Wikipedia talk:Requests for adminship: Difference between revisions

This is not the page to nominate yourself or another editor to be an administrator. To do so, please follow these instructions.

Current time: 05:30:33, 29 September 2024 (UTC)
Purge this page

Archives

2003 · 2004 · 2005 · 2006 · 2007 · 2008 · 2009 · 2010 · 2011 · 2012 · 2013 · 2014 · 2015 · 2016 Most recent Template:Archiveline Template:Archiveline Template:Archiveline

This page has archives. Sections older than 20 days may be automatically archived by Lowercase sigmabot III.

Is it time to call a moratorium (again) on questions that require more than one answer? Specifically, what I have in mind are questions that follow the format of "here's 10 random user names, please tell me how you'd deal with each one individually." My understanding is that, as these questions require more than one answer, they fall foul of the two-questions-per-editor rule. Am I completely wrong here? Exemplo347 (talk) 22:53, 4 May 2017 (UTC)[reply]

I don't really have a problem with them in principle, provided only one person asks such a question and they don't give more than something like 8 usernames. The answers required from the nominee are generally short and don't often require much nuance, so I suspect they don't take much more time to answer than your typical open-ended RfA question, but I'd be interested to hear from someone who's gone through an RfA recently and answered such a question. I do think they're largely pointless if the candidate hasn't expressed the slightest interest in working at UAA though. Sam Walton (talk) 23:15, 4 May 2017 (UTC)[reply]

Question 9 demonstrated a good answer to this :D — xaosflux ^Talk 23:23, 4 May 2017 (UTC)[reply]

Perhaps the answer to question 1 on that RFA is also relevant... Caeciliusinhorto (talk) 08:40, 5 May 2017 (UTC)[reply]

I think that's the best answer to Q1 I've ever seen, maybe I'll plagarize it someday (just kidding don't banninate me) ansh666 09:38, 5 May 2017 (UTC)[reply]

It is generally ok, as long as they don't surpass 10 usernames, otherwise, they are considered to be two questions. RileyBugz^会話_投稿記録 23:39, 4 May 2017 (UTC)[reply]

The thing about the UAA question is the individual responses are short, but in order to actually test the candidate's knowledge you have to cover a few situations, so you need to list a handful of usernames. I was given 12 in my own RfA, which was really fine since I wasn't asked that many questions overall, I was only slightly annoyed that several of the names were really testing the same thing. Maybe it would be different if it was one of those RfAs with 20+ other questions, I can't say. I asked the question myself a while later in Cyberpower678's second RfA and tried to make sure that each of the 7 names I listed was testing a different aspect of the policy, including a couple that weren't really covered by the policy at all but were just current issue questions, to mix it up. I don't think it should count as multiple questions, and besides everyone who runs these days ought to know it's coming. However I do think it should only be asked once in an RfA. Ivanvector (^Talk/_Edits) 00:21, 5 May 2017 (UTC)[reply]

These are fine when someone specifically claims interest in UAA, but the more important question is why it's relevant to "test the candidate's knowledge" of something they don't intend to be involved in. People usually give some gauzy hand-waving response along the lines of "once you're an admin you can use the tools however you want!" but somehow that never prompts quiz questions about template syntax. IMO there's been enough feedback on these UAA questions that a) the askers should be taking the hint, and b) candidates should be able to predict that they are unlikely to damage their candidacies by declining to answer, at least if they're not planning on doing any UAA work. (Personally, I'd be more inclined to support someone who responded to these questions with "I'm not interested in this aspect of adminship and this question is tedious busywork", but I more or less think that about every pop-quiz type question.)

I have to admit I feel pretty similarly about the "what would you do if you found a speedy tag on this hypothetical barely-coherent unsourced substub about a topic that turns out to be notable" question genre, but in that case I know the questions are motivated by the askers' long-standing interest in broader related issues. I suspect that part of my irritation with the UAA stuff is that many of the people asking seem to have little investment in UAA as a process, or history of participation in conversations about newbie treatment, or engagement with RfA issues. Opabinia regalis (talk) 02:18, 5 May 2017 (UTC)[reply]

↑What she said. I'd be far more inclined to support a candidate who told the people who play these nasty WP:NEWT hypothetical games, on what's already a stressful and unpleasant process, exactly where they could shove them. ‑ Iridescent 10:55, 5 May 2017 (UTC)[reply]

Are you talking about me? I normally try and put forward an opinion based on existing contributions, but if I'm on the fence and can't tell, an open-ended question is a way of seeing how they work things out and how to break the logjam in my mind. I realise RfA is unpleasant, but having to read off-wiki comments about how people hate Wikipedia because they got stung by an unnecessary drive-by deletion (random example) is even more so, in my opinion. I agree that the username questions are pretty bone-headed. Ritchie333 ^{(talk) (cont)} 15:47, 5 May 2017 (UTC)[reply]

It's Andrea James again. Hardly a "random example". --NeilN ^{talk to me} 20:02, 5 May 2017 (UTC)[reply]

I don't think you're the only one who's done this, but it was your essays on newbie-biting I had in mind. That suggests a question being asked with a broader purpose in mind, as opposed to the UAA stuff that mostly looks like it's done to show off.

That being said, I don't find those hypothetical substubs very useful in "breaking logjams"; I don't think it's likely to be a good use of admin time to go digging on google to see if this barely-coherent misspelled single sentence might actually be a real subject, nor a good use of candidate time to do it for a completely made-up example where the only clue that the reaction you're fishing for is "oh, this is real after all" rather than "delete, that's nonsense" is familiarity with your previous RfA questions. In fact, I'm not convinced it's a net positive to rehabilitate this kind of borderline useless article even if a willing editor sees it; it communicates to new editors that they can just write crap and expect someone else to do all the real work. (But IIRC, I did that once and then got accused of biting newbies by fixing it myself instead of teaching them how...) And while there's no doubt some over-enthusiastic deletion going on, I find that boingboing article, um, unpersuasive. The whole thing needs a liberal sprinkling of {{cn}} tags. Opabinia regalis (talk) 21:30, 5 May 2017 (UTC)[reply]

While obviously it's all personal opinions, they do match real-life experiences I've witnessed, and one of the guys commenting is a real-life friend of mine (assuming nobody else would use the same handle and avatar). So I wouldn't put {{fact}} tags on everything there. While some Wikipedia criticism is just butthurt people who didn't get what they want and can't cope with it, other criticism is very much fair comment and perhaps representative of the silent majority. And I want all admin candidates to have sufficient clue to recognise that and manage it accordingly. Ritchie333 ^{(talk) (cont)} 11:49, 8 May 2017 (UTC)[reply]

A significant number of questions are posed by newbies who simply think it's cool to participate in such an important meta area. Some of them have been around for all of 140 edits and already have one of those silly I wanna be an admin uboxen on their u-page. Kudpung กุดผึ้ง (talk) 11:51, 5 May 2017 (UTC)[reply]

When the candidate specifically expresses interest in, for example, blocking usernames, then it would be reasonable to expect a question on it with, perhaps, two usernames. Anymore than two is unreasonable and the latest loophole seems to be packing as many into one question as possible under the guise of their being all related. Aiken D 19:12, 5 May 2017 (UTC)[reply]

Would the community really support someone who expressed an interest in blocking usernames but had to rely on a "test" because they had no practical experience reporting UAA violations? We expect editors who are interested in vandalism work to have experience at ANI and AIV. Mkdw ^talk 19:55, 5 May 2017 (UTC)[reply]

(subsection this part to address the more abstract question of too many questions or not) --Hammersoft (talk) 16:39, 8 May 2017 (UTC)[reply]

Too many questions asked at RfA are BS anyway. More and more, I think, they're for the self-aggrandisement of the questioner rather than to allow the candidate to mull answers over for the community to assess. — O Fortuna ^{semper crescis, aut decrescis} 08:08, 6 May 2017 (UTC)[reply]

As a rough heuristic, by the time the candidate has been asked 7 questions, I've usually got enough to go on to make an informed decision, irrespective of whether I've asked any myself (and I do try not to if I can possibly avoid it). Going up to 20+ is ridiculous. Right now I reckon clpo13 could ignore all of the remaining unanswered questions on his RfA and still pass with flying colours. Ritchie333 ^{(talk) (cont)} 11:54, 8 May 2017 (UTC)[reply]

Twenty-three questions is not just ridiculous, it's ******* ridiculous. We ought to consider enforcing a limit, as well as restraining the length of individual questions. — O Fortuna ^{semper crescis, aut decrescis} 12:25, 8 May 2017 (UTC)[reply]

Without denying that twenty-three questions is absurd (because it obviously is), I can't imagine how such a limit would work. Is it first-come-first-served system? That seems like it will just incentivise people asking questions quickly, just to have the opportunity to ask a question (already in this RFA Q10 was asked less than 4 hours after transclusion – less than 3% of the way through the RfA). And neither candidates or voters are well-served by having only formulaic or ill-thought through questions, rather than questions about the candidates actual record – i.e. the kinds of questions which take more time to formulate.

A different potential approach would be to limit who is permitted to ask questions, but I see only one question on this RfA by a user who isn't extended-confirmed, and none by IPs, so I don't see that that would necessarily work either. Caeciliusinhorto (talk) 13:33, 8 May 2017 (UTC)[reply]

Well, questions are optional, are they not? The trouble is we'll need somebody prepared to be a guinea-pig and not bother answering the stock UAA question and ignoring other ones. Either the RfA will succeed and (hopefully) set a precedent, or the first person to write "oppose - didn't answer a question" will get piled on. Ritchie333 ^{(talk) (cont)} 13:44, 8 May 2017 (UTC)[reply]

What about allowing the candidate or nominator(s) to stop allowing more questions when they deem it reasonable? Haven't though this through, just an idea. Sam Walton (talk) 13:48, 8 May 2017 (UTC)[reply]

It's 38 questions now by my count, if you include the two multi-part questions. --Ahecht (TALK
PAGE) 14:02, 8 May 2017 (UTC)[reply]

I was curious about the impact of questions on an RfA. There's a lot of different ways to look at this. I decided to see how many questions there were in an RfA at the end compared to at 24 hours into the RfA, compared to the support percentage at the same break points. I did this for 30 successful candidacies dating to December 2015, and 15 unsuccessful (non-snow, non-not now, only withdrawn or full term) candidacies dating to July 2015. The unsuccessful candidacies data is a bit problematic due to RfAs being withdrawn early in many cases, and 12 of them in the same time period being withdrawn before 24 hours. Nevertheless, I saw some interesting trends:

1. Among successful RfAs; the number of questions increased 56% from 24 hours to the end, and the number of votes increased by 160%. Yet, the support % decreased only 3.7%. This would seem to indicate that additional questions after 24 hours in successful RfAs have no impact on the outcome.
2. Conversely, unsuccessful RfAs on the same metric; 40% increase in questions, 152% increase in votes, but a 13.4% decrease in support. Perhaps there's a correlation between unsuccessful RfAs and questions asked after 24 hours? This is possible. More extensive data would need to be done to support this, but it's tantalizing.
3. The number of questions asked of successful RfAs averages 15.5, and unsuccessful 19.2.

Certainly more data is needed to tighten things up. For my part, the trends here indicate that placing artificial limits on the number of questions, or the time in which questions are allowed to be asked, would seem to be counterproductive. Just some food for thought. --Hammersoft (talk) 16:39, 8 May 2017 (UTC)[reply]

What statistics above don't provide is information on the relevancy of the questions and the experience of those who pose them. The issue of questions has been demonstrated up in much greater detail here. An in-depth evaluation of the questions themselves is reported here.Kudpung กุดผึ้ง (talk) 23:58, 8 May 2017 (UTC)[reply]

RFA is a hazing ritual. HalfGig talk 02:24, 9 May 2017 (UTC)[reply]

• Questioners should be required to write a 100-word essay on why they feel the question is relevant to the candidate. And they should be required to answer all the other questions and have random people criticise their answers. Then they might have some sympathy for the person whose time they're wasting during an already stressful week. HJ Mitchell | Penny for your thoughts? 00:11, 27 May 2017 (UTC)[reply]

  That's not a bad idea. I do feel that for an editor who says that they intend to work on backlogs, a pop quiz question can be quite appropriate. It lets you see how the editor handles an unfamiliar admin area, and whether they can look up the policies and procedures in question and then apply them correctly. (It may just be me, but I have to field a lot of "Where's to one about... ? questions lately) It's only that we are all getting bored with the ones on UAA. One on DYK might have been more appropriate. Hawkeye7 (talk) 01:14, 27 May 2017 (UTC)[reply]

What needs to be done so that a struck-out !vote is not counted in the tally of an RFA? Roger (Dodger67) (talk) 11:29, 15 May 2017 (UTC)[reply]

Remove the # in front of the comment. Afaik the bot only counts comments that are automatically numbered using the # wikicode. Regards SoWhy 11:31, 15 May 2017 (UTC)[reply]

Actually, the above works only if the vote being removed from the tally is the first vote. To remove a vote from the tally for any other vote, one needs to indent the vote, as to do otherwise will break the numbering for voters after the stricken vote. --Izno (talk) 13:08, 15 May 2017 (UTC)[reply]

Correct of course. Further !votes need # to be changed to #:. Regards SoWhy 13:20, 15 May 2017 (UTC)[reply]

Moved from Wikipedia:Requests for adminship/ONUnicorn

 – With respect to the participants, this was never about ONUnicorn specifically, so per some of the suggestions I've shifted this to a more neutral location. Primefac (talk) 19:13, 23 May 2017 (UTC)[reply]

• Is anyone getting utterly sick and tired of these stupid username questions that tell you nothing about whether the candidate can use the tools in a responsible and trustworthy manner? Ritchie333 ^{(talk) (cont)} 16:34, 23 May 2017 (UTC)[reply]

• Same with me, I'm getting tired of this being not put under multiple questions. — JJBers 16:50, 23 May 2017 (UTC)[reply]

• Thirded? It's not "fun" to make up random names and see if you can "trip up" the candidate. Politely asking Smartyllama to retract their question. Primefac (talk) 17:15, 23 May 2017 (UTC)[reply]

• I offer a kitten to the first candidate who replies to a UAA question by saying it's not among their areas of expertise therefore venturing hastily-formed answers would be irresponsible. (I have no idea if that's ONU, maybe she's an UAA pro, but I hope candidates and voters alike view optional questions as truly optional--there's no more reason every RfA candidate must take a UAA quiz than there is to require every admin candidate be, I don't know, a templates expert. What is important to me is candidates knowing where they can contribute effectively and where they're out of their depth.) Innisfree987 (talk) 17:18, 23 May 2017 (UTC)[reply]

• These UAA comments appear on virtually every RfA. I fail to see what the issue is. The "getting in on the fun" was a bit facetious, I'll admit, and if people take objection to that specifically, I'll strike and rephrase, but these questions have a legitimate purpose. Smartyllama (talk) 17:22, 23 May 2017 (UTC)[reply]
• And to add on to that, as Innisfree987 mentioned, this may or may not be an area ONUnicorn wants to work in. If not, that's a reasonable answer. If it is, I would expect them to show some knowledge of the process and how to handle things. Smartyllama (talk) 17:29, 23 May 2017 (UTC)[reply]

• I think the fact that they appear on virtually every RfA is part of the problem; it's being asked just because it's fashionable to do so. The only way to get it to stop is for candidates to decline to answer.--Pawnkingthree (talk) 17:37, 23 May 2017 (UTC)[reply]
• To be clear I was saying maybe ONU has a lot of knowledge about usernames, I simply don't know (ETA: ah turns out she does! regardless I still feel the same way about these questions generally); but as for where she wants to work, she's told us, and it didn't include usernames. If you want to confirm that, perhaps just ask directly if that's something she plans to work on before going on to a name quiz. I share Primefac's sense that it's a bit of a trap to go straight to a question to which the correct answer may be, "I shouldn't answer that question." There's a lot of pressure on candidates to answer everything lest omissions be held against them--recently a voter went to the oppose side simply because a question wasn't answered in four hours! Taking care to avoid putting candidates in such impossible positions is important to getting good admin candidates to be willing to go through RfA. Innisfree987 (talk) 18:03, 23 May 2017 (UTC)[reply]

• fourthed. --Floquenbeam (talk) 17:35, 23 May 2017 (UTC)[reply]

• Fifthted. Kudpung กุดผึ้ง (talk) 18:18, 23 May 2017 (UTC)[reply]

• If people really want to change this, perhaps they should open a dedicated RfC rather than discussing it on every single RfA, as appears to have been done for some time. Smartyllama (talk) 18:33, 23 May 2017 (UTC)[reply]

• (edit conflict) (And you still haven't taken any notice, Smartyllama?) Getting people to stop asking these username questions by direct means is unlikely to happen; they're entrenched, and it would be rules creep. We have to appeal to the candidates. The first candidate who uses Floquenbeams answer or similar will get 50 Appropriate Boldness Support Points from me. Better than a kitten! If they also self-nominate and thereby get my standard 50 Support Points for Independence, there's 100 points, and a whole support, for them right there. You too, Floquenbeam? Let's offset some of the outraged opposes that would follow "your" reply. It could become a movement. Bishonen | talk 18:40, 23 May 2017 (UTC).[reply]

• Ah, and here's the problem with conflicting RfA standards. A candidate (who isn't an obvious pass) would get at least 10 opposes between self-nominating and refusing to answer a question about policy, likely more. Since we more-or-less require 75% support to guarantee a pass, you'd need to find 30+ editors who would support (and who would not support otherwise) to make it "worthwhile" for candidates to buck the trends. RfA is an exercise in being a politician for a week. All successful politicians know their constituencies. If a candidate knows their RfA constituency, they'll see the numbers on doing those things just don't add up. If we want to improve RfA, we can't try to force candidates to tell people with bad questions to screw off - that just makes things worse. An RfC on limiting multi-part questions to, say, three parts would not be instruction creep. ~ Rob₁₃^Talk 18:57, 23 May 2017 (UTC)[reply]

• Well, put me down as another person who's increasingly leaning to automatically supporting anyone who refuses to answer the UAA questions if they haven't said in Q1 that they intend to work in UAA. If we think that refusing to answer the question is likely to cripple borderline candidates, though, and we reject forbidding them due to instruction creep, it seems though the only remaining option is to wait for the question to go out of fashion and another stupid question come in... Caeciliusinhorto (talk) 19:11, 23 May 2017 (UTC) edit conflict on the above thread, was added to the RFA here. Primefac (talk) 19:16, 23 May 2017 (UTC)[reply]

• Can we make this an actual RfC? I know I would support it. I think these questions could all be answered simply with "go read UPOL." They are utterly non-informative and seem to be excuses for editors to try to make up outrageous names. Eggishorn (talk) (contrib) 19:19, 23 May 2017 (UTC)[reply]
• I was sorely tempted to just leave it at "You said they all have no edits, so per the instructions at Wikipedia:Usernames for administrator attention/Instructions, I would probably not take any action unless and until they edit." In my mind, that's all that question really calls for. ~ ONUnicorn^{(Talk|Contribs)}problem solving 19:27, 23 May 2017 (UTC)[reply]
• Sixthed. I think it's become enough of an issue that we should seriously consider topic-banning people who ask the "gotcha" questions on UAA and CSD, except in cases where the candidate has indicated a desire to work in those areas, from the RFA process. The standard "but the admin bit will give them access to the complete toolset so we need to know how competent they are in those areas in case they change their mind!" argument is hogwash; if people genuinely felt this was an issue, they'd be asking questions about editing the MediaWiki namespace (an area where an incompetent admin can do orders of magnitude more damage than they can at UAA or CSD). ‑ Iridescent 19:51, 23 May 2017 (UTC)[reply]
• Were I to run, I would 100% answer in the form of "A) I have no current desire to work here, and B) am mindful-enough to read PAGs before jumping into X or Y corner of the wiki, should I C) wish to participate in this area in the future." --Izno (talk) 20:05, 23 May 2017 (UTC)[reply]
• I'd wielded the bit for years w/o going near UAA. It's pretty simple, now that I'm there. The egregious Spamu and multi user one's are easy enough. The others I don't worry bout unless they are, well egregious. When in doubt, question the wisdom of carrying out an action.Dlohcierekim (talk) 20:46, 23 May 2017 (UTC)[reply]

  ANd what Iridescent said.Dlohcierekim (talk) 20:48, 23 May 2017 (UTC)[reply]

• Sorry if this seems a little too self-congratulatory, but I'm happy to say I refused to answer one of the standard questions (#9 in the diff), and no one opposed me for it. Although I did chicken out and answer the other 45 questions (well that's what it felt like). --Floquenbeam (talk) 21:25, 23 May 2017 (UTC)[reply]

Heh. That's why you came within 113 votes of failing. — O Fortuna ^{semper crescis, aut decrescis} 10:43, 24 May 2017 (UTC)[reply]

• Seventhed. Lepricavark (talk) 22:03, 23 May 2017 (UTC)[reply]
• I'm not going to call out anyone or any RFA in particular, but I really roll my eyes when the questioner treats these as a sort of "gotcha!" moment. I totally understand questioning someone's judgement if they thought FUCKYOURMOTHER was an appropriate username, but I've seen the candidate gets chastised for not recognizing the name of some random living person. Someguy1221 (talk) 22:08, 23 May 2017 (UTC)[reply]
• As I said in the last go-around on this subject, n+1 for thinking these are stupid and endorsing candidates who don't waste their time on this stuff. The "look how clever I am with these made-up usernames!" crowd is apparently right on that awkward threshold of being experienced enough as editors to participate substantively in RfA but too inexperienced/socially unaware to recognize the growing tide of disapproval for their game-playing. That being said, people using RfAs for grandstanding, attention-seeking, and publicly revealing their own poor judgment is hardly a new phenomenon, and this particular genre of timewaster question is neither the most toxic current contribution to the RfA climate, nor the worst of the historical examples of "de facto Q4" questions. Going back far enough, they used to be more blatantly political, like "Would you put yourself up for recall?" or "Are you a teenager?" By comparison, these are just nuisances, not attempts to filibuster the RfA process or force candidates into choosing a side in an ongoing wikipolitical dispute. Opabinia regalis (talk) 01:32, 24 May 2017 (UTC)[reply]

True. I seem to remember it wasn't that long ago that 'Republican or Democrat?' was deemed an appropriate question (for a while, anyway). — O Fortuna ^{semper crescis, aut decrescis} 10:46, 24 May 2017 (UTC)[reply]

• It always amuses me, Opabinia regalis that threads like these often have a tendency to discuss issues as if they've never been raised before. In fact the topic of RfA questions was researched in very great depth a few years ago here and some almost identical suggestions for reform were made. More recently, whether on the surface it looks reasonable enough or not, Q4 here is another clear case of a cluebie trying to be clever. 90/500 would not be an unreasonable requirement to participate at RfA.Kudpung กุดผึ้ง (talk) 20:43, 26 May 2017 (UTC)[reply]

• Can we at least make a rule that there are a maximum of 2 names per question? To pile on as many as have been is needless and inappropriate. Aiken D 19:40, 23 May 2017 (UTC)[reply]
• (ec) I agree that the user name questions I've seen have a "gotcha" quality and serve no useful purpose whatever. Coretheapple (talk) 19:42, 23 May 2017 (UTC)[reply]
• I wouldn't support instruction creep specific to UAA because this is a more broad problem about multi-part questions. I would support an additional requirement that multi-part questions have, at most, three or four parts. ~ Rob₁₃^Talk 21:08, 23 May 2017 (UTC)[reply]
• The problem is not the number of names, the problem is that the question is being asked of everyone as if there was consensus for it to be q4. If someone says they are going to be active at UAA, or they carefully don't but in the past they've done some bad UAA reports, then a UAA question is relevant. But usually it is a distraction from the RFA. A better solution would be to require questions to be dif supported or relevant to the nomination statement or answers to earlier questions. ϢereSpielChequers 21:10, 23 May 2017 (UTC)[reply]
• Trying to fix the problem by limiting the number of UAA examples kind of misses the point (Well, it misses the point *I'm* trying to make; others are probably making different points). You are not going to be able to "fix" RFA by adding new rules, because that isn't the fundamental problem. The fundamental problem with RFA is that many of the voters, and many of the people asking the generic CSD/UAA/block vs ban gotcha questions are [redacted because it would somehow be considered a personal attack on Wikipedians in general]. And not to sound too pessimistic, that problem is unsolvable. --Floquenbeam (talk) 21:19, 23 May 2017 (UTC)[reply]
• I don't think that it's so much the number of usernames, either. At UAA, unless the username is a blatant violation, the explanation from the reporting user gives some context. The RfA question is just a bunch of usernames with no explanation, and no context for the candidate to go by. ​—DoRD (talk)​ 21:30, 23 May 2017 (UTC)[reply]

It's perfectly simple; just make RfA 30/500 or whatever we call it now. That'd clear out all the noobs (and that includes the latest example, which might not be new, but is experientially a noob, it seems). There's no reason anyway for editors who want to edit the encyclopaedia that anyone can edit to also want to vote in its elections; they're not the same thing. — O Fortuna ^{semper crescis, aut decrescis} 21:46, 23 May 2017 (UTC)[reply]

I know it's a controversial stance, but I would fully support WP:BLUELOCKing RFA pages. I certainly didn't understand the nuances of the administrator role, and what it does and doesn't entail, before I would've met the 30/500 criteria. However, I'm not sure that that's the root of this problem. --Ahecht (TALK
PAGE) 21:52, 23 May 2017 (UTC)[reply]

I don't think it would work. Of the UAA questioners at all of the RfAs currently linked from the RfA page, one had their question struck when they were indeffed as a sock; the rest are all currently extended confirmed (and had easily met the 30 days limit when they commented on the RfA; I haven't checked whether they met the 500-edit limit at the time). I'm sure I wouldn't have had anything useful to say at an RfA before I was extended confirmed, but I suspect that enough of the unhelpful comments at RfA are made by EC users that this wouldn't make much difference either way... Caeciliusinhorto (talk) 08:32, 24 May 2017 (UTC)[reply]

I completely agree. I don't think protecting RFA would help this problem, but would improve the level of discourse in general. --Ahecht (TALK
PAGE) 18:35, 24 May 2017 (UTC)[reply]

This. A lot of the questions are like "what do you do if you see username like uuzzuu". The report would usually clearly state that it's possibly an attempted impersonation of [username]. -- zzuuzz ^(talk) 09:00, 25 May 2017 (UTC)[reply]

• Lets make it more general than that. Why not just clarify that for multi-part questions or questions that asks candidates to reply to multiple examples, each part or example counts as a question for the purposes of the two-questions-per-editor limit. No need to single out UAA quizzes. --Ahecht (TALK
  PAGE) 21:49, 23 May 2017 (UTC)[reply]

  I think this would solve a lot of the problem. Of course, there have been fashions in stupid RfA questions before, and many of those would not have been eliminated by this rule. But they differ from the UAA question in an important respect: they take much less time to answer. So while they do get irritating after a while, they're mainly irritating to voters who have to waste braincells on reading an irrelevant question with a prepared response. Whereas the exact usernames someone makes up for the UAA question can't be prepared for, so candidates have to spend time doing things like googling arbitrary addresses to see why the questioner thinks they are significant, searching to see if any of the names given could be reasonably construed as imitating existing prolific users and/or admins (I don't believe I have ever run into Gogo Dodo, and so I would probably have got that one "wrong" unless I was lucky in my searching...), and so on. When they could instead be either a) improving wikipedia, or b) giving thoughtful answers to actually relevant questions on the rfa. Caeciliusinhorto (talk) 07:46, 24 May 2017 (UTC)[reply]

• If a question requires 10 seperate answers, it's not 1 question. It's 10 questions. Calling these 10 questions 'examples' is a clever way to get around the 'max 2 questions" rule, sure, but IMO this should stop. That goes for any question with a list of 'examples', not just the UAA lists. The RfA page clearly states "Multi-part questions disguised as one question, with the intention of evading the limit, are disallowed." If there are multi-part answers, there are multi-part questions. Quite simple really. If it looks like a duck... Yintan  08:30, 25 May 2017 (UTC)[reply]
• You can't fix every problem with making a new rule, then a new subrule, then a new subsubrule. RfA will end up like golf, except the rules will be enforced by cheetos-eating 20-somethings rather than polo-wearing middle-age pricks. The best course is for candidates to feel confident enough to ignore the questions if they lie outside their area of activity. Doing so is likely to attract few if any opposes. --Mkativerata (talk) 08:42, 25 May 2017 (UTC)[reply]
• Back in January 2016, this was added to the header: "Also forbidden are multi-part questions which are disguised as one question, but in effect are really more than one question and violate the two-question limit."[1] This seems fairly clear to me, but I've now changed it to refer to multiple answers.[2] Feel free to revert, but even more, feel free to apply it. -- zzuuzz ^(talk) 09:00, 25 May 2017 (UTC)[reply]

Thanks zzuuzz, that'll sort it; a shame it can't be applied retroactively though  :) — O Fortuna ^{semper crescis, aut decrescis} 09:08, 25 May 2017 (UTC)[reply]

• It always amuses me that threads like these often have a tendency to discuss issues as if they've never been raised before. In fact the topic of RfA questions was researched in very great depth a few years ago here and some almost identical suggestions for reform were made. More recently, whether on the surface it looks reasonable enough or not, Q4 here is another clear case of a cluebie trying to be clever. 90/500 would not be an unreasonable requirement to participate at RfA. Kudpung กุดผึ้ง (talk) 19:02, 26 May 2017 (UTC)[reply]

• Well, to a point. RFA always had a problem with people asking pointless or show-off questions, but the specific trend here—of people raising intentional "gotcha" questions using situations that could never arise in reality (since actual UAA reports and deletion nominations always specify the tagger's concern), in a deliberate attempt to get the candidate to give a wrong answer to the hypothetical situation, seems to be a new and unwelcome development. ‑ Iridescent 19:13, 26 May 2017 (UTC)[reply]

That was researched in depth too, Iridescent, six years ago. There is a catalogue of over 156 questions that were examined and categorised, ranging from the inappropriate to the silly and ridiculous.

The sample 156 questions were taken from a total of roughly 772 questions on all passes, and all full-term fails in 2010. They are not exhaustive and for many questions only one example might have been be provided. Some of them, particularly multiple questions bundled under the guise of one, may not be strictly in the appropriate section, because they fall into several categories.

Kudpung กุดผึ้ง (talk) 20:04, 26 May 2017 (UTC)[reply]

That survey was over six years ago. I'm not sure from your comments here how you think it affects the current discussion, other than to say: "People have been asking inappropriate RfA questions for a long time." Since that would not be a stunning revelation (to say the least) I'm sure that you mean something else that I'm just missing. I look forward to hearing what you do mean. Thanks. Eggishorn (talk) (contrib) 21:03, 26 May 2017 (UTC)[reply]

The irony is, Eggishorn, that the 6-year-old research is as stunningly accurate and relevant today, if not even more so, than it was then. It's a shame people don't take the hint and read it, it's extremely detailed (embarrassingly so) - then perhaps they would not keep re-starting these perennial discussions as if the issues were a brand new phenomenon each time. Kudpung กุดผึ้ง (talk) 21:51, 26 May 2017 (UTC)[reply]

With all due respect, Kudpung, I think people keep re-starting these discussions because, as your list shows, the problem has been around for a long time and just doesn't get solved. Yintan  22:23, 26 May 2017 (UTC)[reply]

Absolutely, Yintan - all talk and no solutions. What was your impression of the 6-year-old research? Kudpung กุดผึ้ง (talk) 22:55, 26 May 2017 (UTC)[reply]

That there is some progress. At least we don't get the plain silly questions anymore. Not usually, anyway. Yintan  22:58, 26 May 2017 (UTC)[reply]

Can someone explain to me why "poor AfD stats" is seen by some people as an argument to oppose a candidate at RfA? Does that mean they want people to not voice minority opinions in AfDs? Do they want people to add a sixth "delete, not notable" vote to unanimous AfDs in order to boost their stats? If a candidate has AfDs where they did not end up on the winning side, this tells us how they behave in disputes that they do not win. This is useful information, as admins typically disagree with each other a lot and then have to accept and uphold decisions they do not agree with. I think whether the candidate agrees with the majority should not be important as long as they have a useful and meaningful contribution to the debate. My personal AfD stats these days are poor, as I do not contribute to unanimous debates at all (other than as nominator) and only comment on AfDs that I find interesting, sometimes as the lone dissenting voice. How does this make me a bad admin? —Kusma (t·c) 12:31, 25 May 2017 (UTC)[reply]

• Interesting point. Speaking as someone who perhaps isn't as familiar with the admin role as he should be, I would additionally ask: given that the role of an admin with regards to AfD discussions is to close them, not vote (which they're certainly entitled to do, but then they wouldn't be the closer), how is how they vote on AfDs pertinent? Is the suggestion that an admin who tends to cast minority opinions on AfDs would be ill-suited to close them competently? If so, that doesn't seem to me like it logically follows. If nothing else, a good admin simply wouldn't close AfDs in cases where they disagreed with the outcome. It seems like there may be somewhat of a failure to assume good faith here. I look forward to gaining additional insight into this subject. DonIago (talk) 13:03, 25 May 2017 (UTC)[reply]

I also think there's a difference between assessing an article or topic for compliance with rules (voting in an AFD) and assessing and carrying out consensus (closing an AFD). Moreover, I don't trust that tool that gives the AFD stats - for example Wikipedia:Articles for deletion/Redwall doesn't show up in my AFD stats for some reason, and it's not the only one. Also, Wikipedia:Articles for deletion/Political party control of United States state legislatures and governors shows up as me casting a redirect !vote instead of a delete as nom, simply because in a later comment I said I'd be fine with it being redirected and bolded the word redirected, which confused the tool (although the tool still thinks I got it "right"). Like Kusma I see little value in adding a sixth "delete, not notable" vote to a unanimous AFD, or a sixth "keep" vote to a unanimous AFD unless I have something new or different than what others have said to add to the conversation. That doesn't mean I couldn't close the unanimous delete and carry out the consensus. The only thing I can think of for why people are hung up on AFD stats is that they may be looking at it as a measure of your ability to persuade others to your point of view. If there are a bunch of delete !votes and you can persuade others to change their minds and successfully get the article kept, or vice versa, that's a measure of your ability to know the rules, apply them effectively, and get others on board with you. But that would be better judged by looking at the content of both the "correct" and "incorrect" AFD comments, not merely the results. ~ ONUnicorn^{(Talk|Contribs)}problem solving 14:08, 25 May 2017 (UTC)[reply]

• While I can sort of understand this, I logically cannot support it in any way. As long as they are backing up their votes with logic, then it should have no bearing on an RfA, much less a negative one—logically backing up one's arguments may actually be a good thing, at least compared to a really high AfD percentage just going "by nom" all the time. RileyBugz^会話_投稿記録 14:12, 25 May 2017 (UTC)[reply]
• I addressed this in the RfA, but I'll repost it here. It all boils down to the fact that for many people, having a number to cite is much easier than actually digging into the substance of an editor's contributions, whether you're talking about AfD "success", edit count, percent edits to various namespaces, featured articles/DYKs, or the various other numbers that people have used as a proxy for suitability for adminship. With AfD stats in particular, as you mentioned, the roll of an editor in an AfD discussion (judging notability) is completely different from the role of an administrator closing a discussion (judging consensus). This leads to a bit of a paradox: if you are planning to run for RfA you would always want to !vote with consensus to improve your stats, so your AfD stats might be an indication of your ability to judge consensus. However, someone who is !voting against the merits probably isn't the sort of person we want as an administrator. On the other hand, if you're truly !voting at AfD to improve the project, and are therefore the kind of person we want as an admin, you would be !voting based on the merits and not on consensus, and the your AfD stats wouldn't have much correlation with your ability to judge consensus. --Ahecht (TALK
  PAGE) 14:16, 25 May 2017 (UTC)[reply]
• There might be a reasonable argument made that someone who sits with consensus in an AFD understands the correct weighting of the arguments made in an AFD--an understanding which he will need as an administrator to close AFDs correctly—as in, without too many false positives, as evidenced by the number of times he ends up at DRV or AN(I) or whatnot—since consensus is not a headcount. I'm not entirely sure I subscribe to this argument (I think some question is begged regarding whether the user in question might have gamed the system by !voting as discussed above [please AGF], or might not have !voted with some significant rationale actually indicating his agreement with what ends as the consensus position), but I think that's the one I might make. --Izno (talk) 14:32, 25 May 2017 (UTC)[reply]

If you are talking about my take on this, I didn't mean to imply an editor running for adminship gaming the system, I just meant that an editor will generally vote on not too controversial noms, which is what a lot of editors (at least I do it) do. RileyBugz^会話_投稿記録 14:44, 25 May 2017 (UTC)[reply]

I did not intend to respond to your comment. --Izno (talk) 14:52, 25 May 2017 (UTC)[reply]

• For the record, I've repeatedly argued that the entire idea of using AfD stats as a criterion for RfA is outrageously wrong, and that indeed any such votes should be discounted by the crats as a matter of principle [3][4]. Fut.Perf. ☼ 17:04, 25 May 2017 (UTC)[reply]
  • I think you are incorrect for the reason Primefac mentions below (and which I mentioned at ONUnicorn's RfA). AfD percentages are a valid criterion but not the way people currently use them. A very high percentage of "mistakes" indicates a user is !voting in a pattern, without considering the article's merits at all (such as !voting delete or keep on every AfD). Conversely, a high percentage of "correct" !votes indicates a user who might prefer to "play it safe" and just !vote in AfDs where the consensus is already clear to increase their "winning" percentage. So we don't need crats to discount such !votes in general, just those who don't use the stats to check whether the candidate is able to make policy- and guideline-based arguments. Regards SoWhy 17:52, 25 May 2017 (UTC)[reply]
• There's nothing inherently wrong about someone having a minority viewpoint and having something like a 60-70% accuracy rate at AFD. There are really two types of "bad" percentages: the first is simply voting delete on everything and being wrong sometimes. The other is what's mentioned above - looking at the situation, making an assessment, and occasionally being wrong. I think "bad AFD stats" isn't a good reason to oppose, whereas "bad AFD stats because 99% of the time they vote delete with no rationale" is a valid reason to oppose. Primefac (talk) 17:08, 25 May 2017 (UTC)[reply]
• The underlying problem is human nature; we are quite willing to look at objective measures and believe they give us wisdom. However. wisdom usually comes from the subjective. We have plenty of ways in which to objectively measure candidates. People refer to those rather than do an indepth review of a candidate's edits. So, if someone has <insert arbitrary % level> corrected votes at AfD, people conclude the candidate must know what they are doing. It's a false metric, but people latch onto it anyway, and there's really nothing we can do about it. --Hammersoft (talk) 18:06, 25 May 2017 (UTC)[reply]
• I agree that stats are unhelpful. There is a big difference between voting delete the day before a footballer has their first team debut, or voting delete on an article that is rescued with paywalled sources and voting delete on an article that could have been saved with a simple websearch. Raw statistics don't show you whether this is a person who makes sensible contributions in edge cases or one who usually votes per nom and occasionally does some deletion nominations that are hard to tell from vandalism. ϢereSpielChequers 20:20, 25 May 2017 (UTC)[reply]
• Another issue I don't think I've seen mentioned here is that AfD stats are particularly poor grounds to judge this candidate since a lot of the AfD's they !voted "Keep" were closed as "No consensus". That means that the end result of the discussion was to keep the article as ONUnicorn suggested but the statistics do not count that as "vote matched result". So, as lazy as the argument is in general, it is even less applicable here. Eggishorn (talk) (contrib) 20:28, 25 May 2017 (UTC)[reply]
• re the afd stat tool getting it wrong, as someone who adds afds to the various discussion lists, the tool sometimes (around 2% to 5%?) assumes i have 'voted' in line with the following editor's 'keep' or 'delete' vote, annoying/surprising but you get used to it:) Coolabahapple (talk) 02:24, 26 May 2017 (UTC)[reply]

  Coolabahapple, speaking as one of the maintainers of the tool, feel free to drop notes on my talk page about errors of this type, since we should definitely fix that :) Enterprisey (talk!) 02:35, 26 May 2017 (UTC)[reply]

  hi Enterprisey, no probs, it just looks like some funny aberration, oh wait, its just that it could ruin my almost(?) perfect record, oh no, from some of the comments above a very high 'success' rate means i won't ever become an admin, phew!Coolabahapple (talk) 02:50, 26 May 2017 (UTC)[reply]

Fads come and go - and so do they at RfA. This year's fashion is an obsessive concern with candidates' AfD stats as if that were all that matters for becoming an admin. ...having a number to cite is much easier than actually digging into the substance of an editor's contributions... (Ahecht). Exactly, and it also often precipitates a raft of unqualified pile-ons. Kudpung กุดผึ้ง (talk) 02:36, 26 May 2017 (UTC)[reply]

• My opinion - I hate to see raw statistics such as these used to make an arbitrary judgement on a candidate's suitability. Whenever I see a !vote based purely on AfD stats, length of tenure, number of articles created, number of posts at one or another Admin board, simple edit count etc. with no evidence that the voter has actually examined the candidate's contribution, I just scroll past that lazy vote, completely discounting it. Exemplo347 (talk) 07:06, 26 May 2017 (UTC)[reply]
• AFD stats can be useful if used correctly. If an editor is out of sync 30% of the time, it could indicate that they are not up to date with policies. But there are a number of other factors to consider before reaching that conclusion. If the majority of that 30% consists of well-reasoned arguments citing policies it's not a problem. Similarly if they were out of touch with consensus years ago but have sharpened up their game in recent years it may not be an issue. In all cases, people need to look beyond raw stats. Valenciano (talk) 07:29, 26 May 2017 (UTC)[reply]
• I was curious about this as well. If someone steps in with reasoned rationale in a contentious AfD that doesn't go their way, that shouldn't be harmful. If anything, this trend would probably discourage candidates from voicing their opinion on anything other than slam dunk AFDs. South Nashua (talk) 17:05, 26 May 2017 (UTC)[reply]

ONUnicorn was on the raw list at the Poll candidate search project. She was vetted then received the post suggesting she take the poll. The poll results were encouraging. She is now at RfA and doing well.

The Poll candidate search caused a spike in entries for April. More poll entries = more RfAs, right?

I am posting here To editors MusikAnimal, SoWhy, Enterprisey, Ritchie333, Mr. Stradivarius, Bri and Samwalton9: and anyone else who can make scripts and that sort of thing that produce names for the raw list.

For years people at this page have spent hours and thousands of keystrokes going on about not enough admins and possible solutions. Well, this might be working. I keep posting again and again with appeals for raw list names. I am mystified as to why there is so little help here.

Please feed the machine with names.

Anna Frodesiak (talk) 22:09, 26 May 2017 (UTC)[reply]

Stay mystified Anna. It's the same at WT:NPR, a lot of talk about problems but when the issues have been clearly identified no one want to take the initiative to implement the solutions, and so the keystrokes continue, but they are all banter, not programming language. Kudpung กุดผึ้ง (talk) 23:02, 26 May 2017 (UTC)[reply]

Hi Kudpung กุดผึ้ง. All I can say is: rats, dang, grrrrrrr, hmmpf, jeez, awwwww, and sigh. Anna Frodesiak (talk) 23:11, 26 May 2017 (UTC)[reply]

Maybe, instead of waiting for people to volunteer their abilities, I will have to knock on doors one-by-one. So... Anna Frodesiak (talk) 23:17, 26 May 2017 (UTC)[reply]

Who here knows of people who know how to populate the raw list? Name names! Thank you. Anna Frodesiak (talk) 23:16, 26 May 2017 (UTC)[reply]

@Anna Frodesiak: I think the big barrier is that "go forth and make a list" isn't something you can tell a script to do. We need parameters on what we want to be on the list. Once those parameters exist, then we can call on a coder to take up the task. Computers are stupid, so we have to be very exact about what we want anything automated to do. How about the following as a basis for a list?

1. Pull a list of everyone who has edited AIV, RFPP, AN, or ANI in the past month. (Feel free to suggest other centralized project-related pages or other means of generating some initial list that we can then reduce. Everyone who's edited in a time period is too broad. We could pull lists from user groups as an alternative to something based on individual pages.)
2. Remove from the last anyone with the sysop or bot flag.
3. Remove from the last anyone without the extendedconfirmed flag. (Note: Checking edit count is likely an expensive operation. That's best done by vetters.)
4. Remove from the list anyone who has been blocked in the past month.
5. Remove from the list anyone who's account was created less than two years ago. (Mostly as the "cheapest" substitute for checking edit count - a check on a single log entry is easier.)

This would pull a broad list of people. There would need to be substantial cuts in vetting. Is this what you're looking for? ~ Rob₁₃^Talk 01:33, 27 May 2017 (UTC)[reply]

As an aside, a smart move that doesn't require a script would be to add everyone who failed an RfA in 2015 or earlier to the list and vet for current activity/improvements upon their initial candidacy. I imagine some could be convinced to go again, but they may not decide to do that without prompting because failing an RfA can be brutal. ~ Rob₁₃^Talk 01:38, 27 May 2017 (UTC)[reply]

Checking edit count is actually fairly cheap, so long as it doesn't have to be exact; there's a column for it in the user table, and it's also available through the API if you really prefer that route.

My concern here is that any given thread on WT:RFA is at least 50% likely to be how horrible an idea it is to use statistics to evaluate candidates, and you're essentially asking us to do exactly that even before they're candidates. —Cryptic 01:58, 27 May 2017 (UTC)[reply]