Wikipedia talk:Requests for adminship/Optional RfA candidate poll/Archive 4

All that text before the entries start
It could use a trim. It got a tad creepy. Anna Frodesiak (talk) 06:51, 29 April 2017 (UTC)
 * I'll make a pass at it, but note since the poll's inception, there has been ongoing discussion on how to keep comments short, constructive, and focused. For better or worse, without curation, editors generally do whatever they want, regardless of what the instructions say. isaacl (talk) 08:12, 29 April 2017 (UTC)


 * Hi isaacl. I figured that if we trim sentences without losing meaning, there would be less clutter and the important bits in bold and red et cetera would stand out better. Anna Frodesiak (talk) 08:20, 29 April 2017 (UTC)


 * Too many contributors either make an RfA-style in-depth review or talk more about how they would vote themselves. The White Arabian Filly poll just went totally OTT. Someone should have put a stop to it. The other problem is that even if one makes a short objective criticism, the candidate retorts with righteous indignation or tries to justify themself. Nothing much we can do about it all. Good edit notices, BTW. Kudpung กุดผึ้ง (talk) 11:09, 29 April 2017 (UTC)
 * Regarding trying to limit the amount of off-topic discussion: how about asking that responders not reply beneath another person's comments? If they really feel a need to say something about someone else's evaluation, they can do it in their own response. Part of me thinks responders should be encouraged to not comment on the viewpoints of others, allowing the candidate to decide how much weight to give each response. But I realize the practical difficulties in trying to enforce this. isaacl (talk) 22:26, 29 April 2017 (UTC)

Thanks, all. That is a huge improvement to all that text before the entries start. Anna Frodesiak (talk) 06:19, 30 April 2017 (UTC)


 * All meta/maintenance areas are a magnet to new and inexperienced users (it's a fact of internet forum life) and ORCP, like RfA itself, is no exception.  Just an observation. Kudpung กุดผึ้ง (talk) 21:21, 4 May 2017 (UTC)

Closing polls
Regarding [//en.wikipedia.org/w/index.php?title=Wikipedia:Requests_for_adminship/Optional_RfA_candidate_poll&diff=779330381&oldid=779253830 this edit closing a poll]: as I previously discussed, in general I don't see much need for formal closures of polls (and I don't see any need to specifically mark a closure as a non-admin closure). I appreciate in cases where there is an ongoing trickle of pile-on views that closing a poll may be useful. Personally, though, I suggest not having a lengthy closing statement as I think evaluating feedback is one of the skills candidates should be demonstrating. If they want someone's opinion on the received feedback, they can ask on the editor's talk page. I also think it makes the poll feel less lightweight and perhaps more intimidating than was originally intended. I know some may think that this poll should be more intimidating to those are not sufficiently self-aware to properly evaluate the received comments (which, on a side note, I do not believe was the case for the candidate in this particular poll). But I believe we are able to discourage this category of editors from initiating a request for administrative privileges in a succinct manner, thereby keeping the poll more informal. isaacl (talk) 15:59, 8 May 2017 (UTC)
 * It clearly does no harm at all. But thanks for discussing my edits. &mdash;  O Fortuna   semper crescis, aut decrescis  16:03, 8 May 2017 (UTC)
 * I agree with Isaac's position that the closing statement was a touch overdone. Let's keep these succinct (of the sort: do the thing, don't do the thing, maybe do the thing), since this is intended as a lightweight process. --Izno (talk) 16:27, 8 May 2017 (UTC)
 * I don't think it did any harm in this case. I am suggesting, though, that it might not be a practice that we want to continue. If there is a consensus to have formal closing statements, that's OK, too. isaacl (talk) 16:31, 8 May 2017 (UTC)
 * I recognize that this closure was in good faith and I have no quarrel with it. But I also think formal closures are not necessary, and in fact I would prefer not to have them (with the possible exception of snow "put this one out of its misery" results). We are not deciding whether they should go for it or not; that is their decision. We are just providing opinions in a somewhat structured way; it's not up to us to tell them what to do. I think we should just archive the discussion after the commentary dries up. --MelanieN (talk) 18:40, 8 May 2017 (UTC)
 * Agree entirely. Sam Walton (talk) 19:41, 8 May 2017 (UTC)
 * I only started the poll in the first place because several people asked me about 'cratship. Maybe if 25 people had voted "10/10 I'll nominate you right now" I might have thought about it, but obviously that wasn't the case and I can't get that excited about it. I think FIM did quite a good close rationale myself. Ritchie333 (talk) (cont)  10:42, 9 May 2017 (UTC)
 * I've seen a lot of "formal closures" as Melanie put it best - I don't really think they're necessary. Archiving old ones will suffice. Patient Zerotalk 10:44, 9 May 2017 (UTC)

Putting it right
Hi User:TonyBallioni. You wrote "...I hate ORCP and think it is actually more broken than RfA in some ways..." How so, and what can we do to fix it? Best, Anna Frodesiak (talk) 02:18, 30 June 2017 (UTC)
 * I drop by about once a week to see if anyone is here and see what is written. Three main ones: first, for about 75% of the regulars here I could guess their opinion on any given candidate before reading it simply based on their general views on wiki-politics. The ratings don't come off to me as how well I think you will do so much as how much I'd personally want you to run and I'm going to encourage you to do so or discourage you accordingly. That's not good. You also have the pile-ons that can make this worse: GoldenRing would never have gotten a good rating at ORCP and Pvmoutside got 8.5. I think both of these factors played into it. Finally, the fact that it has started being cited in RfAs I think is a major negative: it takes the "O" part out of it. Someone who might otherwise pass RfA could get a negative rating at ORCP by the very limited group of people who comment here and have it used against them in an RfA.Now, there are definitely positives to it: about every ORCP you get several very insightful comments and you also get nominators for candidates who they might not have ever thought of, which is a very good thing. The first three things though to me make it a broken process that was not intended when it was implemented as a reform: all together, it makes it seem like a sparsely attended smoke-filled room. TonyBallioni (talk) 02:34, 30 June 2017 (UTC)
 * I don't think this poll has enough visibility yet for it to be considered an essential stop on the way to requesting administrative privileges. As I mentioned previously, I think curation would be needed to try to keep comments focused on evaluating a candidate's probability of receiving approval for administrative privileges. For better or worse, though, English Wikipedia editors who are most likely to contribute to these discussions are typically pretty resistant to changes that would keep them from writing about whatever they want. isaacl (talk) 03:56, 30 June 2017 (UTC)
 * Cwmhiraeth's RfA had multiple comments cite the ORCP as a reason for an oppose, as did Dane's, and Lourdes' was criticized for deciding to test the waters at RfA rather than ORCP and in fact the first question was "why didn't you go to ORCP?". I think this is a very disturbing trend, especially when combined with the relative low-attendance compared to RfA, and my critique on what the comments come off as. TonyBallioni (talk) 04:12, 30 June 2017 (UTC)
 * In the case of Cwmhiraeth, it wasn't specifically the ORCP that was the issue; that just happened to be a convenient place where reasons for opposing existed. Ritchie333 (talk) (cont)  11:44, 30 June 2017 (UTC)
 * It still raises the point that if you are a potential admin candidate who thinks they might have a chance of passing but that it would be a bumpy ride, you are incentivized not to come to ORCP, or even worse, simply not put your name forward at all because why would anyone want to go through two stage bumpy-ride where the first stage can be held over you? At least RfA is posted so a large sector of the community comments and a candidate's supporters will often engage the ridiculous opposes. Here you have neither. TonyBallioni (talk) 13:44, 30 June 2017 (UTC)
 * Just so. That's why I find it hard to summon much enthusiasm for the ORCP-stage of the process. They are two different consituencies that ocassionally overlap. RfAs have a mass of editors, both experienced and inexperienced, and there are enough that a crap 'oppose' rarely meets the critical mass needed to influnce the outcome. Here, the ratio is far narrower, and with less eyes on it, contributors have no pressure to actually do as they are meant to and the candidate at the next stage. &mdash; fortuna  velut luna  15:16, 30 June 2017 (UTC)

See below. Kudpung กุดผึ้ง (talk) 15:22, 30 June 2017 (UTC)
 * For what? &mdash; fortuna  velut luna  15:26, 30 June 2017 (UTC)

I think Requests_for_adminship/GoldenRing is a good example why ORCP may not be a good idea. If he had come to ORCP he probably would have picked up mostly 0/10 and 1/10 due to activity levels, and been dissuaded to try; and if he had tried RfA afterwards there probably would have been a pile-on "candiate ignores consensus at ORCP" as one more thing to attack him on. Luckily he ignored ORCP and went straight for RfA, and managed to pass. Without him doing this there might have been one less admin. I've also seen candidates attacked at RfA for not trying ORCP first.

For some reason many commentators at ORCP tend to "pile-on" far more than those at RfA. Of course it's far easier to do this as it's not one's personal opinion, but instead a prediction of RfA chances. Another reason why this may be the case is that people at RfA tend to get nominated by notable editors, while of course this doesn't happen at ORCP; there are less people to defend oneself. --Jules (Mrjulesd) 16:51, 30 June 2017 (UTC)
 * I think abolishing the number scores/anything bolded would help a bit. It would make piling on and citing ORCP in RfA more difficult because people would have to read the feedback first, which really is what ORCP should be about anyway rather than "pass or fail the pre-RfA". I still think you'd have the issues I described above, but it might lessen them. TonyBallioni (talk) 17:13, 30 June 2017 (UTC)
 * I don't think this is literally true, given the considerably smaller numbers of commenters at this poll versus RfA, but I will agree there are a lot of needless pile-on comments at this poll. As we've discussed previously (sorry, can't find the link right now), while it's true that more people expressing the same opinion gives it more credence, I don't believe there's much value past a few people. The candidate will either take the feedback into account at that point, or won't. Adding a few more "me-too's" won't make a difference. isaacl (talk) 18:11, 30 June 2017 (UTC)

I agree that we should watch for a trend, and start reminding people of the optional nature of this poll if necessary. As of yet, though, I don't think there's enough people considering the poll to be mandatory that it is an issue. By its nature, those taking a poll will generally be those less confident in their ability to self-evaluate. Personally, I think we should be encouraging potential RfA candidates to hone their introspection skills, and part of that is going through the process of rating their own suitability to be approved for administrative privileges, rather than taking a poll.

I'm not sure what can be done about those concerned about ignoring advice. It is a valid question to raise, but of course it depends on the quality of the advice: some perhaps ought to be ignored. Regarding the numeric score, I have previously expressed my dislike for this, but at the time, the consensus felt differently. Perhaps there has been a change? One potentially good thing about the score, though, is that it makes it easier to direct commenters to estimate the probability of success, rather than giving a personal opinion of the candidate's suitability. isaacl (talk) 18:04, 30 June 2017 (UTC)

On a minor side note: I realize some people are just doing it for levity, but the scores with decimal points always stick out like a sore thumb for me. This poll is hardly likely to be accurate within 10%, much less 1%. isaacl (talk) 22:03, 30 June 2017 (UTC)
 * I missed this discussion because I was away from home. It was not clear to me how an ORCP should be closed, so when I thought no further comments were likely in mine I tried to think of some innocuous remark and closed it with the statement "Thank you all for your comments. I will not be running for admin in the near future." It was not the ORCP itself that was seized on by the opposers at my RfA, but the fact that I was running only six weeks after saying this. So I think possible candidates (and reviewers) should be very careful what they say here. Cwmhiraeth (talk) 12:59, 11 July 2017 (UTC)


 * , I don't think this poll was ever supposed to be considered an essential stop on the way to requesting administrative privileges. As  says, it would take the 'O' out  of ORCP. Forcing candidates to run the gauntlet twice would surely kill off any remaining interest to be an admin, besides which, we have enough bureaucracy  on Wikipedia as it is.  For those who  want to  use it, ORCP links back to 647 pages. That  should be enough. There is probably  not another project on Wikipedia that has had so  much exposure. Finally, the only 'privilege' attached to adminship is the one of having to take a lot of vile attacks and other crap and not being allowed to do much about it. Kudpung กุดผึ้ง (talk) 14:29, 12 September 2017 (UTC)
 * Sure—I am in complete agreement regarding the optional nature. I was just disagreeing that this poll is a significant factor in people's expressed opinions at requests for administrative privileges, as it does not match my anecdotal reading of the comments at RfAs (but I have not collected any stats to confirm this). Accordingly, I don't think any candidate would start thinking of this poll as non-optional based on reading past RfAs. As I mentioned in my most recent comment above, personally I think we should encourage potential candidates to improve their abilities to self-evaluate their suitability. But I realize there is more than one way to do things, and some people like feedback forums. isaacl (talk) 17:19, 12 September 2017 (UTC)

side discussions
In the interest of keeping the poll comments concise and focused on the candidate, could commenters consider proceeding with side discussions on this talk page, instead of the poll page? This isn't a full discussion on a request for administrative privileges, so there's no tally at the end of the ratings: candidates can judge for themselves how much weight they should give to a given rating, based on the accompanying reasoning. Let's try to keep the poll lightweight and addressed to the candidate, rather than to other commenters. isaacl (talk) 03:46, 30 June 2017 (UTC)


 * I like the concise thing. Maybe side discussion should be at the user's talk. Anna Frodesiak (talk) 08:37, 30 June 2017 (UTC)


 * FWIW, 356 editors have posted at ORCP. 190 commented 3 times or fewer. The vast majority of posts are made by about 12 regular participants who all appear to be very experienced editors and/or admins. These users tend to respect the recommendations for commenting, while others appear more often to be expressing their own opinion of the candidates rather than making an objective estimate of the candates’ chances at RfA. RfA will never be an entirely  friendly place; to continue to be successful, ORCP needs to avoid becoming a mini, precursor RfA. In my opinion, ORCP should never be cited at RfA. Kudpung กุดผึ้ง (talk) 15:03, 30 June 2017 (UTC)

Is a 0-10 scoring system the best method

 * I would suggest a short bolded summary as an additional option to a 0-10 scoring system, 30-40 words: main strength, main areas to improve, if you would back candidate's RfA in a given time frame, and any additional note. Then the main body can cover this in more detail. I have replaced my score on Chrissie's with this. It is more informative to the candidate, and would remove a completely arbitrary scoring system, which seems to have no methodology. It would seem simpler to simply say whether a commentator would back, with provisions if areas seem weak. RfA is so unpredictable that these numbers are effectively meaningless. As has been said above, Goldring would probably have got 0's and probably 1's. This would seem to be more nuanced. Also this additional option may attract highly constructive comments from collegues who would otherwise feel uncomfortable with opening with an arbitrary figure, basically picked out of the air. Thoughts? Irondome (talk) 17:41, 30 June 2017 (UTC)
 * Yo I agree with you there. Just, FYI, there's a script  that some editors use to get the results they do. Not sure how, or whether it- for example- calculates a 'score' for you based on AAS-style factors- as I have not installed it. Cheers, &mdash;  fortuna  velut luna  17:53, 30 June 2017 (UTC)
 * I just changed mine to a more basic summary. I like this method better than the scores for the reasons I pointed out above. I'm torn between bolding and not bolding because something I do want to make harder is ORCP being used against people in RfA, but I also think a quick and dirty summary could be useful for the candidate. I've kept it bold for now, but might just remove it all together at some point. TonyBallioni (talk) 18:10, 30 June 2017 (UTC)
 * Tony, can we get a consensus going that ORCP is not to be mentioned at RfA, because they are completely separate (to my mind) entities. Any ORCP comments should be struck. Irondome (talk) 18:15, 30 June 2017 (UTC)
 * I'd certainly get behind that, but I don't think it'd be enforceable. Only 'crats can clerk RfA, and to my knowledge they haven't really done so recently. TonyBallioni (talk) 18:38, 30 June 2017 (UTC)
 * Anyone can perform clerking duties at requests for administrative privileges, but bureaucrats aren't likely to be reverted, whereas clerking actions undertaken by others will only stick around if no one objects enough to revert. isaacl (talk) 23:17, 30 June 2017 (UTC)
 * I don't agree with making this a personal opinion. There are too many commenters (and candidates, for that matter) who aren't willing to invest effort into understanding what the community is looking for when granting administrative privileges. For this poll to have value, it needs to be more than a straw poll of whoever shows up at the time: there should be targeted feedback on the various areas that matter to the community when evaluating suitability to be an administrator. isaacl (talk) 18:17, 30 June 2017 (UTC)
 * Merely adding a summary instead of a number does not turn this into a personal opinion. I do not quite understand your point Isaacl. Irondome (talk) 18:24, 30 June 2017 (UTC)
 * "... if you would back candidate's RfA in a given time frame ..." is the phrase that I think poses an issue. That isn't really a factor for ORCP, and when poll respondents heavily consider their own support or lack thereof, we often give a flawed view of the chance of success. Keep in mind that population at ORCP is very invested in finding more admin candidates, so we'll skew more toward support than the general population of RFA. ~ Rob 13 Talk 18:33, 30 June 2017 (UTC)
 * People will do that anyway regardless of what the instructions say because we let our biases show through even if we try hard not to. I'll agree it should not be encouraged, though. Having a summary vs. a number would IMO be a vast improvement over the current process. I suppose the question would be what should be in the summary. TonyBallioni (talk) 18:37, 30 June 2017 (UTC)
 * As I mentioned above, I prefer targeted feedback to a numeric score. I believe though that the feedback should be based on the areas that the community have historically taken into consideration when approving someone for administrative privileges, and not just an opinion on whether or not the commenter would support the candidate. (The commenter may well not support the candidate personally and yet think he/she would be approved by the community, or vice versa.) isaacl (talk) 21:51, 30 June 2017 (UTC)
 * Nonsense. You say ORCP is a "completely arbitrary scoring system, which seems to have no methodology" and guess what? That's what RfA is. "Goldring would probably have got 0's and probably 1's." Exactly. I don't think he should be an admin. Our self-selectors haven't done the best job anticipating the vote of the hoi polloi that shows up at RfA. Taking the number away only makes the system more fuzzy. I think ORCP numbers usually track pretty well to RfA results. Your suggestion that any discussion of ORCP at RfA be prohibited not only seems to indicate that you're trying to protect subpar candidates who show up at RfA after criticism at ORCP, but it's also impossible. Anyone could read those results. Chris Troutman  ( talk ) 18:53, 30 June 2017 (UTC)
 * Thank you for your graceful response Chris. Irondome (talk) 19:16, 30 June 2017 (UTC)
 * What, you were expecting one? :P RileyBugz 会話 投稿記録  20:24, 1 July 2017 (UTC)


 * 356 editors have posted at ORCP. 190 commented 3 times or fewer. The vast majority of posts are made by about 12 regular participants who all appear to be very experienced editors and/or admins whose 'scores', based on their experience as regular participants at RFA, can probably be relied upon.


 * When Anna created this project getting on for two years ago, I welcomed it enormously. I assumed it to be an informal, optional thing, particularly useful since Editor review was deprecated in May 2014. The commenting has recently become sometimes even more detailed than votes at RfA which IMO defies the object of the exercise. There is really no need for a candidate to offer a riposte or vehemently defend their need for adminship, and there is rarely a need for the commenters to argue among themselves unless someone is pointing out an inappropriate or grossly inaccurate statement.


 * If bureaucracy is now to be introduced to tighten this project up, and if the new trend at RfA to produce long discussions on the talk page (or voting page footer) is to be reflected here by inviting commentary to spill over onto this talk page, I think I would lose my interest for it. Kudpung กุดผึ้ง (talk) 03:34, 1 July 2017 (UTC)


 * I agree that there should be a return to the original, stripped down approach Chris. I (almost) always agree with your reasoning on various topics because they are based on common sense and a great store of experience and knowledge in what has worked and what has not. So I would support a back to basics approach, and withdraw some of my comments which would, if taken to their conclusion would inevitably involve more wasted bytes. However there does seem to be a growing tendency for more detailed comments. I cannot account for this. However, I cannot see any great harm in providing a very short initial summary in lieu of the numerical scoring system for comments who may wish to use such an approach, as long as it is clear, simple and acceptable to candidates and the community. Simon. Irondome (talk) 03:53, 1 July 2017 (UTC)
 * Whatever the best way to go about this project going forward is, I think there are enough comments on this page that demonstrate that there is frustration with what ORCP currently is. I also think the current one running has become messier than some recent RFAs, which should not be the case. TonyBallioni (talk) 04:21, 1 July 2017 (UTC)
 * I think there is a key voice missing: what do the candidates think? Are they finding the feedback useful? Is there another format that they would consider to be more helpful? isaacl (talk) 06:20, 1 July 2017 (UTC)
 * There are many editors who like to make pronouncements of their opinions, and who feel they have something important to say, which leads them to expand their comments to greater lengths. English Wikipedia's consensus tradition puts us in a bind, as historically as long as enough people drop into a conversation to insist on doing something, then that's what happens. So if lots of people come by the candidate polls and insist on leaving long comments, then it's pretty hard to say that consensus is against them. In the past, when I tried to gently dissuade commenters from arguing with each other, some responded they thought the discussion was on topic and fine. (When I suggested above that side discussions move to this talk page, it isn't because I want these threads on this talk page, but it doesn't work to tell people not to discuss what they want to discuss: they just proceed anyway.) I recently proposed that commenters should be barred from adding responses below comments from others. Perhaps we can reach a consensus now on eliminating threaded conversation in the poll? isaacl (talk) 06:20, 1 July 2017 (UTC)
 * , are you from the future in an alternate timeline or something? Editor review was depreciated in 2014, so it can't be depreciated 7 years from now in 2024.  I spend too much time on Reddit.  —   Gestrid  ( talk ) 15:30, 1 July 2017 (UTC)
 * Corrected. I spend too much time on Wikipedia. Kudpung กุดผึ้ง (talk) 15:37, 1 July 2017 (UTC)

I've been mulling this over and I think after Chrissymad's poll here, which more or less resembled a traditional "badger the oppose !voter" part of an RfA, that ORCP might have run its course. It was a nice idea to begin with, it was fresh, and people put in one-line brief opinions. Unfortunately I can't think of any way you'd get back to that - I can only see bad things happening if we start moderating the "wrong" sort of comments.

As an alternative, I've made a point this year of searching for RfA candidates and putting forward as many nominations as I can, and added my name onto Request an RfA nomination. Perhaps we could promote that a bit more, though the number of editors that have put forward many successful RfA candidates and know the process is quite small, and it would be extra work on their shoulders. Ritchie333 (talk) (cont)  11:47, 10 July 2017 (UTC)
 * What do you think of eliminating threaded discussion? I don't think it's a magic cure but I believe it would help. isaacl (talk) 16:04, 10 July 2017 (UTC)
 * That might work, both for ORCP and RfA. Each editor gets their one vote, no discussion, no responding to anyone. I'd support it. The ballot box shouldn't be a home to water cooler discussion. Chris Troutman  ( talk ) 19:01, 10 July 2017 (UTC)


 * . I would support a simple number at ORCP, no discussion. If they want to give advice, would they then just visit the user's talk page? Would we care if that happened, or would that be a good thing because the user could end up with interactive guidance? Anna Frodesiak (talk) 23:35, 10 July 2017 (UTC)
 * Guidance is much better than numbers, as it tells a candidate, in one place, what they could improve on and if they are ready for adminship and why. RileyBugz 会話 <sub style="color:#D7000B;">投稿記録  23:38, 10 July 2017 (UTC)

Kudpung puts it well with "...ORCP needs to avoid becoming a mini, precursor RfA. In my opinion, ORCP should never be cited at RfA..."

The problem is all the discussion, right? It can be nasty, pile-onish, long, unhelpful, and worst of all, RfA-like. That last thing is just what we don't want.

The point of ORCP is simply to give someone the odds. That is what numbers do. Getting rid of comments gets rid of an obvious problem and leaves behind the essence of what ORCP is supposed to do.

So, split off the comments. Those can happen in a two-way at the user talk, where discussion can be helpful, guiding, even mentoring. That is better than a one-way dressing down here, where the user hears what's wrong.

Strip down ORCP to what it is supposed to be, a poll. Gallup doesn't ask for speeches. They are looking for an end figure.

I'm going to make a support/oppose proposal below.

Should we make a support/oppose proposal below, or do we need more debate?

Anna Frodesiak (talk) 00:25, 11 July 2017 (UTC)
 * Hello Anna. Hope you're doing well. If you're going to make a support/oppose proposal, then one option may also be, if you think appropriate: "Strip down ORCP to only guidance, specifically excluding the number-driven rating system." Thanks. Lourdes  01:08, 11 July 2017 (UTC)
 * If moving discussion to editor talk pages is just going to result in the same multi-threaded arguing that is going on now, then personally I'd prefer keeping comments here but eliminating discussion. On the candidates' talk pages, most of the onus will fall on the candidates to manage the discussions, and since they're trying to be open to feedback, they're at a disadvantage.
 * I appreciate what you're saying about succinctness. I was going to try to propose a change to the instructions to allow for a brief comment in lieu of a score, but I had difficulty trying to figure out a way to do so while still keeping the instructions lightweight. Asking for a probability in the form of a score keeps the instructions simple. But as I've said before, I find the comments a lot more helpful than an unadorned score, where you have no idea why you've been rated high or low. Yes, politicians look at an approval rating to see if they're doing well, but when trying to improve, they turn to focus groups or more extensive surveys with more questions.
 * Could those who have, in the past, spoken up in favour of a simple score please weigh in? We could benefit from hearing your viewpoints first-hand. isaacl (talk) 01:39, 11 July 2017 (UTC)
 * When I did my poll I felt a bit hindered on this page and took it to a user talk page to discuss further with one of the polled persons. There was a learning point in there for me that couldn't be had here due to requested terseness of the discussion. Of a review of the most-recent discussions, inclusive of a couple of archives (where score average > 1), the only sore thumbs from a length perspective are Chrissymad's and White Arabian Filly's. There are a handful of users who consistently give detailed reviews who might be told to back away from the length (SoWhy comes to mind), but otherwise, all seems well. I do have a few suggestions: 1) Dump this numbers scheme and ask for "Likely Pass", "Discretionary Range", "Likely Fail", and "whatever else comes below Likely Fail" bolded comments. The numbers don't mean anything and aren't used consistently from what I can tell (though I see common ranges; 0-3, 3-6, and 7-9 or so, with rare 10s). 2) Suggest that more than N responses in sequence (as in, A replies to B replies to A replies to C) to a particular poll opinion should be taken to the user's talk page who had the original opinion. I found value in my extended discussion, and I would guess there are the rare editors who might also find value in them (though I note at least one editor who threw up a poll twice and did not like what he heard the second time, recently). I do agree that ORCP probably should not be cited at RFA. There might reasonably be worth in allowing for certain such citations (SoWhy's and HJM's comments here come to mind, even though I cited him/her earlier for exceedingly lengthy comments). --Izno (talk) 02:34, 11 July 2017 (UTC)
 * The most obvious time I would want to cite ORCP in an RfA is in a nomination statement eg : "I've been looking at Joe Blow for some time as a potential admin, and three months ago he ran a poll at ORCP to a favourable response. So let's put that to the test." <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk) <sup style="color:#7F007F">(cont)  09:55, 11 July 2017 (UTC)

I've got a sinking feeling that the latest poll, from InsertCleverPhraseHere, is going to go the way of previous ones. I might have a quieter word with him on talk / email when I get a mo, if he's interested. <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk) <sup style="color:#7F007F">(cont)  12:02, 11 July 2017 (UTC)

Break
Discussion's already dead, but hell, may as well put in my 2c. I think a 0-10 rating system is too broad and too open to interpretation. The people coming here should know what numbers to expect and at the moment what I'm seeing is a mix of: There are a few ways to stop the ambiguity of these ratings, and my top choice would be narrowing it down to a 1-5 voting system styled similarly to those surveys that go like "How do you agree with this statement? Vote between strongly disagree and strongly agree". Look around in the archives. People are giving numbers all over the shop. On some people are giving ratings ranging from 6 to 9, others could have anywhere between ratings of 2s and 6s. What's the difference between a 2 and a 3 or a 7 and an 8? And then decimals come in. What's the difference between a 7.5 and a 7 or an 8 or a 2.5 and a 2 or a 3? You're giving the impression that you want them to run/not run (respectively) so why muddy the waters with unnecessary points (especially 7.173925)? People come here for stone hard opinions and numbers so giving easily digestible content would be the first step to getting this process back on its own two feet. I'm probably talking through my hat, but oh well.  Anarchyte  ( work  &#124;  talk )  07:37, 24 September 2017 (UTC)
 * 0-3: No chance
 * 4-6: Possibly
 * 7-10: Go for it
 * 1: Strongly disagree with you running
 * 2: Disagree with you running
 * 3: Neutral with you running (you have your pros and cons and if you were to run there's no guarantee you'd pass (or fail))
 * 4: Agree with you running
 * 5: Strongly agree with you running
 * Thanks, . The concern I've expressed elsewhere after this conversation died down has been that ORCP has taken the worst parts of RfA and lowered the barrier to people being rude. ORCP typically gets two types of commenters, those who don't know much about RfA but who give their opinion because they can, and those who have strong opinions on what they want adminship and RfA to become. Even though both are acting 100% in good faith, neither group actually provides valuable insight for the candidate about their chances. The thread below started when I saw multiple weeks old overwhelmingly negative polls still up. ORCP has become a place to voice negative opinions about people with zero consequences or odds of getting called out. I think it's outlived it's usefulness for anything other than discouraging those who have no chance of passing from standing, but I also think your suggestions might be a move in the right direction towards trying to see if ORCP is salvageable. TonyBallioni (talk) 13:15, 24 September 2017 (UTC)
 * As I said above, numbers really don't aid in understanding. Bold "do it", "don't do it", and "maybe do it" are enough to get the point across, which is the most important part of the poll. --Izno (talk) 01:53, 25 September 2017 (UTC)

Peer review
I was wondering, whatever happened to it. I dont want to run for adminship, no wish at all. At least for now. (No idea what will happen after an year: I might want to become an admin, or I might get busy enough IRL to semi-retire from the project. I dont have a crystal ball after all.) But I wanted to get a "peer review" of myself, wanted to know where I stand as an editor in general, and what should I change/improve to become a better editor. I already consulted Fortuna IM regarding that a few weeks ago. He gave me an appropriate feedback/review. But that was just one person's opinion. Is there any place where I can get a thorough peer review? Thanks, — <span style="font-family: monospace, monospace;">usernamekiran (talk)  04:56, 14 October 2017 (UTC)
 * The Editor review page describes what happened, but in short, there weren't enough people doing reviews and requests were languishing there for years. Your best bet is to keep asking editors you feel can give you constructive feedback; others who see your request (in whatever venue) might also respond on your talk page. isaacl (talk) 08:49, 14 October 2017 (UTC)

Archive length
, was unaware of the previous conversation. I upped the archive length to 7 days because, to be blunt, I think it is simply cruel to have these archived conversations stay up for two weeks after closing. Most of them of late have been overwhelmingly negative. Keeping them up for two weeks serves no purpose that I can see. Given, I'm also the most openly critical editor of ORCP we have, so I might be a bit biased here, but I think we do need to take into account the human feelings of the people who put their names forward here. TonyBallioni (talk) 06:13, 24 September 2017 (UTC)
 * FWIW I support the higher archive rate. I tried to change it a few months back but was ultimately undone.  Anarchyte  ( work  &#124;  talk )  06:20, 24 September 2017 (UTC)
 * I'm pretty sure I undid the change, for the same reason as described below. isaacl (talk) 06:26, 24 September 2017 (UTC)
 * My reasoning is that this is an informal optional poll, and so participants shouldn't be expected to be monitoring the page daily. Accordingly, I feel one week is too short a time to expect someone to have already read all the feedback.
 * If there is agreement the feedback is insufficiently constructive, we should try to address this issue directly. For example, can we gain consensus on eliminating threaded conversation? How about limiting the length of comments? Or maybe eliminating comments entirely as Anna has suggested? Or should there be a limit on the total number of commenters, after which the poll can be closed? What other ideas does anyone have? isaacl (talk) 06:24, 24 September 2017 (UTC)
 * My suggestion is to mark the entire page as historical and be done with it, though I doubt that will get consensus.The reason I think a week archive is fine is that by the time the poll is archived it's typically been open for a few weeks. An archive rate of one week typically means the poll will be open for at least three weeks, which is still too long in my opinion, but I don't think there's a lesser time that'd be acceptable to set the bot to. TonyBallioni (talk) 06:35, 24 September 2017 (UTC)
 * Actually, Anna has asked in the past if the poll should just be closed, so I think there's a reasonable chance of this happening in the intermediate future. (Side rant: it's another case of why we can't have nice things. The instructions have virtually begged commenters to be concise and constructive, but people do whatever they want.) How about capping the total time for comments at one week, and leaving the poll up for another week? With a fixed duration, participants will know when to expect all feedback to be in, and will also know when the poll will be archived. isaacl (talk) 06:43, 24 September 2017 (UTC)
 * It's not great when the page where people are meant to ask about their chance at passing says "The result of a poll may differ greatly from an actual RfA". I think the page is great on paper, but the rules/guidelines for it aren't apparent enough. If this is to get any better I think we need to put some numbers in stone and 1 week for polling+1 week for archive would be a good place to start.  Anarchyte  ( work  &#124;  talk )  07:16, 24 September 2017 (UTC)
 * I concur. That would also set the bot at a 1 week archive. Once the poll was shut down it'd be the last comment so it would stall the bot for a week. TonyBallioni (talk) 13:00, 24 September 2017 (UTC)
 * Ideally, everyone ought to realize that a small sampling of opinion isn't necessarily indicative of how an actual request for administrative privileges would proceed. However given all the hand-wringing that occurred when the results of a poll did differ from an RfA, the reminder may be helpful for some. There is a significant number of editors who believe that more detailed rules = greater bureaucracy = something they don't want for an informal, optional poll (and given that more rules require more enforcement, it's a fair concern). Any suggestions for clarifying the instructions while keeping them concise are welcome. isaacl (talk) 13:44, 24 September 2017 (UTC)

How about extracting the list of all the people who ever voted here at ORCP and sending them a newsletter telling them to get their act together? Kudpung กุดผึ้ง (talk) 14:28, 24 September 2017 (UTC)
 * Sure, it's worth a try. (They've been implored on this talk page to follow instructions, but guess it can't hurt.) Given the shifting audience of commenters here, a complete list would probably reach a lot of people who are no longer providing feedback; should the list be limited to more recent contributors? isaacl (talk) 15:06, 24 September 2017 (UTC)
 * The problem is not with the participants, I don't think. Or at least not entirely. The purpose of this page is to tell people whether they have a realistic chance of passing an RfA, regardless of whether we think they should pass, and many of the people coming here have no hope at all (but at least we're reducing the number of actual RfAs that are open and shut within an hour). My suggestion is that the candidate should decide when they've had enough feedback and we should expect them to to close it when they've made their decision, and obviously premature polls should just be removed or summarily closed. Likewise, comments that aren't helpful to the candidate in deciding whether to run or not should just be removed. HJ Mitchell  &#124; <span style="color:Navy; font-family:Times New Roman;" title="(Talk page)">Penny for your thoughts?  15:08, 24 September 2017 (UTC)
 * The instructions currently state that candidates can close or withdraw their poll; we could emphasize this a bit. But I also think commenters should exercise restraint; if they see candidates have received sufficient feedback, they can forego adding more comments, and allow the poll to be archived. As for removing comments, unless someone gets appointed to do so and their judgment is trusted, it's tricky to remove all but the most blatantly off-topic posts. Attempts to gently nudge people away from digressions have typically been met with responses that the discussion is relevant. isaacl (talk) 15:20, 24 September 2017 (UTC)


 * Without getting into the meta discussion as to what the problem with ORCP is (or if there is a problem), I've updated the instructions to make it clear how long the poll stays on the page using the time lengths suggested above. If people are fine with this wording, we can up the bot to 7 days, which will give individual polls a week on the ORCP page after being closed before they are sent off to the archive. TonyBallioni (talk) 15:44, 24 September 2017 (UTC)
 * Well, I was waiting for a larger consensus to emerge... Can anyone else interested please comment? isaacl (talk) 16:45, 24 September 2017 (UTC)
 * Yes, I hadn't planned on changing the bot for a few days. I'm always fine being reverted, but also think adding a bold change that we'd talked about here and seeing if there are objections is a way forward. TonyBallioni (talk) 20:33, 24 September 2017 (UTC)
 * I appreciate some people like this approach; personally, once a portion of text is under discussion, I prefer to leave it as is until a consensus is reached, with changes proposed within the discussion thread. isaacl (talk) 20:44, 24 September 2017 (UTC)
 * I believe I have previously advocated for a shorter archival timeline--not because of anything said in specific discussions but because the long timeline causes these discussions to be on this page long past the time they are useful to anyone (especially the poll responders). I think it is quite reasonable to expect a user who has a poll open to be observant of that poll. It is a straw man (or a red herring?) to suggest we require a user to check daily, but they should be willing to check every couple days or even weekly. (Though, if they aren't checking Wikipedia at least daily, their RFA week will surely be hell without considering all the much-maligned behavior of the !electorate.) --Izno (talk) 02:00, 25 September 2017 (UTC)
 * A formal request for administrative privileges and an optional, informal poll are two different things, though. Various people have expressed the desire to keep this poll lightweight, in contrast to RfA. With the current mode of operation, with an open-ended poll, the participant must continually check regularly, without knowing when feedback will stop coming in. Under these circumstances, I personally think it better to keep the poll unarchived for a greater length of time, as it's typical after a couple of weeks to stop checking back as often and some late feedback can get missed. If there is consensus to shift to a fixed-period poll, then the participant will know exactly when to check back, and so the need to keep the poll unarchived is reduced. isaacl (talk) 05:59, 25 September 2017 (UTC)
 * Not directed at your main point, but: I would hope that RfA candidates would be competitive without being so addicted to this website, and so devoid of other things going on, as to check this website every day for any extended period of time. -- Ajraddatz (talk) 06:56, 25 September 2017 (UTC)

Limit on poll duration
Although personally I'd prefer there'd be a consensus based on more participation, in the spirit of a lightweight process, I suggest proceeding with setting a poll duration of 1 week (as already specified in the revised instructions). If anyone has any concerns, please do raise them; thanks! isaacl (talk) 03:17, 3 October 2017 (UTC)
 * I don't object to this; this actually would render a quicker assessment than earlier, which I suspect would be beneficial for prospective candidates. I would prefer wordings that include "may" instead of "should". Also, currently, the closure paragraph reads a bit unstructured (especially the second line). Thanks.  Lourdes  05:13, 4 October 2017 (UTC)
 * I think it is better not to use "may", which I believe will be interpreted in the same way as most Wikipedia discussions: as long as there is some ongoing discussion, the thread should remain open. This will defeat the purpose of setting a deadline for comments, though. Can you expand further on what structure you may consider helpful? The second sentence starts with stating when a poll should be closed, and then lists two cases where it may be closed earlier. Personally, I think the two exceptions can be dispensed with, leaving it to editors to exercise their judgment on when a poll can be closed early. However, I appreciate there are some who may feel differently. isaacl (talk) 17:46, 4 October 2017 (UTC)
 * Proposed: "Polls may be closed after seven days without any closing statement, or earlier if there is unanimous agreement that the candidate has no chance at being granted administrative privileges, or if the candidate has filed a request for adminship." Thanks.  Lourdes  02:36, 5 October 2017 (UTC)
 * How about: Polls are normally closed without any closing statement after seven days. They may be closed earlier if there is unanimous agreement that the candidate has no chance at being granted administrative privileges, or if the candidate has filed a request for adminship. Alternatively, the portion of the second sentence regarding an RfA being in progress could be removed; I don't think this needs to be spelled out. isaacl (talk) 02:49, 5 October 2017 (UTC)
 * Your structure seems good. I agree on your alternative suggestion too. Thanks.  Lourdes  04:46, 5 October 2017 (UTC)
 * I'll just chime in to make this a bit easier. I agree that the final suggestion here is quite nice. —  Insertcleverphrasehere (or here)  05:06, 5 October 2017 (UTC)
 * OK, I've [//en.wikipedia.org/w/index.php?title=Wikipedia:Requests_for_adminship/Optional_RfA_candidate_poll&diff=804092439&oldid=803704231 implemented the proposed change]. If there is any more feedback, please join the discussion! isaacl (talk) 17:57, 6 October 2017 (UTC)

A bit more on archiving
@isaacl and Lourdes. What about obvious SNOW closes like the one we just had? per the above change it should obviously be closed early (as has been done), but when should we archive it? Should we leave it up for a week and let the bot archive or just throw it into the trash where it belongs archive? I'm open to either, but curious what you guys and anyone else thinks. —  Insertcleverphrasehere (or here)  01:51, 16 October 2017 (UTC)
 * Here's a related discussion on removing a poll. In this case, since the candidate removed the poll, it can be assumed that the results have been seen, and so there is no issue in archiving it immediately as a courtesy. If the candidate hadn't removed the poll, then I believe it should be left for the bot to archive (just let it do its job). isaacl (talk) 02:24, 16 October 2017 (UTC)
 * Accordingly, I have archived the poll manually. isaacl (talk) 02:34, 16 October 2017 (UTC)
 * Hello, . Perhaps we could add a line to the instructions section mentioning that snow closes may be manually archived immediately. Thanks.  Lourdes  03:10, 16 October 2017 (UTC)
 * Sounds good. —  Insertcleverphrasehere (or here)  03:25, 16 October 2017 (UTC)
 * No, if the candidate has not had a chance to see the closure, it should be left for the bot to archive. Basically by default, let the bot do its job. isaacl (talk) 03:34, 16 October 2017 (UTC)

Results subpage
The results subpage has fallen behind. I'm so sorry to say that I'm just too busy off-wiki to keep it up. If others want to take over, great. If not, oh well. :) Anna Frodesiak (talk) 17:51, 11 October 2017 (UTC)


 * I think it's important to keep  it  up. It  shows that we're not often wrong. I can't do it because I'm generally  winding  down my  interference in most  meta areas - I  want to  shift  the focus of my  Wikiwork. Kudpung กุดผึ้ง (talk) 09:04, 14 October 2017 (UTC)
 * I will be happy to bring it up to date. Cwmhiraeth (talk) 10:03, 14 October 2017 (UTC)
 * Cwmhiraeth, thank you, thank you, thank you, thank you, thank you! :) Anna Frodesiak (talk) 01:36, 16 October 2017 (UTC)