This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.
So with this year closing out at RFA, we can look back now on a year where many records were set;
The first year ever with 5 months of zero successful RFAs.
The first year ever with 6 months of zero successful RFAs.
The lowest ever average number of successful RFAs per month (0.6).
The lowest ever number of RFAs put forth, successful or not, beating the prior record by more than 40%.
The fewest number of successful RFAs over an entire year.
The largest number of consecutive months without a successful RFA.
The longest drought between successful RFAs ever (at 111 days and counting; the prior record was 76 days, set in 2014).
The lowest number of active administrators ever on record at 434, set 26 September 2021 (minus obvious errors in data collection).
The lowest number of edits to this talk page in the last 10 years, and possibly many more (I only checked the last 10 years).
With the oncoming year, we can now look forward to some statistical possibilities;
Based on the average number drop of successful RFAs over the last 10 years, we might have only 2-3 successful RFAs in 2022.
The same average over the last 5 years; we might have only 5-6 successful RFAs in 2022.
It is possible (though unlikely), based on the decline of the number of RFAs put forth over the last ten years that we may see no new nominations in 2022 at all (though as we can see from the section above, it's likely we'll have a new nomination within the next day, so maybe just one?).
Based on the average decline of admins over the last 10 years, we will lose another 42 admins this year (about 9% of admins).
There have been other "The sky is falling!" threads before at WT:RFA. This is really no different. There may be a fair bit of discussion over this thread, perhaps about what to blame, or what new solutions to invent. However, nothing will come of it, as not much of anything has come from any prior related discussions. Given that it's unlikely we'll see any significant change, this time next year we'll quite possibly be discussing marking WP:RFA as {{historical}}. It won't be marked as historical, but a valid discussion will possibly occur.
The decline of RFA is but a symptom, and honestly a very small one, of the much wider problems. The larger problems are tied into the reality that the Wikimedia Foundation has failed as an organization to help foster the future. The projects are failing. There's many examples of rot and decay at the administrator action level. Commons, for example, now has a nine month backlog at deletion requests. Barring some radical change in how leadership is tooled and conducts management at the WMF, we're eventually looking at the end of Wikipedia as we know it. Maryana Iskander has a very complex problem before her, not least of which is the very high level of distrust by the WMF staff of the WMF leadership. If she fails, it's unlikely another CEO will be able to turn this around. I'm sorry to be the bearer of such ill tidings in this holiday season, but it is as it is. --Hammersoft (talk) 10:50, 24 December 2021 (UTC)
Having been barely active this year (and so, in a way, part of the problem) I'm wondering which administrative activities are in trouble on enwiki? BethNaught (talk) 11:41, 24 December 2021 (UTC)
Unless you've been craftily lurking, you haven't been around much at all for the best part of 5 years, Beth. You've missed out on some juicy bits of scandal around the admin cabal, and some shady double-dealing by various iterations of en.Wiki's high court, but apart from that, the classic WP soap continues like a classic soap for people who watch Wikipedia instead of East Enders, and the WMF continues to spend millions on itself like a classic non-profit and nothing on the volunteers whose work pays their big-tech salaries. Na, you ain't missed much. Nothin' worth writing home about. Kudpung กุดผึ้ง (talk) 11:56, 24 December 2021 (UTC)
This isn't going to result in lots of immediate admin backlogs because we have enough existing admins to keep the lights on, and probably will for a while. But appointing 7 admins a year is not sustainable in the long run as those people eventually leave or become less active. Even if the average new admin stays active for 10 years that will result in us having 70 active admins, which is far less than we do at present. The volume of admin work isn't declining, e.g. the number of AfDs has been fairly stable for some time. Hut 8.5 17:14, 24 December 2021 (UTC)
Seems to me the dearth of RFAs may itself have a chilling effect. A single RFA these days finds itself the dedicated focus of attention of a large, and otherwise largely silent, gallery lying in eager anticipation of the next nomination (and that of an obligatory dedicated thread on Wikipediocracy to boot). To my mind, running for RFA these days seems rather more like running for RFB, say, 10-15 years ago. More singled-out, nerve-wracking attention than most would find palatable on account of a hobby, I think. Tyrol5▸[talk] 17:42, 24 December 2021 (UTC)
(edit conflict):I don't think you are the bearer of ill tidings, HS. You're just stating some year-end stats, and thank you for doing so. I won't be surprised if your post doesn't stimulate much in the way of new discussion here either. The lack of interest on this talk page was very clearly described - with just a tad of comic relief - here, 3½ years ago (OMG!). And nothing's changed. Abs-o-lu-tely nothing... Kudpung กุดผึ้ง (talk) 11:42, 24 December 2021 (UTC)
I think the decline in RfA boils down to (harsh) scrutiny. One part of this is even though people say "adminship is not a big deal" out loud, they do think that it is a big deal — for different reasons though. One of them is responsibility, which is understandable. But some editors still think of it as something prestigious/trophy. I also think (not much sure) that scrutiny also comes because of jealousy: "if I cant pass an RfA, I will find something to make this RfA fail as well" I am not sure, but this certainly is a possibility. In short: (harsh) scrutiny → decline in RfA. "If I cant have it, then no-one else can have it" tendency → scrutiny(?) —usernamekiran • sign the guestbook • (talk) 16:54, 24 December 2021 (UTC)
Quite honestly, I'm not sure anyone is taking the "no big deal" angle seriously anymore. What you're describing could perhaps explain a situation where there are plenty of applicants, but a majority of them is rejected for mostly spurious reasons, but that's plainly not where we're at -- out of the three failed RFAs of the current year, only one came after a process that was both lengthy and thoroughly unpleasant for the applicant. I'm also not so sure how to assess the chilling effect the Eostrix debacle might have on this rumoured potential candidate pool, especially given the unusual circumstances. If you've ever misspelled the word as "all be it" and tend use a lot of parentheses, that might be enough to make you reconsider... AngryHarpytalk 17:17, 24 December 2021 (UTC)
Actually, I think you're wrong, all be it I can see your point (which isn't a bad one (although debatable (and did I mention I have no at all interest in Eastern Europe?))) ——Serial 17:30, 24 December 2021 (UTC)
(You joke, SN, but one day you'll be listed in an Icewhiz SPI with this as the evidence.[FBDB]) — Bilorv (talk) 01:03, 25 December 2021 (UTC)
For what it is worth, I think Eostrix will not have a lasting effect on RfA. Almost an entire year of 2021 RfA elapsed before that debacle. But if it does, perhaps it will be old salts or editors who think themselves worse people for falling for an Icewhiz sock not being so quick to support someone. –♠Vami_IV†♠ 17:38, 24 December 2021 (UTC)
I agree with Hammer's overall assessment that our underlying problem is a decrease in editor capacity site wide, of which our decrease in admin is but one problem. I also agree that this problem is likely to get worse. That said comparing the number of people getting RFA to smoe other PERMs, say autopatrol or if you want a high level one template editor, does, in my mind, show that the community consensus that there is a problem with RfA it self to be correct. Even in a situation where we have declining editor capacity we are still minting fewer new admin than we should. Best, Barkeep49 (talk) 20:44, 24 December 2021 (UTC)
Thoroughly agreed. We are being killed like the apocryphal frog in a frying pan as the heat is turned up in every area of the encyclopedia (i.e. work required relative to capacity gradually increases) and also facing additional unique failures at RfA based on an outdated structure that is simply not suited to the post-exponential growth period of Wikipedia. — Bilorv (talk) 01:03, 25 December 2021 (UTC)
I have written a lot about RFA and caused a lot more to be written about it, but so have others including this excellent exposé 9 years ago by Ed Erhart (also a former E-in-C of The Signpost in the days when it was a sturdy weekly publication), and it's actually surprising who are (or were) the leaders of the pack on this talk page, so let's first dispel once and for all this myth made by the Founder's Adminship is no big deal who ironically stated a few years later implying a kind of volte-face with: "RfA is a horrible and broken process". AngryHarpy makes a couple of cogent points, Tyrol5 is not wrong, and I even have to partially agree with Bilorv's metaphor. I'm just waiting for WereSpielChequers to add his opinion in this thread.
Of course adminship is a Big Deal - and usernamekiran is dead right, especially for those who clearly expressed wanting to become a sysop (even to the extent of adding the 'I wanna be an admin' ubox to their page as one of their earliest edits), or put their names on the poll at ORCP, or self-nominated, or like controversial current candidates. It always has been a damned big deal - why else would it have attracted so many hat collectors, obvious power seekers, pompous pre-teens, and people with a paid agenda, not to mention the number of salaried WMF staff who are also en.Wiki admins. Why else would a desysoping be akin to stoning in the stocks on the village green? It's a catharsis for all the non-involved who come to gloat and throw more stones. We don't get better entertainment since the tragicomedy The Life of Brian (but fortunately that was a work of fiction). Instead of moving forward, Wikipedia has descended into a mediaeval quagmire of animal instinct.
The increasingly abnormal phenomena surrounding Wikipedia governance systems (e.g. ANI, Arbcom, etc) are becoming a creeping normality. No attempts to analyze or reform RfA from WP:2011 right through to RfA Review 2021 (now in its 4th month!) despite their focused approach, have addressed the core issues surrounding the attrition of admins with any success and I'll eat my hat if anything changes in my lifetime (other than the entire WMF/Wikipedia thing completely collapsing or the WMF unable to resist billion dollar bribes and selling out to Big-Tech. Which might happen sooner).
I suspect, in hindsight, that all the talk and reform attempts have in fact even had a further chilling effect on the desire of potential candidates of the right calibre to walk red hot ploughshares for a week and if they win, to shiver and quake in the shadow of Damocles' sword for a few years.
Although it wasn't meant to be, I thought this and it's followups might become a seminal work nearly 4 (yes four) years ago, but while they attracted quite a lot of interesting comment, it was, well, just comment. And still nothing's changed. Abs-o-lu-tely nothing... Kudpung กุดผึ้ง (talk) 04:41, 25 December 2021 (UTC)
oh biggus dickus! (sometimes interpreted as oh crap!)
I think no matter what we do, as long as we don't change the mentality of "If I can't, then he can't", and as long as we don't bring the adminship down from the pedestal, I'm afraid not much will change. —usernamekiran • sign the guestbook • (talk) 05:40, 25 December 2021 (UTC)
In my other life as an academic 20 years ago, I studied the declining number of Catholic priests in the U.S. There were a great many research studies done and some found that one major factor for the decline was that current priests no longer encouraged young men to go into the priesthood and without this encouragement and mentorship, many potential recruits never saw it as a possibility for them.
Granted there are huge differences between a religious vocation and being an administrator on a website, but it could be that not only are there fewer current admins but the ones who are here do not encourage potential RFA candidates to go for it at the same rate they did 10 years ago. I know that after my own RfA, which was bruising, I no longer encourage other editors to give it a try. Or, if someone does ask for my advice, I let them know, up front, that an RfA can be a very painful process and they have to be prepared for this. I think now I'd only encourage editors who seem like a "sure thing" because I really wouldn't wish a contentious RfA on friend or foe. It can be a very unpleasant & turbulent 7 days. LizRead!Talk! 06:27, 25 December 2021 (UTC)
In March I made several comments about RfA at WP:DESYSOP2021, about how creating a desysop process would hugely decrease our number of admins. I remember some "There's no issue" comments, about how the number of active admins had stayed in the 490-540 range for three years. Indeed, in April, the active administrator number dwindled to the 480s. Now it is 465. And how do we reward that decline? By stripping the admin rights from three (or more, depending on which month it is) underactive admins at WP:INACTIVE. To combat this, we… hold an RfA that will not pass. It's a snowball's Power~enwiki's chance in hell well, Wikipedia's hell. Unless a) Power~enwiki's RfA turns out OK (which won't happen, with a 66% only a day into it) or b) there are three successful RFAs in the next month (which is less likely to happen than Power~enwiki's RfA passing) {{NUMBEROFADMINS}} will display 1,062 to start the new year off. And we just closed an RFA reform discussion. What did we do? Revise a question, un-unbundle a right, and create a noticeboard for overturning administrative actions. Would these changes to reform the horrible and broken process help us encourage more people to run, or to retain admins? And how about the discussions here? They never turn out to anything. Look through the archives if you don't believe me. "Nobody wants to run", "Low activity at RfA", "Constantly losing admins", "Create larger candidate pool", etc. 🐔dat (talk) 12:42, 25 December 2021 (UTC)
What would be more interesting would to do an analysis on the # of admin actions a year and the number of active administrators doing 80% of those actions. I'm willing to bet the 2nd number has stayed relatively stable in gross numbers over time. Thanks to advances in technology, automation, and unbundling, we have less of a need for new admins than we did 15 years ago, or even 5 years ago.The 465/500 active admin number is being used a lot, but has that actually had a noticeable impact on the project? For some reason, I don't think so. I know it's heresy on this page, but RfA isn't broken nor is Wikipedia anywhere near a breaking point with the need for administrators. I think creating new ones is good for reasons other than need, but we're doing fine, and the non-crisis we've been having at RfA for ~15 years now is just that: not a crisis. Crises don't last 15 years. TonyBallioni (talk) 13:15, 25 December 2021 (UTC)
Indeed. The reality is that most "active" admins don't do much admin work. Here are the admin stats for the past 3 months. This table doesn't include items such as closing discussions or Main page tasks, but I still think it provides good insight into the current situation. We've got maybe 100 admins doing 95% of the work. -FASTILY 00:24, 26 December 2021 (UTC)
Numerically speaking nearly all deletion is done by three or four admins; I look at the queues multiple times a day when I'm active but they are nearly always almost empty, so I don't delete very much and often find myself edit-conflicting when I try to. I don't think this is a healthy situation, personally. Espresso Addict (talk) 00:38, 26 December 2021 (UTC)
Fastily, thank you for that! What a useful tool that exports to excel! Some actual numbers now:
For the 92 days ending 25 December 2021:
Total Non-Bot actions: 246,915
Total non-bot admins with an admin action: 550
106 admins represent 95.05% of admin actions
33 admins represent 80.34% of admin actions
This of course doesn't represent all that admins do; many are extremely visible behind the scenes, in dispute resolution, etc. It does show the point I was trying to make, though: we are not anywhere near a point where the decrease in active administrators represents a threat to the functioning of the project. In my opinion there are reasons besides this to want more people to go through RfA, but we are nowhere near a crisis. TonyBallioni (talk) 00:45, 26 December 2021 (UTC)
The point that's often made is that we are hugely dependent on a surprisingly small number of admins. (33 is not the whole answer as it fails to capture those active on, say, the main page.) If several of these decide to give up admin work, get de-adminned, retire from the project, or even die, then the project is at risk, and as so few new admins are being trained, there are few people to take up the slack.
I don't personally know how big the risk is; for example RHaworth's forcible retirement only had a temporary effect on the deletion queues. Espresso Addict (talk) 01:28, 26 December 2021 (UTC)
'
2021
2020
2019
2015
Total Non-Bot actions
247,616
213,066
243,335
231,568
# of admins to reach 95th percentile
107
130
127
161
# of admins to reach 80th percentile
33
49
46
65
Total # of non-bot admins with actions
551
553
541
623
Because I'm bored after visiting family for a week, here you go. A comparison of admin actions for the 92 days ending 25 Dcember 20XX over time. Last 3 years and 2015. The 2021 number went up slightly because I accidentally excluded NinjaRobotPirate in my first responses as a bot in the excel filter.Number of admin actions are up from 2015, but the number of admins comprising the 95th and 80th percentiles is sharply down from a year ago and from 2015. What are we to make of this? Probably a lot of factors, but if I had to bet, we've creates a lot of tools to automate things in the last 6 years, which probably accounts for more "productivity", and the other thing is that when there's a gap in people doing the work, someone steps in to do it. RHaworth is probably the best example of this. We're able to do more admin work with less admins than we previously did. Like I said, I'm fully aware that not everything is captured by logged actions, but if we're looking at productivity numbers, this December 25th had the most non-bot logged actions in the last 3 months of the 4 years assessed, and it also had the least number of admins needed to account for the 80th and 95th percentiles, all while the number of admins with a logged action remained relatively stable for the last 3 years.Again, you can have a discussion about whether or not this is healthy on an individual level, but on the easiest to measure metric, our administrators are doing more than they ever have, and it is taking less of them to do more work. That's not really crisis mode for backlogs. TonyBallioni (talk) 02:17, 26 December 2021 (UTC)
I've agreed with you in part on this, I think I've said the same before, but it's mainly only true on the large scale. It's also heavily biased towards routine housekeeping actions like deletion of stale drafts, and ignores a few single actions which some may argue have a greater impact on the project (personally, I'm currently conflicted on how to assign 'value' to an individual admin action). This methodology doesn't account for things on the small scale.
For example, at Wikipedia_talk:Today's_featured_article#TFA_vandalism we were having trouble finding an admin who wants to run a bot to implement the RfC consensus that TFA should be semi-protected. Only a handful of admins can do this, as the criteria is: you need to be an admin, you need to be willing to run an adminbot, and you need to not be ideologically opposed to the concept and/or willing to put that aside. A mix of this criteria narrows the pool to < 5 people, realistically, some of which are semi-inactive or otherwise don't want to do it. This is a high-impact change, though, but would not be measured in your calculations. The admins who are capable of doing it (Anomie, Legoktm, MusikAnimal, etc) are probably not individually 'highly active' by your methodology.
All I'm trying to say is that the flaws of the methodology have to be noted. Overall I'm not so sure anymore that it's a reliable way of looking at the issue. Frankly I'm not sure there is an objective way to measure the question. But the consensus of the community's feelings, per WP:RFA2021 Phase 1, seems to be that there is a problem. ProcrastinatingReader (talk) 02:47, 26 December 2021 (UTC)
Most of the work admins do is maintenance work, though. Yes, there's those of us who are pretty active in the project-space/discussion sphere, but even then that typically corresponds to logged actions to some degrees. On your other points: you're going to have problems with people filling niche rolls regardless of how many we have. The number of admin bots performing admin actions has actually gone up since 2015, by the way, so that example seems like not the best one to use.Yes, any objective methodology is going to be limited since no measure can accurately capture everything admins do. At the same time, logged actions have been the universal proxy for admin activity on Wikimedia since the dawn of time. I think it is probably fair to say that non-logged admin activity correlates with them to some degree for everyone but arbs.I suppose my point is this, if people want to make an ideological point that RfA is broken, that is one thing. I might be sympathetic to it depending on how it is phrased. But if people are trying to make an argument from utility, they need to be prepared to address the strongest counter-argument: the reason there's less people running for RfA and less admins making actions is because there's less of a need than there was in 2015 or 2005. That wouldn't mean that we shouldn't be creating more admins for culture/principle reasons, but it would mean that the problem is a lot less about format and more about the evolution of the project. TonyBallioni (talk) 03:27, 26 December 2021 (UTC)
Are there many non-logged admin actions that are being done more often by bots? I know at DYK the move-to-queue used to be more complicated and now is bot-assisted so that only the actual stuff that needs to be done by a human is left to a human. —valereee (talk) 12:04, 27 December 2021 (UTC)
It is sad that the numbers are this low. I feel the only two things that will have an actual impact is making the process of going through RfA less imposing and improving editor and especially newcomer retention so we eventually get more potential admins. Both are difficult problems so instead I will just share the best pieces of wisdom I heard and learned when I went through the process this spring.
No one is good at assessing how people will look at their own RfA. You can rationalize your poor decisions and become overconfident and you know all mistakes you ever made and can convince yourself you will never make it. Therefore it's good to talk with other people. I would recommend admins you trust that work in areas similar to you. Most admins (me included) are happy to talk adminship and would love to give you their advice if you want it. It may be an awkward first message, but please reach out to someone if you are even remotely considering adminship.
There is a ton of uncertainty about how the RfA will go and feel. There are horror stories and there are RfAs that went incredibly smoothly. You won't know how it will go for you, but if a few admins you trust have confidence in you it's more likely than not you will get a support percentage in the 90s from a purely statistical perspective.
Questions can make or break an RfA. Consider your answer and take no shame in not answering immediately and giving the question a second thought later in the day. Thoughtful and informative contributions is one of the main things many people look for in discussions. There are times when the best course is not to act, to ask for advice or delegate the task to someone else. You are not expected to be a master at everything, just to have a good judgement when you aren't.
I have no idea if that's any help, but my inbox is open if you wish to talk. Happy holidays! --Trialpears (talk) 13:46, 25 December 2021 (UTC)
Ten years ago when WP:2011 was started, a group of over 40 not-quite-so-stupid editors genuinely had concerns that the environment at RfA would soon leave the project with not enough admins. Indeed, RfA are far more rare than they used to be and with most of them passing with very high scores, one could even claim that the venue is somewhat less toxic. However, nowadays with very few borderline or potential no-hope RfA being transcluded, little opportunity is left for the nasty voters to play their games. Their toy 'hate' venue has changed its spots and they've either retired, been blocked, or backed away to their caves. In 2017 the trend underwent a complete reversal and instead of the majority of RfA resulting in failure, in the last 3 years more than twice as many pass, but I will still make an ideological point that RfA is broken.
Nevertheless, things can change, even if excruciatingly slowly. Ten years down the line, the sky still hasn't fallen, admin backlogs, especially ANI, aren't any worse, and even NPP is keeping its backlog down to 'only' 10,000 articles. The series of 5-yearly reform projects have not brought about the changes they sought after, and because (according to one well informed user) the WMF will never allow a secret poll for RfA on technical grounds, en.Wiki is probably going to be stuck with the status quo. In another 5 years or so there will be a huge drop through natural causes in the number of pre-2012 admins leaving even fewer to do the work, but new scripts, filters, and other features (such as perhaps a ban on IP editing) will have arrived that will cope with that too. Consensus can change, and often does. And so can my opinion. Kudpung กุดผึ้ง (talk) 12:27, 26 December 2021 (UTC)
There is certainly a long-term issue. The main short-term risk I see is that the rate of successful RFAs is getting dangerously close to the rate of new ARBCOM members - we should actively try to avoid a situation where voters view a successful RFA as a guaranteed ticket to ARBCOM and oppose candidates on that basis. User:力 (powera, π, ν) 19:43, 26 December 2021 (UTC)
Arbcom could be the next, high priority Wikipedia governance feature to be the target of a huge reform programme, even including proposals for it to be replaced by something completely different. While WP:RFA2011 remained largely a huge research project because its further activities were disrupted by the very toxic elements that it was investigating, subsequent reform projects failed due to their complexity, sidetracking, and lack of initial up-to-date research. The most recent endeavour, though planned and launched with the very best intentions and largely following the model of the 2015 exercise, like its predecessor it almost got stuck on its many proposals. Continuing for nearly 4 months with still some ends to be tied up, overall participation was low considering its importance, and finally, not one of its primary intended goals concluded with consensus for possible RfA solutions. Perhaps if the people who initiated or who contributed heavily to the 7 major reform projects since 2007 (including Scottywong. WereSpielChequers, Biblioworm, Carcharoth, The Blade of the Northern Lights, Ched, Dank, Camembert, Tim Starling, TonyBallioni, Worm That Turned, Barkeep49, and others) could get together and ask ourselves "Where did we go wrong with our RfC?", some of the learnings could come up with ideal solutions for addressing urgently needed restructuring of Wikipedia governance. But it needs to be done while some users with the required institutional memory are still around. Wikipedia is now 20 years old, meaning that most people under the age of 30 don't know a world without it but those of us who were 60 or more when we registered 15 or 16 years ago aren't going to be around forever; enthusiastic newcomers to these discussions probably don't read up on the sinuous history of policies and practices. Anyone wanting a good potted history of the RfA reforms could read the series of articles in The Signpost .
Good to see a lot of familiar faces here. Question for all: would RfC voters be willing to settle for some kind of change that's safer but slower, or will voters reject any proposed solution that doesn't have at least a shot at solving most of the perceived problems quickly? What I have in mind for "safer but slower" is: change the rules so that your first RfA is more or less what we've come to expect, but if that fails, then six months down the line, do something experimental with the 2nd RfA. The obvious disadvantage is: this changes nothing right away, or even within a year, probably, so it's not going to generate any great enthusiasm. But it avoids the many problems inherent in changing everything all at once (including: there will probably be low community confidence in candidates promoted through any new process until the new process has been thoroughly tweaked and vetted, which could take years). - Dank (push to talk) 16:37, 2 January 2022 (UTC)
Good to see you around, Dank. A shortcoming I'm seeing with this particular suggestion is that there really aren't very many initial RFAs at all, let alone second RFAs, these days. Even less than there were in 2011, when many of the folks here (myself included) speculated about whether the yield of new sysops had "bottomed-out." Such an experimental process to be deployed in subsequent RFAs could therefore take a very long time indeed to iterate into a more refined one unless, of course, it would encourage more candidates to step forward, thereby increasing the volume of first, and second etc., RFAs (which I am dubious about, if the initial process will not change). Secondarily, I think there could be concern that an alternative process for subsequent RFAs would be perceived as moving the bar (in either direction) for candidates who initially did not succeed at RFA. That being said, and to your point about gradual, deliberate change, I am partial to running an alternative process in conjunction with the present process, such as annual or twice-annual "elections" (see my comment above for my view on RFA anxiety). Tyrol5▸[talk] 20:19, 2 January 2022 (UTC)
The OP has not even shown all of the failure of RfA. RfA has two purposes: to promote new admins, and to keep unsuitable people out. For many years, it has failed to do the first but done a decent job on the second purpose (and for some of the people who were desysopped later, the issues were raised by opposers in their RfAs). The Eostrix debacle has shown that the RfA crowd (myself included) can nevertheless still be fooled by a clever player. Bottom line, we have made the process extraordinarily unpleasant for good people without making it impossible to succeed for bad actors. If we try just a little harder, maybe we can get to the point where only Icewhiz socks can get through RfA. —Kusma (talk) 23:08, 2 January 2022 (UTC)
I would argue that RfA does not do a good job on the second purpose. Of the 10 most recent admins forcibly desysopped by ArbCom (that I could find RfAs for anyway), 6 had 90% or higher passes. None fell into the 70-79% range. I.e., clear passes; none approached the discretionary range. I would modify the "second purpose" to say "clearly unsuitable". For that, I think it might be ok. --Hammersoft (talk) 19:35, 5 January 2022 (UTC)
RFA 2021 Completed
The 2021 re-examination of RFA has been completed. 23 (plus 2 variants) ideas were proposed. Over 200 editors participated in this final phase. Three changes gained consensus and two proposals were identified by the closers as having the potential to gain consensus with some further discussion and iteration. Thanks to all who helped to close the discussion, and in particular Primefac, Lee Vilenski, and Ymblanter for closing the most difficult conversations and for TonyBallioni for closing the review of one of the closes.
The following proposals gained consensus and have all been implemented:
Revision of standard question 1 to Why are you interested in becoming an administrator? Special thanks to xaosflux for help with implementation.
A new process, Administrative Action Review (XRV) designed to review if an editor's specific use of an advanced permission, including the admin tools, is consistent with policy in a process similar to that of deletion review and move review. Thanks to all the editors who contributed (and are continuing to contribute) to the discussion of how to implement this proposal.
Removal of autopatrol from the administrator's toolkit. Special thanks to Wugapodes and Seddon for their help with implementation.
The following proposals were identified by the closers as having the potential to gain consensus with some further discussion and iteration:
Editors who wish to discuss these ideas or other ideas on how to try to address any of the six issues identified during phase 1 for which no proposal gained are encouraged to do so at RFA's talk page or an appropriate village pump.
A final and huge thanks all those who participated in this effort to improve our RFA process over the last 4 months.
This is the final update with no further talk page messages planned.
01:47, 30 December 2021 (UTC)
don't mention it MediaWiki message delivery, it was my pleasure. I just feel bad that I didn't participated/contributed much. —usernamekiran • sign the guestbook • (talk) 19:01, 30 December 2021 (UTC)
I'm disappointed, but not surprised that few substantial changes came out of this lengthy process. It feels a bit like re-arranging deck chairs on the Titanic. I hope the temporary adminship and optional election process ideas get more discussion. Nosferattus (talk) 21:01, 30 December 2021 (UTC)
The fundamental RFA process itself has changed very little over the decades; you put your name up, any editor interested enough comments and then if pretty much everybody agrees you become an administrator.
What *has* changed is that "any editor interested" has been replaced by an "RFA community" - people who count "being a member of the group who select admins" as "one of the things they do". It seems optimistic to hope that these exact people will solve the problem of their own presence. 2A01:E34:EF5E:4640:3D3E:D342:C9D8:6A8 (talk) 12:45, 1 January 2022 (UTC)
Has someone done another analysis on RFA voting to see what proportion of the !votes in 2021 RFAs were cast by people for whom this is a significant part of their activity? In the 2005-2009 era there were almost always multiple RFAs running, and it was possible that we had a number of peoople for whome RFA was indeed "one of the things they do". Last year we only had seven successful RFAs and most of the time there weren't any RFAs running. One of the successful RFA reforms was advertising RFAS on watchlists, so the typical recent RFA gets far more participation than the ones in our first decade. So my assumption is that we are closer to "any editor interested" now than we used to be. ϢereSpielChequers 21:41, 2 January 2022 (UTC)
In the old "non toxic" days, people expressed an opinion when they had reason to have one, more often than not based on actual interaction or having "seen them around". Failing that, the default - quite reasonable - position was to simply not comment. These days, "members of the RFA community" express their opinion on virtually *all* RFAs statistically speaking baed on - at best - the answers to the standard questions or, just as likely, piling on the prevailing tendancy (this is a supposition on my part, but it might be interesting to learn what proportion systematically comment all RFAs).
I'm sure there are people who diligently go over edit histories looking for good, bad or indifferent and basing their opinion on that but, at the end of the day, since the process really hasn't changed it seems reasonable enough to start wondering if the problem isn't the people involved. 2A01:E34:EF5E:4640:3D3E:D342:C9D8:6A8 (talk) 23:46, 2 January 2022 (UTC)
Yes, people are a problem. One of the things I really don't get is why people who say "I'm sorry but I can't support" then go and oppose instead of simply not supporting. If you want to apologise for opposing, you should say something like "I'm sorry but I need to try to prevent you from becoming a sysop". As to "diligently going over edit histories", that hasn't been half as much fun since the maximum number of entries displayed changed from 5000 to 500. —Kusma (talk) 00:56, 3 January 2022 (UTC)
I didn't know this either until recently, but the limit was quietly removed in June 2020. Graham87 06:00, 3 January 2022 (UTC)
@2A01:E34:EF5E:4640:3D3E:D342:C9D8:6A8: That's actually not as true as you might think. A couple of years back (well into our modern rare-RfA era) when I ran, someone did a review on myself and a couple of other newly-made admins in the form of common or different (!)voters. And there was a huge variance - only about a fifth had (!)voted in all of them, and there were large numbers of individuals who had only participated in one. For mine, for example, I had lots of AfD regulars, because that was my primary field at the time. While I'd love to have a broader data set, perhaps reviewing every RfA (both successful and failed) from 2021, it's not as clear as you thought (until that point, I shared the position) Nosebagbear (talk) 18:55, 3 January 2022 (UTC)
The problem with RfA is there are no second chances. For example, I weakly supported an RfA candidate last year, because I felt they didn't quite have the empathy skills and hoped, given the weight of support, that they could learn on the job. Looking at their contributions, I can't see so much as a typo fix or copyedit in the last three weeks, leading me to suspect my advice was not taken. Unfortunately I can't go back and add further comments; we do have WP:XRV now to at least review some cases. Ritchie333(talk)(cont) 14:49, 4 January 2022 (UTC)
I think there are second chances - after a waiting period of 18+ months which admittedly a high cost to that second chance. I don't know that there are third chances though. Best, Barkeep49 (talk) 19:13, 4 January 2022 (UTC)
Nosebagbear, that was me; here are the stats on that cohort for those interested. The topline result is that across 3 RFAs that attracted 180-219 !votes each, there were a total of 356 unique !voters. MarginalCost (talk) 05:48, 26 January 2022 (UTC)
"Voice your opinion on this candidate" link
Sorry if I missed something (surely I have), but at the current RfA, the blue link "Voice your opinion on this candidate" links to an edit window for the "Questions for the candidate" section. Is this what's intended? JBchrchtalk 18:08, 25 January 2022 (UTC)
My fault. I added the co-nom as a section which changes the section numbers which causes that issue. It is now fixed by changing it into a pseudo header. --Trialpears (talk) 18:11, 25 January 2022 (UTC)
User:Cyberpower678/RfX Report, linked at the top of the page, is not being automatically updated. It appears that it is usually updated by User:cyberbot I, but this is way out of my wheelhouse. I suspect users of that template are not being kept up to date. The template immediately above does seem to be working though. SchreiberBike | ⌨ 02:30, 28 January 2022 (UTC)
I've replaced that page with a transclusion of the template-generated report which appears on this page. If the bot ever comes back up, then it will revert me, but in the mean time the report will be updated without manual edits. * Pppery *it has begun... 17:19, 28 January 2022 (UTC)
A point came up in one of the recent RfAs that we often tell unsuccessful candidates who haven't quite got the hang of adminship (in the view of those opposing) to "come back in 6/12/18 months". It was also pointed out that, at least in the last five years, unsuccessful candidates are extremely unlikely to attempt or pass RfA. The only example I can find is Wikipedia:Requests for adminship/Cwmhiraeth 2, which occurred about three years after the first one.
I've had a dig through unsuccessful candidates from 2017 - 2020. In almost all cases, I think I can put them into the following piles:
Obviously failed RfA, no chance of success (in a couple of cases, the candidate is now indefinitely blocked)
Issues at RfA haven't been satisfactorily addressed, meaning a second run would probably be the same
Candidate has gone inactive and wouldn't pass RfA because of that
RfA failed because of controversy which can't obviously be overcome in any timeframe, leading to subsequent ones being something of a crap shoot
The RfA was so stressful, it's put the editor off running ever again
I haven't further refined these into RfAs where a key opposition was WP:NOTQUITEYET; maybe I'll do that at some point.
Given the lack of evidence, can we stop saying "come back in 'x' months" as it doesn't appear to be factually correct any more? Ritchie333(talk)(cont) 14:37, 4 February 2022 (UTC)
Lourdes passed about 12 months after her first attempt and both of those are within your study period. I think providing a "here's what I'd like to see in order to support you in 6/12/18 months" is helpful. RfA oppose is a rejection of the editor in many ways but that offers a positive way forward rather than "the Wikipedia community hates you" which can be a message some receive even from successful RfAs (admittedly I've exaggerated but only slightly). Best, Barkeep49 (talk) 15:10, 4 February 2022 (UTC)
Lourdes was 4 years ago and AFAICT the only 1-year successful re-RfA since. A very sloppy look (I may have miscounted in one direction or the other on months, that is: if one RfA was in July and the other in January, it's quite possible I counted wrong in one direction or the other, I was just doing this to get a rough idea):
2121
BusterD 2 was 10 years after
2020
Jackmcbarn 2 (who btw made zero edits in the past year+) was after a desysop for inactivity, so doesn't really count as success after failure
LuK3 2 was 8 years after
Red Phoenix 2 was 12 years after
Cwmhiraeth 2 was 3 years after
2019
Floquenbeam, also doesn't really count
Enterprisey 2 was 4 years after
2018
Lourdes 2 was 1 year.
2017
Anarchyte 2 was 1 year
Generalizations are bad was a bit over a year
Cipo13 was nearly a decade
Dodger67 was 2.5 years
Primefac 2 was 1.5 years
I don't think anything before 2017 really is pertinent. But yeah, anything under a year is basically not happening currently, and really 3+ years is currently the minimum and ~a decade currently more common than anything else for a successful re-RfA. Certainly the idea that 3/6 months is realistic has no evidence to support. valereee (talk) 21:43, 4 February 2022 (UTC)
This doesn't tell us, though, that editors cannot improve within, say, six to twelve months such that they can receive consensus support for administrative privileges. To examine this question, we need to look at those who failed (as Ritchie333 did), find those who took steps to improve their shortcomings, and then see why they chose not to make another request, or why they did and it failed again. Additionally, statements on how a commenter might respond to a future request should be interpreted as a personal view, and not a prediction of the future end result. As such, the historical track record is less critical than the specific case of the requester. isaacl (talk) 22:05, 4 February 2022 (UTC)
It tells us that within the past five years no one has. Could they? Sure. Let's circle back when someone does. :) valereee (talk) 22:07, 4 February 2022 (UTC)
As others have pointed out, so few have received administrative privileges at all in the last five years that it's hard to draw definitive conclusions. If no one has even tried to make a second request within a year of the first, then I can't see how any conclusion can be drawn on this specific matter. All the same, I do think commenters should avoid being overly optimistic when making predictions about how they might comment on a future request, and take into consideration the potential effect of the request failing again. isaacl (talk) 22:14, 4 February 2022 (UTC)
Yes, of course, the sample size is too small. But the fact that in the past five years more people have had successful 2nd RfAs after eight years than after one and only one person since 2018 has succeeded after one and that since 2019 no one has successfully re-RfA'd after one tells me we shouldn't be suggesting one year is realistic. Notable: these are not actually sample sizes. They are population sizes. We are actually looking at all incidents. Does that mean it's not possible? No. But it does mean it's not happening. It certainly provides no evidence that a second RfA after under a year is even a possibility for success. It would be good if someone would analyze failed second RfAs and their timing. valereee (talk) 22:30, 4 February 2022 (UTC)
I think we're talking about different things. I agree we shouldn't tell editors that others have come back within a year (in say the last five years) and been approved, but I can't remember anyone saying that recently. The comments I recall are providing encouragement about the potential support the commenter may provide. isaacl (talk) 22:39, 4 February 2022 (UTC)
It used to be that many' editors actually did that, and lots passed on goes 2 or 3. Frankly there have been so few "unsuccessful candidates from 2017 - 2020" I'm not sure such an analysis is much use. It's still good advice, & that is worth giving even if people choose not to take it. Johnbod (talk) 15:28, 4 February 2022 (UTC)
(edit conflict)A failed RfA can be seen by the subject as a rejection of ability. This isn't helped when the comments can be cutting. It's not surprising that people don't want to do a second one, and arguably it's surprising that after failing someone would want to continue to edit. The comments of "not yet, would support after more tenure/experience/shown skills" are generally saying that you would have a good chance if you decide to run again. In theory, this makes sense. In practice, it's usually now or never if you've had the drive to run. Best Wishes, Lee Vilenski(talk • contribs) 15:30, 4 February 2022 (UTC)
Personally I've not the slightest interest in whether someone's got enough "tenure/experience/shown skills". The question is whether they've had enough activity in areas where there is often conflict, so that we can judge their real character. Do they listen to other points of view and take note, or are they pompous know-it-alls? That's all I want to know. Nigej (talk) 15:59, 4 February 2022 (UTC)
The subset of people who run a second (or third time) after an initial run has always been smaller than the group who run a first time. And we're at a moment where few decide to run, so of course only a very small number run a second (or more) time. I think it would be healthy for our community if RfA felt a little lower stakes for everyone. While at the moment it's largely, though not completely true, that it's now or never it was also true you needed a GA/GL/FA if you were running as a content editor (and maybe even at all). Until it wasn't as we've now had a couple (in a small sample size) run that way. So I'd rather see us try to create a healthier atmosphere at RfA and better outcomes than dismissing the theory based on a small sample set of current practice. Best, Barkeep49 (talk) 16:03, 4 February 2022 (UTC)
Oppose. (Geddit?) On two fronts really. Firstly, the number of months a candidate is advised to wait is a subjective assessment made by a review from their own particular perspective, and as long as it doesn't fly in the face of policy, it would be pretty incollegiate to police editors' good-faith opinions; if anything, it would only add to the acrimony that is RfA today. (I think that's something that's generally underestimated: that many of the things we discuss with the intention of making things easier for the candidate may actually make the experience a worse one the more you restrict reviewers' freedom of movement.)The second point is more procedural in nature; I don't see how it can be said in one sentence "I haven't collected all the evidence as to whether X is true or not" and then immediately follow that with "Can we stop doing X because we haven't any evidence it's true". SN54129 16:20, 4 February 2022 (UTC)
Two of those categories have reverse survivorship bias: the base assumption is to try again after you've remained active and worked on resolving concerns. I understand that failing to be approved for administrative privileges may discourage editors from remaining active. We should be trying to encourage candidates who are resilient towards the criticism they will face. On the flip side, we should be doing more to encourage critics to be as graceful as possible in their comments. I know it's very difficult: providing feedback and being receptive to it are some of the hardest parts of life. isaacl (talk) 16:23, 4 February 2022 (UTC)
Telling a community not to say something is practically unenforceable. It's personal advice. Leaky caldron (talk) 16:30, 4 February 2022 (UTC)
Few days late on this, but... I agree with the basic idea that if your main reason to oppose a candidate is NQY, you shouldn't oppose, because they're probably not going to run again when it's QY. RfAs with serious opposition where NQY is the sole rationale are almost unheard of, though, and if we are to generalize from that 'almost' then it seems unlikely such an RfA would even fail going by the crat chat weighings. (Unless the candidate withdraws.) The tricky part is that in all the other RfAs with a mix of NQY and 'other' opposition, the 'other' opposition can be vicious enough that running again is the last thing on someone's mind. To note extreme cases, if an RfA is hostile enough to cause people significant mental harm, the presence of "try again in six months!" opposition comes across as sadistic. Fairly few (not no, but few) RfAs are botched enough to endanger people's real-world health, let alone that of people who weren't even the main victim, so this is an uncommon extreme, but it's an extreme that happens. In the realm of more ordinary mixed-opposition RfAs, the issues remain; mixing "hey, you should do this again!" with attacks on one's fundamental character doesn't tend to work out very well. I assume anyone who's a prolific RfA reader has seen the "everyone should do this every year" exceptions, but those are exceptions. Vaticidalprophet 08:35, 7 February 2022 (UTC)
Something that no one talks about but that I find weird and off-putting about RfA is that an editor with like 6mo experience can tell an editor with two years experience that they're not ready to be an admin, even that it's "not quite yet". How would a new editor know how much is enough experience? I wonder if editors should only be allowed to !vote in an RfA if they have more experience than the candidate. Same with a !voter who's under a sanction or been repeatedly sanctioned in the past--why is their opinion about a candidate given the same weight as editors who have a clean record? It doesn't make sense to me that editors are evaluated by other editors who are, in some cases, inexperienced or under sanctions, and those votes are given equal weight. Levivich 13:54, 7 February 2022 (UTC)
Because it's an advertised public 7-day combined performance review/job interview and your worst work enemy and the intern who started yesterday are both allowed to ask up to two questions. Duh. valereee (talk) 14:12, 7 February 2022 (UTC)
I have the impression (have read every RfA from 2017 on, every "complicated" (significant support for a fail or significant opposition for a pass) RfA from 2015 on, some earlier, every cratchat, pretty much any interesting one several times over) the most...unusual...tenure-opposes don't tend to be from newer editors. If anything, newer editors who take an interest in RfA tend to have lower tenure standards, on account of such editors generally want to be admins at some point and would rather be able to pass sooner rather than later. (There are regular RfA participants with no interest in adminship, but they don't overlap with the ones who are too new to pass RfA.) Generally, inflated tenure standards come from experienced editors who have no gauge of how long two-three-etc years is anymore, or have low opinions of themselves that they generalize to everyone else (I've seen "I'll never support anyone less experienced than me" opposition, which at the time it was stated meant about five years), or similar forms of detachment from what it means to have edited Wikipedia for a year. None of the really idiosyncratic tenure opposes I've seen have been from people who were around for shorter than the candidate, or shorter than the standard they were enforcing. Vaticidalprophet 14:58, 7 February 2022 (UTC)
I also have the feeling that highly experienced non-admins are a big part of the increase in standards. It is only human to compare the candidate to oneself and one's own suitability; this causes me to support candidates who seem to be more experienced than I was when I became an admin (6 months of real contributions, 6500 edits) and may cause other people to oppose candidates who clearly are less experienced than themselves, even if two years and 10000 edits and 100 AfDs is actually a lot of experience. —Kusma (talk) 15:45, 7 February 2022 (UTC)
My two RFAs were only a few months apart and both over a decade ago, but I may be the first person to join this discussion whose first RFA failed but who passed with little opposition at my second attempt. Opposes based on skills I hadn't yet demonstrated, and especially opposes that said come back in x months if you fix this were much easier to handle, and ultimately much more useful, than other opposes. ϢereSpielChequers 20:56, 8 February 2022 (UTC)
Yeah, it seems a little wide. I think we can get rid of a few of the columns. "Status" seems like it could be removed (as it wasn't in the Cyberpower report). Honestly I think "Ending" can also be removed because it's redundant to the "Time left" column, but that's arguable. Also, I think "Dups" can be removed unless there's a "Yes" answer. Pinging the most recent editors of Module:RFX report: Firefly and Mr. Stradivarius. Enterprisey (talk!) 23:06, 26 January 2022 (UTC)
If we're going to remove one of the two columns, I'd suggest removing "Time left", as the current time is always changing. I appreciate, though that it can be helpful for those who might forget to adjust for the time difference with UTC. If we go back to my suggestion of having a legend (displayed only if a new parameter is set, thereby maintaining backwards compatibility), we could abbreviate the "Dups?" column further, so it wouldn't be as wide. I'd suggest deleting it entirely, but I'm guessing the bureaucrats might like it? Maybe it can be made optional, too, and bureaucrats can display the version with the "Dups?" column on their user or user talk page for convenience (or it can be put on this talk page). isaacl (talk) 23:35, 26 January 2022 (UTC)
I think there is a problem with tables which are put in the right of the page. It would not behave the same as left-right table. Thingofme (talk) 10:48, 28 January 2022 (UTC)
I strongly disagree with removing either the Ending or Time left columns - while it's nice to see "oh, this ends on 3 Feb", it's another to know "that's in six days", which is particularly helpful when you're at the beginning/end of a month, or just genuinely don't care to do the maths to figure out how long it's going to be before X date. I do agree that the "Status" column is a little unnecessary, since if there is still time left on the clock it's obviously open (and if it's closed, it's removed from the table). That being said, it's a relatively tiny column that won't likely do much if removed. Primefac (talk) 11:13, 28 January 2022 (UTC) updated 11:24, 28 January 2022 (UTC) following comment by xaosflux
Hmm, that's less than ideal. Thanks for the ping, I'll see if there's a way to make this more responsive. firefly ( t · c ) 11:28, 28 January 2022 (UTC)
Meh, don't think we need to optimize this page for tiny screens. — xaosfluxTalk 11:17, 28 January 2022 (UTC)
An increasing percentage of editors are smartphone only or at least smartphone primarily. Seems less than great for us to meh them away. Best, Barkeep49 (talk) 11:23, 28 January 2022 (UTC)
Meh, don't think we need to optimize this page for tiny screens - we should really be trying to optimize the whole site for mobile screens, considering the amount of editors and readers that primarily use mobile phones (me included). This number will only grow over time, as almost all websites are no longer written as desktop first. Best Wishes, Lee Vilenski(talk • contribs) 11:35, 28 January 2022 (UTC)
Is there any reason this is right aligned, btw? Best Wishes, Lee Vilenski(talk • contribs) 11:37, 28 January 2022 (UTC)
I'm in favor of general things that will help readers, and something to help with "wide right-justified tables" in general is a good idea to work on, seems like a technical problem - not sure how many articles have this issue. Working on "it is too wide" that scrolling could be needed at all, that is what I'm saying isn't worth working on - that is something present across articles everywhere. — xaosfluxTalk 14:10, 28 January 2022 (UTC)
A related stackoverflow thread about this behavior for float:right elements on small resolutions; would be interested in general solutions. — xaosfluxTalk 14:18, 28 January 2022 (UTC)
I just recently found {{wide template}}, which I took to TFD because it's not used much, i.e. it's transcluded in no more than ten. If the TFD discussion results in "no consensus" or "kept", then I may like to use the template for the "RFX report" template in the /Header subpage. George Ho (talk) 06:18, 3 February 2022 (UTC)
I'm not sure I understand: if you think it may be useful, I'm not sure why you proposed its deletion. I still think using a media query is preferable, though, rather than hardcoding a surrounding scrollable box for all situations. isaacl (talk) 06:31, 3 February 2022 (UTC)
I hadn't thought about wide template's potential until now. As said before, it's rarely used to this date, so I can subst the template if "deleted" is the result. George Ho (talk) 07:58, 3 February 2022 (UTC)
Maybe {{scrollable}} alternatively, but its box window makes the template less visually appealing unless modified to allow border and background settings. Would that matter much? George Ho (talk) 22:38, 12 February 2022 (UTC)
Turns out that the "overflow" coding still doesn't resolve the mobile issue. Even the "width wouldn't help much. I think probably the "float" style may have cause the mobile issue? George Ho (talk) 02:18, 5 March 2022 (UTC)
Hello George Ho; I believe I've fixed the reported problem. Can you verify that you are now able to access all of the information rendered by {{RFX report}} using your smartphone? Thank you.--John Cline (talk) 05:57, 5 March 2022 (UTC)
It works on my smartphone. How about yours? George Ho (talk) 06:06, 5 March 2022 (UTC)
As far as I can tell, everything is working as the module intends. I believe the situation is now resolved. I missed seeing your original posting about this problem or it could have been fixed back then; I'm glad I noticed the current activity on my watchlist and was able to help. Best regards.--John Cline (talk) 06:21, 5 March 2022 (UTC)
As I figured, setting float to either left or right makes the template un-scrollable on mobile. Instead, it's scrollable if set to none, inherit, or initial. I tested this in my sandbox. Still, Redrose64 said I need consensus to approve proposed changes to /Header subpage. George Ho (talk) 11:46, 5 March 2022 (UTC)
I filed a Phabricator task (T303107) about the float setting in mobile view. George Ho (talk) 19:31, 5 March 2022 (UTC)
There is no reason I can think of for those pages to be edited after the fact. BD2412T 00:08, 17 March 2022 (UTC)
I think the folks who fix linter errors sometimes edit old ones. DanCherek (talk) 00:15, 17 March 2022 (UTC)
I have seen people change their username for privacy reasons and then edit archives to make their old username less visible. —Kusma (talk) 23:31, 17 March 2022 (UTC)
How often do closed RfAs get vandalised? (Bot edits to fix things that broke in the interim don't count.) —Jéské Courianov^_^va little blue Bori 00:08, 17 March 2022 (UTC)
So does almost everything that is an "archive", however we don't bother with protecting things that aren't being disrupted. — xaosfluxTalk 00:32, 17 March 2022 (UTC)
What xaosflux said. Also, why do you propose for only successfuls? Happy Editing--IAmChaos 00:33, 17 March 2022 (UTC)
@Faster than Thunder: What those people said. Before proposing a solution, please demonstrate that a problem exists that needs to be solved. In short: if it ain't broke, don't fix it. --Redrose64 🌹 (talk) 19:12, 17 March 2022 (UTC)
@Redrose64 and Xaosflux: suppose I fail an RfA. After a couple of months, I start making tiny changes to my failed RfA — one small word change per day. Unless someone has watchlisted the RfA page, my edits would be whitelisted, with good score, and these changes would go unnoticed. Lets say after stealth editing of 3-4 months, I have a consensus supporting me to be an admin. What if I contact the closing crat saying "yo. You closed my RfA as successful 7 months ago, but forgot to give me the tools. Now gimme!" We should start protecting failed RfAs, and AfDs (closed as keeps) for that reason. —usernamekiran (talk) 21:19, 17 March 2022 (UTC)
Huh? Have you heard of page history? Schwede66 21:27, 17 March 2022 (UTC)
Yup, you'd totally fool me with that gambit - go for it. — xaosfluxTalk 21:28, 17 March 2022 (UTC)
If a bureaucrat were to oblige such a request without question, without so much as a brief peek at the history of the page, I'd humbly submit that the issue is less one of wholesale page protection to prevent repetition of a one-off event and more so one of basic competency among our bureaucrats. For the record, the latter is not an issue that I believe exists. I'd not oppose this if a widespread issue of problematic editing to closed RFAs were to arise, but I see no such evidence presented. Tyrol5▸[talk] 22:15, 17 March 2022 (UTC)
Schwede66, Xaosflux, Tyrol5, don't get too annoyed, I think UNK was just kidding. --Floquenbeam (talk) 22:20, 17 March 2022 (UTC)
I did not join Wikipedia to become anything short of a crotchety, process-ridden imbecile. Let me have my fun! Tyrol5▸[talk] 22:24, 17 March 2022 (UTC)
hehe. Indeed I was joking. I should have said "yo. You closed my RfA as successful 7 months ago, but then you had a poop emergency, and later forgot to give me the tools." Seriously speaking, I have undone vandalism in at least two AfDs weeks (probably months) after they were closed. And I think I have seen a similar case with an RfA, not sure though. I agree with rest of the people - no need to protect anything. If any particular page in any namespace becomes target of repeated vandalism, only then we should protect it. Case-by-case basis as we do currently. —usernamekiran (talk) 05:18, 18 March 2022 (UTC)
Indiscriminate protection as proposed will be a net negative. Gnomes and bots often make technical edits to RFAs, AFDs, archives and other closed discussions. Examples are to fix Lint errors and template maintenance following WP:TFD. There are already too many overprotected pages that interfere with this. See this and this which resulted in removal of full protection from over 2,000 pages for this purpose. ಮಲ್ನಾಡಾಚ್ ಕೊಂಕ್ಣೊ (talk) 10:57, 19 March 2022 (UTC)
FWIW, Commons does protect RFAs after they close. --Rschen7754 06:30, 18 March 2022 (UTC)
How do they deal with the wiki syntax changes? At my own RfA, Wikipedia:Requests for adminship/Kusma, the version at closure is broken. There has been about one edit per year to the RfA since its closure, all of them helpful gnoming, zero vandalism. —Kusma (talk) 11:33, 18 March 2022 (UTC)
Commons is Commons; English Wikipedia is English Wikipedia. They have their policies, guidelines and conventions; we have ours. What one project decides to do is not binding on any other project. We have, for example, WP:V, and it is a longstanding policy that we are all expected to observe - but some Wikipedias do not have anything similar, and that's their choice. No Wiki should be told that it must step into line with all the others. --Redrose64 🌹 (talk) 23:47, 18 March 2022 (UTC)
Let's assume good faith here, Redrose64. I'm confident Rschen7754 was offering that as a point of comparison and not as a directive to align with Commons. Ed[talk][majestic titan] 05:02, 19 March 2022 (UTC)
Request for comment on administrator activity levels
Please see the following RfC, suggesting that we increase the minimum activity requirements for administrators to an average of 20 edits per year, over a 5 year period.
So they make 100 edits in the first year and then none for the next 4 years! There has to be at least a minimum number of edits per year --Bduke (talk) 11:03, 18 March 2022 (UTC)
Strictly speaking, since the current condition remains @Bduke, you'd want to make 96 edits in the first year then one edit every other year. However, as WTT said, firstly he specifically notes he'd like a higher standard and was just starting with what would get broad support, and there probably is support for a significantly higher standard. Nosebagbear (talk) 09:52, 22 March 2022 (UTC)
I should perhaps add there should also be a simple way for admins who have been de-admined to get it back when they have again become reasonably active after a break. --Bduke (talk) 11:23, 18 March 2022 (UTC)
@Bduke Can I suggest that you have a look at the actual RfC, where the proposal is in more detail and there is also some suggestion of a higher rate (100 edits each year over 3 years). WormTT(talk) 11:26, 18 March 2022 (UTC)
Stats
I am a former admin, who resigned as I am now over 80 and my memory is not as good as it was. So you have to look at my comments as being from a long term wikipedian who was an admin for quite a long while, or a senile old codger. It seems that we have a lot of admins who are now less active than they were, and a decreasing number of new admins. Is that correct? Do we have any details of the activity and number of admins over the years. I think it is clear that it is now much more difficult to become an admin and that less wikipedians are trying to become admins. Do we have any studies that could show that admin activity overall is just as good or effective as it was say 5 or 10 years ago, or are bad things happening because not as many admins are keeping an eye on them? We probably need a serious study. --Bduke (talk) 06:07, 24 March 2022 (UTC)
Hi Bduke, Re your second question we definitely have a declining number of admins. I started Wikipedia:RFA by month over a decade ago and the decline in new admins is clear. 2021 had just 7, the fewest new admins in any year since records began. 2010 had 75 and that was seen as a huge decline from previous years. Even in 2014, the year when editing levels dropped to something of a low, we had 22 new admins. It is clearly more difficult to become an admin than it was in the 2003-2008 era, for starters since the unbundling of rollback it has no longer been possible to pass RFA just as a "good vandal fighter" - you need to have added some content with inline cites. I'm not convinced that it is more difficult to pass RFA in 2022 than it was in 2010, I see the problem as people not being willing to run until they are passing by acclamation - Most of the recent successful RFAs were by huge margins. The three highest unanimous RFAs, ones with no oppose or neutral !votes, have all been in the last couple of years. As for the issue of admin backlogs, I think the picture is less clear, but in my view the direction is such that a prudent driver would steer differently now rather than continue to drift off the road. ϢereSpielChequers 07:01, 24 March 2022 (UTC)
All true - now the latest one is concluded, 2022's 3 successful Rfa's have totalled 644 supports and 5 opposes! Johnbod (talk) 04:43, 26 March 2022 (UTC)
I'm not sure how I feel being responsible for 100% of those opposes... Sdrqaz (talk) 19:21, 26 March 2022 (UTC)
The opposes.. sucked, to say the least, so I wouldn't worry too hard about it casualdejekyll 23:00, 26 March 2022 (UTC)
So we have a problem. The fact that the few successful applications were highly successful, but that there are very few of them, is not good news. It will lead to not having enough active Admins, even if the few are very good. --Bduke (talk) 07:34, 24 March 2022 (UTC)
Maybe there needs to be a more active effort to encourage people to run for adminship even if they're likely to pass on the low end? Is there any evidence to suggest that admins who pass with less support are ultimately less effective than those who pass with better numbers? I just wonder how much the community may be shortchanging itself by requiring at least 65% support (and then it's not guaranteed even as it's more stressful for the candidates). For instance, if lowering the bar to 50% would result in fifty competent new admins within the next two years...I'm not going to advocate for changing a process I've had little involvement with, but I think there needs to be awareness of the possibility that there's fewer admins because those in a position to make the process easier or harder have ultimately chosen the place we find ourselves at presently. In any case, something I don't think I've seen is any real "admin drive", where existing admins put together a coordinated effort to put forth new candidates. Something else that might be useful could be "admin testimonials", where admins talk about the good and bad experiences they had at RfA (and giving some room for failed candidates to provide their own constructive thoughts on the subject) and why they wanted to be admins and how they've ultimately found the experience to be rewarding. Or, to put it simply, better marketing, because what I don't see is non-admins being given much reason to want to become admins other than their own desires for tools. DonIago (talk) 13:15, 24 March 2022 (UTC)
The existence of oppose votes are problems because it could get into a pile-on oppose votes, and make RfA very stressful. So most successful RfA now have over 95% support rates, some have no opposes like Modussiccandi or Firefly. So we also need to have lower-end admins. Thingofme (talk) 13:18, 24 March 2022 (UTC)
I don't think there is any real correlation between the support percentage at an RfA and how somebody actually performs as an admin. I believe several people have mentioned that opposing Money Emoji (aka Moneytrees) and Galobtter was a mistake. Ritchie333(talk)(cont) 15:13, 24 March 2022 (UTC)
I personally believe the community as a whole has taken WP:NOBIGDEAL and decided to completely throw it out the window. I disagree with that position but that's just what I'm observing. casualdejekyll 14:08, 24 March 2022 (UTC)
I'd be very curious to see how a straw poll on the subject of whether NOBIGDEAL is still valid in practice...and whether it should still be valid in practice...would fare. DonIago (talk) 14:26, 24 March 2022 (UTC)
I believe that this community consensus confirms that Adminship is a desireable position - ergo, it is a big deal. WormTT(talk) 14:43, 24 March 2022 (UTC)
I don't think NOBIGDEAL was anything more than an offhand comment by Jimbo once. In terms of selecting who should have administrator privileges - that's a big deal. In terms of do administrators have more respect and get away with things new users don't - no they shouldn't, being an administrator in itself should not be a big deal in working out what to do. Ritchie333(talk)(cont) 15:13, 24 March 2022 (UTC)
FWIW I have believed for many years that NOBIGDEAL is de-facto WP:HISTORICAL. Ad Orientem (talk) 15:55, 24 March 2022 (UTC)
Another nice way to see the issue is with the RfA candidates by year graph. We can see that both the number of RfAs has decreased significantly but also that RfAs have approximately double the success rate they used to have. That illustrates well both the increased reluctance of people to run and also the overcompensation (to word it somehow) where people only run if they feel they have overwhelming odds of passing. A. C. Santacruz ⁂ Please ping me! 14:30, 24 March 2022 (UTC)
Most of successful RfAs are nominated by someone else who is an experienced editor. Self-nominating RfAs are very few and rare. Thingofme (talk) 14:50, 24 March 2022 (UTC)
But we also frown on those publicly displaying they would be up for consideration through a userbox, thus relying on a small number of very productive experienced nominators. I think publicizing WP:ORCP widely would be really good and I'm sad that wasn't one of the results of the 2021 RfA reform. It would allow those with initiative to give themselves up for consideration and feedback without it actually being a self-nom RfA and would weed out one of the main concerns with Self-nom RAs (the large proportion of WP:NOTYET self-noms). A. C. Santacruz ⁂ Please ping me! 15:09, 24 March 2022 (UTC)
I think these things also compound in that... we frown on self-nominations because people aren't supposed to want to be an admin. But because an RFA can often be a stressful or grueling experience, few people are willing to nominate someone out of the blue (or at all); and when they do, it's only the rare person who will clearly sail through with no stress. The result is very few nominations at all. -- Aquillion (talk) 04:58, 26 March 2022 (UTC)
And at the same time that candidates are not supposed to want to become admins, they are also supposed to talk about their "best contributions" to Wikipedia. Not contributions that show the kind of editor they are or how they could be a good admin – their "best contributions". (I guess it is one way to weed out unsuitable candidates. If you are not culturally and temperamentally conditioned to think about your own contributions in such terms, you'll be even less likely to go through an RfA and still want to contribute to Wikipedia in any capacity afterwards.) --bonadeacontributionstalk 11:48, 26 March 2022 (UTC)
I'm not sure that simply looking at the declining rate of RfA nominations is the best way to determine whether we have a problem. I think it would be more prudent to look at things like the number of active admins over time, and the length of various admin backlogs over time. In other words, we should measure the problem based on whether or not we have enough admins to do all of the admin work that needs to be done around here in a timely fashion. Who knows, maybe we don't really need to add that many admins each year in order to replace those that have left the project for whatever reason. Personally, I haven't seen too many overflowing admin backlogs (although I'm sure there are a few), or any other evidence that we're facing a critical shortage of admins who are available to do all of the admin work around here. Sure, having more admins is generally better, but I haven't seen evidence that we actually have a problem that needs to be solved. —ScottyWong— 16:28, 24 March 2022 (UTC)
ScottyWong you raise a very interesting point. There is a strong possibility that by emphasizing the number of admins rather than other admin metrics we are falling victim to Goodhart's law. Targeting number of admins as a metric equivalent to effectiveness of admin activity rather than taking a more nuanced perspective makes number of admins a worse measure of the state of adminship on enwiki (I'm pretty sure this sentence is unintelligible but you get my point). A. C. Santacruz ⁂ Please ping me! 16:35, 24 March 2022 (UTC)
To clarify my comment above "this" being that there is no problem that needs to be solved. Best, Barkeep49 (talk) 17:15, 24 March 2022 (UTC)
Barkeep49 I agree that the decreasing number of admins is in-and-of itself a problem, but hyper-focusing on that metric might be preventing us from completely understanding the impacts of that problem past a general sense of impending doom. Ritchie333 raises an interesting point about backlogs. Is there some page on wiki that's tracking the average duration of open requests/cases in administrative areas over time? Is there even some way to get that information? A. C. Santacruz ⁂ Please ping me! 17:25, 24 March 2022 (UTC)
With all due respect, just because some people believe there is a problem doesn't necessarily mean there is a problem. I'd be much more interested in seeing relevant stats and metrics that demonstrate the a distinct lack of active admins rather than reading everyone's perceptions and emotional reactions to the question. Additionally, it is possible that that there could be a problem with how RfA works, while simultaneously there is no problem with the quantity of active admins. —ScottyWong— 17:28, 24 March 2022 (UTC)
I agree that it is possible that that there could be a problem with how RfA works, while simultaneously there is no problem with the quantity of active admins. But it's not "some people" that think it's a problem. It's a consensus of editors. Consensus can be wrong, and individuals can disagree with consensus but understanding that there is a consensus and you're in the minority is important in terms of what you can reasonably expect from others. Best, Barkeep49 (talk) 18:10, 24 March 2022 (UTC)
I acknowledge that there is a recent consensus that there is a problem, but I also acknowledge that that consensus was derived in a discussion that did not benefit significantly from any data or evidence to back it up. It was essentially a discussion about "do you feel there is a problem". I'm not even saying that I think there isn't a problem. I'm just trying to steer the conversation towards identifying exactly what the problem is (using data and evidence) and what the most effective solution would be to that problem. My concern is that reflexively saying "we need moar adminz" may not address all of the problems that we hope it will address, if we don't do the due diligence of digging deeper to understand the true nature and extent of the problem. It's also likely that there have been attempts to quantify the problem using data and evidence, and I'm likely unaware of them, so I apologize if that's the case. —ScottyWong— 18:39, 24 March 2022 (UTC)
Whereas I agree we should be looking for data to find areas with shortfalls and to estimate what is a minimum replacement rate, I also think the influx rate is an important stat to monitor. For the community to thrive, editors should be able to take on new roles as they wish, and not feel an obligation to stay in a role because there's no one to replace them. isaacl (talk) 20:17, 24 March 2022 (UTC)
One obvious place where the lack of admins is felt is at WP:AE. If you absolutely insist I could gather some stats on this, but it's obvious at a glance that most AE requests are handled by a tiny handful of admins (El_C in particular was infamously carrying an absurd amount of weight there before they threw up their hands at the situation.) This is bad on several levels - it places a ton of workload a small number of admins; it means we'll see disruption if any of them stop or if anything happens to them; and it gives the views and interpretations of those admins excessive weight. The latter is undesirable even if they are ideal admins - having many voices helps ensure results are consistent and fair. -- Aquillion (talk) 05:04, 26 March 2022 (UTC)
Thanks, these are good steps in the right direction of quantifying the problem. In the administrative backlog category, I see significant backlogs in areas dealing with file name problems and user name problems. While these are valid issues, I don't see them as being critical issues, like closing deletion discussions, addressing vandalism, protecting pages that are being abused, etc. These are also often considered the more "boring" parts of WP, and therefore adding more admins might fail to attract more attention to these areas. WP:CCI is an interesting data point as well. The fact that the backlog dates back 12 years ago suggests to me that the problem here may not be active admin count either. Back in 2010, by all measures we had loads of admins, and we were adding a new admin every 4-5 days on average. If this problem was already happening back when we had a lot of admins, it may be a problem that won't be efficiently solved by adding a lot more admins either. —ScottyWong— 17:36, 24 March 2022 (UTC)
This is kind of what I alluded to in earlier comments...if there's an area of the project that could benefit from additional admins, then perhaps there could be a "help wanted" drive to find admins who wouldn't find those areas boring (or at least not boring enough to neglect), versus the more general "we need more admins" sentiment that may not be doing anything to address the areas of greatest need (in terms of greatest backlog). DonIago (talk) 19:14, 24 March 2022 (UTC)
@Ritchie333: I wouldn't trust that number as a good reflection of the backlog there that actually needs administrative attention; {{Uw-botun}}, {{Uw-username}} and {{Uw-coi-username}} populate the category and instruct the category should be removed manually when the account is blocked, it is decided the username doesn't violate policy, or "this template has been in place for a while with no action". Perhaps we need a bot that removes the category after an account is blocked or after a certain amount of time has elapsed with no action. Ks0stm(T•C•G•E) 20:14, 24 March 2022 (UTC)
TheSandBot (talk·contribs) does that, though I'm not sure why it skipped over the one Ritchie mentioned. DanCherek (talk) 20:20, 24 March 2022 (UTC)
I'm finding quite a few it seems to have skipped over, but then again I didn't even know that bot task existed, so what do I know. There's quite a few blocked and inactive users in that category. Ks0stm(T•C•G•E) 20:29, 24 March 2022 (UTC)
Both CCI and other long backlogs such as CfD can be done largely by non-admins. (t · c) buidhe 21:00, 24 March 2022 (UTC)
A different way of understanding whether we need more admins is to look at how often a given task isn't getting the attention it needs, and what the consequences are when that happens. Some areas of work (XFD) get backlogged fairly frequently, but they're not time sensitive. Others (AIV, RFPP, AN3) may not get backlogged often or for long periods, but when they do get backlogged, it results in considerable disruption. Still other venues (AE and DYK, in my experience; there may be others) could use more admin effort at virtually all times. I would urge anyone who doesn't think we need more admins to watchlist some of these venues, and see for themselves. An aspect of our need for admins that's even harder to quantify is the risk of burnout in active admins. We may not be falling behind on some tasks, but that may be because some admins are doing more than I'd like. For a considerable period, El_C and myself were the only admins handling WP:GS/IRANPOL sanctions. I certainly would have liked to spend less time there, but didn't feel able to step away; and it had an impact on how much time I was willing to spend on Wikipedia even when I had no RL constraints. Admin backlogs may be short at the moment, but that may not be a sustainable state of affairs. Vanamonde (Talk) 15:28, 25 March 2022 (UTC)
A curious observation
I have noticed, over the years, that when an RfA is fully positioned to succeed, and those in opposition are few and far between, the non-acceptance and badgering of that opposition is numerous and exceedingly fierce. In contrast, when those opposing are numerous, there's considerably less outcry (even when the pile forms on frivolous ground, generating considerable toxicity, and discussion rebuttals could serve a mitigating purpose). And so I wonder: what are the factors that cause this to be? Has anyone else noticed this or am I imagining things? Best regards.--John Cline (talk) 04:22, 25 March 2022 (UTC)
Opposition to an almost-unanimous RfA is socially unacceptable, but once there is a pile, adding to it is not frowned upon? Overall, the way we deal with discussing oppose votes shows to me that "RfA is not a vote!" maybe isn't a great way to do things. In a discussion, pointing out flaws in the nomination statement or in the oppose rationales should be welcomed and not dismissed as "badgering". —Kusma (talk) 11:18, 25 March 2022 (UTC)
I've long wondered if we should remove the support section. What the community needs to know is if there are issues, and following that, to allow to have those issues discussed/assessed by the community.. I doubt if that would ever gain consensus though, people want to be able to driveby and chime in/vote. We'll get a zillion arguments that we're stifling freedom of expression, etc. We wouldn't be, we'd just be shifting the style of discussion. But whatever : ) - jc37 23:09, 25 March 2022 (UTC)
You know, I for one think this is actually a good idea. (What RfA needs is a radical shakeup, and this is a radical shakeup, so one politician's syllogism later..)[Humor]casualdejekyll 00:07, 26 March 2022 (UTC)
This is just me spitballing, but what if we only allowed the candidate to respond to opposes? If there's a person that should have an issue with opposes and have the right to ask for further explanation, it should be the candidate in question. And if they have the temperament to be an admin, they should be able to respond seriously and graciously to opposition...it might even be a better sign for the RFA itself if they did that. I suppose I'm thinking about a situation that happened in my own RFB when I responded to an oppose, and that person actually switched to support. bibliomaniac15 22:51, 25 March 2022 (UTC)
Not sure. Often, the main idea behind responding to an oppose it not to convince the opposer, but to prevent the oppose rationale from convincing others. —Kusma (talk) 22:57, 25 March 2022 (UTC)
You've already lost that user. The "badgering" is usually to try and get a good reason. If you see a decent oppose rationale, there's rarely a response to it. When it's either petty or irrelevant, there's going to be a bit of a discussion needed. Best Wishes, Lee Vilenski(talk • contribs) 23:12, 25 March 2022 (UTC)
Often, not always but often enough, a nomination by an excellent candidate will sail through virtually unanimously except for a handful of opposes that are perfunctory, trollish, or just plain stupid. Those kinds of opposes look even more unfair because of the lack of any other, legitimate, concerns about the candidate, and the more conspicuously unfair a comment looks the more people are going to want to object to it. ReykYO! 23:20, 25 March 2022 (UTC)
looks unfair → is unfair. Brilliantly said otherwise. ~Swarm~{sting} 06:01, 26 March 2022 (UTC)
Yes, we cannot have any opinion that differs from the hivemind, so we'll badger them, and call their opinions perfunctory, trollish, or just plain stupid. Those opinions are obviously illegitimate, not worthy of being posted, so we can call on the swarm to harass, badger, and attack those views because that is sure to increase the collegial nature of the project. Or not. GregJackPBoomer! 11:35, 26 March 2022 (UTC)
I don't know if so many people reacted so viscerally to something I do regularly, I'd think about it a bit more. Happy Editing--IAmChaos 12:48, 26 March 2022 (UTC)
Do you have any rational basis for your belief that the editors in opposition haven't thought about it? Or do you believe that somehow those editors will be assimilated by the hivemind? GregJackPBoomer! 13:34, 26 March 2022 (UTC)
I didn't name any editors or list any examples. If you perceive your own behaviour in what I'm describing then that's pretty telling. ReykYO! 20:31, 26 March 2022 (UTC)
It seems like this pattern is exactly what you'd expect, whether it were a Vulcan-like process of logical deliberation, or the pack-animal behaviour of a bunch of angry simians. (I take no view on where that spectrum the reality might actually lie.) If there's a few 'unreasonable' opposes, they're naturally going to attract annoyed comment, and the opposers themselves are in turn furious about that, so it turns into a rolling maul. If there's a large number of more standard-issue ones, there's going to be both less motivation to 'badger', and it's a less feasible group dynamic too. But it's not especially helpful or healthy. I think the community should sensible take more of view as to whether either on the one hand, ridiculous opposes will be given little or no weight by the 'crats, or even disallowed outright; or on the other, that we're operating more of a 'just count the votes' system and not a junior varsity debate one, and hence people should be more restrained on commenting them. (I understand there was a huge RfC on RfA not so long ago, and I haven't dared look for it, so pardon me if I'm rehashing tiresomely old ground.) 109.255.211.6 (talk) 05:52, 26 March 2022 (UTC)
"Numbers of roughs waited, in various avenues in the town, for Messrs. Paget and Morley; but they did not make their appearance. Ultimately, the hustings was got possession of, the flags were torn down, the scaffolding set fire to and totally consumed, the rabble pelted the friends of Mr. Morley with stones, and several persons were seriously injured. The windows of the Express office and other houses were broken. Thousands of persons congregated in the Great Market-place, and the "lambs" armed themselves with bludgeons broken from the hustings. The police for some hours did not make their appearance, and the mob had it entirely their own way."
Unruly behaviour is common at elections when the place is uncivilised and poorly regulated. See (right) for a historical example. Andrew🐉(talk) 21:22, 26 March 2022 (UTC)
As an RfA contrarian the first few opposes are generally the hardest, since you have no idea who will agree with you and you're pretty much disagreeing with hundreds of other people who will judge you for this. There's no real way to fix this besides having the restraint not to actually punish people for dissent seen as "frivolous" because it's a basic fact of human interaction that this is how consensus works. We need people to say things that they truly believe despite it going against what everyone else is saying. Then one of two things happens. Either the person rallies a ton of other people who were thinking the same but were afraid to say it until they heard someone else agrees with them, or the person gets no traction at all and gets brutally annihilated by the combined weight of the fact that their oppose really is something that nobody else would agree with. Does this system kind of suck? Yeah, sure. But it's how human society works in many ways aside from just RfA and we can't change that by overhauling the process. Chess (talk) (please use {{reply to|Chess}} on reply) 04:08, 28 March 2022 (UTC)
My experience, including sometimes being an early opposer, is not along those lines. An effective Oppose is rarely a rallying call for a bunch of people who were afraid to say what was said. An effective oppose brings up some diffs to show that there is a problem with the candidate's editing, and that problem is ongoing and can't be dismissed as an isolated incident. If the case is made well, and on grounds that the community doesn't have consensus to ignore, then others will rally around that oppose position. There are also opposes of the "doesn't meet my criteria" type, and that can easily degenerate into discussions about the criteria rather than the candidate. Much like a backseat full of driving inspectors arguing whether the driver being tested should know how to double de clutch or how much space to give when passing a husky drawn sleigh. ϢereSpielChequers 15:39, 30 March 2022 (UTC)
Voting oppose at an RfA should be hard—at least much harder than voting support. If you're voting support, the nominators (or the candidate themselves) should have already done the work for you in giving a convincing argument that the candidate should be an administrator, so it's easy to simply sign off on the same reasoning. However, if you are voting oppose, we expect you to provide your own high-quality reasoning to support your contrary position, and because RfA is theoretically a consensus-seeking discussion, we would also expect you to defend your position against counterarguments from other editors. Generally, if an RfA is "fully positioned to succeed", that is usually an indication that the opposing arguments are simply not very convincing, hence the perception of fiercer criticism of the opposition. Likewise, if an RfA is closer to the discretionary zone, that usually indicates that at least some of the opposing arguments are convincing. You're correct that we don't always get it right, but in my experience, if the opposition at an RfA is numerous, there's usually at least one or two valid concerns in the oppose section, and it's a matter of whether you believe those concerns are severe enough to warrant an oppose vote. Mz7 (talk) 09:00, 30 March 2022 (UTC)
Lack of candidates
I've been really struggling to find any suitable RfA candidates recently. Over the past fortnight, I've looked closely at about 50 established editors to see if any of them could pass RfA. The only one that I thought had any decent chance of passing right now flat out declined immediately claiming "inexperience" and wasn't interested in persuading otherwise.
A couple of highlights:
An editor who has multiple GAs, regularly deals with vandalism, very good AfD stats, civil and polite - but has only been on Wikipedia for five months and has 6,000 edits.
An editor with a good mix of content and NPP work, committed to the project, but has civility issues.
A longstanding editor who has all the right skills, but made the mistake of dragging a popular editor to ANI for incivility early in their wiki-career, which has stuck.
A very established editor with lots of FAs and GAs, good understanding of policy, but AfD stats are a bit hit and miss.
An editor with a suitable mix of skills and experience who looks good on paper but has engaged in a small amount of paid editing (at least they're up front about it!)
All of these might pass RfA, depending on who turns up. But I don't want to put forward a candidate who's chances are hit-and-miss, as if it fails they'll get disgruntled and it'll look bad on me. Ritchie333(talk)(cont) 13:41, 23 March 2022 (UTC)
You can try discussing it with them first. I recently nominated a failed candidate (they had a content policy challenge recently that I didn't know about, which I think was what derailed them) - other than that I let them know that they were not a shoe-in, but someone that I thought otherwise had a decent chance to pass, but to only bother if not passing wouldn't leave them disgruntled. — xaosfluxTalk 13:49, 23 March 2022 (UTC)
Some thoughts about those descriptors.
We got serious about Firefly after about 9 months of activity and I was willing to run him then. We decided to wait for 12 months which I think was the right decision. But given how his RfA went he'd have stood a reasonable chance then too. But waiting feels like a relatively low cost to pay given the time that would be required to wait after a failed RfA.
The civility is probably a DQ
I don't know what "has stuck" means but if this person has shown themselves to be genuinely helpful to others I think they could still pass. They would just need to be comfortable not passing with near unanimity like most recent passes.
Maybe not a problem if they don't want to close deletion discussions. If they're going to work say AIV and have good stats there, I think that candidate could be viable.
I suspect the paid editing would be a deal breaker too.
Also I have no qualms about putting up a candidate whose variance I see as high, as long as they know what they're getting into - it's baked into my checklist. I'm in the midst of vetting such a candidate now and they've indicated that they're OK with a potentially rough RfA. I don't know where this vet will end (trying to be extra thorough given what I know) but I think as long as everyone is aware of what might happen it's ok ethically/morally to put them forward. After all if I think they're a good admin I have to be willing to trust their judgement about that. Best, Barkeep49 (talk) 14:08, 23 March 2022 (UTC)
@Barkeep49: "has stuck" is an Anglocism which in this context means, X earned a reputation that they have been unable to throw off over the years; opposite of teflon, perhaps :) SN54129 14:21, 23 March 2022 (UTC)
Thanks @Serial Number 54129. I understand the wording but am not sure what it really means in terms of a wiki reputation. I perhaps should have said "What "has stuck" means will matter a great deal in the specifics for this instance but if..." Best, Barkeep49 (talk) 14:38, 23 March 2022 (UTC)
Ah, I see. You mean, what it is that has stuck not just the sticking of the stickiness in the first place? Cheers, SN54129 14:44, 23 March 2022 (UTC)
For what it's worth, I am glad I waited. As Barkeep said, a small "price" for a lower variance. I would recommend that Ritchie's first candidate wait until they hit ~12mo activity, at which point if they keep up the good work they seem to be doing from the description here, they should pass with no issue. Candidate 4 should be fine - I wonder how many voters look at AfD stats these days. Mine is something like 95% delete, with a ~90% accuracy. I could very easily have gotten a "only pile-on votes at AfD, all deletes, deletionist!" vote, but... didn't. firefly ( t · c ) 15:05, 23 March 2022 (UTC)
There are also some issues with the AfD tools, at least in my experience, as it's only showing up AfD's I've started rather than all my votes. I agree with firefly's thoughts on the candidates. I also think that candidate 3 should be alright if its long ago. A. C. Santacruz ⁂ Please ping me! 16:33, 23 March 2022 (UTC)
@A. C. Santacruz: A box probably needs unchecking. You can also, for example, enter the name of your previous account to show the collective total number of nominations made with both. It's an excellent, and colourful, tool! SN54129 18:37, 23 March 2022 (UTC)
Look at what just unfolded with the most recent RFA and tell me why any editor would want to willingly put themselves through that level of scrutiny. We come back once again to the reason why there was so much dissatisfaction with the RFA reform. It seems like the principle of "you need a thick skin to be an admin" is a blank check to be as rude and nitpicky as possible to someone who chooses to run the gauntlet. WaltCip-(talk) 17:24, 23 March 2022 (UTC)
I wish I could say that Sdrqaz's Ifnord's RfA was a blatant disaster the community will hopefully learn from, but I can already tell nothing will ever change. casualdejekyll 14:05, 24 March 2022 (UTC)
Brain burnt. Meant to say Ifnord but mixed up the past two RfAs. I was referring to the whole feeding him to the lions thing that went on. I regret opposing to be honest. It seems to demonstrate the problem that perfectly suitable candidates are opposed over minor issues. casualdejekyll 19:00, 24 March 2022 (UTC)
Many people withdraw their RfA early as they have a lot of opposes and they know for sure that they won't succeed, so also it could be a concern (only have a 90% chance to pass--> run; 70-90% --> NOTQUITEYET.) Thingofme (talk) 14:15, 25 March 2022 (UTC)
I must say, everyone I've contacted recently have been not interested. Best Wishes, Lee Vilenski(talk • contribs) 18:18, 23 March 2022 (UTC)
I tend to err on the side of caution when offering nominations; and even so, in slightly under three years of hunting for candidates, about 80% of people I have approached have declined, of about 25-30. I think it's fair to say that of late, it's been more common for people to not want adminship, sometimes, but not always because they think it's too much responsibility. Previously, a lot of negative responses mentioned the reputation of RFA. Perhaps this is a good thing; if we can sound the "we need more admins" alarm loud enough, some of these folks may change their minds. Vanamonde (Talk) 01:01, 24 March 2022 (UTC)
Some reasons not to run for adminship
As someone who has a previous failed RfA in the Way Back When at this point, I thought I'd offer my own thoughts as to why I'd be reluctant to run again (I may have brought up some or all of this in prior discussions):
Almost all of my work here is gnome-work and I like feeling that I'm making a difference without investing a significant amount of time/effort. Why do I need the tools?
I'm not interested in content creation, and I have every reason to believe that will be held against me even if as an admin I wouldn't focus on areas in which content creation was a primary concern.
I don't know where more admins are especially needed, much less whether those are areas where I would be interested in focusing my attention, much less whether I'd be any good at assisting in those areas (Disclaimer - I haven't done my research lately either). There's not really an "admin help wanted here" list that I've seen, nor have I seen a good listing of "these are places where you might help out if you're an admin". If it's on me to think "I want to be an admin so I can help out in area X", then that's a valid approach, but I think then you're missing out on people who might be willing to help and good at doing so if only they knew what their options were.
As a non-admin, while I certainly try to be civil and such, I don't make any real effort to keep myself entirely beyond reproach...but from what I've seen at RfA, even one rude comment can spawn multiple objections, and I don't trust that someone reviewing my past interactions won't find something they could jump on, no matter how minor I perceived it to be. There's an argument that the best thing you can do to become an admin is not engage with anyone substantively prior to your RfA, and, if that's true, I think that's a troubling commentary on the process.
I don't believe I'm especially well-known here (perhaps mostly at WT:FILM), and as such, if I did run, I could see people being reluctant to support me because I'm not well-known, people I've had negative interactions with pouncing on it (I can think of one name without even trying) and calling out, as I noted above, things I felt were pretty minor as a non-admin but sufficient to compromise an RfA.
I've heard it said that I'm less likely to ever become an admin because I have a template on my user page in which I express interest in potentially becoming an admin. If that's the case, then I think it's terrible that a) that template even exists, and b) that people would both complain about a lack of admins and discard people who expressed interest in being an admin.
Anyway, I didn't see this conversation having much discussion as to why people don't want to go through RfA beyond the usually-mentioned reasons (perhaps there need to be more 'exit interviews'?), so I thought I'd do a bit of a deeper dive. Hope this is helpful! DonIago (talk) 04:05, 24 March 2022 (UTC)
Very reasonable points, but I wouldn't worry about #5 (not being well-known) at all. Also #6 (template) shouldn't be too much of a problem either (especially if you remove it now....and wait a bit). Johnbod (talk) 05:15, 24 March 2022 (UTC)
Indeed, Doniago, to echo Johnbod's point, there is a converse to your point #5; speaking as someone who is probably too well known, I can say that having a low profile is probably a bonus. In fact, pace the successful candidates, I'd say that description has probably applied to the majority of the last couple of years, at least. SN54129 19:16, 24 March 2022 (UTC)
Thank you! #5 is a pretty mild concern. #6 though...why should I have to remove a template saying that I may be interested in being an admin if in fact I may be interested in being an admin? Sure, it's difficult to be an admin without 'politics' becoming a factor (I imagine), but the whole idea that I improve my odds of becoming an admin if I suppress evidence that I'm interested in becoming an admin (I wonder whether someone would call that out during an RfA...) kind of rubs me the wrong way, especially given that we're just talking about a userpage template versus my canvassing or such. DonIago (talk) 05:27, 24 March 2022 (UTC)
I agree that being unknown is no barrier to running. I've had editors I've ended up nominating who were completely unknown to me before their name was brought my way who did very well. That said, I do think having a group of editors who really likes you is an advantage. In looking at our last three successful candidates I think Blablubbs was better known than Firefly who was better known than Modussiccandi and that's why you see their support totals in that order. Similarly I think the difference in CaptainEek's incredibly close RfA were the people who knew and trusted them from their experience together. Best, Barkeep49 (talk) 19:41, 24 March 2022 (UTC)
Regarding the userbox, it is a point that means little without context. On the user page of an editor with a demonstrated lack of maturity and an over-eagerness to act beyond one's capability, it suggests to me a view that adminship is something more akin to an achievement in a video game than license to perform (decidedly unglamorous) back-end website maintenance. On the user page of a generally well-spoken and thoughtful editor who is otherwise demonstrably mature, it suggests to me that such an editor is open to the role if approached (and, indeed, it places a category that may be useful for prospective nominators to find suitable candidates). You will find that a small but vocal minority hold the view that an affirmative desire to become an administrator is "prima facie" evidence of hat-collecting. I think probably the category ought be renamed from "Wikipedia administrator hopefuls" to "Wikipedia editors interested in becoming an administrator" or something similar and that "someday" (which seems needlessly whimsical to me) ought be dropped from the userbox, or at least replaced with something more objective and forthright, like "may consider the role in the future" so as to allow folks to set fort their potential interest in a more professional manner. Tyrol5▸[talk] 00:43, 25 March 2022 (UTC)
Incidentally, there is a slight variation on the userbox you display, something like, "This user is not an administrator but would like to be one someday", rather than "might like to be one someday". My sense is the former evokes a stronger reaction of suspicion, while the latter, which you've chosen, strikes me as somewhat more deliberative. Your mileage may vary, of course. Tyrol5▸[talk] 00:50, 25 March 2022 (UTC)
I honestly can't recall whether I noticed that there were two different versions at the time I chose the one on my page, or whether the other one even existed at the time, but personally I'd never say I would like to be an admin versus I might like to be an admin, as likely evidenced by the other things I've said in this thread. It's interesting to me that the "would" template does include a warning about the potential negative repercussions of applying it, while the "might" template doesn't include such a warning; I wonder whether it should, or whether the "might" template is so much less concerning that one isn't merited for it? DonIago (talk) 01:35, 25 March 2022 (UTC)
You're right, and I very vaguely recall the discussion that led to that note of caution. I'm indifferent. My general sense is that an editor with the clue of someone who might be a good candidate at RFA would probably understand that such a user box might ruffle a few feathers and be able to assess it on its merits. But I'm cognizant of not wanting to "hide the ball" either. Seems to me the better solution is a rewording of the user box and cat, or the selection of a different user box entirely, if one wished to indicate their potential interest on their user page. "Would like to be one someday" or "might like to be one someday" makes it sound like an aspiration to be (or a tentative interest in being) a firefighter or an astronaut or something. "Category:Wikipedia administrator hopefuls" is an offputtingly juvenile title for an otherwise useful category. Tyrol5▸[talk] 20:55, 25 March 2022 (UTC)
Gnomes can definitely pass RFA, but you need some content contributions. At least enough to demonstrate you know how to do an inline cite to reliable sources. Someone with just a DYK is unlikely to pass unanimously, but we don't require unanimous support. I wouldn't worry about one rude comment if it isn't recent, people do hold admins, and especially candidates for adminship to high standards, but an oppose based on a comment from over a year ago isn't likely to derail an RFA unless others can show that there is a pattern that still persists. If the rudeness didn't rise to the level that merited a block then I'd remind you that even formerly blocked candidates can pass if the block is from over a year ago and their behaviour has changed. !voters do like to see some indication that you are getting involved in an area where tools are useful, reporting vandals etc. If you haven't done any content contributions I'd suggest getting involved at AFD - adding references to articles worth Keeping and voting delete or merge if you can't find sufficient sources for us to have an article. ϢereSpielChequers 07:19, 24 March 2022 (UTC)
Thank you for your thoughts!
I should say upfront that this section isn't intended as a 'convince me to run for RfA again' argument. I wouldn't even consider self-nominating, and when an admin approached me about potentially running in 2020, in the end we both felt it likely wouldn't work out well.
If adding cites to reliable sources is generally considered sufficient content creation, then sure, I've done that (I'm pretty sure any editor with almost 95K edits has at some point), but I've seen objections on the basis that an editor hadn't written articles, and while I'd like to believe I'm a proficient editor, writing articles isn't one of my areas of interest.
I honestly have no idea how many of my (recent) past comments might objectively be considered rude or might bite me in the butt if I did run for RfA again. I'd like to think I'm more terse and to-the-point than rude, but I freely admit others might feel differently, and I'm not even saying they'd necessarily be wrong to feel that way. OTOH, I've seen many an admin act in ways that left me a bit gobsmacked...which raises other questions, but I won't get into that here.
I do have zero blocks! :)
I've had some involvement at AfD (especially CfD) including nominations, but I wouldn't count myself as a 'regular' at such, though I think I did precipitate a sea change in at least one categorization schema, if that would count for anything. Indeed, I'd say that contributing to policies/guidelines in areas where I have interest is one of my stronger suits...but you don't need the tools to do that, either.
In the end, I think there's some cognitive dissonance between the community 'requirements' to become an admin and the actual skillset needed to be a competent admin, and it's my perception that the desires/expectations that the community places upon those who wish to be admins are frustrating efforts to elevate those who might be perfectly competent admins but aren't rising to the level needed to pass muster by the community. DonIago (talk) 12:53, 24 March 2022 (UTC)
I don't want to run because if I wanted to subject myself to a week of abusive screaming about what a horrible person I am, I get enough of that at AfD. ReykYO! 19:46, 24 March 2022 (UTC)
My sentiments are similar. Cleaning up after petty ego-tripping is a part of this place that I don't like. Why should I create an opportunity for more of it, and at my own expense? Even if I felt confident that I could commit the time to do admin things (which is less and less true), I have no inclination to subject myself to that. XOR'easter (talk) 03:26, 27 March 2022 (UTC)
This is contrary to the spirit WP:NOTCOMPULSORY... another of Wikipedia's various documents that earnestly document something other than Wikipedia as it actually socially functions (or dysfunctions). Yes indeed, it does seem that the expectation is more and more that to pass RfA it's necessary to do a ton of article writing, a ton of gnomery, to commit to keeping both of those up after being enmopped and to commit to an additional workrate of admin-only tasks. All while having the patience of a plaster saint, and the comms strategy of a politician spending their entire term running for re-election. And if by some miracle a candidate does pass such scrutiny, then subsequently slacks off on any of the above, it's taken as a rationale for why admin standards need to be even higher, in some cases to the point of ending up WP:POINTy. This is not the way to get healthy!109.255.211.6 (talk) 06:32, 27 March 2022 (UTC)
It's good that there's a category for it, but that doesn't describe the kind of help that's needed or give any indicator as to how perennial those backlogs are. There's also the question of whether people who might be interested in being admins and helping with those backlogs are aware that that category exists. DonIago (talk) 12:49, 25 March 2022 (UTC)
I'm not sure that provides as much depth as might be desirable for potential admins, but fair enough. As I mentioned in my OP, I hadn't really looked into it of late. DonIago (talk) 13:29, 25 March 2022 (UTC)
Point #6 genuinely baffles me. Is earnest enjoyment of Wikipedia and open appreciation of what it strives for now a bad thing? Is it really that hard to imagine that not everything is about collecting trophies? I refuse to bikeshed the precise wording of that infobox; treating it as a disqualification is just silly.XOR'easter (talk) 16:43, 27 March 2022 (UTC)
For me there's a #7: I don't want a specific userright that is bundled with the mop by default (CRASH). —Jéské Courianov^_^va little blue Bori 19:28, 27 March 2022 (UTC)
Just because it's bundled doesn't mean that you have to use it. There are an awful lot of rights in the admin bundle, and there are several there that I'm certain that I've never used - such as: Create short URLs; Delete tags from the database; Forcibly create a local account for a global account; Import pages from other wikis; Reset failed or transcoded videos so they are inserted into the job queue again; Revert all changes by a given edit filter; Set user's mentor; View information about the current transcode activity; and others. Will I get de-sysopped for not using those? I hope that better grounds could be found. Will I care if they're removed from the bundle? No. But I do care that some of these were removed from the bundle a few years back, some of which I had previously used without causing problems for anybody. --Redrose64 🌹 (talk) 22:09, 27 March 2022 (UTC)
I have an existential gripe with PC writ large, to the extent I don't want the CRASH badge, full stop, even if I don't use it. In fact, that was the reason why I was deopped in 2010. —Jéské Courianov^_^va little blue Bori 22:16, 27 March 2022 (UTC)
As I understand it, from a quick look at the response to two proposals in the previous RfC there's strong opposition to the idea of any sort of 'admin, junior grade' sort of splitting. Anything that's framed that way seems a long way from ever happening. One might argue that a self-denying ordinance in which one stands in a full, normal RfA, but then voluntarily asks for certain rights to be removed could be different. But first there would have to be agreement to allow that, then the software would have to be tweaked for it to happen, and at that point there might be the concern that de facto it becomes the same thing. "Optional questions to the candidate: #umpteen. Will the candidate agree to my quixotic and whimsical [[User:MeMyself/Admin Criteria]] which requires them to give up some of the rights as I don't think they're up to using them," So I suspect that's still likely not to be a flier. 109.255.211.6 (talk) 03:03, 28 March 2022 (UTC)
That doesn't apply here because it is possible to not be an admin and still have a CRASH badge. That RfC was about unbundling the core parts of the admin userright; this is a peripheral one that has always been available sans mop. —Jéské Courianov^_^va little blue Bori 06:05, 28 March 2022 (UTC)
Wait, what is a CRASH badge and what does it have to do with PC protection? @Jéské Courianocasualdejekyll 22:40, 7 April 2022 (UTC)
In recent years, our RfA processes have become increasingly stringent. You are currently an editor with less than 10,000 edits. Currently, any editor with less than 15,000 or 20,000 edits would generally be doomed to fail. What are your thoughts on this and how would you address the Wikipedia community about this issue?
I was curious to what degree the data on recent RfAs would support the premise of this question. I had previously come across this graphic at WP:RFA Guide showing edit counts for successful RfAs. Unfortunately, it's now more than a full decade out of date! Also, since averages are sensitive to outliers, it's probably not a good aggregation to use here. So I whipped up an updated graphic (right). It shows an individual data point for each RfA, plus a linear regression line of fit (which doesn't look linear because the y-axis is on a log scale).
A few observations:
My candidacy (not shown in the graphic, since it was made before my RfA closed) was the first successful one from an editor with <10k edits since GoldenRing, who had a successful RfA in April 2017 with around 2,400 edits.
That RfA ended with a vote count of 178/88/14 and a ctrl+f for "edit count" brings up 82 results, so it seems their relatively low edit count was a significant point of contention.
Between GoldenRing's candidacy and mine, there were 11 other completed RfAs from candidates with less than 10k edits, all unsuccessful. It's unclear to what degree edit count was a factor in those cases. (See second plot at the right which adds markers for unsuccessful candidates)
The next most recent <10k edit successful RfA was in 2015. And in fact there are multiple such cases in 2015 and every earlier year going back to 2007, which is where my data starts.
Since I know people will be curious (I sure was), the one really low outlier isn't an error. lustiger seth had 42 edits at the time of their successful nomination in 2008. But they also had 10,873 edits on German Wikipedia.
Overall, I think the data makes a pretty strong case for some form of "RfA inflation", especially when compared to the data from 2007 or 2008, when people were already raising the alarms about inflation. But I see it as an open question to what degree this is caused by the community turning away otherwise good candidates because of low edit counts, or if there's a selection effect where good candidates who could pass an RfA are discouraged from even attempting one because of exaggerated conventional wisdom around minimum edit counts (Barkeep49 put forward this hypothesis in response to the question quoted above, calling it "a great example of conventional wisdom about RfA that is wrong"). Since no-one at my RfA raised my low edit count as a point against me, I'm inclined to put more weight on the latter explanation. Colin M (talk) 22:01, 9 April 2022 (UTC)
This is very interesting data, thanks for compiling it Colin M. Looking through this I have two reactions: a comment/question and a suggestion. The comment/question is whether the regression line log-scaled the edit counts? I'd assume not, given the shape of the regression line. You are correct that averages are sensitive to outliers, but linear regressions are also sensitive to outliers in the form of high-leverage points. Consider, for example, the single candidate with over a million edits. It looks like the next closest editor had only a few hundred thousand, and most of the successful candidates cluster an entire order of magnitude lower, between 10k and 100k. In log-space (like the graphs), those high-magnitude observations are closer together and so have less influence on the trajectory of the regression line. It also makes intuitive sense distributionally and intuitively. With regards to the distribution, the data is probably not normally distributed, but instead is log-normal. This is a property of many time-growth phenomena, and log transformations are recommended when working with log-normal data. The log transformation also makes sense intuitively given what we know about RfA. If we add 9k edits onto someone with 100k edits, the difference is unlikely to change much, but add 9k edits onto someone with 1k edits and that could flip the whole thing. In without log-trasformation, those 9k edits are treated the same regardless of where they are, but by log transforming the data we can capture the "diminishing returns" aspect of edit count.As for the suggestion, if you want to continue with this work, I think larger regression models would be interesting. A linear regression predicting number of edits by a successful RfA candidate predicted by year of candidacy but also age of account would be an interesting next step and would allow us to get a sense of the relative contribution of both. A more complicated model, but probably very insightful, would be a logistic regression predicting RfA success based on edit count, year of candidacy, age of account, and their interactions. The effect sizes and significance of effects would help clarify what factors have been most influential. You may not have that data or may not be interested in taking that up given your new wikipedia-related time sink (congrats), but I always love seeing more data. Thanks again for sharing this and hopefully it prompts further insights. — Wug·a·po·des 22:57, 9 April 2022 (UTC)
No, the regression was not applied to the log-scaled counts, and so you're absolutely right that it's not very informative. But the graphing library I was using added it by default, and I didn't bother to suppress it. Maybe when I generate an updated version with an added dot for my RfA, I'll just make it a simple scatter plot. Or it might be more useful to replace the regression line with a line tracking the median edit count for each year. (By the way, if anyone else wants to play with the data, for example to make their own visualizations, let me know. I'm happy to share. It has a few extra facets not shown here, such as separate counts for automated vs. non-automated edits.) Colin M (talk) 23:52, 9 April 2022 (UTC)
Out of thirteen people with more than 1,000,000 edits, only five are admins - Ser Amantio di Nicolao, BD2412, Materialscientist, Bearcat and Tassedethe. Which one had passed the million before their RfA? --Redrose64 🌹 (talk) 23:29, 9 April 2022 (UTC)
While you were indeed under 10,000 edits, crucially the question stated an expectation of 15,000 - 20,000. Running even 15k as the baseline you'd see more people who've passed since GE - who is and was a bit of an anomaly in terms of edit counts. I would be curious what the chart and line of fit would look like since 2016 because I believe the reform effort that year marks a substantial change in RfA culture because of the expanded electorate it encouraged. Thanks for doing this work Colin. Best, Barkeep49 (talk) 23:45, 9 April 2022 (UTC)
FWIW, here are the number of successful RfA candidates having less than 15k edits for the last several years (denominator is total number of successful RfAs for that year):
2021: 0/7
2020: 4/17
2019: 3/22
2018: 1/10
2017: 4/21
2016: 2/16
2015: 3/21
(All of these were in the 10k-15k range except for 1 in 2017 (GoldenRing) and all 3 in 2015.) Colin M (talk) 23:59, 9 April 2022 (UTC)
Thank you enormously for this work, Colin. At least we have some stats now after years of begging (it's unfortunately something I don't know how to do). While they clearly depict the theoretical edit count 'bar', even if the logarithmic line is skewed by some exceptionally high edit counts, they do rely on the baggage the candidates bring with them in recent years where the vast majority of RfA are destined to pass and with healthy consensus, so let's not get too scientific about the results and how they were achieved. A count of 10 to 20K edits was nothing unusual in 2010 - 2012 and is probably more realistic than an exponentially growing bar.
A possible effect is that candidates don't come forward because they believe such a 'bar' would disqualify them. However, the sample size is so small nowadays that my thoughts are more towards a general disinterest in adminship, or the Sword of Damocles that will be hanging over them if they choose to work at the coal face. On the other hand, seeing how rare RfA are these days, they probably believe adminship to be almost unobtainable anyway. There are no indications that the trend will change. Users such as Ritchie333 who actively scout for candidates will know more, but will probably confirm the same as I did when I used to be actively looking for candidates. Kudpung กุดผึ้ง (talk) 01:22, 10 April 2022 (UTC)
The next useful information would be a profile of the voters such as we did
here, almost 11 years ago to the day. Since the 2016 reforms that increased the number of voters two or threefold, today's profile would look very different. Firstly because most RfA pass so those opposing would be very few by comparison, but more importantly it would reveal who are the most regular participants, who are admins, who are just drive-by supporters, and what their own 'qualifications' are for voting at all. That said, on Colin's RFA, for example, with the exception of around five, the participants would pass any personal criterion of mine. IIRC, Scottywong's computer ran all night to get those stats, but more modern technology would probably do it quicker. Kudpung กุดผึ้ง (talk) 08:23, 10 April 2022 (UTC)
I have no idea, A. C. Santacruz. It was probably something he devised himself. Scottywong was quite a whizz-kid for developing useful scripts and stuff but I don't think he has the code for that any more - or the time. You'll have to see if he answers the pings. Kudpung กุดผึ้ง (talk) 09:10, 10 April 2022 (UTC)
I'm pretty bogged down with university work until Summer so I might see if I can build the code then if Scotty can't in the meanwhile. A. C. Santacruz ⁂ Please ping me! 09:39, 10 April 2022 (UTC)
Scotty is definitely around. If he doesn't chime in here, ask him on his talk page. He might be able to give you some insights how he did it. Kudpung กุดผึ้ง (talk) 12:34, 10 April 2022 (UTC)
@A. C. Santacruz: I posted source code for my public tools somewhere around the time that toolserver folded (you'd have to dig through my talk page archives to find it), and many people grabbed it and ported it elsewhere. There was an RfA stats tool that was ported, but it appears that it no longer works. I dug around on my computer and found a version of the code for that tool. I'm not sure how useful it will be since it's 10+ years old, but it might have some good regex and SQL queries that you can pull out of it to make your life a lot easier. I pasted it here if you're interested. —ScottyWong— 16:45, 10 April 2022 (UTC)
No dispute that the minimum edit count required was inflated rapidly in the early years of RFA. However I'm not convinced that the criteria have shifted much in the last decade. Average edit count of new admins is not a good indicator of the threshold for minimum edit count requirement, as the community ages we have more and more longterm editors available to persuade to run for RFA. As for minima, many editors make a sharp distinction between manual and automated edits, four thousand edits by an FA writer might represent more work and more Wikipedia experience than 40,000 edits by someone using highly automated tools. To get a feel for minimum edit count requirements I would look at RFAs where there were opposes on this issue, and trends in RFA criteria by people who publish their own criteria. I'm pretty sure that those who would auto-oppose anyone with fewer than 10,000 edits would be an outlier, and that an otherwise excellent candidate with four thousand manual edits could still easily pass (though one with less than 2,000 edits would likely be a snow fail). ϢereSpielChequers 16:07, 10 April 2022 (UTC)
Non-automated edits only. Since you mentioned the automated/non-automated distinction, I decided to see what the plot would look like if it were limited to non-automated edits. This time I included myself in the graph (I'm the rightmost dot, with a bit under 8k non-automated edits). Overall there still seems to be a trend of inflation, particularly of the lower end of the distribution. Successful candidates with around 4k or fewer non-automated edits are rare over the last decade, especially in recent years, though that doesn't necessarily disprove your hypothesis. (Minor caveat: This graphic excludes a small handful of editors whose edit counts were so high that the XTools API gives up on counting them. But since we're most interested in the floor rather than the ceiling, their omission shouldn't change much.) Colin M (talk) 16:53, 10 April 2022 (UTC)
Thanks for that. We had one unsuccessful candidate a few years back who had over a million edits, but that was a special case. I think your chart shows that there was a change about a decade ago, and there may be a slow increase since, or it may just be an indication of the decline in the number of RFAs generally. ϢereSpielChequers 19:25, 10 April 2022 (UTC)
How to ask an optional question?
In Special:Diff/1084614762, I pondered over the instruction to "Add your question above this comment", which seemed rather odd, but I went along with it. But that was obviously not the right thing to do because my question is now out of order. Rather than try and fix it myself and possibly make things worse, could somebody who understands better how RfAs are formatted just Do The Right Thing for me here? Is the instruction in the comment just wrong? -- RoySmith(talk) 15:12, 25 April 2022 (UTC)
I have fixed it. The problem here is on me because I didn't follow instructions when adding my optional question which was the first one and then everyone (reasonably) followed my lead. Best, Barkeep49 (talk) 15:41, 25 April 2022 (UTC)
Does lack of content contributions really matter when it comes to adminship?
Back when I ran for adminship, one of the reasons I got opposed was because of lack of serious content contributions. Back then, I didn't contribute any content and all I did was yell at people about "breaking the rules".
And over the years, I've seen other users' RFA's get opposed because of lack of substantive content contributions, without giving any other reason.
Now if you would've asked me in 2009, I would've said lack of content contributions doesn't matter, but since being unblocked in 2012 and having actually contributed content or added references to articles since then, I'd say a lack of content contributions does matter, as it's one way to tell whether a user understands what should or should not be done to a sufficient level that they're suitable for the tools. If nothing else, content contributions (whether small or substantive) demonstrate a users understanding of certain policies and guidelines, especially when it comes to BLP and copyright policies.
Even though many users (once they become admins) tend to focus more or mainly on the administrative side of the project, some of the most extensive content contributors I've seen happen to be admins. And quite a few RFA's I've seen where the nominator or co-nominators will talk about how the user they're nominating has many DYK's, GA's and FA's under their belt.
It seems the argument for not supporting an RFA unless the candidate has contributed content or contributed substantive amounts of content is if they're actually here for the project aims and/or actually dedicated enough to the project to be an admin. While patience, temperament, familiarity with community practice, clue, etc, matter too by all means, over the years, I've changed my stance when it comes to admin candidates lacking any content contributions, because I actually did start contributing content after my 2012 unblock.
What has been your observation when it comes to RFA's passing or not passing based on content contributions? —Mythdon (talk • contribs) 17:53, 21 April 2022 (UTC)
It depends on the candidate. The more accomplished you are in other areas the less you absolutely need content creation in order to pass. The more you do work that requires a lot of deletion of stuff the more you need content creation. But also you can run effectively as a content creator these days even if you don't have a ton of FA/GAs to your name. Good content work is an easy way to assure the community that you can be trusted with the tools and that is ultimately what it's about - can you get the community to trust you or not. Best, Barkeep49 (talk) 18:02, 21 April 2022 (UTC)
Did you happen to read the section immediately above this one? It's got lots of data and discussions thereof on this very topic... --Jayron32 18:35, 21 April 2022 (UTC)
That is a great section but edit counts is a completely seperate topic from content creation. People accumulate edits at very different rates including temperaments (do you do many small edits or fewer larger ones) and areas of work (countervandalism editors will generally have more edits than other profiles). Best, Barkeep49 (talk) 18:39, 21 April 2022 (UTC)
Edit counts is the only way we really have to quantify and get data on content creation. There's not a really convenient metric to go by otherwise. --Jayron32 18:42, 21 April 2022 (UTC)
I can think of quite a few metrics, many of which get mentioned at RfAs. Number of FAs, FLs, and GAs. Number of new articles created. Number of articles created that have been deleted. Average edit change in bytes. Even % of edits to mainspace. Best, Barkeep49 (talk) 20:24, 21 April 2022 (UTC)
Very well put together essay :) —Mythdon (talk • contribs) 18:54, 21 April 2022 (UTC)
It's also why we have a number of admins who were, are and will remain, content creators par excellence yet hardly dip a toe into admin work. Happy Days! SN54129 19:03, 21 April 2022 (UTC)
I think you're spot on when you say that "it's one way to tell whether a user understands what should or should not be done to a sufficient level that they're suitable for the tools." That is the reason that content contributions are usually considered important. In the absence of extensive content contributions, a candidate wishing to be successful at RfA will need to provide evidence to demonstrate their understanding in different ways. Showing extensive content contributions is the easiest (and most standard) way of demonstrating this; other ways of demonstrating it are more difficult and will require a bit of creativity to effectively communicate. I was not (and am still not) an extensive content contributor when I passed my RfA. But, I did have extensive experience in all of the various deletion forums, new page patrolling, and I had created a number of tools and bots that provided valuable services. That seemed to be enough to convince most people that I at least kinda knew what I was doing and probably wouldn't do anything extremely dumb. —ScottyWong— 19:29, 21 April 2022 (UTC)
If an editor doesn't have content experience, then they don't have experience participating in (and resolving) content disputes, and if they don't have content dispute experience as a participant, then they won't be able to help resolve content disputes as an uninvolved admin. They won't know who to block, when to protect a page, or when to delete a page. They won't know what to do when faced with an edit war if they've never had someone revert them (or had to revert someone). If they've never participated in a contentious RFC or talk page discussion over some content issue, they won't know the difference between a normal content dispute and bludgeoning, POV pushing, misrepresenting sources (vs. just a good-faith difference in interpretation of sources), or other types of disruption. They can't judge consensus if they've never participated in reaching consensus. "Content experience" means a lot of different things--it's not just creating articles from scratch or having a certain number of FAs--but what Ritchie writes in his essay about "front line experience" is exactly on point. In my experience, there is a direct 1:1 correlation between how good a particular admin is at resolving a content dispute, and how much content experience that admin has. Levivich 19:40, 21 April 2022 (UTC)
I'm not an admin, but I have been around since 2018 and I have participated in a few RfAs. I never did oppose someone for lack of content contributions, but it was something I kept in mind and I used to have a section in my RfA criteria about it [2]. My reasoning involved trusting someone with the autopatrolled right, since it used to be bundled automatically with the admin toolkit. Admins can still technically assign the rights to themselves, so I can see why someone would still find content creation relevant. Ultimately, we are here to build an encyclopedia, so there's that, too. I also agree with the general sentiment of Ritchie333's essay and Levivich's comment above. Clovermoss(talk) 19:45, 21 April 2022 (UTC)
You should be an admin btw Ymblanter (talk) 13:16, 22 April 2022 (UTC)
Reading that was surprising. I'm curious about what prompted you to say that. But there are definitely things I would like to be more confident in and have much more experience in before I would even consider running for RfA. It's not a never, but again, definitely not soon. There are other concerns I have as well, to be completely honest. Clovermoss(talk) 14:50, 22 April 2022 (UTC)
There are a lot of people who do good work for the project, both on the writing and maintenance side, but don't particularly want to be admins. I recognise this group exists, and it's a bit of balance act to convince someone who doesn't want to be an admin, that actually they do. In your case, Clovermoss, I would only get involved in the maintenance side if it's something that particularly interests you; nobody should complain if you just stick to content. Ritchie333(talk)(cont) 14:56, 22 April 2022 (UTC)
I've never really been just someone whose just focused on content. My comments were focused on that because the thread was about whether admins need to have experience with content creation and I thought I'd chime in with my thoughts. Most of my edits in mainspace are relatively gnomish in nature, even if I've technically created enough articles to have autopatrolled. I wrote a GA once, but most of my other content contributions are a cited paragraph here or there.
I already have experience in the maintenance-related areas of the project. In my opinion, not enough to justify running for RfA and not enough to make certain judgement calls. So, in my mind, I'm a WP:NOTQUITEYET. I also have some other concerns that would make me hesistant to run. Clovermoss(talk) 15:27, 22 April 2022 (UTC)
Not all admins need to make all judgement calls. I'm not trying to peer pressure you into running, as your other concerns or lack of want are probably well-justified, but thought I'd point that out. — Ixtal ⁂ (talk) 16:06, 22 April 2022 (UTC)
I agree that the reason we're collectively here is to create content. I consider my meager collection of 6 GA articles to be a more important contribution to the encyclopedia than the 1000's of blocks or page deletions I've done. For some things that an admin does, having a good grounding in content is important. Anything that touches WP:N or WP:V for example. Many of the WP:CSD, and a lot of WP:AfD. But, a lot of stuff is more technical. I don't need to know anything about creating content to recognize blatant vandalism. Or to be an effective sock-hunter at WP:SPI. Much the same is true of the folks who keep the system running. If you're writing front-end code for editors, skins, or citation tools, you'd better have lots of experience producing content, because how else will you know if what you're writing fills the needs of content producers? But there's also a need for people who know about operating systems, databases, routers, and traffic management. One size does not fit all. -- RoySmith(talk) 01:35, 26 April 2022 (UTC)
Edit summary usage?
I just noticed that a standard part of the RfA boilerplate is a link to the Edit summary usage for the candidate. Why? Are we really evaluating candidates on how often they use edit summaries? Can we lose this? -- RoySmith(talk) 20:20, 25 April 2022 (UTC)
there have been opposes/neutrals based on this in the past I think. But this definitely comes up in ORCP. —usernamekiran (talk) 20:32, 25 April 2022 (UTC)
It's one way to see how much an editor communicates with others. Given that it's just a rough measure (simple counting stats can't tell you the effectiveness of the edit summaries), my impression is that it's used as a quick, basic sanity check. isaacl (talk) 22:00, 25 April 2022 (UTC)
Note that mobile replies in discussions do not even allow you to give an edit summary, so editors that use mobile a lot appear to have lower edsums. Not that the difference would disqualify anyone, really, but it is a rougher measure than desktop-only editors realize. — Ixtal ⁂ (talk) 22:03, 25 April 2022 (UTC)
It's sad that not only do we have editcountitis, we also have editsummarycountitis. I just checked; my percentage is 98%. The vast majority of my edit summaries are generated automatically by scripts. Of those that aren't, the majority are boilerplate like "fix". When there's something worth saying, I'll say it. If we're actually getting opposes at RfA based on low edit summary percentages, those people are not doing a good job of evaluating candidates and we shouldn't be encouraging them. -- RoySmith(talk) 22:23, 25 April 2022 (UTC)
My feeling is that only a value very near zero would raise questions. I was going to mention the prevalence boilerplate edit summaries in my previous comment but forgot. I think people are well aware of them though and so aren't going to distinguish between different percentages to any great degree. isaacl (talk) 22:32, 25 April 2022 (UTC)
Gaaah. That's exactly what I'm on about. That kind of stuff is exactly what has given RfA a bad name. -- RoySmith(talk) 22:51, 25 April 2022 (UTC)
My apologies; I clearly forgot about the discussion in this request for administrative privileges. I know my views on edit summaries are in the minority (I find them useful and provide them in all namespaces), so perhaps this bias is causing me to underestimate the number of people who look at this metric. isaacl (talk) 22:53, 25 April 2022 (UTC)
My RfA as well had some opposition based on edit summaries. I'm not sure about what percent I was at at the time, probably around the 60-70% range. I'm going to be completely honest, I still don't see the importance of leaving them (even though I mostly do), but then again I never did things like vandal patrol so I dunno. ansh.666 07:28, 26 April 2022 (UTC)
This was introduced more than 15 years ago, and was part of a cultural change towards more edit summary usage in general. As it is linked from the edit counter, I don't think the extra link from the RfA is really needed anymore. It is less potentially harmful than the "AfD stats" link though. —Kusma (talk) 22:20, 25 April 2022 (UTC)
RfC on display of vote totals
The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion.A summary of the conclusions reached follows.
Withdrawing proposal. Thanks everyone for your thoughtful rationales. Enjoy the rest of the week and good luck editing ^u^ — Ixtal ⁂ (talk) 12:18, 26 April 2022 (UTC)
Should information about the vote distribution of RfAs in progress be displayed at WP:RFA? — Ixtal ⁂ (talk) 22:16, 25 April 2022 (UTC)
No, as proposer - I think it influences editors before they read the nomination and Q&A, so while I think it is important that other people's votes are public and displayed on the request page, we should try and avoid the subconscious effect the vote distribution has on editors that haven't yet read the nomination. Editors should be voting as much as possible by their own standards and be encouraged to do their own due diligence rather than bandwagon, and I think that being shown the vote distribution before even being able to open the request is a minor detriment to RfXs. I understand this is a minor change and not some ground-shaking improvement on the RfA process, but I think it worth proposing. The vote tally inside the request page is very small and a bit out of sight, which makes it so that those that do not want to read it don't have to, so I think leaving that there is alright. — Ixtal ⁂ (talk) 22:16, 25 April 2022 (UTC)
Yes – after I participate in an RfA, the display makes it easy to stay updated on its progress without having to click through and scroll. I'm not convinced that there is enough of a "subconscious effect" to outweigh that convenience, though I appreciate the intentions of this proposal. DanCherek (talk) 22:21, 25 April 2022 (UTC)
DanCherek I like using the report link for that as it gives me a chronological perspective (e.g. has someone recently uncovered some serious faults in the candidate that is swinging the RfA?). — Ixtal ⁂ (talk) 22:24, 25 April 2022 (UTC)
One way to address your workflow is to only change the display on the RfA page, and for you to transclude the table (with the count breakdown) onto a page in your user space. isaacl (talk) 22:35, 25 April 2022 (UTC)
Yes Let's not try to hide publicly available information to make a point. * Pppery *it has begun... 23:25, 25 April 2022 (UTC)
Yes. Careful making RFCs before you're sure that there's support for your idea. Sometimes discussion or an informal survey is better. –Novem Linguae (talk) 02:11, 26 April 2022 (UTC)
Novem Linguae An RfC ensures that the discussion stays on topic, while informal discussions on RfA proposals in my experience end up becoming an "Everyone is proposing their own ideal solution and venting about a different part of RfXs" type of discussion due to how many problems editors have identified with the RfX process. I had a particular idea I wanted to propose here and I think an RfC is an appropriate format. — Ixtal ⁂ (talk) 07:44, 26 April 2022 (UTC)
Yes Happy Editing--IAmChaos 06:59, 26 April 2022 (UTC)
Status quo. I'm willing to accept that there is a "subconscious effect" - but there is the same effect when you make one more click and look at the page. It's the first thing you see after "Voice your opinion" - the current vote total. If we want to deal with the subconscious effect - we'd need to move away from bolded, counted votes - and the community doesn't want that. Alternatively, we could move to secret ballot, more like a vote. but the community didn't want that either. In all, I understand where you are coming from, but making this change without making the others means we'd have all of the down sides (less information easily available) and none of the upsides (possibly improving the RfA process) WormTT(talk) 07:50, 26 April 2022 (UTC)
Keep my deckchairs the way they are. If you want to do something in this direction, go further towards things like secret ballots, or at least do away with having separate sections for supportive and opposing comments. —Kusma (talk) 07:58, 26 April 2022 (UTC)
Leave as is - primarily for the reason given by DanCherek. Additionally, I would imagine this is slightly more likely to nudge people towards support (we have very, very, few borderline RfAs these days) Nosebagbear (talk) 08:05, 26 April 2022 (UTC)
Leave as is I am not convinced there is a priming effect that is being assumed, or even what direction the effect is supposed to be. IMO (fwiw) both the effect size and its direction are likely to vary amongst !voters (see Replication crisis#In psychology and this letter by Daniel Kahnemann for the reasons behind my skepticism). As such the proposal, while well-intentioned, doesn't address any actual issues with RFA. It just makes the process of keeping track of an RFA (mildly) more inconvenient . Abecedare (talk) 08:50, 26 April 2022 (UTC)
Leave as is I use this all the time, especially following the links if there are new minority votes (like a new single oppose or single neutral) to quickly navigate to that section. — xaosfluxTalk 10:19, 26 April 2022 (UTC)
Leave as is because the consensus reflects mostly by the power of discussions, only partially by number of votes. And I could keep track new RfAs often. Thingofme (talk) 10:24, 26 April 2022 (UTC)
Status quo not sure a sufficient problem has yet been identified to necessitate action. SN54129 11:05, 26 April 2022 (UTC)
Leave as is While the problems with RFA are so well known at this point, it's almost cliche to bring it up, this is absolutely not one of them. This kind of change is like rearranging the deck chairs on the Titanic. Utterly pointless and will have no effect on making the situation any better. --Jayron32 11:51, 26 April 2022 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
