Wikipedia talk:Requests for adminship/Archive 184

Archive 180

←

Archive 182

Archive 183

Archive 184

Archive 185

Archive 186

→

Archive 190

So let's talk about it here

Yeah, he screwed up, but ArbCom, yet again, blew it. When I pointed this out, or questioned the decision, I was directed here. Of course, I've been here before, and know the fallacy of such a line of thinking. Whatever helps you sleep better at night. Hey Majorly? How did this page work out for you? Right. It didn't. Fine example of yet another solid admin who, when not receiving the support of the community in doing the right thing and taking care of contentious, tendentious editors, finds himself kicked to the curb by a bunch of ass kissing wet shirts who wouldn't know an encyclopedia from a paper weight. There. I done brung it here. Hiberniantears (talk) 04:02, 15 September 2009 (UTC)

@hen you were told to take it RfA, I can only guess that he was talking about another RfA for WMC. This isn't really the right place for discussing the removal of adminship. \ Backslash Forwardslash / (talk) 07:33, 15 September 2009 (UTC)

Yeah, thank you for your rant. I hope you feel better now. I'm sure 'crats will now rush to give WMC the tools back. Seriously though, try an actual RFA if you feel that strongly about it.--Atlan (talk) 07:40, 15 September 2009 (UTC)

Vote weighting

Just curious, so I thought I'd ask here. Do you suppose that the votes of long-time, established and respected editors carry more weight around RfA in the eyes of bureaucrats, or are all legit, reasoned votes weighted equally? I can think of numerous names of regular voters here (I won't bother mentioning them, we all know who they are) who no doubt produce extremely valid points the vast majority of the time.

Imagine this hypothetical (and I do mean that, I'm not going off any past experience whatsoever) scenario where there are numerous oppose votes by registered editors with decent reasoning, but there is an equally lengthy list of established RfA voters with convincing supports. Would that hypothetical RfA be closed as successful due to the respect, perhaps subconsciously, held by the closing 'crat towards those editors? I appreciate there are a lot of variables in this scenario and it's hard to place accurately in your head, but I'm interested in hearing others opinions. My apologies if this has been brought up before.

Regards, --—Cyclonenim | Chat  18:53, 15 September 2009 (UTC)

Obviously true. Hence the fact that the instructions repeatedly say it's not a vote count. Ten opposes from established editors more than counter 30 supports from three-day-old SPAs. The only alternative would be minimum-edit-requirements for voting, as it done with Arbcom elections, but that's more hassle than it's worth. – iridescent 18:57, 15 September 2009 (UTC)

Whilst I appreciate the obviousness of the case with SPAs, I was suggesting a scenario a little less clear. For example, registered editors = respected editors from elsewhere on the Wiki who rarely vote at RfA, and established editors = those who regularly vote here such as yourself. Maybe it's just me overthinking the situation/the situation doesn't exist. Regards, --—Cyclonenim | Chat  19:06, 15 September 2009 (UTC)

I would sincerely hope that a bureaucrat holds the opinion of any editor in good standing equal to that of any other editor in good standing. Granting extra weight to the !votes of "respected" editors, or worse still - RFA regulars - kind of brings up the whole specter of cliques and cabals. I really, really, really hope not. Shereth 19:48, 15 September 2009 (UTC)

Second that, but I feel the effect might not be voluntary. There may not even be an effect, I can't be bothered to research it, but if a bureaucrat holds certain editors in very high regard, their opinion on an RfA might influence the closing 'crat's decision. Regards, --—Cyclonenim | Chat  20:03, 15 September 2009 (UTC)

I was going to say that I trust our current bureaucrats not to hold any of us in very high regard that their opinion is influenced. But I wonder if I ought not bite my tongue with that comment .. :) Shereth 20:21, 15 September 2009 (UTC)

An editor in good standing that has a crappy RfA argument is easily and soundly trumped by a brand new user that can show that the candidate has posted numerous copyvios, and that's exactly the way it should be. EVula // talk // ☯ // 20:08, 15 September 2009 (UTC)

^ Trust the crats, they're smart people or they wouldn't be crats (Also the only people who can speel their title).--Patton123 (talk) 20:16, 15 September 2009 (UTC)

EVula's comment is exactly what I would have expected to hear and what I trust our crats to do. No matter how well-reasoned someone usually !votes, in those cases where they don't, their !vote is not worth more than any other !vote with a weak or no argument. A crat who would judge !votes based on !voters rather than the arguments is not worth their money (hypothetically speaking) but I am confident that none of them would do so. Regards SoWhy 20:46, 15 September 2009 (UTC)

Yes, but I would say that an experienced editor with a strong argument would have at least a bit more weight than a new user with a strong argument.--Unionhawk ^{Talk E-mail Review} 20:58, 15 September 2009 (UTC)

Where a !voter's experience comes into play is in influencing other editors, rather than the 'crats themselves. EVula // talk // ☯ // 21:01, 15 September 2009 (UTC)

(edit conflict with EVula, saying exactly the same thing in twenty times as many words) I do like the mental image of our bureaucrats sitting there with a calculator and individually weighting votes, but I have to say I doubt that happens. A tight RfA close is always going to be made on the subjective definition of "strength of argument", and unless there's some serious influx of low-edit-count SPAs I doubt the experience of individual editors in the mass of supports/opposes tends to make a difference - it's the overall relevance of the points raised and the level of consensus that the point is an important one.

That said, the experience and reputation does have an effect on the power of their vote: other people participating in the RfA are rather likely to take the opinions of editors they respect into account when making their own votes. I know that if I read through an RfA and find an oppose by an editor I respect the judgment of, I will take the concern seriously when deciding on my own position. Conversely, if I run across an oppose by an editor I know to have a vastly differing opinion from me on a particular point, or to be generally overly picky, or just someone who's clearly unaware of key facts - I'm far less likely to care what they think. So rest assured that people's standing in the community does affect the force of their argument: it just may not be directly via the 'crats, but via the decisions made by other RfA participants based on your view. ~ mazca ^talk 21:08, 15 September 2009 (UTC)

(edit conflict) With an experienced !voter making a strong argument, it's more likely to have "per X" !votes than with an inexperienced or new user. But such !votes do not change the strength of the argument itself. Similarly, multiple "good guy" !votes by experienced users are not the least bit more convincing just because they are made by experienced users. Quite the opposite even maybe because one would expect them to know better... Regards SoWhy 21:12, 15 September 2009 (UTC)

Active admins

Number of active admins

As we touch 890 active admins today, I took the time to create a graph dating from two years ago showing the number of active admins at any given time. As the graph shows, we peaked in January 2008 and have gone downhill since then. Both in terms of cycles and absolute figures, we have never had this low of a level since before August 2007. MBisanz ^talk 06:42, 13 September 2009 (UTC)

Thank you for more of this. Does this mean, then, that the time has come for me to scout out fifty non-admins for the purposes of nominating them for adminship? @harej 06:51, 13 September 2009 (UTC)

Yes, that is what it means, just make sure they are qualified. MBisanz ^talk 06:58, 13 September 2009 (UTC)

"I am experienced and I am sane". Harej, IRC, whenever the hell he ran, replying to me a request in seven words why I should support. Keegan (talk) 08:33, 13 September 2009 (UTC)

Okay, but what about productivity? It seems to me that many admins have over the years specialized and gained experience in particular areas and also that admins who do not run bots now have a host of tools and templates at their disposition that make the technical part of many tasks from closing an Afd over deleting a bunch of pages to declining an AIV or RPP report easier.--Tikiwont (talk) 08:35, 13 September 2009 (UTC)

I actually was going to raise that point but since I always do I waited for someone else. A lot of what was done a couple years ago was non-botted and presently involves run of the mill tasks. Having to hit block, protect, delete isn't quite as necessary in addition to closing discussions, which includes RM (of which harej does many) that don't show up in the +sysop log. Keegan (talk) 08:55, 13 September 2009 (UTC)

That's a very striking graph. What happened about the beginning of 2008 to make the curve change direction so markedly? JohnCD (talk) 10:46, 13 September 2009 (UTC)

Eric Goldman's claims that will fail by the end of 2010 look relevant. Goldman alleges that: there is a conflict between high quality and freedom of editing; WP faces a labour shortage for dealing with vandalism and other chores. The "labour shortage" is apparent in the graph of active admins. We need to know what % of the total admin workload is accounted for by each admin task, as see what is needed to reduce the major time-sinks. --Philcha (talk) 11:38, 13 September 2009 (UTC)

It's back to the levels around the time I was made an admin. I've noticed a considerable shift in how admins do their adminning over this time, particularly in terms of anti-vandalism, which is something I tend to specialise in. I presume it takes up a large proportion of admin time. Back in the day we used to block AOL IPs for 15 minutes at a time, and schools until the end of the lesson. No more! Now we block proxies and schools that spew vandalism at us for years at a time. We even have whole school districts rangeblocked. This has reduced the vandalism enormously. We have auto-disallowed 25,000 instances of poop vandalism in the last six months, not to mention the thousands of other edits now being automatically filtered. We've cut down the number of moves a pagemove vandal can get away with, in fact we've move-protected most of the tempting ones. We have most of the major open proxy ranges blocked, Tor is completely prohibited, and other open proxies get blocked at an unprecedented rate. Any backlogs at AIV have been reduced dramatically. The biggest delay we have these days is caused by all the content disputes and non-vandalism being reported there. We use semi-protection far more liberally on high-vandalism articles, particularly on BLPs and articles like Fat. We've basically got rid of template vandalism on the front page and elsewhere, we've got the image blacklist and title blacklist and two spam blacklists. We've got global blocking and more checkusers blocking repeat vandals. And then there's all the anti-vandalism bots, adminbots, huggle, and twinkle. I've found that the vandalism is so well covered these days that I can get back to focusing on other things, like content. We have lost some freedom of editing. We've found it necessary to anonblock large numbers of IP addresses, and we are far less tolerant of silly edits. But I'm not sure if having more active admins would help lower that number. More sensible admins, doing less, would be a good thing though. -- zzuuzz ^(talk) 12:29, 13 September 2009 (UTC)

We are better off with fewer admins. In an ideal world, we wouldn't need admins at all because ordinary users could always be trusted to do the right thing. Jehochman ^Talk 13:14, 13 September 2009 (UTC)

Im also curious about the notch in the graph around February 2009. I was pretty active then, but I don't remember any incident that caused 40-ish admins to resign and then 40-ish to come back again some weeks later. Was it an ArbCom decision that had only a temporary effect? -- Soap ^Talk/_{Contributions} 13:20, 13 September 2009 (UTC)

I think it was a combination of a few people burning out, as well as a couple of ArbCom decisions, and some frustration with BLP issues. Hiberniantears (talk) 13:38, 13 September 2009 (UTC)

Well, we lost a few due to Jimmy violating blocking policy and then following up by calling people names, but that's water under the bridge now. KillerChihuahua^?!?Advice 13:41, 13 September 2009 (UTC)

That too. Forgot all about that. I would submit that overall, 2009 has been a pretty down year for Wikipedia. Hiberniantears (talk) 14:20, 13 September 2009 (UTC)

Yeah, the irony is funny. Jimbo beats up on the admin corps, arbcom puts them under a microscope, not to mention heckling from the sidelines on both internal and external fora, and everybody puzzles at why admins decide to give up their huge paycheck. Short Brigade Harvester Boris (talk) 15:34, 13 September 2009 (UTC)

Soap, check out the history of WP:FORMER around that time to get an idea. December 2008 and January 2009 saw many administrator resignations, most of which (IIRC) were in protest to the various issues of the project (BLP, admin recall, etc). Lara 18:50, 14 September 2009 (UTC)

I would like to see that graph going back another year and having a scale starting at 0. With the current limited scale a variation of only 11% looks like a shocking decline, but it could be simply an 11% boost during the year then going back to regular levels. I cannot really tell from the data presented. Chillum 15:19, 13 September 2009 (UTC)

It would also be interesting to know the ratio of admins to users: one user in every (X-amount) is an admin. iMatthew ^talk at 15:23, 13 September 2009 (UTC)

I think I just figured it out. If I'm correct, there is approx. 1 active administrator in every 167 active registered accounts. There is approx. 1 administrator in every 6242 registered accounts. iMatthew ^talk at 15:29, 13 September 2009 (UTC)

Fair point chilum, but seen in comparison with overal internet useage levels the trend is even more alarming. Projections suggest this has increased by about 15% these last 18 months, and if we go back further the discrepency becomes even more shocking, though I guess it depends on how you view the efficiency improvements zzuuzz talks about. Imatthew , what date was it when we had one admin per 167 users? FeydHuxtable (talk) 16:05, 13 September 2009 (UTC)

According to this... today. iMatthew ^talk at 16:22, 13 September 2009 (UTC)

Chillum, sadly this is the only continuous dataset that exists. Below is the extended non-continuous dataset, other than this information, there is no other information available by any source that I am aware of. MBisanz ^talk 16:39, 13 September 2009 (UTC)

Date	# of active admins
August 19, 2007	912
March 3, 2007	849
January 18, 2007	814

Thanks for the info MBisanz. If you drop by my talk page with the criteria for "active" being used I think I can make a bot to fill in the gaps. Chillum 16:53, 13 September 2009 (UTC)

RFAs brokenness and the urgent need for a solution

A common view is that the break down of RFA is a primary cause of the declining active admin corps Mat illustrates above. We've talked about negativity, grudge bearing and excessive scepticism. Another problem is the stunning inconsistency in the communities expression about what it wants from successful candidates. For example, lets contrast two recent RFAs scarcely two weeks apart, in both cases with a good, friendly and helpful editor whose arguably only prominent flaw was a lack of copyright knowledge. At Chamal's RFA the candidate gave an imperfect answer on copyright. Several users said they'd prefer that in cases where he wasnt fully conversant with the relevant policy he should undertake to check with an expert before acting. Chamal did eventually commit to checking with an expert, but even before that he enjoyed a very high support %.

While at ImperatorExercitus's RFA the candidate was upfront right from the start that if asked about images and copyright he'd refer to someone more experienced. A position in line with even the harshest opposers on Chamal's RFA , yet his RFA nosedived to 60% - apart from a coupled finding fault with his AfD closures, this was almost entirely based on the complaint that admins need a comprehensive grasp of policy, totally contradictory to the sentiment expressed a couple of weeks before. Such arbitrary outcomes risk making good editors feel the community is against them, no wonder talented folk are abandoning the project.

On the need for more admins

zzuuzz outlines many reasons why admins have increased their productivity, but the fact remains we still frequently have administrative backlogs, with other work queues only kept clear due to the dedication of a small number of highly active admins. Even if we werent at risk from a chain reaction of burn outs caused by key admins dropping out due to overwork coupled with a feedback effect where more resign due to increasing pressure, there's other problems with having individuals feeling compelled to do so much work, not least the effect on the admins themselves. Most of the busy admins im aware of are young, and if their sense of duty causes them to spend more time than they'd like keeping this place ticking over, its unfair as

• they miss out on other areas of life by being on here so much.
• they especially miss out on the chance to express high spirits due to the obsession some seem to have about editors acting maturely including off the main space.
• though being an admin teaches one about mediation and the use of power, the mostly menial duties primarily only offer experience in how to be a good functionary. Not the most profitable use of intelligent young peoples time, especially compared to researching and writing good articles, which develops many of the same skills academics need to write papers that advance their carerr, or which analysts in the commercial or public spheres need to reach the highest levels where they get to advice senior leaders.

The drawback to addressing work load concerns with further enhanced productivity is that as zzuuzz admits it reduces freedom of editing. The knock on effect is to reduce the aggregate amount of editing here, the good as well as the bad.The only serious issue with more admins is that inevitably there will be more bad ones, but apparently the arbcom have streamlined the process for dealing with these and can likely make further improvements if need be.

Possible solutions.

1) Change the default result of RFA to promote, so a consensus will be needed to prevent the candidate becoming an admin. (possibly with some safeguards like a min of 3000 edits) 2) Have an intermediate rank below full admin, possibly with all the same technical powers but not expected to be fully clued up on polices so they wouldnt be expected to intervene as an admin in the more complex disputes. There will be many other workable alternatives.

On the need for urgent action.

When this has been raised previously some requested more stats and studies before any action is taken to reform RFA. In my view Mat's already provided close to the optimal level of empirical data. It might be useful to compare the active admin graph to overall editing levels, global internet use and the easily definable work queues like the XfD backlogs. But the total admin work performance is probably not quantifiable. One can look at queues ANI , but what the raw numbers wont tell us is how many disputes are being resolved in a fair way that helps the parties become collegial editors? And how many editors havent left or became increasingly frustrated as they dont trust admins to fairly address their concerns, and so havent gone to ANI in the first place? etc etc

Let this be understood: if we hold back from fixing RfA until its brokenness is obvious to every single editor, the community will suffer massive disruption and Wikipedia may even loose its no 1 ranking. Either way it will be a tragedy as our noble mission will be substantially delayed. If attempts to reach consensus for a solution are blocked, it might be an idea for a respected admin or crat to raise the issue with the foundation. FeydHuxtable (talk) 15:13, 13 September 2009 (UTC)

I don't think the decline is due to the rate of promotion on RfA. The problem may not be a lack of new admins, but our ability to keep the old ones. It is getting more and more common for admins to be accused of corruption and bias over the slightest issue, and while there always seems to be a pack of people ready to pile on such baseless accusations one cannot count on a defense. Admins have a tough job which they don't get paid for and if they are treated like shit then there are going to be less of them. Just like any other type of volunteer.

Two years ago it was not like this, accusations needed to be supported by evidence or were dismissed quickly. Now it seems that the burden of proof is always on the admin to show that the accusations are false instead of the accuser showing the accusations to be true. I know this has made me consider quiting. You don't fix a leak by adding more water, more in and more out only reduces the quality of the whole. Food for thought. Chillum 15:28, 13 September 2009 (UTC)

I agree that poor retention is likely also a primary cause, though that again is likely related in part to workload. Folk like to be part of a successfull project, and if things are going to pot as we lack sufficient hands to do the admin work with due attention some will sense this and become discouraged. I also agree it would be best not to make it easier to de –sysop admins; the only time I commented on an arbcom case I said the proposed de-sysoping seemed too harsh (even though it was a deletionist) . The process can become more stream lined without making it easier for folk to attack good admins, including ones who make the odd mistake. Im not sure what the project should do to make admins feel more appreciated, hopefully someone will think of something. FeydHuxtable (talk) 16:05, 13 September 2009 (UTC)

• Strong second to what Chillum just said there. Admins have already been promoted, having survived the scrutiny of Rfa; the onus most assuredly must be on the not-as-highly-vetted Random User providing compelling evidence for their claims of Foul Play. One puppy's opinion. KillerChihuahua^?!?Advice 16:07, 13 September 2009 (UTC)

strongly thirded. i didnt mean to sound like making it easier to desyop admins ought to be part of the solution. FeydHuxtable (talk) 16:14, 13 September 2009 (UTC)

Re the presumption in favour of admins (Chillum 15:28, 13 September 2009; KillerChihuahua 16:07, 13 September 2009):

• I've raised a case at ANI where an admin was incivil not once and several times in quick succession. I've seen non-admin users blocked for less. The presiding crat stalled, started a diversionary attack on the conduct of one of the victims, and then said effectively "too much time has passed, it's water under the bridge."
• Another admin seems to have a habit of rainsing that are hasty, excessive and quickly unblocked or much reduced.
• RFA is a process operated by humans, and therefore is fallible. So KillerChihuahua's 16:07, "Admins have already been promoted, having survived the scrutiny of Rfa" (13 September 2009) is no substitute for requiring admins to comply with Wikipedia:Administrators. --Philcha (talk) 18:19, 13 September 2009 (UTC)

It is well known that established editors get away with far more than new editors do. I understand your concerns, but it's a much bigger problem than just admins, the topic of discussion today. Dave (talk) 21:07, 13 September 2009 (UTC)

the "presumption" is in your interpretation, not in either my or Chillum's posts. We were speaking of "innocent until proven guilty" and I merely noted that Admins have passed at least one vetting which shows at the minimum that they are not vandalism-only or SPA accounts. Please refrain from extrapolating more from my post than I placed there. KillerChihuahua^?!?Advice 01:12, 14 September 2009 (UTC)

(od) My take on this (as a new admin whose RfA was described as "the most bat shit insane week RfA has seen" by more than one person): IMO, the following reforms should be implemented:

• The discussion taking place in the !voting section should be between the !voter who made the statement and the nominee. The discussion should be encouraged, but between those two parties only. Any 3rd party interference should be stopped.

I will admit, that in some cases this was beneficial, as a concern raised was responsibly addressed while I was out, so I didn't have to answer it myself. That was the exception, not the norm. The majority of the time it was petty bickering, that served only to make a hostile atmosphere. It resolved nothing.

• All allegations should be supported by evidence. Specifically raising allegations of plagiarism, falsification, etc. must be backed up by evidence, or the accuser is blocked, and their comments stricken. No more poisoning the well should be allowed at RfA.

As for the next part, how to keep admins from leaving. I'm still working my way into the duties. I've been surprised when I've pinged admins "what would you do in this situation". I've come across many situations where the admin thought action was appropriate, but was unwilling to act because they don't want to get involved in the drama that would follow. The day admins are afraid to do their duty, it is a sad day indeed. Not sure how to resolve it, but I think it's a point that should be raised. Dave (talk) 21:07, 13 September 2009 (UTC)

Those sound like good recommendations, as long as they're accepted as a package. Unless opposers align with the practice you outlined, we'd probably loose out if supporters stop engaging with the oppositions arguments. Candidates are often advised to leave it to supporters to counter oppose rationales and some dont respond to them at all. Someone posted to this Meatballlink about why its good to be defended by others rather than do it yourself. Goethe , whoes been estimated to have the highest IQ of any European, said: "Against criticism a man can neither protest nor defend himself; he must act in spite of it, and then it will gradually yield to him". As you say it does sometimes help an RFAs chances when supporters respond to persuasive but challengeable oppose reasons. I've held back on responding to comments in the past , thinking others would see the flaw in the rational, only to see a pile on for the same reason when I next check. Sometimes you dont even have to point out errors of fact or logic, just offer an alternative perspective. When opposers over play their hand it can turn the RFA if the opportunity they provide is skilfully capitalised on. On the other hand I agree that arguing against opposers often backfires, especially when the discussion becomes heated. I might write a user space essay on this one day, with examples to flesh it out. Interesting subject. FeydHuxtable (talk) 13:50, 14 September 2009 (UTC)

In response to the original post, specifically the section "Possible solutions".

1) Change the default result of RFA to promote, so a consensus will be needed to prevent the candidate becoming an admin. (possibly with some safeguards like a min of 3000 edits)

2) Have an intermediate rank below full admin, possibly with all the same technical powers but not expected to be fully clued up on polices so they wouldnt be expected to intervene as an admin in the more complex disputes. There will be many other workable alternatives.

For 1, I don't understand exactly what that means, but I think what would be a great change is to have the community flesh out some basic standards. Some things will always be up to personal preference of the voter, like the amount of content work a candidate has. But some things don't need to be left up to personal preference. Community consensus should dictate things like number of edits (and percentage of those that are automated) and time served, basic experience with administrative tasks, knowledge of policy, etc. Quality of edits require vetting, but the most basic things should be standardized for reasons noted above. To have to similarly qualified candidates have the same issue come up in their RFAs and one pass while the other fairs, completely arbitrarily, it stupid, and it's a great example of how this project is an unorganized mess. Also a great example of why good editors don't want to go through the RFA process. There's no telling how it will go down. So, back to the point, 'crats should have something to go on to determine just how much weight votes should be given. For example, if the community consensus is that 5,000 unautomated edits and nine months of consistent editing is good enough, the Oppose. I think users should have 10k edits and a year on the project. would be appropriately disregarded.

RFA is overly arbitrary, to the point that it's stupid. This, of course, doesn't even take into account the grudge votes. Want to improve RFA and increase the number of active admins? Give 'crats greater power to discount votes and let the final decisions outline those that were dropped. Let enough people know that their reasons for opposing are so stupid and baseless that they were completely ignored and you might find a decrease in the number of opposes that are so stupid and baseless they should be completely ignored. From there you may find more people willing to go through the process. Some of the greatest minds, who would be the greatest asset to our admin corp, may just be those who are smart enough to know RFA isn't worth the time or stress.

As for 2, this has been recommended before. Splitting the various tasks up and having content admins and behavioral admins. Well, that was my idea. There have been others. The problem, or one problem rather, is getting the developers interested in making such a change. You'd have to get the community behind it first, of course; but looking to the future, even if the community agreed on some admin packages, and somehow managed to decided whether or not current admins would remain full admins or somehow be divided among the new ranks, there still is no guarantee the devs would design and implement it. Quite likely, as I was told, they wouldn't bother. Lara 19:24, 14 September 2009 (UTC)

Back to how to get more hard-working admins

WP:ER doesn't put people on the right track to adminship because Wikipedians don't show up en masse unless there's a fight, or at least something at stake. It's not pretty, but there it is, and we have to deal with that if we want more admins. It's not my call, but if it were, I'd say: create a new position called Trainee [The "Assistant" suggestion is more popular, see below 17:15, 14 September 2009 (UTC)], with the userrights to see deleted material and to move pages without leaving a redirect behind. ("Trainee" "Assistant" because we want people to stay focused on what Wikipedia needs, which is hardworking admins, and by analogy to modest titles like "janitor" and "bureaucrat".) I'd propose we not involve other userrights unless/until there's a clear need. We've already got some solid support for making the "view-deleted" right more widely available, because taggers and some reviewers can't do their jobs properly without it, and I want to add RFA voters to that list, too ... which is a more serious problem since not being able to see a candidate's deleted contribs partly disenfranchises them from an important !vote. I mentioned above that there's a legal problem with letting untrustworthy people see deleted material, so this has to be something more than "Yeah, you're okay" ... maybe 3 days at WP:PERM some special page [see below], and a crat won't flip the switch unless it's clear that a serious evaluation effort was made. (It's a Foundation issue that crats are required to flip a switch to let people see deleted material, although I suppose we could change that if we really wanted to.) These two userrights would help people do a better job with the kind of work that might eventually lead to RFA. Even better, this would give anyone who is willing to help a chance to give people feedback on what they've done so far that would put them on the road to an eventual successful RFA; if principles of volunteer management work roughly the same around here as elsewhere, and I think they do, a successful application for Trainee Assistantship, accompanied by RFA-style feedback on what it would take for that candidate to be an admin, is a lot more likely to produce an eventual admin than the same advice delivered at a failed RFA. - Dank (push to talk) 20:13, 13 September 2009 (UTC)

An apprentice admin. Sounds like a good idea. This should be integrated along with Wikipedia:Admin coaching. -- œ^™ 22:47, 13 September 2009 (UTC)

Sounds like a good idea, but how about calling this something like "Superuser" instead of "trainee" or "apprentice"? Those words (i.e., "trainee" and "apprentice") imply that the person is supposed to be continually asking for coaching, instead of simply using the tools as appropriate. --Orlady (talk) 22:52, 13 September 2009 (UTC)

How about (administrative) assistant? The parentheses show that the first word can be omitted. Or "manager" (prob' be confused with admin/sysop though). -- Soap ^Talk/_{Contributions} 00:47, 14 September 2009 (UTC)

I hadn't thought of that. My only preference is for some name that doesn't say "You've arrived!". - Dank (push to talk) 23:22, 13 September 2009 (UTC)

The best way to get more admins and have them stick around is for people not to give them so much crap when they make a call that is within policy; ie, the wolves howling at the moon because they don't like a decision. — Rlevse • Talk • 23:25, 13 September 2009 (UTC)

Agree 100%. J.delanoy^gabs_adds 00:45, 14 September 2009 (UTC)

"(Administrative) assistant" sounds good, too. How do we make it clear that we're taking the risk and giving extra tools to make it easier for them to participate at RFA, tag and edit, and eventually, to help them succeed at RFA? Perhaps these tools should automatically expire in 6 months, or they could apply again at PERM after 6 months, but the tools are gone forever after 1 year? (until/unless they succeed at RFA) - Dank (push to talk) 02:40, 14 September 2009 (UTC)

• There's a related discussion at WT:TWINKLE#Adding userfication to Twinkle; this could be a third (informal) userright to go with Assistantship, some functionality in Twinkle that makes userfication easier. Several people, including Amalthea, are generally positive on the idea but have expressed reservations about making it easier for just anyone to userfy articles. - Dank (push to talk) 13:07, 14 September 2009 (UTC)
• The "browsearchive" right touches on legal issues, and judging by the last comment on the issue by MGodwin, this won't happen. The "deletedhistory" right might be possible, but since libelous edit summaries have often only been deleted and not oversighted, maybe not. Amalthea 13:38, 14 September 2009 (UTC)

I raised the same issues above, and thanks for the link, we should talk with Mike sooner rather than later. The counterargument is: OTRS didn't fall apart in 2003 when a typical discussion to promote at RFA was "Yeah, he seems okay", and it didn't fall apart in late 2007 when we had 30 RFAs going at a time, most of them successful, so I can't see why OTRS would fall apart when we follow the much stricter standards we have now, and only promote to Assistantship if it appears to the community that a candidate is well on their way to a successful RFA ... particularly if the userright is removed after 6 months (on the theory that either we were right, and they pass RFA, or we were wrong and they don't need the extra tools). - Dank (push to talk) 17:25, 14 September 2009 (UTC)

Sounds like a plan! The assistant title sounds more appealing. I see you've struck it out but i like the idea of the six month window to apply to be a full admin. Your idea seems a most elegant solution to the problem of getting more admins without substantially lowering the bar and hence creating additional pressure to make it easier to desyssop (which I agree would be undesirable for many of the reasons above). Would it be possible to get round these legal objections to view deleted if there was an election for assistants (a less hostile one with promote as the default result, maybe thats in line with your thinking on the application process) and perhaps more oversighting of libellous content? FeydHuxtable (talk) 13:50, 14 September 2009 (UTC)

• I recently began a Wikiproject with similar aims, but of a more informal nature - WP:Junior Admin Squad. It recognizes the existing involvement in admin activities of motivated non-admins. It's not yet ready for roll-out, but any suggestions are welcome on the talk page. Delicious carbuncle (talk) 17:01, 14 September 2009 (UTC)

Yeah, I shifted my recommendation from a max of 1 year to a max of 6 months because that keeps us on well-trodden ground; we can just rely on all the input people have gotten in failed RFAs as our guide for how to tell when someone should get Assistantship ... they qualify if they look like the kind of RFA candidate who would be told that they would probably be successful if they do X, Y and Z and come back in 3 to 6 months. On the point raised by MGodwin and Amalthea and you, last night I changed my mind and don't think we should do this at PERM as if it's no big deal ... being able to see deleted material is a serious responsibility requiring some serious community input. Of course, any process is what it is ... if people take it seriously, then it's serious, and if they don't, it's not. But we can at least suggest that it ought to be taken seriously by drawing a connection to RFA and trying to get as many as possible to participate in the election and advice-giving.

I'm encouraged by the response, proposals like this usually require asbestos suits ... are we ready to discuss all the positive and negative intended and unintended consequences, maybe at Carbuncle's page? - Dank (push to talk) 17:11, 14 September 2009 (UTC)

Coaching with Tools

I'm still watching my stuff despite being in retirement, but there was an allusion above that piqued my interest. I'm an advocate of Admin Coaching and think that if it is done right, it is a great benefit. Unfortunately, there are a few (vocal voices) that oppose it. I think an interesting idea would be to introduce the tools to admin coachees. The process would be as follows:

1. An admin would apply to become a coach. This would be done via a simple request to the 'crats. We could establish a minimum tenure/requirements as an admin, but assuming the admin is in good standing, the admin will be allowed to have 1-2 coachees with tools. (The coach might have more, but only 1-2 will actually get access to the tools... this number might increase as we gain comfort with the process or individual coaches.)
2. The coach is then expected to review potential coachees to ensure that the coachee meets the coaches minimum expectations. If the coach is willing to accept a coachee, they simply make the request of the 'crats to grant the rights.
3. The coach will then work with the coachee on the use of the tools and will actually be somewhat responsible for ensuring proper usage of the tools.
4. If the coach looses confidence in the coachee, the coach can request that the tools be removed. If other admins have concerns about the coachees activities, they can ask the coachee to stop using the tools until the issue is discussed between the coachee, coach, and other involved admins. Additionally, admins can request that the tools be removed from a coachee if there is cause and coaches could lose their ability to have coachees if their coachees go unchecked.
5. After a 2-3 month coaching period and with the blessing of their coach, the coachee could formally request that addition of the tools become permanent. At this point, there would be more of an onus to show that the coachee failed to use the tools properly or did so in a destructive manner.
6. Have a limit (say 4 months) that a person can be a coachee with tools. If they complete the 4 month period and can't pass an RfA, then they don't get the tools.

This process would A) make it easier to get the tools into people's hands, B) make it easier to get them out of new admin's hands, C) ensure some mentoring and D) assuming decent coaching, passing an RfA should become a fait accompli. There won't be as much of a question as "what kind of admin will this person be?" but rather this person has used the tools for 3 months without destroying the wiki or going before ArbCOM. If there was going to be an issue, it would have more likely arisen during the coaching than during the RfA itself. I know that this proposal could probably be tweaked as I am writing it on the fly, but wanted to throw it out there....---Balloonman ^{NO! I'm Spartacus!} 19:36, 14 September 2009 (UTC)

• Comment: I have objected in the past to admin schools and coaching which had, oddly enough, non admins as instructors or coaches. I find that... questionable. I think your idea has merit. KillerChihuahua^?!?Advice 20:15, 14 September 2009 (UTC)
  • Agreed. Significant merit. → ROUX ₪ 20:19, 14 September 2009 (UTC)
    • It can easily get a bit in bred under the training school, mentorship umbrella, the pages new adim are good for new admins. I think of adminship as being something you are ready for but unskilled and that you grow and learn into the work, helped along by experianced admins. Off2riorob (talk) 20:24, 14 September 2009 (UTC)
      • What? → ROUX ₪ 20:25, 14 September 2009 (UTC)
      • No offense, but I think the cries of in breeding via admin school are a crock... take a look at my admin coachees. Many of them I routinely butt heads with and do not see eye-to-eye with... Seresin often votes exactly the opposite of me on many issues, Tan has drastically different views, Stephen comes to me for advice but is still his own man, and the others I haven't had too much interaction with since their coaching. But I can proudly say that as far as I know, none of my coachees have had any problems with ANI/ARBCOM/ETC (although I can't say that for all of my candidates.) In fact, if you are looking for Balloonman clones, you are more likely to find people whom I haven't coached mirroring my stance than those whom I have coached!---Balloonman ^{NO! I'm Spartacus!} 20:36, 14 September 2009 (UTC)
• I love it. iMatthew ^talk at 20:27, 14 September 2009 (UTC)
  • Regardless of its merit, I would be totally put off of applying for adminship if it required coaching. This is volunteer work, not school.--Atlan (talk) 20:29, 14 September 2009 (UTC)
    • I don't think that he's saying you HAVE to get this coaching to become an admin, but that it could be a good idea to implement. FWIW, I wholeheartedly agree and think that this is an excellent way to improve a somewhat defunct admin coaching system. Regards, MacMed^talk_stalk 20:32, 14 September 2009 (UTC)
    • Absolutely not. This would in no way become a requirement, but another avenue towards coaching. It would become an easy way to get the tools into people's hands while they proved themselves. It would appeal to people who might be interested, but don't want to subject themself to an RfA. It would also become a way for somebody to find out if it was worth it.---Balloonman ^{NO! I'm Spartacus!} 20:38, 14 September 2009 (UTC)
      • Yeah, I didn't read that well enough. Anyway, they idea seems hard to implement. Wouldn't it overburden the few 'crats we have? Besides, I can already see the "Oppose, didn't have coaching" votes coming a few years down the road.--Atlan (talk) 20:42, 14 September 2009 (UTC)
        • I don't think it would over burden the 'crats we have now nor do I think it would become the norm. You would still need to get people to be willing to be coaches and to take on the responsibility of reviewing their coachees actions. Coaches who don't review their coachees actions would lose the ability to have coachees. In the past, I would guess that we've never had more than 20 active coaches, let's assume that number doubles to 40. Most of those coaches will be one and done, coaching a friend or colleague on a project (thus coaching would not be as institutionalized.) So at any time, optimistically speaking, we might have 40-60 coachees. Even if we rotated in and out of those coachees on a quarterly basis, we are talking about 200-250 additional actions per year---half of which might have occured anyways. The only place where there might be an increase would be on desysopping at the Stewart Level, but still we are only talking about flipping a switch on those people who fail/drop out... so 40-80 people? It wouldn't be that great of a burden. If it were to become the norm, it would be because it was an easier process.---Balloonman ^{NO! I'm Spartacus!} 20:52, 14 September 2009 (UTC)
    • I like it, and I'm normally a (silent) voice against coaching. And the level of schooling absolutely depends on the coach, just as the mentoring at WP:ADOPT depends on the adopter: If I'd be asked to coach someone under this program, I'd normally be content to stalk the coachee's contributions, certainly not have lessons or a program I want to go through. Well, maybe except requiring a certain amount of CSD work, since I expect every admin to be on very sure footing there.
      So yeah, I'm in principle in favor of such a proposal. I was thinking for a moment that there should be some kind of crat discretion when allowing a coach, but that would be counterproductive. Amalthea 22:33, 14 September 2009 (UTC)
      • In my vision, the crat would have two pieces of discretion. First, is the person requesting the coaching an admin in good standing? Second, is the coachee a candidate in good standing that meets some general guidelines on experience. (The coach should check those first.) But yes, this would change the nature of coaching.---Balloonman ^{NO! I'm Spartacus!} 22:48, 14 September 2009 (UTC)
  • me too. Here is the link to the school. I like this it seems good and I was saying before that .. with mentors and personal teachers that nepotism can creep in. Here is the school ..Wikipedia:New_admin_school Off2riorob (talk) 20:31, 14 September 2009 (UTC)
    • If that is a concern, we could make it a requirement that coachees/coaches are not related.---Balloonman ^{NO! I'm Spartacus!} 20:40, 14 September 2009 (UTC)
• I'm in favor of trying this and various other things and gathering data. All you really need is to come to some agreement beforehand on how to tell if it's working or not working, and then watch it and see. - Dank (push to talk) 20:37, 14 September 2009 (UTC)
  • Caveat: there are some options I'm not willing to experiment with. Any promotion to any position of importance, and I don't see how "coach" wouldn't be an important position, has to be a community decision, per WP:CONSENSUS. The crats are only there to decide what the consensus was, not to make the call.
• I strongly support this proposal. Part of the problem with RFA is that it's just such a big leap to go from 'ordinary editor' to 'admin', and voters often feel uncertain whether they can trust an editor with all the tools even if they've had no problems so far. Having a system of 'trial administrators' (which is basically what this is) would give candidates a chance to demonstrate what they would use the tools for and what problems they might have with them, and thus would make things much easier for voters when they came round to the RFA. It would also probably make it easier for good users to pass RFA, by giving them a chance to demonstrate competence with the tools. And I like the whole 'presumption in favour of the candidate' as well, where they get to keep the tools unless there are obvious problems with their having them. Of course, it would be crazy to introduce this process for all candidates; but if it was offered as a voluntary option, I expect many would take it, and the wiki would benefit as a result. Robofish (talk) 21:19, 14 September 2009 (UTC)
• Support Per above, brilliant idea.--Giants27 (c|s) 23:54, 14 September 2009 (UTC)
• Support - brilliant! There's only one problem I can think of: "what does this button do?" Perhaps making sure coachees know exactly what that button does, before using it in a potentially destructive manner... It can all be reversed, but still...--Unionhawk ^{Talk E-mail Review} 00:49, 15 September 2009 (UTC)
• Support and I would like to formally volunteer to become the first coachee ;) -- œ^™ 02:42, 15 September 2009 (UTC)
  • knowing WP, it would take 2-3 months before we had enough consensus to get this through ;-)---Balloonman ^{NO! I'm Spartacus!} 02:53, 15 September 2009 (UTC)
• Before we go further with this, we should probably ask the crats if they'd be willing to promote someone even temporarily on the say-so of the coach, in light of Mike Godwin's objection that was linked above. Wouldn't we need some kind of community consensus that the candidate is at least trustworthy? - Dank (push to talk) 03:38, 15 September 2009 (UTC)
  • I've set up a separate page with a little more fleshed out proposal Wikipedia:Requests for adminship/Coaching with Tools. Yes, this would require more input and a wider community, it would require not only the 'crat approval but also Stewarts (as they might have to desysop more people) but I don't think we would be adding a tremendous amount to the process. As for Mike's concern. I think the main concern is having powers granted out willy nilly, understanding where the concern comes from, I think the foundation would be open to this idea if it was something the community sought. I went ahead and emailed him about the issue.---Balloonman ^{NO! I'm Spartacus!} 03:54, 15 September 2009 (UTC)

Another way to look at it

If you want to read the dates then you will need to zoom in to the full sized version.

I think the graph at the top of this discussion may not be the most enlightening way of looking at the issue. I have graphed the situation in a different way, that is how many admins performed an administrative action in each given day. Instead of counting how many admins have edited without 2 months of a date this only counts how many admins performed an admin action on a given day.

I am not sure what this graph says in regards to the debate happening here but this is I think a better representation of how many active administrators there are over time. The green area is an approximate match to the time period covered by the graph at the start of this section. Chillum 00:28, 14 September 2009 (UTC)

Ooh, pretty, thanks for crunching these numbers. I think this trend still shows a decline, with the peak in late 2007/early 2008 (which is identical to the peak in the graph above). MBisanz ^talk 00:32, 14 September 2009 (UTC)

One thing I notice is a sharp drop every Christmas. Perhaps something about the birth of Christ is driving off our admins? Chillum 00:34, 14 September 2009 (UTC)

(ec)ⁿ I wonder how much a coincidence it is that the downward slide began in earnest during the tenure of the 2008-vintage arbcom (which I affectionately refer to as the Arbcom From Hell). Short Brigade Harvester Boris (talk) 00:35, 14 September 2009 (UTC)

I think you will see the same slow decline after most exponential rises to a peak. I don't think I included enough information in the graph to blame one group of people. The graph does not show the quality of the admin actions or the quality of the admins. Chillum 00:37, 14 September 2009 (UTC)

I can blame a group: The community for not planning for such an eventuality. MBisanz ^talk 00:43, 14 September 2009 (UTC)

By the way, the majority of these actions are deletions. Chillum 00:39, 14 September 2009 (UTC)

Would it be possible to restrict it to mainspace actions? The others aren't terribly important.  Skomorokh  00:50, 14 September 2009 (UTC)

Very interesting, thanks. I noticed the sharp decline around the holidays each year as well. –Juliancolton | ^Talk 00:40, 14 September 2009 (UTC)

Is there anyway you can extrapolate your data out say five years into the future? MBisanz ^talk 00:43, 14 September 2009 (UTC)

I can extrapolate the data 5 years into the future dozens of different ways, each with drastically different results and none of them likely to be accurate. (yesterday the number of graphs-per-day I uploaded was zero, today it is one, by this data this time next year I will have uploaded 365 graphs - http://www.xkcd.com/605/ (read the mouse-over text)) Chillum 00:48, 14 September 2009 (UTC)

Might I ask how you extracted the data here? Also, it would be most interesting, if it's not too hard, to plot another line representing edits or something similar on the same graph; I think that edits per day actually looks pretty similar, reaching a similar peak and then a slow and steady decline. Also, once again, if you have a chance, it would be interesting (though I'm not sure exactly what it says) to plot something like the ratio of active admins to active editors. Just thoughts. Anyway, great graph. Cool3 (talk) 02:52, 14 September 2009 (UTC)

I took the timestamps of every admins action in the logs and put them into a mysql database which then broke it down into days and counted how many admins were active that day. As for edits, I could get the edits of admins but not how many edits in general. If the statistics for how many people a day edit are available then I can plot the ratio between the two. Chillum 14:19, 14 September 2009 (UTC)

There are monthly stats for edits per day, active Wikipedians, etc. here, but I don't know about daily figures. Cool3 (talk) 16:01, 14 September 2009 (UTC)

We don't have quite enough admins because that's a decision that we all made and all enforce, as a community. One trick we use to keep the admin numbers low is a kind of deception; the general principles of passing RFA aren't hard to learn, but every candidate brings different strengths and weaknesses to the table and every candidate is treated differently at RFA, and in practice, it's been very hard for the candidates to guess what the issues and the votes are going to be. If we want more admins, the first step is to stop tripping up the people who are trying to become admins by withholding information they apparently need. WP:ER doesn't work because not enough informed people show up because there's nothing at stake; failed RFAs repel as many candidates as they educate. Another trick we use is not giving people a sufficiently interesting but attainable intermediate target they can aim for. #Back to how to get more hard-working admins is one proposal, but any proposal that tackles these issues is IMO better than the status quo. - Dank (push to talk) 17:48, 14 September 2009 (UTC)

P.S. Another way we trip candidates up is by requiring (for many of them) extensive experience in deletion work, without letting them see some of the occasionally critical pieces of evidence (deleted edits and pages) that would allow them to make good decisions and justify those decisions. - Dank (push to talk) 18:16, 14 September 2009 (UTC)

The problem this graph and all others have is that it doesn't take into account the affect of tools/bots on admin functions. A good admin with a tool/bot can do more in a day today than a score of equally good admins could have done two years ago. The question I'd want to see are lag times at areas needing admin actions. Do we have a significantly larger gap between the time a 3RR request is made an action taken? CSD? etc. If the areas requiring admin actions are not going up, then the number of admins or the number of actions per admin are irrelevant.---Balloonman ^{NO! I'm Spartacus!} 19:42, 14 September 2009 (UTC)

Chillum, I like this graph.

Everyone, one thing I believe I see, (though of course I haven't done a proper statistics test to a 95% confidence level), is that there is a seasonal decline during the northern hemisphere summer. In 2006, it shows as a temporary flattening of a growth curve. In 2007 it shows as a decrease before returning to the same level. In 2008 it isn't as much of a decrease but it shows. In 2009 - the curve doesn't show the post summer period yet. So too soon to tell, but this may well be one of the factors.

Why would we have fewer admins active in the northern hemisphere summer? Vacations and school. (We know our editors are disproportionately from the northern hemisphere.) Some of are on vacation from school and/or work so do less during the summer. And many of the things that need admins to deal with are done by people in school, and thus there is less demand for admin actions in the summer.

I think this chart will be even more informative in October, as we'll be able to see what the stats look like after school is back. But looking at the chart, I'd date the start of the trend downwards at sometime in the first four months of 2008. GRBerry 00:40, 16 September 2009 (UTC)

Another graph

Here's another graph for your consideration, showing the number of promotions at RfA (the red line) and the number of admins lost in that year to inactivity, resignation, or involuntarily desysopping. Please note that the inactivity figure only includes admins who left during that year and never returned, thus it does not take into consideration admins who became inactive for a period of time and then returned. I must say that, frankly, this one is quite troubling. The number of admins we lose is almost perfectly exponential, while the number we promote is declining precipitously. Here is the raw data for lost admins and RfA promotions as well. Please further note that in the data set, a 12-month estimate has been used by extending the data for the ytd. Cool3 (talk) 17:45, 14 September 2009 (UTC)

Year	Admins Lost	RfA Promotions
2002	2	–
2003	9	–
2004	14	239
2005	21	387
2006	54	353
2007	77	408
2008	120	201
2009 (ytd)	165	88
2009 (12 month estimate)	234	124

This is why the community needs to get off its ass, ignore the fearful bad apple admins who protest against an easy desysop process, and force something through that will 1/ improve the RFA process, making it less of a week from Hell and 2/ make stripping the crap admins of their tools less of an act of congress, or ArrbCom as the case may be. Lara 19:03, 14 September 2009 (UTC)

In order for that to happen, everyone would have to take the cotton balls out of their ears and actually listen to the ideas that many editors have about improving RfA. iMatthew ^talk at 19:13, 14 September 2009 (UTC)

While I wholeheartedly agree with Lara that we need to make desysopping a much easier process, which would then make RfA a less strenuous process, I think the graphs above are misleading. Yes, the number of active admins appears to be declining. However, as alluded to earlier, what impact has this had on the functioning of WP? Are the backlogs increasing? If not, why are we worried? Karanacs (talk) 19:47, 14 September 2009 (UTC)

Both administrative and editorial backlogs have indeed been increasing. CAT:SD is almost always backlogged; DYK is just barely getting by these days due to a lack of participating admins and clerks; Category:All unreferenced BLPs is populated by 50,000 pages; the list goes on and on. At first glance it appears everything is running smoothly, but this isn't the case. –Juliancolton | ^Talk 19:57, 14 September 2009 (UTC)

There was a surge in new admins from 2005-07 and I imagine that many of those, like myself, now have less time for Wikipedia or are no longer interested in committing as much time. Meanwhile admin growth has dried up. Part of this is the effect of higher standards. But I also feel that Wikipedia is not as interesting as it once was. This is partly the unavoidable fact that it is no longer "new", and partly the fact that it now has a more antagonistic relationship with its own growth. I think the project has deliberately undertaken to reduce its scope (e.g. across large swathes of the pop-culture area) and limit growth in other ways. While this might help ensure better quality, it lowers growth prospects and in this way makes the project less exciting to be a part of. Christopher Parham (talk) 22:05, 14 September 2009 (UTC)

Agreed!!!---Balloonman ^{NO! I'm Spartacus!} 22:20, 14 September 2009 (UTC)

Well, the project is going through stages. Perhaps we're entering a stage where we've come to focusing on harboring a professional community and creating a high-quality work, where in the beginning, Wikipedia was little more than an experiment and our goal was to create as many articles as we could. I for one think it's a step in the right direction; the project will inevitably continue to mature over coming years, as even now, several years after its creation, Wikipedia is in uncharted territory. Certainly in the two years I've been a contributor, the atmosphere has evolved. Overall, I agree that Wikipedia has become less "fun", so to speak. –Juliancolton | ^Talk 02:42, 15 September 2009 (UTC)

That is a fair assessment, and I concur with your idea that the project is taking on the attitude of a "professional community" as a way of achieving higher quality. Perhaps, though - and I am wholly speculating - in becoming more professional, we become less attractive to the sort of young, high energy individuals who probably lack the writing skills and research capacity of their older colleagues but have more time for and interest in administrative/maintenance tasks. There are a lot of admins who do mainly administrative work in their old wiki-age but I sense that few people become attracted to Wikipedia in the first place without stumbling upon an article they can meaningfully contribute to; between the filling in of blank space as Wikipedia grows and our increasing aversion to low-research-intensity material like plot summaries, it's harder for new users to stumble onto such an article. Christopher Parham (talk) 13:20, 15 September 2009 (UTC)

You raise the bar you lower the number of people that make it. That's not rocket science. The knowledge expectations in RfA are really exorbitant: copyright regulations, the whole nine yards of the deletion process, questions to which no one has a convincing answer (difference between block and ban), trick questions (cool-down block), implicit behavioral expectations which require following RfA for weeks (Oppose: excessive badgering of opposers), and so on and so on.

Besides, editcountitis has also not made the process better. To create a half-page stub with correct templates, basic formatting, and one or two sources one edit is sufficient. Do it this way and you take many years to reach the threshold. But if you don't know where the preview button is you'll be there in six months -- I have yet to see an Oppose: too many edits to boost edit count, as well as a successful RfA for a user with 500 high-quality edits across the spectrum (yes, I do know this exception). The usual response, it takes 2500-3000 edits to gain understanding of policy, is strange. --Pgallert (talk) 10:55, 15 September 2009 (UTC)

I have to correct you: Oppose: too many edits to boost edit count = Oppose: too reliant upon tools."---Balloonman ^{NO! I'm Spartacus!} 18:11, 16 September 2009 (UTC)

This chart is garbage. It is strongly biased to show an up trend of departure in the most recent period, regardless of real trends. The definition of "became inactive and never returned" oversamples people that recently became inactive. Admins who become temporarily inactive do so for varying lengths of time. The definition used will count any temporarily inactive admins that aren't yet back as permanantly inactive. We know that time period exceeds a year (e.g. Tznkai who came back from a long break and has been doing good work). So the 2009 data is totally useless, and the 2008 data is suspect as well. A better definition is needed that lacks that inherent bias - it will have to be one that counts temporary inactivity as a departure in all periods, or one that just doesn't look at departures within a period long enough that I'd still be counted as an admin.

In addition, if existing admins were departing with a constant probability in a given period of time, there would be a growth trend in departures so long as the total admins are growing. (A more reasonable model, of course, has a distribution of probabilities of departure based on tenure as an admin, this is the nature of burnout.) GRBerry 00:49, 16 September 2009 (UTC)

Timeline of events

This discussion has been closed. Please do not modify it.
The following discussion has been closed. Please do not modify it.
Please use Wikipedia Review to shout about injustice or whatever else you came to complain about. This page is for discussing en.wikipedia's process for assigning admin rights. Complex stuff, obviously.Pedro :  Chat  20:48, 14 September 2009 (UTC) If you're interested in some of the unfortunate events that took place in the first quarter of 2008, when the decline noticeably began, there is an interesting timeline of Wikipedia events, maintained externally. Some key things to note: January 2008 Wikipedia moves its headquarters from St. Petersburg to a secret location in San Fransisco. January 6th Jimmy Wales replies to journalist Seth Finkelstein, "Seth, you're an idiot." Seth had questioned the viability of the about-to-be launched Wikia Search. January 22nd Spanking Art Wikia utilizes GFDL content from Wikimedia Commons - An inappropriate wiki site hosted by Wikia, later shut down due to outcry about the content hosted. February 7th Administrator User:JzG makes a number of changes to the biography of Rachel Marsden. February 29th Valleywag reveals that Jimmy Wales has been having a relationship with Fox TV reporter Rachel Marsden. Wales intervened in her Wikipedia biography back in 2006, as was noted by Wikipedia Review, and the intervention reportedly led to an in-person meeting. Valleywag also publishes "transcripts of Wikipedia founder's sex chats" with Marsden. March 1st Former Wikimedia Foundation employee Danny Wool blogs that Jimmy Wales used to boast about several affairs extra-marital affairs during his time at the organization. Wool also alleges that Wales "was certainly not frugal in his spending on his endless trips abroad" and that Wales was "careless" with his receipts while using money donated to the Foundation in good faith. Wool also alleges that Wales spent donation money on a massage parlour in Moscow, spent $650 of donors cash on two bottles of wine, and thought he needed a limousine "because I am like a rockstar too." Associated Press report that WMF Chair Florence Devouard castigated Wales, "I find (it) tiring to see how you are constantly trying to rewrite the past. Get a grip!" March 5th The mainstream media universally cover both Jimmy Wales's Rachel Marsden debacle, and the allegations from Danny Wool that Wales misused money donated to the Wikimedia Foundation. Those are just a few excerpts that would cause me, if I were an admin, to give up on the project. -- Thekohser 19:30, 14 September 2009 (UTC) You don't have to be an admin to quit. hint hint Why would those examples affect you differently if you were an admin? If you hate the project, just go. J.delanoygabsadds 19:34, 14 September 2009 (UTC) These events definitely figured into my decision to resign as an admin. Obviously I can't speak for the others. Short Brigade Harvester Boris (talk) 19:39, 14 September 2009 (UTC) Well, I guess for me, I didn't imagine that something would make you want to resign as an admin, and yet not resign from editing, so to speak. But maybe that's just how I roll... J.delanoygabsadds 19:41, 14 September 2009 (UTC) Hardly unusual. It's entirely possible to think this project can be turned around and it's worthwhile trying to do so, without wanting to be interrupted every ten minutes by random cranks screeching "abuse" and "conspiracy" at every single thing you do. – iridescent 20:07, 14 September 2009 (UTC) Another explanation would be that the number of active Wikipedians started to decline in early 2007 ([1]). It takes a while to become an admin, so there were still 408 RfA promotions in 2007, but there was less "fresh blood" available in 2008 and 2009 and RfA output was much lower (299 for the two years combined so far). Combined with the large number of departing admins, this reduced the number of active administrators. Thus, I would say that if you're looking for causation (which is not to say that there is any one event responsible), you should really be looking in early 2007. Cool3 (talk) 19:53, 14 September 2009 (UTC) Fuck me. Let's use the talk page relating to an obscure in-house process to get some extra permissions on a website to rant about people's off-wiki commentary, James Wales knobbing some bit of totty and load of inconsequential twats who bang on about this bollocks on thier blogs. If I wanted to read gobshites who love their blogs I'd have gone to Wikipedia Review and read the self-reverental and insecurity laden bullshit from Kato and Barbour. Someone please remove this rubbish from this page. Pedro :  Chat  20:38, 14 September 2009 (UTC) Seconded...---Balloonman NO! I'm Spartacus! 20:43, 14 September 2009 (UTC)

Alternative RFA reforms

I have a suspicion that if the admin coaching route is adopted it could wind up being counterproductive and even reduce our number of new admins - especially if it becomes the normal method of becoming an admin. So I'd like to resurrect some of the other ideas for RFA reform that have previously been rejected because not everyone was willing to accept that we had a problem here.

1. Synchronise standards. If this was a job interview you would expect the interview panel to have done some work to agree what standard they were looking for, as it is that partly happens by experience at RFA. But the drawback of that is that we give out mixed messages at RFA with people opposing or supporting based on differing standards and the same information. Too many RGAs degenerate into adhoc discussions such as whether WP:CLEANSTART is compromised by RFA, or whether a candidate with a lot of automated edits can be assessed just on their manual edits. I suggest we set a series of RFA standards for candidates to aspire to and RFA !voters to assess candidates against.
2. Consensus at 75% values an oppose as being worth three supports - hence the ratchet effect that leads people to think I was on the losing side perhaps I should reconsider my standards, even when they were in the majority. Lowering the threshold to 50% would solve this problem, If that's too big a jump for some, lowering the crat discretionary range from the current 70-75% to 60-65% would reduce the ratcheting effect and produce more admins. Looking at those like me who achieved that range in an RFA and made it on a subsequent attempt I'm reasonably confident that we won't break the wiki if we lower the bar a bit. So I propose that the crat discretionary zone be lowered to 60-65% of valid !votes, with any decision outside that range requiring a crat chat.
3. Make Wikipedia:Arguments to avoid in adminship discussions a policy and encourage the crats to ignore !votes that use such oppose rationales.
4. Enforce wp:civility at RFA and strike votes that are attacks without diffs to substantiate them.
5. Survey our active longterm users who are not admins, have clean block logs and have not run at RFA for at least a year and ask them why they haven't run, or haven't run of late. Then use the feedback from the survey to improve the RFA process.
6. Set a minimum standard for self noms. If we agreed that certain criteria must be met before you could self nom at RFA then some of the Dwamah would be avoided and we could focus on things that really mattered - exceptions will happen but they can get someone to nominate them and explain why they should be an exception: So I propose that self noms at RFA must have

1. 2,000 manual edits
2. 6 months tenure.
3. 3 months since their last RFA

Obviously some of these reforms would interact and are best not all done at the same time. But I believe that any or all of these could improve the RFA situation ϢereSpielChequers 14:46, 15 September 2009 (UTC)

I don't think it would help much. Civility rules apply to RFA, also a tenure requirement is a bad idea as it will jst get bigger and bigger.--Patton123 (talk) 15:11, 15 September 2009 (UTC)

I really like the Coaching idea, but I agree its not the ideal solution on its own. For many good candidates coaching wont be suitable. It would be great to have a package with something for everyone. 1) & 6) could be effective and are in line with Lara's suggestion above, maybe we need to guard against scope creep per Pattons point. 3) & 4) could maybe be left out of the initial proposal as they're likely to be unpopular? Like the other two. FeydHuxtable (talk) 15:27, 15 September 2009 (UTC)

• The possibility of reforming RfA is zero. It just isn't going to happen. So, in response, the community has come up with an ever widening way of effectively limiting the need for admins. I look at things like Category:Candidates for speedy deletion and WP:AIV, and don't see backlogs. With the abuse filters in place, along with protected pages, autoconfirmed, etc. the opportunity for vandalism (a frequent area of admin work in the past) has gone down. There's a crap load of automated tools out there now, reducing the work load admins typically shouldered. Plus, there's a large number of bots operating as well. This trend will continue. The need for administrators will continue to decrease. As the need goes down, the standards will keep rising; no need to be non-selective when you only need a few. Make sure they're good ones. This is a natural, evolutionary process. Attempting to change it is like getting angry because the tide is coming in. --Hammersoft (talk) 15:31, 15 September 2009 (UTC)

That's one way of looking at it. Looks like in addition to the graphs above we'll need well presented evidence that admin queues are growing or at least are persistently large, else the argument you've just made will kill the proposal when we take it to Village Pump. FeydHuxtable (talk) 15:48, 15 September 2009 (UTC)

Vandalfighting bots don't block accounts - we still use admins for that. If anything the decratting of an admin earlier this year for running unauthorised admin bots has left us with adminship less automated, while the things that admin wannabees do are getting more automated. So the admin shortage that most now accept exists will tend to get worse through automation as it limits our traditional pool of recruitment. ϢereSpielChequers 15:56, 15 September 2009 (UTC)

If we could agree on standards, vote counting would become entirely unnecessary. A single oppose, for good cause, should be enough to prevent promotion. The trick of course is that nobody can agree on what counts as "good cause". I've seen enough here to conclude that the "community" is hopelessly incompetent at the job of picking admins. Let the crats or somebody do it, with minimal input from the peanut gallery. Friday (talk) 15:35, 15 September 2009 (UTC)

• It's a growing trend; the 'community' isn't really trusted anymore. Protection, autoconfirmed, throttles, the upcoming color coding of articles, etc. New users are treated like shit around here, and IPs are treated like the scum of the earth. This too will expand, and it is a natural progression of the project as it ages. Lots of people are blogging now about the changing nature of the project in this way. The number of active editors is going down too. --Hammersoft (talk) 15:39, 15 September 2009 (UTC)

Completely oppose making arguments to avoid a policy. Those are positional statements that are subject to change. What is an argument to avoid today may be a valid argument tomorrow or might be substantiated at a later date by somebody making a solid rationale. Plus, I don't think everybody agrees with every point in that essay.---Balloonman ^{NO! I'm Spartacus!} 15:49, 15 September 2009 (UTC)

That sounds like a good argument for challenging some points in the essay whilst turning it into policy, and of course like any other policy or process being willing to review it over time to see if it still works. Just as we currently need to review the RFA process.... ϢereSpielChequers 16:07, 15 September 2009 (UTC)

You miss my point. We are an evolving community. An argument to avoid that year may be accepted today as a valid argument because the validity was substantiated or became accepted. Two examples. First, Admin Coaching. There was a time where being admin coached was perceived as having value. When Kurt Weber first made the first "Admin coachee" argument (and yes Kurt was the first I looked it up!) it probably would have gone into the Arguments to Avoid umbrella. Kurt however continued to make his arguments and eventually convinced others that coaching was evil. Second, again a Kurt argument, Self noms. Kurt's prima facia oppose is still the stuff of legend. Without looking at ATA, I'm pretty sure that Oppose per self nom was listed as an ATA. Yet, when Kurt was RfA banned, others were accepting that rationale and actually opposing due to self nom. I suspect that if Kurt had been active here for another 6 months that Self Noms would have become an accepted reason to oppose. The community changes and expectations change... we should not codify the reasons people could give to oppose, that could stymie productive discorse. The 'crats know that "Oppose self-nom" is not an argument that the community embraces. However, if the community were to embrace it, I have no doubt that the crats would acknowledge that change in how they review RfAs.---Balloonman ^{NO! I'm Spartacus!} 14:27, 16 September 2009 (UTC)

I agree that our criteria may change over time, I'm merely suggesting that such changes be done by consensus outside of an individual RFA rather than by policy drift during successive RFAs. And I suppose I'm trying to solve the admin shortage problem by resetting the clock and taking the RFA process back to a time when people who'd now get a snow close were sailing through RFA. ϢereSpielChequers 23:00, 16 September 2009 (UTC)

• Comment on 2 - lowering threshold to 60-65%. If we include this suggestion, would it work better if we combined it with a proposal for "desysop light" that only applies to new admins for their first 6 months? Something like if there is a clear case of the new admin mis using the tools they can be desysoped by a panel consisting of say only 3 arbitrators, with streamlined proceedings that wont drag on. So the fear of bad apples slipping through is addressed, but also we wont make wiki life harder for long term admins? FeydHuxtable (talk) 16:02, 15 September 2009 (UTC)

• I'm not convinced either that new admins are any more likely than old hands to need decratting or that resolving any problems in the decratting process should be linked to RFA. I'm aware that there are some who consider the decratting process broken and who have frequently turned conversations about reforming RFA into discussions about decratting. But I don't agree either that the problems are related or that the decratting problem is anywhere near as severe as RFA. That said if someone did a study of decrattings and detected a problem such as admins with fancy signatures or more than 2 FAs are more likely to get decratted then I would expect that provoke some interest here. ϢereSpielChequers 16:17, 15 September 2009 (UTC)
• (too Feydhuxtable) See Wikipedia:PEREN#Administrative, hierarchical structures. --Hammersoft (talk) 16:27, 15 September 2009 (UTC)

(too WSC) I wouldn't expect a strong causal relationship between difficulty of desyoping and the declining active admin core. If there is one it might even be inverse. I just remember that when lowering the threshold's been discussed before, several expressed a strong concern that bad admins are already a problem and they wouldn't want it to be made easier for new ones to slip through. If we can allay that concern there might be less resistance to the idea of lowering the threshold? But yeah change to desyop need not be linked and if the idea can pass on its own that probably would be for the best.

(too Hammersoft) thanks, yep IronHolds pointed me to that a few weeks back. Im hoping that now we've gathered such strong evidence about the problem, concensus for previously rejected proposals can change, especially as we're introducing a few differences. Already we've had admins like the Puppy who previously opposed this type of reform making positive comments about the idea. FeydHuxtable (talk) 16:37, 15 September 2009 (UTC)

• I’d like to see progress in synchronization of standards. I use the word “progress” deliberately, as complete synchronization is an impossible goal, but asymptotic progress may be possible.
• I support a gradual reduction of the consensus bar. 66%, where one oppose equals two supports has a nice feel to it, but I could live with 65%, and maybe 60%. I have reviewed some of the past requests, and getting into the fifties is too low.
• I’d like to drop the pretense that adminship is no big deal. I gather someone important said it once, so it became gospel, but when I see the detailed understanding of policy expected – it is misleading to pretend it means so little.
• More specificity regarding minimums would be helpful. They should be soft, meaning you can apply with fewer, but the expectation will be more evidence of quality in the smaller number. 2000 edits is too low, moreover, one can have 2000 edits without learning any appreciable portion of policy. Can we add some suggested edit counts for relevant areas like XFD? Maybe 500 edits to XFD, ANI or similar?
• I’d like to see a strong suggestion that participation in RFA is a sign of taking the process seriously. It is possible to read the RFA pages and still not cast an !vote, but ten !votes would be a sign of minimal involvement. (I can understand concern that encouraging !votes might produce poorly researched !votes).SPhilbrickT 18:03, 15 September 2009 (UTC)

Percentages, rules, and limits are fine and well, but in the end, RfA, like any community discussion, boils down to "consensus". –Juliancolton | ^Talk 18:40, 15 September 2009 (UTC)

Is there an admin shortage?

I don't really see huge backlogs. I stopped bothering with AIV because it gets cleaned up quickly. Moves are usually a bit backlogged but that's more because it is hard to figure out consensus in some cases (when in doubt, I wait for someone else to do it and, if no one does, do try to clear it up). CSD goes through cycles (and, IMO, CSD purists are likely responsible for many an RfA failure - just saying!). RFPP is usually fairly up-to-date. The edit warring notice board is almost always current. Given the lengthy discussion above, there must be a humungous shortage of admins. Let me know where - I'll be happy to help out. --RegentsPark (sticks and stones) 02:31, 16 September 2009 (UTC)

CAT:TEMP needs 30,000 pages reviewed by admins for deletion and Wikipedia:New histmerge list needs 20,000 history merges fixed by admins. MBisanz ^talk 02:35, 16 September 2009 (UTC)

Just looking at the content of those two tells me that having many more admins is not going to be a great help. --RegentsPark (sticks and stones) 02:47, 16 September 2009 (UTC)

But only admins can review/delete the pages there. MBisanz ^talk 03:16, 16 September 2009 (UTC)

True. But the work involved seems tedious (which is why new or old admins won't be flocking there in a hurry!) and there doesn't seem to be consensus that there is any work to be done at all on, say CAT:TEMP. Also, there are no instructions on what an admin is supposed to do. For example, I can see that CAT:TEMP has a backlog but what does one do to clear it (assuming that there is consensus to clear it in the first place). Perhaps we should be focusing our energies on creating better instruction sets. --RegentsPark (sticks and stones) 18:16, 16 September 2009 (UTC)

Please note that WP:RM maintenance is not strictly an administrators' task. @harej 02:37, 16 September 2009 (UTC)

Noted. But it is largely done by admins (even though an admin is only required for page deletion, I'm not sure why few non-admins close move discussions). --RegentsPark (sticks and stones) 02:47, 16 September 2009 (UTC)

One thing that stands out to me is when vandal/hoax tagged CSD articles stay around for a few hours (e.g. John Moustafa - recently deleted after tagged for 4.5hrs) - this one actually originally contained an attack too. Not sure what (if any) targeted turn-around time there is for CSD tags... sometimes it's within minutes, but sometimes CSDs are handled days later. But for 3+ hours for this kind of content, it seems a little long. I am wondering if it is less an issue of shortage, and more an issue of timezone concentration.  7  03:02, 16 September 2009 (UTC)

This is partly a timezone issue, or at least certain timezones illustrate our admin shortages more sharply than others. I suspect that I'm sometimes the only admin looking at CSD in the later part of the UK morning. But not all articles tagged for speedy deletion have the same priority, that's why there is no targeted turnaround time at CSD. Subcategories such as attack pages get much faster attention than others, looking at the attacks I've deleted this morning some were up for less than ten minutes but one was up for about an hour. However it isn't just the stuff that's already tagged. One of the downsides of the admin shortage is that there is less time available to look for the stuff that the hugglers have missed, a couple of days ago I deleted a particularly nasty attack page that had been up for over two years. ϢereSpielChequers 12:25, 16 September 2009 (UTC)

Definitely agreed on the timezone concentration issue. Naturally there are fewer EN admins active during my (Japan) timezone. I also agree that different CSD categories should receive different priorities by a wandering admin. Makes sense to clear out attack and copyvio first and save the harmless junk for later. Just thinking that if the average time it took to process the average CSD (in any category) started to dramatically exceed what the community felt was ideal that it might be an indication that more hands might be needed to hold the mop.  7  14:09, 16 September 2009 (UTC)

I disagree with the argument that John Moustafa is a proof that we have an admin shortage. Vandalism occurs all the time on all sorts of pages; it's built into our system. It is not prevented by any number of admins. Even a duplication of admins - which would be a huge effort - would (on average) only cut the time for an administrator to see it by one half. That means, the vandalism would have been up for more than 2 hours. Hardly an improvement worth the effort. This sort of vandalism could have been much easier prevented by flagged revisions. — Sebastian 21:13, 16 September 2009 (UTC)

I'm not sure there's a shortage, per se, but there is turnover as admins burn out, leave the project, "take their ball and go home", stop doing admin stuff, etc. etc. I seem to remember Mbisanz or someone doing an analysis of "active" admins, and I am pretty sure that the number of "active" admins has stayed relatively constant despite the fact that we still keep making new ones; I think we are pretty much at "replacement rate" now, and are not actually growing the admin corps... But I could be wrong on that... Just the impression that I have of the situation. --Jayron32 03:46, 16 September 2009 (UTC)

Yo Jayron, if you think fallen admins are been replenished with the blood of the anointed, well... look up.  Skomorokh  03:48, 16 September 2009 (UTC)

Ah yes, thanks for bringing that to my attention... So we have actually dropped BELOW replacement rate. Which answers teh question as an affirmative yes. We are currently in the process of having an admin shortfall. Or maybe a "market correction"... --Jayron32 03:53, 16 September 2009 (UTC)

What I'm reading so far makes me cautiously optimistic that we'll be able to get at least a small, helpful tweak of some kind to RFA. Previous battles have been along the lines of "RFA is broken, kill it"; this time, almost everyone is being more constructive and open-minded. - Dank (push to talk) 11:53, 16 September 2009 (UTC)