Wikipedia:Requests for adminship/2021 review/Issues

This request for comment is part of a community review of the process for requesting administrator permissions. The goal of this discussion was to evaluate the community perception of current problems. Editors proposed statements about the Request for Adminship (RfA) process, which the community debated in separate sections, focusing on problems rather than solutions. The purpose of this close is to articulate the community’s consensus on RfA’s problems so that the next phase can consider what proposals will best address these identified issues.

As a high-level summary, we divided the proposals into four broad categories: clear consensus, rough consensus, no consensus, and clear opposition. Items in the first two groups (E, F, G, I, J, O, P, and U) will pass on to Phase 2 for consideration of the problems, and crafting potential solutions for them.


 * Clear consensus
 * E: Corrosive RfA atmosphere
 * F: Level of scrutiny
 * I: Standards needed to pass keep rising
 * J: Too few candidates
 * O: “No need for the tools” is a poor reason to oppose


 * Rough consensus
 * G: Lifetime tenure (high stakes atmosphere)
 * P: Admin permissions and unbundling
 * U: RfA should not be the only road to adminship

*: These proposals have a numerical rough consensus but have very low participation relative to other issues presented.
 * No consensus
 * K: Too many questions
 * M: The RfA process is biased against long-term editors*
 * N: The lack of agreed upon RfA criteria is a major part of the problem
 * R: Time commitment for RfA
 * T: Expecting the unexpected*


 * Clear opposition
 * A: There is no issue
 * B: Not enough like a discussion
 * C: Too much discussion
 * D: Administrator is undesirable position
 * H: RfA reflects a declining editor base
 * L: Who participates
 * Q: Splitting the admin role
 * S: Too few trusted and experienced nominators

There is a clear consensus that the atmosphere of RfA is a significant problem (section E). While community oversight of the administrator role is important, editors believe the level of scrutiny is inappropriately high (section F) and that the standards of the community have risen without a clear justification (section I) and with sometimes poor rationales (section O). The combination of these and other problems discourage qualified editors from requesting adminship (section J).

Participants came to a rough consensus on a number of points, and future proposals to address them have a good chance of success as long as opposition concerns are accommodated. Editors largely agree that alternate routes to administrator should be explored (section U) but there are substantial concerns about the development of a two-tiered adminship system or creating a hierarchy within the administrator corps. There is sufficient consensus to explore unbundling (section P) but editors are suspicious of separating the core delete-block-protect tools, and many editors are suspicious that there are any tools left that are worth unbundling. There is a rough consensus that the indefinite tenure of administrators contributes to the corrosive atmosphere and high scrutiny at RfA (section G, cf. sections E and F) but the recent failure of a proposal to impose administrator term limits suggests that proposals to address this point will face wider opposition than seen here.

There is no consensus that the time commitment for RfA is a problem (Section R) but editors generally agree that expectations about candidate availability are too high. Limits on the number of questions are unlikely to succeed as there is no consensus that the number of questions asked is a substantial problem (section K). There is no consensus that the lack of clear criteria is a problem (section N) but there is a rough consensus reaffirming the tacit criteria of civility and reasonableness (see section O, cf. 2015 RfA clerking RfC and bureaucrat discussions). There is some agreement that editors with long tenure are at a disadvantage in RfA (section M), but there is no consensus that this is a problem given the opposition and limited participation. Editors were split on whether the unpredictable nature of RfA was a problem (section T), but participation on that statement was limited (cf. results of N).

Editors reject the ideas that adminship is undesirable (section D), that the problem stems from a declining editor base (section H), or that there are too few nominators (section S). There is a consensus against the claim that RfA participants are ill suited to the task (section L) though these discussions saw relatively limited participation. The current balance of discussion-like process (section B) to vote-like process (section C) seems to be an appropriate compromise as proposals on the issue saw similar levels of opposition.

There is a consensus against the idea that there are no problems to be resolved (section A). In evaluating consensus for future, substantive proposals, editors should consider the strong consensus that problems exist and need to be resolved.

Summaries of each section are provided in the following section. Thank you to all who participated!

Signed,
 * Wugapodes (talk) 19:25, 8 October 2021 (UTC)
 * Primefac (talk) 01:12, 9 October 2021 (UTC)

Summary by section

 * A. There is no issue
 * Participants frame the problems of RfA as existential: failure to recruit more administrators threatens the long-term viability of the project. Evidence discussed shows that Administrators are leaving the project or being desysoped for inactivity, new administrators are being promoted at a record low rate, and the number of articles and maintenance tasks is growing. These problems together produce a chronic net loss of administrators and the shrinking number of remaining admins are tasked with an increasing maintenance burden. As this trend continues, critical maintenance burdens will experience growing backlogs until the project ultimately becomes unmaintainable.Participants were largely divided on whether the tipping point is in the future or whether it is already upon us. A number of editors point out that critical processes like speedy deletion, copyright deletions, and page protection are maintained by a handful of administrators and often only one or two (see bus factor). When these editors take breaks or retire entirely, these processes experience backlogs that can grow very large before anyone gets around to handling them. While this point gets at the relative urgency of a resolution, editors reach a consensus regardless that we are promoting too few administrators for our maintenance needs and that the bottleneck is the RfA process.


 * B. and C. Balance of vote-to-discussion
 * For every editor who opposes making RfA less vote-like, there is an editor who opposes making it more vote-like. In many cases editors oppose both. Looking at the supports, there are more editors who support making it more like a vote than support making it more like a discussion, but both positions are in the minority compared to those advocating for the current balance. Taken together, the bulk of opinions seem to be for keeping the balance as is with a slight skew towards making it more vote-like.


 * D. Administrator is undesirable position
 * Editors don't see the role as undesirable, but it's not clear that the community sees it as desirable either. Even if editors agreed on the perceptions of the role, the general sentiment is that this is not a major barrier to admin recruitment.


 * E. Corrosive RfA atmosphere
 * Editors agree that the RfA atmosphere is corrosive and a major barrier to admin recruitment. Some editors support the idea of moving to a secret rather than open ballot system, but there was also substantial pushback to the idea found in both support and oppose comments.


 * F. Level of scrutiny
 * While community oversight is widely valued, editors point out that the subjective effects on candidates are a serious harm. The community generally views the level of scrutiny as a serious barrier to candidates stepping forward, as the perception of excessive scrutiny discourages candidates. Editors considered whether a kind-of “statute of limitations” for old disputes would be an effective remedy but there is concern that a too-rigid rule would prevent effective and appropriate scrutiny in some cases.


 * G. Lifetime tenure (high stakes atmosphere)
 * There is a rough consensus that the term length is an issue. However, the two main solutions discussed--term limits and community desysop procedures--have recent global discussions rejecting detailed proposals, and those opposed believe that introducing either of these solutions will cause administrators to no longer feel comfortable performing difficult or controversial actions as it may impact their ability to successfully retain their adminship.


 * H. RfA reflects a declining editor base
 * Editors generally agree that this is inconsistent with the data, and even if it were supported by data, editors point out that the issues with RfA only exacerbate the problem.


 * I. Standards needed to pass keep rising
 * Everyone agrees that standards have risen but there's a lot of split opinions on what that means. Most seem to think that it's a good thing that we matured since 2005 when the bar was on the floor, but a lot of comments focused on particular standards that individual editors considered too high or exemplary of unrealistic standards. Editors in opposition raise the point that “standards” are socially constructed by hundreds of individual standards. While participants agree that the rising standards (or at least the perception of high standards) is a problem, they are less optimistic about our ability to address this directly.


 * J. Too few candidates
 * Editors generally believe that we have too few candidates and that this is a problem in a broad sense. That said, the number of candidates is considered a symptom of the identified problems rather than a problem in its own right. While proposals addressing this point might be worthwhile, the consensus is that addressing other problems will naturally resolve this one, suggesting that it may be a viable metric to evaluate success of future reforms.


 * K. Too many questions
 * There is no consensus that the number of questions is a problem. While some candidates may receive questions from many people, editors are generally against a global limit as it could prevent relevant and important questions. The general sentiment is that if a candidate doesn't want to answer a question, they shouldn't. There seems to be no major concern about the two-question-per-editor rule, so proposals relating to the number of questions are unlikely to be productive.


 * L. Who participates
 * Regardless of the qualifications of RfA participants, editors generally reject restricting participation to a limited set of editors. While some voiced support for minimum requirements similar to autoconfirmed and extended confirmed permissions, a popular democratic process still enjoys general consensus.


 * M. The RfA process is biased against long-term editors
 * There is no consensus that RfA is biased against long-term editors. Those who view the system as biased point out that editors engaged in the project for a long period of time have interpersonal conflicts and have a more extensive history available for those who want to find a handful of edits to discredit them with. Others claim that long term editors only have problems at RfA if there are long-term behavioral concerns, and that issues from many years ago are often overlooked if recent behavior shows improvement.


 * N. The lack of agreed upon RFA criteria is a major part of the problem
 * There is no consensus on this point. Editors believe that the community needs wide latitude in order to properly address the issue of trust and competence of potential administrators. Even for those who believe this may be a problem, many are sceptical of whether much can be done to resolve the issue. There is a general sentiment that crats should understand the community sentiment on what rationales are seen as legitimate, and the results of section O and R may help in clarifying how the community views particular RfA arguments.


 * O. "No need for the tools" is a poor reason as we can find work for new admins
 * Editors are generally suspicious of rationales which focus on a "lack of need". Positions on whether to outright prohibit such rationales are mixed however. Editors point out that it may be a nicer way of stating a lack of trust, but the general sentiment is that the administrator toolkit should be granted to any trusted editor regardless of their immediate plans to use every part of it.


 * P. Admin permissions and unbundling
 * Unbundling is contentious, but apparently viable if done thoughtfully. Editors in opposition to unbundling raise two concerns. First is that the tools are part of a tool kit. When an editor has only one tool of the whole kit, they will use that tool even if it is the worst one for the job (see Law of the instrument). The second concern raised was the number of tools left to unbundle. This point is worth considering, but given the potential options discussed whether we have reached the limit of unbundling is not a settled issue.


 * Q. Splitting the admin role
 * Consensus is against this proposal. Opinions on unbundling are better articulated in response to P.


 * R. Time commitment for RfA
 * The support and opposition were largely responding to two different questions. Those in support pointed to the problem of a community expectation that candidates be extraordinarily available during the RfA week. These editors generally agree that candidates should be free to continue with their normal work, school, and leisure without facing opposition for it. Those in the oppose section generally comment on the time to prepare for an RfA and the relative importance of this point compared to others. These editors agree that while time to prepare for an RfA may be substantial, it is dwarfed by the time and work required to even be considered a viable candidate. Others point out that while time commitment may be an issue, it is probably not a major barrier preventing candidates from coming forward.


 * S. Too few trusted and experienced nominators.
 * There is no shortage of nominators. While a handful of editors make many nominations, many members of the community have put forward nominations. A number of editors state that they would like to see more self nominations of qualified candidates.


 * T. Expecting the unexpected
 * No consensus that this is a problem. While there may be unexpected dynamics at an RfA that can be upsetting to a candidate, the possibility of unexpected questions or opposition is not seen as a major barrier to candidacy.


 * U. RfA should not be the only road to adminship
 * There is a rough consensus that there should be more than one way to gain adminship. Editors raised concerns about the lack of specifics, worrying that consensus around a particular solution may not be possible. The lack of specifics was due to the format of this RfC, but future proposals should consider the potential for conflicting preferences of how a new process would function. The most substantive concern with the principle discussed is the potential for a two-tiered administrative corps, and while a numeric minority, proposals should consider the potential negative social impacts that multiple routes to adminship may create. Editors also raise the point that, absent clarification from the Wikimedia Foundation, any proposals will need to keep in mind the existing requirement that administrators must have community approval through an RFA-like process.

Introduction and purpose
This phase of RfA review will largely follow the same format as the first phase of the 2015 review.

The intent of this RfC is to broadly identify the problem(s), if any, with our request for adminship process. 01:37, 29 August 2021 (UTC)

As was mentioned in the RfC statement, we are focusing on broad issues here. Assuming that any of these issues obtain consensus (as described in the "Instructions for closers" section), there will be a second phase that will follow this one, so that we can focus on how to address issues which have community consensus.

Instructions for voters
The RfC is structured as a list of statements about RfA.
 * Editors may add additional statements, but please do not add duplicate or largely similar statements.
 * Editors may support, oppose, or simply comment on the statements in the appropriate sections.
 * To best represent consensus, participants are encouraged to comment on as many potential statements as they are able.
 * Participants are encouraged to stay focused on discussion of problems with RfA. Related discussion, discussion about the process, or general discussion about RfA is encouraged to happen on the talk page.

Instructions for closer(s)
The uninvolved closer(s) shall have the discretion to determine which potential problems attained consensus. They are encouraged to be lenient when determining which proposals should advance and to note the degree of consensus achieved. The closer(s) of phase 2 will be encouraged to consider the degree of consensus reached here when considering consensus for any potential solutions.

A. There is no issue
Given the number of editors, admins, and our needs as a community there simply isn't a problem that needs addressing.

A. Support (agree)

 * 1) Admin actions that are actually important get handled fairly quickly. We do have admin backlogs, but they are in unimportant areas. All in all, we don't have a dire need for more admins relative to the number of editors (we do need more editors though). There are problems with RFA, but we have more urgent problems than "too few admins". Levivich 07:11, 29 August 2021 (UTC)
 * 2) While the current uptake of new admins may become an issue, and may become an issue soon, and while this may not be sustainable, I agree that there is no issue. Chetsford (talk) 23:10, 30 August 2021 (UTC)
 * 3) The sky is not falling. – ♠Vami _IV†♠  00:41, 4 September 2021 (UTC)
 * 4) RfA reform has been discussed ad naseum with no consensus. We must consider that perhaps the current solution is optimal. There's no mandate in life that an optimal solution to the RfA process has to be easy and hassle free for the applicant. It could just be the least undesirable from among a bunch of less desirable choices. If we are to keep RfA an open process, it will always involve comments and ideas from everybody and this makes things hard. I think it's time to start shelving this topic and pay attention to other things. I'm starting to consider the difficulty of RfA as a net positive: If a person can gracefully handle it, it shows they have qualities to make a good admin. Jason Quinn (talk) 01:50, 6 September 2021 (UTC)
 * 5) Currently, it seems like we are mostly keeping up with demand. Jo-Jo Eumerus (talk) 20:45, 8 September 2021 (UTC)

A. Oppose (disagree)

 * 1) I don't think it's a big problem now, but I also don't think the current RfA pass rate is likely to be sustainable in the long run. The number of genuinely active admins (i.e. admins who do admin work regularly) is about 300. |revision-delete|log-delete|restore|re-block|unblock|re-protect|unprotect|rights|merge|import|abusefilter|contentmodel Here are the stats of logged admin actions for the last year. While high profile admin areas have plenty of people processing them, lower profile ones often have one admin or a small group of admins. We've had 6 successful RfAs so far this year and the lowest annual figure so far is 10. Since that figure isn't nearly enough to maintain the figure of 300 active admins, the project is still relying to a large degree on people who passed RfA more than a decade ago when the pass rate was higher.  Hut 8.5  08:50, 29 August 2021 (UTC)
 * 2) The current rate of promotions is just not sustainable in the long run.Jackattack1597 (talk) 11:59, 29 August 2021 (UTC)
 * 3) It is a problem that the vast majority of admins have been here forever, and we don't have a countering influx of new people who become admins after a reasonable (<12) number of months. —Kusma (talk) 12:56, 29 August 2021 (UTC)
 * 4) I think the question of whether we have enough/not enough admins is only a tangential issue to whether or not RfA needs fixing. That is, we could use more admins, especially in certain areas, but I'm not sure we need as many as we once did, and I'm not sure it's a dire issue. But that doesn't mean the fact the RfA process discourages candidacy isn't a problem we should try to solve. If our process is preventing candidacies of people who would be good admins, we should want to fix that even if we had plenty of admins. —valereee (talk) 13:33, 29 August 2021 (UTC)
 * 5) The declining number of admins is deeply concerning. RfA isn't the only cause of that, but it seems like a probable one. &mdash;  Rhododendrites  talk \\ 15:31, 29 August 2021 (UTC)
 * 6) Unsustainable in the long run, and visible in places like "Requests for closing discussions/closing AFDs." Unglamorous stuff that requires a lot of effort and should be treated delicately... Those are the places where things languish and are un-answered for many weeks past soft deadlines. Would more admins help these areas? If only because it means there's less work to each individual admin, leaving more time to focus on the individual tasks which are themselves very taxing! —  Shibboleth ink  (♔ ♕) 21:58, 29 August 2021 (UTC)
 * 7) I think there's very few admins who would not greatly appreciate a shorter backlog to work through. That the tasks are currently getting done doesn't mean that they will continue to get done; and in many crucial areas, particularly in the dispute resolution arena, there's already not enough participation. That said, I think the problem isn't all at RFA as narrowly defined; there's also perception issues around what adminship entails that makes people less likely to want it. Vanamonde (Talk) 07:22, 30 August 2021 (UTC)
 * 8) There are several problems in our failure to recruit new admins. One is that the wikigeneration gap is unhealthy in a volunteer community. We have only 24 admins whose first edit was since the start of 2015, that is not good for community cohesion, especially as many from a decade before that were becoming admins in months. We also lose the editor retention benefit of appointing people as admins - new admins do tend to stay here long term. As for sustainability, if we appoint ten new admins in a year, to maintain a pool of a 1,000 admins half of whom are active at any time, we need the average new admin to remain an admin for 100 years and be active for half of them. Given current human longevity this is an unrealistic scenario.  Ϣere Spiel  Chequers  09:44, 30 August 2021 (UTC)
 * 9) Strongly disagree. The decreased number of candidates shows that there is a problem. Z1720 (talk) 14:58, 30 August 2021 (UTC)
 * 10) Our admin population is well below replacement rate, as evidenced by Desysoppings by month, and the pool of admin tasks is not getting smaller. Furthermore, as WereSpielChequers argues, the admin corps should be constantly refreshed with new blood. — Goszei (talk)
 * 11) Many admin-dependent areas/processes have an effective Bus factor of 1 or 2, a fact which won't be apparent to most folks because the vast majority of admin activity is inherently low-visibility.  Of course, it only becomes obvious when these people take a vacation/quit, and the process in question grinds to a halt.  -  F ASTILY   00:20, 31 August 2021 (UTC)
 * 12) The trends here have been well-described by both firm data and research as well as impressionistic evidence for just about any editor who has contributed consistently in the last decade or so.  The community has made accommodations by streamlining many processes, front-ending control of vandalism and numerous other disruptive editing issues, and somewhat increasing the footprint of wider community discussions for some behavioural issues, but these factors will not continue to mitigate serious deficiencies in oversite indefinitely if the current retention:replacement ratio for sysops continues. Some very important areas are carried on the backs of a relatively small number of dedicated technical experts in the administrative corps. I'm not sure to what extent this militates for adopting this or that change to the RfA process: though I recognize there is a need for some reform, I do think we have a vested interest, as a community, in keeping the applicants for the mop under some degree of significant scrutiny.  That said, the scope of the underlying issue is, regardless, pretty hard to deny. SnowRise let's rap 06:07, 31 August 2021 (UTC)
 * 13) WereSpielChequers and Fastily perfectly describe my top three issues with the current admin situation: bus factor, unsustainability and wikigeneration disparity. — Bilorv ( talk ) 16:18, 31 August 2021 (UTC)
 * 14) Per above. We're still afloat right now, but I'm concerned for the mental toll of the workload on admins in some areas and the mental toll of the RfA process on candidates. These problems are likely to get worse in the future. And then there's the lack of adequate continuity planning (see bus factor comment above). &#123;{u&#124; Sdkb  }&#125;  talk 22:06, 31 August 2021 (UTC)
 * 15) We have articles, which is always increasing, yet our admin attrition means that we're losing admins month on month. I'd rather not get to the point that that we're missing important things and have to risk our processes to gain more admins. WormTT(talk</b>) 15:16, 1 September 2021 (UTC)
 * 16) WereSpielChequers said it well. Additionally, the I'd consider the simple fact that people see AfD like running the gauntlet is damaging in itself. We shouldn't have processes people are uncomfortable going through when they just want to help out. /Julle (talk) 16:10, 1 September 2021 (UTC)
 * 17) Have noticed backlogs slowly worsening (or places like AIV becoming backlogged more frequently) and this is not sustainable in the long run, even with new tools/filters/bots helping extend the workload.  Spencer T• C 22:36, 1 September 2021 (UTC)
 * 18) There is a general trend of fewer applicants and fewer appointments. Together with the attrition rate, this is likely to create backlog problems in the future. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  10:00, 2 September 2021 (UTC)
 * 19) As a smart man once said, "RfA is a horrible and broken process". It's only gotten worse in the years since he said it. — python coder  (talk &#124; contribs) 18:58, 2 September 2021 (UTC)
 * 20) Completely agree with bus factor, unsustainability and wikigeneration disparaty. In terms of the latter, I believe that current active editors are more geographically diverse, and women make up a larger share of our volunteers, and as such I suspect getting more of our "newer" editors to run would help close some of these disparities in our admin corps. FemkeMilene (talk) 20:55, 2 September 2021 (UTC)
 * 21) There is absolutely a problem. It is an objective truth that we are losing admins faster than we are gaining them, and very few people are even willing to attempt to go through the RFA process. New admins provide new perspectives and help make up for admins who leave the project. The current situation is not sustainable long-term. Even now, many areas have admin backlogs or are being propped up by only one or two admins (CCI is a good example). Trainsandotherthings (talk) 23:00, 2 September 2021 (UTC)
 * 22) There is a problem; we are losing admins faster than we are getting them. 331dot (talk) 21:12, 3 September 2021 (UTC)
 * 23) There is a huge problem. I still feel the scars from my RfA six years ago. I'm surprised I didn't quit before the 7 day period was over. But I don't see any simple solutions for it sometimes being an awful ordeal. Some editors seem to see being publicly roasted & belittled as a necessary rite of passage. Because there have been some bad admins in the past, they view every candidate with suspicion and an attitude of "Convince me that you won't abuse the tools". Since becoming an admin, I can't say that I have encouraged any editor to go through the RfA process without a lot of warnings on how damaging it can be. In fact, I might have convinced some editors to choose not to go through an RfA just by being honest about how bad it can be. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 01:52, 4 September 2021 (UTC)
 * 24) We will, at some point, need more admins. We have a lot more to manage than we did in 2007, when I went through RfA, yet we have far fewer admin candidates running. We don't want to wait until it's actually a crisis to address the problem; let's do it before that. Seraphimblade Talk to me 04:03, 4 September 2021 (UTC)
 * 25) There are issues with our current process of RfA which this RfC will undoubtedly bear out. In my opinion, mitigating these issues, where possible, is in Wikipedia's best interest.--John Cline (talk) 05:12, 4 September 2021 (UTC)
 * 26) There are many issues with RfA process, too much drama for something that should not be "a big deal". The process leaves potential candidates stress, anxiety and afflict new ones to be recruited. On the contrary, it should be a positive feedback to be a better editor and a school to how use admin tools properly. People seeks perfect candidates and oppose or complain about minor things, that is wrong, and demotivate many good editors. Lack of more candidates, also, may turn English Wikipedia into a sort of outlaw place and it is dangerous for the project. Carlosguitar (Yes Executor?) 14:22, 4 September 2021 (UTC)
 * 27) I'm in that group of Admins who've been around since the early days and only rarely do much, in part probably for the same reason that currently few people put themselves forward -because the job appears to be being done. In the early days there was also the element that everything WP was new and shiny and we all got to know each other, often in meatspace as well as IRC. I'm not sure how much that might still happen, and it especially applies to anyone wanting to get involved more in the WM projects; it's all been around so long they don't know where to start. From appearances things are being done ok, but yes being able to spread the load and get new blood into administrating WP would be a good thing. Maybe some sort of outreach program? --AlisonW (talk) 14:24, 4 September 2021 (UTC)
 * 28) There's numerous issues – we're losing admins faster than we're gaining them, RFA is a gigantic ordeal, and people are opposing candidates because they are afraid they might become rouge and they would be too difficult to remove. Stifle (talk) 10:42, 6 September 2021 (UTC)
 * 29) Yes I have an uneasy feeling - we could do with more. People need to be be made aware that concerns regarding admin conduct can be made to arbcom. Cas Liber (talk · contribs) 11:59, 7 September 2021 (UTC)
 * 30) I'm motivated to oppose mainly by the obvious difference between editing in areas with active admin attention and those without. Though it can't be measured in admin actions or the size of backlogs, there are definitely some dark corners that could use some light attention. I also want to second some great points: there are likely good admins out there that we are missing out on (valeree), some areas with backlogs (Shibbolethink, Spencer, and I'd add WP:CCI), and need for wikigenerational diversity (Kusma, WereSpielChequers). Firefangledfeathers (talk) 01:59, 8 September 2021 (UTC)
 * 31) We're now starting to see a decline in the number of active admins, which is far more concerning than the long-heralded decline in the total number of admins. The bus factor/Pareto principle problem is palpable: if we lost a dozen particular admins, whole parts of the project would immediately start rotting away. The remedy is more admins, and we're clearly not getting anything approaching that. It behooves us to ask why. Extraordinary Writ (talk) 18:50, 8 September 2021 (UTC)
 * 32) In the long run, absolutely. The supply of admin-hours can't seem to keep up with demand, which has several major negative long-run effects. KevinL ( aka L235 · t · c) 19:15, 8 September 2021 (UTC)
 * 33) Everybody is aging and at some point Wikipedia will face some kind of correction. You see the same admin faces, doing the same work for years and years. When they retire, the correction will come. Probably 10-12 years, perhaps sooner, perhaps longer as everybody is really dedicated. If not fixed by then, then an entirely practical approach will be taken, e.g. editors in good standing who want to do it, will be made admins, outwith the RFA process.   scope_creep Talk  10:41, 9 September 2021 (UTC)
 * 34) Per above. I'm a single point of failure for two critical processes. MER-C 11:06, 12 September 2021 (UTC)
 * 35) The spigot does not produce sufficient flow to keep the lawn watered. --Izno (talk) 05:51, 14 September 2021 (UTC)
 * 36) There is little interest in becoming an admin these days. The answer is to make the (s)election process less 'horrible and broken', and the job itself more appealing and less of a Sword of Damocles - but without letting them run away with the idea that it's part of a climb up a greasy pole. Kudpung กุดผึ้ง (talk) 12:52, 19 September 2021 (UTC)
 * 37) Ostriches in the sand isn't a solution. The number of inputs alone on the subject should indicate that there's at least the PERCEPTION of a problem. Buffs (talk) 21:20, 22 September 2021 (UTC)
 * 38) I think in the long run there will be an issue as evidenced by above oppose voters. As such we need to address any issues that can be fixed with RfA before there are too few admins. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 13:06, 27 September 2021 (UTC)
 * 39) There are various issues which have been outlined below which I and many other editors have supported and expressed their concerns in it. If there was no issue, this RfC wouldn't be held in the first place. TheGeneralUser (talk) 22:23, 29 September 2021 (UTC)

A. Comments

 * It may become a problem, but most of the new admins are being extraordinarily productive DGG ( talk ) 06:52, 7 September 2021 (UTC)
 * What Jason Quinn said in the support section. There are issues, but "the community" is wasting too much time on this. Before you can fix RfA you need to fix the community. Good luck with that. For now, I'm focused on the stuff that's easier to fix, and I wish more editors were too. Still having trouble getting over the fact that there's no consensus we even have a movement. Maybe after we form one, that movement will be able to fix RfA. – wbm1058 (talk) 13:44, 7 September 2021 (UTC)
 * The criterion that lists about 400 'active' editors  '...made 30 or more edits in the last 2 months' , is extremely misleading. For one thing, it's not even based on logged admin actions which don't necessarily figure in an edit count, while other admin tools might record many edit counts from only one mouse click. Kudpung กุดผึ้ง (talk) 23:29, 26 September 2021 (UTC)

B. Not enough like a discussion (should be less like a vote)
Unlike nearly all other processes, RfA is largely a straight up vote with defined percentages for pass, crat chat, and fail. Wikipedia works best when we achieve consensus through discussion.

B. Support (agree)

 * 1) I'm surprised to see this section empty, so I'll go ahead and populate it. Specifically, there's not enough discussion of the evidence base for anyone's vote, and even more specifically, not enough discussion of oppose reasons. (The last time we did this RfA review thing, I said something like "votes should be evidence based" and people showed up to oppose the idea ) The culture against "badgering" is in many ways a good thing - candidates being defensive or making excuses is never a good look - but the effect is often that people seem to want influence over the outcome of a discussion without the responsibility of actually... discussing. Opabinia regalis (talk) 19:02, 31 August 2021 (UTC)

B. Oppose (disagree)

 * 1) Our most political processes (Arbcom, RfA, etc.) have always been more or less done by a vote. There are two reasons for this. One is that when we're dealing with individual editors rather than article content, the application of policies and guidelines tends to be much more subjective. Everyone has good points and bad points, and it is up to the individual voters to decide whether one outweighs the other; policy doesn't tell us what to do. Secondly, the number of participants in an RfA is an order of magnitude greater than that most other discussions. When you have an AfD with 10 !votes, flipping one person's !vote will shift the percentage by 10%, so !vote percentages in normal discussions are very rough and don't mean too much. When you get to RfA-level participation, there is much less sampling error and the percentage starts to mean something. -- <b style="color:red">King of ♥</b><b style="color:red"> ♦</b><b style="color:black"> ♣</b><b style="color:black"> ♠</b> 02:15, 29 August 2021 (UTC)
 * 2) Discussions work best when there's a small group of editors participating and a complex set of possible outcomes. RfA is the exact opposite—large participation and a very binary outcome. The fact it's not 100% a vote is a quirk of our consensus model, and one that seems to work alright (the 'crats are competent, even if their discussions are stressful to candidates), but it's not feasible to try to make it more discussion-like. &#123;{u&#124;  Sdkb  }&#125;  talk 07:31, 29 August 2021 (UTC)
 * 3) Consensus decision making is unsuitable for decisions with only two possible results. —Kusma (talk) 12:57, 29 August 2021 (UTC)
 * 4) It only looks like a vote if you make the mistake of igoring the discussions and the people who consequently change their position during the RFA.  Ϣere Spiel  Chequers  09:46, 30 August 2021 (UTC)
 * You can change your vote in SecurePoll elections like ArbCom or Board ones too, and technically there’s discussion, but we’d certainly call those a vote and not a consensus discussion? ProcrastinatingReader (talk) 09:51, 30 August 2021 (UTC)
 * True, but doing so isn't public and I'm assuming doesn't change the debate in the way it does in RFA.  Ϣere Spiel  Chequers  14:29, 1 September 2021 (UTC)
 * 1) Sdkb has got it right. Consensus decision making is useful in complex tasks like writing an encyclopedia because there is an enormous number of possible solutions and permutations to explore. This is not the case in an election. — Goszei (talk) 22:46, 30 August 2021 (UTC)
 * 2) I agree with WSC, there is already sufficient discussion, you can see how people change their mind and express their thoughts. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:18, 1 September 2021 (UTC)
 * 3) There is discussion. Having more discussion with the number of people involved in the AfD process would make AfD a huge process, probably more time-consuming for candidates than it currently is. /Julle (talk) 16:12, 1 September 2021 (UTC)
 * 4) RfA is a vote. However there is enough discussion to allow voters to see the reasons for other people's votes, especially opposers' votes. On rare occasions, bureaucrats will overrule (blatantly) misguided or troll votes. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  10:09, 2 September 2021 (UTC)
 * 5) I always thought the hysterical obsession with "OMG WE DON'T VOTE" was counterproductive anyway. Stifle (talk) 10:43, 6 September 2021 (UTC)
 * 6) RfA, with the number of participants, is unavoidably a vote. Having it more of a discussion would make it very hard for crats to determine the consensus of an RfA, and with a structured vote format it helps keep everyone on the purpose of the RfA (which is to decide whether to make someone an administrator). Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 09:10, 7 September 2021 (UTC)
 * 7) Nearly all RfAs have some discussion, which I suspect sways many voters one way or the other without them needing to resort to spelling their vote out Cas Liber (talk · contribs) 12:02, 7 September 2021 (UTC)
 * 8) RfA is a vote Most processes here proceed by consensus, but it's a wise arrangement that some processes here be a vote.  At the moment there are two: RfA and the elections for arb com. These are suitable places, because they are decisions that affect not an individual article or the applicability of an individual guideline, but the right to decide objectively  whether there is in fact consensus. They are additionally positions that can see and decide on confidential or hidden data . In such decisions I trust a definite vote more than a vague consensus--it's the basic governmental tradition of all major English speaking countries.  A vote has one absolute advantage: on most really contention matters, there have been many arguments over whether there is in fact consensus for something. A vote is a numerical sum, and is final and unmistakable. A few of the most unfortunate AfDs have been instances where the crats have--in my opinion wrongly--attempted to exercise their claimed power to over-ride a community rote, rather than merely see whether it is sufficient,  DGG ( talk ) 14:25, 7 September 2021 (UTC)`
 * 9) I don't think it's reasonable to want more participation and less of a vote-like atmosphere, and of the two, I'll take more participation. With hundreds of votes, it's unreasonable to expect a discussion process to yield understandable results without an interminable timeframe. I do see pressure for Oppose votes to explain themselves, which I find to be fair. Firefangledfeathers (talk) 02:05, 8 September 2021 (UTC)
 * 10) Mostly per . KevinL ( aka L235 · t · c) 19:16, 8 September 2021 (UTC)
 * 11) We have learned as a community that the "RFA is not a vote" mantra just allows cabalism to trump consensus. Also, if we are to encourage more people to run for RfA, it would really help if people said less there, not more. – filelakeshoe (t / c) 🐱 08:43, 22 September 2021 (UTC)
 * 12) Per King of Hearts and Sdkb. Extraordinary Writ (talk) 03:00, 28 September 2021 (UTC)
 * 13) I didn't want to pile on, but I might as well say that I don't care much for this idea per above at the last minute.  – John M Wolfson (talk • contribs) 18:08, 28 September 2021 (UTC)
 * 14) Both the votes and the discussion itself are equally important for an RfA. Everyone (whether they support or oppose) are encouraged to explain their rationale for doing so that consensus can be achieved by the end of an RfA. Hence, both are definitely equally important. TheGeneralUser (talk) 18:16, 29 September 2021 (UTC)

B. Comments

 * An interesting writing on consensus and RfA is Requests for bureaucratship/Mackensen. I think the nomination statement is worth a read, and as such I don't quite agree with 's comment above. ProcrastinatingReader (talk) 13:45, 29 August 2021 (UTC)
 * In theory, our current process is a hybrid between strict policy-based consensus and a vote: in that !votes are expected to be fundamentally reasonable, i.e. an evidence-based comment on whether the candidate would be a benefit with the admin toolset; but the bar for that reasonableness is lower than in other discussions, where policy allows less latitude with respect to what's reasonable, and so among those who are being reasonable, a numerical threshold still applies. IMHO this is fundamentally a good system, and we should keep it. Vanamonde (Talk) 07:27, 30 August 2021 (UTC)
 * In general, I think the consternation with respect to RFA-as-vote versus RFA-as-discussion tends, in the case of the latter (the way I think about it, at least), to focus on those cases wherein an otherwise qualified, well-rounded candidate's nomination is torpedoed as a result of some perceived misstep; one or two well-researched, extensive opposes give perhaps undue weight to a valid reason for opposition by way of featuring so prominently in the scope of discussion in a manner that precipitates a wave of drive-by, considerably less well-researched opposition that overshadows support by volume, but perhaps not substance. Whether or not that's justified (which, to be sure, is a case-by-case analysis) or whether this is what actually takes place at RFA in some cases, I think the very perception of such contributes to a general hesitancy to launch an RFA. It's unknowable what will precipitate such a snowball effect (though educated guesses and the input of the wise and experienced ahead of an RFA can be helpful), and the best way to avoid such potential for dramatic scrutiny is to shy away from contentiousness altogether (lest some conflict or momentary lapse in judgment be brought to bear at RFA), which may diminish the ability of a level-headed, intelligent editor to contribute, either before or after RFA, an effect no one should find desirable. <b style="color:#C75300;">Tyrol5</b> ▸ [talk]  20:59, 31 August 2021 (UTC)
 * This may be outside of the scope of this stage of the RFC: The public aspect of RFA is a major contributing factor here but, on the other hand, the transparency of the process strikes me as a strength. I wonder if a private ballot process wherein voters are encouraged (required?) to provide feedback (perhaps on a form with comment fields visible only to the candidate and bureaucrats) to the candidate would preserve some benefit of transparency (to the candidate, at least) while perhaps somewhat mitigating the negativity of the public aspect. Much further thought and development would be necessary before this would be anything close to coherent. <b style="color:#C75300;">Tyrol5</b> ▸ [talk]  21:30, 31 August 2021 (UTC)
 * I'm not going to !vote either way on this issue, but I will say that I think the process of voting should be a little more stringent about votes being for valid reasons. There's a reason they are public. Trainsandotherthings (talk) 23:11, 2 September 2021 (UTC)
 * I'm of two minds about this. I do think the opposers make some valid points, but there should be some mechanism to appropriately discount those who provide bizarre or totally off base reasons. Seraphimblade Talk to me 04:06, 4 September 2021 (UTC)
 * I am a bit torn with this section because it requires that believing RfA should be more discussion like also means that it should be less like a vote. The fact is, I believe the discussion can be meaningfully enhanced while not desiring that it become less of a vote.--John Cline (talk) 07:03, 4 September 2021 (UTC)

C. Too much discussion (should be more like a vote)
The extensive discussion makes it unpleasant for candidates who have to watch themselves be dissected by various participants with limited recourse.

C. Support (agree)

 * 1) Having a discussion isn't as much of a problem as having a discussion while voting. I know of no other semi-democratic process where campaigning and arguing is allowed right in the ballot box. —Kusma (talk) 12:59, 29 August 2021 (UTC)
 * 2) Kusma is right. Being dissected by others is unpleasant, but it is necessary, and ultimately a good thing. Having a discussion while voting (even responding directly to other user's votes, which should not be allowed) is the problem, not the discussion itself. I think an ArbCom-like system like Beeblebrox alludes to (discuss, then vote) is ideal to avoid the problems stemming from discussion and voting at the same time. — Goszei (talk)  22:46, 30 August 2021 (UTC)
 * I said "at the same time" in this comment, but what I really mean is "in the same place". I have no objection to a secret ballot concurrent with discussion, as long as the vote can be changed until the voting period ends. John M Wolfson's concern about the length of proceedings in the Comments below is valid. — Goszei (talk) 23:59, 30 August 2021 (UTC)
 * 1) Exactly what Kusma and Goszei said, insofar as I understand what they're saying. Whether we like to call it that or not, this is a vote, albeit it one occurring using an unconventional and flexible standard. And there's too much discussion occurring while it's happening. A more conventional vote, one in which there were no discussion (but in which a discussion period preceded it), would be an improvement. Chetsford (talk) 23:07, 30 August 2021 (UTC); edited 23:32, 30 August 2021 (UTC)
 * 2) Many of our colleagues on other projects make this a straight vote without discussion. I kind of like that idea. Risker (talk) 01:32, 8 September 2021 (UTC)
 * 3) A vote and a discussion are needed, but how it works right now is not great. I quite like Kevin's suggestion of a two day discussion period, followed by five days of voting (and continued discussion). A private ballot would be untenable, but I agree that the immediate vote presents challenges. Having a pre-discussion allows for a good number of questions to be answered first, which allows for more informed voting. Sure, voters can change their vote after questions, buts its dramatic and most folks don't' like admitting they're wrong. CaptainEek  Edits Ho Cap'n!⚓ 21:25, 8 September 2021 (UTC)

C. Oppose (disagree)

 * 1) I also disagree with the idea that we have too much discussion. People should be free to defend their view and challenge others in an effort to convince !voters to their side. But once all the convincing is done, the election should be decided by (super-)majority rule. -- <b style="color:red">King of ♥</b><b style="color:red"> ♦</b><b style="color:black"> ♣</b><b style="color:black"> ♠</b> 02:18, 29 August 2021 (UTC)
 * 2) It's pretty close to a vote already; also per my comment in the previous section. Vanamonde (Talk) 07:28, 30 August 2021 (UTC)
 * 3) I actually think having a narrow gap is beneficial, as some close RfAs have more, shall we say, stupid (!)votes, than others. I think leaving it with "votes if clear-cut, discussionish if not" is a good general principle to have, even if the specifics change Nosebagbear (talk) 11:11, 30 August 2021 (UTC)
 * 4) People's minds often get changed by these very discussions. Unless we move to something more like the arbcom elections (discuss -then- vote), this is as it should be. Beeblebrox (talk) 21:40, 30 August 2021 (UTC)
 * 5) The candidate needs to be scrutinised to some degree (otherwise it's inside baseball and most !voters don't have the information they need about most candidates), which necessitates comments about them. These comments also need scrutiny, due to the extreme hostility, incivility and grudge-bearing displayed by a large proportion of the en.wiki community, particularly towards anyone seen to hold a position of power. And the downside of this is then that reasonable opposers are badgered. I agree that "The extensive discussion makes it unpleasant for candidates who have to watch themselves be dissected by various participants with limited recourse", but making it more vote-like without other changes (like Arbcom's discuss-then-vote) would introduce more problems than solutions. — Bilorv ( talk ) 16:27, 31 August 2021 (UTC)
 * 6) I agree with Kusma and others in the support column that it isn't good to have both discussion and !voting occurring simultaneously, and I'd support trying out no !voting during the first 24hr. But what this section seems to be at its core is getting rid of the discretionary zone, and I'm not sure there's a need for that, per Nosebagbear. &#123;{u&#124; Sdkb  }&#125;  talk 22:10, 31 August 2021 (UTC)
 * 7) Per King of Hearts, I think we've got this about right. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:20, 1 September 2021 (UTC)
 * 8) The RFA process should encourage discussion - otherwise, it's no different than a straight up vote and we shouldn't use the term !vote anymore. Trainsandotherthings (talk) 23:14, 2 September 2021 (UTC)
 * 9) I have on plenty of occasions see people change their minds (both from support to oppose and vice versa) based upon discussion at an RfA. Given that, the discussion clearly can be useful for clarification and should be kept. However, RfA should never be a civility-free zone; editors engaging in discussion there should be expected to follow the behavioral guidelines that apply everywhere else. Seraphimblade Talk to me 04:09, 4 September 2021 (UTC)
 * 10) I think that the amount of discussion is about right. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  12:56, 6 September 2021 (UTC)
 * 11) Although a discussion then voting process like for ACE would be an interesting idea, this means that the process will need to take longer. Plus without using secure poll it would be difficult to fully prevent discussion in the voting period. Also if a concern is raised in the voting period, then this may be completely missed. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 09:15, 7 September 2021 (UTC)
 * 12) No. Things get discussed that need discussing Cas Liber (talk · contribs) 12:03, 7 September 2021 (UTC)
 * 13) Weakly. I actually wouldn't object too much if the first two days of an RfA were discussion or question only, with a five-day vote phase (with discussion and questions still allowed) coming after. Indeed, the fact is, RfAs are essentially decided in the first day right now, and many votes are cast without the benefit of answers to later questions. But mostly per . KevinL ( aka L235 · t · c) 19:17, 8 September 2021 (UTC)
 * 14) Both the votes and the discussion itself are equally important for an RfA. Everyone (whether they support or oppose) are encouraged to explain their rationale for doing so that consensus can be achieved by the end of an RfA. Also, both the discussion and voting should happen and take place simultaneously and gives people an equal chance to vote and discuss at the same time. Hence, both are definitely equally important. TheGeneralUser (talk) 19:04, 29 September 2021 (UTC)

C. Comments

 * The real problem with the discussions is that most of them are about what the criteria should be rather than whether the candidate meets the criteria. Hence the classic problem of RFAs, they can look like taking a driving test with a minivan full of test examiners who are arguing between themselves whether you need to know how to change a flat tyre and can't agree as to which side of the road you should be driving on, while paying little if any attention as to your driving. Most other userrights have at least some of the criteria agreed by consensus and then only changeable by consensus.  Ϣere Spiel  Chequers  09:55, 30 August 2021 (UTC)
 * I oppose having a vote after discussion because I think it would take too long, either objectively or subjectively. We do a similar process with ArbCom and it's this whole thing that's once a year; ditto for Steward elections on Meta. If we do ultimately decide to use, or at least pilot, a secret ballot, !voting should take place simultaneously with discussion, so as not to make RfA a whole multi-stage event that might intimidate adminabili. – John M Wolfson (talk • contribs) 23:39, 30 August 2021 (UTC)
 * I think RfA already sort of is a multi-stage event, with people talking to possible nominators in private, going through weeks or months of trying to boost their CV or sanitise their reputation by increasing uncontroversial CSD noms/AFD NACs/whatever else they feel they have to do, and then actually running. I'm not necessarily in favour of discuss-then-vote but the baseline of intimidation has to already be at a record high. There are people who run impulsively because it'll be over in a week, and it's fantastic when that happens, but they seem to be the exception. — Bilorv ( talk ) 16:35, 31 August 2021 (UTC)
 * That's all fair enough, I'd just rather not have the formal RfA itself become multi-stage and up the intimidation factor even more. – John M Wolfson (talk • contribs) 16:38, 31 August 2021 (UTC)
 * has just firmly hit the nail squarely on the head. Kudpung กุดผึ้ง (talk) 13:04, 19 September 2021 (UTC)

D. Administrator is undesirable position
The position of administrator comes with little benefit and lots of drawbacks (e.g. increased scrutiny/calls for accountability).

D. Support (agree)

 * 1) I don't know that I would say there are drawbacks like "increased scrutiny/calls for accountability" but the admin tools (block, delete, protect) are not very interesting compared to other things (e.g. content writing, coding, finding images). Plus, WP:INVOLVED means an admin can't use admin tools on the articles the admin works on, so adminship is about doing boring things in topic areas I'm not interested in. Compare that 0 upside against any cost (like dealing with an RFA), and it's not worth it. Plus I don't really care to have "view deleted" rights because I don't really want to see the junk that gets revdel'd. Levivich 07:18, 29 August 2021 (UTC)
 * 2) Signing up to do administrative tasks is signing up to do chores. For some editors, those chores are on their own behalf, but Wikipedia needs many more chores to be done on the behalf of others. Typically in a young community, there is a strong feeling of camaraderie in building up a new project from the start that spurs a lot of volunteer efforts. As it grows and ages, it can be harder to scale up the number of people handling chores to match the size of the community. English Wikipedia has mitigated this problem through greater automation and more refined permissions. Nonetheless, to be sustainable, Wikipedia needs a regular influx of new volunteers to alleviate pressure on previously enlisted volunteers and to continue to integrate representatives of new editor cohorts into Wikipedia's governance structure. We need to think about more ways to encourage new editors to sign up for chores. isaacl (talk) 14:54, 29 August 2021 (UTC)
 * I'm not an admin, but I don't think the average admin faces so much more scrutiny that the role becomes undesirable. Especially since the position is more or less a lifetime appointment. -- Calidum  17:00, 30 August 2021 (UTC)  This was meant to be in the section below.  -- Calidum  18:49, 31 August 2021 (UTC)
 * 1) From User:Antandrus/observations on Wikipedia behavior, bullet 32: "When someone screams about "admin abuse", it's most likely true – they're probably abusing admins again. If there's a block involved, expect to see a battalion of sockpuppets in short order, making even more shrill cries of admin wrongdoing." -  F ASTILY   00:26, 31 August 2021 (UTC)
 * 2) It's all about whether you're seen as being in a position of power. When I do NACs, experienced editors treat me like shit; when I'm answering edit requests on heated topics, new editors treat me like shit. The mop isn't inherently a target on your back, but it can be a factor in someone treating you like shit when you exercise some form of power over them. Some admins have decreased scrutiny in the long-term because they get an accurate reputation for doing great things, and when they start misbehaving in ways that other people couldn't get away with, people think "an admin couldn't behave like the accuser is saying". But some potential admin candidates are dissuaded by a correct assessment that for some types of things they want to do, the mop will only increase the anger directed at them. — Bilorv ( talk ) 16:50, 31 August 2021 (UTC)
 * 3) The users who want to be an administrator are the ones that never pass RfA; the users who don't are (often) the ones who would make an excellent admin but never try. 🐔 Chicdat  <sup style="font-family:Times New Roman">Bawk to me!  11:04, 2 September 2021 (UTC)
 * 4) In a sense. I think being an admin is perceived as an undesirable position for a large portion of the community, even for people who would (1) pass an RfA, (2) benefit the community, and (3) enjoy the admin tasks (especially the lower-profile ones that we desperately need more of). Fixing this perception could help. For this group, the "I don't know why anyone would want to be an admin" refrain is actively harmful. My actual experience of being an admin is much like 's: Other than some undeserved respect / people taking my random opinions too seriously, I haven't had problems with being an admin. Best, KevinL ( aka L235 · t · c) 18:40, 8 September 2021 (UTC)
 * I agree with Kevin here, but do not think it's reasonable to support the question as such. :^) Izno (talk) 18:56, 14 September 2021 (UTC)
 * 1) per  Kudpung กุดผึ้ง (talk) 13:08, 19 September 2021 (UTC)

D. Oppose (disagree)

 * 1) It's not really meant to be a 'benefit'. It's a set of advanced responsibilities that are requested voluntarily, ostensibly because one wants to contribute more to the website. There are plenty who want to do that, myself included. There are also plenty who want it for reasons of status, which while unproductive does contribute to overall interest in the position. /Tpdwkouaa (talk) 04:15, 29 August 2021 (UTC)
 * 2) Yes and no, but rather no. I've heard this concern voiced in person to me ("I have no idea why anyone would run for adminship at all"), but that's not an issue to be fixed. Some requests for adminship by very inexperienced users are speedily deleted, officially as "pages unambiguously created in error" per WP:G6, practically per WP:IAR as the criterion was likely not intended for this purpose. I'd like to link to a log of these, but I can't find a way to filter the deletion log for subpages of WP:RFA. Perhaps someone can provide an SQL query. What I'm trying to say is: There are people who mistakenly believe that administratorship is a desirable position for various reasons, and each of them is rather problematic than helpful. There is no need to add "benefit" to administrative positions; those primarily looking for personal benefits shouldn't apply for the position. ~ ToBeFree (talk) 11:00, 29 August 2021 (UTC)
 * @ToBeFree Your wish is my command: query/58099 AntiCompositeNumber (talk) 02:04, 30 August 2021 (UTC)
 * Godsend,, a godsend. Face-smile.svg I've filtered it for log summaries containing "G6" and got a list of over 1000 pages with no false positive immediately visible to me. Most of these will likely have had wrong impressions about the benefits of adminship. That's okay, but we don't need to worsen the situation by intentionally attracting more benefit-seeking candidates. ~ ToBeFree (talk) 20:14, 30 August 2021 (UTC)
 * 1) Certain actions come with extra scrutiny, the position itself doesn't. Other than some undeserved respect / people taking my random opinions too seriously, I haven't had problems with being an admin. —Kusma (talk) 13:41, 29 August 2021 (UTC)
 * 2) There is no requirement to do anything once you are an admin, and you become much more immune to being sanctioned since admins get kid-glove treatment from their colleagues. What's there not to like? That occasionally you'll be criticized by the non-admins as part of the privileged wiki elite? :> --<sub style="border:1px solid #228B22;padding:1px;">Piotr Konieczny aka Prokonsul Piotrus&#124; reply here 02:44, 30 August 2021 (UTC)
 * 3) I don't see the benefit issue as having changed at all in over a decade. What has happened is the greying of the pedia, and while it is normal for teenagers and young people to sit exams and want to gain qualifications, we now have lots of greybeards, many of whom consider themselves to be past such things.  Ϣere Spiel  Chequers  09:59, 30 August 2021 (UTC)
 * 4) I have no perception that there are drawbacks to having admin tools. The drawback is having to undergo RfA to be granted them. Nurg (talk) 23:53, 30 August 2021 (UTC)
 * 5) It really depends on what areas an admin chooses to work in. GABgab 15:55, 31 August 2021 (UTC)
 * 6) No one knows you're an admin unless/until you've done admin work to them, you have a mop icon on your userpage, they went through the trouble of looking at your usergroup log, they remember your RfA, and/or they downloaded the blue highlighter, none of which apply to the vast vast majority of even (especially?) bad-faith users. – John M Wolfson (talk • contribs) 16:14, 31 August 2021 (UTC)
 * 7) <span id="202108311850_Calidum"> I'm not an admin, but I don't think the average admin faces so much more scrutiny that the role becomes undesirable. Especially since the position is more or less a lifetime appointment. -- Calidum  18:50, 31 August 2021 (UTC)
 * 8) Per Nurg. However, I would say that RfA is less desirable among the very people we'd most like as admins. There's the quote about "anyone who is prepared to run for office should automatically, by definition, be disqualified from ever doing so." &#123;{u&#124; Sdkb  }&#125;  talk 22:15, 31 August 2021 (UTC)
 * 9) I think we have to accept this. It may be "not a big deal" or a "janitorial role", but actually, there is prestige to the role, it is desirable, within our community and outside it. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:21, 1 September 2021 (UTC)
 * Firmly - per my comments to Kudpung in the discussion section below. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 08:02, 21 September 2021 (UTC)
 * 1) As WTT said above, there is absolutely prestige to the position and it is desirable. I would tender that the RFA process is what's deterring candidates, not that nobody wants to be an admin. Trainsandotherthings (talk) 23:24, 2 September 2021 (UTC)
 * 2) Not undesirable, I don't think. There are of course upsides and downsides to it, but things like keeping out vandals and spammers is work that has to be done if our project is to remain functional and viable. And ultimately, no admin has to take on any task that they don't want to; it's not like someone can dock our pay if we don't do something. If an admin wants to stick to relatively uncontroversial areas and help out with those that's fine, and if they're at some point willing to go into deeper water that's great too. Seraphimblade Talk to me 00:34, 4 September 2021 (UTC)
 * 3) I fully agree with Nurg regarding where the undesirability lies.--John Cline (talk) 08:34, 4 September 2021 (UTC)
 * 4) There are neither material benefits nor drawbacks to being a sysop. Stifle (talk) 11:24, 6 September 2021 (UTC)
 * 5) There are clearly many people who desire to be an administrator, but billions of people who don't desire to be an administrator. Does this make being an administrator undesirable? By one interpretation, yes. However I don't think that's the intended meaning of the original point. Perhaps a better question is something like: "Do the drawbacks of being an administrator outweigh the benefits to such an extent that editors are discouraged from applying?" In my opinion, no. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  13:08, 6 September 2021 (UTC)
 * 6) If undesirable don't do it. Period Cas Liber (talk · contribs) 12:04, 7 September 2021 (UTC)
 * 7) Even the janitorial aspects are desirable to some, with editors taking on admin-like activities all the time. If given the admin tools, they'd act in similar ways, but with more ability and authority. I agree with those above mentioning the prestige factor. I don't want to deny the kind of 'admin abuse' that Fastily/Antandrus mention admins experiencing; I am not sure what can be done about it. Firefangledfeathers (talk) 02:15, 8 September 2021 (UTC)
 * No, it is gatekeeper process. Once your in, generally everybody is happy your an admin, and happy to run with it. It only the entry that is the hassle. From experience in IT, there is a class of folk who really enjoy the work, quite rightly. It benefits Wikipedia by definition Mankind in general. It is a good thing.   scope_creep Talk  10:46, 9 September 2021 (UTC)
 * 1) I've been an administrator for just under 15 years and I don't feel that I'm particularly under scrutiny or called to be accountable any more than other editors are. If I do something contentious or get involved in something controversial, then sure, I'd expect my actions to be scrutinised that bit more, but that's true of everyone, not just admins. Not all admin actions are mired in controversy; in fact most of it is straightforward grunt work to keep the wheels turning. <b style="color:#98F">W</b><b style="color:#97E">a</b><b style="color:#86D">g</b><b style="color:#75C">ge</b><b  style="color:#83C">r</b><b  style="color:#728">s</b><small  style="color:#080">TALK  12:36, 27 September 2021 (UTC)
 * 2) Per WTT's comments, and the fact that the "lots of drawbacks" (assuming they exist at all) are easily outweighed by the lifetime tenure (barring truly overt misconduct) that is currently conferred to all admins. Extraordinary Writ (talk) 03:09, 28 September 2021 (UTC)
 * 3) I agree that being an administrator requires increased scrutiny and accountability but that is a part of the process. Without a proper community vetting and review, we can't decide who is qualified enough to be an administrator. Administration in almost all areas is mostly a back end process. It is never meant to be an accomplishment and achievement of any kind and because of that it is naturally an undesirable position. It is volunteer role that is taken up editors who are happy to clean up the mess (just like the janitor and their mop). TheGeneralUser (talk) 19:16, 29 September 2021 (UTC)

D. Comments

 * I think there are two ways to interpret this question, and I want to make sure I'm answering the right one:
 * Is being an administrator actually an undesirable position?
 * The secret is ... it's not that bad. It is true that there are certain important administrative tasks that will require you to be in conflict with other editors, and don't get me wrong, those tasks are hard, and I have the utmost respect for the administrators who volunteer to do them. However, at the end of the day, we are all volunteers, and your experience of being an administrator is largely dependent on the things you like to work on. It is entirely possible to be an uncontroversial admin that just deletes obvious CSDs, blocks obvious vandals, and protects obvious RFPP requests—and yes, we do need more of these kinds of admins too.
 * Is there a perception that being an administrator is an undesirable position, and does this perception hinder RfA recruitment?
 * This I think is the more interesting question for this RfC, and I would be interested to hear the thoughts of others. I feel that my perception of this may be skewed—after all, I probably wouldn't have applied for RfA if I didn't think it was a desirable position in the first place. I'm curious to know whether it's more the case that editors do want to be administrators, but they are hesitant to do so because of the difficulty of the selection process, or whether it's more the case that editors don't want to be administrators at all.
 * Mz7 (talk) 02:51, 29 August 2021 (UTC)


 * I wouldn't say it's "undesirable", but I think attitudes towards admins/adminship are significantly more negative than they used to be.  Hut 8.5  13:45, 29 August 2021 (UTC)
 * Mz7's question is spot on; absent that clarification, I don't see how I can answer this. I agree that it isn't actually undesirable, but it is certainly perceived to be. Vanamonde (Talk) 07:30, 30 August 2021 (UTC)
 * Being an admin is an undesirable position if you are a spammer, undisclosed paid editor or POV warrior as it brings extra scrutiny. Logically one assumes we have more of at least two of those groups than we did before Wikipedia became known as the world's most important reference site. There are other people who just don't want the aggro, but I doubt they are more or less common than in the past.  Ϣere Spiel  Chequers  10:03, 30 August 2021 (UTC)
 * MZ7s question is excellent, in fact, it's so good I've had to hold off answering. In a less exciting clarification, am I right in thinking this is "life as an admin is undesirable post-RfA, rather than "being an admin is an undesirable position, factoring in RfA"? Nosebagbear (talk) 11:13, 30 August 2021 (UTC)
 * Mz7's question 2 is spot on. I encourage regular RfA nominators to conduct a survey of RfA candidates who have chosen not to run and ask them why they have chosen not to. We need to determine if candidates are not running because of the RFA process, the perceptions of admin work, a combination of both, or some other reason. Z1720 (talk) 15:05, 30 August 2021 (UTC)
 * Pinging - not sure if Mz7's query above falls within your deliberate broad topics, but thought it worth checking if you'd had a specific thought in mind when you included it Nosebagbear (talk) 20:33, 30 August 2021 (UTC)
 * So this was not an original idea of mine but was instead trying to capture what I've seen other say. I have seen it expressed both ways and was trying to capture that here. That is some say Admin is an actually undesirable position and some say that Admin is seen as undesirable by candidates. For despite having originally come up with 12 problems, and having several more added since launch, I was trying to be broad because in my experience a larger number agreeing on a single concept, even if the nuance is different, because there could be more than 1 solution to any given problem which has consensus. Of course this distinction could also emerge through discussion and (hopefully) be captured by the closer when considering consensus. Best, Barkeep49 (talk) 21:06, 30 August 2021 (UTC)
 * I have often referred to adminship as "the Golden Turd" in that we make such a big deal out of it, but once you have it you realize it's mostly just boring stuff and getting criticized for every conceivable reason, warranted or not. What to do about that is less clear. Beeblebrox (talk) 21:44, 30 August 2021 (UTC)
 * I am not so much worried about the general attitude, which I think is rather positive. I am really worried about extremes. I have seen literally same editor one day say how wonderful admin I am and next day say that I am such piece of shit that they are surprised I have not yet been site-banned, or at least speedily desysopped. And, mind you, I have never been sanctioned by ArbCom or the community.--Ymblanter (talk) 14:52, 2 September 2021 (UTC)
 * And unfortunately I see less users who respect administrators, but I see a lot of users who are incivil or even rude to them, and a lot who are afraid of them and would be rude if they were not afraid of.--Ymblanter (talk) 21:28, 2 September 2021 (UTC)
 * The role of administrator is whatever one makes of it. There are no formal responsibilities, no requirement to use the tools or otherwise act in an administrative capacity in any given situation. To me, to the extent there is negative perception of the role, it is perhaps on account of what the community at RFA expects. My general perception in recent years is that wanting to act administratively upon solely noncontroversial actions (or even just in one or a couple of different areas) when one's time is free is probably not as unassailable a justification for requesting the toolset as it once was -- there may be a general expectation that potential administrators be able and willing to engage in the contentious slog of mopping up the drama boards, for instance, or, if not, just to push themselves generally to become a jack of all trades and master of none, administrative or otherwise, hence the consternation about the acrobatic plate spinning I think a lot of potential candidates feel as though they'd need to engage in to become a viable candidate. <b style="color:#C75300;">Tyrol5</b> ▸ [talk]  17:28, 3 September 2021 (UTC)
 * Just a note that the process of becoming an administrator is part of the cost, and so the decline in candidates is de facto telling us that fewer users feel the cost is worthwhile. isaacl (talk) 14:51, 6 September 2021 (UTC)
 * It's an unijtentional trick question. One could quite legitimately vote 'support' or 'oppose', and that's why I think I'm really ambivalent. Even vote doesn't really show which side he's on and it's more a shrug of the shoulders. Kudpung กุดผึ้ง (talk) 13:18, 19 September 2021 (UTC)
 * I don't agree. Administrator is a "desirable" position - it has prestige, it has power (within our local community). It is not just a janitorial role, nor is it "no big deal". Evidence for this? The number of admins who carry on a handful of edits per year to keep the tool, the high standards and high turnout at RfA, the amount of process it takes to remove an admin bit "for cause", the fact that we have had so many "RfA reform" projects over the years.
 * Personally, I don't worry about my admin bit. I have handed it in when I wasn't using it - but I also understand how the wider community sees the role and denying that fact makes it more difficult to enact change. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 08:01, 21 September 2021 (UTC)

E. Corrosive RfA atmosphere
The atmosphere at RfA is deeply unpleasant. This makes it so fewer candidates wish to run and also means that some members of our community don't comment/vote.

E. Support (agree)

 * 1) The current process is a vote but it is not a secret ballot.  Other similar processes such as the current Trustee elections do use proper voting software which maintains the confidentiality of each vote, as is standard practice in western democracies.  If you don't have a secret ballot, then you naturally get problems such as the bandwagon effect, groupthink and intimidation.  We routinely scoff at elections in North Korea where "turnout is near 100%, and approval of the Democratic Front's candidates is unanimous or nearly so" but such perfect scores are often expected and seen at RfA.  That's not what we get when our ballots are secret –  even the most successful candidates for arbcom get hundreds of opposes.  But our proper elections don't seem to attract so much drama or perennial claims that they are broken and need fixing.  That's because those elections have a secret ballot. But the problem is not just a lack of secrecy.  Oppose voters are forced to go into detail about the candidate's failings.  This then leads to acrimony as the candidate's supporters will naturally argue about this per WP:BADGER and WP:BLUDGEON. Andrew🐉(talk) 09:48, 29 August 2021 (UTC)
 * 2) It dismays me how many people can't seem to put themselves into the place of candidates and show some empathy, or at least listen to what they're saying about it. It's a weeklong performance review in front of the entire company, every paper clip you've ever used is on the table for discussion, and your worst work enemy and the intern who started yesterday can both ask questions. And yet many people who haven't been through it (or who went through it ten or fifteen years ago) seem to think the current atmosphere no big deal. —valereee (talk) 13:37, 29 August 2021 (UTC)
 * 3) Per above. I think a secret ballot is at least worth a try. &#123;{u&#124;  Sdkb  }&#125;  talk 18:42, 29 August 2021 (UTC)
 * 4) I support trying this, per above. &#8213; Qwerfjkl  talk  20:53, 29 August 2021 (UTC)
 * 5) From reading past RfAs and listening to people who gone through the process, it can be a very stressing and upsetting experience. I think it's valid to say that many users edit Wikipedia for the joy it brings to them (for a variety of reasons), and I imagine that most would rather not go through this experience, even if they would make for a valuable sysop. I'm not sure if a secret baallot would improve it, but might be worth a try. Isabelle 🔔 23:38, 29 August 2021 (UTC)
 * 6) RfA can really suck sometimes. Your entire contribution history is put into the spotlight, including all of the embarrassing mistakes that you've learned from. Some of the worst RfA oppose sections overemphasize trivial issues (such as ancient dramas or meaningless statistics like raw edit count), fail to require corroboration for any allegations or claims (in the form of diffs), and fail to provide a safe space for the candidate to respond to concerns. For example, the oft-cited page Advice for RfA candidates states:  Instead of responding to oppose votes directly, candidates often desire to wait for a question that allows them to respond to concerns in the oppose section, a question that may or may not come. Mz7 (talk) 00:08, 30 August 2021 (UTC)
 * 7) It's a rare time I agree with user:Andrew Davidson but he makes an excellent point. This needs to be a secret ballot, following a discussion (the question to arbcom candidate format is pretty good). --<sub style="border:1px solid #228B22;padding:1px;">Piotr Konieczny aka Prokonsul Piotrus&#124; reply here  02:41, 30 August 2021 (UTC)
 * 8) I'm going to have bifurcate myself. I agree that in a significant number of RfAs the atmosphere is corrosive, and in an additional fraction it's abrasive, even if not downright corrosive. Steps to mitigate that are absolutely worthwhile. However I am strongly against a secret ballot. It makes it harder to identify reasons for failure, inherently eliminates Cratchats, would require people who were opposing for a non-obvious feature to note the reason so that others could be aware of it, or risk it going unnoticed etc Nosebagbear (talk) 11:17, 30 August 2021 (UTC)
 * 9) RFA is absolutely ghastly. I have no idea why any rational person would submit to it.—<b style="font-family: Verdana; color: Maroon;">S Marshall</b> T/C 12:02, 30 August 2021 (UTC)
 * 10) The corrosive atmosphere at RfA exists, and that much is indisputable. As for the cause, Andrew's analysis of the effects of a non-secret ballot hits the nail on the head. Like Piotrus, I think we need an ArbCom-like system of discussion before a secret vote. — Goszei (talk)  22:54, 30 August 2021 (UTC)
 * Reproducing a similar comment I made on C: I don't think the discussion has to take place prior to voting as in ArbCom elections, but simply in a separate place from the discussion. I have no problem with concurrent discussion and voting, as long as voters can change their vote before the period ends. — Goszei (talk) 00:02, 31 August 2021 (UTC)
 * 1) Prospective candidates absolutely (and understandably) think that RfA has a corrosive atmosphere, and identifying ways to fix that is certainly appropriate. I do share Nosebagbear's skepticism of the secret-ballot idea, so I hope that better methods of addressing the issue can be identified. (Here's a (probably horrible) idea: community-authorized general sanctions permitting individual bureaucrats to impose RfA topic bans on disruptive participants. While such a power would hopefully be used sparingly, the very threat of it might bring some of our more contumacious editors into line.) But whatever the solution is, the problem is definitely worth discussing. Extraordinary Writ (talk) 23:54, 30 August 2021 (UTC)
 * 2) Every prospect I've suggested adminship to in the past several years has declined with this concern.  -  F ASTILY   00:26, 31 August 2021 (UTC)
 * 3) Most reasons already said by others. I will add that I don't remember the last time the crats did any clerking at RfA except to remove sockpuppet comments etc, and on at least one occasion a crat was defending even that (when removed by a non-admin). I also think the following comment is amusingly accurate: if you can put up with people (like me!) throwing socking accusations around—because RfA is the one place on the entire project where one can not only make such comments without falling foul of the civility brigade but can double down on the accusation with impunity on the grounds that you are *protecting the encyclopedia*. There are many examples of PA at RfA which, if made anywhere else, would be block-worthy. ProcrastinatingReader (talk) 00:43, 31 August 2021 (UTC)
 * 4) Self-evidently, yes.  Unfortunately, it's very difficult to extract these unpleasant elements while leaving important investigation of the applicant's record of contributions and conduct as a community member in tact.  I know it's not a very satisfying suggestion, but I think any change in the culture of the process is going to necessarily have to be more exhortative in tone, rather than involving any particularly strong new prescriptive/proscriptive requirements for lines of inquiry.  That said, it probably would not hurt to have admins taking a stricter stance on violations of existing policies, and likewise issuing quick warnings for borderline (or for that matter, across-the-line but short of a block) comments--and getting real support from the community when wading into such disruption. <b style="color:#19a0fd;">S</b><b style="color:#66c0fd">n</b><b style="color:#99d5fe;">o</b><b style="color:#b2dffe;">w</b><b style="color:#B27EB2;">Rise</b><b style="color:#d4143a"> let's rap</b> 06:30, 31 August 2021 (UTC)
 * 5) Not supporting a secret ballot, but RfA is a cesspit. — Bilorv ( talk ) 17:02, 31 August 2021 (UTC)
 * 6) I also don't support a secret ballot (or any specific proposed solution at this juncture), but whatever word you want to use (corrosive, abrasive, cesspit), it's bad, bad enough that I also question why anyone goes through it. It's like watching people walk across fire or do a polar plunge or eat some nasty bug: brave, but why? Levivich 19:15, 31 August 2021 (UTC)
 * 7) People do not enjoy their RfA because no one enjoys personal feedback - and there's a lot of it, hundreds of people scrutinising your behaviour over years. But that's also fundamental to the system. I don't support a confidential vote at present, but I could be persuaded. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:25, 1 September 2021 (UTC)
 * 8) There is no shortage of controversial RFAs we can look back at to prove this issue exists. Even candidates who get overwhelming support can still suffer from this issue. There are certainly RFA voters who are looking for reasons not to support a candidate, as opposed to looking for reasons to support them. I'm not sure on the remedy, but I am certain this is an issue. Trainsandotherthings (talk) 23:28, 2 September 2021 (UTC)
 * That's me! I look for reasons to oppose a candidate, because my default position is "support" if none of those reasons exist.  If someone's at RfA looking for reasons to support, then it's because "support" isn't their default.—<b style="font-family: Verdana; color: Maroon;">S Marshall</b> T/C 09:52, 3 September 2021 (UTC)
 * 1) I don't support a secret ballot. But I agree with the headline here: RfA is currently a savage process. MarioGom (talk) 13:01, 3 September 2021 (UTC)
 * 2) Absolutely. I don't know what the best way out is. For the consideration of secret ballots, I'd like to hear what former ArbCom candidates think about that process. —Kusma (talk) 13:09, 3 September 2021 (UTC)
 * 3) My RfA constitutes the worst four days and several more thereafter of my now six years on Wikipedia. It was absolutely miserable. Aside from being told that you have problems that you need to work on, which is of course never pleasant, but my RfA was also dominated by running battles between the entire planet and Joe Roe for a now-redacted edit desc and his general attitude towards me, stupid opposes that were then badgered to death by my camp, zealot partisans of me, emotional injury to friends of mine, several of which were those zealot partisans. My experience on the whole was that I felt rejected, of course, but also like a humiliated, mistrusted vagrant. It has led me to think that whatever takes as much of the conversation about an RfA out of a candidate's earshot is the best and should be pursued. – ♠Vami  _IV†♠  00:33, 4 September 2021 (UTC)
 * 4) Speaking for myself, it was a horrible experience and what was the worst was that it was unexpectedly so. It caught me completely by surprise. That's why I have warned editors about not undergoing an RfA unless they are prepared for a bruising. I've noticed that things have gotten better than six years ago but I think that is only because there are fewer editors who want to put themselves through it. Only candidates that are virtually assured of passing will undertake an RfA now where years ago, lots of people would give it a shot just to see if they would pass. I think that the few candidates who have that casual attitude either withdraw or their RfAs are snow closed. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 02:05, 4 September 2021 (UTC)
 * 5) Not a secret ballot, just actually enforce the civility and NPA policies, and then for everyone else, don't throw a fit when someone does that. A secret ballot would deprive candidates (both successful and failed) of feedback, and so long as that feedback is given in a civil fashion, it is quite useful. Seraphimblade Talk to me 04:14, 4 September 2021 (UTC)
 * 6) As I said before: too much drama for something that should not be "a big deal". The process leaves potential candidates stress, anxiety and afflict new ones to be recruited. Carlosguitar (Yes Executor?) 14:24, 4 September 2021 (UTC)
 * 7) It's not so much (any more) that RfA is an incivility free-for-all, it certainly was at one time but folks seem to be behaving lately. It's more that RfA is treated not so much like a job interview where the community teases out the candidate's strengths and weaknesses, it's treated more like an inquisition where the community tries to grill the candidate until they reveal a horrible truth about themselves that disqualifies them from adminship, and at the same time each inquisitor has a different criteria for what is disqualifying. As a result, the process greatly favours candidates who are already comfortable with Wikipedia's gladiator culture, while candidates who are competent don't pass or don't step up in the first place. Ivanvector (Talk/Edits) 18:41, 4 September 2021 (UTC)
 * 8) I agree that RfA is, too often, unnecessarily corrosive and unpleasant. I also agree that this negativity affects candidates and !voters alike, discouraging many from participation.--John Cline (talk) 22:54, 4 September 2021 (UTC)
 * 9) Very much so. Stifle (talk) 11:24, 6 September 2021 (UTC)
 * 10) Yes, it is. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  13:24, 6 September 2021 (UTC)
 * 11) The atmosphere was felt to be unpleasant when I first became an admin, a decade and a half ago. I'm not sure I'd want to stand again if I had to.  ++Lar: t/c 23:58, 6 September 2021 (UTC)
 * 12) Yes - am in two minds about a secret ballot. I think it might lead to lower pass rates and worry about the lack of feedback. Cas Liber (talk · contribs) 12:08, 7 September 2021 (UTC)
 * 13) Hard to watch sometimes. Even if the knock-down-drag-outs are a rarity, the spectre of the few is definitively an obstacle to there being more RfAs. In other parts of this discussion, current admins are mentioning dissuading others from RfA because of the atmosphere. Firefangledfeathers (talk) 02:22, 8 September 2021 (UTC)
 * 14) I agree entirely with IvanVector above. Vanamonde (Talk) 12:54, 8 September 2021 (UTC)
 * 15) I think this is a major reason people give when they decline to run an RfA. KevinL ( aka L235 · t · c) 18:40, 8 September 2021 (UTC)
 * 16) This is perhaps the central and key issue. I don't have a silver bullet solution. I think part of it is a cultural shift: just being nicer. Part of it may be better clerking/moderation. Perhaps just as we have ArbClerks, we need some RFA Clerks. A selected body of 5-10 clerks would allow for some recusals, and provide a much more structured environment. A key issue to solve though would be: what are the standards for clerking? Perhaps we wouldn't even need clerks if we wrote some better guidelines/rules about how to behave during an RFA. CaptainEek  Edits Ho Cap'n!⚓ 21:14, 8 September 2021 (UTC)
 * 17) In agreement with . JBchrch   talk  03:57, 10 September 2021 (UTC)
 * 18) Per Firefangledfeathers. Looking forward to proposals in phase two to improve this like more active moderation, instructions to give constructive feedback when opposing (maybe use a different word than oppose, something less confrontational), or secret ballot. FemkeMilene (talk) 16:31, 11 September 2021 (UTC)
 * 19) Yes, it is worse than it needs to be. --Izno (talk) 18:55, 14 September 2021 (UTC)
 * 20) Isn't that what I had been saying for 12 years and no one would believe me? Now we suddenly have 37 supports on this one! Perhaps this is why WT:RfA has just entered its longest hiatus since the page began. Kudpung กุดผึ้ง (talk) 13:25, 19 September 2021 (UTC)
 * 21) Infinitely worse than it needs to be. Buffs (talk) 19:55, 22 September 2021 (UTC)
 * 22) tesat do kamene ("that I would sign with my blood"). – filelakeshoe (t / c) 🐱 20:17, 22 September 2021 (UTC)
 * 23) I'm an RfA veteran, never an admin. What Vami IV said about afterwards feeling humiliated and mistrusted is dead on. I nearly quit the project after the experience. This aspect of RfA must be rectified, if nothing else. I think the recent run of nearly 100% approve !votes reflects the fact that only those who expect little to no negativity in their RfA are running. ☆ Bri (talk) 17:46, 24 September 2021 (UTC)
 * 24) This is something that has got worse over time and continues to do so. Go back 15-20 years and it wasn't like this, and the process worked really well. So while I'm sure there will be many suggestions about how to improve the process, I'm afraid hardly any of that will make much difference - it's the culture and perception around RfA that's the problem, not the process itself. <b style="color:#98F">W</b><b style="color:#97E">a</b><b style="color:#86D">g</b><b style="color:#75C">ge</b><b  style="color:#83C">r</b><b  style="color:#728">s</b><small  style="color:#080">TALK  12:41, 27 September 2021 (UTC)
 * 25) I'd say that this is the case. For example, a person who wants to oppose a candidate on good reasons may be badgered by those who supported. As such voting against the majority is often something which causes lots of comments and disagreements that the voter now has to deal with. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 13:15, 27 September 2021 (UTC)
 * 26) Like Seraphimblade said above, just enforce the civility policy. If RFA clerks were empowered with a zero-tolerance mandate for civility violations and personal attacks, the RFA process might actually be tolerable. Nosferattus (talk) 17:15, 27 September 2021 (UTC)
 * 27) Support / Agree. Hmlarson (talk) 22:28, 27 September 2021 (UTC)
 * 28) I definitely agree with this. Although a candidate may prepare well for an RfA and most editors act in good faith and give their honest opinions and feedback, sometimes there can be people who act in bad faith and create and increase that corrosive and hostile atmosphere at RfA which can cause immense emotional and mental pain and suffering to the candidates. Something seriously needs to be done about it. However, I don't agree with the secret ballot and believe that both the discussion and votes and everything related to an RfA should be done on-wiki only. TheGeneralUser (talk) 19:38, 29 September 2021 (UTC)

E. Oppose (disagree)

 * 1) One problem with a secret ballot is that you don't know why people voted against you. My first run failed, my second run was a success in large part because I could see what the opposers said in my first run and I could change appropriately. Just as importantly, when we do turn down established editors it is invariably because someone points out a flaw that others then agree is a reason to oppose. If we had run past RFAs as secret ballots there are a number of RFAs that would have passed instead of failed, I'm not sure that would always have been a good thing.  Ϣere  Spiel  Chequers  10:11, 30 August 2021 (UTC)
 * 2) Same here. My first RFA failed, as did my first run at being a functionary. The RFA, I understood the problem and was able to correct it. At the time I first applied to be a functionary we were using a secret ballot. Nobody got elected, and none of us knew why. That's one of several reasons we don't do that by secret ballot anymore. Beeblebrox (talk) 21:46, 30 August 2021 (UTC)
 * 3) Over the last three years (2018–2021), the median RfA candidate (excluding obvious NOTNOWs) received 93% support. Toxic RfAs with pile-on opposes are the exception, not the norm. – Teratix ₵ 06:58, 2 September 2021 (UTC)
 * It's possible — if not likely — that such a high support rate is a symptom of this very issue. If only the absolutely most perfect, most stellar candidates run, and then only if they've several functionaries for nominators, we'd expect to see a high support and success rate, but it's a self-limiting group. ~ Amory <small style="color:#555"> (u • t • c) 16:53, 2 September 2021 (UTC)
 * Certainly that's a possibility, but if that were the case, surely the issue would not be "RfA, at present, has a corrosive atmosphere" but rather "RfA has a historical reputation for a corrosive atmosphere that discourages candidates in the present". I'm more sympathetic to the latter sentiment, but it seems to be distinct from statement E. – Teratix ₵ 07:04, 21 September 2021 (UTC)
 * Run for admin yourself and come back to this opinion. – ♠Vami _IV†♠  00:38, 4 September 2021 (UTC)

E. Comments

 * I am probably an outlier, but I found my RfA ok, and I was treated like shit multiple times afterwards, and this was really a bad experience.--Ymblanter (talk) 14:50, 29 August 2021 (UTC)
 * Maybe there could be a secret ballot with comments allowed, albeit anonymized. So for opposes, a ballot could return Oppose Not enough work in projectspace, concerns with maturity. without the opposer's identity being disclosed. Ideally we could at least see the counts of !votes while the RfA is ongoing so that basketcases get prematurely closed appropriately. – John M Wolfson (talk • contribs) 14:49, 30 August 2021 (UTC)
 * , I really like this idea. It resolves some of the issues mentioned in the Oppose votes above by allowing candidates to still know why they failed, and to be able to work on it. And I think it would remove a lot of the stigma about participating, as everyone won't be watching how you oppose/vote. That being said, I think this would only work as long as Anonymous IP editors were still not able to vote, but were able to comment without voting (and without revealing their IP). Such as comment concerns about temperament, see [diff] even while not logged in. — Shibboleth ink  (♔ ♕) 22:15, 30 August 2021 (UTC)
 * This idea doesn't work when we have socks, grudge-bearers and bad faith !voters as a significant proportion of people interested in RFA (and we do). — Bilorv ( talk ) 16:57, 31 August 2021 (UTC)
 * Let candidates choose whether they want a secret vote or the classical process. See what happens over time. ~ ToBeFree (talk) 17:38, 31 August 2021 (UTC)
 * I can see that. – John M Wolfson (talk • contribs) 18:18, 31 August 2021 (UTC)
 * The corrosive atmosphere on Wikipedia doesn't start at RfA. It's found in places like WP:BOGOF (where editors are accused of helping COI editors) and MOS:INDENTGAP (where editors are accused of failing to be inclusive of editors needing screen readers). Bring your corrosive attitudes to RfA at your own risk. US presidential elections use secret ballots but are highly corrosive. Set up a system with limited scrutiny of candidates (questions sanitized to be non-confrontational, short voting period) or waive people thru despite the corrosive atmosphere at their RfA and expect an increase in traffic at ArbCom's desysop department. – wbm1058 (talk) 21:42, 25 September 2021 (UTC)
 * Until about 18 months ago, I kept an extremely close eye on all things RfA for well over a decade. Nowadays I’m not concerned whether or not en.Wiki will end up with too few ‘’truly active’’ admins to accomplish all tasks for which the tools and responsibilities are needed. I will however repeat this comment of mine from exactly two years ago:
 * ''There is 'badgering' (to harass or annoy persistently – Merriam Webster), and genuine expression of concern that a vote might not conform to our Wikisocial norms. Practicing questionably high criteria or posing trick or irrelevant questions are issues that could be perhaps better addressed on the voter's own talk page where the editor is made to look and feel a fool slightly less publicly. Purely disingenuous, disruptive, or false voting probably ought to be responded to directly on the RfA.
 * In any case, the number of mini threaded discussions being moved to the RfA's talk page is becoming very much more frequent. This is not due to more consistent clerking, but is a result of the steady degradation of the environment of the process, the doubling of participation since the December 2015 reforms, and the classical propensity at Internet forums for everyone to add their two pence.''
 * Anyone who does not read this and follow all its links before running for adminship only has themself to blame if their RfA turned out to be a bad trip whether it passed or not. The data is only held for 5 years but at nearly 20,000 views in that period alone would appear to demonstrate that most candidates have been reading it. Interestingly, a logarithmic graph would probably reveal that the page views are concomitant with a growing lack of interest in becoming an admin and the apathy at WT:RFA. Hardly surprising in today's climate.Kudpung กุดผึ้ง (talk) 12:19, 26 September 2021 (UTC)

F. Level of scrutiny
Many editors believe it would be unpleasant to have so much attention focused on them. This includes being indirectly a part of watchlists and editors going through your edit history with the chance that some event, possibly a relatively trivial event, becomes the focus of editor discussion for up to a week.

F. Support (agree)

 * 1) We need to have some level of scrutiny to adequately vet candidates, but it's very clear from the debriefings that the current level of scrutiny makes RfA a highly unpleasant experience that discourages editors from running. &#123;{u&#124; Sdkb  }&#125;  talk 02:18, 29 August 2021 (UTC)
 * 2) I think we need to have some level of scrutiny, but as a recent RFA attempt, I can attest to the fact that it really feels like you'll get torn apart for weaknesses in areas where you have indicated you don't intend to work. There seems to be a high level of expectation of "jack of all trades", while in reality one person isn't going to be able to check every. single. box. Hog Farm Talk 02:22, 29 August 2021 (UTC)
 * 3) The problem with crowdsourced scrutiny is that some percentage of the scrutiny will be unreasonable. We need to filter out the unreasonable scrutiny. Levivich 07:20, 29 August 2021 (UTC)
 * 4) I think scrutiny is important -- candidates should be very clueful, knowledgeable in areas they initially intend to work, and pretty much unfailingly civil -- and we can't confirm that without someone going through their edit history. I'm not sure it's helpful to have probably dozens of people going through their edit history trying to find something to object to, which I suspect happens for at least some candidates. I feel like something that happened more than X years ago, unless it's absolutely egregious -- socking or something at that level -- should be off limits for questions or comments. —valereee (talk) 13:55, 29 August 2021 (UTC)
 * 5) When a community is small enough that everyone knows each other, they can rely on their personal knowledge to evaluate their level of trust in a candidate. As it grows, though, people have to rely on other indicators and form an impression based on a sampling of the candidate's actions. Evaluating someone in the open can be uncomfortable as it often involves speculation and tentative hypotheses, which may not get proven out in the end but only a fraction of participants may take note of their final dispositions. I think the community should strive to make its evaluation as efficient as it can manage, and determine a candidate's characteristics based on their actions with minimal speculation. Tentative assumptions should be proven or disproven before layering further analysis on top of them. isaacl (talk) 15:17, 29 August 2021 (UTC)
 * 6) I've been asked twice if I wanted to run for admin. The atmosphere of RfA was enough to make me decide I shouldn't try. XOR&#39;easter (talk) 17:19, 29 August 2021 (UTC)
 * 7) based on several past RFAs Eddie891 Talk Work 22:30, 29 August 2021 (UTC)
 * 8) The problem is that mud sticks. I think this can be proven by a correlation of how many years one needs to become an admin, which will probably be a curve. You can't become an admin with just a year or so of history, but if you have 10+ years you'll fail too as surely over that time you've made enough mistakes and enemies that when taken together there'll be enough first that you'll fail. Scrutiny is good, but it has to be limited to the last few years or such. Mistakes of the youth should not count against someone whose last few years were stellar. But most people will assume that those old mistakes will be dragged and used to poison the well and scare others off, and they don't want to bother. This is why many old, experienced contributors who would make good admins are not bothering to help. --<sub style="border:1px solid #228B22;padding:1px;">Piotr Konieczny aka Prokonsul Piotrus&#124; reply here  02:40, 30 August 2021 (UTC)
 * 9) I think the kicker is the variance of the scrutiny. People have had RFAs pass without someone so much as mentioning unbecoming incidents that happened the month that they ran in, and had them fail for ridiculous concerns a decade old or based on the strangest conjectures you can think of (worst case I've seen by far is the candidate who had to watch dozens of people argue about a number in their username and its relation to fascism). But when you're running for RFA, you have to assume it's the latter—that's how risk management works. Because if it is, you have to be prepared to live with the fallout of it. — Bilorv ( talk ) 17:11, 31 August 2021 (UTC)
 * 10) I think this is a common belief and a reason people don't run, though I'm not sure that it's a problem that is solvable. Calliopejen1 (talk) 23:52, 31 August 2021 (UTC)
 * 11) I agree with the statement, but can't necessarily commit to a remedy for the problem. It would be good if there was a sort "statue of limitations", a mutual understanding that incidents X years ago shouldn't be dredged up, but even that needs wiggle room built in. I can conceive of incidents so egregious that I'd realistically never want to see the person made an admin—so I wouldn't want to deny any other participant the right to draw such a line, even if it's not where I'd draw it. --BDD (talk) 15:34, 1 September 2021 (UTC)
 * 12) As Piotrus mentioned, mud does indeed stick. Every editor makes mistakes, it's part of being human. But RFA candidates are expected to be perfect or else face serious scrutiny and criticism, often disproportionate to the severity of whatever mistake they've made, or how long ago it happened (and a perfect candidate will be scrutinized for being too perfect). We should remember that admins are human too and make mistakes, and this is just a fact of life. RFA feels more like a background check for a high level security clearance in an intelligence agency than it does a process for selecting trusted editors who can be given additional tools to help the project. Trainsandotherthings (talk) 23:37, 2 September 2021 (UTC)
 * 13) Don't forget that it can be two weeks in the worst case (bureaucrat chat). —Kusma (talk) 13:17, 3 September 2021 (UTC)
 * 14) Scrutiny is part of the process but it is just incredible to me how some editors will make their Oppose decision on a single, solitary diff that they disagree with. For me, when I was a new editor, I was a little sarcastic and I found casual talk page comments I made when I was new brought up during my RfA to demonstrate that I'd be a difficult and disrespectful administrator. Some editors project a lot of unpleasant emotions on to other people's edits. But when every edit you have ever made is forever logged in your Contribution history, there is no editor that doesn't have baggage. You can't control people combing through your edit history but it would be nice if there was a time limit on how far back the examples are that are brought up during an RfA. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 02:14, 4 September 2021 (UTC)
 * 15) Though I don't know what the solution to this is besides "be reasonable". A major problem a month or two ago is probably a cause for a lot of concern. A minor issue a few years ago is probably really not. People should be able to differentiate those, and realize that over years worth of history, there will be some mistakes and bad calls. We're all human, and sometimes we have a bad day or a brain fart. Seraphimblade Talk to me 04:18, 4 September 2021 (UTC)
 * 16) I don't exactly agree with the framing but definitely with the conclusion. RfA commenters have a pattern of treating any transgression as fatal, no matter how minor or how far in the past, and nobody who has any experience on this project knows if they put themselves forward, what someone is going to dig up from years in their past and frame in a way that fails their RfA. Most editors actually don't find it very enjoyable to have to defend every action they've ever made, just for the privilege of then having to defend every action they subsequently make. Ivanvector (Talk/Edits) 18:52, 4 September 2021 (UTC)
 * 17) I'm sure this is true. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  13:28, 6 September 2021 (UTC)
 * 18) I'm sympathetic to some of the opposers' comments (e.g. WSC and Kevin), but the clincher for me is the comments from recent RfA candidates themselves. Take a statement like this one: the candidate found the scrutiny so "emotionally draining" that she considered withdrawing. And that was for an RfA that passed with 98% support! For controversial candidates, the experience can only be infinitely worse. While it may not seem so unpleasant to the rest of us, the scrutiny is clearly a real factor that's discouraging RfA prospects. I don't know what the solution is, but it's definitely a problem. Extraordinary Writ (talk) 21:54, 26 September 2021 (UTC)
 * The link you included is broken. Trainsandotherthings (talk) 22:45, 26 September 2021 (UTC)
 * Fixed – sorry about that. Extraordinary Writ (talk) 23:02, 26 September 2021 (UTC)
 * 1) Although I agree with the opposers that scrutiny is needed, I and others likely found that their RfAs were often a stressful time. Especially having to think about what might be brought up. As such the level of scrutiny is possibly too much, but perhaps this level is needed for the community to have an effective RfA process. So therefore, I agree with the question but I'm not so sure reducing the scrutiny applied would help make RfA a better place. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 13:18, 27 September 2021 (UTC)
 * 2) I can definitely agree that a proper amount of scrutiny in evaluating candidates at RfA is both necessary and required. The main problem lies in when things/mistakes a person has done no matter how insignificant (even if it was a long time ago) which do not reflect on the candidate's current suitability and ability in becoming an administrator are brought into an RfA just to stir up some kind of drama. That is very real and serious problem. TheGeneralUser (talk) 20:06, 29 September 2021 (UTC)

F. Oppose (disagree)

 * 1) True there is scrutiny, but its reputation is not as bad as the normal reality. Few things are considered so egregious that they would merit opposes two years later - that doesn't mean there won't be any opposes over long dead issues, but they don't gain traction and have other people opposing per them if they are stale.  Ϣere  Spiel  Chequers  10:52, 30 August 2021 (UTC)
 * 2) RFA is a sort of stress test for admins. If you can't face it, you probably should not consider adminship. Beeblebrox (talk) 21:48, 30 August 2021 (UTC)
 * 3) I agree with WSC about reputation vs reality. I really don't like the idea of changing things so we explicitly give less scrutiny to our admins <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:27, 1 September 2021 (UTC)
 * 4) Per my comment on E, the idea that RfA candidates are excessively or unreasonably scrutinised doesn't really hold up when you consider the high median support percentage candidates have received in recent times. – Teratix ₵ 10:22, 2 September 2021 (UTC)
 * 5) Per all preceding really. Cas Liber (talk · contribs) 12:11, 7 September 2021 (UTC)
 * 6) Weakly oppose this as a general statement. Certainly there have been overly scrutinized RfAs, but in my experience, most RfAs face about the right amount of community scrutiny – we have to have some; adminship is now, under our present policies and norms, a big deal. Best, KevinL ( aka L235 · t · c) 18:42, 8 September 2021 (UTC)

F. Comments

 * While I think the perceived scrutiny of the process might keep some people away, I sometimes wonder if the scrutiny is as intense as it is made out to be. I realize RFAs are few and far between these days, but how many of them start out with dozens of support votes before anyone votes in opposition? -- Calidum  16:57, 30 August 2021 (UTC)
 * I agree that the perceived level of scrutiny is enough to cause hesitancy in adminabili; indeed, it delayed my own RfA by almost a year. However, the actual level of scrutiny, based on my RfA and several recent ones, is not enough IMO to cause undue stress in a rational candidate. I think filling in this perception gap would go a long way in fixing RfA. – John M Wolfson (talk • contribs) 23:34, 30 August 2021 (UTC)
 * I'm with this guy. Izno (talk) 18:54, 14 September 2021 (UTC)
 * Re the "stress test" comment above, I don't necessarily think this is true. I'm an admin and tend to work in mostly non-contentious areas. I imagine there are more people like me who may not want to do contentious stuff, but could still be trusted with the tools. Calliopejen1 (talk) 23:52, 31 August 2021 (UTC)
 * My thoughts mostly fit in the support camp. I think the level of scrutiny is appropriate, but misdirected. I don't care much about old conflicts, AfD stats, or stray political comments, which all seem to attract inordinate scrutiny. I am fine with rigorous scrutiny on recent editing and behavior. Firefangledfeathers (talk) 02:27, 8 September 2021 (UTC)
 * The biggest issue is probably that the scrutiny is often misplaced. A lot of the scrutiny seems to be over-detailed into elements that administrators are really given leeway with.  For instance, after becoming a sysop, I was on the wrong side of Articles for deletion/2024 United States presidential election in Missouri.  Nobody cared; one AFD on the wrong side of the consensus isn't a big deal.  If that had happened right before my RFA there would have been much wailing and gnashing of teeth.  For some reason, a number of RFA participants expect more perfection from RFA candidates that they do administrators themselves.  Yes, scrutiny of admin candidates where they plan to work is good, but it seems that a substantial minority expect candidates to not make mistakes, rather than make them and learn from them.  I think this is the main scrutiny problem. Hog Farm Talk 20:14, 17 September 2021 (UTC)


 * ...editors going through your edit history with the chance that some event, possibly a relatively trivial event, becomes the focus of editor discussion for up to a week. This is what generally happens. Taken mostly out of context, such vengeance is blown out of all proportion, but people love a bloodbath - they want to be entertained. ...how many of them start out with dozens of support votes before anyone votes in opposition? – .  Most of them and they haven’t been done with any due diligence. Perhaps an idea would be to delay voting in the support section until some oppose votes have been made. RFA is a sort of stress test for admins. If you can't face it, you probably should not consider adminship. – . Many people suggest this but RfA is not Boot camp (been there, done that), and job interviews are rarely as distasteful or humiliating. Kudpung กุดผึ้ง (talk) 19:51, 26 September 2021 (UTC)

G. Lifetime tenure (high stakes atmosphere)
Because RfA carries with it lifetime tenure, granting any given editor sysop feels incredibly important. This creates a risk adverse and high stakes atmosphere.

G. Support (agree)

 * 1) We need better mechanisms for culling or desysopping. &#123;{u&#124; Sdkb  }&#125;  talk 02:10, 29 August 2021 (UTC)
 * 2) Things wouldn't be so high stakes if it was easier to remove an admin and if there were terms. Levivich 07:21, 29 August 2021 (UTC)
 * 3) We need a desysop process that doesn't require a full arbcom case. I'm leery of a term limits/reconfirmation process; I know of multiple editors I believe would be excellent admins but who would have a hard time passing RfA because they've done a lot of work in contentious areas. I worry that required reconfirmation might discourage admins from making difficult calls or working in difficult areas. If you have to keep reconfirmation in mind, are you going to avoid difficult calls? —valereee (talk) 14:06, 29 August 2021 (UTC)
 * 4) Agreed, this makes sense. This would be a much lesser deal if admins were subject to yearly (bi-yearly?) review and reconfirmation vote, where they'd have to show more than just 'I've been active'. They should write a brief summary of what they did (to prevent people from becoming admins just to have a 'status symbol'/'token of kid-glove treatment when reported to ANI') and answer questions (if any) from the community. --<sub style="border:1px solid #228B22;padding:1px;">Piotr Konieczny aka Prokonsul Piotrus&#124; reply here  02:46, 30 August 2021 (UTC)
 * 5) Since I was on ArbCom, this is always at the forefront of my mind at RfA. It is extraordinarily difficult to desysop admins who are misusing their tools unless they do something completely egregious – not quite deleting the main page, but close. And I invite those saying that it's easy or easier to desysop to try getting the committee to take, and then participating in, a 6-week admin conduct case! As long as that's the case, I am not going to be the only one who is very cautious about supporting people I don't already know well. Attempts to do something about this have consistently failed, so when the same people who fight ArbCom tooth and nail on desysop cases start clamouring for RfA reform, I can't help feeling that it is more about enlarging the old boys' club than dismantling it. –&#8239;Joe (talk) 04:38, 30 August 2021 (UTC)
 * 6) I agree with Joe. Also my comments re WSC below; it’s not just about whether ArbCom will pass a motion to desysop, it’s about the divisive process required to achieve that result. Finally, RFA is the only point where the community directly gets a say on the composition of its admin corps, or the status of an individual admin. A better and simple desysopping process (like the WP:INTADMIN one, EFM, or any unbundled perm, all of which work perfectly fine) would probably make people more comfortable to give prospective admins a chance. ProcrastinatingReader (talk) 10:38, 30 August 2021 (UTC)
 * 7) We treat desysopping like it's a very big deal. To get desysopped, you have to screw up in a way that seriously annoys enough people to escalate to Arbcom, and then you have to fail to type out the mea culpa of automatic sysop exoneration.  And because desysopping is a big deal, admins are hard to remove, and because admins are hard to remove, the community feels the need to scrutinize sysop candidates very closely, and that scrutiny is absolutely harsh.  It's a vicious circle.—<b style="font-family: Verdana; color: Maroon;">S Marshall</b> T/C 12:08, 30 August 2021 (UTC)
 * 8) I agree it creates that impression for some people, I just happen to think those people are wrong. An unpaid volunteer position, where you put yourself forward to help solve certain problems, knowing you will get grief even if you are right, and it comes with an expiration date after which you have to prove you're good enough to keep taking crap from random strangers? No thanks. This would result in fewer admins, not more. Beeblebrox (talk) 21:50, 30 August 2021 (UTC)
 * 9) I agree, and there are several possible solutions; Piotrus brings up some of them. Finding a formulation that is fair and will gain consensus is clearly a tricky task, but I think it is possible. — Goszei (talk) 22:54, 30 August 2021 (UTC)
 * 10) I've always felt that "oohhh you can't have a community desysopping process because it'll get too political/admins will be disincentivized from doing dirty work" reflects a too-cliquey attitude of adminship and becomes a bit elitist; admins are community volunteers, not first responders or combat veterans, and any matter that arises to such a level can be at worst dealt with by the WMF or functionaries. That said, I have no comment at this point about the specifics of such a process. – John M Wolfson (talk • contribs) 23:29, 30 August 2021 (UTC)
 * 11) There are at least two admins who have recently taken actions that, if they gave it as an answer to an RFA question on a previously un-opposed RFA, would get them at least 50% opposition. They remain admins. — Bilorv ( talk ) 17:14, 31 August 2021 (UTC)
 * 12)  Agree, per S Marshall. Calliopejen1 (talk) 23:45, 31 August 2021 (UTC)
 * 13) Fully agree. It's a shame that Requests for comment/Adminship term length failed. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:29, 1 September 2021 (UTC)
 * 14) I'm an admin on a Wikipedia in another language, and my impression is that annual re-elections have been crucial in keeping the threshold for getting new admins low. Given the size of English Wikipedia and the admin corps, our model wouldn't necessarily work here, but comparing the two wikis there is a stark difference in how much people feel is at stake, and the drama involved in removing admin rights from someone (which we rarely do – it happens, but most admins are comfortably reelected): someone getting just 60% support in a recurrent process tends to leave far fewer wounds within the community than trying to get someone de-sysoped in a process specifically to remove the rights, rather than potentially confirm them. /Julle (talk) 16:49, 1 September 2021 (UTC)
 * 15) I still feel like some binding recall procedure or desysop procedure that doesn't require arbcom would be a net positive, but we've seen that it doesn't find sufficient traction (especially among existing admins, understandably). &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 18:15, 1 September 2021 (UTC)
 * 16) The high stakes certainly invite more scrutiny and make reform more difficult, to say the least. Either a probationary period or periodic elections might be something to explore. Trainsandotherthings (talk) 23:41, 2 September 2021 (UTC)
 * 17) I think this is a problem but shouldn't be. Stifle (talk) 11:26, 6 September 2021 (UTC)
 * 18) Unfortunately some !voters do regard this as a reason for excessive scrutiny. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  16:37, 6 September 2021 (UTC)
 * , I am one of those, and can you blame me? Some admins seem okay when elected, but five or ten years down the line it's easier to divorce your partner than it is to desysop an admin. — Alexis Jazz (talk or ping me) 22:53, 6 September 2021 (UTC)
 * How many admins actually misuse their tools five or ten years down the line? I am not convinced that it is a significant problem. Even if it is a problem, a better solution would be to find an easier way to desysop rogue admins. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk] 23:22, 6 September 2021 (UTC)
 * , granted my experience in this field is more with Commons, but in the end we're all human. It's generally not misuse per se, just acting in ways that most users would barely (if at all) get away with. Biting newcomers, being incompetent with tools, failing/refusing to follow policy to name a few things. They would never pass a new RfA, but tenure makes them (nearly) untouchable. I applaud everyone in Category:Wikipedia administrators open to recall. — Alexis Jazz (talk or ping me) 00:40, 7 September 2021 (UTC)
 * 1) Tenure makes sense for roles in which someone may have to rule against their employer, like a judge. Not having to fear for their job is one way to allow them to remain impartial. This doesn't really apply to admins. The WMF can fire admins anyway and they aren't supposed to make highly controversial decisions on a regular basis. Perhaps applicants could be given a choice: RfA to get tenure, requiring (typically) 75% support or apply to be admin for 3 months requiring only 60% support. More people will pass, but after 3 months their adminship expires anyway. It might also prevent occupational burnout. If they do a bad job, they'll never be elected again. If they do a good job, they could easily apply for another 3 months or (as they now have a track record) have a better shot at tenure. — Alexis Jazz (talk or ping me) 23:10, 6 September 2021 (UTC)
 * 2) I agree with this statement, but only in the sense that it's factually correct and if I could get rid of the bad effects without getting rid of the good effects I would. No immediate solutions come to mind, however. KevinL ( aka L235 · t · c) 09:24, 9 September 2021 (UTC)
 * 3) I agree with the statement but only because I think it is factually correct, like Kevin above. I would not agree with term limits, as it encourages admins to play politics where they have to please to keep being re-elected. As such I can see this causing admins to be a lot more cautious and as leaves unblockables unblockable by an individual admin because of the negative implications brought by blocking the user. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 13:24, 27 September 2021 (UTC)

G. Oppose (disagree)

 * 1) It hasn't got any harder to desysop an admin in recent years, if anything it's got easier since admins are now held to higher standards. Therefore tenure and desysopping can't be the cause of the problems at RfA.  Hut 8.5  13:54, 29 August 2021 (UTC)
 * I think it's always been a problem, but it just never showed when Wikipedia was younger because there wasn't a class of legacy admins at that time. &#123;{u&#124; Sdkb  }&#125;  talk 18:37, 29 August 2021 (UTC)
 * 1) Given the grounds on which several people have been desysopped in recent years, I think that if it was ever true that it was difficult to get Arbcom to get rid of a bad admin, it clearly isn't now. I used to think that unwarranted concerns over the difficulty of removing bad admins motivated some Oppose voters, but given the proportion of RFAs that pass nem com that can't be true.  Ϣere Spiel  Chequers  08:38, 30 August 2021 (UTC)
 * It isn’t just about whether ArbCom will flip the switch, it’s about the amount of crap needed to get to that point. Several of the recent cases I presume you refer to (RH, BHG, KP & RexxS) were quite animus. At least one of those cases (BHG) was referred to recently at ANI, where someone said the then-ArbCom’s decision was influenced by the gender of the case subject; these are evidently divisive cases.The proportion of passes is surely not related; we have the highest pass rate ever, but also the lowest attempt rate ever too. That just shows less people want to run, and the only ones that do are uncontroversial passes. It doesn’t suggest whether it’s too hard to remove or not. ProcrastinatingReader (talk) 10:31, 30 August 2021 (UTC)
 * True most of the successful RFAs are uncontentious, my suspicion is that this is because people wait until they are over qualified and few stay around that long. But if the difficulty of desysopping was a cause of RFA's problems you would have a different voting pattern with most passes being narrow ones. The problem about the argument that it sometimes takes a lot of crap before Arbcom will desysop is that some of the recent desysops leave you wondering why it was thought necessary to desysop that admin, and while there are people who won't run because of recent desysops, and arguably we have too few admins to afford to lose more marginal admins, we don't have unsuccessful RFAs that you could point at and say that person would have passed if it was easier to desysop admins.  Ϣere Spiel  Chequers  11:55, 30 August 2021 (UTC)
 * 1) Lifetime tenure does not necessarily equal sustained admin activity. Some admins become less active, some voluntarily resign the tools, some retire entirely. So I'm not sure whether RFA carries the expectation that the candidate will be an admin forever, regardless of lifetime tenure. GABgab 15:13, 31 August 2021 (UTC)
 * 2) This is a widely repeated talking point, but I'm quite convinced this one's a myth. In most cases where an admin is desysopped, or even where it's seriously considered, the people involved had long periods of unproblematic admin work before developing the behaviors that got them in trouble. Though every example seems to have a few people insisting they told us all so back at the editor's RfA, desysoppings are much more about burnout than anything to do with RfA failing to catch problems. If anything, I'd predict that higher RfA throughput - even with "lower" standards - would mean a lower rate of desysopping by mitigating some of the highest-risk burnout situations: working in highly controversial areas, doing a very high volume or a disproportionate share of any specific task, or working alone in a specialized niche with little feedback. Opabinia regalis (talk) 16:30, 31 August 2021 (UTC)
 * 3) Desysopping should NOT get easier. Then the number of admins would decline even more. 🐔 Chicdat <sup style="font-family:Times New Roman">Bawk to me!  11:07, 2 September 2021 (UTC)
 * 4) I've never seen this. When I was on ArbCom, several admins were either "for cause" desysopped, or effectively were because they resigned during the case (which is an "under a cloud" resignation requiring a new RfA). I know some people like the "reelections" idea, but that raises several practical issues. First, if some qualified candidates don't want to do RfA once, how many will that decrease by if you ask them to do it repeatedly? Secondly, any admin who works in any type of controversial area can't please everyone, and so will be racking up opposes every time they make a "close call" close or decision in such a case. Who will be willing to do those? There is something to be said for a deliberative body, like ArbCom, examining claims of misconduct against an admin, and then taking the time to determine if they are well founded or just sour grapes. Seraphimblade Talk to me 16:06, 4 September 2021 (UTC)
 * 5) I don't think the tenure is a problem, on the contrary. Not having to fear a reelection a few months/years down the road allows admins to make difficult decisions. More admins would probably shirk from doing so if they have to consider that these actions, even when correct, might be unpopular and thus hurt them in a reconfirmation RFA. Additionally, since RFA is a grueling experience as pointed out above, more people might decide not to run if they just have to periodically do it again. Regards  So  Why  18:59, 4 September 2021 (UTC)
 * 6) Although I wholeheartedly support community desysop on reasons of principle, I agree with SoWhy that adopting it will not resolve RfA's problems. It might marginally improve the "high-stakes" issue, but that benefit would be substantially outweighed by the fact that adminship would become, for lack of a better word, undesirable: nobody wants the sword of Damocles hanging over their head at all times. If we're going to adopt community desysop (and, to be clear, I think we should), it needs to be discussed on its own merits instead of being tethered to a tangentially related issue. Extraordinary Writ (talk) 22:18, 6 September 2021 (UTC)
 * 7) While I agree that the atmosphere at RfA has come a long way from the days of "no big deal" to the high stakes endeavor it has become, the causal meme of so called "life time tenure" is not the reason and term limits are certainly not the solution. Although I do believe that a community based process to desysop for cause would mitigate !voter apprehension and ease the high stakes strain measurably, The past inabilities to gain consensus for such a process and the fact that it's no more likely to gain consensus now (or even ever) should not be considered as moot. Such a process has it's own potential cost and conceivably (if not likely) could be far greater than its potential gain. I think a fresh approach needs to be considered or else we're probably going to have to live with the fact of high stakes.--John Cline (talk) 22:32, 6 September 2021 (UTC)
 * 8) I think admins are held more accountable now than 10-15 years ago. Desysopping was formerly a rare thing. Cas Liber (talk · contribs) 12:13, 7 September 2021 (UTC)
 * 9) Per Casliber.Risker (talk) 01:29, 8 September 2021 (UTC)
 * 10) I agree with the many Support votes that push for easier de-sysopping, but I don't see those points as strong support for term limits. Given that part of the problem here is the low number of admins, I don't see persuasive enough evidence to override the plain paradox of more automatic de-sysopping somehow leading to an increase in admins. Firefangledfeathers (talk) 02:47, 8 September 2021 (UTC)
 * 11) I reject the implication that admins should be term limited. That would greatly increase bureaucracy, and admins would always be thinking about WikiPolitics in the back of their mind. CaptainEek  Edits Ho Cap'n!⚓ 21:08, 8 September 2021 (UTC)
 * 12) There is no harm in admins continuing to use their tools if they are  genuinely helping the project with them and not just batch deleting CSDs without taking a second look. Some kinds of desysoping have become more frequent only because some arbitrators see themselves as small-town sheriffs. Kudpung กุดผึ้ง (talk) 13:33, 19 September 2021 (UTC)
 * 13) This is basically no longer true. If admin accounts become inactive for a while, they are (semi-)automatically de-sysopped. If admins remain active, then any contentious actions are open to scrutiny and can result in them being de-sysopped if deemed necessary by the community via a discussion at WP:AN and/or ArbCom among other routes. Those proceedings are open, transparent, and easy to initiate. <b style="color:#98F">W</b><b style="color:#97E">a</b><b style="color:#86D">g</b><b style="color:#75C">ge</b><b  style="color:#83C">r</b><b  style="color:#728">s</b><small  style="color:#080">TALK  14:23, 27 September 2021 (UTC)
 * 14) Yes, RfA is a high stakes atmosphere because selecting the right candidate for adminship is itself a very important process. But, I clearly and definitely do not agree to any kind of term limits or annual reconfirmations. Preparing oneself for an RfA and subsequently passing it is already a hell of a tough job and it makes no sense to make a person go through this again and again (that would just be mental torture). This year's Requests for comment/Adminship term length and Requests for comment/Desysop Policy (2021) clears everything. We have Arbcom to deal with problematic admins on a case by case basis and they do a pretty good job at it. TheGeneralUser (talk) 20:23, 29 September 2021 (UTC)

G. Comments

 * 1) I do not believe in term-limited administrators, but we should have a recall procedure similar to Commons. Over there, the way it works is: A discussion is opened on AN or ANI to discuss the admin's behavior. If it appears from the discussion that the admin has lost the trust of a significant portion of the community, then a formal Request for de-adminship can be opened, with a simple majority required to remove the admin. -- <b style="color:red">King of ♥</b><b style="color:red"> ♦</b><b style="color:black"> ♣</b><b style="color:black"> ♠</b> 02:21, 29 August 2021 (UTC)
 * Commons doesn't have an Arbitration committee. Agreed if we didn't have Arbcom we would need a process such as Commons has. But Arbcom has several advantages over the commons system, and we wouldn't want the double jeopardy inherent in having two systems working at the same time.  Ϣere Spiel  Chequers  12:02, 30 August 2021 (UTC)
 * We have community bans, just as we have Arbcom bans. What's wrong with having both for desysopping? -- <b style="color:red">King of ♥</b><b style="color:red"> ♦</b><b style="color:black"> ♣</b><b style="color:black"> ♠</b> 08:22, 31 August 2021 (UTC)
 * There are quite a few reasons, not least of which is the RFAR process has an evidence gathering phase. If you build a community based system that does the same....why, when you already have it? --Hammersoft (talk) 23:17, 27 September 2021 (UTC)
 * 1) Fifteen years ago, adminship was also not term-limited, but it wasn't all that special, as it was something that most dedicated contributors would eventually get. While views about adminship have changed quite a lot, I don't get what term limits (that have not changed) have to do with it. —Kusma (talk) 12:51, 29 August 2021 (UTC)
 * 2) I do not see a problem here. Admins who are inactive get dropped; those who abusive the tools have a process to be removed; and I fear making it too easy to remove an admin would make them less likely to do the work that needs to be done. I am not entirely certain if I should be agreeing or opposing here; to be clear, I support continuing a lifetime adminship with the current methods for removing them, if necessary. Ifnord (talk) 16:15, 29 August 2021 (UTC)
 * 3) I'm going comment to start, because I can't figure out how to explain what I think is the case that "yes, the premise is true, but reversing it would not immediately change the community zeitgeist on it - the interim period could be unpleasant in terms of admin numbers". Beyond that, I think term-limited admins are unwise, for a long reason list see the Community discussion on it. I do think there are good grounds for a Community Recall method, but in the ones proposed for general adoption, it was very tough to find ones that didn't have significant safeguarding flaws. I am reticent to further agreeing to this principle because it has a "and we should change quickly" feeling to it, and am concerned that any option that could reasonably avoid these issues could be found but was not in the multiple prior discussions. Nosebagbear (talk) 11:24, 30 August 2021 (UTC)
 * 4) I support the idea of a community-based desysop procedure, I just have yet to see a setup for one that looks like it would actually gain community consensus and work as intended. Beeblebrox (talk) 21:52, 30 August 2021 (UTC)
 * 5) As shown by WP:DESYSOP2021, any attempt to change the "lifetime tenure" aspect is currently bound to fail. I don't think it's necessarily a problem leading to a lack of candidates, however. ~ ToBeFree (talk) 17:41, 31 August 2021 (UTC)
 * 6) I can support a recall system in principle, but please don't go down the road of mandatory reconfirmation. I think there are plenty of admins for whom that would be a huge waste of our time. --BDD (talk) 15:37, 1 September 2021 (UTC)
 * 7) I in fact supported the last community desysop proposal, but I am not convinced that even if eny future proposal is going to be successful, it could solve the RfA problem. I believe it will just result in us having less admins (and less work being done).--Ymblanter (talk) 15:32, 2 September 2021 (UTC)
 * (These are comments, they shouldn't be numbered?) If we aren't willing to make it mandatory, an RfA candidate should still be able to choose to be subject to a vote-based recall or reconfirmation process.  But if we allow those types of pledges, I think we need to allow complex pledges like "I will not block an ECP user unless I am administering consensus of ANI/ARE or suspect it is a compromised account" as well. User:力 (power~enwiki,  π,  ν ) 21:53, 3 September 2021 (UTC)
 * A candidate can pledge whatever they want. Such pledges are, of course, unenforceable. Same as in real life, but those people have to face the electorate in the next election, whereas admins don't. Regardless, unless crats are going to start desysopping based on broken promises, any pledge (whether about recall or something else) is nothing more than a personal assurance. I for one don't think we should move towards 'binding pledges'. It won't fix RfA, and it will create a disparity between new admins and legacy admins, the former group already more scrutinised than the latter. ProcrastinatingReader (talk) 22:58, 3 September 2021 (UTC)
 * I would be open to having a term-limit for new admins but only to give people an alternative of giving candidates a test/trial period rather than refusing outright. Stifle (talk) 11:29, 6 September 2021 (UTC)
 * Length of tenure and/or term limits do not impact on the voting. They are the last things voters have in mind when they participate at RfA.Kudpung กุดผึ้ง (talk) 19:57, 26 September 2021 (UTC)
 * Support / Agree. I don't think any editor with these privileges should be given lifetime tenure without biennial review or other requirements to ensure they're acting according to Wikipedia's 5 pillars and guidelines. Hmlarson (talk) 22:34, 27 September 2021 (UTC)

H. RfA reflects a declining editor base
The problem isn't RfA, the problem is that we have fewer editors on the whole. To fix RfA we need to increase our overall number of editors and/or retain the ones we have now better.

H. Support (agree)

 * 1) I agree but it's also a general problem that underlies all our other problems, so I'm not sure how helpful discussing editor recruitment is to the issues of adminship and RFA specifically. Levivich 07:23, 29 August 2021 (UTC)
 * While the user base isn't necessarily in decline, it seems sort of the static. The chart below shows the number has largely leveled off, but I don't think five edits a months is a good measure of determining whether a user is truly active or not. -- Calidum  16:52, 30 August 2021 (UTC)  Striking; see below  -- Calidum  18:53, 31 August 2021 (UTC)
 * 1) I do agree with this. WMF do not give any good metrics because "active editors" are not sufficiently active to be admins. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:31, 1 September 2021 (UTC)
 * 2) Full support.  The only way the pool of adminabile editors turned off by the process of RFA would be large enough to make a difference is if we went back to 2004 standards for adminship.  My estimate is that changes to make RFA "nicer" would cause a one-time boost of only 10-20 admins. User:力 (power~enwiki,  π,  ν ) 17:32, 1 September 2021 (UTC)
 * 3) At least to some extent, but not likely something we can fix by changing RfA. &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 18:13, 1 September 2021 (UTC)
 * 4) While the data shows our editor base has been fairly consistent, its still less than it could or should be. 2006 was the height of editors and near the height of RFAs. I think we still need to greatly increase the number of active editors we have. Wikipedia is now way more read, popular, and trusted than in 2006, and yet we have less editors. While this lack of editors is not the cause of RFA problems, recruiting new editors would definitely help to solve the issue. Recruiting and retaining editors remains THE number one challenge on Wikipedia, and everything is linked to it. CaptainEek  Edits Ho Cap'n!⚓ 21:03, 8 September 2021 (UTC)

H. Oppose (disagree)

 * 1) |line|2001-01-01~2021-09-01|(page_type)~content*non-content|monthly This suggests that the number of active editors (broadly construed) has actually remained surprisingly constant over the past decade, and is still far higher than it was in, say, 2004–2006, when RfA was approving hordes of sysops. While it's certainly not a perfect metric, it's enough to convince me that RfA's issues lie elsewhere. Extraordinary Writ (talk) 03:23, 29 August 2021 (UTC)
 * 2) Per Extraordinary Writ and the fact that's clear that there's an abundance of qualified candidates; the issue is just that those candidates aren't running. &#123;{u&#124; Sdkb  }&#125;  talk 07:33, 29 August 2021 (UTC)
 * 3) The stats don't bear this out, see Time Between Edits.  Hut 8.5  13:34, 29 August 2021 (UTC)
 * 4) If the stats were showing that RFA had dropped from the 2007 peak of 408 new admins back to something like the 2004 crop of 240 new admins then this theory could account for a significant part of the drop. But we had just 17 new admins in 2020, a year when the community was much larger than in 2004. Relative community size can't be a significant factor in the decline.  Ϣere  Spiel  Chequers  08:11, 30 August 2021 (UTC)
 * 5) We do need to increase our overall number of editors and/or retain the ones we have now better. But there is additionally a problem with RfA. — Bilorv ( talk ) 17:15, 31 August 2021 (UTC)
 * 6) These are two separate issues. One might even suggest that the difficulty of becoming an admin discourages participation in the site as a whole. -- Calidum  18:53, 31 August 2021 (UTC)
 * 7) I assume this means "Declining number of RfA candidates reflects a declining editor base". I agree with the replies immediately above. It may be one factor but it is not the whole story. Nurg (talk) 02:09, 2 September 2021 (UTC)
 * 8) That's just not what the data shows. — python coder  (talk &#124; contribs) 19:12, 2 September 2021 (UTC)
 * 9) As pointed out by Extraordinary Writ, active editor numbers have been stable. Trainsandotherthings (talk) 23:41, 2 September 2021 (UTC)
 * 10) This does not seem to be borne out by the data. I know of more than a few people who would make fine admins, but have indicated they are not interested in filing an RfA. If we could make that process more collaborative and less unpleasant, that might change. Seraphimblade Talk to me 16:09, 4 September 2021 (UTC)
 * 11) I moved my comment down here. I agree with exactly what Levivich said, but that's not a problem with the structure of RfA, it's just a systemic problem for Wikipedia in general. If we recruit more new editors, that won't inherently "fix" RfA. Ivanvector (Talk/Edits) 21:39, 4 September 2021 (UTC)
 * 12) I'm not sure this is borne out by the data. Stifle (talk) 11:29, 6 September 2021 (UTC)
 * 13) There doesn't seem to be a drop in the number of editors. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  23:28, 6 September 2021 (UTC)
 * 14) Don't think this is the case based on the above stats presented above which disagree with a decline in editors. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 09:43, 7 September 2021 (UTC)
 * 15) Statistically untrue. Risker (talk) 01:28, 8 September 2021 (UTC)
 * 16) "The problem isn't RfA" overstates it. The problem is, at least, both RfA and recruitment/retention more broadly. KevinL ( aka L235 · t · c) 09:38, 9 September 2021 (UTC)
 * 17) Per Risker.--John Cline (talk) 07:52, 13 September 2021 (UTC)
 * 18) We are having problems transitioning active editors to adminship (maybe retaining too), not a "declining" base, per the data. --Izno (talk) 18:52, 14 September 2021 (UTC)
 * 19) The overall editor base can both increase and decrease over time and new editors are joining all the time. Also, I agree that maybe if the overall active editors increase by a huge margin over time, we might get to see some new admins, but I don't think this issue is in any way directly related to RfA. TheGeneralUser (talk) 20:31, 29 September 2021 (UTC)

H. Comments

 * The problem is not so much fewer editors, the problem is an increasing number of spammers and POV pushers (including UPE) who have figured out that obscure areas of a top-10 web resource can be edited and the edits would stay under the radar for years.--Ymblanter (talk) 14:57, 29 August 2021 (UTC)
 * Agreed that's a real part of the problem, adminship involves scrutiny that bad actors don't want. But how significant could it be? Maybe I'm naïve, but I don't think that spammers and POV warriors are anything close to the large majority of the community that could explain the RFA problem.  Ϣere Spiel  Chequers  08:18, 30 August 2021 (UTC)
 * I do not mean spamers run RfA or intend to run RfA, I mean the admin work has significantly increased because of spammers.--Ymblanter (talk) 15:35, 2 September 2021 (UTC)
 * I agree that spammers have increased in number, but anti vandalism has been partly automated so total admin workloads are similar. My point is that the active community includes some editors who aren't running at RFA because heavy scrutiny might show them to be unpaid editors, or they are here as unpaid editors and see adminship as a distraction that might lead to them being caught.  Ϣere Spiel  Chequers  13:42, 13 September 2021 (UTC)
 * It's a complex issue and this is a factor too, but how significant? That requires in-depth studies. --<sub style="border:1px solid #228B22;padding:1px;">Piotr Konieczny aka Prokonsul Piotrus&#124; reply here 02:47, 30 August 2021 (UTC)
 * I don't understand point H. Does it mean "Declining number of RfA candidates reflects a declining editor base" or something else? Nurg (talk) 10:07, 30 August 2021 (UTC)

I. Standards needed to pass keep rising
It used to be far easier to pass RfA however the standards necessary to pass have continued to rise such that only "perfect" candidates will pass now.

I. Support (agree)

 * 1) The standards don't expect candidates to be perfect, but they are far too high and I think that's a large part of the problem. Ideally an experienced non-admin who has done plenty of admin-type work should have an easy time at RfA, at least in a typical case. In practice a lot of such people struggle, I think in particular because of the extremely high content creation standards. Looking at the nominations and answers to the first three questions for the six candidates promoted so far this year four have written multiple GAs, one has written an FL and the remaining candidate has created over 200 articles. Candidates who haven't got that kind of content creation record (e.g. Money emoji) struggle because of it. This kind of standard isn't needed for prospective admins at all.  Hut 8.5  13:32, 29 August 2021 (UTC)
 * 2) Hut 8.5 is correct. The skillset/knowledge required to become an admin does not include the ability to create entire articles or even bring them to GA. It does require enough general editing experience to have a good understanding of sources, NPOV, etc. This includes rewriting articles to remove original research, adding sources and in particular discovering unreliable sources and either replacing them or if there are no reliable sources removing the text. In other words, admin candidates should be able to demonstrate their understanding of our basic policies through their general editing, which doesn't have to involved GAs or creating articles. We aren't appointing admins to demonstrate their writing skills. I can see no reason why the best article writers would automatically be good Admins. Doug Weller  talk 13:47, 29 August 2021 (UTC)
 * 3) Standards have risen. Many very reasonable voters want to see at least some content creation, which I understand -- some of these editors have had interactions with admins who have no experience creating content and felt those admins didn't understand what content creators sometimes have to deal with. And many voters want to see an extremely high level of civility; I'm one of those voters. Those things weren't necessary in 2007. So that part of the statement I agree with. But I'm not sure that means only perfect candidates can pass. —valereee (talk) 14:28, 29 August 2021 (UTC)
 * 4) Admins shouldn't be expected to be perfect... Eddie891 Talk Work 20:19, 29 August 2021 (UTC)
 * 5) * Pppery * <sub style="color:#800000">it has begun...  20:30, 29 August 2021 (UTC)
 * 6) Many prospective candidates decline RfA nomination offers until they have a certain length of tenure (~18 months seems to be the arbitrary minimum these days), a certain amount of edits (or worse, a certain ratio of namespace totals), a certain amount of GAs, a certain amount of AfD participation, a certain amount of project-space participation, a certain period of time without any content disputes, a fleshed-out CSD log, and the ultimate cliche is needing to demonstrate a "need for the tools". Only when a prospective candidate has all of these things do they consider maybe throwing their hat into the ring. I think the reality is that we tend to overemphasize trivial issues at RfA. What's more important are things like temperament and overall good judgment (e.g. can we trust the user not to jump into doing controversial actions right away in areas they don't have as much experience?)—in other words, "not a jerk, has a clue" (from User:TonyBallioni/RfA criteria). Mz7 (talk) 23:50, 29 August 2021 (UTC)
 * 7) Even if the standards haven't risen overall, there's a perception among potential candidates that they have (and that's as good as the same thing). There's significant doubt among most of the people I've talked to, and it's my opinion as well, that candidates who successfully ran based only on anti-vandalism, copyright, or technical experience 3-5 years ago would not pass if they ran the same RfA today. --AntiCompositeNumber (talk) 02:12, 30 August 2021 (UTC)
 * 8) They do, which is not a bad thing. The problem is that mud sticks. As time goes by, every active editor makes some mistakes and enemies. So actually after a few years, almost every editor will be less likely to meet our standards, as there'll be some 'dirt' on their record. Rising standards are ok, but it's crucial to treat all editors fairly, and right now, experienced, highly active editors are at a disadvantage. --<sub style="border:1px solid #228B22;padding:1px;">Piotr Konieczny aka Prokonsul Piotrus&#124; reply here  02:55, 30 August 2021 (UTC)
 * 9) Without getting into whether this is a good thing, or what the standards should be, I think it's indisputable that standards have risen since the early days, though they have arguably stayed the same for a half-dozen years. Vanamonde (Talk) 07:42, 30 August 2021 (UTC)
 * 10) Standards did rise sharply in the early days. In particular after the unbundling of Rollback in early 2008 candidates were no longer able to pass simply on the credentials of being a "good vandalfighter". I'm not convinced that as many !voters today actually spend serious time trawling through a candidates edits, so the real standard today is lower, though there has probably been a small increase in the arbitrary standards such as edit count and tenure. But I'm pretty sure those standards have not risen to the point where an RFA like my sucessful one in 2009 would now fail. What's more of a problem is that people don't want to run until they clearly exceed the tenure requirements and by then in many cases they have moved on to other hobbies instead of getting more involved in this site the way that new admins used to.  Ϣere  Spiel  Chequers  08:27, 30 August 2021 (UTC)
 * 11) It is indisputable that standards have risen. However, I don't think that is changeable through reform (standards are not codified, and exist in the minds of the voters), and I don't think that is a bad thing (they have likely risen for good reason as Tpdwkouaa argues). — Goszei (talk)  23:04, 30 August 2021 (UTC)
 * 12) I was elected to arbcom with (just) under 10k edits. Here is an example of an RfA with people talking about an admin candidate with 10k edits as insufficient. What? Rising standards aren't in and of themselves a problem, but I do think there is (and has been for a long time) a problem with insufficient social disincentive for exaggerated expectations that push the Overton window even if not widely adopted as stated. Opabinia regalis (talk) 16:58, 31 August 2021 (UTC)
 * Yes, there is no set standard, but the community expects near perfection out of potential candidates. -- Calidum  18:57, 31 August 2021 (UTC)
 * Yes, and I think this is connected to the lifetime tenure issue. Calliopejen1 (talk) 23:53, 31 August 2021 (UTC)
 * 1) Per Opabinia - 10k edits is a nonsense. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:33, 1 September 2021 (UTC)
 * 2) Admins are expected to be competent in a wide range of situations, even when explicitly stating a specific need. This is counterproductive – even with the rights bundled together the way they currently are. /Julle (talk) 16:43, 1 September 2021 (UTC)
 * 3) If I was a perfect editor right now (which I'm not) and ran for RFA right now (which I wouldn't) I'd likely get shut down on the basis of "only one GA" and "only 13K edits" and "only been here for a year." But in 2000s RfAs, we regularly elected candidates with <5,000 edits, and no content creation. 🐔 Chicdat <sup style="font-family:Times New Roman">Bawk to me!  11:13, 2 September 2021 (UTC)
 * 4) I'm not sure anyone could disagree that 2000s RfAs had wildly different standards for applicants. There is some merit to increased standards (back in the early years nobody had significant contribution history for the most part) but it has gone entirely too far today. Different community members have different standards for RfA applicants, some bordering on absurd - why is GA or even FA creation so important? You do not need to be an admin to do GA review, or to create GAs. Admins are not involved in the process. As a community, we should really think about what kind of user contributions we look at to determine if someone can be trusted with the mop. GAs and FAs are just one of very many different and important ways one can contribute to the encyclopedia. Yet, a few GAs are more important than making dozens of new articles, major efforts in copyediting or CCI, or other kinds of contributions, in the eyes of some. There's also an arbitrary focus on high edit count, which is an imperfect measure. Just look at my edit history for a perfect example: I have over 2,000 edits, but a large number are from doing assessments and aren't in article space. Trainsandotherthings (talk) 23:50, 2 September 2021 (UTC)
 * 5) The standards seem unrealistically high so we have fewer editors going through an RfA who are relatively new (less than 3 years of editing). It also can't go without notice that not all experience holds the same value for RfA participants...it is not impossible for an editor who is technically oriented (bots, scripts, templates and such) to pass an RfA but they have much more difficulty than a content creator. Actually, everyone who isn't a big content creator generally has a very rough time. A lack of substantial content creation is a factor that will cause some editors to Oppose, regardless of whether the candidate is otherwise promising. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 02:22, 4 September 2021 (UTC)
 * 6) Yes, it keeps raising and something must be done now, before things gets completely broken. People are on wrong path about "minimum requirements" to be a sysop. Carlosguitar (Yes Executor?) 00:04, 6 September 2021 (UTC)
 * 7) Trivially true. Stifle (talk) 11:29, 6 September 2021 (UTC)
 * 8) The required standard is unnecessarily high. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  23:17, 6 September 2021 (UTC)
 * 9) This is true, but not necessarily a bad thing. Given the rise in complexity of our other policies and processes, a large portion of the rise in standards is justified. KevinL ( aka L235 · t · c) 18:44, 8 September 2021 (UTC)
 * 10) The bar is now ludicrously high. Had I run say 15 years ago, I would likely have been able to run 6 months earlier and also breezed through. But under the modern bar, I ran at the earliest possible time and still barely scraped by. The baseline should be: "Has clue, not a jerk", but for the most part that standard has been left behind in the dust. It feels the same as applying to college today vs. 50 years ago. I think one of the biggest misleading things is tenure and edit count. I definitely inflated my edit count using Huggle in the leadup to my run, thinking that without 10k edits I wouldn't pass. But realistically, I think an editor with even just 5k quality edits should be able to be an admin. As a practical note, there is an xTool out there somewhere that tells you the odds you will pass an RFA, and it includes the 10k edit count as a bar. If someone knows the link to that xtools, ping me would you? I think things like that are very harmful.  CaptainEek  Edits Ho Cap'n!⚓ 20:52, 8 September 2021 (UTC)
 * The XTools score is here, with the explanation behind it here. I don't think people take it seriously and given how the score for edit count is calculated, an edit count of 80 seems more than reasonable... Sdrqaz (talk) 21:33, 8 September 2021 (UTC)
 * 1) The only criterion should be "does the community trust this person?". Being deemed generally trustworthy among a community that has "assume good faith" as one of its central pillars means it should be vanishingly rare for an AfC to fail, and AfCs should be encouraging, rewarding experiences for the candidate, full of people saying how trustworthy the person is! <b style="color:#98F">W</b><b style="color:#97E">a</b><b style="color:#86D">g</b><b style="color:#75C">ge</b><b  style="color:#83C">r</b><b  style="color:#728">s</b><small  style="color:#080">TALK  14:41, 27 September 2021 (UTC)
 * 2) I can definitely agree with this. The standards necessary to pass have continued to increase upto such a point that only near perfect candidates normally pass now with a few exceptional cases. No one person can be good or perfect at everything and everyone has a different kind of interest and skill set. A basic criteria is necessary, but every candidate is different and thus needs to be evaluated on their own merit. TheGeneralUser (talk) 20:49, 29 September 2021 (UTC)

I. Oppose (disagree)

 * 1) Given the contemporary prominence of Wikipedia, and the substantially larger userbase and task backlog, I don't see why higher standards should be viewed negatively. I'm also not convinced that they're objectively high, rather than just comparatively to the early days of the website. /Tpdwkouaa (talk) 04:13, 29 August 2021 (UTC)
 * 2) RfA doesn't have standards; individual voters have standards. And the fact that pretty much every RfA usually rapidly accrues 50-70 unqualified support votes in the first couple of days shows that most people's standards are low. What really determines whether an RfA will succeed is whether those who subsequently raise concerns (if there are any) are persuasive enough to overcome that mass of near-automatic support. The standards of those people vary too much to comment on; certainly they allow for a lot of 'not perfect' candidates to pass. As long as RfA remains an unstructured vote, it is meaningless to talk about its standards and futile to try and change them. –&#8239;Joe (talk) 04:59, 30 August 2021 (UTC)
 * 3) Per Joe Roe and others in the "agree" section; I doubt this is that much an issue. I'm far from perfect and got through RfA unopposed, so I doubt any increase in standards post-2010 are, if existent, a bad thing. – John M Wolfson (talk • contribs) 14:55, 30 August 2021 (UTC)
 * 4) Agree with Wolfson, above. Chetsford (talk) 23:03, 30 August 2021 (UTC)
 * 5) I disagree with "the standards necessary to pass have continued to rise".  Surely the standards have risen from those of the 2003-2006 timeframe.  But they stopped rising years ago.  I've actually decreased my tenure recommendation from 24 months of being a "very active editor" to 18 months based on recent RFA results. User:力 (power~enwiki,  π,  ν ) 17:34, 1 September 2021 (UTC)
 * 6) Quite often potential opposition in an RfA can be predicted and planned for. I've seen candidates keen to try without planning to deal with oppostion they were warned would take place based on their editing habits. Cas Liber (talk · contribs) 12:17, 7 September 2021 (UTC)
 * 7) My thoughts echo Beeblebrox below. Standards are definitely higher than in the no big deal era (that paraphrased message was written in 2003, which is before some active editors today were ), but I thought going through the RfAs that standards did peak years ago and the pendulum's swung back the other way, with (relatively) uncontentious RfAs, with one notable exception. I think some of the candidates this year and last year would have attracted more opposition in different years. Maybe that's because it's become more in vogue to badger opposers and neutrals, or maybe it's because standards have actually fallen slightly. Sdrqaz (talk) 20:02, 7 September 2021 (UTC)
 * 8) Standards have been the same for years when one examines the last 400  or so RfA. Those who are not such frequent voters will not have noticed, but the voters, certainly since the December 2015 reforms, are a fickle and volatile crowd. It  just  takes one to  demand a couple of FAs, or 30,000 edits, or 3 years tenure or some other ridiculous criterion and everyone of the pile-on voter crowd follows suite for a while until it gets out of fashion again. Kudpung กุดผึ้ง (talk) 13:55, 19 September 2021 (UTC)

I. Comments

 * 1) There are two types of standards here, and we should be clear what we're talking about. One is the mentality of !voters, i.e. where the threshold between support and oppose lies when a !voter decides how to !vote. This is a cultural issue and is difficult to change. The other is the !vote threshold, which can be adjusted if beneficial. I think we should do a study of the fate of all successful RfA candidates, with support percentage and tenure as input variables and current status as the output variable. The current status could be: still an admin, desysopped not for cause (e.g. inactivity or voluntary resignation), desysopped for cause (e.g. by Arbcom or resigned under a cloud). If there is no significant negative correlation between support percentage and rate of being desysopped for cause, then we can consider lowering the crat chat range to something like 60-70%. -- <b style="color:red">King of ♥</b><b style="color:red"> ♦</b><b style="color:black"> ♣</b><b style="color:black"> ♠</b> 04:16, 29 August 2021 (UTC)
 * , I consider this an excellent proposal. One could also do a variant on the negative, asking: "Is there a correlation between higher unsuccessful vote % and likelihood of successful re-RfA?" If the answer is yes, in my opinion it would further justify lowering the threshold to make candidates who are going to get the mop eventually jump through fewer hoops.— Shibboleth ink  (♔ ♕) 02:33, 30 August 2021 (UTC)
 * I threw together a data set of all RfAs from the last RfA reform (2016) until 2020 and also all people who've been desysopped by ArbCom since 2016. Best, Barkeep49 (talk) 03:35, 30 August 2021 (UTC)
 * , In this set, is the second sheet (ArbCom) the list of admins who have been de-sysop'd for cause since 2016? and in that case, what does "crat?" in that sheet mean? They were de-crat'd by ArbCom? Does the "2nd or more" column indicate those who were de-sysop'd for cause, and had run 2+ RfAs? Thanks — Shibboleth ink  (♔ ♕) 22:29, 30 August 2021 (UTC)
 * , I think "ArbCom" is indeed all admins desysopped by ArbCom for cause since 2016; "crat" refers to RfAs that went to a crat chat; "2nd or later" means that the RfA in question was not that editor's first. Cheers. Extraordinary Writ (talk) 22:37, 30 August 2021 (UTC)
 * , thank you! I think you're right. And hence why it makes sense that crat chats are universally at the lower end of the RfA % in both sheets. Helpful.— Shibboleth ink  (♔ ♕) 22:39, 30 August 2021 (UTC)
 * EW's explanations are correct. "Advanced" shows if they got CU and/or OS (funct) or were elected to Arbcom (Arb). If boht happened I only noted Arb. Best, Barkeep49 (talk) 22:57, 30 August 2021 (UTC)
 * Unfortunately I don't think we have nearly a big enough sample size to do that kind of analysis. Or at least to get one, we'd have to go back years and years to when the project was a different place. But the recent case of an admin who passed RfA at 64% (below the current discretionary range), after a controversial crat chat, and was desysopped in an acrimonious arbitration case just two years later, makes me think that dropping the range any lower than it is now would be more trouble than it's worth. You're going to have a rough time as an admin if over a third of your peers don't think you can be trusted with the mop. –&#8239;Joe (talk) 05:06, 30 August 2021 (UTC)
 * 1) The "you have to have written an FA to pass RfA" thing is a myth, isn't it? RfA votes generally look for a longer editing record than they did a decade ago, sure, but when was the last time one actually failed for lack of high-level content editing?  cites Requests for adminship/Money emoji, but with all respect to Moneytrees (I didn't vote but think they've done an excellent job), the concern there was that he had created one article and made less than 10k edits, not that he hadn't written multiple GAs and FAs and hundreds of articles, and to be honest that's a fair concern. And in the end he did pass. –&#8239;Joe (talk) 04:48, 30 August 2021 (UTC)
 * "You must have an FA to pass RfA" isn't correct, but "You must have a GA" isn't far off the mark. As I've noted of the 6 successful candidates so far this year 5 had written at least one GA/FA/FL and the other one had created over 200 articles. Of the 17 successful candidates last year 13 (76%) claimed credit for at least one GA/FA/FL and 10 (59%) claimed credit for more than one. Of the four who didn't claim an FA/GA/FL one was Jackmcbarn (an former admin), one was Hammersoft (who was ludicrously overqualified at 70k edits since 2007), one was QEDK (who got several opposes for lack of content creation experience despite having written several articles) and the last one was Money emoji (who did pass, yes, but by an extremely narrow margin after the longest bureaucrat chat in RfA history). It certainly looks like people expect a typical candidate to have improved at least one article to that level and preferably several. There isn't any good reason for these very stringent content creation requirements, there is no admin job which involves writing content and very few which involve judging content at that level.  Hut 8.5  08:57, 30 August 2021 (UTC)
 * I think Joe is right in that it's a bit of a myth. I did a skim through post-2018 RfAs and, since 2018, the only ones to fail for content creation were WP:Requests for adminship/1997kB and WP:Requests for adminship/Jbhunley. Of the successful candidates, many did not have a single GA/FA but still passed. Sure, they gathered some opposes for it, but between 10-20 opposes is a natural fluctuation at RfA for one reason or another, and on avg doesn't cause the support percent to drop below 90%. I don't think minority opinions at RfA indicate any systemic problem. ProcrastinatingReader (talk) 11:17, 30 August 2021 (UTC)
 * I don't agree with that at all. The vast majority of people who pass have either written at least one GA/FA or can point to some other big example of content creation instead. People who haven't got that either don't run, fail, or pass with difficulty. Of the unsuccessful RfAs in 2020/1 L293D and Vami IV could point to multiple GA/FAs, and Guy Macon's nomination describes him as a "tireless content creator". 1997kB and Steel1943 failed with content being one of the main reasons. That just leaves CASSIOPEIA and Chongkian failing for any other reason, and content creation probably would have come up a lot more if the candidates hadn't made more glaring mistakes.  Hut 8.5  13:05, 30 August 2021 (UTC)
 * The former two editors did not fail due to content contributions, so I don't see why that's relevant. Steel's outcome wasn't solely due to content creation, per se. Guy might have had content creation, but the opposes were for other reasons.
 * "Creating content" (DYKs, expanding stubs, creating start-class articles, etc) is not the same as the "needs GA/FA". All I'm saying is that the evidence doesn't support the idea that lacking a GA/FA causes an RfA to fail. Only two RfAs in the past three years failed for 'content creation' reasons, and those (AFAICS) had minimal content creation, not just lack of GA/FA. Several have passed without GA/FA: Requests for adminship/Izno, Requests for adminship/EvergreenFir, Requests for adminship/Johnuniq, Requests for adminship/Enterprisey 2, Requests for adminship/JJMC89, etc. They got a few opposes for "fails my criteria"/"no GA/FA" but nothing substantial (Izno passed with 98%).
 * Common wisdom being that GA/FA is required probably explains why most successful candidates happen to have some peer-reviewed content. Trialpears started Featured list candidates/List of countries by Human Development Index/archive3 after he said he was going for RfA soon, and shortly before his actual RfA, so the nomination was likely part of 'RfA prep'. That doesn't mean his RfA would've failed had he not started the process to get it to FL. Precedent suggests it would've passed anyway, but perhaps with less support than it did. Ultimately, "some people will oppose over X" does not mean "RfA is unpleasant due to X" or "your RfA will fail due to X". As such, I don't think this is really a problem. ProcrastinatingReader (talk) 13:26, 30 August 2021 (UTC)
 * The first two RfAs are relevant because they show that RfA candidates often have FA/GAs, which in turn suggests that FAs/GAs are seen as a prerequisite for adminship, at least by prospective RfA candidates. People don't file RfAs unless they think they are going to pass. I assume there aren't many examples of failed RfAs where the candidate has recent blocks. That doesn't mean that blocks aren't seen as a problem in an RfA candidate, because people who have been recently blocked know they won't pass and don't apply. Similarly the fact that there aren't that many RfAs which failed for reasons relating to content creation doesn't mean that content creation isn't a problem (although content creation was clearly one of the main problems in the Steel1943 RfA). And yes I absolutely agree that "creating content" doesn't mean GA, FA etc, but there are a fair number of people who want RfA candidates to have that or something similar to qualify as a "content creator".  Hut 8.5  14:08, 30 August 2021 (UTC)
 * @Hut 8.5, the objection isn't that admin jobs require the ability to write/create content. It's that without having created content, admins may not understand what content creators are faced with. Some content creators feel their interactions with admins who've never had to face the problems content creators face have been affected by that lack of experience. —valereee (talk) 13:32, 30 August 2021 (UTC)
 * Since we don't have a "neutral/mixed" category if people think a "yes compared to 2015, but roughly stable in last 3 years" is a position with caveat that belongs more in "support" or "oppose"? Nosebagbear (talk) 11:28, 30 August 2021 (UTC)
 * I'm right about here. Izno (talk) 18:50, 14 September 2021 (UTC)
 * It seems to me like they fluctuate with time. Before 2007 they were laughably low, and that's what got us a lot of "legacy admins" who have caused a number of problems and been at the center of several arbitration cases. Then standards went up, and up, and up for several years, and seem to me to have relaxed a bit from the peak around 2012-2015. Moving the required percentage to pass also helped. Beeblebrox (talk) 21:54, 30 August 2021 (UTC)
 * We've been billboarding our high standards for years, what else is new? If the goal was to discourage applicants, well, it's worked. Probably for the best as any qualified applicant running with less than 2,000 edits would be slammed with accusations of sock puppetry since they didn't take long enough to get a clue. wbm1058 (talk) 13:03, 7 September 2021 (UTC)
 * It's not a myth. It's very real. It just takes one clueless voter to demand a couple of FAs, or 30,000 edits, or 3 years tenure or some other ridiculous criterion and everyone of the pile-on voter crowd follows suit for a while until it gets out of fashion again. But  there are two conflicting ideas above: That  they  influence the outcome, and that they don't influence the outcome. Kudpung กุดผึ้ง (talk) 14:07, 19 September 2021 (UTC)

J. Too few candidates
There are too few candidates. This not only limits the number of new admin we get but also makes it harder to identify other RfA issues because we have such a small sample size.

J. Support (agree)

 * 1) I agree that there are too few candidates but this is a result of other issues at RfA rather than being a problem by itself. Candidates aren't running because they don't think they would pass, don't want to go through the process or don't want to be admins. We need to look at those problems instead.  Hut 8.5  14:09, 29 August 2021 (UTC)
 * 2) Agree with both. I'd like to know why good candidates turn down an offer to nominate. When I approach someone who just says, "Not interested", it feels intrusive to follow up with a question about the reason. I figure if they'd wanted to share that, they would have. But it would be good to have actual data rather than making assumptions. Maybe we could ask such candidates if they'd be willing to fill out an anonymous survey? —valereee (talk) 15:02, 29 August 2021 (UTC)
 * 3) Part of the problem with RfA right now is that there aren't enough candidates, and a diversity of candidates, to get a good grasp on what the community's standards actually are. Almost all of the recent candidates were successful or unsuccessful by a wide margin. Only 7 RfAs in 2020 and 2021 so far have ended even close to the discretionary range, and that's counting 4 that withdrew early. --AntiCompositeNumber (talk) 02:19, 30 August 2021 (UTC)
 * 4) Clearly a symptom, rather than the root of the problem, but as an observation I don't think this is in dispute: furthermore, I'd say there are too few willing candidates, rather than a shortage of qualified candidates. Vanamonde (Talk) 07:46, 30 August 2021 (UTC)
 * 5) This is primarily a symptom, but not only such. The issue raised in the question is correct, but I think there's another "fewer candidates ups the review of each". Nosebagbear (talk) 11:30, 30 August 2021 (UTC)
 * 6) RfA is one of the funnest parts of Wikipedia, IMO. It's such a shame that they have been reduced to a trickle. – John M Wolfson (talk • contribs) 14:59, 30 August 2021 (UTC)
 * 7) Agree. Z1720 (talk) 15:08, 30 August 2021 (UTC)
 * 8) Agree, and I wonder if this isn't a cause and not just a symptom. More thoughts in J. Comments, below. Firefangledfeathers (talk) 17:13, 30 August 2021 (UTC)
 * 9) Yes, and more. Fewer people running for RfA leads to increased drama at each RfA, because so many !voters, peanut gallery members and crypto-Wikipedians are looking for a fight. Fewer RfAs mean that each candidate has to be more confident in order to think they are worth the inherently big deal that an RfA is. This is not the root cause, but it's half of the vicious circle. — Bilorv ( talk ) 17:22, 31 August 2021 (UTC)
 * 10) Concur with Valereee. It's very hard to draw conclusions on such a tiny data set. Six successful RfAs this year out of eight total. A single data point in those eight RfAs can get construed as "Well, in almost 13% of RfAs this issue arose.." Ok nobody's going to say that, but the lack of data undermines analysis. An idea for a survey as Valereee suggests; use List of Wikipedians by number of edits as a basis for who receives the request, ensuring that the recipients never ran at RfA, and ask questions about why they aren't running, what obstacles they see, etc. That would be quite useful. There's a potential for well north of a hundred respondents, and possibly hundreds. There's a lot of hand-wringing in sections above about this or that being the problem. Nobody really knows. We need data to know. We don't have it. Without, conjecture such as this section above is just conjecture. You might just as well conclude that the lack of administrators candidates is because we're in a solar minimum. --Hammersoft (talk) 15:23, 1 September 2021 (UTC)
 * 11) Agree, but I can't think of a solution. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:34, 1 September 2021 (UTC)
 * 12) This is technically correct.  I'm not sure it's needed in a consensus statement more than "only admins can view deleted content" would be, but it is correct. User:力 (power~enwiki,  π,  ν ) 17:37, 1 September 2021 (UTC)
 * Yes, but unless we're talking about "not enough new editors" in general (which is a very valid thing to be worried about, but bigger than this thread), there are only too few candidates because the candidates we do have either don't want to go through the process or wouldn't succeed. &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 18:11, 1 September 2021 (UTC)
 * 1) Yes, RfAs are a bit like walking on hot coal by choice to prove your commitment, but I believe we are now at the juncture where the sample size is just too small to say that the RfA atmosphere is the only reason that candidates don't pass, at this point, what we need is for more candidates to run - this won't change the inherent nature of RfAs but my conjecture (feel free to name it after me) at this point is that even half-decent candidates can make it through with half-decent noms. The only requirement, from my perspective is people who haven't done some stupid stuff in recent history, WP:BEANS, really. --<span style="font-family:'Trebuchet MS',Geneva,sans-serif"> qedk  ( t  愛  c ) 20:06, 1 September 2021 (UTC)
 * 2) Don't know off the top of my head how many candidates there have been this year but I can say that the number is too low. — python coder  (talk &#124; contribs) 19:16, 2 September 2021 (UTC)
 * 3) An objective truth, as even the disagree votes mention. With a small sample size, it's hard to conclude much of anything beyond that the sample size is too small. Trainsandotherthings (talk) 23:51, 2 September 2021 (UTC)
 * 4) Corrosive RfA atmosphere is one of main reasons why there is few candidates. Carlosguitar (Yes Executor?) 00:16, 6 September 2021 (UTC)
 * Yes, but this is a symptom, not a problem. Stifle (talk) 11:33, 6 September 2021 (UTC)
 * 1) It's fairly obvious that there aren't enough candidates. However I think that we still have enough applicants to draw conclusions. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  23:31, 6 September 2021 (UTC)
 * 2) Cas Liber (talk · contribs) 12:19, 7 September 2021 (UTC)
 * 3) What  said. KevinL ( aka L235 · t · c) 18:44, 8 September 2021 (UTC)
 * 4) Largely per Hut 8.5. Extraordinary Writ (talk) 03:00, 13 September 2021 (UTC)
 * 5) I agree both with the statement here and with Ivanvector's comment down the way: having a small sample size is a symptom, not one of the problems. --Izno (talk) 18:12, 14 September 2021 (UTC)
 * 6) Agree with Stifle here. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 13:26, 27 September 2021 (UTC)
 * 7) I surmise there are two few candidates in part because the processes and environment allow for harassment and incivility. Historically, a significant percentage of the personalities attracted to these types of roles have not been held accountable for ill behavior and the accountability process is yet another pandora's box for harassment and incivility, for which many contributors do not have the time or energy. Hmlarson (talk) 22:45, 27 September 2021 (UTC)
 * 8) This is definitely true. Less and less number of people have been running for RfA since the past several years. Maybe many of them want to become admins, but just do not want to go through the increased scrutiny and the overall process of RfA (mainly the negative aspects of it). We need to find a genuine way to get more people running for RfA and something needs to be done about it. TheGeneralUser (talk) 20:59, 29 September 2021 (UTC)

J. Oppose (disagree)

 * 1) I'm not disagreeing that there are too few candidates, or the self-evident consequences that it limits the number of new admins and makes for a small sample of candidates. But this is a symptom, not a cause. Even if there were more candidates, you could only learn so much by looking at candidatures (successful and unsuccessful). It's more useful putting the horse before the cart and looking at the reasons why there is a low number of candidates, rather than looking at the consequences of there being a low number. (BTW, I don't know why others who think this is not a cause (only a symptom) are putting their comments in the 'agree' section!) Nurg (talk) 00:35, 31 August 2021 (UTC)
 * 2) I don't disagree there are too few candidates. I disagree, however, that "This...makes it harder to identify other RfA issues because we have such a small sample size." What is the point of going through this exercise if that were true? -- Calidum  19:00, 1 September 2021 (UTC)

J. Comments

 * This is largely a byproduct of other issues, and not really an issue in its own right. Hog Farm Talk 02:34, 29 August 2021 (UTC)
 * This could be read in different ways. One is "there's not enough qualified editors who would make good admins to form a candidate pool"; that really doesn't seem plausible to me. Another is "the fact that candidacies are so rare means there's more of a spotlight on them and less information to go off of when deciding how to reform. @Barkeep49, it might be worth clarifying. &#123;{u&#124; Sdkb  }&#125;  talk 07:26, 29 August 2021 (UTC)
 * I have intentionally tried to write these broadly so we didn't end up with many permutations. For this statement I think it encompasses the ideas of "We don't know what to fix at rfa because of low numbers ","We need better recruitment methods", "We need better training/mentoring". Perhaps others as well. Best, Barkeep49 (talk) 14:48, 29 August 2021 (UTC)
 * I have had over a dozen (two dozen?) qualified candidates decline nominations over the last 2-3 years. This is in comparison to the five I have nominated or conominated, who were (eventually) persuaded to accept, sometimes by someone else. Vanamonde (Talk) 07:47, 30 August 2021 (UTC)
 * Does "candidate" here mean a person who runs for RfA, or a person who could run, if only they were willing? Nurg (talk) 10:24, 30 August 2021 (UTC)
 * It means people who run for RFA. We have lots of editors who would probably pass if they ran.  Ϣere Spiel  Chequers  12:08, 30 August 2021 (UTC)
 * If it is true that, as suggested in another section, that today's editing population tends to drift away from Wikipedia before they are both interested in being an administrator and feel confident in making a successful request, then from a supply point of view, we should consider how to get editors to sign up for administrative work while they're still engaged on the site. Giving editors the option to request administrative privileges for a limited period of time (after which they would solicit feedback from the community and mutually decide on continuing with another fixed term, or an open-ended granting of privileges) might be helpful. The community might be more willing to let an editor take on administrative chores knowing that they will be able to review the results, and the editor in question might be more willing to try it out for a brief period, before they decide if they want to continue participating on Wikipedia. isaacl (talk) 15:15, 30 August 2021 (UTC)
 * At least, in the form of "too few nominations", this could be a part of the problem. Here, I am fully in the realm of speculation, and would be happy to be shut down by those with more experience/evidence. Vanamonde's comment above suggests that they've made something like one nomination per month over the past few years. If only 20% of nominees accept, perhaps we should expect nominators to roughly quintuple their nomination rates? Firefangledfeathers (talk) 17:13, 30 August 2021 (UTC)
 * If this is a part of the problem, surely it is testable. If each of the 5 admins supporting this above nominated 3 candidates, then a fifteen-candidate tranche would certainly put the question to rest. There are a lot of proposed explanations here but this seems the closest to an actual hypothesis. Eggishorn  (talk) (contrib) 22:06, 31 August 2021 (UTC)
 * If it were that easy to recruit fifteen candidates, we wouldn't have a problem at all. Extraordinary Writ (talk) 22:09, 31 August 2021 (UTC)
 * If you doubt this and want to test it look for nominators among those who disagree that we have a shortage of candidates. It is a tad perverse to ask those who are aware of a problem to somehow disprove their own position.  Ϣere Spiel  Chequers  14:04, 1 September 2021 (UTC)
 * , Who said I doubted this? Who said "testing" automatically means "testing to dismiss"? Far from being perverse, formulating a hypothesis and testing it is the standard scientific method. Eggishorn  (talk) (contrib) 17:52, 1 September 2021 (UTC)
 * Testing the proposition that "There are too few candidates" could be done in a couple of ways. People could double check the work that I and others have done to compile RFA_by_month, which lists both successful RFAs per month, and total candidates per year. Or they could try making the case that we ultimately don't need as many admins. Emailing a bunch of potential candidates and asking them if they'd like to run wouldn't test the too few candidates theory, it would test the hypothesis that one of the reasons why we have too few candidates is that lots of qualified people don't want to run. Running fifteen candidates ignores the discussion as to whether there are too few candidates and why there are too few candidates, including the testimony from various nominators that we have little difficulty finding qualified Wikipedians who would make great candidates, its just that those people mostly don't want to run. It instead tests various theories such as whether the number of questions per RFA is a result of the small number of RFAs per month.  Ϣere  Spiel  Chequers  18:43, 1 September 2021 (UTC)
 * This isn't a problem with RfA, this is a result of the problems with RfA. Ivanvector (Talk/Edits) 18:59, 4 September 2021 (UTC)

K. Too many questions
The number of questions that candidates receive creates substantial extra work for the candidate without benefiting the community as a whole in most cases.

K. Support (agree)

 * 1) It's not the number of questions so much as the number of unhelpful/ridiculous questions. Levivich 07:25, 29 August 2021 (UTC)
 * 2) What Levivich said. —valereee (talk) 12:46, 29 August 2021 (UTC)
 * 3) Agree. I think part of the issue is how many editors end up asking redundant/unhelpful questions, just to ask. Sometimes I think community vetting of questions might be helpful, but that just adds more bureaucracy... Maybe a shorter question period?—  Shibboleth ink  (♔ ♕) 21:54, 29 August 2021 (UTC)
 * 4) Two dozen questions per RFA is a sympton of the lack of RFAs. Asking questions at RFA is a phase that many potential admins go through and there being few RFAs for such questions. If we were back in the era of a couple of dozen RFAs per month then the number of questions per RFA would fall, even if the average number of RFA questions per month was many times higher. The bigger problem is that few are spending the time to seriously scrutinise the candidodate and ask the sort of diff supported question that makes RFA work. It would be good sometimes to ask questioners why they are asking that specific question of that specific candidate. But again, if there were multiple candidates per week it would become obvious when people ask questions without first spending half an hour looking at the candiate's contributions.  Ϣere  Spiel  Chequers  12:17, 30 August 2021 (UTC)
 * 5) Agree with Lev. Too many contrived hypotheticals, idiosyncratic quizzes, personal bugbears, and roundabout expressions of disapproval. &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 18:10, 1 September 2021 (UTC)
 * 6) The candidate's body of work is much more meaningful than questions; per User:Valereee/RfA questions as well.  Spencer T• C 22:43, 1 September 2021 (UTC)
 * 7) Joining the chorus agreeing with Levivich. Many questions can be trivial, off-topic, or absurd. There should be better standards for what questions are considered appropriate. Trainsandotherthings (talk) 23:53, 2 September 2021 (UTC)
 * 8) Absolutely agree. Rhododendrites said it very well. It is off-putting even to many voters. I usually stop reading after five questions.--Darwinek (talk) 23:55, 3 September 2021 (UTC)
 * Yes, but really only a minor point. Stifle (talk) 11:34, 6 September 2021 (UTC)
 * 1) Yes, there are too many questions. Indeed I think that some questioners asks questions only to make themselves look clever. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  23:35, 6 September 2021 (UTC)
 * 2) I think this is perhaps a bit more important than people realise; it is part of what drives the "you must be always available throughout the RFA" mindset, and most of them do not help voters reach a decision. Risker (talk) 01:17, 8 September 2021 (UTC)
 * 3) Per, and  is quite right too, but yes, there are still generally far too many questions. When I was still an active voter, I used to disregard the voters' questions and vote based on my own research, at least until I saw an avalanche of opposers based on the answer to a question, then I would go and investigate what it was all about. I have only asked questions three times on all the 400  so RfA I voted on - so that's just about how important I think the questions are. When I was doing research for WP:RFA2012, I put together a catalogue of around 200 of the most ridiculous and/or inappropriate questions. That section of the research never got discussed. I think because too many of the people following the research were embarrassed at seeing their own questions in the list. Some silly or superfluous questions even get asked by admins. I can't understand why. Kudpung กุดผึ้ง (talk) 14:26, 19 September 2021 (UTC)
 * 4) Another +1 for the fact it's not just the quantity, but the quality of questions that is an issue here. <b style="color:#98F">W</b><b style="color:#97E">a</b><b style="color:#86D">g</b><b style="color:#75C">ge</b><b  style="color:#83C">r</b><b  style="color:#728">s</b><small  style="color:#080">TALK  14:46, 27 September 2021 (UTC)

K. Oppose (disagree)

 * 1) If you want the responsibility of the tools, you should be willing to put in a bit of work to answer some questions. It's not a huge burden. &#123;{u&#124; Sdkb  }&#125;  talk 02:12, 29 August 2021 (UTC)
 * I was asked a question at my RfA for which I probably spent 2+ hours researching and composing an answer. Not one of them did I spit out an answer off the top of my head and hit publish. I was lucky in that I got only 13 questions; some recently have gotten 27. I don't think it's accurate to dismiss this as not burdensome. —valereee (talk) 21:06, 29 August 2021 (UTC)
 * I will echo Valeree here. There was at least one or two questions on my RFA that I took inordinate time to answer and I am similarly thankful I had only the 16. Many questions are clearly asked from "what the asker cares about" (and/or what the asker is pushing a POV about at any given time) and not necessarily what the candidate needs to answer. --Izno (talk) 18:10, 14 September 2021 (UTC)
 * 1) Most recent RfA's that I've seen have at least one or two questions that go unanswered, usually those later in the process. I've never seen anyone take issue with this, so it seems that nominees have sufficient leeway regarding this "extra work". Besides, the answers are typically the most immediate indicator as to the legitimacy of the nominee's case for the mop. /Tpdwkouaa (talk) 04:09, 29 August 2021 (UTC)
 * @Tpdwkouaa, just six months ago three people commented negatively because the candidate hadn't answered quickly enough a question from an account that turned out to be a sockpuppet. —valereee (talk) 12:58, 29 August 2021 (UTC)
 * 1) I think the questions are what you make of them and aren't the problem in and of themselves. I quite enjoyed answering questions in my RFA, felt they offered an opportunity to explain myself and show what I thought. I'd argue bad questions can be answered quickly and aren't that big of a deal. Perhaps a shift in RFA culture where it's acceptable to not answer all questions would be helpful, but it's not the question's themselves that's the issue there, it's the atmosphere. Eddie891 Talk Work 21:18, 29 August 2021 (UTC)
 * 2) I think back to when it was common to see UAA or copyright quizzes at RfA. Or questions about a candidate's opinion on a particular hot topic policy/guideline question. Those questions were truly over the top. As far as bad questions go, candidates are always free to not answer questions they don't wish to dignify with an answer. However, admins often get plenty of "dumb questions" and how they respond to them can be quite telling. --AntiCompositeNumber (talk) 02:25, 30 August 2021 (UTC)
 * 3) I've chatted to various others of the 2018-onwards RfA crowd, and question numbers seem pretty low down the list of issues. For a start, given that we have a de-facto requirement on the candidate not to speak during an RfA, the questions are the only way for them to voice. As such, more questions is good. And with the loss of UAA quizzes and so on, the idiotic questions usually aren't too time-consuming. Beyond that, as a !voter, I have actually picked out helpful knowledge on how they handle them. I oppose both the idea there are too many questions, and that significant limitations on their topics compared to 2020-onwards should be done. Nosebagbear (talk) 11:33, 30 August 2021 (UTC)
 * 4) It's not the number of questions, although that's very high, so much as the nature of the questions. There are people who come to RfA to act like the interviewer in a job interview.  Asking a trap question at RfA should be grounds for a topic ban.—<b style="font-family: Verdana; color: Maroon;">S Marshall</b> T/C 16:58, 30 August 2021 (UTC)
 * 5) I don't see any problem with a high number of questions. On the contrary, the way that a candidate chooses to respond (or not respond) to "dumb", redundant, or unhelpful questions speaks greatly on their disposition and even-headedness. There should be no banned questions, outside of our regular protections on harassment/trolling/etc. — Goszei (talk) 23:04, 30 August 2021 (UTC)
 * 6) I think we've seen actions to address this, including the two question limit and the work  regularly does in challenging questions that shouldn't have been asked. I don't think the question section is where the most critical problem is. A bad question is something you can at least dispute the premise of in your answer (and people can do this without getting opposed for it), whereas a bad oppose is something you absolutely cannot reply to or you'll get hit with 10 immediate opposes for breaking a social convention. — Bilorv ( talk ) 17:27, 31 August 2021 (UTC)
 * 7) I have never seen the number of questions as a problem. They are a place the candidate can actually address issues, without "badgering". An RfA is something that requires some actual attention, just as, say an article review would be. Being able to show where you're competent (or not) by answering questions is just what we need. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:36, 1 September 2021 (UTC)
 * 8) Neither the number nor nature of the questions is a problem. Actual admins will have to respond to loaded questions and the like, not to mention some pretty silly and ill-founded requests. Someone who lacks the temperament to do that even for a brief period lacks the temperament to be an admin, and often a candidate's skillful handling of such a situation can actually reflect very well on them. Seraphimblade Talk to me 00:03, 4 September 2021 (UTC)
 * 9) I think this situation was improved a lot when the 2 question limit was imposed. Looking at past RfAs, things in the past could get very excessive but the limit has helped. I'll note though that "questions" vary quite a lot from easy questions that encourage a candidate to explain their position on policies to ones that seem designed as "gotcha" questions that focus on a candidate's weaknesses. I don't see a solution to this but RfA participants vary in how much importance they attach to questions and answers. For some participants, a candidate's answers can affect their vote (almost always negatively) but it's clear that other participants don't pay them much attention at all. As an aside, I don't think I've ever seen a candidate who ignored questions pass an RfA so it is important for a candidate to stay on top of them, at least for the first few days of the RfA. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! — Preceding undated comment added 02:45, 4 September 2021 (UTC)
 * No, this isn't a problem. Nobody goes to a job interview and is given the job without any interaction with the interviewer, nor with rehearsed answers to softball questions they've gotten in advance. We've gotten a lot better at clerking away truly ridiculous questions, but otherwise, if you don't want to answer questions, you're not suited to be an administrator. Ivanvector (Talk/Edits) 19:02, 4 September 2021 (UTC)
 * 1) I'd say the bigger issue are the content of the questions rather than the quantity of them. There are many, many questions asked where there is no right answer and not in a good way: however the candidate answers them, they will either get opposed by the questioner or opposed by another editor. There are even more where the question being asked is more about axe-grinding than actual scrutiny of the candidate's suitability, such as the infobox question on Oshwah's RfA. All that being said, administrators have to deal with questions all the time to be accountable, regardless of how bad-faith they may be. I agree with those above that how a candidate answers "bad" questions (or better still, recognise when it is appropriate to decline answering) is valuable for assessing a candidate's suitability; in my opinion they are more important than nomination statements, which are full of puffery. Sdrqaz (talk) 20:19, 7 September 2021 (UTC)
 * 2) I've seen some terrible questions, but I am not sure what we could do to stop them. I support the two question limit, and actions by existing admins to push back on bad questions. I want to add some attention to Nosebagbear's point that this doesn't seem to be a frequent complaint of those who've gone through the process. Firefangledfeathers (talk) 03:33, 8 September 2021 (UTC)
 * 3) KevinL ( aka L235 · t · c) 18:45, 8 September 2021 (UTC)
 * 4) Questions are a key vetting tool. The issue isn't that there are too many, its about what the questions are, and that a single question can make or break you. CaptainEek  Edits Ho Cap'n!⚓ 20:47, 8 September 2021 (UTC)
 * 5) I agree that the number of questions is not the issue (since the limit was imposed). Their content is often the issue. --Izno (talk) 18:06, 14 September 2021 (UTC)
 * 6) Perhaps there are too many questions, but IMO with the 2 question limit this usually isn't an issue. I would agree that the content of said questions is often the issue at hand. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 13:27, 27 September 2021 (UTC)

K. Comments

 * How do these questions help anyone decide whether a candidate should get the mop? Most are silly: What possible answer to "Why do you think people enjoy editing Wikipedia?" would tip the balance for someone undecided about whether to support or oppose? Some are just pointy/rhetorical: what candidate is going to answer "Would you ever block an admin, when necessary, and would your process for doing so be the same process as blocking a non-admin? If not, what would you do differently?" with, "Oh, no, I'd never block an admin, and if I did, I'd definitely use a different process than with non-admins. I'd try talking to them first. Becuz, you know, I'm one of them." Two recent RfAs had 27 questions. 27! —valereee (talk) 20:50, 29 August 2021 (UTC)
 * , This to me is a lot like how supreme court justices are confirmed. We ask all these questions that no one would ever sincerely answer anyway, and all it ends up doing is testing the candidate's ability to correctly mirror what we want them to say.— Shibboleth ink  (♔ ♕) 21:48, 29 August 2021 (UTC)
 * I agree with Shibbolethink that such questions are shibboleths (heh); some are intended to be a crude "select-out" with a right answer to test a candidate on "objective" measures. At the very least they're not quite at the level of the infamous recall question. – John M Wolfson (talk • contribs) 23:18, 30 August 2021 (UTC)
 * — Shibboleth ink  (♔ ♕) 15:53, 5 September 2021 (UTC)
 * wrote above, For a start, given that we have a de-facto requirement on the candidate not to speak during an RfA, the questions are the only way for them to voice. This custom/expectation seems a bit peculiar to me, and not really in line with how we handle discussions elsewhere. For example, we invite article creators to participate at AfD, and there's no guideline that restricts them to leaving one !vote and then standing aside. Instead, they can participate just like anybody else, until they become obnoxious, same as anybody else. (You definitely don't need to be the original creator of an article to get warned about bludgeoning or pointed to WP:DROPTHESTICK, the first law of holes, etc.) I don't have a grand point in mind; it just seems unusual, like a quasi-parliamentary procedure settled upon and never reevaluated. XOR&#39;easter (talk) 01:45, 31 August 2021 (UTC)
 * It's like we have this myth of a George Washington-type candidate who is only running because they've been asked to (and to serve) despite not really wanting to. If the candidate actively (and publicly) goes out of their way to advance the process on their own accord (ie participating in the actual discussion), it's a deviation from that illusion.
 * To continue the earlier comparison, in American politics is was historically seen as improper for a presidential candidate to actually personally campaign for the office (a tradition started by Washington). People didn't like it when someone actually acted like they wanted the job; the logic was they were supposed to be above petty things like campaigning. It wasn't until 1932 with FDR that a candidate even publicly accepted a nomination for president. &#8211;<span style="font-family:CG Times, times"> MJL &thinsp;‐Talk‐☖ 03:11, 31 August 2021 (UTC)
 * To be fair, that tradition extends across the Anglosphere; the Speaker of the House of Commons has to be nominally dragged to the chair. I think it might even have Roman roots with Consuls et al.. The idea is that, like adminship, political offices are supposed to be chores/mops, and any positive effects/prestige are the result of grift and/or attract the wrong type of people. – John M Wolfson (talk • contribs) 16:33, 31 August 2021 (UTC)
 * I think the standard explanation for the Speaker tradition was that it had a relatively high mortality rate, as it involved standing up for the House's supremacy against the monarch. Most people taking up the post (at least in the early days) were genuinely unwilling to take it rather than this feigned modesty. Sdrqaz (talk) 20:24, 7 September 2021 (UTC)
 * That is indeed a factor, but I think its continuance, and the "aw shucks" tradition in general, is based on the observation/meme that the people who actually want power (which is what adminship, de facto, is) the most are the people who should have it the least. That's also why people with "I want to be an admin someday" userboxes rarely ever become admins. – John M Wolfson (talk • contribs) 20:30, 7 September 2021 (UTC)
 * I'm inclined to agree with you (for clarity, my sentence was still on that Speaker tangent). There will always be opposes based on self-nominations or because they were too eager. There should really be a disclaimer that accompanies that userbox, given the side-glances its use gets. An interesting thing I heard from a PERM admin was that they generally tolerated hat-collectors and WP:MMORPG editors because their work in getting to the "grand prize" of adminship meant that lots of work got done. Sdrqaz (talk) 21:55, 7 September 2021 (UTC)
 * Don't think that's true at all. Plenty of admins probably embodied the philosophy of the "I want to be an admin someday" userbox, they were just smarter in their approach and knew it's a bad idea to display it. Same as GN's comment on the talk page just now about potential candidates not wanting to comment in this RfC for the sake of their future RfA.Yours is an argument that anyone ambitious is unsuitable for positions of power, except in the real world (which is the analogy being used here) almost everyone in a position of power has/had high ambition. Nobody falls into the role of POTUS without actually wanting to be POTUS, for example. ProcrastinatingReader (talk) 21:57, 8 September 2021 (UTC)
 * I'm aware that most people actually having power tend to the be the very ambitious, as is anyone not living under a rock. What I'm saying is that the Anglo/Western tradition views that "power corrupts", meaning that the people who really want it without really caring about the "servant" part of "civil servant" are exactly the people who shouldn't have it, but often unfortunately do. It's a very Whiggish way of looking at things, and to what extent this is actually true depends on your politics and who in fact rules; it's really easy to make the case in America and Britain, at the very least. As such, it's considered unsightly to actively say "I want to be the President because I want prestige" (not even Trump expressly said that), as well as earlier taboos described above. Although RfA is far less important, people who want the mop because of its aura and don't know what it really entails, as evidenced by the userbox, are precisely the people who we don't want. In an earlier account I had as a child/young teenager I displayed the userbox, but I had wisened up by the time I created this account in 2018 and, without trying for it in my first year, eventually got the mop. – John M Wolfson (talk • contribs) 22:13, 8 September 2021 (UTC)
 * I don't think that's the case, and, if it is, that it's a bad thing. I think the tradition is that candidates are discouraged only from directly rebutting opposes, the thinking being that any truly-bad oppose can be dismissed by other !voters and that good/acceptable opposes shouldn't be badgered. Of course, directly thanking supports comes across as needy and slightly pathetic, so I think that means that candidates can still participate in general discussion. If it is the case that candidates shouldn't go out of the question area, I think that's actually a bit of cool purdah that creates a sense of solemnity, so I don't think it needs any reevaluation. – John M Wolfson (talk • contribs) 16:33, 31 August 2021 (UTC)
 * absolutely, in the sense that it was never formally agreed, and just "is the way things are". However, on occasions where candidates have started speaking more, it's almost always gone badly - partly because the Community doesn't like it, but also because it makes already acrimonious areas become outright fireballs. When we get to phase 2, perhaps propose we should have a standardised 4th question that is left blank to start but can be answered at any time - "what question would you like to have been asked? Answer it" Nosebagbear (talk) 13:28, 1 September 2021 (UTC)
 * If it goes badly because the Community doesn't like it, then we have an ossified social norm that (a) doesn't match up with how we handle discussions pretty much anywhere else, and (b) doesn't really tell us whether the candidate would be a good administrator or not. If it goes badly because the candidate fans a flamewar, then we've learned something about their temperament under pressure, which I'd say is important to know. If it goes badly because other people escalate the fireball, then we get to see the candidate keep their head whilst others are losing theirs, which is also a helpful observation. Any way I slice it, it's a strange custom, and even if it's turned out to be a good idea in practice, it contributes to the insularity of the process, adding a little to the "activation energy" barrier that inhibits participation. It's one more factor in making RfA this big, built-up thing which nobody undertakes without serious encouragement. I guess when I put it that way, I'm saying that RfA is a microcosm of Wikipedia in general.... XOR&#39;easter (talk) 18:28, 2 September 2021 (UTC)
 * well in a big sense, RfA is not like any other discussion. It's not akin to, say, the proposer of an RfC doing lots of talking. It would be more akin to the proposal doing lots of talking. While ACE is split across lots of individuals, an RfA is functionally unique in being a "let's discuss this editor, generally" - it's not even like the specific example(s) of behaviour within an ANI or ARBCOM thread in that sense. I still make no note as to whether it should be like this, or not, but that it being different to everywhere else is not, fundamentally, an indication of flawed status quo. I do note that whatever the outcomes of Phase 1 and Phase 2, this process can't just change the mindset of the Community by a close. Hence working within it to mitigate the issues in execution, as I mention above. I absolutely do agree that it does add one more component to the difficulty in getting candidates to run in the first place. Nosebagbear (talk) 18:52, 2 September 2021 (UTC)
 * As Nosebagbear states, it's a peculiar custom because RfA and RfB are themselves peculiar. They're the only discussions where we judge people, not articles or proposals (ArbCom elections are more straight votes, and not discussions in the same way) . Think of it like American Idol; it's unseemly for the contestant to talk back to Simon Cowell/Louie/Paula Abdul/etc., and it is likewise unseemly for the candidate to "talk back" to !voters outside of questionspace. I do find that the "what question would you like to be/have been asked" question might alleviate some stress, so long as it's strictly optional. RfA is indeed supposed to be a bit stressful/solemn as a WikiRiteOfPassage, even if as the discussions show it's too much. – John M Wolfson (talk • contribs) 19:02, 2 September 2021 (UTC)
 * This might point to a fundamental issue. I've seen too many bad processes justified on "rite of passage" grounds to trust the idea. It's hard for me, personally, to find the wisdom of thinking about a process as a "rite of passage" when the point is to decide whether or not someone is suitable for a volunteer position. It's basically like saying, "We need a hazing ritual before we will contemplate trusting you." And if that's the mindset "the Community" is generally locked into, then RfA will never change, and the applicant pool will stay small, as people who could do the work smell the hazing ritual and stay the hell away. XOR&#39;easter (talk) 20:23, 2 September 2021 (UTC)
 * I disagree with the premise that candidates are expected to only respond to questions. Generally speaking, I believe the community wants to see that the candidate is receptive to feedback, and willing to explain their actions and reasoning without becoming overly defensive. It's commonplace for admins to need to respond to upset editors in a way that affirms their commitment to collaboration and to correcting missteps as needed. Candidates need to be able to do the same during a request for administrative privileges, and figure out the best way to respond to feedback positively. isaacl (talk) 02:50, 3 September 2021 (UTC)
 * A question should generally be about the candidate's suitability as an admin and regarding any aspect of editing on Wikipedia. I don't think that the issue is with too many questions because there is already a 2 question limit per editor which can be asked at an RfA (previously there was none). The issue is with frivolous, unimportant and rhetorical questions that are sometimes asked by people who want to ask such questions just for the sake of asking. We need a solution for that. TheGeneralUser (talk) 22:41, 29 September 2021 (UTC)

L. Who participates
The kind of editors who participate in RfA are ill-suited for the task.

L. Support (agree)

 * 1) The kind of people who participate in Wikipedia are ill-suited for the task. I would know, I'm one of them. :-) I would say this: it's important that we honestly assess our own limitations when it comes to vetting editors for adminship, and try to create a system that compensates for those limitations. Levivich 07:27, 29 August 2021 (UTC)
 * 2) As a non-admin, I do not have the full picture of what admins do, and therefore do not have a full idea of who should be an admin. The selection of admins should not be an open community process. Instead, it should be given to a trusted group of editors who have experience as admins and have to go through community scrutiny, such as bureaucrats and WP:ARBCOM members. Z1720 (talk) 14:57, 30 August 2021 (UTC)
 * 3) RfA is a thunderdome, and none of the arguing helps anyone. Least of all the candidate that has to watch all of the arguing around and about them in silence. – ♠Vami _IV†♠  03:59, 4 September 2021 (UTC)
 * 4) For permission to vote, I think voters should be subject to the same criteria they demand of the candidates. So fix a set if crieria for adminship, and hold the voters to it. Kudpung กุดผึ้ง (talk) 14:32, 19 September 2021 (UTC)
 * 5) Completely agree with this. Get that bloody stupid thing off the watchlist. – filelakeshoe (t / c) 🐱 08:54, 22 September 2021 (UTC)
 * 6) I like the fact it's open to the whole community to participate in - but the people who actually do participate are generally not representative of the community as a whole. I'm not sure how to remedy this though - it's not as if we can put a random/representative subsection of the community on AfC duty like a kind of jury service, or how we would go about working out what really is "representative" among a community of anonymous internet users. So this is a low priority for me - we should focus on problems we can solve. <b style="color:#98F">W</b><b style="color:#97E">a</b><b style="color:#86D">g</b><b style="color:#75C">ge</b><b  style="color:#83C">r</b><b  style="color:#728">s</b><small  style="color:#080">TALK  14:56, 27 September 2021 (UTC)

L. Oppose (disagree)

 * 1) This is a somewhat odd topic to me. I mean, writing an encyclopedia is a task to which the vast majority of us, bluntly, would be considered ill-suited. Thus too, admin review. Admins receiving Community review is fundamental, so the question should be less "RfA (!)voters are good/bad" and more "Are they a good cross-section of the community". I mean, I'd be fine with some level of protection, but I also don't think that more than a few % in any given RfA are genuinely problematic reviewers. That does mean, that in some of the problematic cases, editors I'd (and I believe, we'd) consider reasonable participators are being dragged into the morass. Nosebagbear (talk) 11:36, 30 August 2021 (UTC)
 * 2) Per Nosebagbear. You go to RfA with the !voters you have, not the ones you want. – John M Wolfson (talk • contribs) 23:15, 30 August 2021 (UTC)
 * 3) I think a pretty broad section of the community participates in RfA !voting. Occasionally there can be issues with bad opposes from editors who have had some disagreement with the candidate, but that's a somewhat separate issue. &#123;{u&#124; Sdkb  }&#125;  talk 22:24, 31 August 2021 (UTC)
 * 4) Strong oppose.  I can't imagine who would be better suited to conducting this process than a broad spectrum of active editors, including both admins and non-admins. User:力 (power~enwiki,  π,  ν ) 17:39, 1 September 2021 (UTC)
 * No, I don't think that's the problem. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 10:36, 2 September 2021 (UTC)
 * 1) While I think that more voters should participate in the process, I don't think there's an inherent problem with the people who participate now. I also agree with the comments of Nosebagbear. Trainsandotherthings (talk) 23:55, 2 September 2021 (UTC)
 * 2) Many RfAs receive over a hundred participants. That's a pretty reasonable cross section of the community, and many are experienced at reviewing candidates. Regardless, we must have the community's approval for those who will be admins in it; no admin would be legitimate absent that. Seraphimblade Talk to me 23:59, 3 September 2021 (UTC)
 * 3) The sort of editor who participates at RfA is the sort of editor who is interested in community governance, frankly a boring and pretty niche backroom topic that has no visible impact whatsoever on the vast majority of Wikipedia users. What would be better, a mix of editors somehow forced to participate who don't know the policies and also don't care? Ivanvector (Talk/Edits) 19:09, 4 September 2021 (UTC)
 * 4) Everyone who wants to !vote should be allowed to. (Otherwise what's next? Request for Votership?) <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  23:47, 6 September 2021 (UTC)
 * 5) Cas Liber (talk · contribs) 12:20, 7 September 2021 (UTC)
 * 6) While I wouldn't be averse to a very mild restriction in participation (e.g., restricting to editors who meet the "extended-confirmed" threshold), I don't really think this is all that significant an issue. Risker (talk) 01:15, 8 September 2021 (UTC)
 * 7) We should encourage as much participation in RfA as possible and continue to advertise on watchlists. I do support heavier moderation/clerking of RfAs. Firefangledfeathers (talk) 03:42, 8 September 2021 (UTC)
 * 8) The answer here is basically "democracy is the worst form of government except all those other forms that have been tried". Unless the community itself has direct control over selecting admins, all sorts of other issues, worse issues, will present themselves, especially contempt for a "dictatorial" admin corps. The current RfA is on track to be one of the highest-attended in Wikipedia history, and that's a good thing. I agree with Risker that some sort of limited suffrage requirement (whether that's extended-confirmed or even auto-confirmed) might be useful, but I'm under no illusion that it would alleviate any of RfA's problems. Extraordinary Writ (talk) 03:58, 8 September 2021 (UTC)
 * 9) Per Risker. KevinL ( aka L235 · t · c) 18:45, 8 September 2021 (UTC)
 * 10) I think making it less democratic is unlikely to help, which is what would happen if we removed certain editors from voting as a fix to this statement. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 13:29, 27 September 2021 (UTC)
 * 11) Editing and building an encyclopedia needs an open and collaborative environment and so does many other areas of Wikipedia (including RfA). Except for obvious trolls and LTA's, every user should be free to voice their opinion, cast their vote and participate in an RfA as they see fit. It would be unfair, impractical and a partiality to exclude anyone from RfA. TheGeneralUser (talk) 21:16, 29 September 2021 (UTC)

L. Comments

 * I think this problem is at least partially caused by advertising on watchlists. It's not that I don't find it a handy notification myself, but over the course of a week I'd see an announcement at Centralized Discussions. The downsides seem greater than the upsides. —valereee (talk) 20:29, 29 August 2021 (UTC)
 * I've seen this be raised as a potential problem elsewhere, but I don't agree personally. At the end of the day, RfA (as it currently exists) is an election about 'community' trust. This means it cannot be dominated or gatekeeped by a clique of RfA regulars, even if that would lead to better results (however that could be measured). Still, IIRC I read that watchlist advertising only began a few years ago, but RfA's output has been more or less the same since 2012, and people have been complaining about RfA since 2006 and maybe even earlier. So I don't think wider advertisement is a cause of problems, since they already existed before. ProcrastinatingReader (talk) 23:10, 29 August 2021 (UTC)
 * Yeah, we tried not advertising RfA widely, and it didn't work. 2015_administrator_election_reform/Phase_I/RfC AntiCompositeNumber (talk) 02:29, 30 August 2021 (UTC)
 * Is this question about the people who comment in RfAs, or run in them, or both? — Bilorv ( talk ) 17:31, 31 August 2021 (UTC)
 * I find this an odd question to answer because I don't think you can generalize about who participates in an RfA. There are the regulars who you see at every RfA but the majority of people voting in an RfA are editors I've never encountered in my 8 years of regularly editing the project. I'm not sure that there is a "type". I do think it's important that there is a minimum level of experience that is needed to vote...I can't recall what it is but I think it is a pretty low bar. It's not unheard of for outside forums to publicize RfAs so I think it's important that brand new accounts & IP accounts are not eligible to participate. I don't think I'm paranoid as there have been attempts to sabotage RfAs in the past and that should be discouraged. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 02:53, 4 September 2021 (UTC)

M. The RfA process is biased against long-term editors
The problem is that mud sticks. It is much harder for a veteran contributor to become an admin, compared to a fresh editor with one or two years of history, since usually the latter has a much more pristine record than the former.

It's hard to become an admin with just a year or so of history, but if you have too many (active) years you'll very likely fail too as the odds are that over that time you've made enough mistakes that people will be "wary" around you (this is particularly true for the unlucky individuals who had the misfortune of getting involved in some controversial areas, content or otherwise (userboxes, anyone?)). As time goes by, every active editor makes some mistakes and, for the lack of a better word, enemies. So as time goes, more and more active editors will be less likely to meet our standards, as there'll be some 'dirt' on their record, dirt that will be brought up at RfA, poisoning the well, creating guilt by association, and encouraging "better safe than sorry" problem avoidance ("why risk giving the oh-so-powerful tools and status symbol to editor A who makes a mistake every three years and has several blemishes on their record? It looks safer to censure them and deny them the tools and instead support this new promising editor B who has been here for one and a half years and hasn't made any mistakes yet, surely they have a promising future!" [despite the record of editor A from their first year and a half being identical...]). I think this can be proven by a correlation of how many years/edits one needs to become an admin, which I hypothesize to be roughly bell-shaped (something like this; of course, outliers happen). Likewise, I have major doubts that many of the admins with long tenure would pass a new RfA as easily as they did when they were still "younglings", or if they were required to go through it again would opt to do so (this could be measured by looking at the success rate of such re-applications).

A rider on here is the issue of community demographics. We seem to be getting fewer new editors than in the early years of the project, and as we are not enough for die-off of veterans to occur, our community is getting 'older' (each year, the edit count and 'age' of an average Wikipedian is increasing). So the ratio of eligible candidates is decreasing compared to the early years of the projects (as there are fewer 'new promising candidates', while most 'veterans' are seen less and less 'promising' each year, since as time goes they are steadily saddled by some past mistakes).

A partial solution would be to limit critique of editors to their last few years, declaring that any mistakes, blocks, whatever that occurred x years in the past are irrelevant, trying to level up the playing field, and encourage veteran editors to apply. That said, it's an imperfect solution (well, nothing will be perfect anyway), and further, I am not sure how to properly apply it (hiding the older block log entries from public view is something to consider, for example). Even if we say that only the last 5 years or whatever matters for the review, it won't prevent people from remembering about the past and talking about it here and there. Once the well is poisoned, it's very hard to clean it up (outside WP:VANISHing, and I bet there are some folks out there who did just that and became admins after discarding their old accounts - and if they do a good job; it's a win-win, I say).

M. Support (agree)

 * 1) It's not biased in the RfA you see on the page, but in the declined RfAs that never come into existence. Longer-term editors are more likely to decline RfA when approached because they have several times the number of embarrassing moments that could get dragged up in RfA. (At least, I can only extrapolate this from people who are open about not wanting to run and why, as I'm of course not privvy to private conversations that don't involve me.) This relates to what I said under F about scrutiny being random and disproportionate. It's not really so simple as bias against longer-term editors and towards shorter-term ones, but bias against some types of longer-term editors that don't exist towards newer ones. — Bilorv ( talk ) 17:36, 31 August 2021 (UTC)
 * 2) Largely per above.  -- Calidum  18:55, 31 August 2021 (UTC)
 * 3) I think Bilorv is correct - it could be argued either way in terms of those who progress to being candidates in the last 4/5 years, but it's definitely holding back multiple good candidates on this basis. In some of those cases I think that concern is not warranted, in others, it may be - and that's just the ones I'm aware of! Nosebagbear (talk) 21:17, 1 September 2021 (UTC)
 * 4) Such is survivorship bias, unfortunately.  – John M Wolfson (talk • contribs) 21:53, 1 September 2021 (UTC)
 * 5) The community as a whole, and specifically the people who routinely vote in RfA, should be more willing to forgive users for mistakes they made long ago. There's a reason statutes of limitations exist in the off-wiki world, perhaps we need something similar on-wiki as well? Forgive the imperfect metaphor, but when a university looks at an applicant's high school transcript, they tend to put more weight on their more recent years (junior and what they have for senior) as opposed to sophomore and especially freshman - but in RfA, it feels like a mistake you made in grade school will be viewed as little different than one you made two weeks before you applied for college. Trainsandotherthings (talk) 00:01, 3 September 2021 (UTC)
 * 6) I agree that RfA is biased against a large contingent of well qualified long-term editors whose greatest trespass was editing with courage, conviction, and clue.--John Cline (talk) 02:38, 5 September 2021 (UTC)
 * 7) I think RFA should be more forgiving and understanding in general. RFA often feels like a "holier-than-thou" competition, which is hypocritical. I lack a clear idea of how to fix this beyond cultural changes though. Wikipedia and the world at large are increasingly critical of the smallest mistake, which I think is really counterproductive. We are not gods, we are people, and thus we make mistakes. CaptainEek  Edits Ho Cap'n!⚓ 20:45, 8 September 2021 (UTC)
 * 8) Yip, it is biased. I have seen many suitable candidates put off by being shown examples of their work, comments they have made and so on, often in the deep past, that often didn't reflect their behaviour now. They grew and evolved. If they're was two comments at different times where the candidate e.g. lost their temper or called somebody a name, then the chances of being even nominated was reduced to nill.    scope_creep Talk  10:33, 9 September 2021 (UTC)
 * 9) Strong support. Too many people are interested in the "perfect" candidate. Our national elections clearly show us that even with huge stakes on the line, there is no such animal. Buffs (talk) 20:08, 22 September 2021 (UTC)

M. Oppose (disagree)

 * 1) Long term editors only have a problem at RFA if they have a longterm problem that hasn't been addressed. There is a myth that people go back and dredge up ancient grievances, but in the rare occasion where people do it does not get traction in the oppose section unless they can demonstrate that it is still a current concern. That might be true of character issues, it won't be of skill issues.  Ϣere  Spiel  Chequers  12:32, 30 August 2021 (UTC)
 * 2) Per WSC, I think opposes based on issues from multiple years ago won't sink an RFA unless the issue is, as they say, "still a current concern" Eddie891 Talk Work 13:24, 30 August 2021 (UTC)
 * 3) Disagree, it is much harder for a new editor to pass an RfA. True, a longtime editor can have more baggage but they also have the benefit of observing and participating in lots of RfAs and have more realistic expectations of the process than a less experienced editor. Do we really have a big problem with editors who've been active for 10 or 15 years being afraid of putting themselves through an RfA? I would think they were interested, they would have tried it during their first 5 years of editing. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 02:59, 4 September 2021 (UTC)
 * 4) I haven't noticed this as a problem. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  23:51, 6 September 2021 (UTC)
 * 5) No. Cas Liber (talk · contribs) 12:21, 7 September 2021 (UTC)
 * 6) Potentially, but not in an actionable sense. In practice, mud from several years ago doesn't stick that much; my RfA was a good (if embarrassing) example of that. I had a lapse of judgment when I was a young teenager; some opposed on that basis, but only a couple. Best, KevinL ( aka L235 · t · c) 18:47, 8 September 2021 (UTC)
 * 7) --Izno (talk) 17:23, 14 September 2021 (UTC)
 * 8) Partly Per : Long term editors only have a problem at RFA if they have a longterm problem that hasn't been addressed, but there are editors who have been absent for years, then come back for a month and think they can get the bit based on 12,000 edits they did 12 years ago. This is a very frequent issue. They don't run, because they've been through WP:ORCP and told how wrong they are. Kudpung กุดผึ้ง (talk) 14:40, 19 September 2021 (UTC)

M. Comments

 * Potentially, but I would like to see the data. Specifically, months of active editing per admin before successfully passing an RfA. ProcrastinatingReader (talk) 11:27, 30 August 2021 (UTC)
 * This is a terrible estimator, a detailed analysis of months of active editing as PR says would be much more useful, but FWIW unless I'm making some silly error: 2020-2021 successful candidates on average made their first edit in 2013. Unsuccessful candidates (removing 3 SNOWish candidates) on average made their first edit in 2012. —valereee (talk) 12:59, 30 August 2021 (UTC)
 * I don’t think the issue is sinking a RFA, but having to go through the process of having those edits brought up. It doesn’t help that there is no real way to defend yourself in a RFA, without sinking it for badgering (maybe there needs to be an accepted way for candidates to respond to !votes like a dedicated section). Air<b style="color: green;">corn</b> (talk) 18:41, 2 September 2021 (UTC)
 * The more sympathetic questioners will routinely ask a question that gets to a particular objection. I don't know how I'd feel about preemptive responses. Izno (talk) 17:24, 14 September 2021 (UTC)
 * This is one truth that won't bear out by analysis of past RfAs. The reason is that these long-term editors, who know of this "bias", would never put themselves through RfA for the hell that they know it would be.--John Cline (talk) 01:46, 5 September 2021 (UTC)
 * I agree with . I will note, however, that perception (on the part of experienced editors) may not be borne out by evidence. I'd like to see some data here. Risker (talk) 01:13, 8 September 2021 (UTC)
 * I'd like to see some data too, but my perception is there's a pretty even tradeoff between opposition based on relative lack of experience for relatively new editors, and opposition based on old disagreements for old editors. Vanamonde (Talk) 12:45, 8 September 2021 (UTC)
 * Agreed, I too would like to see a proper data on this before proposals and solutions can be formed and planned. TheGeneralUser (talk) 23:07, 29 September 2021 (UTC)

N. Support (agree)

 * 1) Over the years I have come to realise this is the cause of two of RFAs biggest problems. Much of the rancour within RFAs is over what the criteria should be rather than whether the candidate meets the criteria, and much of the difficulty in recruiting people to run is that they aren't sure if they meet the unwritten criteria (hence the few who run usually run a year or more after they could have passed). Some of the criteria could be agreed by consensus, and that would detoxify RFA and focus it on the bits where we do want to maintain judgement.  Ϣere  Spiel  Chequers  12:45, 30 August 2021 (UTC)
 * 2) The unwritten evolution of criteria is one of the difficulties of RfA. Do candidates need FAs/edit summaries/copyright experience? If different criteria go in and out fashion, it should be possible to agree on these by consensus. Some of the most important criteria (good judgment) may prove harder to evaluate than others (editcount, experience in admin related areas). —Kusma (talk) 13:21, 30 August 2021 (UTC)
 * 3) Agree lack of consensus on what qualifies someone to be an admin is a problem. We have WP:FACRIT, WP:GACRIT, WP:ITNCRIT and many others but no WP:RFACRIT. (Also, lifetime appointments make some editors criteria very strict, but that's another question.) We would benefit if we were all judging candidates against some standard criteria rather than everyone coming up with their own. Levivich 13:32, 30 August 2021 (UTC)
 * Yes, but like some others, I don't think it's an area where there's much to be done. We could create clearer criteria, but because so many people approach RfA with their own ideas of "what an admin must be," there wouldn't be consensus for anything beyond broad guidelines that admit personal judgment in some form anyway. &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 18:06, 1 September 2021 (UTC)
 * 1) Actually, I think there might be something in this. We don't have "set criteria" and I think that the whole process could work a lot better if there was. In my head, the top question is "Is this person suitable for adminship?". I'd be really interested in an RfC on "community criteria" for what qualifies people to positively answer that question. We have a lot of "personal criteria" essays all over the place, which could be the basis for the RfC. It would give people a starting point in discussions, and could help bureaucrats discount votes that go way outside community norms. Equally, it need not take away from the general issues people may have. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 10:45, 2 September 2021 (UTC)
 * 2) I agree with the issue as stated. With that said, I do not think it is reasonable or possible to force editors to follow an objective criteria, because if we do so, why even have a process at all as opposed to just granting sysop to anyone who asks and meets the criteria? I would like to see some guidelines, with the caveat that while they represent a community consensus, they are not binding. Trainsandotherthings (talk) 00:04, 3 September 2021 (UTC)
 * 3) Absolutely yes. The only official criteria is "can pass RfA", but each person's RfA ends up having different criteria depending on who shows up. That's a problem for RfA but it's a bigger problem for desysopping, because if we haven't agreed on standard criteria for adminship then we have no reference point for when an admin has fallen short of those standards, and this also has follow-on effects for the RfA scrutiny question on this page. Ivanvector (Talk/Edits) 19:27, 4 September 2021 (UTC)
 * 4) Yes this. Stifle (talk) 12:17, 6 September 2021 (UTC)
 * 5) It would be handy to have shortcut-linked community consensus against certain common but bad support/oppose arguments, a la WP:Arguments to avoid in deletion discussions. Even if not enforced as rules, repeated reference to the norms can help with culture change. Firefangledfeathers (talk) 03:57, 8 September 2021 (UTC)
 * 6) I would be very interested to see what Isaacl's proposal would lead to regarding  requirements rather than the hard requirements one might associate with, of all things, autopatrolled permission. --Izno (talk) 17:21, 14 September 2021 (UTC)
 * 7) Absolutely. Kudpung กุดผึ้ง (talk) 14:42, 19 September 2021 (UTC)

N. Oppose (disagree)

 * 1) I don't think this is so much a problem as others are making out. It's considered poor form on here to obsess over edit counts of adminabili, and I think it would be likewise stupid to make any objective/numerical criteria as strict as those on RFPERM or even much beyond the current "formal" criteria; this is especially true if RfA remains a discussion and doesn't become more a formal vote as was proposed and defeated back in 2019. It's best to just get a feel for the place and a good nom.  – John M Wolfson (talk • contribs) 15:04, 30 August 2021 (UTC)
 * 2) I don't believe having a set criteria would alleviate the issues with the RfA process. The tools and responsibilities a sysop has are many and varied, so the set of criterias would be either too broad to cover all the aspects, or not broad enough. Even for other more straigthfoward permissions (or processes, such as GAN), the criterias are sometimes the absolute minimum an editor or article should meet, and they still need to be vetoed by an acting administrator or editor in good standing, who has to comb through their contributions or history to see if there are any pending issues to be solved. Isabelle 🔔 16:52, 30 August 2021 (UTC)
 * 3) While i think some people have truly insane personal criteria, codifying standard criteria is a terrible idea. I wouldn't mind seeing "spam every RFA with the exact same question in an effort to make it de facto standard question" banned, (i.e. the recall questions that have caused much drama for no real return) but that'd probably be a tough thing to get consensus for. Beeblebrox (talk) 21:59, 30 August 2021 (UTC)
 * 4) Each editor is free, and should be free, to make up their own requirements for adminship. Whether voters are working off a detailed and specific checklist (even if that checklist is insane), or just feel good about a particular candidate, is not relevant in a democratic system. Furthermore, we should have a secret ballot to avoid dog-pilling on editors who oppose for reasons that other editors disagree with, which contributes to the toxic atmosphere at RfA and discourages opposes. — Goszei (talk)  23:15, 30 August 2021 (UTC)
 * 5) If the purpose of RfA is to determine whether the community trusts a candidate, it's necessary to give the community a pretty wide berth in expressing its views. Unless editors are using completely unreasonable criteria (in which case the crats will quite appropriately give their !votes little weight), they should be allowed to explain frankly why they don't trust a given candidate, even if their reasoning seems unpersuasive to the majority. Taking away that right would likely result in contrived reasons to oppose, and regardless wouldn't improve the RfA atmosphere: the most controversial RfAs don't involve disputes about the criteria themselves, but instead disputes about whether the editor meets various uncontroversial criteria (e.g. civility). Extraordinary Writ (talk) 17:34, 2 September 2021 (UTC)
 * 6) I'm mixed on this, there is no agreed upon criteria but I'm not sure if that is a big problem. Candidates vary, greatly, in their editing background. I think that is a benefit as we have admins with different talents. I know when I have a technical problem, there are admins I can go to for help. I wouldn't expect them to be writing Featured Articles. If there were accepted criteria, they would have to be so general and unspecific as to not be particularly useful. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 03:04, 4 September 2021 (UTC)
 * 7) It's okay for !voters to have different criteria. (Bureaucrats should filter out the truly irrational !votes anyway.) Moreover, it would be impossible for the RfA crowd to hammer out a set of criteria that would have consensus approval. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  23:56, 6 September 2021 (UTC)
 * 8) Surprised to find myself disagreeing with WSC, WTT, and IvanVector, but so it goes...I really don't think this is a problem. I agree with WSC that a considerable amount of the argument at any given RFA has to do with what the criteria should be, and not about whether the candidate meets them; this is particularly evident in the unpleasantness that habitually follows the handful of opposition, based on very stringent criteria, that is expressed at RFAs for candidates that patently has the community's support. However, I don't think this is a concern with respect to RFAs that could have passed but didn't, or qualified candidates who chose not to run. Close RFA are frequently close because one or two incidents are seen as evidence of not passing a given criterion by a good fraction of the community, and not by the rest. And such RFAs, I think, are far more discouraging to potential candidates than ones that pass with a token opposition based on stringent criteria. Vanamonde (Talk) 12:36, 8 September 2021 (UTC)
 * 9) Per . KevinL ( aka L235 · t · c) 18:48, 8 September 2021 (UTC)
 * 10) Nah. Its good for different folks to have different criteria. That way the criteria can change to meet the needs. It also means that folks who run should be well acquainted with the voterbase by reading past RfA's and know what will be expected. I found it very personally enlightening to know that some voters cared about XfD, that a few folks felt an FA was mandatory, that some folks were mostly focused on CLUE. Even if we created a set criteria, voters would still subconsciously use their own criteria. People remain people. CaptainEek  Edits Ho Cap'n!⚓ 20:39, 8 September 2021 (UTC)
 * 11) I'm not sure that this is a major part of the problem. At the end of the day, RfA is a way to gain the consensus of the community in regards to whether the community trusts a editor to become an administrator. By setting arbitrary criteria this is likely to cause more issues, as people who don't meet the criteria may be fully trusted but those who might meet it could still have some issue which leads to no trust. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 13:31, 27 September 2021 (UTC)
 * 12) Having any hard fixed criteria will just be arbitrary and will become a numbers game. Even if the criteria is evolving (through changing community consensus), it will just reinforce the already high unwritten standards and expectations most people have of RfA candidates. Trust, credibility, good judgement and ability to learn from one's mistakes are more important things than having X number of edits, AfD votes, X number of created articles, etc. TheGeneralUser (talk) 22:50, 29 September 2021 (UTC)

N. Comments

 * In the 2015 review of the requests for administrative privileges process, I made a proposal to focus the discussion on the candidate's pros and cons, and to establish the relative importance between them. In a comment I wrote: Regarding implementation concerns: from the participants' perspective, basically the same discussion would take place as today, but with less repetition, since the threads would be organized by topic, and not in response to votes. The bureaucrats would have to organize their closure process a little differently, but one advantage is that the community's consensus view on the relative weights for a candidate's characteristics is largely reusable from one RfA to another. Thus once a few RfAs were completed, there would be a good baseline, and discussion would be expedited. In essence this would establish a common view on what characteristics the community wishes to see its administrators. I think many people think of numeric cutoffs when the idea of criteria is raised, but they can be more qualitative as well, guiding those with the desired behaviours towards signing up to do chores for the community. isaacl (talk) 15:00, 30 August 2021 (UTC)
 * To clarify, my comments on qualitative criteria were in context of this RfC, and in particular don't apply for auto-granted permissions. Note many editors have published qualitative criteria in their lists of things they look for when deciding to grant administrative privileges, so it's not a novelty. isaacl (talk) 20:43, 14 September 2021 (UTC)
 * I shortened the title of this section from "N. Unlike other userrights RFA does not have an agreed criteria and the lack of this is a major part of the problem" — Wug·a·po·des​ 20:11, 30 August 2021 (UTC)
 * It shouldn't be too difficult to draw up a proposed set of criteria then hold an RfC to gain consensus for it. Some RfC pass, some don't - it's a question of 'Nothing ventured; nothing gained'. There is a wealth of criteria to be drawn on here from regular voters who take the process seriously. It just needs researching and doing. Meeting such a set of criteria would absolutely not qualify an automatic pass at RfA. It is strictly only the basic minium to run for office without wasting the community's time and inviting trollish  votes and criteria. Kudpung กุดผึ้ง (talk) 20:29, 26 September 2021 (UTC)

O. Support (agree)

 * 1) Once an active editor has those extra buttons they are likely to use them, and that's especially true for active content creators who can then be asked to do things such as assess newish editors for rights such as autopatrolled  Ϣere  Spiel  Chequers  12:56, 30 August 2021 (UTC)
 * 2) Agree. I know the argument about security, but we've got hundreds of inactive admins out there with the rights; they certainly have no need for the tools, so that argument feels like an excuse to me. If we're worried about security, let's fix the security issue. Require 2FA for any candidate who can't "demonstrate a need for the tools" or whatever lessens that issue. I never thought I'd be doing half the admin stuff I do, but I started to see places I could help. Adminship for every trusted, experienced, clueful, and unfailingly civil user! —valereee (talk) 13:06, 30 August 2021 (UTC)
 * 3) People who need the tools (you only truly need the tools if you want to do things that you can't convince other people to do) should not be admins. Wikipedia needs people willing to use the tools to help. —Kusma (talk) 13:24, 30 August 2021 (UTC)
 * 4) No one needs the tools. It's the other way around: the community needs editors with the tools (admins). AGFing, when people talk about "needs the tools," they're using the wrong words, and what they mean is something else, like "likely to actually use the tools" or "competent with the tools." If we had a standard rfa criteria, "need" should not be part of it. Levivich 13:38, 30 August 2021 (UTC)
 * 5) I think this might be one of the main bottlenecks of the others are correct. I could see changing Q1 in some way to make it less onerous on someone. – John M Wolfson (talk • contribs) 01:08, 31 August 2021 (UTC)
 * 6) Nobody "needs" the tools, some people use them for this and that, some more than others. So what? This should never be an objection. --<sub style="border:1px solid #228B22;padding:1px;">Piotr Konieczny aka Prokonsul Piotrus&#124; reply here 03:28, 31 August 2021 (UTC)
 * 7) Most certainly. It might be true that a "need for the tools" is not a common reason for RFAs to fail, but I know from experience it's a common reason for otherwise suitable candidates to be overlooked and/or decline nominations. There are many aspects of admin work that do not require technical expertise, but where content experience is exceptionally helpful: dispute resolution in particular, but also main-page work, etc. Vanamonde (Talk) 06:05, 31 August 2021 (UTC)
 * 8) I've always thought the main consideration for adminship should be whether the editor can be trusted not to misuse the tools, and that most other criteria, questions etc should be aimed at elucidating whether the nominee can be trusted. If "need for the tools" is a significant de facto consideration, it should be mentioned at Administrators, but it's not. Nurg (talk) 05:16, 1 September 2021 (UTC)
 * 9) I don't think it's a good reason to oppose. But, if we take a stand against such rationales, it will just be a game of Whac-A-Mole. Over time, there's always been in-vogue reasons to oppose RfA. This is one of them. The core issue from my perspective for an RfA candidate is one of trust. Do we trust the candidate will use the tools appropriately when they use them? If they're not certain about how to use a tool in a given situation, will they seek assistance? If the answer to those questions is 'yes', the need for the tools is superfluous. --Hammersoft (talk) 15:30, 1 September 2021 (UTC)
 * I think we should take a stand against such rationales until we figure out the ones that we as a community agree are reasonable (or rather, have constrained the set such that we know all the ones we don't like :^). I do think the consensus is still out on this one though because of our lax interface administrator user group request process as well as the view deleted permission, which do have a legitimate security aspect to them. Izno (talk) 06:40, 14 September 2021 (UTC)
 * 1) Supporting that it's not a good reason for opposing, but I figure "no need for the tools" just omits the implied "and I have reservations about granting them to this person" so don't think there's much to be done beyond acknowledging it's a bad reason for opposing. &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 18:04, 1 September 2021 (UTC)
 * 2) per above comments on "set criteria" - this would be the sort of thing I'd like to see the community agree / disagree is a valid concern. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 10:46, 2 September 2021 (UTC)
 * 3) Per Valereee. I believe there is a large group of trusted content creators who would use admin rights only occasionally, but who may discover they enjoy admin work or be willing to help out with more urgent backlogs once in a while. The expectation to use admin tools often may form a barrier for potential candidates to put themselves forward; it is for me.  FemkeMilene (talk) 16:24, 2 September 2021 (UTC)
 * 4) "No need for the tools" is a terrible argument in my book. Admins do not necessarily use their tools constantly or even daily. Admin tools are varied, and some admins may never use some of them. Being an admin also means that the community has placed a level of trust in you, which can be an asset in and of itself. And there will always be tasks needing admin attention, I find it incredibly hard to believe that too many admins will mean some of them get bored or never use their tools. Trainsandotherthings (talk) 00:23, 3 September 2021 (UTC)
 * 5) We don't have a limit to the number of admins we can have, so people who can be trusted with the tools should have them. Even if they only use them occasionally, that's one less thing someone else has to deal with. The question is "Can this editor be trusted with the tools?", not "Will they use them constantly?". Seraphimblade Talk to me 23:53, 3 September 2021 (UTC)
 * 6) There are admins like me who rarely actually use the tools unless a situation arises where using the tools helps solve it. If a user has our trust, give them the tools that let them use their judgement when to use them. As long as they remember to read guidelines constantly and try to follow them when making admin actions, they'll be fine and do good for the project. Jason Quinn (talk) 01:43, 4 September 2021 (UTC)
 * 7) "Need" is imho not a relevant criterion. One action an experienced editor is able to do themselves instead of having to ask for help is already a net positive for the project. I rather have a thousand admins only doing one admin action a month than one admin having to do a thousand actions. "No need" might not have yet lead to an RFA failing but it probably stopped a number of qualified candidates from running. We should instead encourage them even if they don't see much need for the tools. Regards  So  Why  19:05, 4 September 2021 (UTC)
 * 8) This is one part of a catch-22 pendulum: users who haven't "built up enough experience in admin areas" have "no need for the tools", but too much experience and you're hat-collecting; either way there will be at least one voter who thinks this makes a candidate unsuitable. These are silly criteria anyway - nobody has experience with the tools before they pass RfA. Ivanvector (Talk/Edits) 19:33, 4 September 2021 (UTC)
 * Lots of people have experience with the tools before they pass RfA, but the experience usually comes from other wikis (where the tools are the same but the rules and social expectations are quite different). —Kusma (talk) 23:25, 4 September 2021 (UTC)
 * I am not sure I've seen a recent "you're a hat collector" oppose at RFA itself. (I do not know if that's survivor bias.) I can think of one editor who no longer edits much here who might have been seen as one if he had run at the time he was super-active but he has since moved on to bigger and better things within the movement.
 * I think most of us think hat collecting isn't just the appreciable gain of permissions, it's doing so quickly and/or without regard for concerns when raised.
 * As for no experience, indeed as Kusma says, I had experience at other wikis before getting my tools here. Indeed, the delete and protect buttons do not change appreciably with even just non-zero review of relevant deletion policies, and the block button changes only in social acceptability regarding non-vandals. Izno (talk) 06:35, 14 September 2021 (UTC)
 * 1) Yes. One reason for new admins, amongst many, is to ensure existing admins can be relieved. Stifle (talk) 12:18, 6 September 2021 (UTC)
 * 2) No individual person "needs" the tools. The statement seems to be shorthand for "The applicant is unlikely to use the tools". Yet given that Question 1 explicitly asks "What administrative work do you intend to take part in?", it is odd that a !voter might think that an applicant won't use the tools. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  00:03, 7 September 2021 (UTC)
 * Now, more than ever, we need more administrators. After being stable for three years, our number of active administrators is quickly dropping. Having a "need for the tools" is simply this: to help the other administrators with the work. 🐔 Chicdat <sup style="font-family:Times New Roman">Bawk to me!  10:34, 7 September 2021 (UTC)
 * 1) Most editors who I trust enough to have the tools will find important admin work just by being around wielding the tools, even if they're "content-focused". KevinL ( aka L235 · t · c) 18:49, 8 September 2021 (UTC)
 * 2) Quite evidently. Redundancy is important. MER-C 19:08, 8 September 2021 (UTC)
 * 3) Strongly agree. I think this encourages folks to be less than truthful about how they would use the tools, or to stretch what they would actually do with them. Nobody actually knows what tools they will like using until they get them. I had never been a sock detector before RFA, but afterwards found it quite engaging. I thought I would work ANI a lot, and it turns out I now mostly avoid ANI. I thought I would do RFPP, but found the actual button pressing boring. For the practical implication, I think it is time to reform or change RFA question 1. CaptainEek  Edits Ho Cap'n!⚓ 20:24, 8 September 2021 (UTC)
 * 4) It's always been dumb reasoning. Having admin tools just makes editing easier. – filelakeshoe (t / c) 🐱 08:47, 22 September 2021 (UTC)
 * 5) Agreed with Levivich + just because someone is doesn't ascribe to your personal views doesn't mean they would abuse the tools (such as blocking someone with a disagreement). More and more, people are assuming the worst of people without evidence to support such an assertion. Example, I supported Trump over Biden. I have friends who supported Biden over Trump. Just because either of us voted for "the guy" doesn't mean we support everything they do or that you worry they might do. Too many are willing to believe prejudice over facts/demonstrated behavior and RfA is one of those places where it's highlighted. Buffs (talk) 20:02, 22 September 2021 (UTC)
 * 6) I think so. The need for the tools to be given to more editors is why RfA needs more candidates. However, "no need for the tools" can often mean no experience in admin areas, which IMO isn't a poor argument but also shouldn't be seen as a automatic deal breaker IMO.
 * 7) Agreed. No administrator is ever required to use the admin tools in any way. Admins take administrative actions and decisions as a volunteer just like editing the articles. A sudden increase in administrative backlog can happen in any area of Wikipedia at anytime and it is always a good idea to have extra admins who can get the required work done whenever it comes up as it decreases the overall pressure on all the other admins. TheGeneralUser (talk) 21:47, 29 September 2021 (UTC)

O. Oppose (disagree)

 * 1) I don't think the data actually supports the idea that "need for the tools" is a reason why RfAs fail. Many RfAs that pass are from editors with a scant or dubious "need for tools", or where their subsequent use of tools did not correspond with their Q1 answer (or their pre-RfA behaviour, for that matter). Happy to elaborate with more systematic evidence if anyone wants (I was trying not to clog up the discussion too much, if it's not of interest to anyone). More generally, I think a lot of "RfA conventional wisdom" is poorly supported by the data. ProcrastinatingReader (talk) 14:03, 30 August 2021 (UTC)
 * 2) I think the underlying concept here is sound, however, "we can find work" for admins presupposes they can be assigned duties like the day shift at Best Buy. Chetsford (talk) 23:00, 30 August 2021 (UTC)
 * No, obviously not assigned duties, but involved in areas with backlogs would be able to request help. I'm sure I'm not the only admin whose admin actions are often in areas I had not thought I would get involved in in my RFA in 2009.  Ϣere Spiel  Chequers  13:55, 1 September 2021 (UTC)
 * 1) I don't see the problem with requiring people who plan to become admins to make requests for admin actions (in some form or another) before gaining to ability to do it themselves, and that's all need for the tools really means. * Pppery * <sub style="color:#800000">it has begun...  01:11, 31 August 2021 (UTC)
 * 2) It's over-used as a canard, but it is a reasonable response to certain answers to RFA Q1 (What administrative work do you intend to take part in?). User:力 (power~enwiki,  π,  ν ) 17:42, 1 September 2021 (UTC)
 * 3) While "no need for the tools" isn't a particularly strong rational for opposing a candidate at RfA, it's not exactly frivolous and I don't think we could (or even should attempt to) disallow its use. In my opinion, a well articulated answer to Q1 is a much better preemptive measure.--John Cline (talk) 20:58, 4 September 2021 (UTC)
 * 4) I'm in the minority here, but "No need for the tools" often has an underlying reason behind it: the candidate has not demonstrated any aptitude in the areas in which they want to work. If a candidate says they want to work at UAA and AIV and have made zero edits to those noticeboards, while opposition may be superficially because they have no need for the tools, the real reason is that they have zero demonstrated experience in those fields. Would an editor who requests autoreviewer at PERM receive it if they had no article creations? Would someone requesting templateeditor receive it if they had zero experience with templates? No, and for good reason. As for the surface-level reason of "No need for the tools", I'm mildly sympathetic to it because if you're going to make ~200 editors look through your edit history and examine you for seven days, it sure feels like a waste of time if all you use the tools for is deleting pages in your userspace or protecting articles you've written. Sdrqaz (talk) 20:39, 7 September 2021 (UTC)
 * I agree with this comment... which still leaves "no need for the tools" as somewhat problematic.
 * That said, there is at least one or two admins from the past 5 years who have not in fact used their tools much if at all. Should they really have them? Izno (talk) 06:17, 14 September 2021 (UTC)
 * I personally see it as an issue of framing rather than any real substantive difference. As harsh as it may be, I don't see the point in their having the tools. Virtually all of what they do in terms of administrative actions can be covered by using suppressredirect when moving pages (instead of creating a redirect and deleting it) or perhaps not violating INVOLVED when protecting pages. Sdrqaz (talk) 16:25, 15 September 2021 (UTC)
 * 1) I really think this is more of a myth than anything else. Take an RfA like that of Ealdgyth: a content creator with little administrative experience whose own nominators had to preëmptively rebut the "no need for the tools" argument. She passed 250/0/2. To be sure, other similar RfAs have been more controversial, but I agree with Sdrqaz that this generally stems from concerns about experience. Even Ergo Sum's RfA passed, and he had practically zero administrative involvement. While there were a few "no need for the tools" comments, most of the opposes referenced valid questions about a lack of experience. So even if we were to outlaw no-need-for-the-tools opposes (which I think is also a bad idea on principle, per my response to part N), they would simply be parlayed into "lacks experience" opposes, which leaves us no better off than when we started. Additionally, I trust that the crats would discount unpersuasive opposes. As is often the case with RfA, the issue here is more with perception than with reality. Extraordinary Writ (talk) 23:48, 12 September 2021 (UTC)
 * I think such a parlay would be OK. It shifts the statement from "I don't trust you" to "you haven't done anything to be trustworthy" (at least in my head), the second of which can be accepted or rejected by the rest of the community regarding general expectation for trust. Izno (talk) 06:19, 14 September 2021 (UTC)

O. Comments
* Agree that it's a stupid reason to fail an RfA, disagree that it's an actual problem in practice per ProcrastinatingReader. – John M Wolfson (talk • contribs) 15:07, 30 August 2021 (UTC) (Striking; see "Agree" column) – John M Wolfson (talk • contribs) 01:08, 31 August 2021 (UTC)
 * It's a reason RfAs aren't run. I've looked at candidates who are civil, experienced, clueful content creators but the work they do seldom requires requests for admin help, and decided not to approach them. If anyone who thinks that's them wants to be the test case, email me. —valereee (talk) 16:13, 30 August 2021 (UTC)
 * I have shortened the title of this section from "O. Though "No need for the tools" is a reason why RFA are likely to fail, it is a poor reason as we can find work for new admins to do" — Wug·a·po·des​ 20:12, 30 August 2021 (UTC)
 * "Doesn't show a need for the tools" is an argument that I was sure would come up against me the last time it was suggested I put myself through RfA. XOR&#39;easter (talk) 20:19, 30 August 2021 (UTC)
 * On one hand, it's a terrible reason to Oppose a candidate because people change what they focus on. Right now, I'm dealing with expiring drafts and restoring requested ones, an activity that I never did before the pandemic. I need the tools now in a way that I didn't when I first became an admin. However, when I look at Admin Stats I can't help but notice that we have quite a few admins who, over many years, seem to have few logged in admin actions. I know not every administrator action is logged but I have to wonder if some of these folks really have any need for the tools. So, while I think it's a valid question to ask, I also think it's impossible to know today what activity an RfA candidate will be doing in the future and be able to predict who will and who won't "need the tools". Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 03:11, 4 September 2021 (UTC)
 * Question, out of curiosity, for admins elected during the time that ~"what will you do with the tools?" has been obligatory: how closely has your admin activity tracked with your response to that question? This gets at both the role of "need" and the relative value of the standard questions. That question in particular strikes me as primarily an invitation to identify the area where the candidate wants an extraordinary amount of scrutiny. We do not unbundle, nor hold people to that statement, not frown on admins taking on additional work, so why can't we assume "a little of this; a little of that" by default without falling into "no need?" &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 15:44, 5 September 2021 (UTC)
 * Mine has tracked at a higher level as part of my work on TemplateStyling everything.
 * I have spent non-zero time at AIV even though that forum was specifically on the list of things I campaign-promised I wouldn't do.
 * Conversely, I have spent non-zero but not-appreciable time at requests for history merge (just the backlog, because Anthony seems to get everything else). Izno (talk) 06:26, 14 September 2021 (UTC)
 * In addition to my !vote above, I'd comment that adminship is sometimes treated as a natural stepping stone by some candidates, after you've been here for long enough and have enough FAs/GAs/DYKs. Someone who is less prone to assume good faith may call that hat-collecting. Sdrqaz (talk) 20:44, 7 September 2021 (UTC)
 * Given that almost every administrator who has passed RFA having explained what they'll do with those tools actually moves on and does completely different things, it's been a dumb question for a very long time. There are almost no 2+ year admins who still only do what they said they'd do at RFA. Risker (talk) 01:07, 8 September 2021 (UTC)

P. Admin permissions and unbundling
There is a large gap between the permissions an editor can obtain and the admin toolset. This brings increased scrutiny for RFA candidates, as editors evaluate their feasibility in lots of areas. The community should have an WP:UNBUNDLING conversation to determine which, if any, admin tools can be given to non-admins.

P. Support (agree)

 * 1) . As proposer. Z1720 (talk) 14:25, 30 August 2021 (UTC)
 * 2) I know it's a perennial but mark me down for supporting the unbundling of delete/viewdeleted (which requires an RFA per WMF legal, I believe), and block/protect (which would not, I believe). I see the logic in bundling block and protect together, but I don't see the logic in bundling delete/viewdeleted with block and protect. Levivich 18:45, 30 August 2021 (UTC)
 * 3) Right. Some are not controversial, some are. There are people I'd trust with deletion help, but not with blocking (due to temper or whatever, for example). --<sub style="border:1px solid #228B22;padding:1px;">Piotr Konieczny aka Prokonsul Piotrus&#124; reply here 03:29, 31 August 2021 (UTC)
 * 4) A curveball: I don't think technical unbundling can go any further (you need to see revdelled edits sometimes to make accurate blocks/unblocks), but we could have social unbundlings where you can only use the tools for certain circumstances. It's insane that someone who wants the tools just for copyright, just for CSD/AFD or just for blocking obvious NOTHEREs has to go through a person designed to scrutinise people who might wrongly close an ANI discussion over blocking one of our most well-known editors, or make a bad involved block of usually productive editors who are edit warring. If you had a rule along the lines of "someone with the tools for X reason is prevented from ever blocking someone who has made at least 100 edits" then they could do 99% of what they want to with the tools, an improvement on 0%, and enforcement of violations would be completely obvious and straightforward (if the crats were willing to desysop). — Bilorv ( talk ) 17:48, 31 August 2021 (UTC)
 * 5) Bilorv's words exactly. ProcrastinatingReader (talk) 19:39, 31 August 2021 (UTC)
 * 6) Per Bilorv. I don't see such a strong nexus between between deletion and blocking that they necessarily have to be tied together. This may be a perennial proposal, but that doesn't mean we shouldn't consider it moving forward. -- Calidum  19:59, 31 August 2021 (UTC)
 * 7) I think Bilorv has a great idea here. I am an admin and don't really get involved in any high-drama areas because I'm not interested in it. I bet there are others like me who could do a great job and don't need mega-scrutiny. Calliopejen1 (talk) 00:00, 1 September 2021 (UTC)
 * 8) I specifically would like the "viewdeleted" privilege unbundled from the admin toolkit. User:力 (power~enwiki,  π,  ν ) 17:56, 1 September 2021 (UTC)
 * 9) There are plenty of people who see the various rights as too interrelated to justify giving someone one tool when they couldn't be trusted with another... but that means those people would be opposing a regular RfA anyway, and there are enough people who are willing to consider granting unbundled rights that this may make a nontrivial difference. &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 18:02, 1 September 2021 (UTC)
 * 10) Indeed, I am actually pretty disappointed by the perception (which is always cited during the ArbCom desysop case, but is also relevant for RfA) that an admin should not have limitation on tools. This is counterproductive.--Ymblanter (talk) 16:05, 2 September 2021 (UTC)
 * 11) Agreed that we should consider some more unbundling, though I am somewhat conservative on what permissions might be unbundled. The more we unbundle, the more bureaucracy and red tape and processes we create. There is something to be said for having many tools under the single admin process. With that said, there are certainly some admin tasks that would require much more trust than others, and those that are on the lower end of trust needed should be considered for unbundling. Trainsandotherthings (talk) 00:27, 3 September 2021 (UTC)
 * 12) The block tool and the delete/undelete/view deleted revisions tool are different and should be evaluated separately. Whether to empower a user to block/unblock is a decision about their temperament and judgment in social issues and matters of user conduct. Whether to empower them to delete/undelete/view deleted revisions is a decision about their editorial judgment in matters of content. The only overlaps are the basic trust threshold and ability to gauge community consensus. I can't follow the objection to unbundling those two, and I can see clear benefits to doing so. What this means is that I feel that bundled adminship should still exist and be the norm but users ought to be able to apply for unbundled VDR (view deleted revisions), unbundled full access to the delete/undelete tool, and unbundled full access to the block tool, separately. Non-admin blockers would not be allowed to unblock where a sysop has indicated that they've blocked because of now-deleted edits, except in the case of a positive community consensus to unblock. Addendum: both of these would be RFA-like processes with a community discussion and vote, not ask-a-sysop permissions. DRV regulars are examples of the people who would need VDR only.—<b style="font-family: Verdana; color: Maroon;">S Marshall</b> T/C 08:57, 3 September 2021 (UTC)
 * 13) I am a big fan of unbundling tools. The tools that have been unbundled so far have proven the worthwhile workability of this approach. In my opinion, the only ability that should be reserved for admins alone is the ability to add or remove the unbundled tools to users when good cause exists I could be convinced otherwise but I'd definitely support seeing this advanced to phase 2 for further discussion.--John Cline (talk) 17:40, 4 September 2021 (UTC)
 * 14) Unbundle. I actually don't fully agree with Bilorv: if block/unblock is unbundled, sure, a user who is a blocker but can't view revdelled edits may not always be able to place a block. On the other hand: this is also true for admins as they can't always place a block without looking at checkuser data. Obvious vandals and NOTHERE can be blocked without any need to check out revdelled edits. But the other side is more interesting: if a user applies for admin-without-block, they won't be scrutinized over whether they wouldn't incorrectly block someone. (and yes, on another project I know both admins who are likely fine with block/unblock but shouldn't have delete and admins who are generally fine with delete but should definitely have block taken away from them) Both blockers and non-block admins arguably need the right to protect pages, but since protect can be undone without much drama or difficulty (unlike a block or deletion) I wouldn't expect that to be much of a problem. The opposition always brings the same argument: "if I can't trust them with one I can't trust them with the other". By that logic every admin should be a superadmin: bureaucrat, checkuser, interface admin, Arbcom member, steward and member of Jimmy's secret book group. If you are right (you're not, but if), unbundling would have zero effect on the outcome of an RfA. No harm, so the only reason not to try is because you think it won't help. And if we never try, it can never be proven to work. — Alexis Jazz (talk or ping me) 20:05, 4 September 2021 (UTC)
 * The expectation is that unbundling will make it harder to become a full admin, as you need to obtain several of the unbundled permissions one by one, so there is clear harm in trying. —Kusma (talk) 23:22, 4 September 2021 (UTC)
 * I don't understand why you say that a potential admin candidate would have to acquire the unbundled tools one by one before becoming an admin? I would certainly oppose any such requirement should it come up in phase 2 (if this unbundling provision achieves consensus to advance). I don't see what would prevent a candidate from RfA whether they had any of the unbundled tools or not. What have I misunderstood regarding your comment above? Best regards.--John Cline (talk) 13:46, 5 September 2021 (UTC)
 * More and more de fact requirements have been added to RfA over the years, so I expect this one will also be added (people mentioning that they might do some deletions will be asked why they haven't applied for the unbundled version already). Additionally, unbundling reduces pressure on the community to pass enough admins. I think the templateeditor right has made us lose dozens of admin candidates already. —Kusma (talk) 20:47, 5 September 2021 (UTC)
 * Thank you, I understand and agree. I believe our differences are a matter of interpretation. To illustrate, the dozens you see as lost, I see as spared the insult of the de facto rancor they missed for not having to run. And the reduced pressure to pass admins will be met with a reduced pool of candidates to exert pressure on. Incidentally, before the TE flag, tech savvy editors who ran RfA endured some of the most contentious RfAs I've seen and I think the dozens who haven't had to run are a testament to unbundling success. RfA is our Humpty Dumpty. The problem is that we deny its unrepairability instead of moving toward the new things we need.--John Cline (talk) 07:58, 6 September 2021 (UTC)
 * 1) I doubt it'll work, but I'm not dismissing it out of hand. Stifle (talk) 12:20, 6 September 2021 (UTC)
 * 2) Yes. Good editors already know consensus, and how to find consensus as well as any administrator. Good editors already in effect protect pages (protecting good work is part of the editorial job), and good editors already in effect block the doings of poor editing and vandalism (otherwise, the pedia would be even more of a mess, and we even have a bot that does that), and of course editors in good standing already decide deletion. Good editors already cure copyright problems, etc. Since these tasks are already done by good editors, we should make it possible for those who want an extra button to do a task, and if someone wants all those buttons go for it, just like now. And just like now, when one messes up with a permission, call the user account. Alanscottwalker (talk) 19:42, 18 September 2021 (UTC)
 * 3) Block, Protect and Delete form the core administrator rights and it is very unlikely that they will be unbundled. However, I am open to reviewing unbundling of other userights which are currently available only to administrators on a case by case basis. TheGeneralUser (talk) 22:02, 29 September 2021 (UTC)

P. Oppose (disagree)

 * 1) This is a perennial proposal throughout the years, but the consensus seems to be that the things that could be unbundled at this point already have.  – John M Wolfson (talk • contribs) 15:09, 30 August 2021 (UTC)
 * I still have to oppose this. While "social" unbundling might work (and I understand has on other wikis), I feel that it requires too much of an honor system from enwiki. Also, "view-deleted" has to be restricted to admins per WMF Legal, if I'm not mistaken my apologies, I misunderstood that comment, but I still feel that suppression works fine enough for deleted material unsuited for even admin viewing. – John M Wolfson (talk • contribs) 19:11, 1 September 2021 (UTC)
 * 1) I disagree for the same reasons I expressed in the  section below. Mz7 (talk) 18:45, 30 August 2021 (UTC)
 * 2) I have been involved in several, detailed, considerations for carving off narrow further userrights from the admin toolkit, and none have worked. I don't think there's appreciable changes that could be made. Additionally, I believe evidence suggests the opposite - the scale of unbundling has made it harder and harder. Back when you needed the admin toolkit for most tasks it was easy/easier to get it, but now that it's dominated by variants of the "big 3" (deletion, protection, blocking), it's actually harder. Nosebagbear (talk) 20:53, 30 August 2021 (UTC)
 * , correlation does not imply causation. Yes, adminship was easier to obtain in the early days, but not because rights were bundled differently. (I'm not sure what, if any, additional rights a regular admin might have had in the early days anyway?) It was easier because the community was smaller. When everyone knows each other, it's easier to appoint admins. — Alexis Jazz (talk or ping me) 20:16, 4 September 2021 (UTC)
 * Before rollback was unbundled, it was relatively easy for vandal fighters to become admins because giving them single-click access to rollback was seen as beneficial. Now it is much easier to get rollback, but much harder to become an admin. The community of highly active editors wasn't significantly smaller back in 2006, but it was far more open and welcoming. —Kusma (talk) 23:30, 4 September 2021 (UTC)
 * , fine, so temporarily re-bundle rollback and see if we get more admins. If it were retroactive (wouldn't recommend that though), I'd lose rollback and still wouldn't apply for admin. — Alexis Jazz (talk or ping me) 11:26, 6 September 2021 (UTC)
 * 1) As I say every single time this comes up, the tools are a kit. They work together to help admins quickly deal with situations. Protection, deletion, and blocking in particular go hand-in hand. Also, if all you can do is block, that's all you will do when protection might be a better option, and vice versa. Beeblebrox (talk) 22:02, 30 August 2021 (UTC)
 * Seems like you're thinking of only a few use cases (RFPP, ANEW). Why does one need block/protect to handle XfDs? Why does one need block/protect/delete to view deleted revs (eg for abusefilter creation)? Ditto for editing the sysop-protected modules or the interface (WP:IADMIN pre-requisite)? Why do you need block/protect/delete to move protected pages following an RM? ProcrastinatingReader (talk) 22:07, 30 August 2021 (UTC)
 * What if, when reviewing any of those discussions, you see an obvious trolling sock account, but all you can do is delete? You can't delete a troll. Any active admin can tell you that random encounters with disrupotive users are at least as common with noticeboard reports and deletion discussions. Beeblebrox (talk) 22:27, 30 August 2021 (UTC)
 * Fair enough, but you could still close the discussion then request an AIV to deal with the troll, or ignore that particular discussion and go onto a different one. It's an extra step but I feel like the same can be said about many currently unbundled perms. What if you're an EFM but some revs of the LTA filter were deleted (due to abuse etc) and now you don't have all the information to tune the filter accurately? Or if you've got a filter at disallow, but can't edit the disallow message because that requires editing the sysop-protected MediaWiki namespace? A template editor who wants to push a change to a TE-protected template which also requires an update to a sysop-protected template, so they gotta make an edit request? NAC-closing a TfD as delete but then having to tag it as G6 for a sysop to do the actual deleting? Or even a non-admin rollbacker who sees a troll and can only keep pressing "revert" until an admin comes along to block.
 * The list of such examples is probably quite long, but my point is that it seems this issue is already present (in at least some cases) in any unbundled perm short of sysop. No such editor can deal with all aspects of a problem they may come across, but nevertheless the unbundling is seen as a net productivity improvement in most cases. ProcrastinatingReader (talk) 22:46, 30 August 2021 (UTC)
 * 1) This has been repeatedly rejected, and for good reason: Maslow's hammer comes to mind, as does the fact that "if you can't trust them to block, you can't trust them to delete" etc. It could also increase the number of "no need for the tools" opposes, since one's need for each right would have to be justified separately. It's possible that there are a few rights (e.g. editing fully-protected pages) that could be unbundled, but that would be unlikely to resolve the problems with RfA, which is why we're here. Extraordinary Writ (talk) 03:13, 31 August 2021 (UTC)
 * 2) Per EW, if I trust someone to block or delete or protect, I also trust them to know when they shouldn’t. Eddie891 Talk Work 10:58, 31 August 2021 (UTC)
 * 3) There are two major tools in the admin toolkit that (in my opinion) cannot be unbundled. 1) Deletion (including undeletion, view deletion etc) - This tool cannot be unbundled without WMF approval, and last I saw they required an RfA-like process to have access to this tool (i.e. significant scrutiny). 2) Blocking (including unblocking, partial blocking etc) - This tool is the reason for the community divide between admins and non-admins - admins can block the non-admins, therefore have a "power" over them. I cannot see a situation where the community will allow this to be unbundled without an RfA-like process.  Any other tool in the kit (including page protection) I do believe may gain traction in unbundling, but without those two, unbundling becomes a perennial proposal. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 10:52, 2 September 2021 (UTC)
 * deletion, (without undelete or viewdeleted access - these are what require the rfa-like process) may be unbundled if supported by the community - though most workflows here where one may delete generally presume you can reverse yourself. Examples of this setup include the the "closer" group on ruwiki and the "curator" group on enwikiversity. — xaosflux  Talk 15:02, 2 September 2021 (UTC)
 * Deletion without undelete or viewdelete? Yes, I suppose that could work. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:34, 2 September 2021 (UTC)
 * 1) Nope. You should not have delete without undelete, you should not have block without unblock, you should not have protect without unprotect (it is fundamentally unwiki to not be able to fix your own stupidity). Whoever has block should have protect and vice versa, to avoid use of the wrong tool. Whoever has unblock must have viewdeleted. Perhaps you can give people delete/undelete only, making them ask others for SALTs and blocks of attack page creators, but I just don't see a large number of people who can be trusted with delete/undelete but not with the rest of the tools. "Delete" NACs of CFDs/TFDs that use WP:CSD could be avoided by this, but the complete lack of a CAT:CSD backlog doesn't point towards a need to avoid them. —Kusma (talk) 10:56, 3 September 2021 (UTC)
 * 2) When all you have is a hammer, everything starts looking like a nail. There certainly might be stuff like rollback and account creation that can successfully be "unbundled", but sometimes protection can be, for example, a better solution than blocking. But someone who only has access to the block tool might be tempted to use that anyway, because that's all that's in their hand. I cannot think of any conceivable circumstance where I would trust an editor with, say, the block tool but not the delete tool. They're trustworthy or they're not, period. Seraphimblade Talk to me 23:40, 3 September 2021 (UTC)
 * 3) What Seraphimblade said. Everything that could be reasonably unbundled has been but the last "big three" (blocking, deleting and protecting) should not and there is no conceivable circumstance in which a candidate could be trusted with either of those but not the others. Regards  So  Why  19:08, 4 September 2021 (UTC)
 * Greetings SoWhy. The conceivable circumstance regarding the issue of trust is that acquiring these tools through RfA requires the commitment of nearly irrevocable trust where acquiring the unbundled tools requires a reasonable trust that can easily be revoked upon cause. In my opinion we are much more likely to achieve a consensus for more tool unbundling whereas I don't think we're anywhere near emplacing an expedient process to desysop if necessary. That leaves one as a managable risk and the other with no margin for error. Regarding your mention of the "big three" in tools, you didn't mention the ability to add and remove flags which I Invision as the future of RfA and the most important role of an admin in the unbundled Wikipedia that we ought to discuss. Please consider supporting this provision so we can discuss it further in phase 2. Best regards.--John Cline (talk) 12:14, 5 September 2021 (UTC)
 * That does however not address my comment. The question was, in what realistic scenario is there a candidate which e.g. can be trusted to block but not to delete or protect? Additionally, as has been pointed out, these tools go hand in hand when performing many tasks, e.g. SALTing a page title after deleting the page or blocking an editor and deleting their contributions. The technical ability to add or remove flags can easily be regulated by requiring processes to assign them, just as crats are not allowed to hand out the sysop-bit to just whomever they like. Regards So  Why  14:58, 5 September 2021 (UTC)
 * I think I can easily imagine a user (not to provide real examples here) whom I would trust blocking obvious vandals but not blocking experienced users who are on the edge of being incivil or edit-warring. Whereas this can not be technically enforced by unbundling, but can be done by imposing limitations. I also can imagine an administrator who would be doing protections fine, but would make systemic mistakes concerning civility - and, consequently, civility blocks - and this can be a case for giving the protection tool but not the block tool. Though I think limitations would work better - and would actually help RfA, if they are instituted in a formal way. Right now, a candidate can pledge not deleting articles, pass an RfA and start deleting, and therefore users would oppose the candidate if they have any suspicion that deletions might cause trouble.--Ymblanter (talk) 15:21, 5 September 2021 (UTC)
 * Those are not unbundling ideas though but ideas akin to the "adminship lite" proposal that was floated previously. I would also oppose those for different reasons but it has nothing to do strictly speaking with unbundling the tools. Even the hypothetical candidate with a history of making mistakes regarding civility would require the block tool to handle obvious vandals. Regards So  Why  16:53, 5 September 2021 (UTC)
 * 1) I think we've unbundled everything that should be unbundled.  I definitely do not think anything related to deleted content should be unbundled.  It's called a tool kit for a reason. Risker (talk) 01:05, 8 September 2021 (UTC)
 * 2) The more we unbundle, the more bureaucracy emerges around managing each permission. KISW: Keep it simple, Wikipedia. CaptainEek  Edits Ho Cap'n!⚓ 20:20, 8 September 2021 (UTC)
 * 3) It's hard to split out the tools, as their use is inherently tied together and the level of trust needed for each would be the same, rather than "trust to delete but not block". The ability to view deleted edits has been crucial for me to make block decisions, or having blocked someone creating attack pages, those need to be deleted as well.  Spencer T• C 02:21, 14 September 2021 (UTC)
 * 4) This is a perennial. I think we've reached the limit if what can be unbundled and there is no compelling reason to do any more. IMO the lobby for unbundling comes from people who want more rights but who know they either wouldn't pass RfA to get the full tool set, or they are too scared to face the music for 7 days and nights. Kudpung กุดผึ้ง (talk) 14:47, 19 September 2021 (UTC)
 * 5) We have already unbundled the least controversial tools. Any more unbundling will lead to accusations (some warranted, some not) of incompetence in view of the absence of community consensus to grant the tools. At least this level of drama should be avoided with RfA. <b style="color:#808000">Axl</b> ¤ <small style="color:#808000">[Talk]  08:24, 28 September 2021 (UTC)

P. Comments

 * I can see both sides in this proposal. I'm mostly active at WP:RFPP or WP:RFP/ACC or in making admin closures, and I can count on one hand the number of accounts I've blocked (four). On the other hand, unbundling increases the overall complexity of Wikipedia as a system which is a larger problem and one we should avoid. Chetsford (talk) 22:55, 30 August 2021 (UTC)
 * I am still trying to work out details for a "no-private-information-admin" role. Right now, "deleterevision" grants both revdel and unrevdel permissions.  If we're not going to have a consensus of "if it's not oversightable, it doesn't matter if it's in the revision history" it would be problematic to have a corps of admins without the ability to do their own revdels.  Deleting pages, on the other hand, is separable from undeleting; we could create an AFD-closer role with access to delete.  Since "closing AFDs" is consistently the least-backlogged admin task, that's not too exciting from my point of view. User:力 (power~enwiki,  π,  ν ) 21:39, 3 September 2021 (UTC)
 * I do not agree that we another conversation about unbundling. I do agree, as phrased in Q, that most admins work in one or a few areas, at least to start. The statement in Q is something of a solution-seeking statement, but I do like the solution as a possibility for this case. Passive administrator actions such as view deleted revs might make it difficult to judge whether someone has stayed within their bubble. That all said, I don't know what this really gains us over the campaign promises made at RFA itself already today. --Izno (talk) 06:14, 14 September 2021 (UTC)

Q. Splitting the admin role
Most admins work in one or a few areas. RFA candidates are evaluated for the area of the project where they plan to use the tools (DYK, AfD, copyright, etc.) but are then are given lots of permissions that have little or nothing to do with their area of expertise. The admin toolset should be divided up so that candidates apply for the toolset useful in specific areas, and only for use in that area. This will lower the candidate scrutiny as editors will be evaluating the candidate's experience in a specific area, instead of the whole project.

Q. Support (agree)

 * 1) As proposer. Z1720 (talk) 14:38, 30 August 2021 (UTC)

Q. Oppose (disagree)

 * 1) Per my comments on proposal P. – John M Wolfson (talk • contribs) 15:09, 30 August 2021 (UTC)
 * 2) I disagree. In my view, the abilities to delete, protect, and block are closely interconnected, and they form the core of the administrator toolset. If you have one, you often need to have the other two in order to respond effectively to an administrative task. We can and have unbundled other permissions within the admin toolset, but I believe it is very unlikely that these three will ever be separated from one another, and I would caution against excessive energy being spent on ideas along this line. Mz7 (talk) 18:39, 30 August 2021 (UTC)
 * 3) This is just unbundling again, so see my above oppose. Beeblebrox (talk) 22:03, 30 August 2021 (UTC)
 * 4) See my response to part P. Extraordinary Writ (talk) 03:14, 31 August 2021 (UTC)
 * 5) This is unbundling done badly. User:力 (power~enwiki,  π,  ν ) 17:45, 1 September 2021 (UTC)
 * 6) No way. If an editor is not suitable for one area of adminship, they are not suitable for adminship, period. 🐔 Chicdat  <sup style="font-family:Times New Roman">Bawk to me!  11:21, 2 September 2021 (UTC)
 * 7) This is a nice idea in theory, but I fear the effect would be to create multiple RfAs. Reforming one is hard enough, imagine reforming 5 or even 10! Admins may also change the areas they work in over time, and making them go through more applications is a pointless burden. As power~enwiki said, this is unbundling done badly. Trainsandotherthings (talk) 00:29, 3 September 2021 (UTC)
 * 8) Absolutely not. This is a wiki, you don't need permission to edit, and similarly you shouldn't need to ask for a bunch of extra permissions every time you wish to admin in a new area. If you're competent enough you'll figure out new things. —Kusma (talk) 11:00, 3 September 2021 (UTC)
 * 9) One of the things most critical about an admin candidate is trust in their judgment. That would include trusting them to know when they are not sufficiently well-informed about something to use the tools. If we can't trust their judgment they shouldn't have any admin access, and if we can this is a nonissue. Seraphimblade Talk to me 23:36, 3 September 2021 (UTC)
 * No, no, no. When I was a new editor, I spent a lot of time on noticeboards and during my RfA, that activity, which I had viewed positively, came back to bite me, hard. I got the message. Unless one has some technical skills, it's very likely that admins will rotate among a variety of activities. They could become a checkuser. They could work in resolving disputes. They could focus on blocking vandals and obvious sockpuppets. They could volunteer for ArbCom or spend time on arbitration enforcement. But, after a while, it gets old and you move on to other activities. The activities I spend time doing today are different than those I did a year ago, two years ago, and longer ago. An RfA is not a pledge to spend ones time doing any one activity for the rest of ones time on the project and forcing that would just result in more admins leaving admin work. This is a volunteer activity, not a work contract. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 03:22, 4 September 2021 (UTC)
 * 1) Per oppose on P. and basically all of the above. Regards  So  Why  08:00, 5 September 2021 (UTC)
 * 2) RfA should remain an all or nothing affair (with regard to the admin toolkit). Anything less needs to come by some other means.--John Cline (talk) 15:47, 5 September 2021 (UTC)
 * 3) Cas Liber (talk · contribs) 12:22, 7 September 2021 (UTC)
 * 4) I think we've sliced off everything that should be sliced off the "admin-only" profile. There is a limit to the number of hoops people will jump through, and doing this just means that there is less resiliency in the admin corps. Risker (talk) 00:46, 8 September 2021 (UTC)
 * 5) Per my comments on proposal P. Kudpung กุดผึ้ง (talk) 14:49, 19 September 2021 (UTC)
 * 6) I don't think this would be helpful and would also cause extra bureaucracy. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 13:37, 27 September 2021 (UTC)
 * 7) I am open to the idea of unbundling userrights which are currently available only to admins, but I do not agree with splitting up the admin role. Areas of interest and expertise can change over time and an editor always learns through their practice and mistakes (including admins). TheGeneralUser (talk) 22:11, 29 September 2021 (UTC)

Q. Comments

 * This, but not this? I do think this is part of the problem with our overall RfA process, but that it is that the role of sysops has shifted as the editing base has expanded.  The ability to make technically breaking changes has always been part of sysop, but the parts dealing with conduct management of other editors have greatly grown as the editing base has become more complex and sizable.  While it is still possible for someone to pass RfA if they only satisfy one of these major silos (technical capabilities, volunteer management) candidates have to have a very strong argument for one without the other.  This seems to result in candidates needing to be evaluated on both fronts, even if they never intend to work in one of them. —  xaosflux  Talk 14:56, 30 August 2021 (UTC)
 * I agree this is one of the problems, but I suspect many admins help at least occasionally in areas they didn't at the time of their RfA think they'd end up working in. —valereee (talk) 16:24, 30 August 2021 (UTC)
 * How is this different from P? &#8213; Qwerfjkl  talk  18:48, 30 August 2021 (UTC)
 * P is asking if we should have an RFC on creating new permissions that are currently in the admin toolset, similar to how Rollback was created a few years ago. Q is to split up the admin role into specific responsibilities using the RfA application structure, but the discussion would only focus on the candidate's abilities in the place they are applying to. Z1720 (talk) 19:05, 30 August 2021 (UTC)
 * Isn't this the same as P? And more difficult to do in practice. I'd hat this as a duplicate proposal. --<sub style="border:1px solid #228B22;padding:1px;">Piotr Konieczny aka Prokonsul Piotrus&#124; reply here 03:30, 31 August 2021 (UTC)
 * Am I understanding that this is roughly "You get the full toolkit in X area" of the site? I understand how that might work as a "DYK Admin", you can promote, delete, and block in the DYK area - but how could it work as a "Copyright Admin" - that's instantly the full encyclopedia, as long as there's a copyright issue? I'm not really seeing how this could work. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 10:56, 2 September 2021 (UTC)

R. Time commitment for RfA
Potential candidates are deterred by the perceived time commitment to prepare for RfA, justify their request, answer questions and shepherd their request through.

R. Support (agree)

 * 1) I can see this; I was a serious adminabile from, on and off, ~October/November 2019 until I actually passed in October 2020. I think much of the backlog is trying to request a nom who never gets back to them. That said, I don't think the week-long RfA itself can be meaningfully truncated, and this seems more a combination of the various factors above. – John M Wolfson (talk • contribs) 01:05, 31 August 2021 (UTC)
 * Yes, due to a combination of factors, some of which are individually listed above (e.g. too many questions, justifying "need for the tools", high level of scrutiny). It's not that editors don't have the time, but that there is an opportunity cost for those who like spending most of their WP time editing articles. I have no idea for how many people it is an issue, but it is for some. Nurg (talk) 04:46, 1 September 2021 (UTC)
 * 1) Agree that it's an issue. My impression is that someone doing an RfA really needs to clear their schedule for an average of at least a couple hours every day during that week. Unfortunately, most Wikipedians have lives outside of Wikipedia that they can't necessarily commit to being available for a whole week like that (jobs, school, family, health issues, other volunteering, etc.). That said, the reason it takes so long is entirely covered by the other discussion items here, so I don't think there's a real exigency for this one. &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 17:57, 1 September 2021 (UTC)
 * 2) I've definitely seen RfAs gather immense opposition because a candidate wasn't available every day during the week when it had nothing to do with their suitability. Like Rhododendrites says, that's just a normal feature of many people's lives. You don't know what's going on in someone's life or career so you can't tell them that "being free for a week isn't that big of a deal". There are people who have to respond to emergencies without warning, or where they're free in chunks but never for seven consecutive days. That someone isn't free for several hours in every day of a week that they can plan in advance is not even remotely close to someone not being active enough (over a much longer period of time) to be useful as an admin.  Unfortunately, this issue is widely prevalent across Wikipedia, where there is an assumption that people will be active within 24 hours or less at any given moment. With gaming edit wars that admins refuse to enforce blocks for, discussions where a day's silence will be taken as opposition and so on, we systemically disadvantage people with more hectic real life commitments from making contributions in many areas, and drive many of them away. I can't name a single editor who just spends a few hours each Sunday improving the encyclopedia, but I can imagine a large number of people would be able (and willing) to sustain such an editing habit but no more. — Bilorv ( talk ) 19:18, 1 September 2021 (UTC)
 * 3) Being able to clear my schedule for 2-3 productive hours every day for 7 days in a row just isn't something I could ever guarantee, and I'm probably luckier than average in terms of how much I can shuffle my commitments around. XOR&#39;easter (talk) 23:32, 1 September 2021 (UTC)
 * 4) RfA is a major time commitment. Many Wikipedians have jobs, school, or other major time commitments. I don't think anyone's boss would give them a week off because "I want to go through the admin application process on Wikipedia." There should be some way to either reduce the time commitment, or spread it out over a longer period. People with busy off-wiki lives should not be penalized for it when it comes to RfA. Trainsandotherthings (talk) 00:32, 3 September 2021 (UTC)
 * 5) Per Bilorv. I'm not sure that time commitment for the RfA week is that big a deal in itself - several hours a day is overkill, though it's likely you won't get much other on-wiki business done that week. But there is definitely a lot of systemic bias around availability. For a project whose main task can easily be done in an on-and-off pattern, many aspects of our internal dynamics are dominated by people who not only have a ton of spare time, but have it continuously for months to years, often throughout the day, and implicitly expect similar levels of availability from others. Opabinia regalis (talk) 05:13, 7 September 2021 (UTC)
 * 6) Candidates shouldn't be expected to be available around the clock for the entire duration the RFA is open. -- Calidum  17:40, 7 September 2021 (UTC)
 * 7) This will be a problem for a somewhat small fraction of potential candidates (poor health, busy family life). A bigger problem may be the time commitment of all !voters together, which may dissuade  moderately active potential editors from running, as they feel their future admin actions don't justify the roughly 100h editor time (assuming average half hour per !voter) of people deciding whether to support. FemkeMilene (talk) 06:54, 12 September 2021 (UTC)
 * I'm going to have to strongly disagree with that; every successful RfA since 2015 has had at least 100 supports; my own, which was hardly active, had 154. Also, it only takes rather 5-10 minutes at most for the average !voter, not half an hour (though there are useful sleuths such as Kudpung who take up to 2 hours deeply researching the candidate), and even if that adds up such costs are distributed, meaning that I don't feel the costs of other !voters taking their time when I cast mine. – John M Wolfson (talk • contribs) 16:21, 12 September 2021 (UTC)
 * 1) I was deterred for multiple years after a bad review (probably reasonably bad) at ORCP related to justifying my then-future request. I know several other candidates and would-be candidates have been deterred by other factors in the list here specific to the RFA period itself, because they have told me so. (I am not sure what to do about the problem; I did not find RFA a particularly time-draining experience.) --Izno (talk) 06:08, 14 September 2021 (UTC)
 * 2) Agree entirely. No one needs to be present at their RfA for more than an hour once in 24. The problem is that an RfA is such an exiting process that most candidates stay glued to the computer 24 hours a day to catch every vote and comment coming in. I know I did, and that was my fault.  And if it is the reason, then the lack of preparation is a justifiable part of the weeding out process.  It can be seen all the time at WP:ORCP just how reluctant potential candidates are to prepare themselves. If they haven't got the time to read WP:RFAADVICE, then they haven't time to do some of the time consuming tasks that are required by adminship. Problem is that too many admin potentials see the job as just a button mashing MMORPG. Kudpung กุดผึ้ง (talk) 15:05, 19 September 2021 (UTC)
 * 3) Yes, but as editors can choose when to transclude their RfA they can pick and choose when to start it. However, regardless of when this takes place it still takes up a lot of time, especially for an editor who is generally very busy and so is active for short periods every day. However, I don't think adding extra time for the RfA would help necessarily as they get to choose when to go. Dreamy <i style="color:#d00">Jazz</i> talk to me &#124; my contributions 13:40, 27 September 2021 (UTC)

R. Oppose (disagree)

 * 1) I just don't think this is likely to be a decisive factor for editors considering whether to run the gauntlet. Even it it were, I doubt we could do anything about it (aside from shortening the seven-day period, which is a non-starter). I agree with JMW above that this likely stems from issues being discussed elsewhere (e.g. too many questions, high level of scrutiny) and from factors that are beyond our control (e.g. delays in finding a nomination). Extraordinary Writ (talk) 03:22, 31 August 2021 (UTC)
 * 2) Unless any data to back this up exists, I doubt it. If someone doesn't have the time to be active in their RfA, they are not likely to have time to do any serious work around here anyway. --<sub style="border:1px solid #228B22;padding:1px;">Piotr Konieczny aka Prokonsul Piotrus&#124; reply here 03:31, 31 August 2021 (UTC)
 * 3) I note its in the question header this time, and I think so correctly - while there certainly is some preparation time, and it varies, it wasn't too long for me. But the perceived time may indeed be significant. And what we can do to fix that (both in truth and perception, we should). However, I am in oppose because I'm not sure how many actual candidates we lose for what I might call "avoidable time take-up". Reading the last 30 RfAs before mine was somewhat time-consuming, but also made me more confident. My nom was extremely busy, but I knew that when I asked, so wasn't a surprise. I wouldn't have thought of the normal noms there was normally a huge time issue these days. Nosebagbear (talk) 13:15, 1 September 2021 (UTC)
 * 4) This RfA reform cannot fix every single possible issue with RfA. We should focus energy on the most high-impact and likely-to-gain-consensus issues. While time commitments of RfA may well be an issue, I'm not convinced they're a major one. ProcrastinatingReader (talk) 11:45, 3 September 2021 (UTC)
 * 5)  This is not something I see as an urgent issue that needs a new rule to solve it. If we're not proposing a solution this is just noise. Beeblebrox (talk) 16:03, 3 September 2021 (UTC)
 * 6) I don't see preparing for an RfA to be a burden. After 15+ years, it's a pretty predictable process and the typical problem is candidates who do no preparation at all. The "cost" is not time & energy but the psychological cost of being a focus of scrutiny for seven very long days. For me, it was an eternity. It's a long time to be on the hot seat for an entire community to discuss. If I could make any recommendation, it would be to shorten the RfA period to 3-5 days. Right now, it is a marathon and people quit along the way because it can be exhausting, mentally and emotionally. Nothing much changes in Day 6 and Day 7 so just shorten the period of torture and you'd have fewer people drop out because of how unpleasant things are. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 03:29, 4 September 2021 (UTC)
 * 7) The time commitment is very reasonable, it's not like we're asking people to put their lives on hold to park in front of their keyboards continuously for a 168-hour marathon. Prospective candidates who are deterred by having to answer questions on their own schedule for one whole week are the ones who should be deterred. Ivanvector (Talk/Edits) 19:39, 4 September 2021 (UTC)
 * 8) I don't think this is a particular problem. Stifle (talk) 12:23, 6 September 2021 (UTC)
 * 9) Cas Liber (talk · contribs) 12:23, 7 September 2021 (UTC)
 * 10) I suspect that, realistically, a lot of the complaints about "time commitment" aren't really about the RFA itself, they're about the work a good candidate should do *before* RFA.  That would be reviewing their own edits, giving thought to any areas that have been problems, interpersonal confrontations, history of work in relation to deletions, ensuring that they have contributed quality content. That takes a fair amount of work right there. If that pre-work is done, then RFAs generally go fairly smoothly. Risker (talk) 00:44, 8 September 2021 (UTC)
 * 11) Largely per Risker. The time needed to amass the skills to be a decent candidate is considerable. The additional time required to prepare for RFA is trivial. Furthermore, I'm fairly sure there isn't even a perception issue here; this hasn't been mentioned by anyone who has turned down the offer of a nomination from me (I may, of course, have a non-representative sample). Vanamonde (Talk) 12:25, 8 September 2021 (UTC)
 * 12) Per Risker. KevinL ( aka L235 · t · c) 18:50, 8 September 2021 (UTC)
 * 13) My RfA took no more time than I usually spend in a week on Wikipedia, and I had a very spicy and time consuming RFA. Admittedly, I probably put at least 20-30 hours into Wikipedia each week anyway. But I think most folks looking to be admins are already putting in at least an hour a day, which should be more than enough for all but the craziest RFAs. As further proof that it didn't take too much time, I RFA'd during finals week my last semester of college, and was able to balance great grades on my finals, writing a thesis, and passing RFA. (Though I do recommend others not repeat my questionable move...) CaptainEek  Edits Ho Cap'n!⚓ 20:07, 8 September 2021 (UTC)
 * 14) Sure, there is a huge amount of time and energy involved in preparing oneself for RfA, but that is a part of the process and I don't see anything wrong with it. TheGeneralUser (talk) 05:21, 29 September 2021 (UTC)

S. Too few trusted and experienced nominators.
There is a lack of qualified RFA nominators. &#8211;<span style="font-family:CG Times, times"> MJL &thinsp;‐Talk‐☖ 02:31, 31 August 2021 (UTC)

S. Support (agree)

 * 1) As originator. According to my research of the RFA process in the last two years, you only have a small handful of really trusted and really experienced RFA nominators. In my opinion, more active RFA nominators means more RFA candidates, and that means more admins. &#8211;<span style="font-family:CG Times, times"> MJL &thinsp;‐Talk‐☖ 02:31, 31 August 2021 (UTC)
 * 2) In my opinion on J, above, I stated my belief that the low numbers of both nominees and candidates should be seen as a cause, not just a symptom, of the problem. I believe MJL's evidence presented below is powerful proof. Firefangledfeathers (talk) 02:37, 31 August 2021 (UTC)
 * 3) Per my comments on R, above.  – John M Wolfson (talk • contribs) 12:49, 31 August 2021 (UTC)
 * 4) I take "qualified nominator" to mean "someone with the experience to know who to ask to run and how to ask them". There don't appear to be many who are able and willing. Where this fact sits in the feedback loops of cause and effect that might be at play here, I don't know. XOR&#39;easter (talk) 18:05, 2 September 2021 (UTC)
 * 5) I agree there are too few nominators. I think the data (below) backs that up. I'd like to see more admins nominate candidates. Levivich 16:17, 4 September 2021 (UTC)

S. Oppose (disagree)

 * 1) Ritchie, Tony, Barkeep49, and other regular nominators do the community a tremendous service by seeking out qualified candidates and coaching them from an RfA. However, if they were to stop, I think prospective candidates would not find it difficult to find other competent nominators amongst our ~470 active administrators and hundreds of other well-respected regular editors. –&#8239;Joe (talk) 09:38, 31 August 2021 (UTC)
 * No, this is not the case. ~ ToBeFree (talk) 17:51, 31 August 2021 (UTC)
 * 1) I think Joe Roe is correct - there are others who seek out potential RfA candidates, but those I ask (for example) have almost already been asked by one of the names of the regular group. The limitation is willing candidates, not nominators. Should we get more candidates, or lose some of our nominators, others could easily step into the breach. It's not the limiting factor. Nosebagbear (talk) 13:00, 1 September 2021 (UTC)
 * 2) The appearance of a lack of nominators is only because of the minimal need due to a lack of RfAs.  This is a symptom, not a cause. User:力 (power~enwiki,  π,  ν ) 17:46, 1 September 2021 (UTC)
 * 3) Per my comments below ~ Amory <small style="color:#555"> (u • t • c) 18:39, 1 September 2021 (UTC)
 * 4) Per my comments below. Vanamonde (Talk) 06:49, 2 September 2021 (UTC)
 * 5) Qualified nominator should not be a thing. We allow self noms for a reason. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 10:58, 2 September 2021 (UTC)
 * 6) We have some very valuable community members, such as those Joe mentioned, who have done an outstanding job with seeking out and mentoring RfA candidates. There are many things broken with RfA, this is one of the few things that works well currently. Candidates may also self-nominate, which completely avoids this issue. Trainsandotherthings (talk) 00:34, 3 September 2021 (UTC)
 * 7) I don't think I've ever checked who the nominator is, at RfA.  I generally don't bother to read the nomination statement.  My method with candidates I don't know is to check a random sample of their contributions and then read the opposes to see if there's anything substantial there.—<b style="font-family: Verdana; color: Maroon;">S Marshall</b> T/C 09:48, 3 September 2021 (UTC)
 * 8) Plenty of admins have nominated themselves. Beeblebrox (talk) 16:02, 3 September 2021 (UTC)
 * The last successful self-nom was Floquenbeam, two years ago. The last successful self-nom of someone who was not a former admin was JJMC89, Christmas 2018, more than 40 RfAs ago. Recently, it is not "plenty". —Kusma (talk) 21:31, 3 September 2021 (UTC)
 * 1) In my experience, the folks who nominate editors for an RfA are the people who are there, encouraging them to do it. I'm not sure what other "qualifications" are necessary besides community respect and support for the candidate. I haven't volunteered to nominate anyone because, frankly, I probably discourage more people from going through an RfA than I encourage. It was an unpleasant experience for me, and for my wonderful nominators, and so I do not encourage any editor to undergo an RfA unless they are mentally prepared for it. I found that after considering it, editors I've talked to usually decide against having an RfA. I will say that having supportive nominators is what got me through my RfA. They can be an invaluable resource for a candidate, especially one who is not a sure bet. I always look at who is nominating a candidate and it does affect my opinion of the candidate when I see who is there, encouraging them. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 03:38, 4 September 2021 (UTC)
 * 2) I've seen offers to nominate from many people other than those listed. I think the most frequent nominators just put more time into finding people, while the others would still happily nominate if asked. I mean, are there any admins here who would not nominate a good candidate? &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 12:26, 4 September 2021 (UTC)
 * 3) I don't see how this is a problem with RfA. Are there a deluge of voters basing their opinions in the number of nominators, or which signature is in the nomination section? Not in my experience. Ivanvector (Talk/Edits) 19:46, 4 September 2021 (UTC)
 * 4) There are ample numbers of experienced editors who have clue in this area to assess and assist potential candidates Cas Liber (talk · contribs) 12:24, 7 September 2021 (UTC)
 * 5) Heck, even I have co-nominated someone this week. Risker (talk) 00:18, 8 September 2021 (UTC)
 * 6) I think these days every candidate who has a realistic chance at RfA can easily find half a dozen respected admins to nominate them. Also, what Risker said . KevinL ( aka L235 · t · c) 18:52, 8 September 2021 (UTC)
 * 7) There is absolutely no shortage of experienced, qualified nominators. Even since I retired I still get at least one request a month in my email. I usually tell them to go and ask, , etc., They are either not candidates of the right calibre and they haven't done their homework, or they got a good score at ORCP and now they are looking. Kudpung กุดผึ้ง (talk) 15:20, 19 September 2021 (UTC)
 * 8) The problem isn't that there are "too few nominators", the problem is that having a nominator is now seen as necessary. I remember years ago there was a troll who left the same comment on every self-nom, calling it prima facie evidence of power hunger. Turns out the community ended up agreeing with him! – filelakeshoe (t / c) 🐱 08:49, 22 September 2021 (UTC)
 * 9) Enlisting more nominators strikes me as akin to rearranging chairs on the Titanic: while it might be useful in and of itself, it's not going to resolve the far more fundamental issues with the process. Extraordinary Writ (talk) 22:04, 26 September 2021 (UTC)
 * 10) There are just enough trusted and experienced nominators for someone who needs one (however, I agree that having more types of different and diverse nominators would surely be a good thing), but I don't see how that is a fundamental problem for RfA itself. Having a nominator is optional and people can self nominate themselves too. Like other people have said above, this is because of less number of people running (a symptom) and not the cause. TheGeneralUser (talk) 04:02, 29 September 2021 (UTC)

S. Comments

 * See statistics and interesting facts below (data does not include self-noms):
 * There are six users who have made at least four RFA nominations since 2019: Ritchie333 (18), Barkeep49 (8), Amory (6), TonyBallioni (6), Vanamonde (5), and L235 (4). These represent the Top 6 RFA nominators.
 * 29 of the 40 successful RFAs within the last two years involved a (co-)nomination from one of the Top 6.
 * The success rate for a nomination involving a Top 6 nom was 91% versus 67% if none were involved.
 * Ritchie has twice the amount of successful nominations as Barkeep49 does.
 * There has never been an instance where a candidate with the support of at least 2 of the Top 6 nominators has failed to gain enough support.
 * Cheers, &#8211;<span style="font-family:CG Times, times"> MJL &thinsp;‐Talk‐☖ 02:31, 31 August 2021 (UTC)
 * Compelling evidence. I urge you to uncollapse it. Firefangledfeathers (talk) 02:37, 31 August 2021 (UTC)
 * Done. &#8211;<span style="font-family:CG Times, times"> MJL &thinsp;‐Talk‐☖ 02:45, 31 August 2021 (UTC)
 * This may be true, but I don't think the reliance on nominators is a good thing. I would love to see some successful self-noms again. —Kusma (talk) 06:12, 31 August 2021 (UTC)
 * Speaking as someone on this list, I don't think these data show a dearth of nominators being a serious problem. I began systematically searching for candidates a little while after Ritchie did, to the point where I'd frequently find he'd preempted my offer of a nomination to someone by several months. But if a candidate gets multiple offers, I suspect they're a standout candidate to begin with, often one who has to be frogmarched to RFA. Therefore the correlation between multiple "top 6" conominations and a high passing percentage has more to do with candidate quality than any added value to a nomination. Don't get me wrong; I think any extra effort to find candidates by trusted community members is a good thing. But Of the ~25 candidates I've offered to nominate, all but five have declined, and I don't think there's a huge difference in qualifications between those who accepted and those who didn't. I'd rather focus on the reasons why candidates declined; which are, a disinterest in maintenance work, a horror of RFA, and/or a feeling of not being qualified/not needing the tools. I'm obviously not going to ping those folks here, but I will ping ; perhaps they'll have more to add. Vanamonde (Talk) 11:41, 31 August 2021 (UTC)
 * I've been largely out of commission thanks to COVID parenting, but I'm fairly in agreement with this. A nomination from an experienced and trusted editor vouch for you is obviously valuable, but I think at least half of that is just that someone(s) put in the effort to seek out, find, and otherwise cajole candidates.  There are dozens of candidates out there that don't know it or don't care or don't see the value, so if I'd had success as a nominator, it's because I really tried.  There is not a dearth of qualified nominators, there is simply a dearth across the board.
 * Doing so, I might add, only makes the process harder. We should be running more candidates, such that having a specific nominator is meaningless.  A high success rate at RfA means we're not running enough people. ~ Amory <small style="color:#555"> (u • t • c) 18:39, 1 September 2021 (UTC)
 * It used to be that people who were willing to nominate would get a trickle of emails from potential RFA nominees many of whom were excellent candidates. As the section above shows, our active nominators now are people who put a lot of effort into finding potential candidates. We may have a shortage of people in that group, and there is a risk that they all look in the same bits of the pedia and miss qualified Wikipedians elsewhere. But I'm pretty sure we have a lot of nominators around who are in my position, willing to nominate good candidates if approached but rarely being asked for a nomination.  Ϣere Spiel  Chequers  12:11, 31 August 2021 (UTC)
 * I honestly dislike the idea that, were I (or anyone else) to take up regular nomming, they would spend a lot of time treading over the same ground as other people in assessment of ad hoc requests for noms and headhunting for the others. An rfa-noms-l@wiki.org email list or something would be great on the one hand but on the other hand, eventually it would become sufficiently public that you might as well put it on wiki. --Izno (talk) 06:05, 14 September 2021 (UTC)
 * I don't think it's treading over the same ground because nominators are not all the same. A candidate saying no to one nominator doesn't mean they won't say yes to another, and likewise for nom requests. Levivich 03:50, 15 September 2021 (UTC)
 * Yes, that is true, but I still don't want to spend time reviewing someone for X hours if I know someone else has done so in recent times. And while I like to bug certain users offhand that they should run, I don't think potential candidates like being approached all the time. Izno (talk) 15:36, 15 September 2021 (UTC)

T. Expecting the unexpected
The only known thing about RFAs is that they are rather unpredictable. It isn't clear for a given candidate what is on and off the table regarding their edit history. The things that have the potential to sink an RFA can almost never be guessed beforehand. &#8211;<span style="font-family:CG Times, times"> MJL &thinsp;‐Talk‐☖ 02:44, 31 August 2021 (UTC)

T. Support (agree)

 * 1) As originator. &#8211;<span style="font-family:CG Times, times"> MJL &thinsp;‐Talk‐☖ 02:44, 31 August 2021 (UTC)
 * 2) Yes, I raised this issue twice above before reading this. A huge problem. — Bilorv ( talk ) 17:49, 31 August 2021 (UTC)
 * 3) I think this is probably right, based upon my perception of the RFA process as discussed in my comment at Issue B above. The unpredictability of RFA is something that itself probably does not have a solution, though I am hopeful its effects can be mitigated by finding a way to improve the perception of the RFA process itself among potential candidates. <b style="color:#C75300;">Tyrol5</b> ▸ [talk]  21:23, 31 August 2021 (UTC)
 * 4) I agree this is an issue, but I'm not sure we can do very much about it, other than telling candidates to expect the unexpected and know if they have any skeletons in their closet. Trainsandotherthings (talk) 00:35, 3 September 2021 (UTC)
 * 5) True for me. I thought I was going to be quizzed about the difference between a block and a ban, copyright law and username violations and instead I was bombarded with sarcastic talk page comments I made as a new editor. The fact that anyone would take these remarks so seriously as to search out these diffs and present them two years later was something I was completely unprepared for. I know that I had forgotten about them and matured as an editor. And if I had to go through a second RfA today, after 8 years of editing, I have no idea what someone would base their Oppose vote on now. I know some candidates just sail through their RfA but for others of us, it's a wild card. You just don't know what edits from your past someone will base their objection on. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 03:47, 4 September 2021 (UTC)
 * 6) *nod*. A set of "submit us feedback on these specific characteristics" questions would help prepare everyone for the appropriate assessment of a candidate, such as "Quality content", "Admin areas edited", "Temperment", and so forth. --Izno (talk) 06:01, 14 September 2021 (UTC)
 * 7) I agree in principle that RfA's are by nature somewhat unpredictable. Most things and issues that could have the potential to affect an RfA are found out through Optional RfA candidate poll. However, RfA is a community review process and this is to be expected as there are still unknown's that can happen in an RfA. TheGeneralUser (talk) 03:32, 29 September 2021 (UTC)

T. Oppose (disagree)

 * 1) Expect the unexpected seems like it applies to every area of controversy on Wikipedia. I just don't see how this identifies a clear problem or a path to a possible solution, unless the intention of this issue is to throw our hands up and say "nothing we can do," which I disagree with. (I'm assuming it's not to say that we need clearer rules for what's considered a valid reason to oppose, since that's already covered above.) &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 12:20, 4 September 2021 (UTC)
 * 2) No. Often issues come up that a nominee should have planned for and come up with a solution. Lack of content? Write some audited content. Participate in some admin-related debate. Come up with a plausible way you'll use the tools. Plan to make up with people you've had disagreements with. This is why having an experienced nominator is important.  I've only nominated editors I was confident of that no foreseeable issue would crop up like this. Cas Liber (talk · contribs) 12:29, 7 September 2021 (UTC)
 * 3) What Casliber said. I can't think of an RFA in recent years that wasn't relatively predictable, or that did poorly because of something that could not have been reasonably anticipated. Risker (talk) 00:16, 8 September 2021 (UTC)
 * 4) Mostly what Cas said. Nominators generally do a good job of talking this part through. Best, KevinL ( aka L235 · t · c) 18:58, 8 September 2021 (UTC)
 * 5) Any candidate who has read WP:RFAADVICE and gone through ORCP, and found a nominator, will learn soon  enough what  their chances are. There can and will be some exceptions, but  nowadays most  RfA are fairly  predictable before they  are transcluded. Kudpung กุดผึ้ง (talk) 15:23, 19 September 2021 (UTC)
 * 6) Basically per Rhododendrites. A few more points: Even accepting arguendo that RfA is unpredictable (and Casliber et al. have made a reasonable argument that it generally isn't), it would be more of a symptom than a problem. And more broadly, I'm not convinced it's a bad thing: unless we introduce fixed automatic RfA criteria (a perennial bad idea), there will always be some degree of discretion that can't be distilled into objective and predictable standards. Most of the uncertainty can be resolved by talking to your nominator, reading old RfAs and advice pages, introspection, etc.; the amount that remains is implicit in any pseudo-democratic system. Extraordinary Writ (talk) 18:42, 24 September 2021 (UTC)

T. Comments

 * I agree that it's a factor, I disagree that we can do anything about it. – John M Wolfson (talk • contribs) 12:44, 31 August 2021 (UTC)
 * We could decide to evaluate candidates only against a previously agreed set of criteria, not come up with new ones during a running RfA. —Kusma (talk) 15:52, 31 August 2021 (UTC)
 * We could, but I think that would be stupid per my comments on part N, which this seems otherwise a bit similar to. – John M Wolfson (talk • contribs) 15:53, 31 August 2021 (UTC)
 * If there were a set of things that were very obviously off-the-table when reviewing candidates, then I think that'd be helpful at least. Like, if people didn't mention things that happen off-wiki while evaluating candidates, as a general rule, that'd improve the tone of RFA at least. &#8211;<span style="font-family:CG Times, times"> MJL &thinsp;‐Talk‐☖ 01:39, 3 September 2021 (UTC)
 * Such is life. Unless we convert adminship to an automatically granted permission with fixed, open criteria (which we shouldn't due to the abuse/gaming potential), this won't change; as long as humans are involved, there will be mistakes and all kinds of other surprises. ~ ToBeFree (talk) 17:55, 31 August 2021 (UTC)

U. RfA should not be the only road to adminship
Right now, RfA is the only way we can get new admins, but it doesn't have to be.

U. Support (agree)

 * 1) There are several other possibilities for mechanisms to get new admins (using the verb "get" to keep it vague for now). There's also the possibility of setting up recall/desysop procedures which would only apply to people who received their rights through that alternative mechanism. Perhaps more than all of the rest, this seems like it's worth an experiment at least. &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 18:23, 1 September 2021 (UTC)
 * 2) It's time for alternatives, because the status quo is clearly inadequate. GABgab 20:10, 1 September 2021 (UTC)
 * Another idea: Bureaucrats are authorized to appoint a given number of admins, who are then subject to an RFA (details tbd) following a certain probationary period. GABgab 14:36, 2 September 2021 (UTC)
 * 1) I do agree with pushing for an alternative, as the public opinion of RfA is so low, and has been for so long. A completely new process - which meets the WMF scrutiny requirements and which has community buy in - is like to gain potential good candidates who are refusing to run RfA simply because it is RfA. <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 11:03, 2 September 2021 (UTC)
 * 2) I wasn't sure about agreeing with this point simply because I couldn't see an alternative path to the RfA that would be sufficient for the community, but Rhododendrites' suggestion below is quite interesting, and I'm sure other editors will come up with more, so I feel this is worth pursuing. Isabelle 🔔 16:11, 2 September 2021 (UTC)
 * 3) Absent an explicit statement from WMF Legal that the current, highly idiosyncratic RfA system is the only acceptable path, this is an idea we should explore. It's a vague question, and I'm not in love with any of the suggestions offered so far. Our fundamental task here is to build an encyclopedia, and our responsibility is to create the best possible social support system for doing that, not to worship an established practice simply for being established. XOR&#39;easter (talk) 18:39, 2 September 2021 (UTC)
 * 4) I agree we should consider this idea. 2001 Wikipedia was a very different place than 2021 Wikipedia is, and if making a bold change might help, I'm all for trying it. I would want any potential RfA replacement to be tested repeatedly before being implemented. Trainsandotherthings (talk) 00:38, 3 September 2021 (UTC)
 * 5) I agree. --John Cline (talk) 01:57, 3 September 2021 (UTC)
 * 6) Worth exploring alternatives. Nurg (talk) 02:26, 3 September 2021 (UTC)
 * 7) Worth thinking about. If we can't reform RfA, add an optional extra process. —Kusma (talk) 11:02, 3 September 2021 (UTC)
 * Strong support now. Adminship isn't a big deal, but RfA is, so adminship shouldn't be connected to RfA. —Kusma (talk) 11:30, 6 September 2021 (UTC)
 * 1) Put everything on the table. ProcrastinatingReader (talk) 11:34, 3 September 2021 (UTC)
 * 2) I've seen multiple workable proposals. —valereee (talk) 18:31, 3 September 2021 (UTC)
 * 3) Agree. Other options should be explored. I hope editors are willing to give up the power they have at RFA to decide new admins in order to benefit the project. Z1720 (talk) 20:57, 3 September 2021 (UTC)
 * 4) Agree. Just because current admins had to endure an RfA doesn't mean that every future administrator should have to run the gauntlet. There must have been proposals, based on merit or achievement and less on the RfA beauty contest, that been put forward. They should be revisited. Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 03:50, 4 September 2021 (UTC)
 * 5) Agree. Another thing we can consider is automatically handing out the bit to stewards. MER-C 12:57, 4 September 2021 (UTC)
 * 6) Current RfA is all or nothing. Of course, few resilient candidates will try 2nd or 3rd time, but would be good if we had a alternative method to sysop. Carlosguitar (Yes Executor?) 14:12, 4 September 2021 (UTC)
 * 7) Worth exploring alternatives. Levivich 16:11, 4 September 2021 (UTC)
 * 8) Not without community scrutiny and consensus, but RfA is not the only way to get there. Seraphimblade Talk to me 17:54, 4 September 2021 (UTC)
 * 9) At least in principle... Vanamonde (Talk) 12:42, 5 September 2021 (UTC)
 * 10) Yeah, worth exploring the alternatives. I've sometimes wondered if we could just have an annual election for commissioners who would recruit and appoint, then supervise, support and develop, and if needful dismiss, sysops.—<b style="font-family: Verdana; color: Maroon;">S Marshall</b> T/C 19:09, 5 September 2021 (UTC)
 * So basically similar to how CU and OS are currently appointed by ArbCom? Regards So  Why  19:37, 5 September 2021 (UTC)
 * That's approximately what was in my mind, but I'm not sure whether Arbcom actively supervise and develop CU and OS. The details should probably wait for phase 2.—<b style="font-family: Verdana; color: Maroon;">S Marshall</b> T/C 19:53, 5 September 2021 (UTC)
 * I don't know what you mean by "develop" but ArbCom is responsible for supervising CU and OS as per Arbitration Committee/CheckUser and Oversight. Regards So  Why  08:20, 6 September 2021 (UTC)
 * I mean coach, mentor, and encourage reflective practice.—<b style="font-family: Verdana; color: Maroon;">S Marshall</b> T/C 15:41, 6 September 2021 (UTC)
 * 1) Makes sense to at least explore other options. -- Calidum  15:29, 8 September 2021 (UTC)
 * 2) An alternative process has the potential to work better than the current process, or at least increase the number of admins overall.Jackattack1597 (talk) 22:33, 11 September 2021 (UTC)
 * 3) no reason why we can't e.g. have an arbcom style process also. Galobtter (pingó mió) 19:57, 12 September 2021 (UTC)
 * 4) Some years ago, I was involved in drafting a proposal to establish a new commission that would appoint editors as administrators following a thorough review. The proposal, Administrative Standards Commission, sadly failed, but I still strongly believe an alternative process should be established.  RGloucester  — ☎ 22:46, 12 September 2021 (UTC)
 * 5) I struggle to think of a different system that doesn't introduce one of the known problems of RFA (for whatever set that may be as established above), but I think we should leave the door open for wacky solutions like "send an email to AdminsRUs@bureaucrats.com to be considered" anyway. --Izno (talk) 05:59, 14 September 2021 (UTC)
 * 6) So many of the major problems above (e.g. E, F, I) are almost implicit in the RfA system, so it's certainly worth talking about alternatives. (I agree with Seraphimblade above that any feasible solution would still require some form of "community scrutiny and consensus", lest we end up with the even worse problems of unaccountability and contempt for the admin corps.) Extraordinary Writ (talk) 18:40, 18 September 2021 (UTC)
 * 7) Sure - we could give, I dunno, crats, more discretion to hand out sysop flags as long as they were given the same discretion to remove them from people who don't go through a community RfA, in case anything goes wrong. – filelakeshoe (t / c) 🐱 08:54, 22 September 2021 (UTC)
 * 8) Viable alternative. I like it. Buffs (talk) 20:05, 22 September 2021 (UTC)
 * 9) Alternative systems should be considered, though I don't know of any that would work. — Danre98 ( talk ^ contribs ) 22:42, 23 September 2021 (UTC)
 * 10) I agree we should explore alternative methods of gaining administrators. We have non-election methods for checkusers and oversighters, and those are positions of even greater sensitivity in some regards.- gadfium 20:31, 28 September 2021 (UTC)
 * 11) True, there can be more than one way. Open to reviewing alternative proposals and solutions. TheGeneralUser (talk) 22:57, 29 September 2021 (UTC)

U. Oppose (disagree)

 * 1) WMF Legal requires that adminiship be granted only after community review, which would essentially "devolve" back into RfA IMO. – John M Wolfson (talk • contribs) 19:08, 1 September 2021 (UTC)
 * I am neutral on this proposal (as on others because I am trying to facilitate this process) so this is not an endorsement, but I think there is a large gap between an acceptable Community Review for the foundation and the "autogrant RfA based on hardcoded criteria" which is what that thread proposed. So if this idea gets consensus I think there could be ideas which would pass muster with Legal that aren't RfA exactly as we know it today. Best, Barkeep49 (talk) 19:16, 1 September 2021 (UTC)
 * what barkeep said. The last time I saw this perspective articulated it was that a "community process" would be required. There are lots of ways we could document mechanisms to produce new admins that would constitute "community process". Just for example, having the elected bureaucrat team be able to appoint X admins which would then be subject to recall if they lose community trust (again, just an example). My sense is they really just want to make sure some judgment/oversight is involved rather than be something that's more automatic or superficially considered. Perhaps or someone else from WMF Legal could clarify what exactly would be required? &mdash;  Rhododendrites  <sup style="font-size:80%;">talk \\ 19:23, 1 September 2021 (UTC)
 * @Rhododendrites, I don't recommend pinging the general counsel. If you really believe that you need someone from that department to comment at this stage (although it seems obvious to me, and apparently to most editors, that there is nothing inherent in the concept of community review that requires that community process to be carried out through single process?), then I would recommend sending an e-mail message to legal@undefinedwikimedia.org instead of pinging the department head personally. Whatamidoing (WMF) (talk) 18:25, 10 September 2021 (UTC)
 * wmf Legal has veto power here, and, at least as far as I've seen, hasn't been entirely clear with what's required (based on this and past discussions). I think it makes sense for the entity who makes the rules to give us parameters first (whether here or before a more specific subsequent rfc) rather than require lots of volunteers to spend lots of time working on something based on ambiguous guidance and then cross our fingers it doesn't get rejected.
 * on the subject of pinging, this is a bit tangential but it occurs to me there's a frequent issue of "we need wmf input to move forward. Who do we contact?" followed by some guesses, often including someone who doesn't work there anymore, etc. Sometimes not knowing means pinging the dept head which, sure, isn't an ideal default. What if we had an exception to our role account rule for certain wmf staff contacts to allow easier involvement of wmf staff into community discussions? An account to ping for certain things (understanding some things do need to be off-wiki) which that dept could handle or assign as they see fit. In this case whatever statement they make would need to be on-wiki (if not here, on meta), rather than by in a response to an email to me personally, so email would only be to alert them to this discussion. Really, that should be possible with a ping, no? &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 19:18, 10 September 2021 (UTC)
 * You can ping me to any discussion, but if you believe that you need Legal's attention, you will find it more useful to send an e-mail message to them. That's the workflow that seems to be the most effective way to reach that department, and I think we can expect that address to work for many more years, no matter who runs the department.
 * (The English Wikipedia has used at least three processes to appoint admins over the years, and at least the first two were used simultaneously. Consequently, I would be surprised to hear them say that having more than one process was inherently problematic now.) Whatamidoing (WMF) (talk) 17:00, 11 September 2021 (UTC)
 * 1) Even assuming some form of non-full community, but non-criteria, methododology (perhaps as Rhod suggests), I'd oppose this. In many failed and close RfAs, issues are found by individual editors (rather than being directly obvious to a large fraction of the 200+ participants). They can then be considered by the fuller group. Shifting to a far smaller group ups the chances of missed issues. More importantly, I feel that having direct, rather than indirect, community assessment of admins is beneficial and would be very reticent to do away with it. Nosebagbear (talk) 21:06, 1 September 2021 (UTC)
 * [Merged & tweaked from below] So, it's not an issue that the proposal has no specifics - that's a phase 2 area. But I believe it can be opposed on equally general principles, as well as the most specific argument I raise above. I believe that direct Community selection of admins is the only way to ensure buy-in and acceptance of admins. As we see in most RfAs, we have a vast array of what we consider acceptable standards for an admin - even putting aside the extreme outliers, it's a very broad range. Any alternate method would inherently wash array the chance to both adequately present and actually have power behind differing opinions to the narrower selection of any alternate means (whether group, criteria, or so on).
 * You're not wrong, but can we consider something? How many of those deep-dive-diffs that nontrivially shift the balance or momentum of a nomination are really evidence that the person would not have been a good admin. It seems at least worth considering that those cases when nobody sees any problem until one person finds a suspect diff, leading to a bunch of other people jumping ship, may fall into one or more of the "issues" covered by other sections on this page. I say that as someone who has both been swayed by such diffs and who has dug up concerning issues which shifted some opinion in RfAs myself. In the moment, it's hard to justify supporting after seeing some of those, but in hindsight, there are at least some cases when I'm not really so sure I/we did ourselves a service by trying to make the candidacy about those diffs which are sufficiently not part of that person's day-to-day presence to be missed by others. Not sure. I guess what I'm saying is sometimes we may get wrapped up in those unburied diffs, and maybe if people we entrust people to do a careful evaluation of a candidate and they don't find the buried problem, then it's probably not going to result in a terrible admin? Alternatively, scrap the rest of what I'm saying and we can just say "after the bureaucrats announce their appointments, there's a one-week period during which the community can highlight issues the crats might've missed, and it's up to the crats whether any of them are deal-breakers." &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 21:35, 1 September 2021 (UTC)
 * 1) This might fly; however, as someone who has failed at RfA, I've come to respect the process for its usually well-attended and significant community vetting. Even though I thought I would make a really good admin, I can't help but think that the community's drive to disagree with me and to exclude me was the correct decision. For my part, it seems to be far better to have a system that might exclude potentially good admins than to have one that might let too many potentially bad candidates through. With all its present challenges, our system of community vetting at RfA is anything but lenient! I suppose my main concern here would be to have a system where admins are first appointed pending community review, only to have that system evolve into making it extremely difficult for the community to say "no" due to pressure "from above".  P.I. Ellsworth &numsp;- ed.  put'r there 21:22, 3 September 2021 (UTC)
 * The easiest way to prevent bad admins from being appointed is to not appoint any new admins at all, which is pretty close to what we have right now. —Kusma (talk) 21:26, 3 September 2021 (UTC)
 * lol, have to admit that ever since my RfA I've had this recurring thought of Jimbo finding me and saying he wanted me to be an admin, so just shut up, take the job and be quiet about it. So I would continue my gnomish ways and not make any sysop waves for a good while. It's a weird thought for me when you consider that my desire to be an admin has always been just a bit outweighed by my desire not to be one. ;>)  P.I. Ellsworth &numsp;- ed.  put'r there 21:41, 3 September 2021 (UTC)
 * Be kinda like Einstein writing me asking for help with an equation or Harrison Ford calling me up asking if I want to be in one of his films. How wild izzat!  P.I. Ellsworth &numsp;- ed.  put'r there 14:53, 4 September 2021 (UTC)
 * 1) This is a non-starter. The Foundation requires RfA or a process more or less identical to it for granting certain permissions included in the admin toolset. We could discuss unbundling some permissions or rebundling the sensitive permissions in a way that wouldn't require RfA, but we will not get admins in the sense of what an admin is today without RfA. Ivanvector (Talk/Edits) 19:50, 4 September 2021 (UTC)
 * Our new process would just need to involve the same level of scrutiny as RfA. Which only says "must involve talking about the candidate" if you look at RfAs from 2004. —Kusma (talk) 07:03, 5 September 2021 (UTC)
 * 1) I can't see this as the right solution. Stifle (talk) 12:26, 6 September 2021 (UTC)
 * 2) This won't get off the ground Cas Liber (talk · contribs) 12:57, 7 September 2021 (UTC)
 * 3) So one dangerous, winding, accident-prone mountain road isn't enough, and you want to build another one? No thanks. At least we have lots of experience driving on the road we have. If we build a safer road everyone will abandon the more dangerous road, and there may be a traffic pileup on the new road. We've developed more roads to desysopping (bureaucrats built a new road); I'm not convinced that's a good thing. – wbm1058 (talk) 15:00, 7 September 2021 (UTC)
 * Your analogy is not apt. You would rather see people driving on a dangerous road where some end up driving off the cliff, because establishing a safer road might cause everyone to flock to that and cause a traffic buildup? Your analogy implies that a traffic clog is a bigger problem than people flying off a cliff. ProcrastinatingReader (talk) 15:07, 7 September 2021 (UTC)
 * Well, consensus seems to be that traffic clogs at requests for admins to perform tasks are a bigger problem than the large volume of editors driving off cliffs (filter out all the proxies and there's still a lot). Look at the top tier of the most active editors and you'll see several who have driven over cliffs. Some have survived albeit with disabilities. – wbm1058 (talk) 15:43, 7 September 2021 (UTC)
 * 1) There's no alternative suggestions here to which RFA could be compared. Until someone develops those alternatives, and is willing to advocate for them, and they can be demonstrably equivalent to RFA for the purposes of accessing deleted information (the WMF standard, which applies to 100% of projects), this is just pie in the sky. Before worrying about this, I'd be more interested in seeing the results of the French Wikipedia method of appointing CU/OS without direct intervention by an arbitration committee. Risker (talk) 00:14, 8 September 2021 (UTC)
 * is that w:fr:Wikipédia:Comité_de_nomination? ProcrastinatingReader (talk) 18:27, 8 September 2021 (UTC)
 * Yes. Risker (talk) 18:51, 8 September 2021 (UTC)
 * 1) I want to see a good alternative before I support this as a general idea, because I can't think of a good alternative myself and I've been trying for a few years. Adminship isn't just some buttons these days. We can't deny that adminship comes with a substantial grant of social capital and influence, and far-more-than-technical authority (e.g. DS authority). As long as that's the case, a pre-adminship community review seems like the only acceptable system that I can think up. I would love to be convinced otherwise, though. KevinL ( aka L235 · t · c) 19:01, 8 September 2021 (UTC)
 * 2) The supporters are very low on specifics. If not asking the community, how else would we choose admins? By panel? Wikipedia just loves panels, look at ArbCom! By automatic criteria? Sure fire way to get LTA's as admins. If this is to be concurrent with RfA, that would be untenable. One would have to inherently be seen as easier than the other, and folks would then always use the easier option. If it were to replace RfA, why not just state as much? As long as admins have the power to govern the community, there must be consent of the governed, and thus an RfA like process. CaptainEek  Edits Ho Cap'n!⚓ 19:55, 8 September 2021 (UTC)
 * We're supposed to be low on specifics. This is the identifying issues RfC, not the find solutions one. I see no reason why two different processes mean that one of them is easier for everyone. That will depend on the specifics. —Kusma (talk) 20:08, 8 September 2021 (UTC)
 * Specifics for this topic, at this stage of discussion, is a trap. "Agree" only means "it's not impossible that there's an option other than RfA to get new admins" and/or "it's worth talking more about this as a way to get more admins." There's no specific proposal here otherwise. Nothing is going to be enacted if this finds consensus because it's only a proposal to consider this among other avenues to fix our admin shortage, such as it is. &mdash; Rhododendrites  <sup style="font-size:80%;">talk \\ 20:28, 8 September 2021 (UTC)
 * (both Kusma and Rhododendrites's points). I understood (based on Barkeep's overview) this is a survey to see the community finds to be real issues, and specifics on how to solve these issues to follow in between phase 1 and phase 2. It may well be that several of these issues have no real solutions. For example, the corrosive atmosphere issue (most supported ATM) seems to have no substantial solution bar fundamental change. ProcrastinatingReader (talk) 01:16, 9 September 2021 (UTC)
 * Assuming that statement A does not get consensus - because if it did that would end the process as the community's saying there's not actually a problem - the idea is that there will be a short period between the close of this and the opening of phase 2 where people can brainstorm and refine their proposals so that they are ready to go when Phase 2 opens. We are at the "conceptually is this a problem?" phase. So conceptually is it true that RfA should not be the only path to adminship? Conceptually (to pick another idea at random) is it true that "lack of agreed upon criteria" is a problem? It is possible, indeed I expect it, that there will be consensus that something is true but no consensus on how to fix it and so no change will be made. The intent of doing it this way is that when we get to phase 2 we will avoid opposes based on "this isn't actually a problem" because there will already be consensus it is, instead we will get supports/opposes based on the perceived merits of the idea. Best, Barkeep49 (talk) 15:21, 9 September 2021 (UTC)
 * So, it's not an issue that the proposal has no specifics - that's a phase 2 area. But I believe it can be opposed on equally general principles. I believe that direct Community selection of admins is the only way to ensure buy-in and acceptance of admins. As we see in most RfAs, we have a vast array of what we consider acceptable standards for an admin - even putting aside the extreme outliers, it's a very broad range. Any alternate method would inherently wash array the chance to both adequately present and actually have power behind differing opinions to the narrower selection of any alternate means (whether group, criteria, or so on). Nosebagbear (talk) 20:31, 8 September 2021 (UTC) [Merged with my first bullet above]
 * Could you please combine this or move this under your previous comment? As it is, it makes it look like your vote counts twice here, as both this comment and your previous comment are numbered. Jackattack1597 (talk) 22:30, 11 September 2021 (UTC)
 * 1) Every candidate should go  through  some kind  of community review whatever changes are made to  the  actual  system or voting  process. Kudpung กุดผึ้ง (talk) 15:26, 19 September 2021 (UTC)
 * 2) I don't want or like cabals. A more centralized process will make connections and "knowing the right people" a far greater part of the RfA process than they currently are now. People will schmooze and a self perpetuating network of admins will likely evolve, much like how appointed positions outside of Wikipedia are far more likely to have "outsiders". It's also harmful from a diversity standpoint, given that when these unofficial networks develop they often tend to include people who are like them and exclude people who aren't like them. This happens a lot with "network-based recruitment" in industries, which generally tend to increase rather than decrease segregation. I believe we should reduce opportunities for this rather than increase them, and so far this proposal generally appears to promote a more "closed" and discretionary system for admins.Rhododendrites raised the interesting point that "there's also the possibility of setting up recall/desysop procedures which would only apply to people who received their rights through that alternative mechanism". This would entail the creation of a "two-tier" system of adminship, with the higher-tier inevitably being the admins who are not subject to this process. Chess (talk) (please use&#32; on reply) 21:14, 22 September 2021 (UTC)
 * 3) I don't see how any other road to adminship would help RfA at all. 🐔 Chicdat <sup style="font-family:Times New Roman">Bawk to me!  11:14, 28 September 2021 (UTC)

U. Comments

 * One crazy idea would be to copy the ArbCom election format: Hold elections at a given period, allow candidates and questions, have a secret ballot with checkuser tests, and then just promote the 20 people with the highest vote percentage, as long as it is above 50%. Crazy low you say? It's enough for ArbCom, and the WMF is ok with giving these people access to checkuser/oversight. We need a procedure that gives us a reasonable number of new admins every year, and RfA isn't that procedure. —Kusma (talk) 11:10, 3 September 2021 (UTC)
 * That would cap our annual admin intake at 20, which is roughly what the current system outputs (see Requests for adminship by year). It might be slightly less unpleasant but I'm not convinced it would fix any other issues. ProcrastinatingReader (talk) 11:40, 3 September 2021 (UTC)
 * The "20" isn't set in stone, and there could be several rounds of this per year. Also, my suggestion is 20 guaranteed admins plus whatever RfA comes up with. —Kusma (talk) 12:10, 3 September 2021 (UTC)
 * I don't like not even discussing things because WMF Legal might possibly shoot them down. As a community, we should state what we need and then adapt it should we accidentally hit a legal requirement. (If there is a legal requirement: I still haven't been able to find out whether there is anything wrong with the way we deal with IP edits, as WMF Legal refuse to tell anyone). —Kusma (talk) 07:07, 5 September 2021 (UTC)
 * If this idea gets consensus I have some level of optimism I can get a general statement from Legal about to inform thinking for phase 2. Best, Barkeep49 (talk) 14:48, 5 September 2021 (UTC)
 * I'd agree with that (that Barkeep individually could get some comment from Legal on the topic - I have confidence in the relationship there). <b style="text-shadow:0 -1px #DDD,1px 0 #DDD,0 1px #DDD,-1px 0 #DDD; color:#000;">Worm</b>TT(<b style="color:#060;">talk</b>) 15:37, 6 September 2021 (UTC)
 * If this idea gets consensus - while it's leaning support, several of the opposes seem to be based on the assumption that it will not be allowed. Since it's also all but certain to come up in the future, too, perhaps you could seek out that statement either way? &mdash;  Rhododendrites  <sup style="font-size:80%;">talk \\ 13:58, 7 September 2021 (UTC)
 * First it will be easier to get the statement if there's already a consensus I can point to. That is it's not Barkeep49 asking, it's enwiki. Second we are at the stage where we identify issues. If this is an issue seen by the enwiki community even if it turns out we cannot do one of the ideas sketched out here, it may inform people's willingness to consider certain kinds of changes within the RfA structure. Best, Barkeep49 (talk) 14:18, 7 September 2021 (UTC)

V. Formal moderation of RfA
RfA discussion would be significantly less corrosive if we installed moderators with formalized duties, rights, and responsibilities.

X. Support (agree)

 * 1) Having followed RFA and Arbcom matters for a very long time, it is my observation that the most serious problem is one of accuracy. There are successful RFAs for admins who do poor quality admin work later.  There are good people who are turned down at RFA or who are not approved until near the end of their time of significant contributions to Wikipedia (which is finite for us all, alas).  This, I believe, is the most important problem for us to solve.  Uninvited Company 20:24, 28 September 2021 (UTC)

X. Oppose (disagree)

 * 1) Oppose both procedurally, since you posted this way too late (less than 24 hours before closing), as well as substantively, since the problem is largely too few RfAs rather than, as this implies, too many.  – John M Wolfson (talk • contribs) 20:32, 28 September 2021 (UTC)
 * 2) Oppose, for the same reasons as John M Wolfson. Trainsandotherthings (talk) 22:07, 28 September 2021 (UTC)

Y. Support (agree)

 * 1) Standards for the amount, kind, and quality of writing for new admins are ever increasing. This has little to do with the realities of most admin work beyond ensuring a sufficient respect for the difficulties and importance of good article work.  We need admins who are compassionate, evenhanded, careful, respectful, trustworthy, and slow to anger -- regardless of how many articles they've been able to get to FA or GA status.  Standards are weak and fairly static in this area, with too much willingness to overlook poor behavior and bad judgement.  Uninvited Company 20:24, 28 September 2021 (UTC)

Y. Oppose (disagree)

 * 1) I don't know if that actually affects many RfAs, and if so whether that's a bad thing. Even people with virtually no content (which is bad, IMO) pass just above the 75% mark, and the people who want an FA(!) or higher from candidates generally get brushed off.  – John M Wolfson (talk • contribs) 20:32, 28 September 2021 (UTC)

General Discussion
Please consider posting any general discussion to the talk page.