Wikipedia talk:RfA reform 2011/Sysop on request

WP:ALTRFA

Content of the above proposal was merged into this discussion. My76Strat (talk) 02:30, 2 September 2011 (UTC)

RfA reform

The reason I removed the Rfa reform transclusion is because this proposal is actually totally separate and needs to be developed as an alternative. My76Strat (talk) 03:32, 2 September 2011 (UTC)

am putting that back - it is part of that reform process in a peripheral way, which is clearly identified as such in the transclusion - the idea is not to orphan this proposal.--Cerejota (talk) 03:34, 2 September 2011 (UTC)

That's fine, should it be so prominent? My76Strat (talk) 03:49, 2 September 2011 (UTC)

Its the standard way to do that type of thing. Have you ever been in a meta-process? Its pretty hard to follow and messy. :P--Cerejota (talk) 03:53, 2 September 2011 (UTC)

No I yield to you in that regard. Also I anticipate many good comments will come to suggest improving this draft so I am open to reason. I am keen on following the guidance set fort by Jimbo in the linked discussion. Is there any part of the merge that you disagree with other than the transclusion? My76Strat (talk) 04:01, 2 September 2011 (UTC)

This seems like a great idea. The current system is largely broken, so I think this would be a welcome alternative. Lately, I've had a very hard time convincing anyone to run for adminship, even people who have been active for years. They just don't want to go through the spanish-inquisition-like process. Kaldari (talk) 04:11, 2 September 2011 (UTC)

I agree, please stay involved, we have a lot of momentum. Jimbo is not only behind this idea, he is in front of it, we just need to develop the most thoughtful alternative which we can devise. I am very optimistic. My76Strat (talk) 04:15, 2 September 2011 (UTC)

(edit conflict)The spanish inquisition was nicer - you could do an auto da fe and be saved... in an RfA all the skeletons come out and then are seen as a license to pile-on grievances old and new, with no possibility of repentance or conversion.--Cerejota (talk) 04:17, 2 September 2011 (UTC)

Interim period

The proposal talks about a period of time and/or log actions. We need to define that period. I suggest 6 months and 1000 log actions. Any other suggestions? My76Strat (talk) 04:35, 2 September 2011 (UTC)

I think that would be ok. Enough to show you know how to mop, and enough to show you mop right.--Cerejota (talk) 06:20, 2 September 2011 (UTC)

Comments

Note: Please remember this section is for discussion of a proposal, not an actual proposal. There's no need to support or oppose, but any suggestions would be welcome. Worm^TT · (talk) 09:55, 2 September 2011 (UTC)

I do not support this proposal for these reasons:

• Creating another userright only fuels the myriad of rights-eager newbies that seem to have increased of late.
• I certainly do not want these "admins-in-training" to be able to see deleted material. There's a reason it was deleted in the first place, and if the community doesn't trust someone enough to pass him/her at RfA, then they shouldn't be able to see it.
• Being an admin is not difficult. It's not like we really need to have apprentices for it. I mean, if you can't figure out the basics of the admin tools in a short time, then you shouldn't be an admin. Babying people with this intermediate period is only going to increase bureaucratic nonsense when we start arguing over whether an apprentice did something right or not, etc.
• Why is this needed, again? Because there are very few users who would meet the criteria, have a desire to be an admin, and be approved, but not be ready for RfA for some reason.

/ƒETCHCOMMS/ 05:06, 2 September 2011 (UTC)

Thanks for your input. I think we can accommodate many of your good suggestions. The limited tool set is not defined so perhaps we should shuffle some of the suggested elements. If you are more concerned about their ability to see deleted material, we should move that out of their purview. Need is subjective and I can't express need in terms relevant to your perspective. I believe the need is sufficient. I suppose that remains to be seen. Nevertheless I would like to incorporate as many reasonable suggestions as possible, and I hope you will become an advocate. Again, thanks. My76Strat (talk) 05:18, 2 September 2011 (UTC)

(edit conflict)Thank you for your comments. Your position is valid, even if I disagree with some of the assertions. However, some of your comments lead me to believe clarification is in order: this process will only leave RfA (in whatever form it exists) for appeals, re-sysop, and other such escalations, and for those seeking adminship before they meet the criteria. This actually creates a path to admin that eliminates RfA as the prefered path to admin - as such, those who request this path are considered to be trusted by the community, or rather, this trust is delegated to Bureaucrats as per a set criteria.

To address your points:

Creating another user-right only fuels the myriad of rights-eager newbies that seem to have increased of late. - if you read the proposal, it specifically only allows this useright only to users who have made 6,000 edits over two years (what the service awards call an "experienced editor"). I think such an user has demonstrated a commitment to the community that goes beyond "rights-eager newbies". And in fact, there are pleanty of cases of "rights-eager newbies" getting adminship (specially in 2007-2008) via RfA. So as an objection, it is also an objection for criteria-less RfA (which is what we have today).

I certainly do not want these "admins-in-training" to be able to see deleted material. There's a reason it was deleted in the first place, and if the community doesn't trust someone enough to pass him/her at RfA, then they shouldn't be able to see it. the proposal is open to such calibration of what rights are actually given - however, as deletion is mostly not a result of a dispute resolution process but of CSD, I think this right is important in training an apprentice admin in being a janitor - this toolset is also essential in fixing copy-and-paste moves, fixing redirects, etc - of which there is a significant backlog. One point you seem to miss, however, is that these "admins-in-training" under the proposal would have the implicit trust of the community, they are in a path to adminship, simply being tested out in a limited capacity to show being worthy of more trust.

Being an admin is not difficult. It's not like we really need to have apprentices for it. I mean, if you can't figure out the basics of the admin tools in a short time, then you shouldn't be an admin. Babying people with this intermediate period is only going to increase bureaucratic nonsense when we start arguing over whether an apprentice did something right or not, etc. speaking for myself as proposer (there is another proposer who might or might not share this view), I agree with you completely both in the technical issues (MediaWiki is very easy to use), and the bureaucratic squabbles that might result from supervision of apprentices. However, it is proposed that it be easy to remove an apprenticeship, easy to appeal to RfA, and largely up to Bureaucrat discretion to handle issues with tools. The side-stepping of the RfA as a process for adminship will actually reduce bureaucratic squabbling - particularly, the broken RfA process will be deprecated, and ArbCom will not handle apprentice tools issues. That will remove a lot of the bureaucratic squabbling around The Mop. The issue of community trust then transfers to Bureaucrat selection.

Why is this needed, again? Because there are very few users who would meet the criteria, have a desire to be an admin, and be approved, but not be ready for RfA for some reason. While the reasons are a subject of contention I won't cover here, and your point of view is held by some in the community, the objective data shows a dramatic drop in open RfAs (93% in 3-4 years) with a stable decline rate of ~60% - this means that while the success of the candidates has not dropped or increased, the amount of candidates has. This is understood by many, including Jimmy Wales, to mean there is a problem with the RfA process. Many also feel that there is crisis in terms of available admins - and certainly the lack of new successful RfAs will lead to longer-term issues with attrition, as the rate of desysop and retirement exceeds the rate of new admins. This proposal is a radical proposal to address these issues by focusing on the real problem: no one wants to run the gaunlet RfA has become - even editors who should be admins and who would probably get a 90%-10% result. While other proposals try to "fix" the RfA process (and I am not against that), I believe that even a "fixed" process would fail to address the real problem: adminship has become a big deal, rather than a mop, and that making it a right acquired by objective criteria rather than subjective "trust" and by demonstrated behavior rather than a popularity contest, the mop can be returned to what it should be, a way to improve the encyclopedia, not a merit badge. --Cerejota (talk) 06:18, 2 September 2011 (UTC)

• I disagree with the premise that 6,000 edits/two years means a user isn't rights-eager. I certainly know of a couple who are. In fact, I would be highly interested in someone making a db query to list all qualified users who have made at least one hundred edits in the past 30 days.
• I don't have a problem with deleting pages. I do have a problem with undeleting them and viewing deleted material. And if I trust someone enough to view deleted pages, I trust them to block and protect and everything else, too.
• I think that only increases the likelihood of a squabble. It's also easy to block users, appeal blocks, and review them at an admin's discretion. But that's filled with drama. People think that even semi-admin tools are a status symbol, inherently. So if they're removed involuntarily, there's going to be drama. Just how every online community works. If you give them the opportunity to become "elevated", people will grab for it.
• I disagree that the real problem is people not wanting to run for RfA. The problem is twofold: some nasty and stupid people commenting in RfAs, and the candidates taking it way too personally. Admins get harassed all the time, and if someone is going to pull the "RfA made me sad!" card, it better be for something pretty bad. Adminship is very much about taking the heat, the criticism, the constant scrutiny of your work. If you want to change RfA, change that first, because making it easier to get the tools doesn't make it easier to explain why you used them to do something.

/ƒETCHCOMMS/ 14:28, 2 September 2011 (UTC)

• Oppose, this is an "Admin light" proposal which was turned down many times before, with many good reasons. I would support Sysop on request for full admin responsibilities, though. --Pgallert (talk) 09:51, 2 September 2011 (UTC)
• Fully agree with Pgallert's comment above. I was disappointed to see that this was not a simple proposal for 'adminship per request' Jebus989^✰ 09:59, 2 September 2011 (UTC)
  • Uh, "adminship per request"...I think I've heard of this before, it sounds awfully similar. Steven Zhang ^{The clock is ticking....} 10:45, 2 September 2011 (UTC)
    • On the first preview I wrote 'adminship per simple request' but then realised I had already used simple in the same clause. Nevertheless, I assume my meaning is entirely obvious Jebus989^✰ 15:35, 2 September 2011 (UTC)

Speaking as proposer, this is not admin light at all. The idea is to generate an alternative to the current path to adminship, not to generate a permanent class of "admin lights". --Cerejota (talk) 20:18, 2 September 2011 (UTC)

That is a very good distinction. Elements of this path are: a reasonable request, a review of the request, decline or grant the apprentice rights, (if declined follow a defined recourse to redress) the apprentice admin demonstrates fitness during the apprentice period (currently suggested as 6 months and 1000 log actions); removal of the apprentice right is equivalent to removing rollbacker, successful apprentices are then granted full administrator privileges with the only caveat that they are bound to a predefined recall process which will likely go through RfA (we need to define this recall process). So that is the kind of path we are trying to develop. An alternative to the current monopolized path which many avoid, and some refuse. My76Strat (talk) 21:37, 2 September 2011 (UTC)

• Question Would someone who has had a failed RFA be eligible under this proposal if they meet every criteria? If so, would the closure/content of the RFA have any impact? Would there be any waiting period between the failed RFA and applying under this proposal? Monty₈₄₅ 22:33, 2 September 2011 (UTC)
  • That's a good question, we should define those parameters as well. I would think in opposing someone at RfA the option of seeking adminship via the apprentice path might be suggested. We should define if a recent RfA is a factor or not. My76Strat (talk) 22:40, 2 September 2011 (UTC)

Hybrid

This idea seems to me to blend various elements of two ideas, admin lite and an alternative route to adminship. I'm not convinced that it combines them in a way that creates something better than either, in some respects it may even be worse. Admin lite with stricter criteria than adminship doesn't entirely make sense to me. I'll continue to work on both the other ideas, but whilst I'm not currently prepared to support this proposal I'm happy to give it some hopefully constructive feedback. ϢereSpielChequers 13:42, 2 September 2011 (UTC)

This is a draft proposal, subject to improvement. You have good insight in your comment and we need to incorporate as much of it as possible into the draft until we have the best achievable collaborative proposal. My76Strat (talk) 14:17, 2 September 2011 (UTC)

As I see it, this proposal is a structured version of "Bureaucrats appoint admins", not of "admin lite". The only reason this proposal doesn't touch the final adjudication of full adminship is because there seems to still be support for RfA - even if it is a reformed RfA - and as such, this proposal seeks to not contradict that consensus. However, I would be opposed to a "pure" system of "Bureaucrats appoint admins" because it would be come a festival of ass-kissery and arbitrary blackballing, and also make crats develop patronage, which can have nefarious consequences. If this "apprentice admin" idea becomes policy, and works, it could become the model for full adminship. I have no problem with that. We then would have RfA only for re-sysop and for those who do not meet the time/edit criteria.--Cerejota (talk) 20:17, 2 September 2011 (UTC)

Jimbo quote

If you re quoting Jimbo as supporting the idea of unbundling blocking for IPs and editors with less than 100 edits then you really ought to link to a diff of his ϢereSpielChequers 13:42, 2 September 2011 (UTC)

I erred in that statement. It was actually you who suggested this aspect. You are right that I should have linked the diff instead of relying on memory (sometimes faulty). I will correct anyplace where I wrongly attributed the comment to Jimbo. I still think it is an idea which has merit. My76Strat (talk) 14:32, 2 September 2011 (UTC)

Glad you like it, I'm beginning to think we should run it on its own as sensible tool to give to recent changes patrollers. See Wikipedia_talk:RfA reform_2011/Radical alternatives#Unbundle_block_.2F_unblock, though obviously I have no objection to seeing it incorporated in other schemes. But you might want to consider moving this proposal into two complementary schemes, an alternative route to full admins for the well qualified and a limited right for those who can be trusted to block vandals. ϢereSpielChequers 15:23, 2 September 2011 (UTC)

Deletion

Deletion is contentious. Some members of the community think that it is good to delete loads of goodfaith but poorly formatted, unreferenced and barely notable new articles as soon as they appear. Others think we need a longer process that salvages far more of those articles and their editors. Unbundling deletion is bound to be contentious, and some will want it restricted to "Full Admins". ϢereSpielChequers 13:42, 2 September 2011 (UTC)

That is clearly one of the oldest debates in wikipedia, and not just English Wikipedia, but that is not a problem that this proposal addresses or even attempts to address. Deletionists and inclusionists and those in between, are to be afforded the privilege of the tools if they meet the criteria. There is always DRV, and of course, admins can revert any action by an apprentice admin without it being a wheelie - that would take care of zealots on either side removing CSD templates or deleting articles outside of CSD criteria. And repeated misbehavior means automatic removal of the privilege of requested tools, so there is incentive to behave. --Cerejota (talk) 20:09, 2 September 2011 (UTC)

Can the deletion powers be reduced further? --Σ _{^talkcontribs} 01:57, 10 September 2011 (UTC)

If there is concern that they should be, we will modify the proposal to reflect a consensus desire. My76Strat (talk) 03:20, 10 September 2011 (UTC)

I think it will be difficult to get consensus for any admin lite proposal. But an admin lite proposal that included the delete button would be extra awkward. Limiting it to newbie's articles only is unlikely to assuage opposition as it is these deletions and the treatment of goodfaith newbies that is most contentious about deletion. ϢereSpielChequers 17:22, 23 June 2012 (UTC)

Criteria

Continuity

12 months continuous contributions is OTT. The community expects that longterm editors who suddenly resurface refamiliarise themselves and get up to speed with recent policy changes before receiving userrights. But historically it has been completely unbothered by admins, editors and RFA candidates taking months off. Many actually see it as a positive if the candidate disappeared for a couple of months before their exams. You might consider Active in at least 12 different months including the last two instead. ϢereSpielChequers 13:42, 2 September 2011 (UTC)

I tend to agree, contributions can be seasonal (ie summer recess for students etc), or life can get in the way (work, etc). However, this is also a way to minimize merit badging - so if we do not do this, we need other criteria in its place--Cerejota (talk) 19:46, 2 September 2011 (UTC)

I have no objection to a minimum criteria of 18 months tenure, 6000 edits, at least 12 of the previous 18 months active (an active month is > 100 edits) and at least the previous 4 months consecutively active. I think that is a fair criteria. My76Strat (talk) 21:46, 2 September 2011 (UTC)

That seems overly complicated. Ironholds (talk) 10:54, 4 September 2011 (UTC)

Do you have a suggestion which would constitute a criteria and be less complicated? My76Strat (talk) 19:51, 4 September 2011 (UTC)

• Try WSC's suggestion. Ironholds (talk) 20:56, 4 September 2011 (UTC)

Tenure

2 years is a longtime on the wiki. Personally I'd be happy to appoint admins with as little as 6 months experience if they are suitable. But the community expectation is more like 12 months, any less and you will get a significant number of opposess. If you are going to make this an alternative route to adminship I would suggest Active in at least 12 different months instead. Adminlight however should require less. ϢereSpielChequers 15:46, 2 September 2011 (UTC)

Then RfA. The "on request" nature - while long tenure is not a signal of worthiness, it is a signal of commitment. If someone with less demonstrated commitment wants to mop, let them RfA.--Cerejota (talk) 19:46, 2 September 2011 (UTC)

A major problem I can see with this is that if someone goes for RfA who doesn't meet the apprentice criteria, people are going to object at RfA that they didn't qualify for the smaller set of tools, so they obviously must not be good enough to get the whole enchilada for life. Whatever minimum requirements are set here will inevitably become a de facto minimum for regular adminship. I know that doesn't make complete sense, but I fear it will happen anyway. Richwales (talk · contribs) 06:23, 9 September 2011 (UTC)

Editcount

Raw editcount is a notoriously poor way to evaluate editors, though clearly some use it at RFA. Six thousand edits is nearly double the count of several recent new admins. That might make sense for an alternative route to adminship for over qualified candidates, but it makes no sense for an admin light proposal. ϢereSpielChequers 15:46, 2 September 2011 (UTC)

See my comment above. I oppose editcount as criteria for nearly everything, but we already have it as a criteria to become auto-confirmed and for the service awards. It is an imperfect measure, but it is the best we can come up with. However, those with less edits can still RfA. Again, this about the criteria for a request without RfA.--Cerejota (talk) 19:46, 2 September 2011 (UTC)

• The service awards have no validity and do not demonstrate trust or give access to tools - autoconfirmed, meanwhile, is the lowest possible userright. They're hardly comparable to adminship. Ironholds (talk) 20:57, 4 September 2011 (UTC)

• The full measure of trust will be determined by the conduct during the apprentice period. The apprentice user right will be easily revokable. The criteria is an effort to preclude request's from users who would otherwise SNOW close for NOTNOW concerns at RfA. The intent of stringent criteria is in recognition of the administrators unique roll. While adminship is not a big deal, misuse of administrative privileges can quickly become a very big deal. With these things in mind, what suggestions would you like to ensure are considered? My76Strat (talk) 21:47, 4 September 2011 (UTC)

• We're trying to accomplish two conflicting goals here: higher requirements in order to be allowed to bypass the RfA gauntlet, but a smaller set of abilities (that can be taken away more easily) because apprentices don't have to go through an RfA. My guess is that RfA !voters will fixate on the smaller set of abilities, and (as I said above under "Tenure") whatever requirements we might set for apprentices will unavoidably become a de facto minimum for passing an RfA. Richwales (talk · contribs) 06:32, 9 September 2011 (UTC)

Blocking

In the proposal, it is suggested that Essentially, "Apprentice admins" would have all the tools needed for cleanup, maintenance, and general vandalism protection... In dealing with vandalism, one of the key abilities that an admin has beyond regular editors is the ability to block vandals. I would argue that G3 deletions and semi-protections as interventions against vandalism is are far less common then blocks. I realize that blocking is probably the most sensitive admin ability in the eyes of the community, but I think the toolset is seriously incomplete without at least a limited ability to do so. I would propose that Admin apprentices be given the ability to block IP contributors and Editors who are not auto-confirmed. Auto-confirmed is a very low threshold, and most good faith editors should be able to make it that far with out being blocked by even an seriously overzealous apprentice admin. Further, they could be specifically prohibited from denying talk page access, so that an unblock request would be available, and a full admin would be able to review it. I'm not sure block duration restrictions would really be useful, though block severity (in terms of autoblock etc.) should be discussed. Monty₈₄₅ 14:56, 2 September 2011 (UTC)

I take issue with the concept that IPs are a sub class of editor. They may be treated as such, but they do fantastic work. I know it's easy to see them as vandals, but it's not true. Just because most vandals are IPs does not mean that most IPs are vandals. Whilst I'm sure it'll be done with the best of intentions, giving editors the ability to block IPs is not necessarily a good thing. Worm^TT · (talk) 15:07, 2 September 2011 (UTC)

Most IPs are not vandals, same goes for newbies. But telling the difference between a vandalism only account and a goodfaith newbie or IP is easy and almost all recent changes patrollers master it quite quickly. Deciding when to act against a long established but temperamental editor is not so easy and never uncontentious. ϢereSpielChequers 15:15, 2 September 2011 (UTC)

Right, the community seems very concerned about the admin candidates potentially blocked major, long term contributors, by over zealously applying the block policy. My intention is not to discount the valuable contributions of IP editors and newly registered editors, it is to give the ability to deal with the bulk of vandals while trying not to step on the third rail of RFA politics. Monty₈₄₅ 15:33, 2 September 2011 (UTC)

Autoconfirmed is probably a bit low, sometimes vandals will get above ten edits and a few days before they are spotted. I'd suggest 100 edits as far more than most vandalism only accounts get away with, but far less than a vested contributor would have. But otherwise we are thinking along similar lines. ϢereSpielChequers 15:15, 2 September 2011 (UTC)

While it certainly is true that there are vandals beyond the autoconfirmed mark, and some will intentionally build autoconfirmed accounts with the purpose of using them for vandalism, I think they are a sufficient minority that requiring apprentice admins to send them to AIV would not substantially undermine the apprentice admin's ability to deal with most vandalism, and it would provide the broadest protection available against block even minimally established editors. Monty₈₄₅ 15:33, 2 September 2011 (UTC)

I have no problem with adding the ability to block/unblock non-autoconfirmed users, but I do think block/unblock of autoconfirmed users has a DR component that if left in the hands of someone who has not gone through a process of scrutiny and vetting, will lead to much more drama. I purposely left this out of the proposal to generate discussion.--Cerejota (talk) 19:49, 2 September 2011 (UTC)

Thoughts on this proposal

Having read through this proposal, I think this may work; however, it does look a little like admin coaching, and I do know that several admins are against admin coaching. I do have a couple concerns regarding this proposal, though. They are:

1. One of the qualifications one must meet in order to be an admin apprentice is "Having been an editor in good standing for at least 1½ years with 6000 edits..." The qualifications to run for full adminship are much lower, and since RfA candidates have passed with less than these qualifications (1 1/2 years with 6,000 edits), in my opinion, this proposal won't see as much attention as it potentially could. I do realize there could be a reason for this unusually high qualification, so I'll wait for a response before I go into this further.
2. One of the suggestions under "Request process" is as follows: "RfA Advised because - most criteria met, but there are borderline issues that require community consensus, editor then has the option of going to RfA for appeal. This is a discretionary recommendation of the Bureaucrat only to be done when there are clear borderline issues (for example, concern with technical ability, not enough "Wikipedia" namespace participation etc)." If the potential candidate is not able to meet the high requirements of apprenticeship because of these kinds of problems, an RfA would only address these problems a second time, and would most likely fail. "Request process" also reads: "If it is RfA Advised because, the editor can appeal to RfA immediately, or can attempt a request again after the because issues have been addressed and sufficient time has passed." Again, if the editor went to RfA immediately with these issues, then it is likely they would not pass. I would suggest the bureaucrat not refer the candidate to RfA if he/she doesn't meet the qualifications; rather, they should tell the potential admin apprentice why they don't meet the qualifications and how they can improve so they can meet the qualifications.
3. I also have a problem with the following: "Repeated requests after time has passed that are RfA Advised because can be considered as permanently needing an RfA." If the candidate continuously fails to become an admin apprentice, then it's likely they will continuously fail to become a full administrator through the RfA process. As we all know, the RfA process is harsh, probably harsher than this proposal would be, so the candidate would likely become discouraged after a certain amount of tries and completely retire from Wikipedia (we've already seen this in many cases through the regular RfA system, and if this new proposal goes through, we may or may not see an increase in discouraged editors. This is due to the fact that, in addition to the high possibility of being rejected on RfA, theoretically, they were rejected several times by this proposal, and constant rejection isn't exactly good on someone's self-esteem).

This proposal does have its flaws, but that is expected; after all, it is new. I hope my concerns, and the concerns of other editors, can be taken care of so that this proposal can be tested. The Utahraptor^Talk/_Contribs 15:27, 2 September 2011 (UTC)

Yes that is the desire. In fact contributors are encouraged to be bold in making changes. We can revert and discuss as contentions arise, but achieving the best proposal is the goal. My76Strat (talk) 15:52, 2 September 2011 (UTC)

I agree with the co-proposer's comments fully, to address the points raised:

1. If this were a requirement for a community vetted consensus, I would agree this is excessive. Since the process as proposed is largely Bureaucrat discretion, a brightline criteria is needed. Community consensus on the service awards is that this level constitutes an "experienced editor". That is why this is the criteria.

1. I see the point and that is a possible scenario, but there are other scenarios, for example, a user that had problems when they began in wikipedia but has since long reformed in the eyes of the Bureaucrat, but the Bureaucrat feels uncomfortable with making a call. In this sense, I think many candidates might fail, but many will pass. I think even at the current level (60% fail - 40% pass). My expectation is that his would be a rare outcome of a request, but rare is not impossible.
2. I agree, but I don't think this is a judgement we can make for the user. WP:CIR tells us people would understand this. We must address the possibility of disruptive and incompetent behavior to empower the process to deal with. Experience shows that discretionary policies are drama magnets - see WP:EW but 3RR is generally undramatic except for those being blocked. This begs for a brightline so that malicious users or good faith incompetents can be handled in an uncontroversial fashion.

I hope I have addressed your points, and of course, these are simply rationales, not intended to block any possible alternatives that are proposed.--Cerejota (talk) 20:05, 2 September 2011 (UTC)

Thanks for the response. My concern with point one was that potential admins wouldn't consider apprenticeship due to the high edit count and time requirements, and instead they'd turn to Wikipedia:Request an RfA nomination or even nominate themselves at Wikipedia:Requests for adminship. In my opinion, gaining partial adminship rights does require a lot of experience here on Wikipedia, but it shouldn't exceed the minimum requirement of full adminship rights. Point two ties in with this high apprenticeship requirement, in that if the high requirements of apprenticeship are not met, then the lower (but still substantial) requirements of full adminship may not be met, either. I understand there would probably be exceptions if this proposal went through, so we'd have to see what happens with a test run. Point three was addressed well, and I no longer have any concerns regarding it. The Utahraptor^Talk/_Contribs 20:33, 2 September 2011 (UTC)

Well taken, but my point in this proposal is the exact opposite: RfA (or some other sort of community vetted adminship) can be pursued by anyone with very minimal criteria. Requests of tools - with a low bar of refusal and little discretion, on the other hand, should have a higher bar of criteria to account for the lack of community vetting and the narrow discretion. I think that is a simple and fair formula: those who do not meet the criteria but want to become admins still have other options. This also what makes this proposal different from other "admin light" proposals: it is a way to encourage experienced editors to become admins, not a way to make adminship easier.--Cerejota (talk) 21:48, 2 September 2011 (UTC)

Alright I think I'm up to speed now, thanks for the explanation. All of my concerns have been addressed. Hopefully we can get this proposal tested so we can work out any bugs that aren't noticeable at the moment. The Utahraptor^Talk/_Contribs 22:15, 2 September 2011 (UTC)

I actually quite like the look of this idea, though I thought I wouldn't as (at first glance) the length of tenure and quantity of editcount seemed strange. But, having read through the comments here, I can now see the reasons for those. It's certainly a good set of foundations to build on, though it's likely to have quite a few tweaks before it can go live. Pesky (talk …stalk!) 08:45, 3 September 2011 (UTC)

Presumptions of RfA vs. admins-in-training

An interesting proposal, but just one thing I see I'm a bit shaky on:

The proposal lists a variety of things which admins would be allowed to do but apprentices wouldn't. When we have an RfA, we ask why someone wants to be an admin: the exact question is "What administrative work do you intend to take part in?"

What if we had someone whose only interest is in doing admin work that falls outside of the domain of what apprentice admins can do? Some people might only be interested in doing wikignomery at WP:RM, say, and want to become an admin solely so they can handle RM requests that are move protected. If they are unable to move articles that are move-protected during the apprenticeship, how will the community assess their ability with that tool? Someone might simply want to work the refund desk but if they can't get to deleted edits, how are they going to show their compentence or incompetence at their chosen administrative work?

Ideally, we want to be able to see them in action without observer effects. Imagine if we just gave them all the tools and said "okay, two weeks, show us what you've got" - they'll behave themselves for two weeks, then once they pass, they hit WP:ANI and start wheel-warring and giving over-the-top blocks to all and sundry.

So, I guess my questions are:

1. How do we deal with observer effects? When someone is taking a driving test, they drive differently from when they don't have someone peering over their shoulder.
2. If we are going to have an apprenticeship type system, how do we ensure that the rights that get given to the apprentice line up with the sort of work they might want to do?

Otherwise, interesting proposal. It reminds me of how Wikiversity handles adminship. I'm not sure whether that's a good or a bad thing! —Tom Morris (talk) 14:23, 3 September 2011 (UTC)

I think before this proposal is ready for an RfC we will have tweaked the sub-set of apprentice tools. This to address the valid comments that have been given, like yours, and the ones that are to come. To minimize the "observer effects", this proposal requires all administrators who receive the administrator right through this process to be bound to a sensible recall criteria, which still needs development. The idea is to ensure that conduct after the apprenticeship is censurable, and removal of the right is far less cumbersome than what exist currently for candidates successful at RfA. My76Strat (talk) 15:08, 3 September 2011 (UTC)

Time limit?

I would add an "up or out" provision. After some long period, apprenticeships expire; the idea would be that the apprentice should either apply for full adminship within (say) five years, or subside into ordinary editorship. I suppose that if somebody specifically wanted another five years of apprenticeship, they could ask for that explicitly, explaining why. Septentrionalis PMAnderson 20:48, 7 September 2011 (UTC)

This is a valid consideration. We definitely want to encourage editors who request apprenticeship to be active in using the tool-set given. And we want as many who are fit for sysop, to achieve that level of user right. 5 years seems overly generous; I'd suggest 1 year. Of course additional comments can help hone a good consensus for this limit, and I look forward to them. My76Strat (talk) 21:13, 7 September 2011 (UTC)

I actually think six months would work. If you think about it, six months is a pretty good amount of time for an apprentice admin to get the hang of their tools, and if they still feel they haven't fully gotten the hang of it, they can apply for another six months of apprenticeship. But if this were implemented, we'd have to have a limit as to how many times an admin apprentice can renew their apprenticeship, say four times (two years). After this limit is reached, they can either apply for full adminship or return to regular editorship. What does everyone think? The Utahraptor^Talk/_Contribs 21:27, 7 September 2011 (UTC)

That model could work. Currently the apprentice period is set at 6 months, but there is also a requirement of 1000 log actions. The limit would exist related to how long we should wait for the log actions to accumulate. And there could be other ways yet to consider. I just can't impress enough to each contributor to this discussion, that we want to emerge with the best, thoughtful proposal that we can achieve. Each persons comment provides the blocks we need to build this proposal, and reach that goal. Thank you for yours. My76Strat (talk) 22:01, 7 September 2011 (UTC)

I think the log action requirements should depend on how active the potential admin apprentice is. If the admin apprentice is making an average of 200-300 edits per month, we could set the log action requirement at 1000, or if that's too high we could cut it in half at 500 actions. If the admin apprentice is making an average of 300-500 edits per month, we could set the log action requirement at 1500, or if that's too high we could set it to 1000. And so on. Below is a table that can explain my thoughts better than words can:

Log action requirements

Average edits per month	Number of log actions
100-200	250-500
200-300	500-1000
300-500	1000-1500
500+	1500-2000

--The Utahraptor^Talk/_Contribs 22:12, 7 September 2011 (UTC)

I might make this a matter of bueaucrat discretion; after three renewals, say, the apprentice must specify why they are requesting yet another. If the 'crat is not convinced, that is an RFA because (and in this case, the RfA may well succeed, unlike most examples). Septentrionalis PMAnderson 22:18, 7 September 2011 (UTC)

Unless I see an objection to this, I'm going to include it. Septentrionalis PMAnderson 21:04, 9 September 2011 (UTC)

I think you opened this thread with a valid point. Whether we reach the final wording or not, with your edit, it will certainly be good to initiate prose in the draft, that show this is a contingent for possible refinement. Even before you posted this thread, it had occurred to me that if there were no stipulation otherwise, a candidate could languish in the apprentice state, if even by choice, simply by not achieving the required log actions. So I was glad to see the discussion bring this to bare, and will be as glad to see the prose begin to take shape. Happy editing - My76Strat (talk) 23:19, 9 September 2011 (UTC)

(edit conflict)I have no objection. It actually solves the issues of "merit badging" and forces the question: If you are real about adminship, show it :) However, the table is WP:CREEPy to my eyes - other than a time limit, I would advice bureaucrat discretion - for example, distracting a good admin apprentince from a GA/FA drive to meet a log quota is counter productive, also having issues IRL etc. --Cerejota (talk) 23:29, 9 September 2011 (UTC)

Two thoughts

Tool separation

I was considering this possibility as it relates to the notion of a subset of tools for the apprentice. And it certainly relates to trust as well, in so much as this is ultimately a position of trust. Rather than necessitate the need for a new grouping of tools, bind the apprentice to an agreed ethic on bond of word, and remove the right upon the first, and any, breech. My76Strat (talk) 23:59, 9 September 2011 (UTC)

That would lead to IAR situations and in extremis apprentices being desysopped for good actions that were beyond their authority, or scope creep as apprentices do more and more of the actions that were supposedly reserved for admins. The latter might not be a problem, the former would be. My preference would be that the tools were hardcoded as to what they could and couldn't do. ϢereSpielChequers 17:37, 23 June 2012 (UTC)

The criteria, and trust

I want to include a note that effects a clause, directly upon the application for apprenticeship, where the applicant affirms that they do meet the criteria, or have disclosed in advance where they request waiver. This is another provision where a breech of trust should not be tolerated, even if an element is discovered much later, even into the full administration. Even a diligent review can miss something, and there is every reason to believe in the good faith affirmations of another. If something is missed, and then found, it won't be at the reviewers peril, but to the one who defrauded the application. As an aside, I would like to reconsider some of the things we state can never have happened, and allow for the discretion of a waiver, upon circumstances of mitigation. My76Strat (talk) 23:59, 9 September 2011 (UTC)

A sensible recall

One provision of this proposal is that all apprentices who achieve adminship by this means are bound to a standard and sensible recall. We need to develop this, so add thoughts. My76Strat (talk) 03:22, 11 September 2011 (UTC)

Any apprentice who is recalled must either

• No longer apply for apprenticeship (possibly it would be permissible after two years)
• Or, alternatively, mention it on the renewal application; after all, there will be unreasonable recalls.

Two methods spring to mind:

• Six editors in good standing may force a recall,
• Six editors in good standing may force an RfA

These do not differ much, since the apprentice may always ask for the mop back by RfA.

Looking at recall policies of admins open to recall may provide other ideas. Septentrionalis PMAnderson 00:03, 12 September 2011 (UTC)

One of the reasons why alternative recall methods don't get consensus is that many of us think that Arbcom for all its flaws is the most sensible recall process yet devised. I would suggest that we divorce the creation of admins from the desysopping arrangements. The two are very different subjects and debating both in the same proposal is practically guaranteed to divert any debate into areas where one can't build consensus. ϢereSpielChequers

We need to do this

I developed an essay which I believe will augment our objective; at WP:ADMAN. I need as many people who know an ADMAN (read the essay, it's short enough), to copy the substitutable message to their talk page. You can check the link which shows users who have been contacted already, to reduce duplication. Otherwise we need to generate interest, and this essay can help. I have had positive interactions with the ones I have linked so far. Seriously, please! do this; for the good ones that you know. If nothing else, you've told them a nice thing that was also true. My76Strat (talk) 03:22, 11 September 2011 (UTC)

Wrong place...

This is bizarrely grouped under RfA reform 2011, Radical alternatives. Nothing particularly radical here. It is more or less what the process should be - experienced editor in good standing with clean history wants to do more to build the encyclopaedia, requests tools from (group of) someone with better standing. Group needs only take 5 minutes to look at a bit of edit history (talk page history usually highlights any potential issues) and grant rights or not. If editor subsequently abuses tools, simply remove them. Just the next step on from requesting rollback. Too many people have turned RfA into a bureaucratic process and made it the BIG DEAL it is not supposed to be. --ClubOranje^T 08:09, 30 September 2011 (UTC)

I think you are exactly on track with the spirit of this discussion. It will help as more people express the intuitive nature of viewing adminship in this way. --My76Strat (talk) 09:08, 30 September 2011 (UTC)

Admin Apprenticeships

From reading the comments above and also the reform I do believe that this is a good idea. You have maybe several more experienced administrators who are willing to make sure that the apprentice is making the right moves shall we say. The more experienced administrators will be able to like admin coaching watch over them but as an apprentice you are there to learn the rules of being an administrator and this needs to be made clear that the guidelines if this was to be how it was to be then they an experienced editor with a minimum of say 6 months at least and that they haven't had any problems where administrators have had to get involved. I do admit that as an apprentice mistakes will be made and the more experienced administrators will be there to point them in the right direction if they don't listen to it then the apprentice needs to think clearly why they have become an administrator and that they are trusted user. Corruptcopper (talk) 18:11, 30 September 2011 (UTC)