The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
Summary of results | ||||
---|---|---|---|---|
Question | Support | Oppose | % support | Result |
#A1: Advertise RfAs with a site banner | 0 | 35 | 0 | Failed |
#A2: Advertise RfAs with a watchlist notice | 85 | 12 | 88 | Passed |
#A3: Advertise RfAs on WP:CENT | 41 | 7 | 85 | Passed |
#B1: Disallow threaded discussion on the main RfA page | 16 | 46 | 26 | Failed |
#B2: Limit the number of questions...by any individual editor | 82 | 24 | 77 | Passed (with conditions) |
#B3: Limit the number of questions...of any given candidate | 1 | 34 | 3 | Failed |
#B4: Clerking at RfA | – | – | – | Separate RfC |
#C1: Expand discretionary range to 65% | 74 | 30 | 71 | Passed |
#C2: Expand discretionary range to 60% | 21 | 44 | 32 | Failed |
#C3: Expand discretionary range to 50%+1 | 0 | 27 | 0 | Failed |
#C4: Abolish the discretionary range completely | 0 | 27 | 0 | Failed |
#D1: Upper limits on opposition | 5 | 36 | 12 | Failed |
Requests for adminship (or RfA) is the process by which administrators are selected by the community, (theoretically) via the process of consensus. However, the RfA process has been subject to a great deal of discussion over the past few years, and there is general agreement among the community that the process has problems. Therefore, many attempts have been made to fix the process, including (but not limited to) WP:RFA2011 and WP:RFA2013. In 2014 and 2015 combined, there were four months with no promotions (such months were completely nonexistent between 2003–2011; there was one in 2012, and none in 2013). In a recent RfC, there was over 75% support for the notion that we are not producing enough admins. Therefore, a new project was recently started: WP:RFA2015. The project is divided into three phases. The first phase consisted of an RfC (October 15, 2015–November 15, 2015), in which we successfully attempted to identify the actual problems with RfA. The following concepts passed the Phase I RfC: (1) RfA needs more participants; (2) The load on admins should be eased [perhaps via automated mechanisms]; (3) RfA is a hostile environment; (4) The discretionary range [the range of support in which bureaucrats can use discretion in closing RfAs] is too narrow; (5) The standards at RfA are too high; (6) We need active clerking at RfA; (7) The standards at RfA should be defined; (8) It is too difficult to remove admins [the discussion here was actually very close, so it was left as "discretionary", or optional, for the next phases]. Proposals (2) and (8) are not covered in this RfC, since they are not directly related to RfA. Proposal (6) will be handled in a separate RfC, which is currently on hold until this one ends.
This RfC is the successor to the Phase I RfC. The goal of this RfC is to reach actual solutions for the problems identified in Phase I. Participants will examine the proposals set forth on this page and indicate whether they support or oppose those proposals in the proper sections. Comments on specific proposals should be placed in the comments section for that proposal, while general comments about the RfC should be placed the talk page. Since this proposal may have substantial effects on the RfA process, it will be as widely advertised as possible. Thirty days after the opening of the RfC, it will be closed. The closer will determine which proposals attained consensus according to the process described at Wikipedia:Consensus#Determining consensus. Problems for which a solution is not found in this RfC will be further discussed at a later point.
The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
A1: Advertise RfAs with a site banner
|
---|
A1: Advertise RfAs with a site bannerOne option to increase participation in RfAs is to display a site banner (to logged-in users only) advertising ongoing RfAs. Support A1
Oppose A1
Comments on A1
|
Another option is to post a notice announcing current RfAs on MediaWiki:Watchlist-details, which will display the notice on all watchlists.
Finally, we could also advertise ongoing RfAs at Template:Centralized discussion. It is a heavily watched page, with over 400 watchers.
{{User:Cyberpower678/RfX Report}}
, that carries an automatically updated list of RfAs. If the intention is to put RfA notices on a template which is transcluded in major fora, we should just add this one instead.
BethNaught (
talk) 21:00, 29 November 2015 (UTC)
If this proposal were implemented, all threaded discussions which contain more than one comment will be moved to the talk page. The first reply will be left, so that users who see the vote and the subsequent reply will know the initial topic of the transplanted discussion.
"Discussion is good and moving the discussion somewhere else achieves nothing". I couldn't disagree more, sometimes discussion is good, sometimes it's a complete waste of time. A massive three-page rant on ANI between two IPs on everything to do with the Israeli-Palestinian conflict is a "discussion" for the pair of them, but a colossal pain in the backside for everyone else. Similarly, anyone blocking Eric Corbett will likely receive a "discussion", but probably one that many would prefer to avoid like the plague. Ritchie333 (talk) (cont) 12:16, 18 December 2015 (UTC)
A primary source of stress at RfA may very well be the often large amount of questions asked. Therefore, the community might wish to consider limiting the number of questions that any given editor can ask of a candidate. If you support this proposal, please specify in your !vote what you believe the limit should be. (Example: "Support a limit of [x] questions.")
I want to suport this. The questioning has gotten absolutely absurd. Apparently there si now a sizable contingent of users who believe they have crafted the perfect set of questions that must be posted at every single RFA. I find this extremely obnoxious and a net negative to the already lousy environment at RFA. When I ask an RFA question, it is specific to that candidate, not just something I dreamed up that I imagine will be pertinent at every single RFA. That being said, I worry that this could limit follow-up on personalized questions if the initial answers seemed to miss the point or were otherwise insufficient. I'll give some more thought I guess. Beeblebrox ( talk) 21:24, 29 November 2015 (UTC)
Agree that questioning has sometimes been excessive, but if this restriction goes through we'll also have to devise a regulation to rule out "cluster bombs" where many different questions are bundled into one as parts (a), (b), (c), (d), (e)... : Noyster (talk), 22:55, 30 November 2015 (UTC)
I feel like supporting this but I am hanging on the fence for several reasons already brought up. Besides for the reverse effect, my biggest concern is that the follow-up questions will be limited. I am willing to support two questions at most, but think that there should still be unlimited follow-up questions. Gug01 ( talk) 20:12, 4 December 2015 (UTC)
It's been suggested that I should consider standing for RFA, but (1) I (like probably most editors) would get some opposes (e.g. I'd fail "You need at least a couple of GAs, and preferably an FA or two..." - I specialise in cleanup rather than adding content) so (however well I answered the questions) I couldn't be sure of passing. (2) Answering the questions could take some hours (say 20 questions at 30 minutes each). (3) I can't be sure that I'll have the time available in the next week (real-life happens). (4) Those hours spent answering questions (e.g. doing research into areas of wp I've no current intention to work in or considering all the possibilities of a hypothetical scenario) could be spent doing something else in wp. Thus, I choose to do things from my (very long off-wiki) Wikipedia to-do list rather than stand for adminship. There may be hundreds of experienced Wikipedians in a similar position. DexDor (talk) 13:26, 5 December 2015 (UTC)
The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
B3: Limit the total number of questions that may be asked of any given candidate
|
---|
As an alternative to (or in addition to) B2, we could also limit the total number of questions that can be asked of any given candidate. For example, if we set the limit at 15 total questions, any questions beyond the 15th question would be disallowed. If you support this proposal, please indicate in your !vote what you believe the limit should be. (Example: "Support a limit of [x] questions.") Support B3
Oppose B3
Comments on B3
|
After some consideration, it has been decided that the specifics of clerking should be handled in a
separate RfC.
Biblio
worm 01:07, 4 December 2015 (UTC)
|
---|
Note: For the purposes of these proposals, a "clerk" is a user who maintains order at RfA. Per Wikipedia:2015 administrator election reform/Phase I/RfC#M: Active clerking at RFA, clerks are already authorized to carry out the following tasks:
Depending upon which other proposals pass in this RfC, their tasks may also include moving all discussion to the talk page and enforcing the limit on the number of questions. In this section, we will discuss who should carry out the clerking tasks. Comments on B4
B4.1: Bureaucrats should be the clerksWe could exclusively authorize bureaucrats to perform the clerking tasks. Support B4.1
Oppose B4.1
Comments on B4.1
B4.2: Clerks should be appointed and supervised by bureaucratsAs an alternative to B4.1, we could require that clerks be appointed by general agreement of the bureaucrats (perhaps in a discussion on WP:BN). Once appointed, clerks would also be supervised by bureaucrats, who would have the authority to override any action a clerk performs. Support B4.2
Oppose B4.2
Comments on B4.2"Supervised by" maybe. "Appointed by" no. Crats are only supposed to take action when there is a pre-existing policy or consensus that compels them to do so. Making them the ones to select who clerks RFA would change what type of persons we expect our crats to be, so this is a non-stater as far as I am concerned. Beeblebrox ( talk) 21:26, 29 November 2015 (UTC) As per Beeblebrox. Perhaps some sort of crat-mentoring program? Only in death does duty end ( talk) 10:49, 30 November 2015 (UTC)
B4.3: Clerks should be an elected bodyAlternatively, we might consider electing RfA clerks. The requirements for being elected need not necessarily be high, and "elections" could take place rather informally on WT:RFA. (Perhaps in a similar manner to the way edit filter managers are chosen.) Support B4.3
Oppose B4.3
Comments on B4.3
B4.4: Any editor should be able to clerkFinally, we could allow any editor to clerk, if they are willing to volunteer to perform the tasks. Support B4.4
Oppose B4.4
Comments on B4.4
B4.5: Any admin should be allowed to clerkAs a slightly less exclusive option than B4.1, we could simply allow any admin to perform clerking duties. Support B4.5
Oppose B4.5
Comments on B4.5
-- Kudpung กุดผึ้ง ( talk) 20:31, 1 December 2015 (UTC)
|
We could expand the discretionary range to 65–75%, making it a 10% range, as opposed to our current 5% range (70–75%).
Consensus at RfA is not determined by surpassing a numerical threshold, but by the strength of rationales presented. Historically, most of those above 75 percent approval pass and most of those below 70 percent fail. However, the actual decision of passing or failing is subject to the bureaucrats' discretion and judgment..." [1] So, the discussion in this section is over whether to change either or both of the figures "75" and "70" in this guidance? Simply changing these numbers does not change what has happened "Historically". Note that a higher figure is given elsewhere: "
However, as an approximate guide, you are likely to pass if you achieve at least 80% support. Nominations which receive less than 70% support are unlikely to be successful, except in exceptional circumstances." [2] "
Most RfA's with a final tally of 80% support or more will close as successful, while those under 70% will generally not pass. There have however been important exceptions, with candidates passing with as low as 61.2%. [3] The 70–80 'grey' zone is subject to the bureaucrat's discretion after taking into account the quality of the arguments made by the !voters, the strength of comments in the 'neutral' section, and after discounting any !votes they consider to be invalid. In extremely close calls, an extension to the 7-day !voting period may be accorded, or a discussion ('crat chat) may take place among the bureaucrats." [4] – Wbm1058 ( talk) 15:32, 1 December 2015 (UTC)
The percentages for the 19 new admins so far in 2015, ranked from highest to lowest:
100 1 |
98 2 |
97 3 |
97 4 |
97 5 |
96 6 |
96 7 |
95 8 |
94 9 |
91 10 |
88 11 |
88 12 |
87 13 |
85 14 |
84 15 |
84 16 |
81 17 |
80 18 |
74 19* |
The percentages for the six RfAs which concluded as "no consensus" (Withdrawn, WP:SNOW and WP:NOTNOW not included):
74 1* |
67 2 |
66 3* |
66 4 |
65 5 |
60 6 |
* = 'crat chat
There isn't evidence in 2015 RfAs to support the narrowing of the "rule of thumb" or "historical" range from 80–70 to 75–70, as no candidates finished with between 75–80% support.
And the defacto discretionary range already appears to go down to 65, or even 60, as everything below that level was "snowed". 'Crat chats were conducted for three of the seven RfAs that closed in this range; perhaps more such chats would have been held if supporters had lobbied harder for them.
Wbm1058 (
talk) 18:08, 16 December 2015 (UTC)
The current guidance:
Consensus at RfA is not determined by surpassing a numerical threshold, but by the strength of rationales presented. Historically, most of those above 75 percent approval pass and most of those below 70 percent fail. However, the actual decision of passing or failing is subject to the bureaucrats' discretion and judgment..."
shall be changed to:
Consensus at RfA is not determined by surpassing a numerical threshold, but by the strength of rationales presented. Historically, most of those above 75 percent approval pass and most of those below 66 percent fail. However, the actual decision of passing or failing is subject to the bureaucrats' discretion and judgment..."
Note that it is still true that most of those between 66–69% fail, but this change makes clear the expectation going forward that failure in this range will not be nearly inevitable, even if it's still an uphill battle.
The current guidance:
However, as an approximate guide, you are likely to pass if you achieve at least 80% support. Nominations which receive less than 70% support are unlikely to be successful, except in exceptional circumstances.
shall be changed to:
However, as an approximate guide, you are likely to pass if you achieve at least 75% support. Nominations which receive less than 66% support are unlikely to be successful, except in exceptional circumstances.
The current advice:
Most RfA's with a final tally of 80% support or more will close as successful, while those under 70% will generally not pass.
shall be changed to:
Most RfA's with a final tally of 75% support or more will close as successful, while those under 66% will generally not pass.
And:
The 70–80 'grey' zone is subject to the bureaucrat's discretion after taking into account the quality of the arguments made by the !voters, the strength of comments in the 'neutral' section, and after discounting any !votes they consider to be invalid. In extremely close calls, an extension to the 7-day !voting period may be accorded, or a discussion ('crat chat) may take place among the bureaucrats."
To:
The 66–75 'grey' zone is subject to the bureaucrat's discretion after taking into account the quality of the arguments made by the !voters, the strength of comments in the 'neutral' section, and after discounting any !votes they consider to be invalid. In extremely close calls, an extension to the 7-day !voting period may be accorded, or a discussion ('crat chat) may take place among the bureaucrats."
But, again, under exceptional circumstances, 76–80, and indeed 60–66 are still subject to the bureaucrat's discretion. They are just much less of a 'grey' zone, and much closer to a black or white zone.
The colors used in the RfX Report are changed as below:
Current | ····· | Proposed | ||||||||||||||||||
100% | 100% | |||||||||||||||||||
99% | 98% | 97% | 96% | 95% | 94% | 93% | 92% | 91% | 90% | 99% | 98% | 97% | 96% | 95% | 94% | 93% | 92% | 91% | 90% | |
89% | 88% | 87% | 86% | 85% | 84% | 83% | 82% | 81% | 80% | 89% | 88% | 87% | 86% | 85% | 84% | 83% | 82% | 81% | 80% | |
79% | 78% | 77% | 76% | 75% | 74% | 73% | 72% | 71% | 70% | 79% | 78% | 77% | 76% | 75% | 74% | 73% | 72% | 71% | 70% | |
69% | 68% | 67% | 66% | 65% | 64% | 63% | 62% | 61% | 60% | 69% | 68% | 67% | 66% | 65% | 64% | 63% | 62% | 61% | 60% | |
59% | 58% | 57% | 56% | 55% | 54% | 53% | 52% | 51% | 50% | 59% | 58% | 57% | 56% | 55% | 54% | 53% | 52% | 51% | 50% | |
49% | 48% | 47% | 46% | 45% | 44% | 43% | 42% | 41% | 40% | 49% | 48% | 47% | 46% | 45% | 44% | 43% | 42% | 41% | 40% | |
39% | 38% | 37% | 36% | 35% | 34% | 33% | 32% | 31% | 30% | 39% | 38% | 37% | 36% | 35% | 34% | 33% | 32% | 31% | 30% | |
29% | 28% | 27% | 26% | 25% | 24% | 23% | 22% | 21% | 20% | 29% | 28% | 27% | 26% | 25% | 24% | 23% | 22% | 21% | 20% | |
19% | 18% | 17% | 16% | 15% | 14% | 13% | 12% | 11% | 10% | 19% | 18% | 17% | 16% | 15% | 14% | 13% | 12% | 11% | 10% | |
9% | 8% | 7% | 6% | 5% | 4% | 3% | 2% | 1% | 0% | 9% | 8% | 7% | 6% | 5% | 4% | 3% | 2% | 1% | 0% |
The cliff-drop from yellow at 70 to orange-red at 69 is replaced with a more gradual transition through a new yellow-orange range. Wbm1058 ( talk) 01:09, 22 December 2015 (UTC)
Alternatively, we could expand the discretionary range to 60–75%, making it a 15% range, as opposed to our current 5% range (70–75%). This proposal is a sort of "middle ground" for those who believe C1 is too little of an expansion and that C3 and C4 are too radical.
The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
C3: Expand discretionary range to 50%+1
|
---|
We could also expand to range to 50%+1–75%, so that it would then become approximately a 25% range, as opposed to our current 5% range (70–75%). If implemented, bureaucrats could use discretion for any RfA between a simple majority (50%+1) and 75%. Support C3
Oppose C3
Comments on C3
|
The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
C4: Abolish the discretionary range completely
|
---|
Finally, we could simply abolish the notion of a "discretionary range" completely and permit the use of discretion in all cases, regardless of what the support-oppose ratio is. All percentage-based measurements would become irrelevent to the closing of RfAs. For instance, if an RfA gained 45% support, a bureaucrat could still theoretically close it as successful if the oppose rationales were extremely poor in comparison with the support rationales. Theoretically, this is already the case, but the de facto range is 70–75%. This proposal is suggesting that we expand the de facto range to encompass all percentages, and bureaucrats will be expected to weigh arguments and use discretion at all times. Support C4
Oppose C4
Comments on C4
|
The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
D1: Upper limits on opposition
|
---|
To fix the issue of high standards that often vary from person to person, we could set an upper limit on oppose votes of certain types. There are certain basic statistics considered in RfAs (tenure, total edit count, recent number of edits, and content creation). We would set the upper limit for one of these statistics at a certain point. If a candidate who runs for RfA meets or exceeds this upper limit, a participant could not oppose the candidate because of a perceived deficiency in that category. (It's not as complicated as it might sound.) For example, suppose that we set the upper limit for edit count at 10,000. (To reiterate, this upper limit is for opposers, not candidates.) After this, a candidate with 12,000 edits runs for adminship. In this case, someone could not oppose the candidate because they personally require 20,000 edits; the candidate exceeds the upper limit. As another example, suppose that we set the upper limit for content creation at 2 DYKs/1 GA. A candidate who has 3 DYKs runs for adminship, and therefore someone could not oppose the candidate because they think all candidates should have at least 1 FA, since the candidate exceeds the upper limit for content creation. In the event that a participant casts an invalid vote, it may be "flagged" as an invalid vote, and the user who cast it will be notified and given the opportunity to change it. If it is not changed, the bureaucrats will discount the vote. Finally, two things should be noted about this proposal: (1) It would still be fine to support a candidate who does not meet or exceed the upper limit. The upper limit simply dictates the point beyond which someone could not oppose a candidate. If the upper limit for tenure is set at 1.5 years, someone could still support a candidate who has only 1 year of experience. However, in this case, someone could also oppose the candidate for not having enough experience, since they are below the upper limit. (2) These upper limits would not prevent participants from opposing candidates for an entirely different reason outside the basic statistics covered by the upper limit. For instance, even if a candidate exceeded all the upper limits on basic statistics, they could still be opposed because they're uncivil, were recently warned or had an article deleted, etc. These aspects of the candidate are not covered by upper limits. We will discuss this general idea and also discuss what the upper limits should be. If you support or oppose the general idea, indicate your opinion in the proper sections for D1 below. Under D2, we will discuss what the upper limits should be. Editors should only support one limit for each category (tenure, edit count, recent edits, and content creation). For instance, under the "Tenure" section, you shouldn't support both the 1 year and 1.5 year proposals. If you support the idea of limits for that category but don't agree with any of the primary ones presented in the proposals, indicate your support under the "Support limit other than above" category and state what specific limit you support. If you oppose a limit for that particular category (e.g., you think that it should remain open to voters' discretion), place your opinion under the "Oppose limit for this category" section for each category. If the proposal for the general idea (D1) fails, then all proposals under D2 fail automatically as a result. Support D1
Oppose D1
Comments on D1
D1.1: Determining the upper limitsIn this section, we will discuss what the upper limit should be. Only supports are allowed in the D2 section, and assuming that D1 passes, the proposals that obtain the most supporters in this section will pass. (If D1 fails, all proposals in this section automatically fail as well, since the general idea behind these more specific proposals did not pass.) Please support only one proposal in each category (D1.1.1, D1.1.2, D1.1.3, and D1.1.4). D1.1.1: TenureTenure is defined as the total amount of time one has been on Wikipedia, starting at the point of the candidate's first edit. The primary options for this proposal are: (1): 1 year; (2): 1.5 years; (3): 2 years. (If you support a different limit, place your !vote in the "Support limit other than above" section and specify which limit you support.) If you support one of the primary options listed above, please place your !vote under the "Support limit for this category" section and indicate which one you support in your !vote (Example: "Support option (1).") Support limit for this category (D1.1.1)
Support limit other than above (D2.1.1)
Oppose limit for this category (D2.1.1)
Comments for this category (D2.1.1)
D1.1.2: Total edit countTotal edit count is defined as the total amount of edits a candidate has performed since they began editing Wikipedia. The primary options for this proposal are: (1): 5,000; (2): 7,500; (3): 10,000; (4): 15,000. (If you support a different limit, place your !vote in the "Support limit other than above" section and specify which limit you support.) If you support one of the primary options listed above, please place your !vote under the "Support limit for this category" section and indicate which one you support in your !vote (Example: "Support option (1).") Support limit for this category (D1.1.2)
Support limit other than above (D1.1.2)
Oppose limit for this category (D1.1.2)
Comments for this category (D1.1.2)D1.1.3: Recent number of editsThis proposal is different from D2.1, since these proposals concern the amount of recent edits, not the total. If a candidate meets or exceeds the number the upper limit for this category, they may not be opposed because of "lack of recent experience." The primary options for this proposal are: (1) 300 edits in the past 6 months (avg. 50 edits/month); (2): 600 edits in the past year (avg. 50 edits/month); (3): 600 edits in the past 6 months (avg. 100 edits/month); (4): 1200 edits in the past year (avg. 100 edits/month). (If you support a different limit, place your !vote in the "Support limit other than above" section and specify which limit you support.) If you support one of the primary options listed above, please place your !vote under the "Support limit for this category" section and indicate which one you support in your !vote (Example: "Support option (1).") Support limit for this category (D1.1.3)
Support limit other than above (D1.1.3)
Oppose limit for this category (D1.1.3)
Comments for this category (D1.1.3)
D1.1.4: Content creationThese proposals are about how much content the candidate has created and its relevance to RfA. The primary options for this proposal are: (1): No content; (2): [X] non-stub article(s) created/expanded (In other words, candidates will be exempt from content-related opposition if they create a certain amount of non-stub articles, regardless of whether or not it's recognized content; if you support this proposal, indicate what limit you support. Example: "Support proposal (2) with a limit of two articles."); (3): 2 DYKs or 1 GA; (4): 2 GAs or 1 FA. (If you support a different limit, place your !vote in the "Support limit other than above" section and specify which limit you support.) If you support one of the primary options listed above, please place your !vote under the "Support limit for this category" section and indicate which one you support in your !vote (Example: "Support option (1).") Support limit for this category (D1.1.4)
Support limit other than above (D1.1.4)
Oppose limit for this category (D1.1.4)
Comments for this category (D1.1.4)As worded, "created/expanded" is given lip service but only "created it all" or "created most of it from a stub" really count. A Wikignome who rewrites one paragraph on each of 10,000 articles and does a great job of it is counted as less of a contributor than someone who creates 7 GAs and 3 FAs. -- Guy Macon ( talk) 00:01, 30 November 2015 (UTC)
|