Wikipedia talk:Trial adminship

Note: for an alternative proposal, see User talk:Doc glasgow/provisional adminship.

Initial comments

• • Only if Special:Undelete (with respect to viewing deleted revisions; undeleting is already logged) becomes a logged action, or the risk of abuse is too high. --ais523 16:08, 2 February 2007 (UTC)
    • That's an interesting point. But note that the really Bad Stuff tends to be oversighted, and thus is not viewable that way. >Radiant< 16:12, 2 February 2007 (UTC)
      • Am I the only one who does things this way? When I delete something at CAT:CSD, I go back and use undelete to see if the creator has been warned and see if it has been recreated recently and needs to be salted. I don't know how much I would like having that trivial action filling up logs. Also, if we don't trust someone not to mess with stuff, they shouldn't be an admin. --BigDT 16:22, 2 February 2007 (UTC)
        • You're absolutely not the only one. Anybody who does content review/restoration at deletion review would spam the logs. I'd actually rather see all users be able to see that there were deleted revisions, and perhaps the edit summaries, just not what the page's actual content was. -- nae'blis 18:36, 2 February 2007 (UTC)
  • I really don't like this at all - there's entirely too much risk of sockpuppet/meatpuppet admins this way. Really, the biggest change we need is a cultural one - we just need to quit !voting against people because of reasons other than whether or not we trust them with the tools. If someone isn't a prolific editor, who cares? As long as you can see, through the body of evidence available, that they are trustworthy and that they have sufficient experience that they would know when to hit what button, !vote for them. --BigDT 16:22, 2 February 2007 (UTC)
    • And what exactly would be the problem with having sockpuppet admins? If somebody has multiple accounts and they're not disruptive, so what if they're admins? If they are disruptive, we block them. >Radiant< 16:27, 2 February 2007 (UTC)
      • Great. You block one and the other five all gang up on you and accuse you of blocking a respected administrator. --BigDT 16:46, 2 February 2007 (UTC)
        • Yeah, I get that scenario, but anything we propose is a trade-off to begin with, so I tend to ignore slippery slope arguments. First, how often are respected admins blocked to begin with? And second, the above case would be found out by the checkusers, and all six accounts deopped and blocked for disruption. We're not exactly stupid about that :) >Radiant< 16:49, 2 February 2007 (UTC)
• I don't immediately see any downsides to this idea. Whatever problems can happen in this scenario are problems we already have, right? Friday (talk) 16:51, 2 February 2007 (UTC)

(Was this meant to be closed? If so feel free to revert...) RxS 17:15, 2 February 2007 (UTC)

• A great idea. Having apprentices is a time-tested method for developing skilled artisans. I'm just starting to watch RfD discussions and that seems to be an imperfect process as well. I hope that you succeed. I would be more inclined to seek admin status under this program. --Kevin Murray 23:15, 4 February 2007 (UTC)

More thoughts

A few:

1. The suffrage probably needs to be amped up/adjusted a bit. One month is nothing in Wiki-time, and 1000 edits isn't a whole lot, either. My suggestion would be 3 months and/or 1000 major edits/2000 overall edits.
2. Not sure how to treat blocks on this one. Not all blocks are created equal (we wouldn't treat a 3RR the same as a vandalism block, for instance), and some blocks are rescinded right away, but are still logged as blocks. Clarity is key.
3. Probation ends, and confirmation begins. Probationary admin shows no preponderance to abuse of the tools, no poor decisions. 12 users then come along protesting promotion due to a conflict from six months earlier. Does the confirmation process only deal with the prior month, or the whole body of work?

Just talking out loud at the moment. --badlydrawnjeff talk 17:19, 2 February 2007 (UTC)

I'd like to see would-be probationers being required to find a sponsor. Someone is going to have to mentor them, investing a significant amount of time in the would-be-admin, so the endorsement of one existing administrator willing to mentor the candidate seems like a reasonable minimum for submitting an application. Otherwise we could have admins who have little confidence in editor X mentoring editor X. That's not fair to either of them. In this case, I don't see why any editcountitis, or a buggins' turn time served requirement, would need to be written into the process. If the sponsor trusts editor X, and is willing to work with them, that should be enough to begin a discussion. They're the one who'll be doing the work, and will probably end up picking up the pieces if it all ends in tears. Angus McLellan (Talk) 17:26, 2 February 2007 (UTC)

Excellent idea. --badlydrawnjeff talk 17:30, 2 February 2007 (UTC)

Seems like a good idea, but I'd be afraid of folks who have left Wikipedia and stated that they would like to disrupt it...whether at Wikipedia Review or elsewhere. And they would be happy to befriend an admin for sponsership. The probabtion period (as written right now) would seem like a small price to pay for getting (fairly) quick admin rights here. I can see a steady stream of disruptive admins coming through, doesn't seem worth it. Unless there's a way to avoid that. RxS 17:35, 2 February 2007 (UTC)

You think that other folks won't mention this? We'd still have a discussion period. I imagine that "X is a meatpuppet/sockpuppet of (banned user Y)" would replace "not enough wikispace edits" as the most common thing you'd hear. Angus McLellan (Talk) 17:57, 2 February 2007 (UTC)

The possibilities of a wikipedia review style troll gaming the system should not be feared. 1) They'd have to give us 1,000 good edits first - then they'd do a little damage - which we'd revert and speedy desysop them. The net effect of that isn't a big loss. 2) For every 1 gaming troll, we'd probably produce 50 good admins. The net effect of that is a big gain. Why are we so risk averse?--Doc^g 18:06, 2 February 2007 (UTC)

Presumably because WP:AGF is ignored, except when it's a handy club to beat your opponent over the head with. Angus McLellan (Talk) 18:10, 2 February 2007 (UTC)

A couple comments, first I'm pretty sure someone with bad intent wouldn't announce their plan in advance so we'd have no way of knowing who was doing something like this without checkuser evidence. As far as damage goes I don't know...how much damage could an admin do in 2-10 minutes? Wasn't there some issues with deleting (and restoring) large history filled articles affecting server performance? I'm can't speak authoritively on that, but I do remember some issues in the past...maybe when VFD got deleted/restored? Anyway, I think it's worth considering. I'd hate to see the whole thing turn into some "top this" game. RxS 19:39, 2 February 2007 (UTC)

As Doc Glasgow said, you'd have to do quite a lot of valuable work for the evil hivemind in order to undertake a small amount of vandalism. Clueful vandalism (like the recent fake WP:AES stuff, which could have been much worse) is probably easier, and more damaging, than deleting some random pages for lulz. Anyway, as Friday said, the current system prevents this how, exactly? Angus McLellan (Talk) 20:34, 2 February 2007 (UTC)

I don't think that this proposal (as it stands now anyway) calls for quite a lot of valuable work, it's easy to rack up a couple thousand edits without doing much at all..and anyway the point is that I'm not talking about deleting some random pages. I'm talking about affecting server performance by deleting a page with a very large history (for example). Real "clueful" vandalism can have a real impact on usability...I'll have to find some of the discussing about this. Maybe there's a dev reading this that can tell me if I'm full of beans, ahem. RxS 22:26, 2 February 2007 (UTC)

Except the real clueful vandals do the worst possible thing -- inserting large amount of copyrighted or false material into Wikipedia on a subtle basis. This doesn't require the admin tools at all, and presents a serious threat to Wikipedia, rather than simply disrupting editing for a few moments. Christopher Parham (talk) 22:30, 2 February 2007 (UTC)

Doc, we are so risk adverse because of the potential to cause great harm to the image of the community. As I said before the wider will not differentiate between a trial admin and the full thing. Admins are supposed to be trustworthy people, ambassadors to the community because they are highly visual and therefore any bad admin, probationary or not has the potential to do great harm in their dealings with new contributors. Viridae^Talk 21:30, 2 February 2007 (UTC)

What if each person has to have no non-accidental blocks and has to get the approval of at least one admin. --TeckWiz^Talk Contribs_@ 21:36, 2 February 2007 (UTC)

Considered that while I was writing. That is just going to triple the workload for the full admins, thereby invalidating the influx of new ones. How many articles get tagged for speedy deletion without cause (ie they dont fit into WP:CSD), and how many people get reported to WP:AIV without cause, either without being warned properly or on occasion, for a content dispute. I can envisage a massive rise in the tools being used in content disputes in which the probationary admins are involved, and therefore wheel wars, missues of protection ability etc. We don't to change more admins overall, we just need less of a focus on a persons strong ability to write articles (in peoples RfA standards) and therefore allowing more wikignomes to go through to help clear backlogs. Viridae^Talk 21:46, 2 February 2007 (UTC)

I wouldn't count myself as agreeing with G____ and I____; many sorts of contributors are needed to keep things running, and specialisms are no bad thing. That said, without articles there'd be no encyclopedia. And any editor can write articles, if they try. Doc Glasgow wrote several fine pieces lately, and what's more, he confessed to enjoying it. My view is that gnomes have as much chance as anyone else under the current system, and would probably be even better served by the proposed one. After all, how many edit wars can you get into stub-sorting, and where are the opportunities to be wildly incivil? Angus McLellan (Talk) 23:20, 2 February 2007 (UTC)

Why not have a trial of trial adminship?

All we need is a bcrat who's willing to promote on these terms, right? Do we have a volunteer? Friday (talk) 01:37, 3 February 2007 (UTC)

it is agaisnt policy at the moment and there has not been enough discussion/no consensus reached. Viridae^Talk 05:28, 3 February 2007 (UTC)

We'll be in a better position to judge what works or doesn't, with an actual case or two to look at. I don't see that it's against policy- the crats already promote at their discretion. Since we're talking about something only a crat can do, it's really up to the consensus of the crats, not the rest of us. Friday (talk) 18:47, 3 February 2007 (UTC)

Sign me up as the first whenever this hits consensus Friday. I think everyone would agree that I would be the prime target for this no doubt :) semper fi — Moe 22:06, 3 February 2007 (UTC)

Some revisions

I really like the way this is going. I've fleshed out the proposal a bit, while still trying to maintain KISS. In particular, I've added:

• A limit on mentorship/sponsorship of one trial admin per serving admin. This will help keep serving admins from being bombarded with requests, I think, as well as assuring the community that a trial admin can be properly monitored.
• A requirement that an serving admin have three months experience before being a mentor. (I wouldn't object to this being longer.) It seems logical that a brand-new admin shouldn't be able to sponsor his/her best friend on the day after he/she becomes an full admin.
• A mechanism for evaluating "barely used the tools at all", which was lacking in the proposal.
• A requirement that the mentoring admin make a recommendation on full adminship. This is probably a formality, but it also clarifies the process by which a trial admin undergoes a final comment period (which begins once the recommendation is made).

I realize that the above may be excessively bold, but I think the proposal needed a bit more meat. I ask that if someone has a problem with the above, they remove only what they object to, not the entirely edit, and bring that part to this page (of course) for discussion. -- John Broughton (☎☎) 01:46, 3 February 2007 (UTC)

Dependance on admins

I sorta like the idea of requiring an admin to mentor- admins have already been vetted by the community once. So I added a third option where an admin could object to promotion. If we're already depending on the judgment of admins, let's depend on it again. Friday (talk) 01:50, 3 February 2007 (UTC)

I'd change:

At that point, the user gets regular adminship unless (1) at least ten registered users (per suffrage above) object to that, citing specific actions as their reason for objecting, in which case a regular RFA decides, or (2) the user has barely used the tools at all, in which case the editor probably doesn't need them or (3) a serving administrator has objection for good cause.

to

At that point, the user gets regular adminship unless (1) at least ten registered users object to that, citing specific actions, or other good cause, in which case a regular RFA decides, or (2) the user has barely used the tools at all, in which case the editor probably doesn't need them.

The current third basis is rather too like a "black ball" in the process, and if "other good cause" can be a reason, it should apply to bother editors and administrators who raise objections. An administrator who can find a good cause for denying promotion will surely find nine other editors who'll take the objection seriously, if it is indeed good. Should the "specific actions" be limited to administrative actions, and/or to the probationary period? Angus McLellan (Talk) 18:39, 3 February 2007 (UTC)

I don't like the bit wanting a specific number of editors. Since this is a "fast track" option, and regular RFA can still be used, I don't see a problem with there being only a small barrier to rejection, just as there's only a small barrier to becoming a provisional admin. But, maybe this all just needs simplified- maybe anyone can object, and if the objections are for reasonable cause in the bcrat's opinion, they don't get the permanent promotion. Friday (talk) 18:49, 3 February 2007 (UTC)

I don't think this should be an RfA workaround. I can, however, see this as something a permanent admin candidate could use on an RfA. -- Selmo ^(talk) 05:39, 4 February 2007 (UTC)

On apprenticeship

The concept of apprenticeship has a long and successful history. While this proposal calls for mentoring, for a relatively brief period of time, which isn't the same thing, one aspect of this proposal that makes it so attractive is that is does encourage and permit an informal or even semi-formal "apprenticeship to the community" type of requirement for interested editors. By that I mean the following:

• While not required, admins probably will set up "gateway" requirements for those interested in having that admin be their mentor for a trial adminship: at least X mainspace edits, Y comments in XfDs, Z months of experience, etc. (I think it's safe to say that any admin willing to be a mentor is going to lay out requirements if he/she lets it be widely known that he/she is accepting requests.)
• A admin who focuses on (say) CfDs might be interested in mentoring only those who have a large number of comments in CfD discussions. And certainly admins who constantly are dealing with backlogs will be motivated to find one or more promising editors who can help with the backlog, and if these are not easy to find, will want to make it known that he/she is seeking someone with a certain experience (with image copyrights, with username problems, with CfDs, whatever). And any such "help wanted" signs will in turn encourage editors to commit to a particular type of work in order to become an attractive candidate for at least that particular admin.
• Nothing would prevent an admin from setting up his/her own apprenticeship program - editors could be asked to apply for "pre-trial review" status; the admin would screen out anyone he/she thought wasn't that promising, and then would expect an editor "admitted" to such a program to help out in area X for (say) three months, after which (if the work was satisfactory), he/she would be nominated by that admin to be a full-blown trial admin.

Finally, I note that another really nice thing about letting each admin decide whether to mentor, and what his/her requirements are for candidates is that it lets "1000 flowers bloom." By that I mean that there aren't going to be rigid barriers that apply to everyone (insert note about no one with less than 2000 edits having a succesful RfA in the past six to 12 months); rather, each admin can specify what might be seen as their personal apprenticeship requirements, even if those are general (e.g., an edit count that excludes welcome notices, or participation in at least 30 RfCs). That in turn might remove the diviseness that seems to be in current RfA discussions about what admin requirements should be. -- John Broughton (☎☎) 14:54, 3 February 2007 (UTC)

Absolutely! Nicely said. Friday (talk) 18:50, 3 February 2007 (UTC)

More thoughts

This is a absurdly awful suggestion. In theory, it's an excellent way to get good editors promoted without suffering the disgrace that is RfA at present. In practice, I'm concerned about the following potential problems though

1. It's got cabal written all over it - I can just see apprentices being nominated by an admin for their beliefs and outlook rather than their ability to help run the project.
2. It's much too personal, RfA is sucky enough without making rejection much more personal when an admin overlooks you for apprenticeship or turns you down flat.
3. Due to 1&2 above, it's giving admins far too much power to pick and choose who they have as admins in future.
4. And because of 3, above, there's a real chance apprentices could be victims of association, imagine an apprentice who was nominated by an admin who during the apprenticeship is desysopped, they stand a real chance of being opposed come confirmation time. What happens in those circumstances anyway - would the apprentice admin be desysopped or not ?
5. Bureaucratic discretion is going to open up a whole new can of worms, if we're still proposing a limited number of apprentices, it raises the question over cabalism and/or favoritism of prospective apprentices or admins, which seriously cannot be good.

-- Heligoland 19:14, 3 February 2007 (UTC)

TINC. And providing that the community has the last word, and anyone rejected by the new scheme has the right to appeal to the community via the current RfA, then there is little danger of this.--Doc^g 22:13, 3 February 2007 (UTC)

Sadly, that's not good to be good enough, I fear. If there's one way of becoming an admin that is easier and/or more clique-ish, then it's going to create a two-tier admin system and risks dividing the admin community. The fact there's going to be some sort of quota system will only make this situation worse, human nature being what it is, I'm seriously concerned about all of the points I've made above being made and widespread disruption being the result when the accusations start flying. I've not even mentioned how useless these apprentice admins will actually be, not doing anything remotely contentious to make their confirmation more likely, we need a process that gives the tools out to those who need them, with the mandate to use them fully, and this apprentice system just won't do that. -- Heligoland 22:50, 3 February 2007 (UTC)

Since when is doing contentious things what an adminship is about? Quite the opposite IMO- the best admin is the one who's not even noticed because nobody disagrees with what he does. An apprentice admin who does new page patrol and quietly deletes hundreds of junk pages is far from useless. Friday (talk) 23:00, 3 February 2007 (UTC)

So that's about all they would be able to do. There's no chance of being able to delete the thousands of orphaned fair use images, deal with page protections, remove spam or block users without someone complaining. Who was the last admin you saw without someone complaining on their talk page ? There's only a few candidates that can go through the RfA process at present without people bringing up varying degrees of concerns. Like it or not, people will find plenty to complain about with these apprentice admins. -- Heligoland 23:10, 3 February 2007 (UTC)

I'd just like to say that despite my concerns above, I'd like to see a trial run. Since it's unclear how many months an apprentice admin would work before their confirmation, I'd suggest running for a couple of months if it's going to be sort of 28 days from being given the tools to the confirmation, and 6 months if it's going to be a 3 month apprenticeship, so we get to see how the nomination and confirmation process works for the first tranche of apprentices, and how the whole process affects the second tranche of apprentices after it's all up and running. -- Heligoland 23:36, 3 February 2007 (UTC)

I think this is one of the best proposals that have been made in a long time, though I'm not sure if it would pass. This eliminates the "can't be trusted" and "not experienced enough" votes. I'm sure there are plenty of users, including myself would like to try this out. However, some basic requirements should probably be that the user has been here for six months, has over 2,500 edits, and has never been blocked (except by accident) After all, we don't want to waste time mentoring a user that has no chance at getting enough support to become a permanent admin. Also, I don't think it should last that long. Probably one month maximum for a trial. And if the admin does anything against guidelines on purpose, they are disqualified from the process and immediatly desysopped. --TeckWiz^ParlateContribs_@ 00:10, 4 February 2007 (UTC)

Your criteria are ridiculously high, and defeat the whole purpose. When I started this debate, my intention was that we could give the tools, on a probationary basis, to people whom under the current system the community wouldn't have enough knowledge of to be willing to trust unconditionally. If you've got 2,500 edits, 6 months experience, and a clean block log, then you'll sail through an RfA as it now stands. This idea isn't for you. The point is to give a trial to people who currently would be regarded as having too few edits, too little time, or a more shaky record. We let them try out under supervision, so that the community can then judge them on their performance instead of counting edits etc. Many of them will turn out to be good. --Doc^g 00:37, 4 February 2007 (UTC)

Doc, I had 3,000 edits, 10 months, and no block log when I ran, and I was still declined. --TeckWiz^ParlateContribs_@ 00:42, 4 February 2007 (UTC)

I don't know you at all. But I'd say that if you have that type of record, the community do have enough to judge you on. That judgement may be wrong, but it is still more plausible than thinking someone can't be a good admin because they've only got 1,000 edits. Actually, they can. Let's enable them to prove it.--Doc^g 00:50, 4 February 2007 (UTC)

TekWiz, use me an as example. I have almost 13k edits, plenty of experience in almost every area of Wikipedia, and a couple blocks from 6 months ago that arguably should have never happened to begin with - and one of the blocking admins is Doc Glasgow, who said he'd support me if I went through RfA again. If you know me at all, you know I'll never pass RfA - I'm "too inclusionist," people hold grudges about an association I severed ties with more than a year ago, and I tend to rub some established people the wrong way at times because I'm not afraid to get my hands dirty. It doesn't mean that, given the opportunity, I'd be one of the best admins this site has ever seen, I simply won't meet the completely arbitrary RfA standards. I haven't bothered with RfA since then, and I'm not even sure I'd go along with this, but your standard pretty much eliminates anyone who's ever been involved in a conflict where someone steps in with a banhammer, even for a short time. i use myself as an example because I don't mind the scrutiny, but I know a dozen people who should probably be considered who have a more dented record than myself. RfA as it stands rewards people for the wrong reasons. I know I could set up a sockpuppet tonight, do some gnoming for three months, and get promoted, but established people with something to offer probably shouldn't have to do that for something that allegedly, is "no big deal." --badlydrawnjeff talk 01:01, 4 February 2007 (UTC)

Doc - I rather like this idea if it were to be used to promote bloody good users in circumstances where the current RfA system has failed them and the community, but not where it's going to be used for people who, due to inexperience and a lack of knowledge, wouldn't stand a snowball in hells chance of passing the current process. I'd support it almost unconditionally if it could be invoked immediately following a second unsuccessful RfA, where the support rate was between 60/65% and 75% and where the RfAs were spread over at least 3-4 months (basically, not failing an RfA to access this process). -- Heligoland 01:25, 4 February 2007 (UTC)

What about people who do have sufficient experience and knowledge, but due to lack of edits or time on WP wouldn't pass through the current process? Those people would seem to be an ideal target group for this process. Christopher Parham (talk) 01:48, 4 February 2007 (UTC)

These are actually the people I'm aiming at. If it helps a few battlers who'd otherwise struggle, fine. But when I instigated the debate it was about broadening adminship out to the many potentially good uses on whmo the community lacks enough information to judge. Not particularly to give a back door to those it knows only too well. I was looking to see many dozens of new admins by a more pragmatic method, not a couple of old hands.--Doc^g 01:51, 4 February 2007 (UTC)

So we're ignoring users with lots of experience and knowledge who have upset a few people along the way, and promoting new users with little or no experience. That's fair and sensible, not. I know adminship is not supposed to be a reward, but it clearly is and is clearly viewed as such by many experienced editors, and if we start promoting this way, there's going to be a few disgruntled editors about. -- Heligoland 02:05, 4 February 2007 (UTC)

You're missing the point. This isn't about 'fair'. And this isn't about disgruntled people. This was about benefiting the encyclopedia by helping us identify masses of good admins without having to wait until they had 3.000 edits that we could assess. If this becomes about finding a backdoor for a few controversial old-timers to get the reward that the community has seen fit to deny them, then it is doomed.--Doc^g 02:18, 4 February 2007 (UTC)

We need to find a way to get those people who need the tools through the RfA process, dedicated users who have been working away in a corner for many months, improving the project. It's not a bad idea picking newer users to be admins, but there's a hell of a lot of good users who's knowledge can't be replaced by a new user with a month of editing and a few hundred edits. -- Heligoland 02:34, 4 February 2007 (UTC)

The ONLY way of convincing the community to sysop people it knows well, is to convince the community that it wants to sysop them. There's no other way. Actually, if someone with 2 months and 1500 edits is tried out as a probationer, and the community likes what they see and confirms it, there is no reason to suppose they won't be as good as someone who has blotted their copybook in the past. Now, if a probationary scheme allows the community to decide to give some older hands a chance, I've no objections. But I'm uninterested in RfA reform if that is the main purpose.--Doc^g 02:39, 4 February 2007 (UTC)

Wow, this last thread made me totally swing the other way about supporting this, in which I will not. semper fi — Moe 02:30, 4 February 2007 (UTC)

Fine, whatever. But if you don't give reasons, your opposition won't count for much.--Doc^g 02:39, 4 February 2007 (UTC)

But of course it won't count for anything. I bet you a user with a clean block log and 1,000 edits could say oppose and it mean more than mine. semper fi — Moe 02:48, 4 February 2007 (UTC)

Well, if he gives a good reason, and you give none, then rightly yes.--Doc^g 02:50, 4 February 2007 (UTC)

I meant that he didn't give a reason either. Both of us simply said oppose. I guess in your eyes his would be more valid. semper fi — Moe 02:54, 4 February 2007 (UTC)

No, why should it me? Actually, I ignore all unreasoned !votes.--Doc^g 03:01, 4 February 2007 (UTC)

*sigh* I guess you didn't catch the comparision I was trying to make with the simple !vote to trial adminship, did you? semper fi — Moe 03:04, 4 February 2007 (UTC)

Now, if a probationary scheme allows the community to decide to give some older hands a chance, I've no objections. But I'm uninterested in RfA reform if that is the main purpose. With all due respect, I think this proposal does both: some admins will look for newer folks (presumably those working in an area, like CfD, or image copyright problems, where they think help is badly needed); some will endorse more dented users who they think will pass the initial process (that is, no "compelling reasons" for a 'crat to reject them.) Some will pick editors who with whom they've gotten to know during a wikiproject; some will be open to anyone meeting the criteria they lay out (experience in different areas, no blocks, whatever). In short, we're likely to see a whole bunch of different types of editors going through trial adminships. -- John Broughton (☎☎) 23:25, 4 February 2007 (UTC)

On objections

1. It's got cabal written all over it

I can just see apprentices being nominated by an admin for their beliefs and outlook rather than their ability to help run the project.

Well, if you believe that the current 1000+ administrators are a cabal, then sure, they'll be nominating more of themselves. If you believe, as I do, that there are some cliques and there are some admins who support each other, but that by and large most admins just do their own thing, then no, letting existing admins replicate themselves probably won't change the balance here.

More importantly, the proposal can be tweaked to address this concern. First, it's probably not a big deal to limit any admin to no more than one endorsement every (say) three months. Second, something I'm going to propose below, the proposal can be modified so that if an admin screws up in who he/she endorses, that admin loses the endorsement right for a while.

2. It's much too personal

RfA is sucky enough without making rejection much more personal when an admin overlooks you for apprenticeship or turns you down flat.

We have no idea how the "market" for trial adminship will develop, but I'm willing to speculate. First, I think admins will make it clear what their criteria are (edit counts, no blocks, etc.), and won't "flatly reject" anyone who meets those criteria. Second, I'd guess that editors interested in becoming an admin will put something on their userpage (edit boxes are available now) that not only says they are interested, but gives a profile - articles they're proud of, work in (say) the 10 areas of most interest to admins looking to endorse someone, etc., and that some admins will use this to approach candidates. Third, as noted above, the standard RfA process will continue to exist. And fourth, it's a big world out there - I refuse to believe that any editor who is truly qualified is going to be "turned down flat" if an editor who meets specified (by each admin, not standard) qualifications approaches (say) a half-dozen randomly chosen admins. (If that happens, then the project has been chosing all the wrong people to be admins).

3. Due to 1&2 above, it's giving admins far too much power to pick and choose who they have as admins in future.

What this project needs is another 1000 or 2000 admins, and soon. That would wipe out every admin-related backlog we have, would allow admins to be much more proactive with sock puppets, to do more complicated investigations, to more actively go after vandals - in short, to make this place work a whole lot better. My guess is that the system could well end up sweeping up just about every qualified editor out there - if even half the admins endorse just four editors in the first year, that's (I'm guessing) at least four times the current rate. In other words, it could well be the reverse of what you say, that adminship truly becomes "no big deal".

4. And because of 3, above, there's a real chance apprentices could be victims of association,

imagine an apprentice who was nominated by an admin who during the apprenticeship is desysopped, they stand a real chance of being opposed come confirmation time. What happens in those circumstances anyway - would the apprentice admin be desysopped or not ?

First, there are what - five or six admins desysopped per year? Second, the point you raise should be addressed by a changes in the draft policy:

If an admin is desysopped, any trial admin that is being mentored by that desysopped admin either (a) will be mentored by another admin, if one volunteers, or (b) will become a regular editor again, and will be the exception to the "only one trial adminship" rule.

(I'd add that to the policy, but this is nit-picking; let's save these kind of details after the broad policy is further discussed.)

5. Bureaucratic discretion is going to open up a whole new can of worms,

if we're still proposing a limited number of apprentices, it raises the question over cabalism and/or favoritism of prospective apprentices or admins, which seriously cannot be good.

This proposal gives bureaucrats only the discretion to decide if "compelling reasons" have been raised or not when the endorsement an editor is being considered, prior to trial adminship beginning. If we can't trust bureaucrats, exactly who can we trust, to decide this? Bureaucrats can't limit the number of new admins unless they get extraordinarily picky (e.g., "only used edit summaries for 92% of major edits - FAIL").

And again, there is the regular RfA process. What I really fail to understand, Heligoland, is that you had 33 support votes in your December RfA, before you withdrew, and while I hardly know most of the 1000+ admins by name, I spotted several of them who gave you support endorsements. If this proposal passes, all you'd have to do is drop a note on their pages saying "With the new trial admin proposal in place, I hope you'll consider me if/when you start doing mentoring, given your vote in support of my RfA in December. Thanks." No admin who supported you is going to come back to you and say "I've decided I don't want you as an admin". In fact, given the suggested wording, no admin actually has to come back at all, but I'd bet that a couple would, if only to say that they would certainly be happy to be a mentor to you within the next couple of months. Plus you're just about the perfect candidate for a trial adminship - the only reasons you were opposed in the December RFA were basically lack of experience - in the eyes of some editors - and that's absolutely not a compelling reason (in my mind, and I'm comfortable saying that I don't believe that a bureaucrat would buy that either, particularly two months later). So I urge you, for both personal reasons and the good of this project, to reconsider your objections. (And yes, I concede that you've raised a couple of goog points; these can be addressed by tweaking the proposal.) -- John Broughton (☎☎) 22:57, 4 February 2007 (UTC)

I think that was an amazing reasoning, John, and I think this belongs on the RFA talk page also. I was also opposed like Heglioland, because of lack of experience. --TeckWiz^ParlateContribs_@ 23:19, 4 February 2007 (UTC)

John - if you think some of the admins who supported me in December would mentor me, why hasn't one come forward and just nominated me now. The theory is fine, but it's how it's going to work in practice that I'd like to see (and I would like to see a trial go ahead though for fairly obvious reasons, I'd be quite happy to recuse myself from the program). -- Heligoland 23:35, 4 February 2007 (UTC)

• Pretty much sums up my thought on this proposal. Any finer provisions to remedy this would be shot down as instructions creep, so I'm pretty skeptical that this would even take off in the first place. - Mailer Diablo 11:26, 7 February 2007 (UTC)

Two changes for consideration

Limiting the number of mentorships

Proposed addition:

After an admin has formally endorsed an editor and that editor is being considered by the community for a trial adminship, the endorsing admin may not endorse another editor within the next 90 days.

Discussion: This limits any admin to mentoring four trial admins per year. The primary reason for this restriction is provide some assurance to those who might worry about an admin rapidly building his/her own personal cabal by doing 12 mentorships per year. -- John Broughton (☎☎) 23:16, 4 February 2007 (UTC)

Instruction creep.--Doc^g 00:03, 5 February 2007 (UTC)

I like it, actually (or something close to it, like 6/year). If the point is mentorship, then there should be some sort of relationship there, not just racking up points on the Wikipedia:Administrator family tree. Especially as this has the potential to make the nominator position much more powerful on the eventual confirmation RFA, they should be doing a good and thorough job. -- nae'blis 16:36, 6 February 2007 (UTC)

Losing the right to endorse editors for trial adminship

Proposed addition:

If a trial admin has his/her adminship revoked by a bureaucrat, the endorsing admin loses the privilege of doing new endorsements of editors for six months. If someone who has become an admin through the trial adminship program is desysopped within a year of gaining full adminship, the endorsing admin also loses the privilege of doing new endorsements of editors for six months.

Discussion: desysopping of a someone, whether a trial admin or a full admin, is not trivial. If this happens, it indicates that the endorsing admin needs to thoroughly review his/her selection process, and would benefit from some time to consider what went wrong, and how other admins are deciding who to endorse. Removing the privilege of endorsement provides the admin with that time, since he/she will not have to evaluate candidates or do mentoring for a while. -- John Broughton (☎☎) 23:16, 4 February 2007 (UTC)

Gah, no. :Instruction creep, and paranoia --Doc^g 00:04, 5 February 2007 (UTC)

I second that. Let's assume good faith -- Selmo ^(talk) 02:24, 5 February 2007 (UTC)

Good idea, but shouldn't be an RfA workaround

At that point, the user gets regular adminship unless (1) at least ten registered users (per suffrage above) object to that, citing specific actions as their reason for objecting, in which case a regular RFA decides. At that point, the user gets regular adminship unless (1) at least ten registered users (per suffrage above) object to that, citing specific actions as their reason for objecting, in which case a regular RFA decides

I don't think this should be an RfA work around. RfA has high standards for a good reason. Sure some oppose !votes based on opinion, such as the editor doesn't have a billion edits but the main purpose is to evaluate the editor's ability to administrate the project. Through this method, they get adminship scott free.

I would, however, like to see this as an (optional, perhaps?) prerequisite for RfA. In the nomination, the candidate can include sysop actions during the trial period, which would aid !voters decisions. Obviously, it shouldn't be as complex nor should it have high standards, but it should have an expiry date. -- Selmo ^(talk) 02:50, 5 February 2007 (UTC)

RfA is broken though, that's the problem, and it doesn't matter how much proof you give that a user will be a brilliant admin, if someone wants to oppose for a stupid reason, nobody deals with it. If quite a few people jump on, the user isn't promoted. Really rather stupid as those admins that have the most support are the arrogant ones who end up being desysopped. We need, I think, RfA clerks who will remove baseless oppose !votes and force support and oppose reasons to be relevant to adminship. -- Heligoland 03:06, 5 February 2007 (UTC)

• This idea has its pros and its cons, but to me, the cons seem too great. It seems like an easy back door into adminship. The RfA process does have flaws, yes, but that doesn't mean it should be this easy. There's no evaluation or peer review... all an individual needs is to be good in the eyes of one admin. Users are not infallible, and neither are admins. That's why there's a peer review process to begin with. .V. _[Talk|Email] 05:01, 5 February 2007 (UTC)

• I agree that this should not be an RFA workaround. That process has been tested and challenged, deleted and reconfirmed over years of effort by thousands of users. What the mentorship/trial adminship should do, in my opinion, is answer the concerns of "will go crazy using the tools", "not enough experience in admin-related duties", "not sure that this user can handle dealing with blocks appropriately". It should decided NOT be an end-run. -- nae'blis 16:39, 6 February 2007 (UTC)

Against

I agree with many objections that have been previously raised. Shaundakulbara 05:44, 5 February 2007 (UTC)

Admins are not Bureaucrats

This is the fundamental flaw in this proposal. Administrators are janitors, not Bureaucrats. This process seems to indicate that administrators are somehow better at choosing which user can and can't be trusted with the tools. Sure, the process has the initial probabation endorsement in which past deeds can be brought up, but if the user merely makes minor edits and has caused no trouble, there is really no way to prevent them from getting the tools. Thus, we are assuming that admins are better than regular users at picking who should have the tools. This is not the case. Admins are janitors. They do not have an inherent ability to make good decisions regarding whether a user can be trusted with Administrator tools.

The notion that we desperately need more admins to clear out the backlogs is silly. If backlogs were an absolutely critical concern, they would be cleared out already. Since there are undoubtedly dozens of admins online right now and few of them are working on backlogs (or so I assume), I think it's safe to say that the backlogs are not a critical priority. If they are, then current admins should be encouraged to clear them out.

The current RFA process does a fine job. We shouldn't make a new process - instead, use the old process more effectively. Increase the number of nominations, increase the number of eyes looking at each nomination, and increase the quality of the pool of candidates (to increase the success rate). All three of these things would be easy. Increasing the number of nominations is trivial - just be on the lookout for good admin candidates and nom them. Increasing the number of eyes should be trivial - we have thousands of editors who never look at RFA. Get them involved in the process. Increasing the quality of the candidates is entirely trivial. The quality of the candidates is already increasing every day. The core group of users maintaining Wikipedia is constantly getting better at understanding and following policy.

If a trial admin can't get through RFA, then they obviously are not admin material. If they can get through RFA, then RFA should be used. RFA is not broken. --- RockMFR 05:50, 5 February 2007 (UTC)

Additional concerns

While in principle I understand the desire to streamline adminship, I cannot agree with this as the means to do so. Although my own RFA recently closed no consensus, I would not be comfortable turning to this procedure as a route to the mop; I would rather address the issues the community saw with my participation and its breadth. My objection is not without reason:

• 1. Insufficient oversight. Why do good articles not have the kind of editorial clout that featured articles do? Because they are less rigorous, decided ultimately by one editor. This program is based heavily the decisions of one admin. Admins are able to have the role they do in policy enforcement and dispute resolution because they have gone through a process that mandates some measure of trust and respect from the community. While I would stop short of directly considering adminship to be akin to "featured" content, we should ask ourselves carefully if the project benefits from creating a "good admin" program.
• 2. Like chooses like. Admins are people, too, and despite our best efforts, admins have their own biases and preferences, especially in areas of Wikiphilosophy. A deletionist admin might view with appreciation the work of others of like mind, even those who might be more aggressive than themselves. But it doesn't take resorting to a slippery slope to envision that this process, especially over repeated generations as once-trial admins get the mop (and the ability to have their own trialees), could yield administrators with more exteme views that would -- or should -- ever survive an RFA. This isn't really the same thing as a cabal, but it is still problematic.
• 3. Potential for abuse. It is easier to trick one admin than to trick the community. No one is going to pretend that there have not been problems with seated admins who go through the full process. It is inevitable that there will be more problems with trial admins. Recent evidence suggests that there is motivation for (and active interest in) fostering abuse. And of course, there are ways admins can cause significant problems.
• 4. Effects on bureaucrats. At the same time this places an increased workload on the shoulders of the bureaucrats (who must add the tag at the beginning of the trial, then add or remove the tag at its end), this decreases the role of 'crat discretion in promotion, instead vesting more of that power directly in admins' hands (and expressly restricting the reasoning 'crats are to employ in the decisions they do make). I am not certain this is what is intended for the bureaucrat system. Also, as a technical objection, bureaucrats are not to directly demote, which presents a difficult situation in the event that someone is considered to have failed their trial period.

If there is a problem with RFA (and I am not entirely certain that there is, despite my own failure to achieve a consensus for promotion in part due to limited editing), I can only feel that an effort to improve RFA is the answer, rather than the creation of a new process that might be seen as an end-run around established policy.

Respectfully, Serpent's Choice 08:18, 5 February 2007 (UTC)

Focus

There are two different issues getting conflated here. The first is that RFA is (allegedly) too harsh on novice users and that we don't give people the benefit of the doubt if we don't know enough to judge them by. Fixing this would give us quite a lot more admins. The second is that there are a few long-standing users that don't have a snowball's chance at RFA, and that believe that this is RFA's fault as opposed to their own fault. This cannot really be fixed as long as RFA is a community process. Therefore it is probably best to focus on the former, and restrict these trials to people who haven't been RFA'd yet. >Radiant< 11:04, 5 February 2007 (UTC)

Actually, it occurs to me that what this proposal might simplify to is an effort to limit the use of certain arguments in RFA. The trial adminship allows for a discussion before the month-long mopping, but that discussion prohibits "speculative reasoning" or "'lack of participation in process X'," requiring solely arguments "based on the user's past deeds." Presumably, this has the goal of being permissive to novices while continuing to act as a barrier against longtime contributors whose past contributions have been deemed inappropriate for the office. Then, once the trial has been granted, full adminship (under this proposal) seems virtually guaranteed except in the case of gross failure to uphold the position or willful malfeasence. Thus, it is that first evaluation that is akin to RFA ... and that would become like RFA is now, except to whatever degree those restrictions were held to by commentors. Without directly addressing the merits of these restrictions, would it be better to simply attempt to reform the criteria of RFA without the implementation of a further layer of bureaucracy (no offense to the bureaucrats!)? Serpent's Choice 11:58, 5 February 2007 (UTC)

• The problem is that it's not really possible to reform the criteria of RFA, since it is a place where people give their opinion. The more an outside party (e.g. 'crat) determines which criteria/opinions are "valid", the less it becomes a community process. And if you "forbid" certain kinds of reasoning, that simply encourages people to not state their reasoning and give a blank "oppose" vote. It's a matter of culture; quite a lot of people want to watch RFA and comment on candidates, and they quite often see candidates they don't actually know, so this compells them to think of a way to judge those anyway. >Radiant< 12:30, 5 February 2007 (UTC)

• I agree with Radiant! here. I'm not sure that, say, badlydrawnjeff should be ineligible for this, provided he can find a sponsor, but the main problem is addressing the "not enough experience" canard. Addressing users who are perceived as being too hard on certain candidates is something that should be done within the context of RFA proper. -- nae'blis 16:42, 6 February 2007 (UTC)

Checkuser Safety

I'm a little concerned that as this proposal centres around editors who will be new, have low edits counts or where they've had little interaction with other users, it's an ideal way for sockpuppets of banned users to be granted access to everything they need avenge their block. I know the outrage this is likely to cause, but we would be dealing with users who may not have made enough edits to raise suspicion. Any thoughts ? -- Heligoland 13:09, 5 February 2007 (UTC)

• Yes, but those banned users would still need to do enough good work first (thus in effect they're helping us) and I'm not sure how they could avenge their block with admin tools. What are they going to do, block the people they don't like? Delete their userpage? Yes, they could, but they would find themselves reverted, blocked, censured and deopped very quickly, and not necessarily in that order. >Radiant< 13:14, 5 February 2007 (UTC)
  • Your assuming they would go on a huge damage spree, rather than use the tools discreetly to cause trouble and damage that could take weeks or months to fix.-- Heligoland 14:28, 5 February 2007 (UTC)
  • You're also discounting the fact that it becomes very easy to acquire multiple admin sock-puppet accounts. Normally this is difficult because you have to create multiple accounts with 1000's of constructive edits. This process shortcuts that. This means that it's no big deal to lose the admin account either if you do want to go rogue. Why we want to be giving admin status to people that haven't proved themselves is beyond me. —Doug Bell ^talk 05:56, 6 February 2007 (UTC)

No RfAs allowed...

I don't like this statement:

An editor who has already had an unsuccesful RFA is not eligible for a trial.

There are plenty of admins who had there first RfA failed. However, subsequent RfAs led to the mop. Secondly, lots of great admin candidates currently on review have plenty of support (75%+) dispite failed RfAs. The majority of users with zero RfAs are newcomers.

As I said above, this should not lead to permadminsip. All users who get access to the tools via this process, no matter how good they were, should be desysoped. They can later post their admin action log in the introduction statement of an RfA when applying for the tools permanently.

In short, excluding editors with failed RfAs are unfair, especially since some of them (like myself) applied for adminship before this came out. Thus, they never get a chance to try this out. -- Selmo ^(talk) 20:20, 5 February 2007 (UTC)

I'm fine with this not resulting in a "permanent" adminship, as long as the trial adminship is of indefinite length. Friday (talk) 20:24, 5 February 2007 (UTC)

• It's for the same reason that an article that has been on AFD is not eligible for PROD. If there was an RFA discussion and there's no consensus for promoting the candidate, then if somebody decides to "trial" the candidate anyway people will see that as a backdoor promotion, ignoring consensus, and become upset. >Radiant< 10:44, 6 February 2007 (UTC)
  • Even when the candidate's "failed" RFA was when they had 37 edits? This seems like instruction creep... if the person's previous RFAs are disclosed and someone still wants to mentor them and no one throws a spanner of objection, what's the problem? -- nae'blis 16:44, 6 February 2007 (UTC)
    • Agreed. There should be a threshold for change upon more experience. Take for example the many newcomers who wind up jumping the gun not knowing what to expect, only to find out that 200 edits is not enough experience, and their RFA has been taken off as "failed". 5,000 edits later, they have a greater grasp for policy, and an understanding that mirrors senior admins. Limiting trials in this manner will give some the edge over others in a seemingly unfair light. SynergeticMaggot 20:39, 6 February 2007 (UTC)
      • Okay, good point. If a user fails RFA for lack of experience, trialling would be very appropriate. The idea behind this was that some long-term users fail RFA because of controversy, because many people in the community really don't trust them. I suppose the canonical example is User:Wik (now banned, but earlier he was a good example of someone with an extreme amount of edits and lack of community trust). These people should not get a trial run. >Radiant< 09:18, 7 February 2007 (UTC)

Are the mentoring admins responsible for those they mentor. If a mentor has two consecutive apps that crash and burn soon after gaining adminship do they lose the ability to mentor. —comment added by IQ72^(t/c)

Seriously flawed

Scenario: A vandal creates a new account. The threshold, strongly worded in the proposed policy, requires that any editor can apply, and that only compelling reasons not to give probation can be considered, not lack of experience. So the vandal gets to be a probationary admin in a couple of days with little constructive work required. At this point, with little invested, the vandal could go rogue and muck things up pretty well before getting shut down. But let's say they have just a little more patience and spend some time being a good admin. It's not much investment, as the only real threshold is to use the tools enough so as not to be judged as having barely used them. A month later, with a few hours invested, the vandal is an admin.

Problem: By making people admins with such little effort required on their part, the "value" of the adminship is greatly reduced. The effort it takes to acquire the record of participation is one of the things that keeps most people from rash behaviour once they become admins. We still get an occassional problem admin as evidenced by those admins desysopped, but the ratio is pretty low. I see that ratio going nowhere but up if this proposal becomes policy.

This is simply a bad idea. —Doug Bell ^talk 06:08, 6 February 2007 (UTC)

Agreed with Doug. DS 22:31, 6 February 2007 (UTC)

I disagree. I think this could work, once all the details are hashed out. Your scenario disregards this part of the first sentence: ...if the editor can find a serving administrator who is willing to endorse the editor's application... The way I interpret the proposed policy, as currently written, individual admins will be able to set their own criteria for who they will take on as their trial admin. I think this is a workable alternative to the current messy RfA process. I also think, as suggested above, a trial should be conducted to see how viable this would be. I volunteer to be a trial admin if anybody will endorse my application.--William Thweatt ^Talk | ^Contribs 08:59, 6 February 2007 (UTC)

That is, in fact, part of the problem with this proposal. Every admin (indeed, every RFA participant) likely has a differing viewpoint of what the requirements for the mop should be. While consensus has changed in some regards at RFA over the years, all admins have been promoted through the same process. This proposal would unmake that, with some would-be mentors cleaving to a harsher standard than RFA, and others to an almost trivial one (as evidenced by all the "my RFA guidelines" userspaces pages that get linked occasionally). How quickly would it take before a dispute resolution issue reached AN/I or even ArbCom with claims that a block or article deletion was invalid because it wasn't done by a "real" admin or that it was done by an admin who was promoted by a "diploma mill" mentor?

Furthermore, in my mind, the idea that anyone arguing in favor of this trial process should be volunteering to be given the first (or any!) trial adminships is an archetypal conflict of interest (and there have been several such requests already). Serpent's Choice 09:13, 6 February 2007 (UTC)

Nominations

Unless I'm misreading it, this system would depend on self-noms. One could say "Hey, you look like you'd make good use of admin tools, go bug an admin about trial adminship", but it seems an awkward way to go about it compared to a straight nomination. Would/should one be able to nominate another user for trial adminship? Oldelpaso 21:29, 6 February 2007 (UTC)

Comments of Natl1 (Talk Page) (Contribs)

I think this is a great idea, trial adminship. Also, I think that we could make an admin to be able to downgrade a trial admin by doing the following.

• Creating a new access level (Wikipedia:User access levels) called "Trial Admins" with all the same tools as "Admins".
• Adding "demote trial admin" to permission list which would allow admins to demote "Trial Admins" only.

Cheers--Natl1 (Talk Page) (Contribs) 22:41, 6 February 2007 (UTC)

I think keeping a list like WP:LA of current "trial admins" would be better than pushing the devs into adding new MediaWiki functions. semper fi — Moe 01:48, 7 February 2007 (UTC)