Jump to content

Wikipedia talk:Requests for adminship/Archive 183

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 180Archive 181Archive 182Archive 183Archive 184Archive 185→Archive 190

Cache of RfA questions

Have fun. Also, please add to that list. @harej 22:36, 31 August 2009 (UTC)

What about a version for RfB? Maybe add a section on RfB questions?--Unionhawk Talk E-mail 15:52, 1 September 2009 (UTC)
I think MBisanz already has such a cache. @harej 20:02, 1 September 2009 (UTC)
I thought that was literally a cache of all past questions. Protonk (talk) 04:09, 2 September 2009 (UTC)
Hmm...that's a good idea... bibliomaniac15 04:22, 2 September 2009 (UTC)
User:MBisanz/Qs is more some pointed questions on past events and proposals, not the broader stroke harej is painting. MBisanz talk 04:38, 2 September 2009 (UTC)
Considering the broader scope of my page, would it be okay if we merged pages? @harej 06:55, 2 September 2009 (UTC)
You can copy content from my page, but I'd rather leave it intact as an MBisanz-specific sandbox. MBisanz talk 19:28, 2 September 2009 (UTC)

numbering of optional / additional questions

Some people ask multiple questions. They'll number them Xa, Xb, Xc, etc. This obscures the true number of questions asked at RfA. Please can we have some other system of numbering questions? It's fine to give different parts of the same question a new number. Xa and Xb are fine as X and Y.
Also, I've started seeing people using additional questions from instead of optional questions from - please can we re-inforce the optional again? NotAnIP83:149:66:11 (talk) 11:08, 1 September 2009 (UTC)

Can you explain we need to keep track of the true number of questions? Dekimasuよ! 14:42, 1 September 2009 (UTC)
I agree with Dekimasu here. Why do we need to know the number of questions? –Juliancolton | Talk 16:12, 1 September 2009 (UTC)
In that case, why are they numbered at all? Giving them a true number makes it easier to grep RfA and count the number of questions, and makes it easier to keep stats about question-loading of candidates. NotAnIP83:149:66:11 (talk) 16:53, 1 September 2009 (UTC)
The numbering serves as a convenient reference for participants. It is far easier to say "Oppose, candidate's answer to Q4 demonstrates a lack of understanding ..." or "Fantastic response to Q6, shows this candidate has loads of clue!" rather than "I am dissatisfied with the candidate's answer to User:XYZ's question about the BLP policy". Shereth 16:59, 1 September 2009 (UTC)
Yes, obviously. And so if they're going to be numbered why not just use a new number for each question? I'm not seeing why this is contentious? NotAnIP83:149:66:11 (talk) 17:37, 1 September 2009 (UTC)
I don't believe it is contentious per se; rather, it is being viewed as unecessary to dictate a certain numbering scheme. Shereth 17:49, 1 September 2009 (UTC)
Ah, I see now. So, re-word my original post to "Please consider using Q1, Q2, Q3 etc instead of Q1a, Q1b, Q1c. There's no urgent reason for doing this; it'd just make counting the number of questions asked per RfA a bit easier. Obviously, feel free to ignore this and number your questions however you like." NotAnIP83:149:66:11 (talk) 18:21, 1 September 2009 (UTC)
Often times the reason is because the subsequent questions are related to the same scenario, or are follow-ups or otherwise connected to each other. A lot of them could easily be condensed into one question, but for sake of answerability, readability, and referencing are broken up. Suffice to say, there are bigger issues. ~ Amory (user • talk • contribs) 19:47, 1 September 2009 (UTC)
You're right, it's not a big thing. But many times questions have abc sections just because they're being asked by one person. There's often very little actual connection between the questions. NotAnIP83:149:66:11 (talk) 10:47, 2 September 2009 (UTC)
If one is asking a follow-up question, and subsequent questions have already been asked, then in order to not use either a peculiar ordering or some kind of decimal, it is easier to use "13b, 13c", etc. –xenotalk 19:53, 1 September 2009 (UTC)
That's a great reason for using the abc format. Look at a few recent RfA to see that people asking follow ups sometimes don't number the questions at all. NotAnIP83:149:66:11 (talk) 10:47, 2 September 2009 (UTC)

For that matter, why are we using '''X:''' Question instead of # Question? @harej 20:05, 1 September 2009 (UTC)

I think it's because of the A: part. testing:
  1. Question 1
A: Answer1
  1. Question 2
A: Answer2
Yep, it's because of the answer.--Unionhawk Talk E-mail 23:15, 1 September 2009 (UTC)
Actually, it works if you do it like this:
  1. Question 1
    A: Answer1
  2. Question 2
    A: Answer2
- Kingpin13 (talk) 01:20, 2 September 2009 (UTC)
It's hard enough keeping the numbering on point in the discussion session. Also, this means if someone adds a question in the middle, the numbers will re-number themselves accordingly, possibly making comments already made no longer refer to the right number. –xenotalk 19:34, 2 September 2009 (UTC)

Additional questions vs. optional questions

This point that NotAnIP had has not really been addressed. It's probably just a minor change in language to indicate that they are supplements to the default three questions asked of all candidates. I think it should be taken for granted that these questions are optional, as all questions are or should be. @harej 07:11, 2 September 2009 (UTC)

Yes, it's just semantics. Some people might forget to put "Optional" but in truth the question is optional, and it isn't, all at the same time. –xenotalk 19:34, 2 September 2009 (UTC)

Question

How do you put a question on a RFA? Or can I even do that?Abce2|Aww nuts!Wribbit!(Sign here) 21:59, 1 September 2009 (UTC)

Yuh, you can do it. Anybody is free to ask questions. Just copy the format of other users, or add {{subst:Rfa-question|Number|Question}} where you want your question to go (replace "number" with the number of your question, and "question" with your question) :). Best, - Kingpin13 (talk) 22:02, 1 September 2009 (UTC)
Okay, thanks.Abce2|Aww nuts!Wribbit!(Sign here) 22:04, 1 September 2009 (UTC)
No problem :) - Kingpin13 (talk) 22:05, 1 September 2009 (UTC)

Activity update: under 900

Hurray! RickBot's new tally is 898 active admins. This is the first dip below 900 since RickBot started its daily updates more than two years ago. I recently took a look at the failed RfAs of 2009 and I'd like to point out that the rising RfA standards only account for the loss if we assume that it's preventing valuable candidates from applying. We've lost 46 active admins since January 1st 2009. If we'd promoted anyone with, say, 60% support we would still face a net loss. Pascal.Tesson (talk) 16:40, 30 August 2009 (UTC)

While I've only been moderately active at RfA recently, there close to zero doubt that the problem is not too high a bar for new admins. Is there a place this is being discussed? Is there a way to do a survey of recent departures to find clarify the nature of the problem? Has someone done a longer term study to determine how losing 46 in nine months compares to longer term averages? In other words is the rate of loss normal or above normal? If it's normal, then are there too few candidates interested in the bit? How does the loss of active admins compare with the loss of active editors? Should there be a survey of existing admins to see if some are close to quitting and what might change their view?--SPhilbrickT 17:10, 30 August 2009 (UTC)
"How does the loss of active admins compare with the loss of active editors?" +1. This is always the key point. There's no major managerial problem if the numbers go down in unison. Dekimasuよ! 17:15, 30 August 2009 (UTC)
Actually, what needs to be level is the number of admins with respect to the total editing activity: whether that activity is the result of more or fewer editors is irrelevant to a certain extent. In fact, to state the obvious, the number of admins should at the very least follow the number of admin actions needed to keep the project running smoothly, but we can't measure the latter precisely. Mbisanz recently gave interesting examples of admin backlogs that are getting way out of hand.[1] As far as attrition statistics go, here's a bit of (imperfect) info:
I'm not sure how to interpret that data but I'm absolutely certain that it's worrisome. We've started off-wiki outreach programs to attract more quality editors and we might want to start thinking about how to attract or develop admins. Pascal.Tesson (talk) 18:48, 30 August 2009 (UTC)

{unindent} The problem with only having two years of data is that we have no sense of context for it. Perhaps '08 was low on the desysopping or resigning front; maybe '08 we had a higher-than-average promotion rate (that data could be found relatively easily, though). Numbers are just numbers. Are there CSD backlogs? Are we truly suffering by having fewer promotions? If not, perhaps the project is just (naturally) finding its sysop equilibrium. Doom and gloom is premature, perhaps. EVula // talk // ☯ // 19:44, 30 August 2009 (UTC)

A bit off the point, but I've noticed as of late that CSD is indeed perpetually backlogged. Same goes for WP:TFD, WP:IFD, WP:RFD, WP:RM, etc. –Juliancolton | Talk 19:52, 30 August 2009 (UTC)
Is there a place that lists those backlogs, and how large/urgent the backlog is (preferably in a convenient, easy-to-read tabulated form :))? I'm now in gradschool and I don't have nearly as much time to edit as I once did (and certainly not as indepth with articles/research); however, I feel I could make my Wiki-time more efficient if there were some backlog page that basically gave a list of admin areas in particular need of attention. Lazulilasher (talk) 20:44, 30 August 2009 (UTC)
Well, you can just look at the pages in Category:Administrative backlog for example. Regards SoWhy 20:46, 30 August 2009 (UTC)
That's pretty much what I was looking for. Thanks, Lazulilasher (talk) 20:49, 30 August 2009 (UTC)
(←) You are most welcome :-) Back to the topic at hand though: Anyone who thinks we need more admins, should consider approaching people who have professed their willingness to become an admin and nominate or coach them. This list for example lists more than 250 active editors who have done so, I am sure there are some good candidates on it. Regards SoWhy 20:58, 30 August 2009 (UTC)

To EVula: doom and gloom may indeed be premature but reflection is certainly in order. We can't wait until we notice serious problems before thinking of solutions. To answer your basic question: there is growing evidence that some admin backlogs are becoming problematic (see my link to MBisanz' earlier comments). The notion of "sysop equilibrium" is, with all due respect, nonsensical. If by "equilibrium" you mean "stabilized at a certain level" or "sufficient for the current needs" then that's not what the current data suggests. But, and correct me if I'm wrong, I think you mean "equilibrium" as in "not too many, not too few". If we had 3000 responsible, experienced editors with sysop tools, we'd be all the better for it and if we can find reliable ways of attracting, developing and keeping good admins, we should invest time and effort for that. In particular, a larger admin corps would help curb the perception of sysops as all-powerful potential tyrants. Pascal.Tesson (talk) 22:46, 30 August 2009 (UTC)

One problem with the backlogs, as I've previously said, is that of timezones; for obvious reasons members of the admin corps tend to be from English-speaking countries in specific timzones, and as a result you've got certain hours where very few admins are "on duty". One idea would be to focus on candidates from say, India - not afford them any additional weight as a candidate, but make an effort to seek out candidates from certain areas. This might have the additional bonus of reducing any perceived systematic bias within the corps. Ironholds (talk) 23:38, 30 August 2009 (UTC)
There are 20,000 entries on Wikipedia:New histmerge list that have not been touched for weeks, so I do not see how the timezone issue is creating these backlogs. MBisanz talk 00:21, 31 August 2009 (UTC)
Part of the issue there is that it's a new page. There's no one who was working there and stopped; there was no way to gather that data before, so it started out backlogged that much. It's just that almost no one has started working there yet. That's a different thing from saying that no one performs history merges; they've been going on at WP:SPLICE and WP:RM for a long time. Dekimasuよ! 01:53, 31 August 2009 (UTC)
It was just a (small) part of the problem - I never claimed it was the entire thing. Speaking as someone normally up at 3-4am GMT I have noticed a complete dropoff of admin activity as most western nations head to bed. Ironholds (talk) 00:33, 31 August 2009 (UTC)
I see, you're saying we could partially solve the problem by making you an admin. Worth considering. - Dank (push to talk) 01:38, 31 August 2009 (UTC)
Well it would put someone around at odd hours and add to the admin cabal, so everyone is happy. Seriously though, I just meant people from the wee hours, not meself. Ironholds (talk) 01:40, 31 August 2009 (UTC)
Oh I know you weren't saying it is the only factor, I'm just pointing out that by some measures the backlogs are worse than ever (CSD moving from 60 to 200 instead as bad as CAT:TEMP going from 5,000 to 30,000). MBisanz talk 02:08, 31 August 2009 (UTC)
I would expect a roughly proportionate decrease in need, no? I suppose it is possible that vandals are more prevalent int he wee hours of the US morning, but I'd like to see some stats before assuming that geographical scope is the top or even a pressing priority.--SPhilbrickT 00:39, 31 August 2009 (UTC)
This is OR and getting away from the topic a bit, but when I wake up somewhere between 6 and 7 EDT, there's usually a few counts of vandalism on my watchlist stretching back to the 2-4 range; during the day, vandalism to those pages definitely doesn't last that long. There's a decrease in overall enWiki activity in the latest of hours, but I bet some vandals think it's worth it to wait up late while sysops make the smart choice and get some sleep. I don't know that it's a problem, per se, but it definitely plays a role. ~ Amory (user • talk • contribs) 01:35, 31 August 2009 (UTC)
There's also the problem, of course, that many areas where the wee hours for us are the early morning/evening/lunch break for them speak sufficient english to work their way around the mediawiki interface. It might not necessarily be that there are admins waiting with bated breath until Pedro goes to bed, it could just be that vandals come from odd timezones even if admins don't. Ironholds (talk) 01:40, 31 August 2009 (UTC)
It's safe to assume a proportional decrease in page views at that time of day. During the day someone might vandalize an article, and the next person to view the page (a minute later) might revert. During the night someone might vandalize an article, and the next person to view the page (two hours later) might revert. If that's the case, simply lasting longer isn't a major concern; vandalism lasting longer doesn't necessarily mean it's been more effective vandalism. Dekimasuよ! 01:59, 31 August 2009 (UTC)

<<-- I don't think the time zone issue is a problem. There aren't that many admin actions so urgent that two hours make a difference. "Speedy deletion" doesn't mean "has to be deleted pronto" (with the exception of attack pages of course). I don't really care if a spam page or a copyrighted text is up there for 24 hours if it maintains a db-tag (and the edit-filters help keep the tags in place). What we want to avoid is the sort of situation I saw as a new admin where this backlog sometimes grew way out of control. The problem is not the backlog itself but what it creates: as backlogs grow, admins end up doing longer stints of admin-only work and inevitably grow both frustrated and sloppy. I have little hope for the history-merge backlog: attacking it is the ultimate thankless task and it's hard to prioritize it over other admin tasks. Anyhoo, we shouldn't worry about the time-zone thing unless we get evidence that it creates significant delays for AIV, attack-page deletion or urgent protection requests. Pascal.Tesson (talk) 02:59, 31 August 2009 (UTC)

The real question, in my opinion, is not the number of admins, but rather the number of admin actions performed (helpfully logged at WP:ADMINSTATS). I've gone back to April, which is all that it is possible to conveniently do by hand (someone skilled at scripting should have a look), and there were a lot less actions performed in August (93,279) and July (92,506) than June (117,931), May (104,576) and April (129,471). Of course the shifts in number of actions are much more pronounced than those in number of active admins, and with such a small sample it's hard to put forward anything vaguely concrete. Cool3 (talk) 04:28, 31 August 2009 (UTC)
There are normally less logged actions during the summer because kids have better things to do than vandalize Wikipedia from homeroom. J.delanoygabsadds 04:35, 31 August 2009 (UTC)
And then during the school year, we can have an admin snap and ban every high school in America. @harej 05:38, 31 August 2009 (UTC)
Fewer actions, not less. Stifle (talk) 08:25, 31 August 2009 (UTC)
The number of admin actions seems like a reasonable metric but it might be misleading because it doesn't match the number of necessary admin actions. Consider the two extreme cases. If we had 1800 active admins instead of 900, the number of admin actions would most likely rise. We could expect more history merges from the backlog, shorter protection/unprotection cycles, etc. In other words, a very large admin corps would potentially address every little bit of admin-work to be done and we'd reach some sort of ceiling that truly reflects the project's needs. On the other hand, if we only had 450 active admins, the number of admin actions would necessarily plummet regardless of the number of admin-actions that should be made. My gut feeling is that we're approaching the point where the limiting factor on the number of admin action is the number of admins. Another reason for taking the admin-actions stat with a grain of salt is the work of admin-bots. I don't mean this as a cheap shot to MzMcbride but in March 2009, his automated scripts performed over 40000 deletions (no, I did not add an extra 0, forty thousand!). Pascal.Tesson (talk) 15:11, 31 August 2009 (UTC)
I think this is an important point... yeah there may be fewer admins today than there used to be, but there is a certain degree to which fewer admins are needed. First, I think the front line vandal fighters are better equipped to deal with more issues than they used to be. Various tools endable vandal fighters to stay on top of the issue and report problematic users quicker than they could a year ago. Some of those tools allow non-admins to perform tasks that emulate some admin functions and some admin functions (namely Rollback) have been passed on to others. Other tools allow admins to perform admin tasks much quicker than it used to. When I changed my name back to Balloonman, there was a major snafu with some of my subpages. The 'crat who changed my name and I were working on cleaning up those pages when I said, "This is crazy, I bet you somebody more familiar with tools could do this a lot quicker." I made a call out for talk page stalkers and Xeno came to our rescue. He used a bot to do in a matter of minutes what it would have taken me hours to do manually---and I later made a second request to Xeno perform another admin function that again would have taken me hours to do manually, but because he used tools did so in a matter of minutes.---Balloonman NO! I'm Spartacus! 20:47, 31 August 2009 (UTC)
That difference is attributable to Slakr's open proxy blocking bot going down the first week of July. MBisanz talk 18:08, 31 August 2009 (UTC)
The RFA problem was raised during a session at Wikimania, (probably this one). Apparently the German Wikipedia is in even worse state than us with 10,000 edits becoming a de-facto minimum. Apart from the good editors who've stopped editing after an unsuccessful RFA, we should consider the editors who quite sensibly won't run the gauntlet with RFA in the state it is now. As for zero doubt about the current bar being too high - at that session of Wikimania no-one disputed that the bar is now so high that practically none of the EN admins present could now pass an RFA. ϢereSpielChequers 19:48, 31 August 2009 (UTC)
I looked back to see who made that stupid comment about the height of the bar, and I see it was me. I’ll stand by my comment but plead poor wording choice. I did not mean to suggest that the bar is set just right, but there’s no question I said something like that. I’ll try again.
It’s also possible I misread Pascal’s point--I read it as saying we have a net loss of 46 admins since 1 Jan. There were 157 failed RfAs in that time period, so to just remain even, we needed to promote 46 of these 157. I didn’t do a scientific study, but I doubt that dropping the rate to 50% would pick up 46 additional admins. 50% may win a political election, but it isn’t consensus. So my comment about the bar wasn’t meant to say that the bar is exactly at the right place, but that to lower the bar enough to keep the admin count constant, which I interpret as dropping below 50%, is not a feasible option.SPhilbrickT 20:42, 31 August 2009 (UTC)
I didn't say that we should have promoted 46 of these 157 failed RfAs. To the contrary I wanted to point out that the number of viable candidates among these 157 was below 46 and that we would have had a net loss of admins even if we'd lowered the bar. Pascal.Tesson (talk) 22:15, 31 August 2009 (UTC)

Back when I created the category system for RFA, it pointed to a longterm trend that about 44% of RFAs were successful (there is a historical post with the figure here) and with 86 promotions in 158 RFAs so far this year, that gives a promotion rate of 54%. The issue isn't that we are promoting fewer people who run, it is that in 2007 we promoted 408 people to admin and in 2006 we promoted 353, while rejecting 512 and 532 respectively, while this year we are on track to promote 115 admins and reject 96. MBisanz talk 21:07, 31 August 2009 (UTC)

So basically, the problem is that the rate at which current admins become inactive, is higher than the rate at which new admins are enlisted. This would be due to less people applying for the tools than in previous years, because I can imagine the rate at which admins become inactive is at least somewhat constant over the years. I think the main issue here is that RFA has an image problem due to (perceived) constantly inflating standards, that make prospective candidates less inclined to apply. Regarding myself, I find that I am always some steps behind current RFA standards. I'd likely pass with my current level of contributions some 2 years ago, but I'm not so sure now. I can imagine other editors feel the same.--Atlan (talk) 21:47, 31 August 2009 (UTC)
Are you only including RFAs that run to completion? Otherwise I'm pretty sure the numbers for this year are too low. Looie496 (talk) 22:26, 31 August 2009 (UTC)
All of them use the same metric and are comparable. MBisanz talk 03:13, 1 September 2009 (UTC)

I believe that many people fail to seek adminship because they do not believe it is worth the trouble to meet the ever-rising standards. That is part of the reason for the lower number of requests. I also believe that a smaller fraction of the new editors who are attracted to the project are likely to make good admin candidates. In particular, those candidates who are involved in multiple non-Wikipedia related tasks are unlikely to be able to maintain the now-necessary "activity levels." The Uninvited Co., Inc. 20:44, 2 September 2009 (UTC)

I have been following this discussion tacitly so far. In particular, I note the comments by SoWhy, WereSpielChequers and UninvitedCompany. Wasn't there a big survey about the RfA procedure several months ago by Ultraexactzz? I don't recall seeing any conclusions. Axl ¤ [Talk] 23:40, 2 September 2009 (UTC)
I maintain that RfA is not the problem. Age of the community, editor drift and slowing editor growth are the problems. Protonk (talk) 21:48, 3 September 2009 (UTC)

For those who like visualizations, here is a graph of the active admin count (taking the monthly average for each month) for the last two years. tools:~cbm/admins200909a.png The rate of decrease is about 6/month, and has been relatively constant for more than a year. — Carl (CBM · talk) 12:23, 4 September 2009 (UTC)

Strong or weak oppose makes a difference?

Apparently it does matter if you put 'weak' on your vote at RfA. See Taxman's close of MZMcBride's RfA. I'm not contesting the close (I voted oppose, but it rightfully passed). I'm just concerned that 'weak' is being given consideration. I think I'll mark my votes as "strong" from now on, just to be sure I'm heard. --Hammersoft (talk) 14:10, 4 September 2009 (UTC)

I should hope it makes a difference, otherwise what is the point? Moreover, if b'crats were ignoring such qualifiers, then what other parts of the discussion are they allowed to ignore? Since it is a discussion, not a vote, how we qualify our support/opposition ought to be every bit as valid to the discussion as the rest. Shereth 14:16, 4 September 2009 (UTC)
(edit conflict) It does, and should do. That said, were I a bureaucrat, I would pay more attention to the weak !votes than the strong ones, as in close cases, the importance in determining the collective will is at the margin rather than the extremes (that is, it is better to have weakly-supported administrators than polarizing ones). Reading the weak and netural comments can highlight the issues at the core of the debate and thus enable the bureaucrat to better judge the concerns of the opposition.  Skomorokh  14:17, 4 September 2009 (UTC)
(edit conflict) Personally, I'm against the whole weak/strong thing, but Taxman's point is solid. If there is not only consensus for a pass, but those opposing state themselves that their own oppose is not very well grounded, that does give some weight to the support view. It can't really work the other way (Everyone would say their own is strong, so nobody's would be). Everyone thinks their opinion is the most important, so saying that your own view is strong won't account for much. Having the self-awareness to note that your view is weak, however, is an entirely different beast. ~ Amory (user • talk • contribs) 14:21, 4 September 2009 (UTC)
I'm of the opinion that the strength of a vote should be determined by the arguments therein, rather than explicit statments in bold font. Still, Taxman has a valid point. –Juliancolton | Talk 01:25, 5 September 2009 (UTC)
That is why I do not give bolded statements. @harej 01:29, 5 September 2009 (UTC)
Let's not conflate efficacy with intensity.  Skomorokh  02:49, 5 September 2009 (UTC)
Note that this is not an opportunity for editors to type "support" in size 50 font, and with flashing colors. —Dark 02:52, 5 September 2009 (UTC)
Oppose, typographical emphasis can be helpful. Short Brigade Harvester Boris (talk) 02:57, 5 September 2009 (UTC)
As a bureaucrat, I can honestly assure you that I in no way, shape, or form care about the bolded phrase in front of a person's RfA comment. "Strong Oppose User was blocked for 3RR last month" is no different than "Weak Oppose User was blocked for 3RR last month." Arguments matter, phrases do not. EVula // talk // ☯ // 04:12, 5 September 2009 (UTC)
Disapointing, EVula. As an actor one would have assumed you knew the difference between a stage whisper and an actual whisper. It's not complex (well obviously it is I guess) to work out that opposes that use SCREAMING ALL CAPS or BIG FAT BOLD comments are more heartfelt than others - ditto supports. Assuming that Wikipedia is a textual medium (which I believe it is) then bolds, colours, size are simply the only way to add emphasis. Pedro :  Chat  22:19, 5 September 2009 (UTC)
I think EVula was getting at "no one cares about your emphasis; it's the argument that counts." @harej 22:25, 5 September 2009 (UTC)
Yes, - And of course EVula is wrong for the reasons I outline above. Pedro :  Chat  22:34, 5 September 2009 (UTC)
But... but... but... WP:EVULA is policy! (X! · talk)  · @994  ·  22:51, 5 September 2009 (UTC)
Ah - yeah........ Pedro :  Chat  23:00, 5 September 2009 (UTC)
Hey, I never once mentioned the value of vocal projection when it comes to RfA participation. ;)
All kidding aside, you do raise a valid point. However, if the entirety of someone's argument is in their bold phrase, that means they have an amazingly piss-poor argument. At most, the bold phrase is an indicator as to the tone of their argument, but it isn't, or at least shouldn't be, the end-all/be-all. It should be the TLDR version, nothing more. (I do believe you are correct about the use of bold and italic formatting to take the place of otherwise vocal inflection modifiers in a text-based medium; that is definitely how I use them, but "This user was blocked six months ago for a 3RR violation" simply cannot be properly summed up in a two-word bolded phrase, not matter what the adjective prefix may be) EVula // talk // ☯ // 05:40, 6 September 2009 (UTC)
And I always thought EVula was a sucker for colorful text. eh... Nevermind. :) —Dark 13:51, 5 September 2009 (UTC)
Maybe it was colorful "conversation" ... ;P — Ched :  ?  14:04, 5 September 2009 (UTC)
Out of curiosity, would you give more weight to "Oppose, user was blocked for 3RR last month" than "Neutral, user was blocked for 3RR last month"? –Juliancolton | Talk 15:21, 5 September 2009 (UTC)
Or better yet, make it six months ago: "Oppose, user was blocked for 3RR six months ago." vs "Neutral, user was blocked for 3rr six months ago." vs "Support, user was blocked for 3rr [but that was] six months ago." Strength of support does play a role in helping to determine how the author of the post gives weight to their own argument. Yes, the argument matters more, but in this forum, the weighting can help determine how people view it. I generally see weak oppose/weak support as "I'm opposing/supporting, but if it goes the other way, it won't bother me too much."---Balloonman NO! I'm Spartacus! 15:46, 5 September 2009 (UTC)
I have faith in the intelligence of the average RfA participant to put their !vote in the correct place. If someone opposes because a 6-month old 3RR block, that's valid; if they support because the 3RR block was 6 months ago, that's valid too (same goes for if they go neutral with the comment). So, to answer your question, yes, I'd give more weight to the Oppose !vote, if only because of the section it was placed in. If there was additional commentary, I'd weigh them differently perhaps, but given this exact example, yes. EVula // talk // ☯ // 16:58, 5 September 2009 (UTC)
EVula, you should accept that Juliancolton, and Balloonman have proven that your this comment is invalid. "Strong Oppose (meaning: if the RFA passes, I'll be disappointed) User was blocked for 3RR last month" is different than "Weak Oppose (meaning: if the RFA passes, it won't bother me) User was blocked for 3RR last month." While determining consensus, bureaucrats should pay at least some attention to 'strong' and 'weak'. AdjustShift (talk) 20:04, 5 September 2009 (UTC)
Or more precisely, "Strong Oppose, (meaning: I find this action to be of serious consequence) User was blocked for 3RR last month" is different from "Weak Oppose (meaning: I find this action enough to be of minor concern) User was blocked for 3RR last month." The descriptor of the vote (whether Support/Oppose/Neutral/Strong/Weak) gives context as to how the poster views the issue and should be weighed in evaluating consensus.---Balloonman NO! I'm Spartacus! 04:39, 6 September 2009 (UTC)
I have to accept no such thing! Just because you think I'm wrong doesn't mean that I actually am; see item number two on WP:EVULA. :P
To be perfectly blunt, whether someone will be bothered by an RfA's passing or not is totally and completely irrelevant (referencing your "I'll be disappointed" and "it won't bother me" examples, contrasting Strong and Weak opposition). With the exception of 100% supported RfAs, there's always going to be someone that disagrees; my job, as a bureaucrat, is not to care about personal feelings, but to gauge consensus. I gauge consensus off of the actual arguments being levied, not on adjectives. "Super-Duper-Times-One-Million Oppose The editor doesn't like cats" is a mind-numbingly moronic argument, despite (or perhaps because of) the bold phrase, whereas "The candidate says that they plan to work on AfD requests, yet has shown very little participation there in the past year" (note the lack of a bold phrase) is a solid argument; three guesses as to which one a bureaucrat would give credence to, and which would cause the 'crat to roll their eyes at. If an editor can't make their argument (and thereby their position) clear with their actual text (again, arguments matter, phrases do not), they should try brushing up on their writing skills; we've got an entire encyclopedia with more than 3 million options to choose from. ;) EVula // talk // ☯ // 05:54, 6 September 2009 (UTC)
While a Vorporal Sword +15 of Support is meaningless, the weak/strong/support/oppose/neutral do have a bearing in how the poster views the evidence. One thing that is unmistakable is that different people can view the same evidence differently and reach different conclusions. The weak/strong help provide context to how the argument was intended.---Balloonman NO! I'm Spartacus! 21:31, 6 September 2009 (UTC)

If we accept that neutral == 'dont care', and give that a value of 3, then we can say oppose == 2 and strong oppose == 1, and support == 4 and strong support == 5. Crats just need to total the NotVote values, and divide by the number of editors, and compare the result to number of editors * 3. Then we can have ranges for pass, fail, and tricky. This could be scripted and there could be an RfACloseBot to promote candidates. NotAnIP83:149:66:11 (talk) 22:03, 5 September 2009 (UTC)

Or, in a more sensible system, strong oppose is -3, regular oppose -2, weak oppose -1, neutral 0, weak support +1, regular support +2, and strong support +3? Then we can take the square root of it and multiply it by Avogadro's number. @harej 22:16, 5 September 2009 (UTC)
1.04306551x1024 per WP:EVULA. ~ Amory (user • talk • contribs) 00:19, 6 September 2009 (UTC)
Though it seems silly to accept the RfACloseBot suggestion as an actual way of doing things (considering the fact that the bold phrase can just as easily be misleading), your addendum above suggests that neutral !votes aren't important in gauging consensus; that is wrong. They aren't as important to bureaucrats as, say, the Support and Oppose sections (since we couldn't really have an RfA without somewhere for the !votes to go, could we?), but on a close RfA, the arguments presented in the Neutral section can help swing a decision from one end to the other. If nothing else, there are arguments presented in the Neutral section that may, in turn, influence participants; one bit of evidence may cause one editor to be neutral, but another (perhaps more stringent) editor to oppose. EVula // talk // ☯ // 06:02, 6 September 2009 (UTC)
Where do weaks fall into that system? hmwitht 00:16, 6 September 2009 (UTC)
And then divide by c3… --Izno (talk) 00:21, 6 September 2009 (UTC)
THIS IS RfA - THERE IS NO ROOM FOR THE WEAK HERE. CRUSH THEM, MASH THEM, CHEW 'EM UP AND SPIT 'EM OUT. NotAnIP83:149:66:11 (talk) 22:01, 6 September 2009 (UTC)

Proposal: No rationales in RfA votes

Hello everyone. A big problem in RfAs is the whining and complaining over other users' rationales (in both the support and oppose sections). In the ArbCom/CU/OS elections, users are only supposed to add their signature to either section, and leave any comments they have to get out, in the discussion section. I think we should do that here. Maybe you'll say, "No, then we're not giving any advice to the candidate." Do the signatures in the ArbCom/CU/OS elections give any advice to the candidate? No. And it seems to not be a big problem. Maybe you'll say, "RfA is not a vote, only using your signature removes the discussion aspect." Well, if we hold our ArbCom/CU/OS elections in a vote, and can call it a vote, why can't we do the same for RfA? What are your thoughts on this? iMatthew talk at 22:56, 1 September 2009 (UTC)

See WP:PEREN. It's been long established that direct voting is a bad idea, particularly since it turns it into a "who is most loved" contest at best. With people not having to give rationales everyone who's ever had a gripe against the candidate will line up to hit the oppose section. Ironholds (talk) 23:02, 1 September 2009 (UTC)
How does that make ArbCom/US/OS voting OK? Those elections should be taken more seriously, and according to what you just said, that can happen there. iMatthew talk at 23:24, 1 September 2009 (UTC)
Indeed, and at no point did I claim otherwise - indeed, in my opinion ArbCom elections are essentially "who do we like best". Ironholds (talk) 23:28, 1 September 2009 (UTC)
Ironholds hits the nail on the head. Just because Arbcom is a farce doesn't mean we have to turn the rest of the project into a farce just to be like them. – iridescent 23:30, 1 September 2009 (UTC)
Why would we? RfA is clearly satire, not farce, and as such is the superior form of comedy. → ROUX ₪ 23:33, 1 September 2009 (UTC)
When did not having a valid reason to oppose ever stop anyone?—Kww(talk) 00:05, 2 September 2009 (UTC)
Oppose - Kingpin13 (talk) 23:51, 1 September 2009 (UTC)
I might have laughed if you spelled "oppose" right. :) iMatthew talk at 23:56, 1 September 2009 (UTC)
Oppose - Too many admins currently oh wait a minute... MLauba (talk) 00:03, 2 September 2009 (UTC)
That being said, since WP:PEREN gets quoted an awful lot these weeks, consensus can change. MLauba (talk) 00:04, 2 September 2009 (UTC)
@IMatthew My apologies for robbing you of a good laugh :D. It's the tea see, makes my hands shake. Anyway, seriously, I don't think we should adopt this purely because it's what another part of the project does. I've gotta agree with Ironholds about the ArbCom/US/OS nominations being "who do we like the best". And I also think if we adopted this it would make RfA much less welcoming (that's saying something ;D). And harder for an editor who finds something seriously bad in the candidates history, which everybody else has missed, to point it out. It stands out more who you are addressing if you comment under their !vote, then if you comment in the discussion section. - Kingpin13 (talk) 00:07, 2 September 2009 (UTC)
Rather than trying to emulate ArbCom's essentially unhelpful approach, we should try to make the ArbCom "elections" a discussion instead. Last time I checked, WP:NOTDEM was still policy and rightly so. We don't want to have good candidates barred from adminship just because a lot of people scream "I don't like you" without being able to find a fault in the candidate's contributions nor would we want bad candidates pass just because they amassed a fan club but are otherwise completely clueless. ArbCom's approach essentially means that the most popular candidates will get chosen which might not always be the most competent ones (and could ultimately lead to an ArbCom composed of popular but completely clueless editors). I do not think that's the approach we should have to adminship. Regards SoWhy 07:15, 2 September 2009 (UTC)
In the same vein (and that's bound to become another perennial proposal if it isn't already), maybe we should really reconsider, again, ending the segregation of the support / oppose / neutral sections and have one single continuous voting block. That would foster the discussion aspect and diminish the vote side. MLauba (talk) 08:09, 2 September 2009 (UTC)
Commons does this, and it works okay due to the small numbers of participants in requests over there. It was tried here in Wikipedia:Requests for adminship/Moralis, but the closing bureaucrat found it difficult to interpret a consensus. Camaron · Christopher · talk 11:24, 2 September 2009 (UTC)
Thanks, Wikipedia:Requests for adminship/Moralis was what I was thinking of indeed. That being said, if the only drawback is that it makes it harder on the 'Crats (who are after all also being evaluated by looking at their AfD closures, which uses a similar format), I'm not sure that should stop us. MLauba (talk) 11:49, 2 September 2009 (UTC)

I'm concerned the original editor thinks that discussion is whining and complaining. Maybe something could be worked out to allow notvoting and discussing to co-exist more happily. Perhaps any discussion of a not vote goes to the talk page? NotAnIP83:149:66:11 (talk) 10:42, 2 September 2009 (UTC)

The discussion that comes with the votes helps the crats understand what the arguments are and how to weight the votes. Some voters get into fist-fights with other voters, some say "because of (thing that the candidate can't do anything about)", some say "because of (thing that later turns out not to be true)". Crats are typically tight-lipped about how they weigh things, which is good up to a point, but we've gotten hints through the years that there are some rationales they don't find persuasive. - Dank (push to talk) 12:59, 2 September 2009 (UTC)
Does anyone have any evidence of NotVotes being weighed heavy or light by the crats? A link to any RfA in this year would be fine. NotAnIP83:149:66:11 (talk) 15:57, 2 September 2009 (UTC)
Wikipedia:Requests for adminship/Davemeistermoab is the most obvious recent example. – iridescent 16:05, 2 September 2009 (UTC)
And I haven't seen a counterexample over the last 20 months that I've been watching ... that is, an RFA where the result seemed to depend on counting votes that you would think would be discounted. - Dank (push to talk) 16:54, 2 September 2009 (UTC)

Eliminating the rationales in RfA participation removes the "high-level editor review" aspect that can be helpful to a receptive candidate. Even a successful RfA can have helpful feedback in it (by both opposer and supporter). EVula // talk // ☯ // 16:04, 2 September 2009 (UTC)

I often write fairly verbose rationales and would find it frustrating to have my hands tied. Although I don't like to brag about it, I think I can take a large share of the credit for tipping a recent RfA from unanimous support to an ultimately failing score by discovering things the candidate had done that weren't really that impressive. If I had been forced to just type "oppose" people would be left wondering what my reason was. Likewise, when I vote support on an RfA that has almost no chance, I like to give good reasons for my support, even if they have no chance of reversing the course. -- Soap Talk/Contributions 17:08, 2 September 2009 (UTC)
Wikipedia:Requests for adminship/TenPoundHammer 7 is a good example of an RfA where both Support and Oppose !voters take the time to explain their reasonings. I suppose this tends to happen on 'controversial' RfA's more often than 100 percenters. -- Soap Talk/Contributions 19:25, 2 September 2009 (UTC)

(←outdent) Am I missing something blindingly obvious in this example of votes at the 2008 ArbComm elections? The comments are kept brief and there are limited exchanges occasionally, not the mere signature stamping I got the impression from above. -- Александр Дмитрий (Alexandr Dmitri) (talk) 21:44, 2 September 2009 (UTC)

Most of them are indeed signature stamping, and the comments are either brief or useless; "In trying to determine how to vote, I asked Jesus, what would he do, and as you can see, this is what he told me. If Carcharoth is good enough for Jesus, he's good enough for me.", for example. Ironholds (talk) 22:03, 2 September 2009 (UTC)
Not the best of rationales I admit. My point was that discussion was not strictly forbidden in the election. A moot point now that I have just read that ArbCom elections will be held via secret ballots. -- Александр Дмитрий (Alexandr Dmitri) (talk) 22:36, 2 September 2009 (UTC)
A spectacularly bad idea motivated by worrying about the wrong problems, IMO. It's not like anyone has ever made convincing accusations that arbcom was too secretive. But I don't have too much room to complain since I didn't get off my ass and make a statement on that RfC. Protonk (talk) 13:25, 3 September 2009 (UTC)
  • At my recent RFA, I was actually able to turn an opposer to neutral by pointing out some flaws in their rationale. Now, it wasn't such a close decision that it made the difference between me getting the mop or not, but I think it demonstrates the usefulness of explaining your position, and the usefulness of replying to flawed reasoning. I also pointed out where someone else had posted something that was a flat out lie, and they struck it out although they did not change their vote. This at least stopped other users from using it as a convenient oppose rationale. All this talk about "badgering opposes" is a lot of BS as far as I'm concerned. If you attack every one of them and insult them and make baseless accusations, you will invariably fail the RFA, which is what your opposition wants anyway, so there's no real problem in that case. Although it ultimately does come down to the raw percentage in most cases, this is supposed to be a discussion. How many RFAs have been littered with "per whoever" votes. If you have no idea why some are supporting and some are opposing, you may miss out on a point or detail that escaped your attention. Although I am a strong supporter of maintaining civil discourse, I also think we should be able to speak our minds. As long as it not done in an overly disrespectful manner there's nothing to be lost and much to be gained by thoroughly discussing the reasons a person should or should not be an administrator. Also, all the lazy sycophants who only ever vote based on the opinion of someone whose ass they wish to kiss would have nowhere to go anymore. Beeblebrox (talk) 02:21, 3 September 2009 (UTC)
I think it essential for the process that opposers give a valid reason and be prepared to discuss their reasoning. In my last RFA I responded to four of the ten opposers, three of whom struck or reversed their position during the RFA and the fourth tried to half an hour after it closed. Such dialogue makes the difference between this being a job interview and it degenerating into an election or popularity contest. ϢereSpielChequers 22:39, 10 September 2009 (UTC)

Research results

I performed a study of administrator activity, principally using WP:LOGACTIONS, what follows are my process and results.

The total number of admins who have been admins since the logging system was introduced is somewhere around 1,836, I used the rule of thumb that if an admin has performer two or fewer actions, it was probably either an admin who went inactive before logging was introduced or an error. That left a sample of 1,797 admins with three or more actions. I also removed all bot accounts which brought it down to 1,789 admins in the sample.

In the sample, the average admin performed 4,034 actions, with 349 performing more than that number and 1,441 performing less than that number. The median number of actions for the entire sample was 934.

The top decile of admins by actions collectively performed 67% of all actions, the top quintile performed 81% of all actions, and the top 5% performed 54% of all actions.

The bottom decile of admins by actions collectively performed 0.04% of all actions, the bottom quintile performed 0.3% of all actions, and the bottom 5% performed 0.009% of all actions.

The top 50% of admins performed 96% of all actions and the bottom 50% performed 4%.

Obviously this method fails to take into account things like edit protected requests, editing the MediaWiki namespace, things that occurred in WP:OLDLOG and things like Arbitration Enforcement and General Sanctions. Also things that emanate from the penumbra formed by the admin tools such as RFC closing, mentoring, etc are not recorded.

I can do more analysis if people are interested. MBisanz talk 01:12, 5 September 2009 (UTC)

What is your conclusion, Herr Bisanz? @harej 01:15, 5 September 2009 (UTC)

Almost perfect Pareto principle. --JayHenry (talk) 01:26, 5 September 2009 (UTC)

I think the missing piece of analysis is a chart of active admins over time using this metric. People have been wringing their hands over the reduction in admin count. If the admins that are leaving are coming out of the active sector, that's a different problem than finding that admins that are leaving never performed administrative work anyway.—Kww(talk) 14:12, 5 September 2009 (UTC)

Well said. MBisanz: can you (or we) drill down into this data to look at what percentage of actions specific admins were responsible for? Alison comes to mind as someone who gave up the mop, and probably left a massive gap. If the top 5% account for over half the admin actions taken, the individuals represented in that top 5% each represent a fairly critical stone in the project's foundation. Hiberniantears (talk) 15:51, 5 September 2009 (UTC)
I'll try to do more of this later, but just to answer your specific and easy question, Alison is the 156th most active admin and is responsible for 0.1% of all admin actions. MBisanz talk 16:32, 5 September 2009 (UTC)
But what with people who ran scripts on their accounts? For example, MZMcBride's 800,000+ deletions will probably distort those numbers quite significantly, won't they? Regards SoWhy 16:54, 5 September 2009 (UTC)
The difficulty is determining who ran scripts when and how did they run them. It just doesn't seem practical to try to quantify that difference. MBisanz talk 23:59, 5 September 2009 (UTC)
It would be an interesting drill down on the data, but I think MBisanz would probably go out of his mind trying to pull that off. Importantly, using scripts for admin tasks is still doing admin work, and that work would not be done in that volume if that admin had not been promoted to admin based on their record. Granted, MZMcBride's use of scripts for admin tasks landed him in hot water, but this doesn't take away from the fact that they were legitimate admin tasks carried out by an admin. I see scripts on admin accounts as different from editors using Huggle or some other automated tool, because anyone can do that. Anyone can't be an admin though. Likewise, while we can look at the volume of administrator action and break it down by admin, it is far more complicated to look at each of those tasks and judge whether they were a correct or incorrect application of the mop. So if MZMcBride has two million admin actions, it doesn't matter how he did them, simply that they were done through him. The MZMcbride saga might actually be a good test case because of his removal of the mop for a period of time (appologies to MZ for citing this example). Was there a decline in admin actions during this period, and a related increase in backlogs, or does the admin corps really function as more of a hive and adjust to personelle changes rapidly? Hiberniantears (talk) 17:00, 6 September 2009 (UTC)
I don't think there was an increase in backlogs in that case since MZ's automated scripts did a job that has no backlog associated with it and in fact has become unaccepted to do (if I remember correctly), so there should not be a correlation. Still, I think it would be good if there were a way to split off fully automated (= bot) edits made by an admin through their account from those numbers. Using scripts is legitimate when supervised, no question but if a script runs unsupervised, then it's a bot and should be treated as such in those statistics, otherwise they give a wrong impression (for the same reason admin bot accounts are not included). Regards SoWhy 19:45, 7 September 2009 (UTC)

"the top 5% performed 54% of all actions"? Wow. I make that 90 people. Rd232 talk 22:03, 6 September 2009 (UTC)

Yep, 89 people to be exact. MBisanz talk 03:50, 7 September 2009 (UTC)
89.45 people, to be too exact. Rd232 talk 23:06, 10 September 2009 (UTC)

Here is a chart of the month of last admin action for the 101 most active sysops as of today. MBisanz talk 01:50, 8 September 2009 (UTC)

RfB Questions

A month or two ago, a few people who's previous RfA/RfB's have been used as questions at RfB's expressed concerns about their candidacies being used. Essentially, while they didn't say it, they didn't like being reminded of the incident or the incident being highlighted for everybody over and over again (Especially, as one of them explicitly asked to have his RfA blanked.) At the time, I suggested depersonalizing the RfA/RfB's that were being used. With MBiz's permission I have gone through his list of RfA/RfB candidates and made the following changes:

    • Changed the name of the candidate to simply "candidate" with a numeric.
    • Changed the name of the nominator to simply the "Nominator" with a numeric.
    • Changed known nicknames/shorthand to Candidate/Nominator.
    • Changed the month to December.
    • Changed the year to 2099.
    • Removed any indication as to the actual way the RfA/RfB was closed.
    • Removed any closing comments made by the closing 'crat.

While these changes will not prevent somebody who is motivated from figuring out who the candidate is it should be enough to avoid the constant reminders. Furthermore, by doing this, we actually find out how the person thinks. Previously, people could look at the nomination and see how it was closed and simply had to decide if they agreed or disagreed with the closure. Now they actually have to think about it. The examples can be found here.---Balloonman NO! I'm Spartacus! 04:34, 6 September 2009 (UTC)

It's BLP1E applied to RFA! Sounds good. @harej 04:37, 6 September 2009 (UTC)
When do I get to be an RfB question!? :( On a more serious note, I like this idea. Good work. –Juliancolton | Talk 04:41, 6 September 2009 (UTC)
Thank you. MBisanz talk 04:48, 6 September 2009 (UTC)
You're certainly keeping busy in your retirement Balloonman! :) Excellent idea, let's hope people use it. Dean B (talk) 05:47, 6 September 2009 (UTC)
Impressive work. Though it's still very easy to track down who it is, at least the names aren't constantly being used. –xenotalk 05:53, 6 September 2009 (UTC)
Great work. - Dank (push to talk) 12:33, 6 September 2009 (UTC)
That's looking nice. Perhaps you'll come out of official retirement? Useight (talk) 15:38, 6 September 2009 (UTC)
This is probably the longest list of one-indent agreements I've ever seen on this page. Good work! Regards, --—Cyclonenim | Chat  15:56, 6 September 2009 (UTC)
Agreed. :-) Also, good work Balloonman! Hiberniantears (talk) 16:44, 6 September 2009 (UTC)
Indeed, I think this is the first time we've agreed on something related to RfA! –Juliancolton | Talk 16:46, 6 September 2009 (UTC)
This looks good. Thanks for your work, Balloonman. ···日本穣? · 投稿 · Talk to Nihonjoe 17:36, 6 September 2009 (UTC)
Now that you've done this, I think we all think "Why wasn't this done sooner?". Do come out of retirement. --Patar knight - chat/contributions 03:46, 7 September 2009 (UTC)

Nominating yourself

If you want to nominate yourself, do you just add yourself to the list of candidates? And do you need to add yourself to any other lists? --154.20.103.216 (talk) 23:35, 12 September 2009 (UTC)

See instructions here. However, note that only registered users are eligible to gain adminship. –Juliancolton | Talk 23:56, 12 September 2009 (UTC)

Desysop

I am having a difficult time of why it is so hard to de-sop admins, perhaps the illogical programmers need to invent a new code of conduct that has a low degree of sophistication.. South Bay (talk) 00:58, 13 September 2009 (UTC)

The biggest issue is that we cannot reach any agreement on how to get it done, if at all. In the meantime, egregiously bad admins can be sent before the Wikipedia:Arbitration Committee. @harej 01:34, 13 September 2009 (UTC)
Admins are desysoped when they need to be. I have not seen a case where it has been difficult to do so once it has been decided that is the thing to do. Chillum 01:58, 13 September 2009 (UTC)
It's the "once it has been decided" part which is troublesome. The lack of a clear policy which would mandate desysopping for whatever reason means that it always goes to community discussion, which basically means an ANI argument dominated by the same alphas whose self-preservation demands that precedents are not set. There have certainly been far more incidents where an admin has done something which would indicate unsuitability for the mop than have there been desysoppings. Chris Cunningham (not at work) - talk 02:08, 13 September 2009 (UTC)
There is a proposed policy for desysopping at WP:RFDA. Disclosure: I wrote it. → ROUX ₪ 02:57, 13 September 2009 (UTC)
(ec) See also: section above. TL; DR: Some feel that being an admin and doing your job competently is sufficient grounds for desysopping. ;-) As this is the contingent we would rather not see desysopped, we have a bit of a problem which we haven't been able to solve, up 'till now. --Kim Bruning (talk) 02:59, 13 September 2009 (UTC)
Across all wikis, right? O:-) --Kim Bruning (talk) 03:21, 13 September 2009 (UTC)
Can we make this retroactive to the admin's own RfA? Short Brigade Harvester Boris (talk) 03:25, 13 September 2009 (UTC)
I don't see very many sysops that need to be desysoped but are not. Is there any case where the community has made a clear wish that an admin be desysoped and it not happened? Chillum 03:33, 13 September 2009 (UTC)
Define "clear." Short Brigade Harvester Boris (talk) 03:40, 13 September 2009 (UTC)
Here's one definition: "In the eyes of the editors who feel they have been done wrong by the admin in question, it is clear that admin should no longer wield the tools". Can't poke holes in that, can you? ;) Franamax (talk) 04:43, 13 September 2009 (UTC)
Wow, you just used "clear" in your definition of "clear"... That's a bit like dividing by zero. Jafeluv (talk) 16:57, 13 September 2009 (UTC)

I know of only one instance in which an admin had the conditions for recall, written by that admin herself, met, which was not honored. The admin re-wrote the conditions so that the new conditions assure they will never be met, by the simple expedient of characterizing all discussion about said admin's actions as "having a dispute" which disqualified the concerned party from then having their voice count on a recall. In short, if you followed policies and tried to discuss with the admin first, you lost your voice should the admin fail to listen to reason. That admin is virtually retired now, showing every few days or weeks to make a couple of redirects or other innocuous edits, so no harm is being done by her at this time. Other than that, I don't know of any admins who "needed" to lose the bit who did not subsequently lose it. I may disagree with other admins; this is not the same as thinking they abuse the tools. KillerChihuahua?!?Advice 12:36, 13 September 2009 (UTC)

The Aitias model

The problem of the difficulty of de-sysopping and proposals to alleviate it are perennial, and discussions on WT:RFA about it 99% of the time lead nowhere. However, desysoppings have occurred more frequently recently, through ArbCom '09. Most of these have been remedies at the end of long and arduous cases, but in one case, the desysopping of User:Aitias, it was by motion at least in part in reaction to a user conduct RfC which demonstrated a lack of community support for the editor.

I wonder, perhaps if this could serve as a model for future desysoppings – a widely-participated and extensive user conduct RfC on an administrator, with an appeal for a motion by the Committee. Now it's true that the Aitias action had been preceded by a full arbitration case that came close to removing admin privileges, and the argument could well be made that there would be insufficient due process or careful examination in a RfC/U -> ArbCom Motion model. However, I think this would be to overlook to fact that recourse to ArbCom is only to be made when community processes have failed to resolve the issue in question. It is not extravagant to imagine that under this model, editors could work through the issues and reach a definite consensus on whether the administrator still enjoyed community support, thereby giving ArbCom a clear mandate to act. This solution has the potential to erode the vast and unjust democratic deficit that exists currently whereby the community is entrusted to grant administrator rights, but has no voice whatsoever in removing them, and is essentially at the mercy of the administrator caste. Thoughts?  Skomorokh  18:32, 13 September 2009 (UTC)

In the Aitias desysop motion, we had the advantage of a prior case, under which we retained jurisdiction. Having said that, and speaking only for myself, the existence of a very clear response to the RfC certainly made it very easy to support the motion to desysop. This may be a good way to give the Committee the community input that really helps to make such a decision; one of the challenges we have faced in some of the cases revolving around administrator behaviour has been the paucity of community involvement. As I say, I can't speak for the Arbitration Committee as a whole, but seeing such clearcut community opinion was genuinely valuable. Risker (talk) 18:51, 13 September 2009 (UTC)
I think desysopping used to be a problem, but it's not the problem that it used to be, because Risker and the others have shown many times this year that they're willing to desysop when there's community support for it. Go team. - Dank (push to talk) 19:02, 13 September 2009 (UTC)
Not always, for instance WMC's desysop mere hours ago. I believe he still enjoys a wide margin of support from the community, yet he made one very bad block during the case, and plus a few occasions of misjudgments on the evidence page = Wikipedia losing a generally very effective administrator. Even when it's handled without a full case, there's still some room for improvement. For instance, just because Aitias's desysop was handled via motion, that doesn't mean it was a quickly handled situation; it took two user conduct RfC's and one full ArbCom case for them to finally stop and say "OK, enough chances." It was way overdue. But I do approve of ArbCom '09 overall. Master&Expert (Talk) 02:51, 14 September 2009 (UTC)
  • Skomorokh, I had proposed something very similar just a week or so ago. ArbCom didn't manage to pass a desysop motion until they had absolutely no choice; the desysop of Aitias was long overdue. With a consensus like the one on his second RfC, we should practically be allowed to just bring it to a steward to desysop. Because ultimately, the community should make the decision on whether or not somebody is to be an administrator. Master&Expert (Talk) 02:39, 14 September 2009 (UTC)
I certainly sympathise with your position that there ought to be a more direct way for the community to remove administrator access, and that bureaucrats ought to be trusted to judge consensus to desysop as they are to sysop, but that has been proposed and proved less than unequivocally supported in the past. What my point here is that there already exists a trusted means of desysopping that could potentially have a lot more community involvement. ArbCom often lament in considering to accept cases that little community effort has gone into resolving issues. By stepping up to the plate and applying ourselves to just and thorough RfC/Us to put before the Committee, we can achieve the ends desired – abusive administrators disempowered – though the means may still be circuitous. Unlike the majority of desysopping reforms, this is not a proposal that needs consensus to implement; it's already available, and significantly underused. Regards,  Skomorokh  02:53, 14 September 2009 (UTC)
Indeed, this seems like an incredibly obvious method of desysoping. An RfC which either demonstrates whether or not there is consensus for desysopping, which is then presented to the Arbitration Committee, which should make a cursory check to ensure it is appropriate. How is not the simplest solution? \ Backslash Forwardslash / (talk) 14:14, 14 September 2009 (UTC)

Soon (as in, within the next couple of weeks) ArbCom are supposed to come up with a model for Arbitrator recall. Whatever that model is, it should be applicable to administrator recall as well - and it should obviate the need for ArbCom to be involved in routine desysopping. EdChem (talk) 03:17, 15 September 2009 (UTC)

Some incredibly useful info

Currently, 2223 people are watchlisting this page! –Juliancolton | Talk 20:36, 10 September 2009 (UTC)

Two and half times as many as The Beatles. Deserted Cities 20:45, 10 September 2009 (UTC)
But only about half as much as ANI. –Juliancolton | Talk 20:51, 10 September 2009 (UTC)
Maybe RFA and ANI should both be given founder status, on that score they both outrank Jimbo. ϢereSpielChequers 22:55, 10 September 2009 (UTC)
Yes :D --Dylan620 (contribs, logs)help us! 02:17, 11 September 2009 (UTC)
Okay, somebody stopped watching since this was posted. WP:BEANS please! Useight (talk) 23:38, 10 September 2009 (UTC)
That's probably a desirable outcome. Wisdom89 (T / C) 01:50, 11 September 2009 (UTC)
Touché. Useight (talk) 02:23, 11 September 2009 (UTC)

An interesting tool. According to it more people are watching my user page than the Main page? Perhaps because the latter is rarely directly edited. Camaron · Christopher · talk 20:34, 11 September 2009 (UTC)

Try this link for the Main Page. ;) Acalamari 20:38, 11 September 2009 (UTC)
That's better! I had a feeling something was up. Camaron · Christopher · talk 21:30, 11 September 2009 (UTC)
I think this is the best link, though. ceranthor 22:26, 11 September 2009 (UTC)

This is intriguing - I learn that [108 people find my user page of some interest! Warofdreams talk 00:23, 13 September 2009 (UTC)

Wow, I had no idea I was this popular. ;) œ™ 00:56, 13 September 2009 (UTC)

How does this work? I thought that wathclists were invisible on the toolserver. Graeme Bartlett (talk) 00:45, 13 September 2009 (UTC)
And doesn't that circumvent that Special:UnwatchedPages is admin-only? Using that tool one can find out that I have 145 stalkers but one can also find unwatched pages (using a script maybe to try a bunch of article names), can't they? Call me paranoid but I see some use for vandals in that... Regards SoWhy 08:46, 13 September 2009 (UTC)
I was thinking along the same lines, that perhaps the same data source would make a useful list of pages with 0 watchers. Luckily most vandals do not want to put in a lot of work to find those article. And an article that no one looks at will not be nearly so attractive to a vandal as prominent one. Graeme Bartlett (talk) 11:28, 13 September 2009 (UTC)
I'm almost certain the tool uses a MySQL query to find the data. I tried it out myself on the Toolserver's replicated database (NSFW), limiting the results to the first 1000 entries in the table, and it yielded a list of page names and their namespaces. I'm not perfectly sure here, but I'm going to assume that each occurrence of a page name indicates that a user is watching that page (so if a page name appears four times, four people are watching that page), and the tool appears to agree with me. So in theory, you need only Toolserver access to get a list of all unwatched pages, not adminship. Hmm.... The Earwig (Talk | Contribs) 14:22, 13 September 2009 (UTC)
On a completely unrelated note, less people watch the letter 'N' than any other letter! The Earwig (Talk | Contribs) 14:33, 13 September 2009 (UTC)

Due to concerns over security this feature now only produces a numeric result if at least 30 people are watching the page. See User talk:MZMcBride for details. Camaron · Christopher · talk 17:41, 15 September 2009 (UTC)

Success Rate of RfA

I've crunched a few number on the overall success rate of RfAs since 2004. Based on [[2]] and [[3]], there have been a total of 3552 RfAs of which 1676 were successful and 1876 were unsuccessful (there were also admins promoted in 2003, but I could not find readily available data on unsuccessful RfAs in that period). This means that over the course of RfA history, 52.8% of RfAs have been successful.

The surprising fact, though, is that this number has plummeted in the last two years. In 2008, only 33.9% of RfAs were successful and in 2009 so far only 36.5% of RfAs have been successful (2008 was the worst year on record and though there were nearly twice as many RfAs in 2008 (593) as 2004 (302), less admins were promoted in 2008 (201) than 2004 (239). This drop in the promotion rate, is perhaps in large part to blame for our current admin shortfall.

If, in 2008 and 2009, we had promoted administrators at the same rate as the historical average (52.8%), we would today have 151 more administrators. In other words, if we kept promoting at the average level, then our number of admins would probably be about the same as it was in 2007, and we would presumably have shorter backlogs and a lot less moaning and groaning about the lack of administrators. It is also, in my opinion, a reasonable assumption that a higher success rate would encourage more candidates to try an RfA, further driving up our numbers.

What would it take to get back up to 52.8%? The answer is actually fairly simple. There have been 241 RfAs so far this year of which 88 have been successful. At 52.8% success, we would have promoted 127 candidates (39 more). There have been 42 RfAs so far this year that failed with more than 50% support. So, if we simply promoted all candidates where a simple majority of !voters favored promotion, we would have promoted 130 candidates this year (53.9%) and we'd be on track to deal with the plummeting number of administrators. Cool3 (talk) 19:06, 14 September 2009 (UTC)

Out of curiosity, would it be possible to determine how many of the unsuccessful nominations were closed prematurely in accordance with WP:SNOW? –Juliancolton | Talk 19:17, 14 September 2009 (UTC)
But consensus means something, not just at RFA but all over the project, and on probably every WMF project. If you can't get even 70% agreement on anything, there's usually a reason. On top of that, in any approval vote almost anywhere, you're going to have more people willing to support than oppose, it's just politically more savvy to be nice to people than to kick them in the knee, so if only 55% are supporting, that's not a good sign in general. I don't want to rag on people who have failed at 55%, but I think if you go back and look at the RFAs, you'll find significant concerns. Thank the wikigods that we somehow manage to hold on to people who are willing to oppose bravely and convincingly. Rather than make the test easier, I will always favor helping people to be the kind of candidates who can pass the test, and only if that approach is failing would I consider something more drastic. - Dank (push to talk) 19:23, 14 September 2009 (UTC)
I don't actually have the data to back this up, but it is my impression after spending a fair amount of time looking at older RfAs that people were a lot more willing to support back in 2004 than they are today, and that there has been a generally rising level of opposition at RfA ever since then. It is also a well-known fact that standards at RfA have risen exponentially over the last few years (there was a time when 500 edits and 2 months of activity were considered sufficient experience to promote). RfA has over the last two years in particular become much bitterer and we see less people trying and less people getting promoted. A solution exists, which is to simply adjust the bar needed slightly. Sure, when a candidate fails at 55% there are often significant concerns, but a majority of people didn't find them a big enough deal to oppose over, and we simply need more administrators. Cool3 (talk) 19:38, 14 September 2009 (UTC)
I don't want to name names, but I'm not at all comfortable with the idea of promoting the people who got 55% over the last two years. Not at that time, not with that record. I'm on board with the frequent proposal to lower the discretionary range to 65%, but only on condition that we get very specific about what it is that would drop someone into that range but still make them a desirable candidate. For the people who have passed with less than 70%, most recently User:Davemeistermoab, the reason they passed was usually that a large part of the opposition was saying, "Sure, they're qualified in other respects, but there's a particular thing that I think is important that they're not good at," and the closing crat decided that wasn't enough to stop promotion. - Dank (push to talk) 20:01, 14 September 2009 (UTC)
It's no mystery that perceived requirements for becoming an admin have gone up over the years. I say perceived, because the only real requirement is competence, which is subject to opinion. However, the community as a whole doesn't become more competent over time, as editors constantly quit and new ones take their places. Therefore, I think the amount of editors that meets these constantly rising standards (and also applies for adminship, not every competent editor wants to) is steadily declining. I don't think the answer is to lower the discretionary range. That would be countering the community's higher standards, rather than deal with them. I've seen proposals to proactively seek out prospective admins, which I think is a good idea.--Atlan (talk) 20:12, 14 September 2009 (UTC)
Here's an idea for someone with some bot savvy. Perhaps a bot could be created that would run through all users and create a list of users meeting a certain set of criteria that might make them good administrators candidates. Say for example, flag users with at least 3,000 edits, 6 months of experience, and at least 50 edits in the last month. Obviously, we'd still have to check over the list ourselves and nom accordingly, but it might be a good starting point (there is WP:HOPEFUL but some people there only have 200 edits or some such). Cool3 (talk) 00:40, 15 September 2009 (UTC)
(ec) Dan, Not quite. Those were the opposes that were kept. The opposes that were ignored were the ones based from of accusations of falsification and plagiarism, that were later shown to be false accusations. The opposes on the grounds of inexperience, were legit, and were not ignored.Dave (talk) 20:15, 14 September 2009 (UTC)
Imagine a user who clearly states that "I will not work in area X", who then gets 40% oppose because "Does not understand policy of area X" - I'm happy with that editor to get the tools, with the understanding that they are very very careful if they ever do use the tools for area X. How would RfA handle this? Would RfA -or wiki- be ready for editors who say "I'll never work in X, and you can de-sysop me if I ever do"? NotAnIP83:149:66:11 (talk) 20:21, 14 September 2009 (UTC)
That seems an argument for unbundling, maybe? Not every X is unbundleable, but some are. ++Lar: t/c 21:48, 14 September 2009 (UTC)
Occasionally, when I look at the RfA of an admin who's made a particularly poor decision in area X, it turns out that they stated at RfA that "I will not work in area X." In other words, those sorts of campaign promises tend to help more than they hurt at RfA - they assuage concerns. But they suffer from the common flaw of campaign promises - no real mechanism of accountability once the election is over. MastCell Talk 21:54, 14 September 2009 (UTC)
The analogy to Wikipedia:Requests for adminship/ImperatorExercitus 2 is most definitely flawed. Not only were there concerns about copyright knowledge issues, but there were also concerns regarding clarity and effectiveness of communication, plagiarism, and closing AfDs not in accordance to WP:NAC. Those would all be reasons to oppose by themselves. On the other hand, we do have some precedent for limited adminship; perhaps we could expand that further for future cases by adapting the meta temporary adminship model? NW (Talk) 22:56, 14 September 2009 (UTC)
We should definitely incorporate those precedents when we present our plan to the community. As we seem to have a consensus emerging here Wikipedia:Requests for adminship/Reviving the active admin corps has been created so we can keep track of the different strands of our overall reform effort. FeydHuxtable (talk) 12:47, 15 September 2009 (UTC)
What, we have a plan now? The only thing I read on this page, are a bunch of threads showing statistics about admin decline and Balloonman's suggestion to reform admin coaching. Admin coaching is not part the RFA process, despite it being discussed here. No one has presented any feasible plans to reform RFA.--Atlan (talk) 13:37, 15 September 2009 (UTC)
We have the beginnings of a plan, still in the early stages. Balloonman's suggestion, which is a crystallisation of Dank's idea, is actually related to the RFA process. If you read the details it offers an alternative track for editors to be granted the tools, with initially no RFA needed. The concept is that after proving the can use the tools well under mentorship they'll have an easy time passing the RFA which will still be needed to become a regular admin. Also they'll have gained experience and clue thats hard to achieve without having the tools. FeydHuxtable (talk) 13:52, 15 September 2009 (UTC)
One thing to remember in any comparison is that in the early days, the only tools were the ones that admins had. If you wanted to be efficient and were a regular editor, you needed the tools. Over the years, tools like Twinkle and godmode-lite have put some of the features into the hands of regular users. That doesn't account for the declining percentage of approved candidates, of course, but it may account for the fact that the number of admins as a percentage of all users is also not keeping pace. -- DS1953 talk 13:57, 15 September 2009 (UTC)
Retrieved from "https://en.wikipedia.org/enwiki/w/index.php?title=Wikipedia_talk:Requests_for_adminship/Archive_183&oldid=1260843296"