Wikipedia talk:Requests for adminship/Optional RfA candidate poll/Archive 3

This is an archive of past discussions on Wikipedia:Requests for adminship. 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.

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

→

Archive 7

Possible page move

This is Wikipedia:Requests for adminship/Optional RfA candidate poll rather than Wikipedia:Optional RfA candidate poll and I think that makes that edit notice impossible to remove, right? I haven't looked into it much. If so, should this page be moved so that it isn't a subpage of Wikipedia:Requests for adminship? Anna Frodesiak (talk) 03:15, 24 October 2016 (UTC)

@Anna Frodesiak: Came upon this post while stalking your contributions. Are you referring to Template:Editnotices/Page/Wikipedia talk:Requests for adminship? If so, I'm not currently seeing the edit notice show up on my editing window for this page. I'm on a Mac and I've just tested this page through Chrome, Firefox, and Safari. In fact, I don't see any edit notice when I edit this page. Airplaneman ✈ 03:53, 24 October 2016 (UTC)

I don't see the applicable group edit notice, Template:Editnotices/Group/Wikipedia:Requests for adminship, as back in 2015, Template:Editnotices/Page/Wikipedia:Requests for adminship/Optional RfA candidate poll was created as a blank page (well, with a comment) to override the group notice. isaacl (talk) 03:57, 24 October 2016 (UTC)

Hi Airplaneman and isaacl. I'm mystified. I'm sure I saw it not two days ago. Sorry to have wasted your time. Anna Frodesiak (talk) 04:18, 24 October 2016 (UTC)

Poll transclusion

Does this poll need one of those RfA transclusion thingies so people can put it on their userpage to see if there's a new entry to check out? Is this fixing it when it's not broken? Anna Frodesiak (talk) 21:11, 4 November 2016 (UTC)

fixing it when it's not broken? Yes. --Izno (talk) 01:32, 5 November 2016 (UTC)

A new column in the poll results table and the timespan between ORCP and RfA

From my talk page:

Hi Anna, just a quick note - my addition to the ORCP stats is going to massively tip the averages. I note your advisement of this at the top of the page, but I wouldn't be opposed to removing me from the table or omitting me from any calculations -- samtar ^{talk or stalk} 21:07, 4 December 2016 (UTC)

Samtar! Congrats! :) :) :) I was just this minute adding a "Notes" column to the table and was about to add "ORCP Nov 2015" or something like that. Thoughts? Anna Frodesiak (talk) 21:09, 4 December 2016 (UTC)

Thanks Anna :) Yeah I think that'll help at least explain why I'm such an outlier! -- samtar ^{talk or stalk} 21:17, 4 December 2016 (UTC)

(talk page stalker) Perhaps the date of the poll closing should be shown in the table as well as the RfA date, to help gauge how much of each candidate’s ‘track record’ may have been established during that interval.—Odysseus 1479 21:26, 4 December 2016 (UTC)

Let's continue this at the poll talk page. Anna Frodesiak (talk) 21:36, 4 December 2016 (UTC)

Samtar, you're more than an outlier. You're an ORCP success story. In one year, from a lousy ORCP rating to an outstanding RfA result is great. You did what was recommended in the poll and it paid off. Kudos to you. Anna Frodesiak (talk) 22:00, 4 December 2016 (UTC)

So, where were we....

Anna Frodesiak (talk) 21:36, 4 December 2016 (UTC)

Yes. Odysseus, that's not a bad idea. Thoughts? Anna Frodesiak (talk) 21:37, 4 December 2016 (UTC)

Actually, maybe not dates, but rather add time between poll and RfA (and get rid of the notes column for now)? Anna Frodesiak (talk) 22:00, 4 December 2016 (UTC)

Users giving feedback

Folks, forgive me if I sound incredibly cynical for a moment, but I've been seeing a trend on this page that worries me a little bit. The trend is that of admin hopefuls, or seeming admin hopefuls, commenting a lot on ORCPs of other hopefuls; and my somewhat pessimistic reading of this is that some folks are seeking to bolster their own future candidacy by sounding old and wise here. Now the same thing occurs at RFA too, but in my view that's a) less of a problem because we honestly want to gauge everybody's trust in the candidate, and b) because the question being asked of people who comment is the relatively straightforward "do you trust the candidate." Here, on the other hand, the question is "do you think the community as a whole trusts the candidate," which in a way takes a lot more experience to answer. What I'm seeing here is a whole lot of people with a few thousand edits and some months of experience saying things that often sound like boilerplate. I'm honestly not sure whether this is an issue at all, and if it is what should/can be done about it, but I did want to bring it up. Regards, Vanamonde (talk) 08:41, 8 December 2016 (UTC)

I don't think it's an issue at all, and even if it were there's nothing we could do about it. That is, unless we elected a panel of judges who would be the only ones to vote. I trust the people who put themselves here for consideration are smart enough to sift out which people have opinions more worth considering and which are rather less so. --Hammersoft (talk) 15:11, 8 December 2016 (UTC)
I agree that it can be setting up inexperienced hopefuls for a fall. Nonetheless, as I mentioned previously, I believe being able to evaluate the credence of each comment is an essential ability for someone seeking administrative privileges. I appreciate it's a bit of a catch-22: those who aren't sure if they've adequately demonstrated their qualifications may also be less able to judge the context of the feedback. Unfortunately without putting more overhead in place, which I don't think anyone is eager to do, I don't have any suggestions to mitigate this issue. isaacl (talk) 04:23, 9 December 2016 (UTC)
I think the larger point (that I've noticed at AfC and NPP as well) is that we may in principle agree to having new users to join Wikipedia but we don't want them where we edit. We then install measures to chase the n00bs away which seems like a lot of work for experienced Wikipedians to spend time on. I support ladder-kicking behavior but I don't think it's worth our time to worry about. If a clueless editor seeks adminship and gets advice from both experienced editors as well as less-experienced echo chamber hopefuls and they foolishly listen to their fellow hopefuls, perhaps they deserve to get publicly clowned. Chris Troutman (talk) 05:27, 9 December 2016 (UTC)
Isaacl's comment is spot-on. Users who lack the experience to critically evaluate comments here lack the experience to administrate. However, since the objective here seems to be to encourage qualified candidates to run, and much less to discourage unqualified candidates, I both don't think this is a problem and don't think we can or should do anything about it. A separate but possibly related phenomenon here (purely my own observation) is that this informal polling tends to overestimate a candidate's chances, if anything. I can't back that up with stats at the moment, it's more of a gut feeling. Ivanvector (^Talk/_Edits) 13:06, 9 December 2016 (UTC)
This poll underestimates (slightly or perhaps even drastically) the chances a user will pass RFA in the general case. The results page is perhaps enlightening. --Izno (talk) 13:49, 9 December 2016 (UTC)
- You may be correct, but, as with all deliberations in Wikipedia, it depends on the turn out, on the number and quality of contributions. Digging into an editor's past and assessing its potential takes real work, and few are wiling or able to invest this type of energies on a regular basis. Perhaps something should be done to encourage/reward the volunteers who keep the poll honest and relevant. [User:Caballero1967|Caballero/Historiador]]^⎌ 14:10, 9 December 2016 (UTC)
- You know, looking at the poll results and the subsequent RfA results, it seems rather obvious that the poll mechanism does a pretty good job of predicting RfA passage. This process being more lightweight than RfA, yet accurate...perhaps we should dispense with RfA and replace it with this. Of course, do that and this becomes just as bloated and absurd as RfA. --Hammersoft (talk) 15:00, 9 December 2016 (UTC)
  A decent job at predicting passage for users who run RFA (and sample size may indicate that these results are not to be trusted). I think that should be emphasized, because I would suggest that a significant number of contributors are deterred from running by this process--much like how RFA deters users from running as well. The goal of this process is not one-fold as Ivanvector suggests, but two-fold: it also is indeed to deter contributors from running who may not be suitable. --Izno (talk) 15:09, 9 December 2016 (UTC)

Hammersoft, thanks for reminding us of the link to the poll results. It vindicates the exercise. And Izno, no doubt about it; this optional poll should have deterred a good number of potentially weak candidates from appearing unprepared at a RfA. [User:Caballero1967|Caballero/Historiador]]^⎌ 15:31, 9 December 2016 (UTC)

Like I said when the results page was set up: there are so many variables, including small sample size issues and external factors like the overall collaborative environment, that I don't think a statistical analysis is convincing. However I think a simple qualitative test is more than adequate: does this poll seem to be guiding editors towards the best outcome for English Wikipedia and themselves, whether that is attaining administrative privileges or continuing to work on tasks not requiring them? To date, this appears to be the case. isaacl (talk) 17:04, 9 December 2016 (UTC)

Izno, I wasn't so much interested in the stats; RfA by default now is pretty much a poor sample size. If we want to talk statistics, we'd need to do some research to see how many utter failure RfAs were attempted by people who shouldn't have done so both before this poll existed and after. I don't know of any data to suggest any deterrence or lack thereof. Caballero1967, Someone else linked it above, I was just the 2nd, but thanks :) --Hammersoft (talk) 15:36, 9 December 2016 (UTC)

I posted some truncated stats on that at WT:RFA not too long ago, just 2015 vs. 2016, and not really considering ORCP's impact on results (didn't check whether candidates stood at ORCP, etc). There were 19 RfAs closed as NOTNOW or SNOW in 2015, 9 in 2016 (32.8% of 58 vs. 32.1% of 28). In absolute numbers there were fewer candidacies from clearly unsuitable candidates, however in terms of the proportion of overall RfAs the numbers are exactly the same. I don't think it can be said from stats whether there were fewer clearly unsuitable candidacies in 2016 because ORCP dissuaded candidates, or just because of the continuing downward trend in RfA candidacies overall. Ivanvector (^Talk/_Edits) 19:28, 9 December 2016 (UTC)

This is pre-ORCP, but complementary to Ivanvector's observations: I looked at RfA closures from 2008 to mid-2015 (when I compiled the data) and found that the proportion of total RfA candidacies that are closed as NOTNOW/SNOW has been surprisingly stable over that period, despite the dramatic and monotonic decline in absolute numbers of candidacies. This is a really underappreciated fact about the overall decline in RfAs - it appears likely that a) nothing anyone had tried before mid-2015 to reduce the number of no-hope candidacies has had any meaningful effect, and b) whatever process is causing a decline in candidacies is affecting no-hope candidates just as much as good ones. Opabinia regalis (talk) 20:16, 9 December 2016 (UTC)

*Great job with those numbers, Ivanvector and Opabinia regalis. They are forcing me to qualify, better yet, amend my thoughts about the survey numbers, which Izno and not Hammersoft , rummaged for us. We need more data to confirm the effectiveness of the alternative poll. But Opabinia regalis' nifty graphics bring out the sharp downturn in sysop candidates. They are there, in full colors, and honestly, it is frightful. We don't need Einstein's IQ to predict the horrid WP future with a dwindling army of admins facing a spike in data, users, and of course, more problems to fix. Even if we automatize even more, the shortage will pinge us hard at one time or another. For me, this is a greater problem, and if the alternative polling can be turned into a "school" rather than just a poll, we might just contribute, even when just a bit, to the stabilization of those plunging numbers. [User:Caballero1967|Caballero/Historiador]]^⎌ 21:30, 9 December 2016 (UTC)

Unlikely. My vast powers in statistical analysis (which, ironically, leads me to refuse to posts my research ;) ) shows that Wikipedia has been in decline for a very long time. I've written about this several times before (and not just there). We will not be averting the decline witnessed by this chart. That's very much a long term trend that small and even medium scale solutions will not avert. You might as well stop a dam break with a cup. We will occasionally see dead cat bounces, but the overall trend is rather permanent now. We're now down to 523 active administrators, and slowly but surely that number continues to decline. The question isn't how do we stop it. The question needs to be what do we do in response to it. People aren't asking that question yet, but it must be asked. How do we evolve into a post-administrator era? That's what we're facing. --Hammersoft (talk) 21:46, 9 December 2016 (UTC)

Distribution of registration dates of active admins.

Distribution of time period from registration to successful RfA, by year of RfA.

Caballero, there are some caveats to that data that make the apparent problems a bit less urgent. I only looked back to 2008 in part because there were formatting changes that made it harder to parse the older RfAs, but also because rollback was unbundled from the admin toolbox around then, and is widely believed to have reduced the demand for admins who specialized in vandal-fighting. I don't think we have a particularly clear idea of how many admins we "need" to sustain critical functions - people like to point to backlogs, but many of those are artificially self-inflicted.

A bigger problem - and more to Hammersoft's point - is the growing wiki-generation gap that is increasingly dividing the administrative class from the "ordinary" users. On the one hand, the numbers of new and active accounts have been fairly stable lately - but the path from newbie to established user to viable admin candidate is broken. As a rule, active admins originally registered many years ago; all of seven currently active admin accounts were registered in 2013 or later, and three of those are bots. This is, obviously, completely unsustainable. Opabinia regalis (talk) 00:57, 10 December 2016 (UTC)

Opabinia regalis, are you pointing to the cause of the trend that Hammersoft studies? You note a fissure in a generational continuum that before has ensured a steady influx of new admins. Could the problem be related to the way that the RfAs have been conducted lately? Is it actually unrepairable? You seem to think that if the trend continues conditions would become unsustainable. [User:Caballero1967|Caballero/Historiador]]^⎌ 19:48, 10 December 2016 (UTC)

Yes indeed Opabinia regalis, This is, obviously, completely unsustainable, but there are some things that can't be solved with pure math. One of them is the silly behaviour at RfA which for some odd reason is almost fully tolerated by the community with total impunity (IIRC only one person was ever sanctioned for their behaviour at RfA). They talk about it a lot of course, but they carefully avoid doing anything about it. Anyone who has spent a lot of time recruiting new candidates knows that this is the main reason given for not running for office. It doesn't explain why troll RfA follow the same graph. Also however, a lot of troll RfA don't figure in the stats because nowadays we simply delete the transcluded pages before voting stats.

Adminship has become an even bigger deal today due to the rarity of RfA. The voter profile has changed significantly and especially since Jan 2016. To be of any use in an argument, the stats that need to be extracted and summarizes are the ones done in 2011 at Voter_profiles and adding an extra column for their accuracy of prediction based on {{U|Scottywong}}'s tool. I'd do it myself but I don't have a clue how to do database searches. Scott has retired, and nobody else, despite many, many calls for help, will do it. Maybe the community is scared of opening a can of worms. Perhaps Ivanvector or Opabinia (whom I have often stated is a whizz-kid at this kind of thing) can have a go. Maybe even MusikAnimal could help too, though this particular topic area may not be up his alley and he also has a very heavy workload on other important development stuff.

Such a table/stats would also reveal who the serial opposers are (we know this already but to mention them by name in a text gets warped complaints about PA, jus like it does anywhere when anyone mentions a misplaced comment by a user). Hammersoft asks 'How can we stop it?' The answer is blatantly clear, to repeat my worn out 6-year old mantra: Fix the voters and RfA will fix itself.Kudpung กุดผึ้ง (talk) 03:17, 11 December 2016 (UTC)

I'm good with Excel but useless at scripting. I already started adding to my data for past years, when I'm done with that I can post it somewhere so someone with more skill can do something exciting with it. Depending on what kind of data we could get out of Scottywong's tool I could probably do something interesting with a vlookup, but I'll have to look into it. Ivanvector (^Talk/_Edits) 13:17, 11 December 2016 (UTC)

Just for clarity, has this thread, "Users giving feedback" relating to ORCP, been hijacked to analyse RfA in the wider sense (yet again and with another thread already active elsewhere for 3 weeks)? Leaky Caldron 20:01, 11 December 2016 (UTC)

It emerged (here) out of concerns about potential candidates (rather than experienced editors) offering advice to other potential candidates in the poll. [User:Caballero1967|Caballero/Historiador]]^⎌ 21:35, 11 December 2016 (UTC)

It's reasonable to wonder whether the poll is having any deterrent effect on premature candidacies; that was the prompt for the rest of the stats. (The answer to that question, as best anyone's looked, seems to be "probably not, but that's fine, because nothing else ever has either".) But mostly those are my fault; I should really post that stuff at WT:RFA at some point, though I rather like that the poll talk page is much more practical and solutions-oriented. Opabinia regalis (talk) 03:25, 12 December 2016 (UTC)

Improving the results table

I suggest:

Name
ORCP info including end date
RfA stuff

...because ORCP is chronologically first, and the timespan between ORCP is relevant, it now seems. Thoughts?

Convenience link: Wikipedia:Requests for adminship/Optional RfA candidate poll/Results

Anna Frodesiak (talk) 23:14, 9 December 2016 (UTC)

Actually, I'm still trying to figure out how to do that so it doesn't look odd. Let's put this on hold for now. Anna Frodesiak (talk) 01:48, 11 December 2016 (UTC)

Done - revert if you don't like it. :) Anna Frodesiak (talk) 23:27, 13 December 2016 (UTC)

Poll removal

Regarding this edit restoring a removed poll: although my personal preference would be to archive polls, I don't feel strongly about someone just removing their poll completely if they wish to withdraw. Should we copy the poll to the archive, or just leave it deleted? isaacl (talk) 18:19, 11 December 2016 (UTC)

It should be archived. Some may believe that a discussion not of their liking should be vanished. Wikipedia, more wisely, archives. Chris Troutman (talk) 20:20, 11 December 2016 (UTC)

Obviously without revision deletion there was no issue of having the discussion vanished. As a courtesy I have archived the poll manually. isaacl (talk) 22:02, 11 December 2016 (UTC)

It should be archived, I agree. It was not trolling, so there is no DENY. It is important for the record. Send it immediately to the archives as a courtesy. Anna Frodesiak (talk) 21:30, 11 December 2016 (UTC)

I suggested to that editor that he may wish to withdraw his poll. My bad for not telling him how to do it. Kudpung กุดผึ้ง (talk) 01:04, 12 December 2016 (UTC)

Fair enough. I'm in favour of keeping it in the archives unless he strongly objects or others think it is best to remove it. Anna Frodesiak (talk) 04:14, 12 December 2016 (UTC)

Just archive it. Kudpung กุดผึ้ง (talk) 12:08, 12 December 2016 (UTC)

Archive by default. However if a user specifically asks for their poll to be blanked we should honour the request. It's in the page history anyway. Ivanvector (^Talk/_Edits) 12:58, 12 December 2016 (UTC)

Results table dates

I just realized that the start and end dates for the poll and subsequent RfA are a bit of a mess. A plan? Anna Frodesiak (talk) 02:06, 20 December 2016 (UTC)

Okay, I fixed all the dates to be poll end and RfA start.

Now, the table is sorted by RfA. Should it be sorted by poll end dates?

Anna Frodesiak (talk) 00:01, 22 December 2016 (UTC)

Okay, I sorted the entries by poll end dates. Anna Frodesiak (talk) 06:13, 24 December 2016 (UTC)

Advice for RfA candidates

Regarding this edit made to WP:Advice for RfA candidates: I have started a conversation at Wikipedia talk:Advice for RfA candidates/Archive 1#Optional candidate poll. Any feedback is welcome. isaacl (talk) 02:46, 20 December 2016 (UTC)

Initiating a poll

Regarding this edit: personally, I believe only the candidate should initiate a poll. The key factor is that candidates must be receptive to hearing candid feedback on their chances at receiving administrative privileges. I feel this should be done on their initiative, not that of others. isaacl (talk) 00:33, 20 December 2016 (UTC)

I agree. I see very little gained in allowing someone else to start an ORCP poll for someone else. Mkdw ^talk 01:13, 20 December 2016 (UTC)

So what to do? In this case, should we leave it considering there was no objection and they were likely in agreement or are friends?

Should we add something to the instructions page about it or avoid instruction creep and just remove future ones on sight and post at usertalks? This is pretty rare, after all. Anna Frodesiak (talk) 01:33, 20 December 2016 (UTC)

I agree this is bad practice. I support a norm that we revert these third-party nominations. If we do that, however, we'd also need to add something to the instructions page as we'd create hurt feelings with editors that honestly didn't know better. It's instruction creep but without instructions we have situations like this. Chris Troutman (talk) 01:43, 20 December 2016 (UTC)

It would be very easy to simply change the wording in instructions section to clearly state, "the candidate should add their name below". Anyone other than the candidate doing so will have failed the instructions and thus a procedural close could be immediately be implemented. Mkdw ^talk 01:55, 20 December 2016 (UTC)

Note that the instructions were amended earlier today. I for one don't see any problems in others nominating a suitable editor, but that will definitely have to have been consented to. Looking at the big picture, we need more admins, and if somebody is too shy to throw their hat into the ring, then there won't be an RfA. If somebody else asks and gets permission to ask on their behalf, we may get an additional admin out of it. So what's the harm? Schwede 66 01:56, 20 December 2016 (UTC)

How does this relate to ORCP? If someone is "too shy to throw their hat into the ring" then there CAN be an RFA because RFA still allows third party nominations. We're talking about third party nominations at ORCP which is neither a precursor nor a mandatory prerequisite to RFA. There's a difference between 'nominating' someone for an RFA, and simply unilaterally putting someone up for a public opinion poll. Mkdw ^talk 02:09, 20 December 2016 (UTC)

If the potential candidates are too shy, the encouraging editor should speak to them and build up their confidence. There's no upside in trying to surprise them with a candidate poll or a request for adminship privileges. The serendipitous surprise should the trust shown by support for an RfA. isaacl (talk) 02:15, 20 December 2016 (UTC)

"There's no upside in trying to surprise them with a candidate poll" - I fully agree. Schwede 66 02:39, 20 December 2016 (UTC)

Additionally, if we start allowing nominations here at the poll, then we're likely to see the same thing that happened at RFA whereby self noms attract opposition; it may have the exact opposite affect of attracting new candidates because they need to find someone to nominate them for just the poll. Mkdw ^talk 02:17, 20 December 2016 (UTC)

Given that the editors commenting at ORCP are all very civil and mature, I really do not see that risk arising. Schwede 66 02:39, 20 December 2016 (UTC)

For this instance, since the candidate has agreed to continue the poll, it can be left alone. The instructions already use "you" throughout; unless a trend emerges, I don't feel there is a need to amend them. isaacl (talk) 02:20, 20 December 2016 (UTC)

The instructions as they stand now state, "Users should not list other editors unless they have been given permission to do so." Mkdw ^talk 03:26, 20 December 2016 (UTC)

Apologies for being unclear; I see no need to amend them from their state prior to the edit under discussion. isaacl (talk) 03:41, 20 December 2016 (UTC)

Based on the discussion, I propose reverting the edit under discussion. If there are any further comments, please reply. isaacl (talk) 16:53, 22 December 2016 (UTC)

As there are no further comments, I plan to proceed with the proposal to revert the specified edit. isaacl (talk) 17:09, 27 December 2016 (UTC)

Without that sentence nothing on the page directs users specifically on whether they can or should list others. While it's true that the 2nd and 3rd paragraphs use "you", I don't think it's entirely obvious that you shouldn't start a poll for another user without that addition. I'm all for rewording it to more explicitly say that you shouldn't list another user at all, but I think a specific sentence makes it less ambiguous either way. Sam Walton (talk) 17:14, 27 December 2016 (UTC)

If there seems to be a trend of people nominating others, then we can revisit the need for further instructions. Honestly, it doesn't make a lot of sense for editor A to use this procedure to start a poll on editor B's suitability for receiving administrative privileges. (If Wikipedia:WikiProject Admin Nominators were to be revived, it might make sense in that context. However, the participants in that WikiProject were reluctant to put candid discussion of potential candidates in Wikipedia, out of a concern of discouraging those who weren't selected to be nominated.) isaacl (talk) 17:25, 27 December 2016 (UTC)

Criteria discussion

Perhaps the extended discussion in Primefac's poll can be moved to this talk page, or elsewhere? There are lots of better venues than within a potential candidate's poll. isaacl (talk) 15:45, 2 January 2017 (UTC)

Personally I think that discussion is very relevant to my thoughts on whether/how to proceed with an RfA. Plus, it's not completely unrelated to the original post. Primefac (talk) 23:09, 2 January 2017 (UTC)

Isaac, personally, I do not find the discussion too long at all and the venue is perfect for it. We need to understand that this is a very informal but albeit (in most cases) a very useful project. Kudpung กุดผึ้ง (talk) 04:49, 3 January 2017 (UTC)

As you know, the original intent of this poll was to allow people to give a rating and a quick comment based on their own judgment. What makes the middle of a potential candidate's poll a better place to ask for someone to gather statistics on RfA candidates than Wikipedia talk:Requests for adminship or Wikipedia talk:Administrators? isaacl (talk) 02:09, 13 January 2017 (UTC)

P.S. Note you've pinged someone else in your comment. isaacl (talk) 02:19, 13 January 2017 (UTC)

Do we need the invisible comments in the preload?

I was looking at the most recent few polls, and there are sometimes large vertical gaps above and below the list of comments. These are caused by the invisible comments present in the preload comment. I think these instructions are redundant to the ones at the top. If we still want reminders in the editing window, we can just add an editnotice. What do you all think? Enterprisey (talk!) 01:50, 13 January 2017 (UTC)

Using an edit notice is another approach, though I'm uncertain of their effectiveness, which is why I opted to include the instructions directly within the preloaded text, rather than request an admin or template editor to create an edit notice. There have been numerous comments on this discussion page about making sure commenters know that the rating is an evaluation of the candidate's probability of receiving administrative privileges, and is not a personal rating. That being said, although I'm slightly more hopeful that inline comments get read, I know there are many who'll ignore them as well. However given that extra white space isn't a big issue, personally I prefer to have the inline instructions. isaacl (talk) 02:16, 13 January 2017 (UTC)

It's always a problem trying to get Wikipedians to read instructions wherever they are. It's a shame that some users use ORCP just to be as spiteful as they are on RfAs where they vote. Sometimes a somewhat longer post is necessary, but as long as it comes with sensible advice, that's fine. But to simply say 'You have my support' is not helpful, while anything approaching 'I'm gonna get you, if I have to dig up every bit of dirt I can find on you' are little short of harassment and such users eed to be asked politely to stay away from ORCP. Yes, the invisible instructions need to stay. Kudpung กุดผึ้ง (talk) 04:24, 22 January 2017 (UTC)

I agree completely that the people giving responses to the polls need to read the instructions. However, I don't think the instructions need to be in form of invisible comments that make the layout of the page look a bit weird. Look at ANI: there aren't any invisible comments, just a series of large, brightly-colored edit notices. Enterprisey (talk!) 04:52, 22 January 2017 (UTC)

Invisible statement

The editor added it in this, his first and only edit. Why do we think he was trying to remove it? Why are we keeping that invisible code? Why not remove the invisible code and reveal the statement? Anna Frodesiak (talk) 07:46, 20 February 2017 (UTC)

I thought it's useful to show the editor's statement (edit summary: "we should at least unhide the application so that it makes sense when it gets archived") but my edit was reverted ("I disagree. The editor doesn't recognize html and there"). In its hidden form, the item doesn't make sense. If there is a need to demonstrate the editor's failings (even though it's a concept that I struggle with), then maybe we show the statement and add a note that it was us who made it visible. Schwede 66 08:45, 20 February 2017 (UTC)

Indeed. I suggest we remove the invisible marks. Anna Frodesiak (talk) 09:08, 20 February 2017 (UTC)

No. That is as much a part of the editor's statement as is the other words typed in. I don't want to edit war over the point, but leaving the entry as the editor made it is the spirit of WP:REFACTOR. Your suggestion to unhide the comment changes how other editors see the comment, which I think is wrong. Chris Troutman (talk) 14:35, 20 February 2017 (UTC)

Nope as per Chris - The editor either didn't realize it was hidden or chose to hide it - Either way they chose to do nothing about it and they had ample oppertunity to do something, personally I think it should be left so infuture when they ever retry for RFA editors can look at it and decide for themselves (I don't mean it as in to shame them but as I said they had plenty of time to remove it). –Davey2010^Talk 16:34, 20 February 2017 (UTC)

You make good points. I didn't think of it quite that way. Okay, fair enough. Let's leave it invisible. I just posted this: User talk:Sanjev Rajaram#Your Optional RfA candidate poll entry. If he speaks up in favour of visible, we can change it then. Best, Anna Frodesiak (talk) 21:28, 20 February 2017 (UTC)

It appears our friend's intention was to have it visible, but had trouble making it appear. :) Pinging User:Sanjev Rajaram so he can see this post. Anna Frodesiak (talk) 10:51, 21 February 2017 (UTC)

Thanks Anna your too kind :)

And to the rest Honestly I had no idea why it was invisible but being that my self-nomination was already deemed unpassable at this point in time I'd say it's already a mote point unless making it visible will somehow improve my chances. --Have a great day , Sanjev Rajaram (talk) 16:11, 23 February 2017 (UTC)

User:Ruigeroeland

I would like to suggest User:Ruigeroeland.

He's the most active person in our Lepidoptera-project (for years). Has a good oversight, stays calm & is friendly (that's not the case of all Admins-here !). Regards

I'm so tired (talk) 08:24, 29 March 2017 (UTC) (with ping to: Ruigeroeland (now I pinged you - do the rest yourself, Ruigeroeland) )

Close?

So am I expected to close this? It doesn't look like I'm winning any noms, and I'm not going to self nom. I'm also patently not going to intentionally change my editing habits just to satisfy a fairly shallow criteria that doesn't seem to put a lot of value into the nature of the content. I don't know how long these normally run, but I'm not sure this is really going anywhere. TimothyJosephWood 21:47, 5 April 2017 (UTC)

These items are generally closed by one of the active RfA participants. They often stay open for a month, which may seem a lot longer than what happens elsewhere on WP. Schwede 66 22:19, 5 April 2017 (UTC)

I've always felt a little undecided on closes. Obviously the subject of a poll should be able to close their poll, but should we arbitrarily close other users' polls based on some notion that a consensus has formed? Sam Walton (talk) 22:29, 5 April 2017 (UTC)

I don’t think consensus per se has much relevance here; rather the basis for the decision should be whether or not the querent would likely receive any more useful feedback were the poll to remain open.—Odysseus 1479 23:13, 5 April 2017 (UTC)

Would you like me to close it @Timothyjosephwood:? Anarchyte (work | talk) 23:48, 5 April 2017 (UTC)

I'm fine with it. I'm fine with reopening if someone wants to go past "here's my evaluation your last 50 edits," but if not it's pretty apparently not terribly helpful. TimothyJosephWood 00:06, 6 April 2017 (UTC)

Honestly, I expected people to dig into my history and give a fairly broad approximation of myself as a user, but I've not found that at all. In fact, I've not found any really critical appraisal of my contributions whatsoever. TimothyJosephWood 00:22, 6 April 2017 (UTC)

Alright, I'll close it the moment I've got some time. Also, if you'd like a better indication, seek out specific editors that opinions you'd like and say exactly what you'd like to know. Anarchyte (work | talk) 00:39, 6 April 2017 (UTC)

It's already closed? TimothyJosephWood 00:45, 6 April 2017 (UTC)

ah, didn't see that. It got closed in the same minute I wrote the first message. Anarchyte (work | talk) 01:41, 6 April 2017 (UTC)

There's no set time for a poll to run; the bot archives a poll after 14 days without comments. isaacl (talk) 02:27, 6 April 2017 (UTC)

Finding qualified admin candidates

Another idea is some sort of algorithm that finds users with current activity, plenty of edits, participation at certain boards, and a clean block log. The generated list of editors could then receive talk posts suggesting they consider using the poll. Thoughts? Anna Frodesiak (talk) 04:07, 6 April 2017 (UTC)

Ha! I like the addition of 'a clean block log' (obviously, it stings, deservedly so, but it stings so!) :D — O Fortuna^{velut luna...} 04:19, 6 April 2017 (UTC)

I like this idea. I know that a lot of experienced folks, such as Samwalton9, have been doing something similar manually, and that resulted in the surge of nominations this January. So long as we're not assuming that every suitable candidate will fit every one of these criteria, I think this might be really helpful. Vanamonde (talk) 05:26, 6 April 2017 (UTC)

Does https://tools.wmflabs.org/apersonbot/aadminscore/ fit the bill? Apparently I score 1,387.2 - no idea if that's good, bad or indifferent. Ritchie333 ^(talk) ^(cont) 08:16, 6 April 2017 (UTC)

Good thinking. That might work. Enterprisey made it, I think. Maybe he can suggest a way for it to generate a list. Anna Frodesiak (talk) 08:26, 6 April 2017 (UTC)

There are a couple of other features that could be added to the score, the obvious ones being AfD score (if we could get the tool to return JSON that could be implemented the same way created pages are at the moment), CSD logs (check red versus blue), PROD logs (same) and edits to AIV, RPP and UAA. If I ever get time (hah!) I might pull the source code and see what I can do. Ritchie333 ^(talk) ^(cont) 08:30, 6 April 2017 (UTC)

The tool really needs to explain the significance of the score. I just got 1393.8 but is that a pass, fail or meh? I already am an admin. Roger (Dodger67) (talk) 08:43, 6 April 2017 (UTC)

As far as I can tell, according to the code :

Edit count : (Log(e) of edits * 71.513 - 621.0874) * 1.25, unless the account has less than 350 edits, in which case -250

Block log : 140 for never blocked, previously blocked is (0.1977 * ( days since last block - 92.3255)) - ( 10 * number of blocks), up to a maxmimum 100, -100 is blocked now, -500 is indeffed now

Account age : (Log(e) of days account has been active * 91.482 - 544.85) * 1.25, unless the account is less than 43 days old, in which case -250

User page : -50 if red, -10 if redirect, 10 if live

User rights : Add 500 for admin, otherwise add 25 for each of abusefilter, checkuser, add 20 for each of autoreviewer and templateeditor, add 15 for filemover, add 10 for accountcreator, add 5 for each of reviewer, rollbacker

Pages created : log(e) of mainspace creations (article and redirect, doesn't matter) * 36.07161 - 68.8246, all multiplied by 1.4, down to a minimum of -140

Activity : Log(e) of average edits per month * 30.41375 - 138.48563, all multipled by 0.9 down to a minimum of -49.1

I don't know how the numbers were worked out, maybe just trial and error. Ritchie333 ^(talk) ^(cont) 09:00, 6 April 2017 (UTC)

Interesting tool. As far as I can tell, basically any score higher than 500 indicates a good candidate since checking any sysop will automatically add 500 points. This corresponds with the fact that Enterprisey wrote that his tool is an adaption of Scottywong's tool and that tool computes scores up to 1000 (but doesn't work on admins) and says anything above 500 is good. Regards So Why 09:13, 6 April 2017 (UTC)

Everyone's favourite admin candidate^{[citation needed]}, Cullen328, scores 779 Ritchie333 ^(talk) ^(cont) 09:17, 6 April 2017 (UTC)

Enterprisey got 577.4 Points in his own tool. Maybe this is a sign that you should nominate them?

Regards So Why 09:42, 6 April 2017 (UTC)

SoWhy, if you look in the original source code, there are a bunch of (usually linear) interpolation tables with a bunch of statements like "edit count X is guaranteed subscore Y", and so on. I fed the tables into Excel, made it generate formulas, and put them directly into the code. This way, my tool and Scottywong's tool should have the same subscores for each one that they share. Enterprisey (talk!) 05:27, 7 April 2017 (UTC)

@Enterprisey: I appreciate that your tool doesn't attribute edit counts on a linear scale, using the natural logarithm instead. Good thinking :) A suggestion; the tool gives me credit for 179.14 points for having created 234 article space pages. Yet, most of those are redirects. Is there anyway to differentiate redirects and actual articles? Also, I agree with Ritchie333's suggestions of somehow including the editor's history at AfD, CSD, AIV, UAA, PROD, and RPP. I know, I know, feature bloat :) Still, that would give a better picture, and would remove undue weight of a particular category. --Hammersoft (talk) 14:35, 7 April 2017 (UTC)

CSD added! The other ones are definitely harder to quantify (except PROD, I guess), but I'm working on some sort of API for that AfD tool. As an alternative to AfD, I think XfD might cover more areas, but it's in a bit of a beta state now. Enterprisey (talk!) 04:17, 8 April 2017 (UTC)

@Enterprisey: CSD logs are a little fickle - you may wish to exclude the latest 7 days of data, as for those of us who do a lot of image tagging many of the links should be blue. (I also archive mine, but I don't expect you to programatically detect that!) – Train2104 (t • c) 04:54, 10 April 2017 (UTC)

I've long wished for something like this. When I'm looking for admin candidates it's usually these kinds of stats that are my first stop. An automatically generated list would be super helpful. Sam Walton (talk) 08:49, 6 April 2017 (UTC)

The pages created by the user doesn't ever load for me when checking anyone. Anarchyte (work | talk) 09:29, 6 April 2017 (UTC)

It takes a while. If they created many pages, it takes a loooong while. Regards So Why 09:36, 6 April 2017 (UTC)

I had that problem too. I gave up. Also, I got +500 for already being an admin. That could be fixed. :) Anna Frodesiak (talk) 09:42, 6 April 2017 (UTC)

I left it open on my page (I've only made like 51 articles) for around 2 hours and it never updated. Sure it's not broken? @Enterprisey: Anarchyte (work | talk) 12:17, 6 April 2017 (UTC)

Try the API query directly. That's really slow for me at the moment. Could you try the API that is my suggested fix and tell us what the difference is? Ritchie333 ^(talk) ^(cont) 12:30, 6 April 2017 (UTC)

Yes...subtract 500 ;). Lovely tool btw, @Enterprisey:. Lectonar (talk) 11:02, 6 April 2017 (UTC)

The page creation score is using the API to get a list of all new pages the user has created and throwing them in the bin, except for the count. That's a waste. @Enterprisey:, try changing the ucprop=title in line 180 of aadminscore.js to just ucprop= and that should speed it up. Ritchie333 ^(talk) ^(cont) 09:51, 6 April 2017 (UTC)

Done. Thank you for the suggestion! Enterprisey (talk!) 05:29, 7 April 2017 (UTC)

MusikAnimal and Mr. Stradivarius are also good at this sort of thing. Maybe we can persuade them to get it to generate a list of the top 100 or something. I can then get a snazzy new ad made up and post at their talks. It just might work and we could get a bunch of new admins and have a new well to go to. :) Anna Frodesiak (talk) 09:42, 6 April 2017 (UTC)

My initial thoughts are that this would probably work best as a tool on Tool Labs, and that it would take a non-trivial amount of time to do it well. For example, the wait time that people experience could be dramatically reduced by batch-processing people's scores and caching the results, but then deciding how to store the (millions of?) scores and how to update the cache requires some careful designing. In other words, this is something that I would be interested in doing in the future, but probably don't have time for right now. — Mr. Stradivarius ^{♪ talk ♪} 13:39, 6 April 2017 (UTC)

Just a one-time list of users meeting some specified requirements in terms of easy to find things like edit count, recent activity, time since they became active, a clean block log, not an admin already, would be useful! This needn't be so complicated as to require generating admin scores for every user. Sam Walton (talk) 14:08, 6 April 2017 (UTC)

Yes, something like that, Wikipedia:List of Wikipedians by number of edits. The query only needs to be done once, then the cached result can be served up to everybody. It only needs to be updated about once a week, if that. An offline script could slowly gather the data, process into a list, and then get a bot (ThreesieBot?) to upload it. Ritchie333 ^(talk) ^(cont) 14:24, 6 April 2017 (UTC)

@Samwalton9: In that case, I might be able to put something together using a simple SQL query on Quarry, which would be a lot less work than making a standalone tool. How would you define "recent activity" and "becoming active"? To make those things into an SQL query we will need to put the definition in terms of numbers of edits and period(s) of time. — Mr. Stradivarius ^{♪ talk ♪} 22:26, 6 April 2017 (UTC)

@Mr. Stradivarius: Lets say they've made 10,000+ edits, have been editing for 2+ years, have made more than 100 edits in the past month and more than 1000 in the past year, are not currently an administrator, don't have a block from the past 3 years? It would also be nice to look at AfD % and edits to AIV/RFPP/etc, but I don't think we can do that with Quarry. Sam Walton (talk) 22:56, 6 April 2017 (UTC)

@Samwalton9: Might want to be careful with the "don't have a block" criterion as some folks cough cough have accidental blocks on their record. EvergreenFir (talk) 23:58, 9 April 2017 (UTC)

@EvergreenFir: Ha, indeed. But this is meant to be a way of collecting 'many probably qualified candidates' rather than 'all qualified candidates', and the extra possibilities from accidental blocks are likely outweighed by the extra effort that would be needed to go through genuinely previously blocked users on the list. Sam Walton (talk) 00:03, 10 April 2017 (UTC)

@Samwalton9: I certainly understand. Perhaps block length can be factored into an algorithm as, presumably, mistaken blocks would be reversed quickly? That might be a bit of work though. EvergreenFir (talk) 00:05, 10 April 2017 (UTC)

@EvergreenFir: Good idea - definitely something worth considering if possible. Even if it was a genuine block that was reverted after a few hours for some reason it probably wouldn't be so bad as to sink an RfA. Sam Walton (talk) 00:08, 10 April 2017 (UTC)

@Samwalton9 and EvergreenFir: I'd definitely chime in with EF's suggestion, as somebody with an edit-warring block (which was not entirely undeserved) who passed an RFA last year. I also know of at least one successful admin who had two blocks before they passed RFA, both not accidental. I think the duration of a block might be a good parameter to filter by; time since last block might also be good, though I don't know if it's technically difficult to do. Vanamonde (talk) 05:25, 10 April 2017 (UTC)

I've manually created a list of high quality candidates using the tool for screening, at User:Bri/Admin candidates. It is a work in progress. - Bri (talk) 04:03, 7 April 2017 (UTC)

Hi Bri. It looks great. I've taken the liberty of adding a section below it with AfD links. I hope that's okay. A copy of the same is at User:Anna Frodesiak/Pink sandbox‎. Anna Frodesiak (talk) 06:27, 7 April 2017 (UTC)

I think we could generate a better list of potential candidates. The above suggestions are great, but are missing an important element; who are the non-admin people who are most active in areas where admins frequently work? Looking at it from this perspective, I looked at editors with the most contributions to WP:AIV (see [1]) and instantly, we see User:Cahk (not pinging him as I'm just using him as an example, and haven't actually vetted him for adminship anyway). He's been here over 10 years, never blocked, 29 non-redirect article creations, 44k edits, a TON of CSD noms a lot of which appear to be accurate, etc. Enterprisey's tool scores him at 809. For comparison, the average admin promoted this year scored ~700. Here's another one, found by looking at the most active contributors at WP:UAA; User:Drm310; He's been here over 10 years too, never blocked, 11 articles created, almost 40k edits, a solid CSD log, and a score of 779. There may be quite a few more people like this that can be readily found by targeting searches on areas where admins do the most work, and where promoting people would have the most bang. Finding someone via the wide net approach might find people who score well, but really wouldn't have much use for the tools. Just a thought. --Hammersoft (talk) 15:20, 7 April 2017 (UTC)

Good points, Hammersoft. Maybe this is naturally turning into a group effort where we keep a subpage of editors to approach with a post suggesting they use the poll. The entries could be added from those who have been watching good potential admins, and also from those who generate lists from some query. Some sort of further vetting or system could move those names to a "ready to approach" section. Others could then post about the poll or even start a dialogue. Anna Frodesiak (talk) 17:06, 7 April 2017 (UTC)

I agree with Hammersoft; I checked my own score out of curiosity; it's 830.9, but in fact I have almost no involvement in areas where admin tools are needed, and if I were to be nominated that would no doubt cause some opposition. Starting with users who are active in admin-related areas is a much better way to go. Mike Christie (talk - contribs - library) 17:48, 7 April 2017 (UTC)

So what about some script that looks for lots of posts in those areas? Anna Frodesiak (talk) 17:53, 7 April 2017 (UTC)

We wouldn't dream of letting this fizzle out, would we? Anna Frodesiak (talk) 05:44, 9 April 2017 (UTC)

I am late to see this and my question will not matter in most cases and may not in the few others. I noticed in the discussion above that the score for never being blocked is 140. The most that someone who has been blocked can get back to over time is 100. There have been a few candidates who have had one erroneous block. Can the tool screen for that or is someone who is erroneously blocked for perhaps a few minutes consigned to a permanently lower score? Or does it not make a big enough difference to matter? Donner60 (talk) 08:25, 23 April 2017 (UTC)

A way forward

I will boldy set something up:

I will make a subpage here. The subpage will have sections. The first section will be raw, unvetted names. Sections below will be headed "Good AfD stats", "No recent blocks", etc... Volunteers can vet, moving items from section to section. Those items that arrive at the last section are ready for a talk post suggesting they use the poll. Sound good? Anna Frodesiak (talk) 23:42, 9 April 2017 (UTC)

I hope this will get the ball rolling. If and when scripts are made to vet for AfD stats, etc., we can modify or abandon the assembly line flow method of vetting.

Here is the link: Wikipedia:Requests for adminship/Optional RfA candidate poll/Poll candidate search

Anna Frodesiak (talk) 00:03, 10 April 2017 (UTC)

Several names have already made it through the vetting and are ready to receive a talk page post. Please check Wikipedia:Requests for adminship/Optional RfA candidate poll/Poll candidate search/Poll candidate post and say at that talk page if it is ready to send? Anna Frodesiak (talk) 07:50, 10 April 2017 (UTC)

Avoid biting comments

Please folks, don't bite. Someone who needs work in various areas should be told so, politely, and with good guidance and encouragement. Bite people here and they may leave the project. Encourage them and they may improve and eventually become admins. Anna Frodesiak (talk) 06:25, 29 April 2017 (UTC)

I share the concern over commentary of this nature, which has been particularly common in the last few weeks. To be honest, I see this as linked to a concern I expressed earlier; which is that this page is shifting away from its intended function of giving accurate advice to the candidate, and that this is the result of changes in the body of people commenting, and how they view this page. Vanamonde (talk) 06:42, 29 April 2017 (UTC)

I agree, Vanamonde. I think the way to fix this is in the instructions section. I've added something. Please help improve it. Cheers, Anna Frodesiak (talk) 06:48, 29 April 2017 (UTC)

(edit conflict) For instance: two of the most hostile comments of late came from users who have six RFA votes between them. I have nearly fifty, and I still hesitate to comment, and do not give numerical ratings...Harsh as it may seem, I'm wondering if we should limit the folks who can comment here, because the instructions at the top of the page do not seem to work. Post script: Yes, the modifications to the instructions are a good thing, but I think my concern still stands... Perhaps an edit notice may help? Vanamonde (talk) 06:52, 29 April 2017 (UTC)

(edit conflict) They're certainly not half as harsh as they will experience at RfA, so I don't think we should exaggerate the problem. After all, the only people we (should) be inviting are experienced editors; they are surely unlikely to walk away after being snarked at at a board they never intended to visit in the first place perhaps.

But I agree that the commentary has lost focus. More and more editors seem to be forgetting that the poll's purpose is to gauge the mood of an RfA in the context of that particular individual. But we are getting an increase in the kind of 'good editor, I'll vote for you' comments. These are singularly unhelpful and merely waste the candidate's time, as they will give them a false impression of what to expect. — O Fortuna^{semper crescis, aut decrescis} 07:04, 29 April 2017 (UTC)

Hi Vanamonde. An edit notice would be great. Giant red letters in a trimmed-down top part would be great. I would stop short of restrictions to who can edit this. Also, I doubt the community would give consensus for that, especially considering this is not an out-of-control problem, and there are alternatives. Anna Frodesiak (talk) 06:58, 29 April 2017 (UTC)

I've made an edit notice: it's here. Anybody's who is watching, please feel free to modify it. I didn't make the lettering red, I thought it might be a bit too scary. @Anna Frodesiak: I can perhaps see why such a limitation may be a problem, so let's see if we can address this another way. How about adding something like the following note the instructions: "Before posting a comment, please consider whether your comment is adding something to the discussion. This is not a vote. Avoid pile on votes that do not help the candidate" Does that get the message across? Vanamonde (talk) 07:22, 29 April 2017 (UTC)

No need to focus on pile-on comments; the issue is that people do not follow their original instructions, re: being a barometer for RfA not a reflection of individual opinions. — O Fortuna^{semper crescis, aut decrescis} 07:28, 29 April 2017 (UTC)

It looks good, Vanamonde. I like the giant letters. I'll leave it to you and others to make any further modifications. You all have great judgement. Anna Frodesiak (talk) 07:29, 29 April 2017 (UTC)