drama, which would not be balanced out by sufficient benefit. --Elonka 16:06, 5 October 2008 (UTC)

That's your opinion, and I for one heartily disagree. I think there could hardly be more drama, and hardly more wasted time than right now. And I think the benefits would not only not have to balance out any alleged drawbacks which I don't see at all, particularly in light of the current situation with its suboptimal status quo, the benefits would indeed be the direct and immediate relief from nothing but the current utter lack of a process like this one. Everyme 16:14, 5 October 2008 (UTC)

I concur with Scott's assessment as well. I also will note that Jehochman's recollection of the Archtransit case isn't necessarily the only interpretation of events. Despite Archtransit's apparent flaws, there was still a substantial portion of the community willing to extend him a mentorship opportunity barely a week before the ArbCom desysopped him. Scarcely a week before he was desysopped by ArbCom, a majority of participants at Archtransit's RfC endorsed a temporary suspension of tool use coupled with a mentorship program in lieu of desysopping. I've mentioned this further up the page, but it bears repeating. There wasn't anything approaching a community consensus to desysop Archtransit, right up until the moment he was caught socking. The notion that ArbCom dithered for a month in the face of a clear community request is just not accurate. TenOfAllTrades(talk) 16:16, 5 October 2008 (UTC)

If our RFC process had teeth, I think the results at Archtransit's RFC would have been different. Had it been an option to send them back to RFA, I think a lot of people who have gone for that. Archtransit would not have passed a second time. Of course, we can't know until we try such a process. The argument, "Prove something new will work before we can try it," ensures that all proposals fail. Jehochman ^Talk 16:20, 5 October 2008 (UTC)

TenOfAllTrades's summary fits my recollection- Jehochman's account seems to rather exaggerate his own role in the matter. Archtransit was exposed by checkusers for having used multiple accounts, whilst it is true that Jehochman was one of those who generally felt Archtransit's conduct was erratic and cause for concern (unsurprisingly as Archtransit had blocked him) I don't believe he had suspicions that multiple accounts were being abused. Or am I mistaken there Jehochman, did you raise such suspicions on-wiki or to checkusers? WJBscribe (talk) 16:22, 5 October 2008 (UTC)

I suppose on reflection it was Jehochman's block of User:Congolese fufu (which turned out to be one of Arch's socks) that lead to Arch blocking him, which was what resulted in the level of scrutiny Archtransit ultimately faced. So yes, I supposed "flushed out" may be an accurate description. WJBscribe (talk) 16:36, 5 October 2008 (UTC)

I was just going to ask whether you ask out of personal curiosity or out of a conviction that this is relevant to the debate at hand? Everyme 16:39, 5 October 2008 (UTC)

Curiosity. It was, I admit, a tangent. WJBscribe (talk) 00:17, 6 October 2008 (UTC)

Archtransit had a very odd response to my action against his socks. Archtransit apparently was Derek1x, who has done quite a bit of socking. Much mischief can be avoided by detecting and blocking socks, which was my general intention at the time, though I had no specific information to link Derek1x to those particular socks. Jehochman ^Talk 17:58, 5 October 2008 (UTC)

all proposals fail — Jonathan, this in itself poses the most serious problem. Maybe it's just me, but in what I've read on this page, I can't help but recognise the same basic patterns and recurring reasonings of opposition as with many other proposals that might shake up the status quo and open up the way to a better, more evolved compromise (or yes, possibly the way into galloping drama, darkest chaos and Social Darwinism). Everyme 16:32, 5 October 2008 (UTC)

Perhaps we should compromise and make this proposal an optional process. Would the bureaucrats support that? If it proves to be an effective process, we can revisit the idea of making it mandatory. Jehochman ^Talk 18:00, 5 October 2008 (UTC)

I think you might be on to something. Optional how?

What if we went ultra-simple and just say:

1. We have RFCs like normal.
2. If an RFC should find a consensus, then a 'Crat may, if they so choose, recognize that consensus, and the result be desysopping.
3. All such decisions are, of course, appealable to arbcom and the board.

It's not the destination we should ultimately reach, but maybe it is better to take things slow and incrementally. --Alecmconroy (talk) 18:13, 5 October 2008 (UTC)

Response to response to doubts:

• Why do you say it took a month to desysop Archtransit? The RfC was opened on February 4 and he was desysopped on February 19. That's two weeks, same as in this proposal. No one here has produced any evidence that arbcom has moved unusually slowly for the kinds of straightforward cases this system is designed to handle. They have moved very slowly for complicated cases involving long-standing admins with deep roots in the community, but I doubt this process could have handled those any better.
• As for the idea that "If a sysop behaves neutrally with respect to content, the good faith editors on both sides will appreciate their involvement," it is sometimes true and mostly not, partially because a neutral role isn't necessarily a centrist one: if an article is deleted or a particular topic is permanently excluded, the admin can claim impartiality but not necessarily neutrality. This is first and foremost a problem with BLPs, particularly those with editor subjects and worse, editor fans.
• Lastly, for drama: don't you fear that what will happen most of the time is that the results of an RfC will be inconclusive, the bureaucrats will choose to do nothing, and then will be hounded for their inaction? With arbcom, a case can be closed, formally and finally. This system has nothing to take the place of that. I fear it will make being a bureacrat a much more unpleasant task, not because of what they do but because of what they don't do. Chick Bowen 19:18, 5 October 2008 (UTC)

"Hot Zones"

One major objection to this kind of and any kind of stronger potential sanctions on administrators is that it will "Reduce the number of people willing to wade into hot zones of conflict".

Frankly, I have to say this is a grossly flawed argument, and those making it have not thought the consequences through as much as they think. Yes, strengthened sanctions on problematic administrators would reduce the number of administrators who 'wade into' conflicts. However, what has not been established is that it would be a bad thing if that happened.

From my past experience, administrators 'wading into hot zones' when they are either not prepared to or incapable of acting fairly and appropriately, have lead to worsening of those conflicts. That they continue to do so is because there have never been any consequences to them for making a conflict worse.

We should be warding off administrators from jumping into 'hot zones' without preparation. If this recommendation does so, the it is an additional benefit not a detraction. --Barberio (talk) 18:18, 5 October 2008 (UTC)

I think the argument about "Hot Zones" is that it doesn't matter how well-informed, how prepared, how capable etc. you are. If you step into some such areas of conflict, you are inevitably going to annoy a lot of people. Some admins may, I grant you, annoy less people than others but there's no admin, no matter how saintly or skilled, who could avoid pissing off a good fraction of the disputants - and the effect only increases the longer an admin stays involved. CIreland (talk) 18:27, 5 October 2008 (UTC)

I completely agree that wading into a conflict about which you are not knowledgeable and not prepared to put in the work to make yourself familiar with the various factions is asking for trouble, and administrators have clearly done that in the past, thus increasing the drama levels rather than solving any problems. Whilst I don't believe that such administrators such be sanctioned as such, they should learn from their errors. As stated above though, there are certain battlegrounds where making any administrative decision at all instantly makes you unpopular with large swathes of editors, regardless of whether it is the correct one or not. However, as has been highlighted by recent RfARs, there is a need for neutral admins in many areas where the existing active admins are tainted - sometimes unfairly and sometimes not - with suspicions of bias towards one side or another. If people are scared away from these areas, they will, logically, not improve - and may become worse. Black Kite 18:47, 5 October 2008 (UTC)

In conflicts where no administrator would be capable of "solving the issue" without causing community disruption, they shouldn't be trying to in the first place.

No. Seriously. Administrators are not supposed to be getting involved in community disputes.

These kinds of issues were always supposed to go through WP:Dispute Resolution, not be handled unilaterally by administrators. It specifically says in all kinds of places that Administrators are not referees or judges.

So this does seem to be an argument based on "But then Administrators might not want to do the things they weren't supposed to be doing in the first place." --Barberio (talk) 19:30, 5 October 2008 (UTC)

You are right that community disputes should go through dispute resolution. The problem is that the hottest of the "Hot Zones" have already been through dispute resolution and all the way to arbitration. Typically, arbitration has added the area to WP:SANCTION, many of which specifically mandate admin adjudication of behaviour and the editorial process in those areas where dispute resolution has failed is likely to continue to fail. CIreland (talk) 20:46, 5 October 2008 (UTC)

Barberio, sysops sometimes have to apply blocks or lock articles. I am not certain how many times one, the other, or both (or multiple) parties to a disagreement have complained about what they felt was the "wrong version" or "unfair treatment", but saying it's somewhere between aleph-null and aleph-one does not feel like a gross exaggeration. Hot one involvement may have nothing to do with content, but applying a block or a lock will often be (mis)viewed by participants as a content statement, and the "offending" admin will be complained about on AN, AN/I, COI, and e-mails have been sent to the foundation lists as well. It can already be seen on RfA that otherwise acceptable candidates have been torpedoed by content related issues, and not their ability to act properly in accordance with guideline and protocol. I am loathe to extend what may amount to a weapon to any clique on wikipedia. Barberio, if everybody here acted fairly and maturely, with wikipedia's best interests as opposed to their own, I'd be much less opposed to your position. Then again, in that case, there may have been no need for this proposal either. Thus, although the majority, and even vast majority, of wikipedians do edit responsibly and for the greater good, unfortunately, I, and apparently others, believe that the minority does pose a potential threat to wikipedia stability. I still maintain that this is ArbCom's purview, and if they need more efficiency, they should be the ones responsible for setting up any related subcommittee whose advice on desysoping they may choose to rubberstamp. -- Avi (talk) 20:27, 5 October 2008 (UTC)

• Since I threw this objection up quite vocally, I'll respond here. To me, the response that "admins aren't supposed to be involved in community disputes" is not connected to how wikipedia works in most of these areas. I am not talking about garden variety problems where an occasional look at content/conduct will come about from a 3RR report or where an admin will just protect the "wrong version" for a couple of days while people cool off. I am talking about content areas where groups of editors have been at it for months or years, where real life differences of opinion across whole communities are being expressed here, where multiple steps in DR have been taken multiple times, where sockpuppets are used extensively. I can list a dozen or more of these sorts of articles. Most of those remain close to what we would like them to be only because there are admins willing to step in and make decisions that make everyone unhappy. They make blocks and topic bans, they force discussion of wording changes, they fish out socks, etc. It is a very necessary and very unpleasant task. Some admins are capable of doing with aplomb and can make everyone happy. Most are still capable of mastering the content, managing disputes and enforcing policy in a competent manner. A small fraction are incompetent at this particular job to some degree. The problem is that the view of most editors on the page of the 2nd and 3rd type of admin is about the same: everyone hates them. If we implement a de-sysoping procedure with teeth, we need to be sure that we hazard only the third type, and not the second type. If it appears that a reasonably competent admin will face the loss of the bit for working in an unpleasant topic area, they will probably decide to just not work in that area. Will admin coverage of editor malfeasance disappear? No. We will still have AN/3, AN/I, etc. However the method of handling disputes will become much more cumbersome and the distribution of coverage much more sporadic. Content presented to readers will suffer. Good faith editors will leave and tendentious editors will determine the changes made to these sorts of articles. I'm not saying I oppose this proposal wholly, but it needs some sort of provision to assure admins that if they aren't engaging in gross misconduct they won't face a real threat of desysopping. Protonk (talk) 16:49, 6 October 2008 (UTC)

I concur. Admins shouldn't be penalized for being willing to make the hard calls-- calls where no matter how perfectly you do your job, someone will be mad.

The phenomenon I _think_ Barberio is kind of trying to discuss is where admins who have strong views ABOUT the hot zones jump in and get just as hot and tendentious as the participants, only they're participants with admin tools.

Such admins might exist, but even if they do, they're outnumbered 100-to-1 by good admins who are doing their jobs well by trying to even-handedly apply policy. And we need some way to ensure that this new process wouldn't get used as a cudgel to repeatedly batter the people who aren't doing anything wrong. --Alecmconroy (talk) 18:18, 6 October 2008 (UTC)

Right. I agree that is his point. However, in order to deter that, we should be careful not to deter the ~60-70% of admins who are basically good at the job but might ruffle feathers. Also, we should consider that the "penalty" for rushing into a contentious article and making poor judgments should not be desysopping. I just think that the overall proposal needs some safety mechanisms. As I noted way above...I don't really have a solution. It isn't an easy problem. Protonk (talk) 20:09, 6 October 2008 (UTC)

Makeup of panel-Abuse of Process

While I mainly support this proposal, I would like to see either a steward or Arbitrator serving as advisors to the panel, to give assistance with issues that might crop up in their respective areas of expertise. What do others think? Geoff Plourde (talk) 18:30, 5 October 2008 (UTC)

Also could an abuse of process provision be added? Geoff Plourde (talk) 18:33, 5 October 2008 (UTC)

Regarding the first suggestion: I think that would probably be unnecessary bureaucracy; the position of Arbitrator is not really a natural choice for a group whose function is to advise on community-centred (de-)adminship matters. Anthøny ✉ 16:45, 6 October 2008 (UTC)

i just mean as resources to handle things that might require their expertise. Geoff Plourde (talk) 03:43, 8 October 2008 (UTC)

Remove "admonish"

If the community is unhappy with someone performance to the point of taking them to a venue like this that in itself is a warning. I am uncomfortable with the idea of one crat distilling the community comments into a single admonishment. otherwise, excellent proposal. Viridae^Talk 22:36, 6 October 2008 (UTC)

Bad Idea

I actually disagree with the idea of giving crats the right to remove sysops and its only because this wiki does not have a de-sysop policy and until a policy is made which ensures users lose their rights due to inactivity and abuse, I will not support such an idea. Stewards have been doing a great job in terms of removing rights and we should stick with that and since the last decision on whether to de-sysop a person lies with the arbcom, I would disagree with it even more. Instead of trying to remove sysops, why not try to improve the RfA process which is all botched up. We only have an average of 1 sysop right removal on enwiki per month and its mainly due to editors getting bored or pissed off at something and retiring, so i don't see a need for the crats to be able to de-sysop them when it can easily be done and logged on the meta log without much drama because believe it or not, if crats on enwiki had the right to remove sysops, we will have probably more than 5 de-sysops a month which does not look good. there is a saying.. "why fix something, not broken"... --Cometstyles 11:55, 7 October 2008 (UTC)

I totally agree. Regards, —DerHexer (Talk) 11:57, 7 October 2008 (UTC)

Simpler sollution

Since it is admitted that the number of times that there's going to be a consensus to desysop is going to be extremely small, I propose the following, which does not drastically vary from current processes, nor give crats any new powers.

If at the end of an RFC, there's an overwhelming feeling of no confidence in a sysop, then a motion should be raised at RfArb. The motion would simply be "on the basis of this RfC, the community has no confidence in User:Example being a sysop, and they shall be desysopped". Arbcom then have three options - 1) to agree with the motion on the basis of community confidence 2) to disagree 3) to invite the opening of a formal case, where things are more complex. In any rare cases where there is an evident " community consensus to desysop" in the RfC, arbcom are then able to desysop without the time and paperwork of a full case.--Scott MacDonald (talk) 12:07, 7 October 2008 (UTC)

This is a good idea. I quite like that it doesn't require any policy changes to go ahead. (We were this close to doing almost exactly what Scott proposes with Archtransit after Arch started ignoring the terms of his mentorship, and that was more than a year ago.) It doesn't create a new bureacracy; it doesn't dramatically expand the authority of sitting 'crats.

I don't even necessarily think that a new fast-track version of Arbitration is necessary. I expect that any RfC the ArbCom would take seriously would already have some pretty explicit findings of fact built in to form the template of a decision, and allowing ArbCom a couple of weeks to review evidence and have sober second thoughts probably isn't a bad thing. If the parties are particularly concerned about an admin's conduct, they can always request a temporary injunction to desysop the admin for the duration of the case. TenOfAllTrades(talk) 12:33, 7 October 2008 (UTC)

If we could write that into policy and thereby coax ArbCom to look at cases of "lost confidence", rather than the higher standard of "clear abuse", that would be an acceptable solution. I would also like to see ArbCom pass motions stating "no abuse" when there iwas a clear consensus in that direction. We need a way to get a decision on these RFCs to help reduce drama and encourage better administrative practices. Jehochman ^Talk 12:35, 7 October 2008 (UTC)

How would you define "an overwhelming feeling of no confidence", tho? Actually, I cannot come up with any recently desysopped admin where the community had an overwhelming feeling of no confidence in his/her abilities as a sysop, so I'd say that this solution really wouldn't change anything. --Conti|✉ 12:38, 7 October 2008 (UTC)

It is the same standard as the current proposal demands - a community consensus. I suspect that, in fact, that will only ever exist in clear-cut cases anyway, which is the weakness of the current proposal.--Scott MacDonald (talk) 12:41, 7 October 2008 (UTC)

I would support this Simpler Solution. I confess that, for ideological reasons, I do strongly prefer the original formulation where "1) community forms consensus, 2) 'crats assess and enforce that consensus, and 3) Arbcom can overrule". But I have to admit there have been some very strong pragmatic objections raised when it comes to that model, so ideology may have to take a back seat to practicality on this one. Given that, I think Scott's "simpler solution" does make a lot of sense: "1) community consensus goes directly to 2) Arbcom assesses, enforces, or overrules". No new powers on the 'crats, no new pressure on the 'crats, but the fundamental mission of letting the community give some feedback and oversight to admins and to arbcom. Good call.

And as Jehochman says above, I too think we should write things in terms of "community confidence" not necessarily "abusive misuse of power". It's an inescapable fact that sometimes the community will conclude a given user shouldn't be an admin-- but that doesn't automatically mean we believe the user has been an abusive admin who has, in bad faith, misused admin powers. It's apples and oranges-- and we shouldn't let a few bad apples automatically tarnish the reputation of the valued oranges who, through no fault of their own, just aren't working out as admins. --Alecmconroy (talk) 13:24, 7 October 2008 (UTC)

To Jehochman: I don't think we need to write it into the policy. Wikipedia policy (with a few exceptions like the privacy policy which are 'etched in stone' by the Foundation or Jimbo) is descriptive rather than prescriptive. If the community is able to come to a clear consensus that a particular administrator shouldn't have the tools – even in the absence of a pattern of abuse, per Alecmconroy's comment above – then the ArbCom should be willing to respect that position. The trick is to be absolutely clear in the RfC about what the community is asking for, and utterly explicit how WP:IAR may apply if necessary. If the new process works, then we can look at any tweaking of policy wording that might be necessary.

To those who are concerned that it will be difficult to achieve 'community consensus' that a user shouldn't be an admin — well, that's not going to be changed by any other formulation of this proposal. Either the community will be able to come to a consensus or not. I think it's safe to say that a regime wherein the community could demand (and get) a desysopping without consensus (by a simple majority vote, say) in the absence of clear abuse would be a disaster. TenOfAllTrades(talk) 14:46, 7 October 2008 (UTC)

It's worth a shot. It can't hurt anything. The reason it won't help, though, is that even with obviously incompetent sysops, you'll never get an overwhelming vote of no confidence. There's just too many people around who will say "Oppose desysopping- so-and-so is a good admin." This is why my preferred way of handling this is to let the decision-makers (be they crats or arbcom) decide whether a reasonable editor would have no confidence in the sysop in question. Leave numeric voting out of the picture, and maybe you're onto something. Friday (talk) 16:20, 9 October 2008 (UTC)

Semi-radical idea

Here's an idea, altho I can't decide whether it would be helpful or harmful. Start supporting almost every RFA- even for those candidates you're not convinced are well suited to the job. Then, if large numbers of incompetent sysops start running around causing problems, one of two things will happen: 1) Arbcom will quickly resolve cases of incompetent sysops, or 2) Ideas like this one will start getting wider acceptance. Either option would fix the problem, in the long term. The question is whether the short-term harm would be something we can live with or not. Friday (talk) 16:28, 9 October 2008 (UTC)

• WP:BEANS, obviously, but people can do a LOT of damage with the tools - do we really want to take the risk? Black Kite 16:41, 9 October 2008 (UTC)

In order to justify an otherwise useless new bureaucratic procedure (because apparently we can't find any candidates for desysopping right now) we need to deliberately harm the encyclopedia by generating a pool of incompetent admins? Good plan. If you truly can't decide whether your idea would be helpful or harmful, then you've obviously lost track of the purpose of the project. Even if your destructive suggestion were to work, it wouldn't work. Creating a speedy, efficient process for desysopping people who are so obviously incompetent that they wouldn't have passed a normal, unmanipulated RfA isn't going to do anything useful.

Friday—go sober up, then reread WP:POINT and WP:GAME before you come back. I hope you'll be embarrassed by your suggestion in the morning. TenOfAllTrades(talk) 16:42, 9 October 2008 (UTC)

It's daytime here. I didn't say support every RFA- just most of them, even if you're unconvinced. One possibility is that people are generally reasonable and will learn by doing, and they won't actually cause many problems. If this is how it works out, it's a win. I submit that this idea is no more preposterous than letting the public edit an encyclopedia. A good many editors are apparently already doing exactly what I've suggested here, and the wiki hasn't exploded yet. Friday (talk) 17:00, 9 October 2008 (UTC)

To put this another way.. I've been saying for a couple years that I'd be way more willing to support RFAs if I thought we had effective means of dealing with bad admins. Maybe I've had it backwards the whole time. The number of people saying "OK then, show me the bad admins" supports the idea that I've got it backward. Maybe the way to get there is to go ahead and start being lenient, and see if problems occur. Then, we fix those problems as they occur. This is how most of the rest of Wikipedia works, and somehow it doesn't result in disaster. I'm not suggesting anyone support a candidate they know is bad- only that they take chances on those they're not sure of. If we're lucky, nothing bad will even happen. Friday (talk) 17:08, 9 October 2008 (UTC)

I think there is a deal of sense in what Friday is saying. Not that we should sysop bad candidates, but that the argument for having a desysopping process is that there are a lot of borderline candidates - that might make good sysops but the community either don't feel that there's quite enough evidence to make the call, or that the evidence has a shadow of a doubt, and if we sysop, we can't easily reverse the sysopping, unless the poor sysop steels the lunch money or is caught with kiddie porn. If the community could be sure that the odd wrong decision to sysop could be reversed, then it would take a few more (educated) risks in sysopping - and (mainly) we'd get more good sysops. I've always thought that enforced recall mechanisms (which is what this is) were a awkward way of doing that - and they are always likely to run into significant opposition. They tend to say "in order to sysop a few more borderlines, we need the power to desysop ANY of our 1400 admins" - and frankly as attractive as that might be to some, it opens up disproportionally large cans of worms. There is another option. That is a more limited process that allows crats to reverse a sysopping decision with borderline candidate cases (or those candidates we've not got much to go on) under certain conditions. Various options spring to mind - what if we said "if a candidate has over 50% support, and much of the opposition rests on inexperience or uncertainly rather than strong objection to recent past behaviour, a crat may offer the candidate provisional sysop status". Provisional sysop status would mean that the decision may be reversed at any time in the next (say) 4 months, if any two crats agree that the experiment seems to have failed - the candidate can always appeal such a reversal by launching a fresh RfA. Anyway, just a thought. It would lead to more admins, without having to find a way to cut this knot.--Scott MacDonald (talk) 17:53, 9 October 2008 (UTC)

"There are a lot of borderline candidates" Pedro :  Chat  20:09, 9 October 2008 (UTC)

Always depends on how we define border. But let's focus on how we grant powers to more borderline candidates with a safety valve of being able to reverse the trial. (Oh and when I say borderline, I don't mean well-known people who fail to get consensus due to opposition, I mean places where people just are not sure, but might be willing to see a trial, if reversible.--Scott MacDonald (talk) 20:17, 9 October 2008 (UTC)

Noted. Pedro :  Chat  20:27, 9 October 2008 (UTC)

Why not allow our bureaucrats to reverse any close decisions that turn out badly? For instance, open the 60% - 80% range to promotion, but reserve the option of un-promoting within 6 months. Jehochman ^Talk 22:53, 9 October 2008 (UTC)

Something like that. We'd need some finesing - firstly I think it should take the say-so of more than one crat, secondly I'd say users who would otherwise fail RfA should be given the option of accepting this "provisional adminship" - if they don't like the idea, they can wait and try another RfA later as at present. It would be an offer made in the case of close fails.--Scott MacDonald (talk) 23:03, 9 October 2008 (UTC)

I really don't understand why promoting someone is the one action in the entire interface that cannot be reversed. Every single other thing can. (Heck, even a rouge page merge could be undone, even though it would take forever to do). Admins can only be removed by a Steward. I can't understand the lack of ability to reverse. -- how do you turn this on 23:05, 9 October 2008 (UTC)

Who can technically flick the switch isn't the point. We are always going to desysop so few people compared to the number we sysop that restricting the power to stewards isn't an issue. However stewards only act under orders, the question is then whether there should be more ways of giving the order or not. One may have need of more capital judges than of hangmen.--Scott MacDonald (talk) 23:26, 9 October 2008 (UTC)

If an RFA is in the 60 - 80 percent range, the bureaucrats who are available have a discussion and decide whether to promote provisionally or not. If there is a provisional promotion, and subsequent problems, they can get together a second time and discuss whether to remove the bit. The discussions are informal, organized however the bureaucrats like. The point is to avoid wheel warring over a provisional promotion. As long as one 'crat represents to a steward that there is consensus to desysop a provisional appointment, the steward can take action. Jehochman ^Talk 23:30, 9 October 2008 (UTC)

New proposal - provisional adminship

Where an RfA would normally fail, but the candidate has 60%+ [a clear/comfortable majority in] support, and where in the opinion of the closing bureaucrat the opposition is largely due to uncertainly about the candidate, rather than informed opposition, then the bureaucrat may offer the candidate "provisional promotion". If the candidate accepts provisional promotion, then during the provisional period, any bureaucrat may instruct a steward to desysop if any two bureaucrats agree the candidate has proven to be unfit. Candidate appeals are made to the community through a new RfA.--Scott MacDonald (talk) 23:43, 9 October 2008 (UTC)

Sounds good to me. For policy, we won't specify the percentage. That has to be implicit. Jehochman ^Talk 23:45, 9 October 2008 (UTC)

The advantage is that the crats do not get power to desysop anyone, unless that person has chosen to accept that power over them. No one is compelled to.--Scott MacDonald (talk) 23:49, 9 October 2008 (UTC)

Tweaked to remove the arithmetic. Suggest "a clear majority" or "a comfortable majority" understood as "about 60% - but with crat room to manoeuvre"--Scott MacDonald (talk) 23:52, 9 October 2008 (UTC)

I very much like this idea. RFAs close in all different kinds of ways, some with clear and unanimous consensus to promote. Some are closer, on the borderline cases. Those are the ones that bureaucrats get discretion over. To reduce the issue of very few people being promoted, we can promote people we otherwise wouldn't have, and let them have a period to use their new tools, under the watchful eye of a bureaucrat (or another admin even). If there are issues, bureaucrats can discuss and remove. Really, all RFAs with more support than oppose should be considered, if the opposes are all really wishy-washy ones (like not enough edits). That way, we can really see if they'll mess up. And in all honesty, if someone managed to get more supports than opposes, we can assume they won't be awful, maybe not as clueful as some people. Remember, the end result isn't always a reflection on the candidate. -- how do you turn this on 00:01, 10 October 2008 (UTC)

Commons has gone down this line. [1]. Not sure how it is working.--Scott MacDonald (talk) 00:04, 10 October 2008 (UTC)

I do believe that's the first time they have done something like that. I expect it's working well. The difference is on Commons, the Bcrats there simply did it. I don't know how well that would work here, frankly (hence this discussion, I guess.) -- how do you turn this on 00:08, 10 October 2008 (UTC)

It might be useful if some of the Commons:Bureaucrats that participated in the recent Commons:Administrators/Requests and votes/SterkeBak/Bureaucrats discussion could comment here about their rationale regarding this process. Cirt (talk) 00:10, 10 October 2008 (UTC)

To generate more discussion, can this proposal be posted at Wikipedia talk:Requests for adminship with noticed at WP:BN and WP:AN? This looks like a step forward that could gain consensus support. Jehochman ^Talk 00:16, 10 October 2008 (UTC)

Question: What length of time should the "provisional period" be for? Also, at the recent Bureaucrats discussion at Commons, that particular RfA ended at about 73%, but that was after canvassing. Cirt (talk) 00:19, 10 October 2008 (UTC)

@Cirt: that can be decided by the Bcrat(s), depending on how the RFA went. @Jehochman, good idea. I suggest you do. -- how do you turn this on 00:21, 10 October 2008 (UTC)

Marvelous idea. This would work well. --Alecmconroy (talk) 05:59, 10 October 2008 (UTC)

Upon reflection-- Suppose I'm commenting on an RFA and I would oppose full adminship, but support provisional. How could I register that? Just !voting "Oppose" would suggest I think the people shouldn't be an admin at all. I suppose I should say "Support provisional adminship", and the 'Crat would be left to balance that as an oppose vote when considering full adminship but as a support vote when deciding provisional adminship". --Alecmconroy (talk) 07:37, 10 October 2008 (UTC)

Now crossposted to Wikipedia talk:Requests for adminship. Jehochman ^Talk 08:19, 10 October 2008 (UTC)

To clarify, please discuss this further on Wikipedia talk:Requests for adminship NOT here.--Scott MacDonald (talk) 08:48, 10 October 2008 (UTC)

Passing on extended Bureaucrat rights

Whilst I believe there should be far, far better and easier methods of removing admins rights I disagree that the current crops of Bureaucrats should have these rights;

• None were "elected" based on a background of this level of trust
• Many are inactive
• Many have assumed rights that were not on the table when they gained the bits
• and frankly there are some who I would simply not trust to have this right.

Proposal (If we want to do this properly);

• All Bureaucrats who wish to have the additional right to remove +sysop run a new community request granting it.
• All other crats can stay as they are.
• Failure to be granted the removal right does not remove the crat flag.
• All future WP:RFB's are based on receiving all rights including de-sysop so as not to create two tiers.

Yes, I know usurp/rename was placed to the 'crats without debate but that was a technical thing and is hardly as politically problematic as a de-sysop. I'm extremly uncomfortable with granting some 'crats a desysop right when their current legitimacy is shaky at best. Pedro :  Chat  23:21, 9 October 2008 (UTC)

Seems reasonable. There is no need for any sort of technical exclusion, and the class of bureaucrats is small enough that we can track this by hand by creating a list somewhere. Jehochman ^Talk 23:27, 9 October 2008 (UTC)

Agree, doesn't as such need to be a technical software solution. All Bureaucrats can be trusted not to abuse. Pedro :  Chat  23:29, 9 October 2008 (UTC)

You say "frankly there are some who I would simply not trust to have this right". Either you trust them or you don't? -- how do you turn this on 23:33, 9 October 2008 (UTC)

No. I trust all our current active 'crats to not abuse features. I do not, however, trust all our crats (active or otherwise) with a de-sysop button. I believe some crats are far to over-zealous, others have no knowledge of current community norms, and others yet are best suited to the bits they currently have - in that I believe they wouldn't go tampering for fear of losing their bit. Bluntly, I trust all 'crats not to abuse de-sysop because I know some of them realise they're on the edge of losing a whole lot more - so by creating a "whitelist" as Jehochman intimates above they are kept out of the desysop arena by virtue of the potential to loose the other bits as well at arbcom. Pedro :  Chat 

I have to agree with Pedro; a lot of bureaucrats weren't "elected" ("!elected"?) to this position, and I wouldn't expect the community to broadly grant additional powers to a group of people that requested different abilities. (for example, my RfB was successful because people agreed I could be trusted to gauge consensus on RfAs and that I'd help with rename backlogs. Neither trait lends itself directly to de-sysopping, and I would have no problem asking the community for the additional "right".) I also think that there being so many dormant 'crats is also an issue, as is the fact that some of them have been around longer than most of the community; I still vividly recall overturning a bureaucrat who indefinitely blocked an IP for month-old vandalism. That's not someone I'd trust to handle demotion privileges to, and the community should have more opportunity to voice its opinion than just once. EVula // talk // ☯ // 19:15, 17 October 2008 (UTC)

I think if this was to be implemented, it would be a perfect time to reconfirm all current Bcrats. Then we don't have to have a tiered system, which I really dislike the sound of. I mean, you either have all the rights with ability to use them, or you don't. I don't like the idea of some having a tool but not being allowed to use it. It simply doesn't make sense. -- how do you turn this on 23:39, 9 October 2008 (UTC)

Why do I need to be reconfirmed less than five months after my RfB? And there have been three more 'crats added since mine. A full-out reconfirm for everyone is a waste of everyone involved. I'm not opposed to reconfirming bureaucrats (there are significantly fewer of us, so it would be the same nightmare as reconfirming admins), but doing all of us at one time doesn't make much sense. Once a year sounds like a nice, round number. EVula // talk // ☯ // 19:15, 17 October 2008 (UTC)

OK, you're right. How about reconfirming all Bcrats who have been one more than a year? -- How do you turn this on (talk) 19:20, 17 October 2008 (UTC)

I think doing all the really old bureaucrats first (like TUF-KAT, who hasn't ever used the 'crat bits as far as I can tell, and Tim Starling, who in my opinion shouldn't technically be a bureaucrat but have all those rights as a dev). Then we can work up to the more active ones, creating a somewhat tiered system. EVula // talk // ☯ // 19:26, 17 October 2008 (UTC)

Straw poll

This is a straw poll to determine whether there may be consensus to implement this proposal. It does not violate WP:PNSD as there has been no discussion for a week. Please sign (only) under one of the headings below. Place further comments and discussion in the discussion section.

Yes, this proposal should be implemented

1. Stifle (talk) 15:16, 17 October 2008 (UTC)
2. Jehochman ^Talk 16:21, 17 October 2008 (UTC)
3. Synergy 16:25, 17 October 2008 (UTC)
4. Tombomp (talk/contribs) 17:32, 17 October 2008 (UTC)

   Alecmconroy (talk) 17:56, 17 October 2008 (UTC) changed to oppose based on lack of current consensus

5. Malinaccier (talk) 21:21, 17 October 2008 (UTC)
6. EVula // talk // ☯ // 21:49, 17 October 2008 (UTC)
7. -- penubag  (talk) 23:11, 17 October 2008 (UTC)
8. PhilKnight (talk) 00:11, 18 October 2008 (UTC)
9. — Athaenara ✉ 02:02, 18 October 2008 (UTC)
10. Giggy (talk) 02:07, 18 October 2008 (UTC)
11. Davewild (talk) 10:55, 18 October 2008 (UTC)
12. --Barberio (talk) 12:42, 18 October 2008 (UTC)
13. HiDrNick! 19:51, 18 October 2008 (UTC)
14. Celarnor ^{Talk to me} 02:55, 19 October 2008 (UTC)
15. Captain panda 22:28, 19 October 2008 (UTC)
16. Bstone (talk) 22:57, 19 October 2008 (UTC)

    Jayron32.talk.contribs 00:17, 20 October 2008 (UTC)

17. NuclearWarfare ^{contact me}_{My work} 01:07, 20 October 2008 (UTC)
18. Protonk (talk) 02:15, 20 October 2008 (UTC)
19. Sarcasticidealist (talk) 03:36, 20 October 2008 (UTC)
20. LamaLoLeshLa (talk) 04:01, 20 October 2008 (UTC)
21. Everyme 05:06, 20 October 2008 (UTC)
22. --Caspian blue 05:51, 20 October 2008 (UTC)
23. Wizardman 06:38, 20 October 2008 (UTC)
24. ~ mazca ^t|c 10:02, 20 October 2008 (UTC)
25. Eluchil404 (talk) 23:14, 23 October 2008 (UTC)
26. WilyD 09:24, 26 October 2008 (UTC)

No, this proposal should be tagged rejected

1. TenOfAllTrades(talk) 15:37, 17 October 2008 (UTC)
2. Mr.Z-man 16:09, 17 October 2008 (UTC)
3. Tom Harrison ^Talk 17:29, 17 October 2008 (UTC)
4. --Hut 8.5 17:37, 17 October 2008 (UTC)
5. Andrew Lenahan - Starblind 17:49, 17 October 2008 (UTC)
6. Spartaz ^Humbug! 18:58, 17 October 2008 (UTC)
7. Little Red Riding Hoodtalk 19:53, 17 October 2008 (UTC)
8. Philosopher ^{Let us reason together.} 21:45, 17 October 2008 (UTC)
9. Throwawayhack (talk) 18:13, 19 October 2008 (UTC)
10. --Scott MacDonald (talk) 20:45, 19 October 2008 (UTC)
11. - auburnpilot talk 22:40, 19 October 2008 (UTC)
12. — Coren ^(talk) 00:31, 20 October 2008 (UTC)
13. Chick Bowen 00:32, 20 October 2008 (UTC)
14. From reading the discussion, there appears to be no consensus for this to happen. Daniel (talk) 01:16, 20 October 2008 (UTC)
15. ~ User:Ameliorate! ^{(with the !) (talk)} 01:30, 20 October 2008 (UTC)
16. OhanaUnited^{Talk page} 02:28, 20 October 2008 (UTC)
17. I don't think there is consensus for this based on the discussion above. The current system for desysopping seems to work fine as it is, too. ···日本穣^{? · Talk to Nihonjoe} 04:00, 20 October 2008 (UTC)
18. --MZMcBride (talk) 04:15, 20 October 2008 (UTC)
19. Personally support proposal. Agree consensus is, however, not found for it. Pedro :  Chat  06:48, 20 October 2008 (UTC)
20. BJ^Talk 10:11, 20 October 2008 (UTC)
21. -- Avi (talk) 13:57, 20 October 2008 (UTC)
22. Changed my mind. I like the idea, and support it myself, but I can also see, after reading this page, that the consensus does not exist for it. Given the wording of the straw poll, I must put my vote here. Jayron32.talk.contribs 14:01, 20 October 2008 (UTC)
23. No consensus in the discussion so why is there a straw poll? KnightLago (talk) 14:03, 20 October 2008 (UTC)
24. Briangotts (Talk) (Contrib) 14:26, 20 October 2008 (UTC)
25. David Gerard (talk) 14:55, 20 October 2008 (UTC)
26. Rjd0060 (talk) 14:57, 20 October 2008 (UTC)
27. I find it extraordinary that the only reason I was aware of this proposal for an enormous change in policy was a post on WikiEN-l. I don't oppose in principle, but there needs to be a genuine consideration of the options and how best to pursue them. I am not convinced this is the answer. In fact, I am convinced that anything with "request for comment" as a required step is not the answer. Sam Korn ^(smoddy) 15:18, 20 October 2008 (UTC)

    There were notices at various central locations, and there is a proposal pending to announce this discussion at MediaWiki talk:Watchlist-details. Jehochman ^Talk 16:55, 20 October 2008 (UTC)

28. Clearly no consensus. --jpgordon^∇∆∇∆ 16:40, 20 October 2008 (UTC)
29. There are several different proposals on this talk page, some of which seem better than the proposal on the table. For example, I think Scott MacDonald's #Simpler sollution solves the problem in a less complicated way. ·:· Will Beback ·:· 18:55, 20 October 2008 (UTC)
30. Grsz^Review! 22:03, 20 October 2008 (UTC)
31. Yet to see a compelling argument that there are problems with the current system, let alone that this would be an improvement of any kind. Christopher Parham (talk) 00:02, 21 October 2008 (UTC)
32. Uh, whut? Aside from the lack of consensus, as I believe most users don't even know about this, there isn't exactly a great deal of faith in the RfC processes, at least not from me.--Tznkai (talk) 15:04, 21 October 2008 (UTC)
33. The proposal discourages desysopping the sysops; it delays the solution when it's needed. The sysop race enjoys too much immunity already, another perk is over the top. NVO (talk) 15:08, 21 October 2008 (UTC)
34. --PinchasC | £€åV€ m€ å m€§§åg€ 23:10, 22 October 2008 (UTC)
35. This proposal is too young yet for me to comfortably support. We should continue to develop and iron out everything, soliciting a yet wider pool of editors' opinions from which to evaluate consensus. The apparent surprise at this proposal suggests to me it is too early for implementation. Back to the drawing board for a while, I recommend, although I am loathe to oppose simply because this proposal has "not yet been around for long enough;" but I do think in this case that it's too early. Good work thus far, though. AGK 22:21, 23 October 2008 (UTC)
36. The proposal lacks consensus and should be marked rejected; however, I personally do support the concept. Xymmax ^{So let it be written} _{So let it be done} 14:48, 24 October 2008 (UTC)
37. I personally strongly support some proposal of this form, but more work and discussion is needed before the such a proposal can be implemented. But there's a definite kernel of a wonderful idea in this that should be pursued, ironed out. I'm convinced there will be a time when a proposal of this sort is implemented-- but not right now. --Alecmconroy (talk) 08:49, 25 October 2008 (UTC)
38. As is, but with a view to further discussion and proposals. LessHeard vanU (talk) 16:33, 26 October 2008 (UTC)
39. Stormie (talk) 02:02, 28 October 2008 (UTC)

Comments/Discussion

The concerns I raised were not answered. No agreement was reached on how 'consensus' would be defined or determined (beyond 'let the 'crats figure it out'). No pressing need for this process has been established. Unnecessary and unprecedented expansion of 'crat authority. No consensus resolution of these issues seems likely in the foreseeable future. Etc. TenOfAllTrades(talk) 15:37, 17 October 2008 (UTC)

To be fair, I haven't seen "consensus" defined or determined specifically for promoting; why should it be any different for demotion? ;) EVula // talk // ☯ // 19:30, 17 October 2008 (UTC)

Really? Go ahead—just try granting adminship to someone whose RfA has less than 70% support. I dare you. You get a fairly free hand between 70 and 75% or so, and then you're pretty much back to rubber stamping votes above 75%. Watch the screaming if you fail a candidate at 80%. While those bounds aren't codified anywhere (that I'm aware of), you'd have to be a lot dumber than we gave you credit for for you not to be aware of them. Nobody in the community gets overly bent out of shape if you flip a coin on a 70-75% promotion (as long as you don't admit to doing it). The rest of the time, barring extensive sockpuppetry or the very rare extenuating circumstance, the role the community has given you amounts to being a reassuring face on a vote-counting robot. (Sorry.)

For a Request for Desysopping (RfD), you'll almost never see one that isn't controversial. There will be deeply dug-in, experienced contributors on both sides. The evidence will be more complicated. There will be ridiculously hot tempers. The RfD proposed here isn't the simple thumbs-up or -down of RfA — this process will have six(!) different possible outcomes to choose from. Except in the cases where the RfD is obviously vexatious, I doubt you're ever going to see 70% of the community (or even 40%) agree on one of the options.

What does that make this process? WP:RFC/ADMIN. A prelude to Arbitation (since no firm decisions will come out of it, and 'no consensus' cases will just end up on RfArb) with the opportunity to abuse the 'crats thrown in. TenOfAllTrades(talk) 22:48, 17 October 2008 (UTC)

Don't worry, I'm exactly as dumb as everyone gave me credit for. ;)
The 70%/80% mark was never determined or written out anywhere; it was arrived at as per the community's wishes. If the de-adminship process is as clear-cut as RfAs can be (where a significant number don't require the coin flip; by the way, how'd you find out about that?), I think we can proceed in a similar manner. However, mine is just one opinion; I'd like to go ahead and see this move forward if only because I think it would address one of the more fundamental issues with the RfA process, where adminship is a neigh-unloseable "honor" so everyone gets their panties in a bunch over every little mistake a candidate makes. If we make it easier to remove the bit, the community may (again, my opinion only) be a lot more liberal with the bit granting, which I feel would be a net positive. EVula // talk // ☯ // 02:26, 18 October 2008 (UTC)

Extensive discussion, no consensus. Tom Harrison ^Talk 17:30, 17 October 2008 (UTC)

I'll just point out that believing it should be implemented and believing it should be tagged rejected are not mutually exclusive. I can believe something is a good idea, but if there isn't a consensus for it, then it has to be rejected for now, no matter what I think. --Alecmconroy (talk) 17:57, 17 October 2008 (UTC)

I can see how this can benefit the encyclopedia and there's really no downside. I'd be more for this proposal if it allowed bureaucrat's could remove bureaucrat flags too. -- penubag  (talk) 23:11, 17 October 2008 (UTC)

Note that if you want to implement this you need a supermajority to achieve project wide consensus. Spartaz ^Humbug! 08:29, 18 October 2008 (UTC)

Agree with Spartaz - there needs to be wide input from the community. Should we add a note to MediaWiki:Watchlist-details? PhilKnight (talk) 11:10, 18 October 2008 (UTC)

Yes, I'd support that but we need to get consensus somewhere (VP?) to do this and it looks too technical for me. Spartaz ^Humbug! 13:29, 18 October 2008 (UTC)

On a purely technical note, granting a particular usergroup the ability to remove a particular flag is trivial to implement. Bureaucrats being able to desysop (and decrat) is actually the MediaWiki default. Happy‑melon 18:00, 18 October 2008 (UTC)

I'm against linking a 240+ kb talk page on watchlists (not against linking the project page, though). — Athaenara ✉ 18:21, 18 October 2008 (UTC)

To anyone just joining our discussion already in progress, I'd point them to my above section where I ramble about all the reasons I support this proposal. --Alecmconroy (talk) 19:01, 18 October 2008 (UTC)

To anyone just joining our discussion already in progress, I'd point them to the entire discussion above. I hope that there aren't people voting without reviewing the existing discussion, or the reasoning behind, for, and against it. TenOfAllTrades(talk) 18:32, 19 October 2008 (UTC)

I would support this proposal, but only on the understanding (as the thread I beagn above) that current 'crats have to run a process / WP:Request For Desysop Approval / or whatever before being "white listed" to desysop. As this has not been fully discussed I'm hesitant to support the proposal at this time. Pedro :  Chat  21:31, 19 October 2008 (UTC)

I largely agree with this point of view; In essence, what those who support this proposal in its current form are saying is "I trust all the sitting 'Crat with this power". If that doesn't fly, I think the next step would be to do as you suggest-- go through them one by one and bestow the power if the consensus exists. --Alecmconroy (talk) 22:40, 19 October 2008 (UTC)

If you have consensus for that, fine. But some of us are opposed to it in any form.--Scott MacDonald (talk) 22:42, 19 October 2008 (UTC)

Of course-- no no, this is a big policy change and the burden is on we who support the change to generate a consensus for it. It may fly in it's current form. It might have to be altered as Pedro suggests or in some other way. Or, it may not fly in any form at this time. --Alecmconroy (talk) 22:53, 19 October 2008 (UTC)

I still don't support this. For my reasons, see Response to response to doubts above. Chick Bowen 00:36, 20 October 2008 (UTC)
In response to Pedro's vote up there: a straw poll is consensus, or at least should be. Tombomp (talk/contribs) 07:20, 20 October 2008 (UTC)

I think a close reading of the discussion shows no consensus. A straw pole is really another method of gaining the data to see if consensus has been achieved. My vote is simply a statement that whilst I generally like the proposal (with the fairly large reservation about 'crats having not been appointed on the basis of desysop tools as above) I also don't believe the discussion has elicited consensus. A list of names under two headings may or may not add clarity to this, but doesn't beat a reading of the debate IMHO. Therefore I have noted that whilst I may be in broad favour of the idea I am opposed at the same time, as there seems to be no community will. Apologies if I'm not really making my stance clear. Pedro :  Chat  07:27, 20 October 2008 (UTC)

I'm in a similar boat; I think the headings on the straw poll should have been differently labeled for just this reason-- whether you support the proposal and whether you think we've achieved consensus are two completely separate issues-- it's not clear which one this straw poll is trying to measure. I think Pedro and I are of the same basic opinion on these questions-- but I answered it as if it was trying to gauge personal support, so I listed myself in the support column, while Pedro answered it as if it were trying to assess whether consensus has been found, and so listed himself in the 'rejected' column. --Alecmconroy (talk) 09:48, 20 October 2008 (UTC)

Some of the votes are really strange - the poll isn't asking you if you think there is consensus, its asking you your opinion on the proposal to determine if there is consensus. Opposing it because you don't see a consensus from the discussion makes absolutely no sense. Avruch ^T 15:15, 20 October 2008 (UTC)

It makes perfect sense. The poll asks whether the proposal should be implemented or should be tagged as rejected. While some may personally agree with the proposal, they see a clear consensus on this page which says it should be tagged as rejected, thus, they oppose. - Rjd0060 (talk) 15:17, 20 October 2008 (UTC)

To back up what Rjd0060 said, the poll asks if the policy should be implemented or rejected. We implement or reject policy based on discussion, which generates a consensus. We don't "vote" to implement or reject a policy on its own merits, without accompanying discussion such votes are meaningless. I changed my vote because, upon reading the page, I interpret that too many people oppose, with reasonable justification, the proposal, and based on that it should not be implemented. I personally think the policy would be a good idea, but that's not what the straw poll is about. The implementation of the policy should lie outside of my personal opinion of it, and instead rest in the existance of consensus. --Jayron32.talk.contribs 15:48, 20 October 2008 (UTC)

I see Avruch's point - if people vote based on how they think others feel, or what they think the status quo is, and not what they want it to be, how does a poll ever determine if there is a popular consensus to update the status quo?

These "oppose because no consensus" votes are comparable to someone saying, "I really want McCain to be President, but so many people are voting for Obama, I think I should vote for him too." Does that make any sense? I hope it doesn't.

Now, on the other hand, if you're saying, "I like McCain's promise of not raising taxes, but there are other more important reasons to vote for Obama, so I will", that would be different. If Pedro, et al., are saying that they have a personal conflict about endorsing the proposal, then their "vote" makes sense, despite the way they've explained it.

There is another problem with the explanation of "votes based on perceived consensus." If everyone is supposed to vote only on what they think the consensus will be, how does any voter ever know what the consensus really is?

I think we should vote the outcomes we, after weighing the options ourselves, think best, and let "consensus" be determined once there is a suitable sample population of votes. Non Curat Lex (talk) 17:16, 20 October 2008 (UTC)

There is a fundemental difference here. There is already pages of text above that show how people feel towards this proposal, and there is not any consensus above for its implementation. we should not base policy on votes but on discussion. To short circuit the discussion by saying "ignore everything that everyone else wrote above, lets just base this on a vote" is what the straw poll essentially does. The analogy to a winner-take-all election like the American presidential election is not valid to this appliction. --Jayron32.talk.contribs 17:21, 20 October 2008 (UTC)

Indeed. Even if I supported the proposal (I do not), there obviously isn't consensus to make it policy and thus it should not be. — Coren ^(talk) 17:36, 20 October 2008 (UTC)

Again, it goes back to the object of the poll - which, to my mind and based on the prompt above the poll itself, is to figure out in an efficient manner what people think of the proposal. Nothing about the poll leads me to believe that the question is "What do you think of the discussion of the proposal." Clearly the wording leaves much to be desired, since its thrown so many people off - essentially we have people voting on separate questions. Maybe the best idea is to freeze it, rewrite the questions and start over. Avruch ^T 17:42, 20 October 2008 (UTC)

I'm not suprised, Avruch, that you have been unable to comprehend my comments as they are not well explained. However I think others have explained it better for me!. The community runs the project. My personal reading of this debate is no consensus. My personal opinion is I quite like the idea with reservations. The straw poll was a poor idea as it was ill defined on what we were really polling to be honest. That being said, I take my view we are poling the outcome of this page. In that case it really is easy - there seems to me no community will to support this proposal. Therefore I cannot support it. As an admin I am the servant of community will. My personal thoughts or aspirations are, by defenition, second to the way I should comment on a poll to change something quite fundamental to our community when that community does not want it. Pedro :  Chat  20:47, 20 October 2008 (UTC)

Voting based on your interpretation of consensus, instead of your support or opposition to the proposal itself, makes some sense if you pay attention only to the section headers - but the prompt above the poll makes it clear that its purpose is to determine consensus, not a consensus interpretation of consensus. See my point? Avruch ^T 17:39, 20 October 2008 (UTC)

Let me just add two things. First, I do not disagree with anyone who is saying that (a) there is a lack of consensus for THIS proposal and (b) that we do not make policy decisions by popularity contest, by by decision. I strongly agree with both statements.

My comments are limited to the more general subject of voting in general. Although Voting based on "perceived conensus" does not make sense to me. Although decisionmaking is not merely a product of popularity contests, sometimes it is necessary to gauge the popularity of a proposal, as in this straw poll. Raw approval scores also sometimes seem to be a factor in review of a discussion, such as for an RfA, where anything under 80% approval seems generally to be inadequate, regardless of the explanations attached. When such a vote presents itself, people need to be able to express their actual beliefs. If they are confident, they can explain it. If they are conflicted, they can explain it. Sometimes, it is tough - this proposal calls for community members to weigh competing interests and come to a conclusion; this may require "hard thinking." If it becomes popular for people to just respond based on "perceived consensus," what happens is that people can duck the hard thinking and abdicate to a "perceived consensus." Nevermind that "perceived consensus" is a perplexing, perhaps paradoxical phenomenon, it is is a poor paradigm, that will impoverish our voting by allowing people to give "votes" or comments that add little in the way of new thinking about tough or close issues to our votes.

Second, practically speaking, I would make it clear that a vote based on perceived consensus is the same as, or just slightly less weighty than a "per nom" or "per above" response (as I understand it, these types of responses are generally deemed disfavored, right?), when it comes to evaluating the outcome of a discussion. Non Curat Lex (talk) 21:01, 20 October 2008 (UTC)

Agree with many of the above. The poll is useful in that it attracted wider attention-- this proposal is a "big enough" change to our policy that it absolutely must have widespread attention. Even if they all agreed, 8 people are not sufficient to enact like this. Now that we have about 50 people commenting, we were at least starting to get into the right ballpark.

The other thing we can do with these eyeballs is start getting ideas for what sort of a community-removal-of-admin-rights would result in consensus. For some people, the answer will be "none", but I bet most respondents will say something like "Well, I could go for something like this, if ________ ". --Alecmconroy (talk) 02:32, 21 October 2008 (UTC)

Result

I think {{rejected}} has it. Stifle (talk) 09:13, 29 October 2008 (UTC)

The checklist for new proposals

Here you go: Wikipedia:De-adminship proposal checklist. Any proposal should be able to respond to these questions. TenOfAllTrades(talk) 18:25, 26 October 2008 (UTC)