Wikipedia talk:Requests for adminship/Optional RfA candidate poll/Archive 2

From WikiProjectMed
Jump to navigation Jump to search
Archive 1 Archive 2 Archive 3 Archive 4 Archive 5

A particular entry

-Jkudlick asked what he thought of his chances. IJBall responded with "...this page is not to be used for "Editor review"...". I could be wrong, but I am not sure I agree with that. Why shouldn't everyone feel welcome to add their name. After all, they could get valuable feedback that could set them on the right track. Have we decided that entries such as Jkudlick's are out of scope? Anna Frodesiak (talk) 07:56, 12 January 2016 (UTC)

Yes, we did, AFAIAC. And the topic above this one is about putting that in writing on the page. But I feel that we did decide that this page isn't for generic "editor review". That's certainly what I think, and I'm pretty sure it's what Kudpung thinks. I'm assuming most others agree. But these kinds of "I'm not ready for RfA now – please tell me what I need to do to become an Admin in 6–12 months..." entries are not appropriate here – editors who want that need to go to their favorite long-term editor's/Admin's Talk page and ask for that kind of assessment there. This page needs to be focused on polling on "What are my RfA chances right now?" Otherwise, what's the point – we might as well just move this over to WP:Editor review... --IJBall (contribs • talk) 08:15, 12 January 2016 (UTC)
As to Jkudlick, the specific statement was: "I doubt I have a snowball's chance right now, but I'd be interested in seeing what other editors feel I need to do to improve my chances." That's asking a speculative question about what they might need do in the future: the candidate already concedes they're a WP:NOTNOW. It's just not appropriate for ORCP, IMO. --IJBall (contribs • talk) 08:18, 12 January 2016 (UTC)
There have already been a lot of WP:NOTNOW editors ask for feedback already. I think that as long as they are asking for that specific feedback on their chances for adminship, they should get it from those who are willing to provide it. It doesn't look like the project is being Inundated with requests, so it doesn't look like it is a practical problem. --I am One of Many (talk) 08:31, 12 January 2016 (UTC)
No – previously we've had a bunch of clueless NOTNOW's wander by, and they got 0/10's and were told they aren't ready. Jkudlick isn't like that: Jkudlick's a self-aware NOTNOW. What value is giving Jkudlick a bunch of 0/10's going to serve?! And the top of the page specifically states (in bold lettering, yet): "Please read Wikipedia:Not now before adding your name to this list." This isn't editor review – this is supposed to be responders' assessment of candidates' current chances of passing RfA. (And if it starts to deviate from that, I suspect a number of us will lose interest...) --IJBall (contribs • talk) 08:47, 12 January 2016 (UTC)
I must disagree, my friend. :) Where did we agree that the poll is only for "my RfA chances right now"? There have already been lots of entries with helpful responses like "...work on this a bit more and in a few months you'll be ready..."
Also, this poll may be serving multiple functions right now. It may be stopping SNOWs/NOTNOWs and it may have helped a couple of now-admins to take the plunge. Why can't it have other functions too? As editor review is defunct, isn't there a need? And why should a single editor give all the guidance at their talk. The magic of the wiki is at work: lots of people all giving input.
So, there seem to be two issues:
1. Objection to a response more detailed than a simple number. (I could get on board with limits here, but I don't mind what is happening either.)
2. Entires asking about their chances at a time other than "not right now". (I'm fine with that. Responders say what is needed. That is a good thing.)
Anna Frodesiak (talk) 08:50, 12 January 2016 (UTC)
All I can tell you is that I won't be interested anymore if ORCP starts allowing "What do I need to do to become an Admin in 'x' months?"-type entries. Currently, candidates are getting numerical results on their current chances, and advice on what they might do to improve their chances. If ORCP morphs into advice on how NOTNOW's can learn all the shortcuts to becoming Admins... well, I couldn't be less interested. --IJBall (contribs • talk) 09:11, 12 January 2016 (UTC)
I'd have to agree with IJBall here - there's no harm in a "3/10 - focus on blah..." style response but ideally I think polls should be created only by editors looking to RfA in a maximum of 1-2 months time -- samtar whisper 09:37, 12 January 2016 (UTC)
Kudpung and you have expressed your views on this, but otherwise I don't see a consensus agreement, particularly since Kudpung deferred to Anna's judgment. Personally, I too think it makes more sense not to conflate a quick poll on suitability to be an administrator with general feedback, but as a matter of practicality, if there are editors currently watching this page who are willing to provide feedback, it may be better to let it continue. Volunteer participation is a tricky thing to manage; even a small change such as shifting discussion to another page (for example, a revived Editor Review page) can kill it off. Anyone not interested in providing general feedback can ignore such requests, and if enough of them are ignored, then they may abate in frequency. isaacl (talk) 17:28, 12 January 2016 (UTC)

Lest anyone think I'm a "monster" here, I've gone ahead and given Jkudlick an in-depth assessment at their Talk page. (IMO, that's how cases like this should be handled!) --IJBall (contribs • talk) 18:12, 12 January 2016 (UTC)

Thanks for giving Jkudlick an in-depth assessment. Anna Frodesiak (talk) 23:12, 12 January 2016 (UTC)

As for what happens here, I defer to the community and do not own this, of course. The only thing I own at Wikipedia is User:Anna Frodesiak/Basement. However, Dr. Blofeld rents it. He can paint it or do horrible things to child admins there if he wants. That is in his lease. Anna Frodesiak (talk) 23:12, 12 January 2016 (UTC)

WHAT??!!! 78.26 (spin me / revolutions) 23:24, 12 January 2016 (UTC)
Yikes! --IJBall (contribs • talk) 23:37, 12 January 2016 (UTC)
  • My thoughts on ORCP are well known. I concur 100% with IJBall who echoes those thoughts entirely. I contend that anyone who joins Wikipedia with even a fleeting thought of wanting to become an admin someday, has no business ever becoming an admin - or at least not for a great many years until they have reached the age or maturity level of one who at least has the age of majority and then a good deal of clue more. As much as I despise the divisive behaviour of the editor who has the Guinness Book of Records for block logs, they hit the nail on the head many years ago when they complained about "the children who run this site". My full realisation came when I closed down the old version of the CVU/A and kicked out the bunch of 7th graders who had turned the section into a Social media/MMPORG and getting nothing else done. I don't have anything against youngsters - I spent a near 40 year career working with them - but Wikipedia is serious stuff and has to be kept so, and that's why I shudder when I see a teenager's user page looking like a teenager's bedroom wall, and keep seeing discussions that start with "Hey, Bro" and contain a ton of 'cools' and 'awesomes'. Kudpung กุดผึ้ง (talk) 13:55, 13 January 2016 (UTC)
Good points, Kudpung. I do agree with "...anyone who joins...even a fleeting thought of wanting to become an admin someday..." So, how about some balance here? I respect, and am listening to, you and IJBall. Maybe instructions should lay out the scope more clearly. Something like "...intended for experienced editors who intend on running in the near future...not editor review..." I also see a dwindling number of new entries. Maybe this is because those who wanted to know, have already added their names. Maybe it is because the page is getting full of big paragraphs. I suggest we modify the instructions to limit scope in terms of who should add entires, and urge responders to stick to a "very brief" response. Thoughts? Anna Frodesiak (talk) 23:27, 15 January 2016 (UTC)
I think the existing instructions already cover these points, but certainly the language can be tweaked. I feel, though, that those using this page to solicit reviews are either not reading the instructions, in which case modifying them won't matter, or are ignoring them. To address this second group, as I suggested above, I think it may be helpful to add a sentence providing some advice on how editors can seek feedback on what they can do to improve their contributions. isaacl (talk) 00:05, 16 January 2016 (UTC)
I agree wit Isaacl, the instructions on the top of the page are clear enough - we don't want admins who can't/won't read basic instructions, but Anna's "...intended for experienced editors who intend on running in the near future...not editor review..." certainly might be worth a try. We need to nip such entries in the bud with 'Sorry, this forum is for genuine prospective admin candidates only.'
If the number of entries has dropped, I think it's probably because they've been reading the kind of answers others have been getting. In which case, the page is also doing its job. Kudpung กุดผึ้ง (talk) 05:24, 16 January 2016 (UTC)
OK, I have made some changes based on Anna's comments and mine in the section above. If there are any objections or further suggestions, please chime in. isaacl (talk) 15:52, 16 January 2016 (UTC)
Looks good! Anna Frodesiak (talk) 20:26, 16 January 2016 (UTC)
I don't really think the number of entries has dropped in a significant manner. We can decrease the archiving time to say 21 days of inactivity to clear the page faster, if desired. I think we shouldn't overanalyze every hiccup and burp: contributions will always have some degree of randomness, and there will always be editors who ignore instructions and do what they want. It can all be managed appropriately without a lot of fuss. isaacl (talk) 15:52, 16 January 2016 (UTC)
Good points too. :) Anna Frodesiak (talk) 20:26, 16 January 2016 (UTC)
In response to Anna's concerns about "dwindling candidate" postings here, I do wonder if the auto-archiving of "stale" candidate assessments should be cut from 720 hours (30 days) to 360 hours (15 days) – doing so would make the ORCP page less massive with content (and potentially intimidating). If anyone misses one they want to comment on within the 15 days, they can always "unarchive" back to the main ORCP page to comment. Thoughts?... --IJBall (contribs • talk) 03:43, 17 January 2016 (UTC)
I suggested 21 days, but 15 days is probably OK. The main constraint is allowing enough time for the candidate to see the poll results before it gets archived. isaacl (talk) 04:51, 17 January 2016 (UTC)

Now, as for my future at this talk page, I think I should stay out of discussions. You guys are smart and neutral and can take it in the right direction. I feel tinglings of ownership and I just don't trust my neutrality. I shouldn't be involved here. However, I still would like to advertise this from time to time at various wikiprojects, unless you have strong objections. I mean, it can't hurt to seek out those great potential candidates, right? Anna Frodesiak (talk) 20:26, 16 January 2016 (UTC)

Although you've not been shy about expressing your views on how this page should be managed, you've also been very accommodating of the points offered by others. So while I understand your uneasiness, I believe you have encouraged a free exchange of opinions and shown a great deal of willingness to defer to larger consensus. Your guidance as the poll's creator has been invaluable in providing a central focus for this page, something that could easily get lost if you stepped back. While personally I hope you continue to participate on this talk page, in any case thanks very much for your persistent efforts! isaacl (talk) 01:27, 17 January 2016 (UTC)
Thank you, my friend. :) I am a strong believer that pages find their way over time. I see a dozen or more editors who have helped shape this thing and there are 99 watchers. I have every confidence that you and others can take it in the right direction. If I really feel compelled to weigh-in on something, I will, otherwise, I will watch from the sidelines. Anna Frodesiak (talk) 04:39, 17 January 2016 (UTC)
Anna, I don't think you should feel any guilt of ownership whatsoever. What you are feeling is a sense of responsibility for ensuring that the project you created stays broadly on the lines of the spirit in which you created it. Much in the same way as I look upon WP:RFAADVICE today. You should continue to provide your input. Kudpung กุดผึ้ง (talk) 06:14, 17 January 2016 (UTC)
Hi, Kudpung. I don't feel guilt, really. The difference between what you say I am feeling and ownership is very fuzzy. Dangerous territory for me. :) Besides, who says it should "stay" along any lines I intended? The community knows best. I want to see how it changes shape without my input. Then the proof will be in the pudding. .....chocolate pudding. I want chocolate pudding. Anna Frodesiak (talk) 22:45, 17 January 2016 (UTC)

Random advice...

Is it time to close Cullen328's entry? Up to this point, I've only purposely archived clear NOTNOW or NOTYET cases. Cullen328's would be the first "obvious pass" case that I've ever tried to close. But, at this point, are any more 10/10 appraisals really useful?! – It would seem it's time to close this one with like a "Please run! (duh!)" close summary or something. So does anyone object to closing that one?... TIA. --IJBall (contribs • talk) 17:05, 24 January 2016 (UTC)

Given that this page seems overprotected (50 discussions in a few months for a lightweight process!), I have no objections. Esquivalience t 03:16, 6 February 2016 (UTC)
As I said before, I don't see a need for a formal closure or a closing summary: candidates ought to have no problem understanding the feedback on their own. Although I have no particular objection to archiving, personally I'd as soon let the archiving bot do its job. isaacl (talk) 04:42, 6 February 2016 (UTC)

Overly high ratings

Is there some way to handle situations where some ratings may be a tad high and encourage candidates to try RfA before being ready? Anna Frodesiak (talk) 11:23, 16 June 2016 (UTC)

Over time it might become self correcting as people realise that their ratings jibe with the RFA crowd. We could speed that up by gently pointing out the de facto standards at RFA to some who are over enthusiastic. Or we could just accept that a candidate who gets around 60% might get very high ratings from the majority and an oppose from the minority. It could of course be inherent to the process that candidates whose RFAs would have passed if things hadn't been found by !voters might get several thumbs up from this process but fail under greater scrutiny. The only solution I see to that is to increase scrutiny of the candidates here. Not sure how you do that though. ϢereSpielChequers 15:58, 16 June 2016 (UTC)
If you have a particular commentator in mind, you might want to have a private word with them to remind them that the figure is supposed to be an estimated probability of passing, rather than the rating they'd personally give the candidate's work. There are plenty of people I'd personally support who I'd consider highly unlikely to pass, and vice versa, and I get the feeling a lot of people are just giving personal assessments. ‑ Iridescent 16:10, 16 June 2016 (UTC)
For reference, here is the previous discussion on this topic. I agree with modifying the instructions to indicate that the rating should be an estimated probability of a candidate passing, and not the reviewer's personal rating for the candidate. isaacl (talk) 19:24, 16 June 2016 (UTC)

WereSpielChequers, increase scrutiny, yes, but maybe it is the responders who require the scrutiny. Could we add to the disclaimer that candidates ought to scrutinize who is rating them? Should we respond at the entry or here if we think a rating is too high (nah, but just brainstorming)? Should we have a section called "Past results" with links to old poll entries and their RfA?

Iridescent, I did have a private word. I found it impossible to help the individual understand that the poll is about their objective likelihood of passing and not about needing more admins or their desire for the candidate to pass.

Good thought, isaacl. Maybe we could add to Wikipedia talk:Requests for adminship/Optional RfA candidate poll#Responders (possibly in bold and not horribly written) something like: "...keep in mind that your rating is about your opinion of the objective likelihood of their passing and not about your desire for the candidate to pass, your views on needing more admins, or anything else..."

Anna Frodesiak (talk) 22:20, 16 June 2016 (UTC)

In case a "past results" section ends up being wanted, here is a list of those who used the poll and then ran for RfA:

Anna Frodesiak (talk) 23:22, 16 June 2016 (UTC)

The current instructions do have a caveat that "actual RfA results may differ greatly and that opinions given here may be based on only a cursory assessement." I don't believe scrutiny is the right word to use, but the experience of the reviewer should be taken into account when a potential candidate is interpreting the feedback provided. Unfortunately, editors who submit their name to this poll typically feel uncertain about their ability to self-evaluate their suitability for adminship, and so may also lack sufficient discernment to place the comments received into context. Regarding providing more guidance for determining a rating, I think it is a good idea. isaacl (talk) 02:19, 17 June 2016 (UTC)
I think some of those commenting are missing the point and treating it like the old Editor review, giving candidates their own personal rating based on how they like the candidate rather than taking the wider view of how others are likely to view the candidate in the less forgiving environment of RFA. Even though it does say that at the top, maybe it needs reworded to make it clearer. Take the latest review (Oshwah) as an example, good candidate but 9/10 and 10/10 suggest a virtually unopposed nomination, whereas in reality there will be opposes based on judgement concerns over GAN reviews and also insubstantial content contribs. (I'm seriously unconvinced that Windows Push Notification Service is a good article.) Valenciano (talk) 14:38, 17 June 2016 (UTC)
I've gone ahead and added a little bit of bolding to some of the responders section. I know people who are already here probably won't notice, but perhaps it'll help any new responders? Feel free to revert an' all -- samtar talk or stalk 14:45, 17 June 2016 (UTC)
I've added another sentence to that same section, hopefully it will make it a little clearer. Omni Flames (talk) 02:16, 18 June 2016 (UTC)
Thank you! Good job. Anna Frodesiak (talk) 23:28, 18 June 2016 (UTC)

I was pleased to see NinjaRobotPirate remind a user about the purpose of the page. I'm remain concerned. I still see users pitching and persuading. Is there anything more that can be done? Should we establish an accepted way of handling this? Talk page posts? Reminders like NinjaRobotPirate did? Anna Frodesiak (talk) 23:28, 18 June 2016 (UTC)

Putting notes on the poll page encourages discussion between commenters on that page, which I think is opening it up to becoming a venue for argument, and that concerns me. (*) Perhaps a note on this talk page, with a ping to the editor in question, would be better. The main problem, though, is that it's really hard to keep people from doing what they want to do, ignoring previous consensus either inadvertently or deliberately, particularly with the drawbacks of Wikipedia's consensus tradition.
(*) I think it might be desirable to try to limit discussion between commenters. Discussion with the potential candidate is more in tune with the intent of soliciting feedback; I don't think the poll subject should have to sift through people arguing with each other about their opinions. However I don't really know how to draw the boundary in a way that would gain consensus (I suspect I'm in the minority in wanting discussion threads to keep the person being reviewed at the centre of the conversation), plus I don't know how to enforce any defined boundary (see the previous paragraph). Moral suasion from the one(s) who launched any given initiative (in this case, Anna) is the only approach I've seen that has achieved some limited effectiveness. isaacl (talk) 02:42, 19 June 2016 (UTC)
Another example of people doing what they want to do: I really don't think it should be necessary to close a poll for the sole purpose of keeping people from repeating what has been said about a subject's lack of suitable qualifications—but unfortunately it seems some editors just really want to comment, even if there's nothing new to add. isaacl (talk) 03:01, 19 June 2016 (UTC)
First, 20 Wikipedians giving a 7/10 rating is more meaningful than two Wikipedians doing the same. Larger polling numbers should give a better statistical view. Secondly, too hopeful n00b editors ought to get their pile-on opposes here rather than RfA proper. We do no service to candidates by sparing their feelings. Finally, this problem ought to fix itself. I would think that candidates given overly optimistic comments here and get trounced at RfA will let those optional poll commenters know how they resent being given a bad steer. It won't take long before the glass half-full crowd learns their lesson. Chris Troutman (talk) 11:58, 26 June 2016 (UTC)
My example was not for 7/10 ratings, but for cases where the subject lacks suitable qualifications. I'm not concerned about sparing anyone's feelings; as I've said before on this talk page, I'd prefer that polls not be closed with a closing statement, but simply archived automatically by the bot. I just don't see any utility in adding more 0/10 ratings without new feedback, after a point. isaacl (talk) 03:16, 27 June 2016 (UTC)

A subpage for results

Should a results log be kept somewhere? Maybe a subpage? Maybe a section? (See the red and green thing above.) The section idea doesn't sit well with me for some reason. However, it could be highly informative to potential candidates. Thoughts? Anna Frodesiak (talk) 23:28, 18 June 2016 (UTC)

I think it would be interesting to keep a couple things in a table:
  • User name
  • Date of open/close/whatnot (month?)
  • Mean score
  • Median score
  • Standard deviation
  • Whether the user RFAed within N months
  • Outcome of RFA
--Izno (talk) 23:45, 18 June 2016 (UTC)
Adding on to that, we may want to use {{Graph:Chart}} to make a (mini-)histogram of the votes. Enterprisey (talk!) (formerly APerson) 00:53, 19 June 2016 (UTC)
I think a subpage is appropriate (see previous discussion). Anyone interested in tracking this info is encouraged to start doing so! (Note there were concerns about trying to draw any conclusions from things like means, medians, and standard deviations, given the large number of variable factors; anybody is of course free to pursue this if you like, but be forewarned it will probably be hard to convince anyone that there is a statistical significance.) isaacl (talk) 02:08, 19 June 2016 (UTC)
@Isaacl: Noted; I'm more interested for comparison of the average score here and the percentage support at RFA (noting, of course, that RFA is most certainly not a vote... ;). --Izno (talk) 11:55, 23 June 2016 (UTC)
Given that around a hundred people typically comment on an RfA, whereas the number of commenters here is around ten or so, I think it will be difficult to determine any statistically significant correlation beyond a broad trend. Additionally, the participants in this poll vary a lot, which means any change in the standards used will have a large effect on the average score. Another complicating factor is that the poll result is supposedly a probability (though whether or not everyone has treated it that way is unclear). The result at RfA, though, is a support percentage that probably has a slight tendency to pull away from the middle of the discretionary zone, so a linear mapping between the two is unlikely. isaacl (talk) 12:44, 23 June 2016 (UTC)
I think its value might be to show potential candidates, especially those who are borderline, what happened in other borderline cases. Anna Frodesiak (talk) 22:56, 23 June 2016 (UTC)

I've started Wikipedia:Requests for adminship/Optional RfA candidate poll/Results. Anna Frodesiak (talk) 21:42, 22 June 2016 (UTC)

Not much to go on yet: only 7 entries. FTR, ORCP was "wrong" twice: "underrating" Amakuru, and "overrating" Anarchyte (though the latter is arguable – Anarchyte's ORCP score was only "74.3%", so I'd argue that ORCP was actually fairly on the mark even here...). --IJBall (contribs • talk) 07:46, 26 June 2016 (UTC)
Should a link to this "results" page be made from the ORCP page? Pinging Isaacl, Izno and Anna Frodesiak for their opinions on the question... --IJBall (contribs • talk) 07:49, 26 June 2016 (UTC)
I'll leave that to others to decide. I did not even wish to involve myself further at this project, but I felt it necessary. Anna Frodesiak (talk) 08:07, 26 June 2016 (UTC)
As for the poll being wrong, I'd say it mislead the candidate unfavourably only once so far with Anarchyte, and to me, that matters. Anna Frodesiak (talk) 08:09, 26 June 2016 (UTC)
I don't see a reason to transclude it, but no issue making a link. --Izno (talk) 17:02, 26 June 2016 (UTC)
Well, I do. That's why I transcluded it. As the discussion indicates, the table is meant to show both candidates and commenters how the results here have tracked with RfA. After all, this page is meant to help editors make decisions and these metrics are helpful. Chris Troutman (talk) 17:36, 26 June 2016 (UTC)
Given that we expect to have a large number of candidates who subsequently run, it is not practical to the purposes of this page to have the table transcluded. As I suggested above, a link is sufficient for that purpose. Even a summary of the results from previous poll instances I would more than happily accept. But transcluding the table will frankly overwhelm this page. --Izno (talk) 17:42, 26 June 2016 (UTC)
I think I'm with you. To transclude, we'd need to do a "rolling" list a la Wikipedia:Requests for adminship/Recent, and that just adds complexity to the idea. I'd rather just link to Wikipedia:Requests for adminship/Optional RfA candidate poll/Results from the main ORCP page. --IJBall (contribs • talk) 17:48, 26 June 2016 (UTC)
We're not yet overrun with candidacies where the table is overwhelming the page. Yes, we could get there eventually and a rolling log or an average stat would be helpful then. Displaying the table on the main page has far more utility than a link as there's less chance a misguided editor would click-thru. I would rather cross that bridge when we get there than remove utility in the here and now. Chris Troutman (talk) 17:52, 26 June 2016 (UTC)

I would actually prefer it if the page weren't a running indicator. I think it would be more useful to have all of the RFA tries listed there, which is why I think it will be unwieldy later regardless.

I don't know of many misguided editors who are a) able enough to run RFA and b) not able to use their backspace button to get back to this page. I think we can very clearly couch some language here of "here's a results page [link]; typically, responders here have a good pulse on whether you will pass RFA; the majority of cases previously which passed a later RFA received an average score here of X or greater, with Y correlation". We can also post something back at the results page (if even the breadcrumbs in the top left don't get the user where he wants to get back to) about that type of stuff. --Izno (talk) 18:15, 26 June 2016 (UTC)

Yeah, that actually sounds good to me – I'd prefer we do it just like that. --IJBall (contribs • talk) 18:57, 26 June 2016 (UTC)
OK, I've added a link to the "Results" page from the main ORCP page. I've left out the suggested statistical analysis for now (I'm not sure 7 results is enough to say much of anything anyway...), but anyone can feel free to refine and edit what I've done there. --IJBall (contribs • talk) 00:11, 27 June 2016 (UTC)
I think a link somewhere would be useful to keep track of the results page, rather than having to dig it out of the archives of the talk page. (Personally, I wouldn't put the link in the "Instructions" section; I believe it is best to keep the instructions as concise as possible.) I don't see transclusion as being required. Even if there were any statistical significance to the figures (something fairly uncertain), the whole point of this poll is not to be another RfA and so the focus is not on "here's the level of support you've received from the commenters". Instead, candidates should consider the feedback they have been given as a supplement to their own judgments on their states of readiness, based on the various RfA advice pages and their own experience with the RfA process. Regarding the degree of credence that should be given to any given comment, the small number of commenters means that the candidate should be determining this on an individual basis for each comment. I see this as an essential qualification for someone being granted administrative privileges. isaacl (talk) 03:50, 27 June 2016 (UTC)

It needs a bit of filling-in. Cheers. Anna Frodesiak (talk) 11:40, 9 July 2016 (UTC)

@Anna:  Done Omni Flames (talk) 11:53, 9 July 2016 (UTC)
Thank you kindly, Omni Flames. :) Anna Frodesiak (talk) 22:42, 9 July 2016 (UTC)

A match column

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

Isn't part of the purpose of the table to show when the poll results got it right, or not? Should we make a righthand column for that? The red and green may confuse people. I see two reds for Supdiop with the poll getting it right. And Amakuru's 65% likely at the poll ended with a 98.9% RfA success. Did the poll get it right? A poll could end with "70% likely" with a "65% RfA fail". So what then? I can't get my head around this one. Brainy people needed to comment on this one. :) Anna Frodesiak (talk) 01:02, 10 July 2016 (UTC)

@Anna Frodesiak: I'd say that the poll got it right in pretty much every case except Amakuru. In the RFAs of Anarchyte, Jo-Jo Eumerus and Peacemaker67, I think that the poll was partially right, so perhaps we could mark those with the lighter green color? That's just my view. Omni Flames (talk) 01:11, 10 July 2016 (UTC)
The color coding is a little weird. RfA results are the percentage of editors which !vote in support, but the ORCP average is the average probability of passing RfA. These are not the same thing, so comparing them seems odd. I don't really have a great solution, but maybe set a hard cut-off for ORCP above which we consider an editor "likely to pass" and below which we don't? That makes it at least somewhat comparable. If ORCP predicts 100% chance of passing and the RfA passes at 80%, that's still a "correct" result for ORCP. ~ Rob13Talk 07:02, 10 July 2016 (UTC)
Good thinking, Rob. Yes, maybe something binary with no colours. It could show a match or mismatch only. But then, yes, we'd need a hard threshold at the poll, and by definition that would be 50%, right? But it isn't. It's probably something like 75%, i.e. below that and the candidate would likely fail, etc. Could historical results actually help us determine what the de facto threshold is, and then could be adjusted on an ongoing basis for each new entry? But then wouldn't that skew previous result reporting. I'm getting a brainache and have to lay down now. :) Anna Frodesiak (talk) 07:11, 10 July 2016 (UTC)

I had to call in the big guns. Anna Frodesiak (talk) 07:16, 10 July 2016 (UTC)

My above suggestion would require a pure subjective judgement call. We'd have to seek some type of rough consensus on at what "average" we recommend someone accept a nomination at RfA. Setting aside my suggestion, the real way to solve this would be to track each individual contributor's ratings and how they correlate with success at RfA so potential candidates can evaluate who's opinions really matter. I have a feeling the individuals without much correlation would get annoyed when that was published, though. There's also an unsolvable issue of bias here; those who receive somewhat low ratings may choose not to go to RfA even if they would have passed, so we're missing some of the "worst" matches. Any measurement of our success rate will over-estimate as a result. Short of randomly compelling certain random ORCP candidates to go for RfA, there's no way to "fix" the self-selection bias informed by the ORCP feedback. ~ Rob13Talk 07:25, 10 July 2016 (UTC)
(edit conflict)It seems to me that we are talking about type I and type II errors or false positives and false negatives. The poll describes the likelihood of passing RfA so a false positive (type I error) occurs where the poll predicts the candidate will pass but they fail and a false negative (type II error) occurs when the poll predicts the candidate will fail but they actually pass RfA. False positives risk much more damage to candidates, IMO. I would include in the extra column when the poll was accurate and when inaccurate, whether the error was type I or II. Of the nine reviews included on the results page, seven were accurate, one was a type I error, and one was a type II error. EdChem (talk) 07:34, 10 July 2016 (UTC)
Again, of course if anyone is interested in pursing this type of analysis, feel free to continue. However as has been raised in previous threads and by BU Rob13, there are many problems with trying to derive significant information from the data, and so my personal feeling is that it may be more fruitful to invest effort in other analyses. For example, it may be useful to try to work up an analysis of the characteristics of RfA candidates. isaacl (talk) 17:39, 10 July 2016 (UTC)

Chart for number of editors using poll

I made it but it needs work. The numbers correct, though.

Anna Frodesiak (talk) 01:26, 21 July 2016 (UTC)

Could someone pls change the July total to 5 and start Aug with 1? I have no clue how to do it. Thanks! Anna Frodesiak (talk) 08:34, 13 August 2016 (UTC)

 Done Mduvekot (talk) 14:54, 13 August 2016 (UTC)
Thank you so much, Mduvekot! :) Anna Frodesiak (talk) 00:35, 14 August 2016 (UTC)

Regarding revert to page by IP

I just reverted an edit to the page by 79.71.182.75 (talk · contribs · WHOIS) using an AGF rollback (diff). But, I just checked to see if the mentioned "Willow q" was a currently registered user, and it turns out they are: Willow q (talk · contribs). I am unsure what to do with this information, so I am bringing it here. (Also, I retract the WP:AGF in the edit summary of the rollback.) ~Cheers, TenTonParasol 05:22, 17 August 2016 (UTC)

Possible spamming, by me of all people

That's right. :) I've asked at this talk before to be sure it is okay. I periodically post a dozen or so of these at different Wikiproject talk pages. Is this appropriate? Best, Anna Frodesiak (talk) 00:53, 14 August 2016 (UTC)

I can't possibly see how that can be construed as spam. But as an admin you get get accused of all sorts of ridiculous things by people who have an axe to grind. --Kudpung กุดผึ้ง (talk) 04:15, 14 August 2016 (UTC)
Looks fine to me, nothing wrong with trying to get a few eyes on the project. --Church Talk 04:25, 14 August 2016 (UTC)

Thank you for the input Kudpung กุดผึ้ง and Church, and sorry for the blated response. I continue to post at these wikiprojects. The response thins. I wonder if this poll is petering out. Anna Frodesiak (talk) 23:19, 3 September 2016 (UTC)

Results page format

Convenience links:

It now looks like this. I've removed the confusing red and added a column showing RfA result. Now, visitors can simply look at figures. Is this better? Anna Frodesiak (talk) 20:56, 31 August 2016 (UTC)

I sort of miss the color coding, but not too much. Otherwise, it looks great! Enterprisey (talk!) (formerly APerson) 21:01, 31 August 2016 (UTC)
I miss it a bit too. :) Anna Frodesiak (talk) 21:07, 31 August 2016 (UTC)

Fun facts

RfA success rate percentage by quarter (2011 1st to 2016 3rd). The red line indicates the average. The green arrow indicates the start of this poll. (Source figures are at the image description field)

Okay, there are no facts. The chart seems to show no increase in percentage of successful RfAs.

Whether or not more people are running because of this poll is hard to tell. Without the poll being in existence, the number of RfAs may have been lower or the same. So, yes, no facts. Anna Frodesiak (talk) 04:09, 4 September 2016 (UTC)

That red line looks awfully high. Maths is not my long suit, so maybe somebody should check it. Thanks. Anna Frodesiak (talk) 06:16, 4 September 2016 (UTC)
Checked and confirmed as correct. Anna Frodesiak (talk) 06:34, 4 September 2016 (UTC)
Interesting find, Anna Frodesiak. I don't think a conclusion can be drawn as to whether or not more people are choosing to participate within the poll either. I also cannot see any correlation between the introduction of the poll and increasingly successful RFAs. Once again, interesting indeed. In short, does this mean that there is no way of telling if these candidate polls are actually beneficial? Patient Zerotalk 09:15, 10 October 2016 (UTC)
The benefit, if any, would likely be a dip in the number of non snow-closed RFAs. People who've been here a week and jump straight in to RFA without reading the instructions are still going to happen, but if this process has a use it's in filtering out people who genuinely feel they're ready but want to know if a mistake they made three years ago, or a lack of participation in the Timed Text namespace, is likely to be held against them. Is there any way to filter the numbers by edit-count? If this process is serving a useful purpose, I'd expect to see a statistically significant drop in "unsuccessful RFA from editor with a high edit count" with "unsuccessful RFA from editor with a low edit count" and "successful RFA" both remaining roughly constant. ‑ Iridescent 09:45, 10 October 2016 (UTC)
Ah, thanks for clarifying that Iridescent - I'm interested to see if that can be done, also. Patient Zerotalk 10:09, 10 October 2016 (UTC)
The key issue goes back to Anna's observation: it's hard to tell what would have happened without the poll, given the many other influencing factors (for example, experienced editors are likely to be disproportionately deterred from requesting adminship due to dissatisfaction with the editing environment, compared with those with less experience). In addition, the decreasing number of requests for adminship makes drawing any statistical conclusions a challenge. Probably the most straightforward measure of a positive influence is the number of people who get feedback (favourable or otherwise) in a poll, and then later proceed to attain approval for administrative privileges. Because the seeds sown now may take some time to germinate, the full impact might only be visible after a few years. 15:19, 10 October 2016 (UTC)

Helper script needing signature

Hello. Would the person who set up the automatic "add your name below" thingy please make the potential candidate's signature appear once saved? This would make keeping track of monthly participants easier. Cheers. Anna Frodesiak (talk) 23:59, 9 October 2016 (UTC)

I altered Wikipedia:Requests for adminship/Optional RfA candidate poll/preload to include the current date in the heading. isaacl (talk) 04:52, 10 October 2016 (UTC)
Perfect. Thank you so much, isaacl. :) Anna Frodesiak (talk) 09:05, 10 October 2016 (UTC)
Retrieved from "https://mdwiki.org/wiki/Wikipedia_talk:Requests_for_adminship/Optional_RfA_candidate_poll/Archive_2"