Wikipedia talk:Request for Admin Sanctions

Since forming WP:WER, the most common concern I hear among the members has been to find a fair and responsive system of dealing with administrative sanctions. I wish I could take the credit, but this is nothing more than taking the best ideas of many participants, and trying to put it together into a system that uses existing infrastructure and doesn't require a new layer of bureaucracy.

Of course, only the Arbitration Committee has the power to desysop someone, and our goal is not to take that away, but instead empower them by doing some of the heavy lifting without taking away any authority. For admins, they need a system that will not allow for frivolous claims or burden them to the point they are unwilling to help in controversial areas. And of course, to the non-admins, it has to be a system that is accessible and gives them a voice. Once in a while, we need more than the binary options of "do nothing" or "take it all away", and this accomplishes that, while putting reasonable limits on the sanctions.

While no system is ever perfect or has everything that everyone wants, and I feel this proposal represents a consensus of the different ideas I have heard over the years, and is a great step in equality for all editors to participate in the process. It is flexible enough to change as we need it to, yet it is workable from day one as it is. It might not have every "feature" that everyone wants, but it can evolve over time and I truly think it is the best possible compromise that is fair to everyone, both admins and non-admins. I appreciate your consideration, and I thank you for your ideas. Dennis Brown - 2¢ © Join WER 21:39, 29 July 2012 (UTC)

This here is mostly Dennis's work, although I did give a hand with some editing for clarity and lightening up the process a bit, as well as advise him a bit on how to best weave ArbCom in (in my capacity as cynical ex-arb consultant-slash-grump). I think the general concept is quite workable, as it provides a venue to review administrators' behaviour if it becomes problematic without making an easy hate forum for airing random grievances.

I believe it has sufficient checks and balances to prevent it being abused, but remains lightweight enough to be considerably less onerous than having to poke ArbCom. It does give Bureaucrats a great deal of power at the close; but it's not unprecedented or unreasonable: the task remains gauging consensus, and it is exactly analogous to an admin closing a RFC/U for an editor. — Coren ^(talk) 21:18, 29 July 2012 (UTC)

Discussion

• Overall this is good but might the page be better titled Request for Administrator Conduct Review (I know its long but there's an pressumption of guilt in "request for sanctions"), . Also is there a better term than "keep" (ie "no action")? Otherwise I think this might work. The only thing I'd like to see included is that genuine dispute resolution attempts should have been tried (ie the sysop must have ignored outside input counseling them to alter their behaviour) before jumping straight to sanctions--Cailil ^talk 22:24, 29 July 2012 (UTC)

• I'm open to many points including the name, but I think it is assumed that other methods have tried and failed (ie: WP:AN), which is why it requires at least two admins to sign off. When it gets to this point, the parties are asking for sanctions of some type, even if only admonishment, not just discussion. This allows all sanctions in the grey area, which were always available but never used. And you are right, Keep might not be the right word. I might tweak that a bit. Dennis Brown - 2¢ © Join WER 22:31, 29 July 2012 (UTC)
• I changed to Affirm for now, but that isn't chiseled into stone. What makes me not want to use "no action" is that it might be the consensus but with an admonishment or still a sanction like "avoid deleting in CSD for 3 months" or "get mentoring on AFD closing". It is still a Keep/Affirm but not "no action". I want to keep the options and sanctions very open for the 'crat. But I am still open minded here, just my thought at this time. Dennis Brown - 2¢ © Join WER 22:35, 29 July 2012 (UTC)

• Just from the perspective of someone who works at sanction/enforcement related boards (WP:AE)"no action is warranted here" is regularly used in discussion. It'd be my first response to a good faith request that just doesn't deserve anything. I do see where you're coming from but again if the sysop is being admonished that's an action. Right now all the !Votes are for punishment of one kind or another there needs to some "reject request" option even after it has been validated (I know technically if it's been validated it shouldn't be groundless but I can still see situations where valid good faith requests don't require any action).
  Also, even though I see where you're both coming from, Fluffernutter has a point being blocked for administrative misconduct should be higher up the chain than desysoping. We shouldn't be blocking for breaches of 3RR or ArbCom Enforcement - we have processes for that. Any action (that I can imagine) an admin being blocked for here would have to involve deliberate misuse of the tools, not a mistake, not bad judgement, not an F*bomb. Blocks are supposed to be preventative, if we block an admin for abuse of the tools we're saying that they're unfit to wield them.--Cailil ^talk 22:58, 29 July 2012 (UTC)

• This isn't limited to just abuse of tools, although that is likely always going to be at least a part of it. The key is to not try to define every possible situation, and let the closing 'Crat have discretion as to the best solution. Since the bar is very high for 'Crats, and judging consensus is what they do, they would have a high degree of trust from the community to do this. Dennis Brown - 2¢ © Join WER 23:16, 29 July 2012 (UTC)

• Making no comment about the worthiness or lack thereof of this proposal in general for now, I gotta say...what universe do we live in where "desysop" is considered a "worse" and "stronger" sanction than "block"? It seems clear to me that a block is a much, much stronger sanction. A fluffernutter is a sandwich! (talk) 22:38, 29 July 2012 (UTC)

  Block duration is limited here, but losing the bit is forever. The goal was trying to give the 'crat flexibility in determining consensus, not to declare the order of severity. Dennis Brown - 2¢ © Join WER 22:45, 29 July 2012 (UTC)

  But the proposal actually says, and I quote (emphasis mine): "A Desysop vote being the strongest possible sanction, may also be considered a vote for the lesser sanctions of a Block or Suspend for the purpose of determining consensus." If an admin is getting blocked, I think it rather goes without saying that a desysop is in the cards, whereas a desysopping needn't, and usually doesn't, lead to a block. I think you're actually speaking from a position of the adminship-is-a-big-deal mindset that you're trying to address when you intimate that being demoted is worse than being kicked off your company's property entirely (to borrow Coren's analogy). A fluffernutter is a sandwich! (talk) 23:04, 29 July 2012 (UTC)

  Words can be changed. Many admins would consider being desysoped a much worse action than a 1 week block. Again, the key is giving flexibility. I might tweak that a bit. Dennis Brown - 2¢ © Join WER 23:10, 29 July 2012 (UTC)

  I agree that a decision to block or not isn't quite in the same spectrum as a decision to suspend or remove administrative privileges. Loss of administrator status can be triggered by evidence of a lack of judgment, whereas a decision to block generally is intended to prevent harm to Wikipedia's collaborative environment. While the two can of course be related, it isn't necessarily the case: an admin that made poor decisions in good faith may warrant a removal of administrative privileges, but not a block. Thus I think the decision to block should be handled separately from the decision to remove administrative privileges. isaacl (talk) 23:44, 29 July 2012 (UTC)

  For the same reason, I expect, that being fired is "stronger" than an unpaid suspension from your job would be. — Coren ^(talk) 22:46, 29 July 2012 (UTC)

  Per my comment above - perhaps the suspensed !vote would better suited to this purpose than a "block" !vote--Cailil ^talk 23:03, 29 July 2012 (UTC)

• I've modified the section of negative vote priorities, per Fluffernutter's concerns. I've also added the idea of "No Action" as a substitute for Affirm. Again, we aren't married into the details, we are here to iron them out. Dennis Brown - 2¢ © Join WER 23:13, 29 July 2012 (UTC)
• I definately like the idea in principle for several rasons.

1. If the community can be trusted enough to elect an admin then they should also be trusted enough to vote one out too if they start acting up.
2. I like the idea of adding another level of remediation before Arbcom. For reasons I won't get into here this to me is a great thing.
3. This is in line with Jimbo's ideas that it should be easier to become and admin and be easier to get the tools taken away if they do something they shouldn't.

• I would suggest a couple things though:

1. Allow a suspension of rights or just an admonishment as well as a complete revokation. I could see a number of occasions where it might be better to just give them a temporary break from admin duties rather than a straight demotion and it would also keep from having to go back through the RFA gauntlet again later.
2. We may want to restrict this to Editors who are Autoconformed or better, no IP's or brand spanking new accounts due to the types of actions that Admins perform. Kumioko (talk) 01:22, 30 July 2012 (UTC)

• You are right, but those options already exist in this semi-final version. There is a suspension of rights, up to 6 months, as the Suspend option. Admonishment is also possible as one of the Affirm options for the Bureaucrat, and he could even require a mentor or to avoid an area for a fixed time. It is very flexible. And it would use the same criteria as RfA, rather than inventing new ones, so that means autoconfirmed users only. That has always been one of my top goals, to offer options to the closing 'crat other than all or nothing. The goal isn't to lose admins, it is to have a fair system with options short of stripping the bit, to keep admins yet deal with problems. Dennis Brown - 2¢ © Join WER 01:31, 30 July 2012 (UTC)

Great to know thanks. I didn't see that but glad to hear its already in there. Kumioko (talk) 02:16, 30 July 2012 (UTC)

• I think we should maybe look at what could constitute several other standard options. I do like the Suspend one, but I;m also thinking of something like topic ban (i.e. an area or areas where an admin mustn't use their tools, while allowing them full use of the tools in areas where they haven't had problems). I can see immediately that many people will have the "Either we trust them or we don;t" reaction, but I'm thinking of Real Life™ parallels here: I know several people who I would trust with my life, but not with either my car or my credit card! Pesky (talk) 03:27, 30 July 2012 (UTC)
  • This system already allows for that, by design, even if closing as an Affirm, and it is at the discretion of the closing crat. They can impose any restriction they deem necessary, subject to review by ArbCom, who can unilaterally remove or add sanctions themselves without adding extra process. I absolutely agree that sometimes you might want to say "Bob, you need to stay out of CSD for a while, maybe get mentoring on delete policy" or similar, if his actions are causing problems. This is the key to this policy vs. others is that it isn't an all or nothing plan, and being flexible makes it more likely to be fair to the admin and the community. I did add a line to make that more clear: They may impose other restrictions or sanctions as the situation warrants. Dennis Brown - 2¢ © Join WER 11:20, 30 July 2012 (UTC)

"Involved" question

Note, I came here after seeing reference to it on Dennis Brown's talk page. I don't understand the idea that the RAS " requires certification is by 4 persons, 2 of which must be Administrators, none of which have been previously involved with the admin in question". Why would an editor with no previous involvement suddenly decide to open and certify a proceeding against another editor? This is not a requirement for RfC/U, RfA, Arbitration requests, or anything similar. I would think that, basically by definition, anyone who was interested and knowledgeable enough to certify such a Request would certainly already be involved (in both a "plain reading" and WP:INVOLVED sense). Am I misunderstanding, or does this need to be changed? Qwyrxian (talk) 23:20, 29 July 2012 (UTC)

• That was a tough one to address from the beginning. The goal is to prevent two former socks who were blocked from putting the admin up. That might need tweaking, but the goal is to prevent frivolous actions being started. Ideally, it should be outside observers who have seen the bad actions, but I'm open to any wording you can provide that will protect the admin from false or overstated claims and still allow fair participation. Again, the details are here to be ironed out, using suggestions from others. Dennis Brown - 2¢ © Join WER 23:26, 29 July 2012 (UTC)
  • Personally, I would have preferred verbiage along the lines of "who have not been directly affected by a sanction for that administrator in the past $INTERVAL". — Coren ^(talk) 23:31, 29 July 2012 (UTC)
    • Changed to and no one who has been directly affected by a sanction for that administrator in the past six months. but the six months would be purely a decision of consensus. That does explain it better. Dennis Brown - 2¢ © Join WER 01:18, 30 July 2012 (UTC)

question

Is this something that came from:

• Wikipedia:Requests for Comment/Community de-adminship proof of concept/Proposals? — Ched :  ?  00:28, 30 July 2012 (UTC)
  • Yes and no. I've actually been working this up for quiet some time, but only recently figured out what the consensus might be, and what might actually work without adding a new layer of bureaucracy. Dennis Brown - 2¢ © Join WER 01:15, 30 July 2012 (UTC)

Block as an option

Should the option for block be at the discussion, or should that a separate discretionary tool that is used outside this process? Dennis Brown - 2¢ © Join WER 02:10, 30 July 2012 (UTC)

Typo'

Probably, "Administrators accept that they are held to a higher standard of conduct than non-editors" should read "than non-admins" or "than other editors." --Anthonyhcole (talk) 03:26, 30 July 2012 (UTC)

Yes, that was clearly a mistake (I've fixed it). This is the kind of bizarre thing that gets left over when you tweak a specific wording over several revisions and you change the way a statement is framed.  :-) — Coren ^(talk) 03:53, 30 July 2012 (UTC)

Bureaucrats

Once certified, a single individual should not be able to vacate (end early) a discussion unilaterally. Even if the individual is a bureaucrat or an arbitrator. - jc37 05:22, 30 July 2012 (UTC)

I disagree. Bureaucrats have the highest standard for approval and should act as a circuit breaker when needed. This will stop (or actually prevent) frivolous claims, because they know it can be stopped in it's tracks, and the Crat stopping it has full authority to give sanctions to any bad faith nom. It is a summary dismissal with prejudice. Doing so will be a rare thing, but I think that option must remain, in a non-political and respected position and only 'Crats fill that high standard. No individual Arb can stop the process, per this policy, only the entire committee acting as one. This is to protect the admin from abuse and unnecessary and unfair process. Dennis Brown - 2¢ © Join WER 10:58, 30 July 2012 (UTC)

It's been stated many times that the trust for arbitrators is when in committee, not as particular individuals.

There is a LONG standing tradition that the only individual who can act in this way is User:Jimbo Wales, and he has increasing reduced his presence in this.

I understand that you're trying to use Bureaucrats as a gatekeeper, but that should be done through the certification process.

Perhaps if you rephrase it as "an early close". That way, arbcom review still applies. - jc37 16:18, 30 July 2012 (UTC)

ArbCom review always applies. This entire process is essentially a proxy for ArbCom, who has the power (as a committee) to start, stop, suspend, void and nullify it. Only ArbCom has the power to desysop, after all. Dennis Brown - 2¢ © Join WER 16:45, 30 July 2012 (UTC)

"Vacating the process" suggests that at that moment the process is stopped, and so arbcom would not be reviewing.

Also, the community is not and never should be considered a "proxy for arbcom". Arbcom serves the community (and the project thereof), not the other way round. - jc37 16:53, 30 July 2012 (UTC)

A bit nitpicky about my casual use of terms, aren't we? This would still be an ArbCom venue at it's heart, or at least under it's control. This just allows the community to play a role in the process and reduces the workload. Regardless of what this process comes out as, or if anything is vacated, Arb can completely set it aside and do as they please. Because desysop is an option and only ArbCom has that authority, it has to be 100% consistent with their obligations or it has no authority. We aren't trying to take anything away from ArbCom at all, and this actually offers them several advantages in dealing with problems. It is about filling the gap between a full Arb case, and nothing. Dennis Brown - 2¢ © Join WER 20:58, 30 July 2012 (UTC)

Might seem off the wall, and I know both proposals operate under this assumption, but is it necessary that arbcom have ultimate power in this venue? Is there something fundamental about arbcom that can't be circumvented by community consensus, or in the least, work parallel to it (without necessarily needing it to effectuate the community's conclusions)? Shadowjams (talk) 05:30, 31 July 2012 (UTC)

It isn't optional without a lot of policy changes that are never going to happen. Currently, only ArbCom has the authority to take the admin bit away (well, I think stewards can temporarily, and Jimmy can) and most people agree that this is a good thing. Additionally, there is no way that a proposal to remove that authority would get consensus. It would be purely an academic exercise, and after years of exercises, I'm ready for a real result. Often, it isn't about "the best proposal", it is about "the best proposal that will achieve consensus", and you iron out the details through the normal editorial process. In the past, there have been lots of proposals but no consensus. This is the idea here, about being flexible and fair enough that everyone can agree. Not perfect, but universal enough to get acceptance. Dennis Brown - 2¢ © Join WER 19:02, 31 July 2012 (UTC)

Entrusting with the tools

As this is essentially the reverse of an RfA, it should focus on usage of admin tools and the related responsibilities alone. This shouldn't be dealing with whether the individual should be blocked or topic banned. Those are things which assess the individual as an editor, not as an admin.

The question is simply:

• Does the community trust this individual with the tools and responsibilities of administrator?

It's the exact same question of RfA.

This should not be some type of editor review, which only applies to admins. - jc37 05:22, 30 July 2012 (UTC)

Again, I disgragree. This is not limited to use of the tools. The problem with other methods is that they don't look at the situation holistically. If I am telling someone they are topic banned at ANI, am I using the tools? But aren't I still acting as an admin? Being an admin is more than using the tools and the system must acknowledge that. You are stating a binary option, "tools or no tools". This system is designed specifically to NOT be binary, but instead flexible enough to deal with any situation. This is not a reverse RfA. We use similar methods because they work, more or less, but the end result may very well be to get the admin to take a short term break, or simple admonishment, or nothing at all. There must be more than binary options or it will not work. That said, that doesn't mean it has to have infinite options, and there are already limits in place. Dennis Brown - 2¢ © Join WER 11:01, 30 July 2012 (UTC)

I think Dennis the point myself Fluffnutter and Jc37 are making is this: the only reason one has to block a sysop here is for a breach of trust. Look at the list of those who have been de-sysoped. Abuse of the role (whether tools based or status based) has always resulted in de-sysop + sanction.
We should not be creating a board where any & every issue with a sysop can be raised. We have SPI, we have 3RR, we have WQA, AN/i & ArbCom - breaches in these categories are covered by emergency de-syoping - there's no need for this place to rubber stamp these or to bureaucratize them.
This board should only deal with things that call the overall trustworthiness of a sysop into question and that don't assess them with regard to editing (except where the two overlap ie WP:INVOLVED). I do understand why you don't want to get into unwieldy and unnecessary specifics, but this is necessary its the heart of the raison d'être of the board.
Furthermore it brings up the question of reasons to de-sysop and I'd suggest reviewing the de-sysop proposal checklist with regard to whether this board will consider de-sysoping "for a broader range of offenses" than has historically been the case (that might be thorny but as far as I can see is implicit in the creation of this board)--Cailil ^talk 14:59, 30 July 2012 (UTC)

As I've indicated below, removing the block option may be preferable, but at the same time I don't want the process to devolve into a binary situation. I would leave any blocks required to the discretion of the closing crat or other venues in that example. I may just go tweak that now. I have removed Block from the options as an official choice. Dennis Brown - 2¢ © Join WER 15:57, 30 July 2012 (UTC)

I pretty much agree with Calil's comments above.

Also, when not using the tools, admins are to be considered "just another editor". So to have any sorts of sanctions in an admin specific process, would create the idea that admins are not "just another editor". Or, another way to say it: any sanction or review that can be assessed of any editor, should not be a part of this process. So to reiterate, we should be dealing with the community's trust of entrusting with the tools. As Calil notes, there are many other, more appropriate, venues for dealing with editorial sanctions. This process should be kept focused. - jc37 16:18, 30 July 2012 (UTC)

Again, many "admin actions" don't require the tools themselves, such at imposing some type of topic ban or admonition. This shouldn't be for pure editor issues, I agree, but we can't limit it to tool usage, and it has to encompass any action that is "admin like", even if tools aren't used. Dennis Brown - 2¢ © Join WER 16:26, 30 July 2012 (UTC)

Closing discussions and assessing CSD? Deletion (and now move) closures have a review process. If the person shouldn't be closing discussions because the community doesn't trust their judgement (or bias) then the result is desysop. I'm trying to think of any other things that admins do that don't require tools. - jc37 16:35, 30 July 2012 (UTC)

Ok, I made a change that I think addresses the concerns here, three options basically Affirm (same as no action), Sanction and Desysop. Sanction meaning anything that isn't desysoping, but could be a topic ban, temporary bit stripping or other sanction that seems fit. Affirm can still be with admonition. Basically, it is a way for people to say "He shouldn't be getting off scot-free, but he shouldn't be stripped of the bit either", and keeps the process from devolving into a binary option. Dennis Brown - 2¢ © Join WER 16:31, 30 July 2012 (UTC)

If the goal is to keep it from being binary, then how about a "disapproval" section. "We don't like what happened, but we don't think it was enough to justify removal". Sort of like arbcom admonishment. - jc37 16:35, 30 July 2012 (UTC)

The "sanction" option covers that, and allows for more options. Dennis Brown - 2¢ © Join WER 16:44, 30 July 2012 (UTC)

Topic ban is an editorial sanction, not an admin one. Temporary desysop as a "sanction" is "punitive not preventative" and goes against the spirit of long standing Wikipedia policy. What other sanctions are you suggesting? - jc37 16:49, 30 July 2012 (UTC)

I disagree, which I guess you are used to by now, but it isn't a guess, it is based on real world experience here. My RfA is a good example of where I promised to stay out of CSD deletions for 3 months and get mentoring. No different than this. It wasn't punitive, it was the community requesting it to prevent mistakes. Telling an admin to not admin for 3 months and work as an editor isn't punitive, it is preventive. If they are a worthwhile admin, but are stressed to the point that it affects their judgement, then you are preventing damage yet addressing the real problem: stress. That doesn't mean they are an evil person, or incompetent, just that they need to step away for a bit. Even a good boss will send a stressed employee home for a day instead of firing them, so this isn't without precedent in the real world. Again, real world problems are not binary, nor should the solutions be. Dennis Brown - 2¢ © Join WER 16:58, 30 July 2012 (UTC)

That's what NOTNOW is used for. and suspension as you use the phrase is indeed a punitive action.

Back to trusting the community - Do you honestly believe that if someone is desysopped due to clearly being stressed or whatever, comes back in 4 months and asks for the tools back, while showing how they no longer fall under that situation they won't recieve them back easily? I can link to a bunch of RfAs where that was true. And most were drama-free.

I think the place where you and I disagree is that you see deadminship as a big deal. Something to be avoided at all costs. And it's not. Adminship is a toolbox we carry. If it's removed, we are still the same person, the same editor that we were while carrying the tools. - jc37 17:07, 30 July 2012 (UTC)

Deadminship is as big a deal as adminship is. Being one of the largest websites in the world, we are fooling ourselves if we think it is really "no big deal" like it was in 2002. If anything, this system makes it easier to make it less of a big deal, as currently, desysoping IS a big deal here, requiring a bunch of processes ending at ArbCom, many months later. THAT is a big deal. This, less so. Dennis Brown - 2¢ © Join WER 17:13, 30 July 2012 (UTC)

Continuance

It looks like the "continuance" option in the validation section is redundant to "reverse and remand". Is there something I'm missing? — Mr. Stradivarius ^{(have a chat)} 05:33, 30 July 2012 (UTC)

• Reverse and remand would send it back to the community. It will seldom be used, it is purely procedural, say if they think the issue should be "unclosed" and open for discussion another week, or other instructions. It is flexible. A continuance does not send it back to the community, and means they are keeping in active and under review at ArbCom, but need more time to decide. Rare, but some cases might be that complicated so they need the option as a way to tell the community that more time is needed. Dennis Brown - 2¢ © Join WER 11:04, 30 July 2012 (UTC)

Restrict

I've added the option to vote "restrict" at the RfAS discussion to the "process" section. It looked like leaving it out was an oversight, but feel free to change it back if it was intentional. — Mr. Stradivarius ^{(have a chat)} 05:36, 30 July 2012 (UTC)

• I'm wondering if we need to simplify all of it into "Affirm"(including with no action), "Sanction" (keep bit but with some restriction or admonishment, explaining in the !vote comment) and "Deysop". Just thinking out loud. Dennis Brown - 2¢ © Join WER 14:31, 30 July 2012 (UTC)

  A suggestion: Just affirm and desysop. Leave the "sanctions" (if any) for arbcom review (such as advised or admonished). Per my comments above, bans and blocks should not be directly part of this discussion. - jc37 16:18, 30 July 2012 (UTC)

Example page

I have created a rather absurd test/example page Wikipedia:Request for Admin Sanctions/test (with myself as the subject of sanctions) for the purpose of giving a visual aid, and to develop the format itself. Please feel free to modify as needed, it is a very rough start only. Dennis Brown - 2¢ © Join WER 16:12, 30 July 2012 (UTC)

Just read it.

If they need "time away from the tools", then the result should be desysop. They can always re-request the tools from the community. - jc37 16:24, 30 July 2012 (UTC)

Again, that is too harsh and narrow and doesn't fit every situation. Ask Bwilkins. The real world is not so binary. Dennis Brown - 2¢ © Join WER 16:43, 30 July 2012 (UTC)

If anything, allowing this to be open to any sort of proposed sanctions by the community would be "harsh". The benefit of the binary process is that it cuts out a lot of drama-filled nonsense. - jc37 16:46, 30 July 2012 (UTC)

Perhaps I have greater faith in the community and the closing Bureaucrat than you do. Dennis Brown - 2¢ © Join WER 16:52, 30 July 2012 (UTC)

One of the greatest strengths (and greatest banes) of the community is its variance of opinion. And this often comes out in what some may label as "drama". Anyone spending any amount of time at any of the other processes where editor behaviour is assessed should see this clearly. There were reasons why the arbitration committee was created. And (so far) this process seems to be ignoring that. - jc37 17:00, 30 July 2012 (UTC)

Not true Jc37. An ex-Arb was involved from the start to try to make it a 100% Arb verified and approved system, with them having 100% control over the entire process, and the power to stop it at any time. Not every case needs to be here, and I would expect that often enough, ArbCom will simply move the case to their arena, and they of course have the full authority to do so. This just gives them options, without adding new departments or extra bureaucracy. There is no "head" of RfAS, after all. Well, other than ArbCom itself. Dennis Brown - 2¢ © Join WER 17:09, 30 July 2012 (UTC)

The reason I italicised "committee" is that it's given responsibilities resides in it as a group, and not as any individual. Anyway, we're discussing this same concept throughout several threads. - jc37 17:17, 30 July 2012 (UTC)

Something that I like about it is the way the community discussion is divided into "No action", "Sanction", and "De-sysop" sections. I think that's, broadly speaking, the right way to go, at least for that part of it. In this case, "Sanction" means something less than, or other than, de-sysop. From what I remember from the last proposal, there is some merit to the argument that that "middle" category should also include "boomerang" sanctions directed at those who started the process, as well as some merit to the argument that doing so would stifle the process. Glass half-empty; glass half-full. --Tryptofish (talk) 18:28, 30 July 2012 (UTC)

Concur with Tryptofish but on reading again I believe "sanction" really is the wrong term (right idea) and is leading to a lot of confusion between what you envisage Dennis and what the rest of us understand as being reasons to de-sysop. Something like "probation" or "lesser measures" would be better (and agin must ONLY be related to the user's role as a sysop, whether tools are used or not, not as an editor). Again a name change is in my view necessary per WP:BEANS--Cailil ^talk 20:31, 30 July 2012 (UTC)

Interesting point. I freely admit that my terms and some of the details need polishing, even if I feel I have a good bead on the overall structure. I truly appreciate the constructive criticisms. As for a word beside Sanction, I would have to break out the thesaurus. In essence, we are saying the person doesn't need to be stripped of the bit, but is culpable in some misdeed, so Sanction does fit and is generic enough to not imply anything in particular, but there may be a better term. Would love to hear some ideas from others for it. Dennis Brown - 2¢ © Join WER 20:46, 30 July 2012 (UTC)

I've added some extra text, feel free to clean up any errors or lack of clarity. Dennis Brown - 2¢ © Join WER 21:27, 30 July 2012 (UTC)

Yes, Cailil is right about needing a better name for it. Maybe Sanctions other than de-sysop? Or, maybe better, Actions other than de-sysop? --Tryptofish (talk) 15:25, 31 July 2012 (UTC)

That is a pretty long name for a !vote, however. Other if we must, but I would try to stay away from !votes with so many words. I still lean to Sanction as I think it is the most appropriate, but I would leave that to a consensus, like all things. Dennis Brown - 2¢ © Join WER 18:10, 31 July 2012 (UTC)

Yes, you're right about it being long. Maybe: Other actions? --Tryptofish (talk) 20:26, 31 July 2012 (UTC)

I'm still chewing on the ideas. I love single, concise words when possible. Maybe Action or something that denotes movement. Still thinking. Dennis Brown - 2¢ © Join WER 00:13, 1 August 2012 (UTC)

Let me add one more note: If I was an admin at RfAS, I would want the word Sanction as an option, as I might be afraid that a milder word might look too "mild" and actually incur more desysop votes because people want sometime to happen that is more than a neutral action. In contrast, participants might be more likely to say Sanction if they just wanted a strong action but no desysop. If I were there, I wouldn't want the middle option to look so weak that no one thought it was strong enough, seriously, so there is some value in "truth in advertising". Not expecting to be in that position, but trying to look through the eyes of someone who was, that is what I see. Just another observation. But still trying to find other words. Dennis Brown - 2¢ © Join WER 00:23, 1 August 2012 (UTC)

I actually like the word "restriction" rather than "Sanction", simply because "Sanction" has maybe a bit too much of a slap-in-the-face feel about it. Restricting the use of admin tools to (for example) stop them using them in relation to a given subject area, or a given editor, or whatever, has more of a preventative feel about it. Like a topic ban or an interaction ban, but not as strong; it wouldn't stop them editing in that area or interacting with that editor, just stop them using the admin tools in those situations. I just have this "thing" about words, and "sanction" sounds too punitive, whereas "restriction" seems more preventative. It's like the difference between a block and a topic or interaction ban for non-admin editors; it still appreciates the good work they can do outside of their problem area. Pesky (talk) 03:14, 1 August 2012 (UTC)

That has a nicer ring to it and worth consideration. Dennis Brown - 2¢ © Join WER 04:54, 1 August 2012 (UTC)

No confidence

• Motion of no confidence

This is essentially what this page should be doing. A call for the question of confidence (trust) in the editor retaining adminship. - jc37 16:27, 30 July 2012 (UTC)

You and I just strongly disagree. There has to be a grey area here, and we have to recognize that sometimes people need to just step away. Being so binary is harsh and perhaps insensitive to the accused, and it doesn't reflect the reality that not all situations are black and white. There is no way I can support any system, by any proposer that is purely binary, and I'm very confident that most admins won't support it either, which is why we have nothing now. There has to be checks and balances and allowances that not all cases are black and white, as well as protection from frivolous claims. In the end, either we get a system that can be accepted by most people as fair and flexible, or we have what we have now, nothing. Dennis Brown - 2¢ © Join WER 16:40, 30 July 2012 (UTC)

Why is it "harsh"? The whole point of the certification process (and the tiers of: bureaucrats close followed by arbcom review) is to prevent such nominations. - jc37 16:44, 30 July 2012 (UTC)

It is harsh because of the sanctions, black and white. On and Off. Good and Bad. Sysop or Desysop. You keep seeing this as a process for desysoping, and that isn't what it is. It is a process for sanctions, and desysoping is just one of the options, and hopefully not the most common one used. The goal isn't to get rid of admins, it is to fix problems. Dennis Brown - 2¢ © Join WER 16:49, 30 July 2012 (UTC)

Then I will unfortunately oppose this, as then this is just creating a layer of bureaucracy where none is needed. If you want to create a sanctions process for all editors, that might be interesting to look at, but I think singling out admins for such a process is an incredibly bad idea. It sets up the idea that admins are not "any other editor", among many other things. - jc37 16:58, 30 July 2012 (UTC)

Admin are not "any other editor". They have special tools and responsibilities. Wikipedia is one of the largest websites in the world, and being an admin is a tremendous responsibility. We need a way to deal with problems that is both responsive and sensitive to the needs of the admin themselves. We don't need to draw a line in the sand and make the admin prove his worth, we need to accept that sometimes, the middle ground is the right solution, and HELP admins with fixable problems, to fix them. Otherwise, it can become a witchhunt and we lose very good admins, and others won't want to become an admin in the future. This is why I can't support binary processes that ignore real world issues. I'm sorry the process doesn't have your support, but I have to stick to what many, many people here have told me about a solution, and develop one that the entire community might actually accept. Otherwise, it is just an exercise, and I'm only interested in real results. Dennis Brown - 2¢ © Join WER 17:05, 30 July 2012 (UTC)

Well, wow.

Admins aren't "just another editor"? Sorry, but that's long standing policy.

Please take a read of WP:ADMIN, for just one example.

This has been re-affirmed MANY times by arbcom as well.

If there is any view which will sink this proposal, it's setting up the idea that admins are some special class of editor, and by extension, requiring some special type of process to sanction them.

They aren't, and there shouldn't be such an "exclusive" process. - jc37 17:13, 30 July 2012 (UTC)

WP:ADMINACCT makes it pretty clear that they are not "just another editor". They are an editor that has a higher degree of accountability and a larger tool set. They do have responsibilities that extend beyond that of "just another editor". Dennis Brown - 2¢ © Join WER 17:16, 30 July 2012 (UTC)

Then my apologies, but you mis-read.

Admins have extra accountability, not "higher" accountability. Due to having "extra" tools.

It says so in the very first sentence of WP:ADMINACCT

Read the history section of WP:ADMIN in particular.

This has long history, and has been reaffirmed many time by Jimbo Wales and arbcom. So this isn't merely my assertion. - jc37 17:25, 30 July 2012 (UTC)

Then it is only a matter of interpreting the sentence "just another editor", but as you have made it clear that you do not support this proposal, I'm not sure what purpose is served by laboring the details. Dennis Brown - 2¢ © Join WER 17:29, 30 July 2012 (UTC)

I support the idea of a deadminship process. I don't support any process which singles out admins for editorial sanction. I don't support any process which would seem to suggest that admins are some higher rank of editor.

With that in mind, perhaps the way forward would be for me to show a process I would support, similar to how you have. - jc37 17:34, 30 July 2012 (UTC)

To be clear, this process is not for sanctions of purely editorial functions. I thought I made that clear. It is for any admin function, whether or not it requires use of the tools. Not all admin "functions" require use of the tools, and excluding those actions that are clearly admin in nature wouldn't be fair. Dennis Brown - 2¢ © Join WER 17:42, 30 July 2012 (UTC)

(de-dent) quoting from the proposal:

• "...to consider sanctions against administrators who are claimed to have shown a pattern of misbehavior or abuse either in using the administrative tools and/or in purely editorial roles."

and/or in purely editorial roles - That would seem to be counter to what you just said.

Admins are admins because they have "extra" tools, and extra responsibilities related to those tools. That's all. No special status in the community, no "higher" ranking. - jc37 17:50, 30 July 2012 (UTC)

Changed to "who are claimed to have shown a pattern of misbehavior or abuse in performing administrative duties, with or without the use of admin tools." which is what I meant, and it was admittedly worded very badly. The key is to not limit it to tool usage. Purely editorial issues should be handled at ANI, that I agree with, ie: 3RR, incivility/NPA where it is clear it isn't as an admin, like on an article talk page. I don't argue that. Dennis Brown - 2¢ © Join WER 17:56, 30 July 2012 (UTC)

Oh, and you really must not know me, or you wouldn't be preaching to the choir about admins being "better than" anyone. A look through my talk page archives makes it pretty clear how I feel about that. Ask anyone. Dennis Brown - 2¢ © Join WER 18:02, 30 July 2012 (UTC)

Ummm, I am a bit confused here. Sure there is a policy that says 'admins are like any other editor' but it is wrongheaded and totally false in practice. I have seen several discusions where admins have been acting in ways that would/should lead to a normal user being blocked. In each of these instances people have said 'we do not block admins' or 'blocking admins is pointless because they can override the block' (If someone did that I think it would be prima facia evidence they should not have tools). So in short a class of users who are not subject to blocks and whose actions must rise to the level of an ArbCom case before they can be disiplined seems to be a special class of user to me. Am I somehow wrong in this view? Jbhunley (talk) 00:04, 1 August 2012 (UTC)

There are many editors (including admins) who say and/or do things which others may deem block-worthy. Spend any sort of time at WP:AN/I and I have little doubt you'll see examples where you thought a particular editor should have been blocked for their actions.

Anyway, yes. Admins are "just another editor". Their comments carry the same weight as anyone else's comments in a consensual discussion. And they can be blocked like any other editor. - jc37 00:25, 1 August 2012 (UTC)

As someone who has worked regularly at ANI for a great period of time, I can say that in theory, yes, but in practice it is seldom seen, until it gets to ArbCom. Again, a significant concern to many involved at WP:WER and myself as well. To say that there is not at least some degree of inequity is to either be blind to it, or to be fooling ourselves. Even an ugly little truth is a truth, nonetheless. Dennis Brown - 2¢ © Join WER 01:53, 1 August 2012 (UTC)

Immunity

• "Additionally, users are not eligible to certify an RfAS for an administrator who has issued a sanction that they have been directly affected by, for the six-month period after the sanction was put in place."

Coren, Sandstein, and I just closed an RfC concerning WP:V. WP:V directly affects every editor. So it looks like we would be immune to this process for the time being.

I'm being somewhat facetious here to point out that while I understand what the intention was, this restriction is incredibly subjective and so is problematic. - jc37 17:58, 30 July 2012 (UTC)

Closing WP:V isn't a "sanction", per usage at Wikipedia. It isn't a negative consequence imposed on someone. Dennis Brown - 2¢ © Join WER 18:00, 30 July 2012 (UTC)

lol As I said, I was being mildly facetious. Though that RfC could be considered an editing sanction of a kind. In this case it was a content-related RfC, rather than a directly behaviour-related RfC.

Regardless, the sentence I quoted would seem to be rather open to gaming. - jc37 18:23, 30 July 2012 (UTC)

Granted, those on the other side of consensus might see it as a sanction. If you are starting to warm back up to the proposal, you can always suggest actual improvements. Dennis Brown - 2¢ © Join WER 18:25, 30 July 2012 (UTC)

There may be some truth to that, even then, but the principle remains. If you can't find at least four unaffected editors who agree that the decision you feel has affected was wrong enough to consider sanctions, then it probably wasn't no matter how strongly you feel about it now. — Coren ^(talk) 19:37, 30 July 2012 (UTC)

For this to work certification must have a timelimit like RFC/U. Also will there be an option for 4 unaffected users to reject at this point?--Cailil ^talk 20:32, 30 July 2012 (UTC)

That is a good point. Throw some numbers out and lets see what it looks like. That probably isn't going to be a major problem, as they would likely have talked about it plenty, but I agree that it needs a time limit, if only for proper procedure. As for 4 to reject, I don't know. I have it so ArbCom (as a whole) or any 'Crat can stop it on a dime. I would think that would be enough since anyone one person or group can petition a Crat, but I'm open to hearing more on why 4 to override would be beneficial. Dennis Brown - 2¢ © Join WER 20:40, 30 July 2012 (UTC)

Fair enough re: reject. How about 48 hours like any RFC/U and yeah if the decision to come here was made at AN/i that should be ample time for good faith certification--Cailil ^talk 20:58, 30 July 2012 (UTC)

That sounds reasonable. Perhaps we should just copy the text and modify it slightly from RFC/U, instead of reinventing the wheel. Dennis Brown - 2¢ © Join WER 21:14, 30 July 2012 (UTC)

Wikipedia:Requests for removal of adminship

• Wikipedia:Requests for removal of adminship (talk)

Taking these discussions and past concerns of the bureaucrats into consideration, I cleaned up and streamlined the process, while removing editorial sanctions from this process, and replacing terms with those used by existing processes. Enjoy. - jc37 21:03, 30 July 2012 (UTC)

Centralize this discussion

This is a remarkable example of the hazards of decentralizing discussions. While I appreciate that this page has been written in good faith, it's not helpful to the community to have two competing proposed policies as well as an RFC on the same topic all at once. I urge the author to consider folding this into the current RFC rather than further dividing the community's ability to focus on the central concepts. Risker (talk) 04:26, 31 July 2012 (UTC)

I started a comparison, and right below it [l2] a debate thread at the RfC. I hope we use that space to start, as per Risker's concerns. Shadowjams (talk) 05:11, 31 July 2012 (UTC)

I respect your opinion Ann, but there are advantages to stand alone proposals as well, including more rapid development. Coren and I both felt it was appropriate to be bold by moving it direct into wikispace, and it can be discussed in other venues. It has already moved along fairly rapidly. Dennis Brown - 2¢ © Join WER 09:47, 31 July 2012 (UTC)

In fact, even this discussion about the need to centralize the discussion is taking place in two different places! I've commented in more detail at the other one, wherever that is. --Tryptofish (talk) 15:13, 31 July 2012 (UTC)

Just a thought: why not, in the process of centralising it, present the various options (if they're markedly different) as options, like the recent RfC on verifiability? Get some input on what might be the best process? Pesky (talk) 03:36, 1 August 2012 (UTC)

Certification requirements loophole

This is a small issue, and one that hopefully never comes up, but it's worth fixing now, early. That is, an admin who knew they were going to be put up for this process could simply sanction (temp block) everyone that was certifying them (including admins) and by the language of the proposal as it is right now, they could not be certified. We need to put in something to stop gaming the system. Frankly I don't think the "no sanctions from admin within last 6 months" piece is necessary; if a certifier had a clear grudge then that becomes part of the debate and the consensus process can deal with it appropriately. There are plenty of other protections against spurious nominations, including 2 admins, and tenured editors. Shadowjams (talk) 05:14, 31 July 2012 (UTC)

• What if we simply changed the wording to say "should", meaning the bureaucrat is free to consider that the process may be an overreaction to a recent event, thus he can shut it down if it clearly is, yet at the same time, it would allow bypassing it as a hard and fast requirement. ie: it is not optimal to have involved editors, but it is allowed. Dennis Brown - 2¢ © Join WER 11:19, 31 July 2012 (UTC)
• I've went ahead and made that change to "should", so it isn't strictly enforced, but is still discouraged. Usually, if it is worth the case being brought up, there will be 4 unaffected people who are willing, but there might be some unforeseen reason why that isn't the case, and the Bureaucrat can determine that. Dennis Brown - 2¢ © Join WER 15:45, 31 July 2012 (UTC)
  • I'm pretty damn sure that an admin that suddenly started blocking editors that were about to certify a request against them would be so egregious that their bits would be removed summarily within minutes. — Coren ^(talk) 18:27, 31 July 2012 (UTC)
    • "Their bits would be removed" ... ouch! That really does sound a bit harsh, lol! Pesky (talk) 03:19, 1 August 2012 (UTC)
      • Again this is covered via the existing emergency de-sysoping system already in policy. Anyway such radical misconduct would clearly be headed for ArbCom--Cailil ^talk 17:37, 3 August 2012 (UTC)
        • And some of us not blocked, like say arbitrators who have revealed their private e-mail addresses?, would almost certainly get e-mail from one or more of the blocked parties, which would raise obvious red flags. Having said that, it might not be a bad idea to somehow, somewhere make a clear statement that, if you think your personal block was in this sort of punative way, to publicize one e-mail account somewhere, like maybe the ArbCom mailing list, which would ensure that someone in a position to do something would definitely be aware of it. Maybe. but that address would likely get a lot of silly e-mails from vindictive people as well. John Carter (talk) 19:08, 3 August 2012 (UTC)
          • Seriously, if an admin went rogue and started blocking, any Steward or 'crat could de-bit them in a second as well. This is reaching into the .0001% range. If we have the other 99.999% covered, I feel pretty good about it. This is no different than if any admin went off his nut and starting blocking willy-nilly outside of the process, there are redundancies build into Wikipedia to deal with that kind of stuff, we don't have to labor it within this process. Dennis Brown - 2¢ © Join WER 20:25, 3 August 2012 (UTC)

Request for Admin Review

The various proposals past and present are a bit confusing, but this seems promising. However "Request for Admin Review" or similar would be a better name, to allow more room for, and emphasis on, admin learning from feedback given through the process. Sanctions are just one possible outcome, and the possibility of other useful ones shouldn't be downplayed by making them effectively a "fail" in relation to the Request for Admin Sanctions intended outcome. Rd232 ^talk 17:47, 31 July 2012 (UTC)

• I'm certainly not fixed on a name "A rose by any other name would smell as sweet", but in previous discussions, there were concerns about "Review" as there already is a review process for admins, and the purpose of this process is that at least 4 people are seeking sanctions, but I would love to hear ideas. Dennis Brown - 2¢ © Join WER 17:56, 31 July 2012 (UTC)
  • Wikipedia:Administrator review, what I was talking about. Dennis Brown - 2¢ © Join WER 00:15, 1 August 2012 (UTC)
• We have long established if perhaps underused processes of editor and admin review. While I think they are imperfect processes they are aptly named. Reusing that name for what are essentially recall petitions or confidence motions would exacerbate the unhealthy trend for us to obscure our processes in jargon. It would also mean that we needed to rename both the existing admin review process and also the editor review process. We really couldn't use review in one sense for editors and another for admins, if we did we'd be bound to have some people treating those who've had an editor review negatively on the grounds that "there's no smoke without a fire". I'd like to see both RFA and if we are foolish enough to introduce it any CDA become more like a review with an acceptance that we are all imperfect, and are here as volunteers not paid staff. But I don't see that such a change can be achieved by simply rebranding a negative and adversarial process. Too many RFAs and I suspect pretty much any CDA are unpleasant hazing ceremonies that bring out the worst in us, rebranding it as a review won't solve that. ϢereSpielChequers 14:21, 3 August 2012 (UTC)

Self-sanctioning admins

I note that the above possibility is not dealt with in the page. Honestly, I can and do think, in some situations, this can happen. I have myself said that I will desysop myself if some admin whose judgment in such matters I respect indicates I should. Possibly, some sort of Request for Sanctions might be the way someone I respect chooses to get that point across, by a request for sanctions including de-sysoping. If I were to receive such as the first request I have seen for my desysoping, I would, I presume, accept that as the indication of at least one admin whose judgment I respect indicating I should be desysop'ed. Granted, this is maybe a non-standard situation, but I can see how it might be possible that, when faced with good evidence in a request from individuals who might be perhaps more respectable than the editors with whom one has had a disagreement, an admin might admit error and agree to sanction themselves. If that were to happen, how would this process deal with it? John Carter (talk) 16:55, 1 August 2012 (UTC)

Excellent question :) This policy was designed to make this easy and fast. Any bureaucrat has the authority to shut down the process for any valid reason, per the policy itself. So if the admin says "You guys are right, I would voluntarily submit to a 3 month break from the bit.", and any Bureaucrat determines that this is consistent with the current consensus, he can (and likely should) close the process and implement the voluntary restriction. The bit stripping is temporary, but still "under a cloud" for the purposes of other policies, but it is agreed in this process that he WILL be given back the bit at the end of the three month period. So it is filed on Monday, he agrees to voluntary sanctions on Thurday afternoon, an hour later the process is closed as "Sanction (voluntary)", and life goes on, drama free and with a genuine consensus. This is one distinct advantage here, since Bureaucrats have the tools and authority to make the actual bit changes, and real experience to determine consensus in regards to admins.

On the other side of the coin, any Bureaucrat can stop the process 4 days in because "it is obvious there isn't enough evidence to show abuse and a consensus to this effect is not likely to happen" in the rare instances that this would be the most drama-free way to handle a situation, such as a giant misunderstanding. This does put some extra responsibility on the Bureaucrat, but it is 100% consistent with their regular roles here anyway, judging consensus as to the admin bit. Additionally, ArbCom can step in at any time and do the same thing (or overrule it), but giving this ability to any Bureaucrat makes the system more lightweight and responsive, less political, and much less drama-filled. Because they are Bureaucrats, it isn't very likely to be abused and ArbCom can overrule in an instant if it was. There are several layers of redundancy here, yet it is very simple and doesn't create any new departments or positions. It is simply a process.

It uses real consensus, it is flexible and fair to everyone, highly adaptable and responsive, has many ways to reduce the drama, and adds no new layers of bureaucracy. Wish I could say I thought up all this myself, but I can't. I am pretty convinced it is the most fair and drama-free system we could design. Dennis Brown - 2¢ © Join WER 17:54, 1 August 2012 (UTC)

• Update: I've added two words to the Modification of process section that make this more explicit, adding "or closing" as one of ways to close early. It was always the intent (and I agree it wouldn't be used that often), but it wasn't clear enough before. Dennis Brown - 2¢ © Join WER 18:14, 1 August 2012 (UTC)

Seven days.

I added Otherwise, the case should be open for discussion for seven days after being certified. in the Mod. section, as that was the obvious intent but had never been inserted or questioned. Dennis Brown - 2¢ © Join WER 22:03, 1 August 2012 (UTC)

Timing

I've had two RFAs and used to nominate other candidates. My advice to anyone running at RFA is to pick a week when your real life commitments will allow you to be on wiki a lot, especially in the first half of the week. An hour before and after work might not be enough to stay on top of a complex RFA. I also encourage people to save their RFA when it is ready to transclude, but not to transclude it until the start of a long session. RFAS is radically different from RFA as the timing of its start is in someone else's hands other than the subject of the RFAS. Our admins are unpaid volunteers and it is neither fair nor reasonable that they can at any time have a 7 day process launched against them, especially if the people starting that process are by definition disposed against them and may even be timing the RFAS for maximum inconvenience. A clearcut example would be to file it in the runup or during someone's exams. Arbcom may have flaws but it does recognise this issue and allows both for longer process overall and for reasonable requests for delay. Since Arbs are neutral whilst those filing an RFAS are anything but, there is some logic in a system that allows for parties to request delays and extra time, but even the Arbcom process has flaws, I've seen it used by third parties to fish for real life info and even to dispute that people have valid real life commitments that take precedence over a voluntary unpaid activity. Yes the consequent longer process is a reason why some people are impatient with Arbcom and want a quicker process. But there is a valid reason for that length of process. ϢereSpielChequers 09:46, 3 August 2012 (UTC)

• I agree that can be an issue, even at ArbCom. With RFAS, any Bureaucrat can suspend or extend the process unilaterally, and it doesn't have to be the closing 'crat. Perhaps that needs to be more explicit in the policy, but this gives them the opportunity to control the process greatly. Maybe the crat should be able to suspend and restart the same process, for those reasons. Keep in mind, I would expect that ArbCom will come in and consider this enough process that they will take the case themselves if it was overly complicated. The 7 days is the usual term simply because having it open ended would be problematic (dramah), but if the admin in question requested it stay open, it should stay open. The whole idea behind the 'crat control is to make it fair to the admin himself, and flexible enough to insure he has the chance to look at the claims without bludgeoning him. The Bureaucrat isn't the judge here, except maybe of consensus, but they are the one who makes sure the process is fair, and I think most people feel pretty comfortable with 'crats being trusted in that role. Dennis Brown - 2¢ © Join WER 12:07, 3 August 2012 (UTC)
• I added a new paragraph in Modifications that should be more clear, and gives the community the power to step in and stop the process as well. Fairness is key, and we do need to bend over backwards with the best of faith. Of course, this is subject to further modification here. Dennis Brown - 2¢ © Join WER 12:19, 3 August 2012 (UTC)
  • Thanks for your changes, they do go some way to addressing my concerns. However it still leaves the timing largely under the control of those preparing the case, and if there is any delay then there is a risk of others piling into the draft or its talkpage - something we've occasionally seen with draft RFAs. If you are going to have a CDA process then my suggestion would be that the start date and time is agreed between the subject and an uninvolved crat, with said crat having the discretion to require the admin in question to to avoid certain topics or tool use until the end of the RFA. ϢereSpielChequers 12:43, 3 August 2012 (UTC)
    • Have to think about how to deal with that. I think that as long as the 'crat has broad power to control the time, the timing issue is 98% solved, and even the community has the power to indef. suspend the event in an unusual case (death in family, illness, etc.) and there is never a way to get it to 100%. As to getting the admin to agree to the time, it is hard to do, as they could just stall and stall. I'm open to ideas, but I can't think of any off the top of my head better than letting any 'crat decide. That is the key here, multiple 'crats can actually be involved, no single one is tied to the process. One might delay 1 week, yet another closes it. And if I were the crat and I decided to temporarily suspend the process for a 1 week, I would full protect the page and talk page automatically. That would prevent any piling on. Dennis Brown - 2¢ © Join WER 14:39, 3 August 2012 (UTC)
    • I added a note about full protection and crats. They were already authorized to do so, but this just gives guidance. As for semi-protection, that should already be understood that any admin can do that if it is needed due to socking or vandalism. Technically, any admin can full protect the page as well if the process was suspended, and nothing here prevents that. Dennis Brown - 2¢ © Join WER 14:44, 3 August 2012 (UTC)
      • The difficulty is that one person's stalling could be another's genuine real life issue, and without forcing people to disclose personal information we can't tell the difference. To make this truly an RFA in reverse you really do need to put the timing into the hands of the subject of the RFAS. Now if you have someone who has been hyper active for months and they offer you a date three weeks away for the RFAS people might get a little narked, especially if they continue to be very active for that time. But we have two safeguards, firstly people who stall but continue to be very active risk being crucified when the RFAS finally starts, and secondly if you have a neutral crat in the process they could impose reasonable limitations for the interim period. So if people are unhappy about an admin's out of process restores the crat could decide that the admin shouldn't restore pages until after the RFAS. Alternatively you could just give up on the idea of making CDA be simply RFA in reverse. ϢereSpielChequers 23:44, 4 August 2012 (UTC)
        • Well, it isn't an RfA in reverse, I've tried to make that clear. It isn't a vote of confidence, it isn't about people's opinions of the admin in general, it is a review of specific behavior only. It uses a similar structure only because the structure is familiar and easily understandable. If someone goes to ArbCom, they don't make an appointment. If they say they can't this week, but keep editing, obviously it was a bad faith delay so you go ahead and start. I don't worry so much about some of these at we are smart enough as a community to know when our chain is getting yanked. You can't make anything 100% perfect, so you do the next best thing. We can work with the admin, but we can't let them dictate the time table. Again, no different than an ArbCom case. Dennis Brown - 2¢ © Join WER 23:58, 4 August 2012 (UTC)
          • Well its a quarter of the default length of an Arbcom case, and lacks the safeguard of allowing the parties to request extensions that the Arbcom procedure has - not that they necessarily follow that procedure. So I would say that the timing issue makes it substantially different to an Arbcom case, and heavily stacked against those people who edit one evening a week. This rather brings into question the purpose of CDA, if you want a CDA because Arbcom is flawed then the ambition should be to come up with something better than Arbcom. What is the point of designing something worse? ϢereSpielChequers 08:04, 5 August 2012 (UTC)
            • No one has claimed that perceived Arbcom flaws are the rationale for this process. This would be a function of Arbcom, not a way around it. This is also much deeper than a simple deadminship, as the real goals isn't to strip the bit (a regular arbcom case would be preferred for very obvious cases) but is to instead find solutions to admin issues, with bit stripping being only the most extreme outcome. Dennis Brown - 2¢ © Join WER 09:06, 7 August 2012 (UTC)

Process

It's looking really great overall, nice work. However, I disagree with the requirement of two certifying admins.

1. It is highly unlikely that any user could get two admins to certify except in the most egregious circumstances.
2. I think it could be damaging to the admin culture to force them to take such strong stands against fellow admins.

We don't need an admin to nom for RfA, you can even self-nom. Nor are we required by policy to earn even a single admin's support at an RfA, so we shouldn't need an admin, let alone two admins to nom for RFAS, IMO. You would be better to require 5 users in good standing than 2 admins and 2 non-admins. Any thoughts? Why am I wrong? ~ GabeMc ^{(talk|contribs)} 03:29, 7 August 2012 (UTC)

Because almost all the people involved in drafting this are admins? --Demiurge1000 (talk) 04:16, 7 August 2012 (UTC)

Right, I guess now we'll find out if they are willing to consider suggestions from non-admins. ~ GabeMc ^{(talk|contribs)} 04:26, 7 August 2012 (UTC)

That sounds a bit like you are assuming bad faith Demiurge, and honestly, I would expect more from you. The bar has to be high enough to prevent frivolous claims, yet reasonable enough to be useful. It only requires that half the certifying persons are admins, which isn't nearly as difficult to get in an obvious case as you would think. Dennis Brown - 2¢ © Join WER 09:03, 7 August 2012 (UTC)

My points here still apply. If the case is such that it would be easy to get two admins to certify, then I suggest that case would be appropriate at RFC/U. Its my understanding that RFAS is being suggested in lieu of RFC/U. Also, Dennis what about point number 2? Do we really want a situation where only admins can certify the process of sanctioning other admins? Yeah, I know, right now its two and two, but that still means there will be no certification without admin support. Do we require admins to get 2 certifications from non-admins when they indef a user? No, we do not. This issue would seem to suggest more double standards between admins and non-admins. Further, "The bar has to be high enough to prevent frivolous claims" would seem to suggest you are ABF here as well. Why would we assume that non-admins would file "frivolous" claims, but admins would not? Are admins incapable of poor judgement? ~ GabeMc ^{(talk|contribs)} 22:13, 7 August 2012 (UTC)

Also, I think if this process is set-up properly then "frivolous claims" would lead to a WP:SNOWBALL, wouldn't they? ~ GabeMc ^{(talk|contribs)} 22:16, 7 August 2012 (UTC)

We indef non-admins at AN/I without requiring any admin support, why should certifing a RFAS require two admins? As I said above, frivolous claims would quickly snowball and end up affirming the good standing of the admin, so I'm not sure what great harm would come from a few frivolous claims. Do we require an admin to certify an AN/I report? Why not? ~ GabeMc ^{(talk|contribs)} 02:15, 10 August 2012 (UTC)

Admins can currently be indef'ed at ANI, or without ANI, or by a steward, bureaucrat or another admin. Yes, an admin can indef another admin. This is only to deal with admin related issues, not editor conduct. It is in addition to, not instead of, and only for admin related activities. Most admins don't get indef'ed or block, likely in part due to the vetting process to become an admin. And by design, RFC/U can NOT end with sanctions, regardless of whether the U is an admin or not. Dennis Brown - 2¢ © Join WER 02:19, 10 August 2012 (UTC)

Unless I missed it, I'm not sure you have addressed the issue of a double standard. Why can users be indeffed at AN/I without prior admin certification but admins (as this RFAS currently reads) cannot be put in any jeopardy of sanctions via this process without two admins to certify the RFAS? ~ GabeMc ^{(talk|contribs)} 02:31, 10 August 2012 (UTC)

Dennis I think this is a great idea, and is exactly how I think it should work. Originally I was supportive of the two admin certification, however reading the above comments I'm inclined to agree with GabeMc. What about, instead of having to have two admins (plus the other two) the standard is (from the second sentence in Process):

"Two of these users must be administrators or be non-administrators with no less than 8000 edtis, and all certifiers must be in good standing (Bureaucrats have discretion in defining what is and what is not good standing) and have had a registered user account for a period of no less than one year."

The 8000 edits can be moved up or down I was just using it as an example. This way the standard is a little lower and doesn't require two admins against another, but still requires two editors with some experience. Since there is still a requirement to produce evidence within 48 hours and that ArbCom or a crat can stop it at any time. I have also added in a quick statement on crats deciding good standing and not - to avoid any issues which may be brought up.

Also I think it would be helpful if "prejudice" where defined a little more exactly. Callanecc (talk • contribs • logs) 17:33, 11 August 2012 (UTC)

I don't know. I would tend to let that be up to the consensus here. One of the advantages of requiring it to be admins is that there is a degree of vetting involved in becoming an admin, and not all 8k editors are equal, so not sure how that will be taken as a requirement. I'm not opposed to modifying the standard, but not sure if that is the best line in the sand. I would have to hear how others feel about that. Part of requiring admins is you can assume they have a higher degree of understanding of the policies than "any random editor", and for this plan to work, we need admins themselves to agree and it offers a higher degree of protection from bad claims. I think the idea that "an admin won't certify against another admin" is overblown and not true. Some may be, but most admins do not want others to be acting in bad ways either, as it makes our jobs more difficult. If an admin is to the point that he needs sanctions, I'm pretty sure it wouldn't be hard to get four other admins to sign off, so two isn't that high of a standard. Dennis Brown - 2¢ © Join WER 14:41, 13 August 2012 (UTC)

I've got to say, I don't like the idea of admins being the gatekeepers to removal of adminship. I've been involved in the last couple of recalls on the encyclopedia, and watched the one prior to that. As such, I know that admins will step up and state that other admins should step down, but there is a perception out there that "admins protect admins". All unblocked editors should have equal voice in this community. There's no point in having this harder to access than an RfC - which requires 2 certifiers who have tried to address the issue before hand. None of this good standing editor with a long tenure. I've seen unblocked editors who've been around for years with thousands of edits behaving like ruddy idiots, and I've seen editos with 100 edits talking a lot of a sense after a week. The only time period I might accept is that the account should be registered prior to the incident occuring. Worm^TT(talk) 14:59, 13 August 2012 (UTC)

Agree with WTT. The process, if it's restricted to admins as gatekeepers, will be seen as biased or unfair. It may not be either, but limiting participation in that way will create the impression that it is, and once that happens... Intothatdarkness 21:14, 13 August 2012 (UTC)

Then what would you propose? I think 2 editors is too low, as we also don't want the process to be too easy that any disagreement turns into a call for sanctions. That isn't going to be accepted by the community either, nor would it be fair to admins, who often have to do things that are not popular and can't have the threat of having their time eaten up for simply doing the right thing. Even if the process shows they purely acted proper, it is stressful and wastes time if it isn't a real issue, so the bar must be high enough to be reasonable for both sides of the issue. Dennis Brown - 2¢ © Join WER 16:05, 16 August 2012 (UTC)

From a "stress level" perspective, this process would not be any more stressful than an RfC... I would suggest that we should follow suit. Two editors who have attempted to discuss the matter with the admin. If you like, one must be uninvolved, in a similar sense to INVOLVED, but that should be defined in the process. If you are worried about bogus reports, have a snow close clause. The community can see off a bogus issue easily. Worm^TT(talk) 16:42, 16 August 2012 (UTC)

Clearly the bar needs to be high enough to be reasonably. I'm just pointing out that if it's an admin-only process it will be viewed with some skepticism by a percentage of the non-admin community. If that's considered acceptable, then go with it. WTT's thoughts above are good ones. I'm not wedded to any specific ratio of admins to non-admins or panel numbers, but there should be balance and as much fairness as can be mustered. Intothatdarkness 16:12, 17 August 2012 (UTC)

Stay with four editors, but one or two must be uninvolved (as Worm said)> Regarding a snowball clause there already are a number of provisions there to prevent bogus claims: evidence has to be submitted within 48 hours, and the community for an arb or crat can end the discussion at anytime. Callanecc (talk • contribs • logs) 12:40, 18 August 2012 (UTC)

• Also, to include uninvolved users would require some talk page canvassing would it not? If an editor needs two admins to certify, then they will be shopping this around to 20+ talk pages, creating debate/discussion/arguments all over talk, and presumably a number of people would insinuate that the candidate for sanctions needs to be notified of each such talk page thread. What am I missing here? Any thoughts? ~ GabeMc ^{(talk|contribs)} 21:57, 25 August 2012 (UTC)

Has this effort officially lost it's steam? ~ GabeMc ^{(talk|contribs)} 07:39, 2 September 2012 (UTC)

• I haven't posted in a while. Many, many people showed an interest in something exactly like this, until it was on the table. Once there were a couple of actual proposals to choose from, the silence was deafening. All I can conclude is that the community is not ready for this type of action yet. Dennis Brown - 2¢ © Join WER 22:44, 2 October 2012 (UTC)

  I found exactly the same with the proof of concept. Worm^TT(talk) 07:37, 3 October 2012 (UTC)

  I don't see it as a failure, but I do see it as a clear indication that people are not wanting change, and the relative apathy and lack of participation at the larger proof of concept discussion, as well as this and Jc37's proposal (both of which were reasonably distinct and fairly well developed) is a clear indication of having no consensus for any kind of change. In the marketing business I sometimes find the same issue, people speaking out about wanting something that doesn't exist, but when it becomes available they are no where to be found. Dennis Brown - 2¢ © Join WER 13:47, 3 October 2012 (UTC)

I only just heard of this for the first time. Dennis, you should try publicizing it again. — Hex (❝?!❞) 17:51, 27 November 2012 (UTC)

Perhaps in time. Lots of lessons were learned along the way, but at this time there isn't enough support to implement the changes. In the meantime, many of us are working on other admin related issues, such as RfA reform and similar, with the same goal of making adminship closer to the non-admin community. Dennis Brown - 2¢ © Join WER 14:21, 28 November 2012 (UTC)