Wikipedia talk:Requests for comment

RFC statistics

If you've ever been curious about past RFCs, see User:BilledMammal/List of RFCs. I think that most of the links won't work, but they should all be findable in the archives via the timestamp. The data is not clean (for example, several of the old RFCs attributed to me were just me making repairs to broken RFC questions), but a few general findings might be fun:

• It's a lot of RFCs, but it's less than it used to be: Almost 17,000 RFCs are listed in this data set, going back to 2007. If you want to see the oldest ones, they were on the main page until about August 2005.
  • This list shows 752 RFCs in 2023, 913 in 2022, and 995 in 2021. This is down significantly compared to previous years: 1,249 in 2020 [beginning of the pandemic], 1,311 in 2015, and 1,215 in 2010.
  • We are currently on track for a record low (around 700) for this year.
  • In the 2010s, we ran about three new RFCs per day. This year, we will average about two new RFCs per day.
  • The drop in RFCs is interesting in part because I've seen comments saying that more RFCs are greeted with complaints about being unnecessary, because there are too many RFCs.
• Most people are first-timers: About 9% of RFCs were unsigned. Almost 6,000 logged-in editors started more than 15,000 (~90%) signed RFCs. Something around 250 RFCs (1.5%) were started by IPs.
  • Most people have very little experience with starting RFCs. 60% of editors in this list created one RFC. Another 15% have started only two RFCs. 90% of editors in this list have started four or fewer RFCs; they account for half of all signed RFCs.
  • Many RFCs were started by someone who was probably trying to read and follow the directions on this page for the first or second time.
  • However, some people start a lot more RFCs than normal: George Ho has started about 200 RFCs, though only a few since 2017. Robert McClenon and Snooganssnoogans have each started more than 100 RFCs. The counts fall off rapidly from there. Only a dozen editors have started 50+ RFCs (pinging Cunard, Binksternet, GoodDay, SMcCandlish, Helper201, The Four Deuces; note that some high-volume RFC starters are no longer editing at all, and others have been encouraged to find a different way of contributing to Wikipedia), and another 30 editors have created between 25 and 49 RFCs. The numbers really add up: Just 25 editors account for about 10% of the signed RFCs. Just 1% of the people starting RFCs have created 15% of RFCs; each of them has created at least 20 RFCs. If you've heard the saying that 20% of people do 80% of the work, we're not quite that skewed overall; here, 20% of people start 60% of RFCs.

WhatamIdoing (talk) 04:50, 29 September 2024 (UTC)

• Thank you. I will certainly look at these statistics. I have been a professional user of statistics, and I usually like to look at statistical reports. I have started a lot of RFCs because most of them have been the quasi-resolution of content disputes that I was mediating. Robert McClenon (talk) 17:24, 29 September 2024 (UTC)

  I think that's typical for some of the higher-volume RFC starters. As I said in the first paragraph, not every RFC attributed to me in this dataset is actually "from" me. I've also been asked to be the person who starts RFCs, since I'm extremely familiar with the process. WhatamIdoing (talk) 19:42, 29 September 2024 (UTC)

Most of the RfCs I've started have been in the course of acting as a guideline shepherd, to resolve questions about the wording or interpretation of a guideline (or occasionally policy), most often an MoS page. I don't often start RfCs that pertain to something mainspace-localized (e.g. the wording of an article's lead, or which photo should be used in a bio article). I do respond (via WP:FRS) to many such RfCs, though. They seem to serve a useful function when done properly, but the noob factor is palapable, as very often WP:RFCBEFORE is ignored, or the RfC is nowhere near neutrally worded and is trying to force a point that some inexperienced editor isn't "winning" on in a prior and short-circuited discussion. I'm not really sure what could be done to curtail that. Perhaps RfCs should require someone to second them before they are listed as RfCs by the bots that do that work. (The concern is that RfCs draw in editors from all over the project, and are thus expensive of editorial time and attention more broadly that just resolving matters more locally on the pertinent talk page with a regular discussion among the editors most interested. But they couldn't be done away with entirely, because they are a site-wide safety value, a check-and-balance against WP:OWN / WP:CONLEVEL problems.)  — SMcCandlish ☏ ¢ 😼  09:55, 29 September 2024 (UTC)

I've heard that RFCs used to require a 'second', but I'm not sure that's true and I haven't looked. Generally, I find that when the question is obviously biased, the community manages to handle it perfectly well anyway. Perhaps, when the need is to tell the POV pusher that the answer is really, truly, absolutely, unquestionably no!, then there's an argument to be made about the biased question being more effective. If you get to ask: "Shall we do it my way, which will result in rainbows, butterflies, peace, and love, or shall we do it the bad, wrong, horrible way that causes poverty, war, and oppression?" and people reject your proposal anyway, then maybe you'll finally get the message. WhatamIdoing (talk) 19:53, 29 September 2024 (UTC)

I haven't begun many RFCs this year, as the content disputes I was involved with in my area of interests over nearly 19 years, have been settled. GoodDay (talk) 15:07, 29 September 2024 (UTC)

I was surprised to see that I am one of the top editors for starting RfCs.

I always try to phrase RfCs simply and unbiased. Perhaps it would be helpful to review RfCs so ensure they are properly worded.

RfCs are useful for articles that a small number of editors dominate since it brings in a wider range for a second look. But they are not very useful if the issues are complex, because most respondents,in my opinion, are not willing to spend much time. TFD (talk) 15:51, 30 September 2024 (UTC)

A little while ago, an editor (NatGertler, might that have been you?) said that getting a larger quantity of editors responding to a question has to be balanced with getting comments from people who actually know what they're talking about. RFCs that require either specific prior knowledge or spending more than a few minutes reading will not appeal to most potential respondents. We have plenty of editors who are qualified to opine about whether A or B is a preferable opinion (e.g., this image at the top of the page or not?). We do not have plenty of editors who are qualified to answer questions that require specialized knowledge. If we can get an RFC dominated by the small number of people who know what they're talking about, then that might be a good thing. WhatamIdoing (talk) 22:06, 30 September 2024 (UTC)

If it sounds right to you, then it probably was me. If it sounds wrong to anyone, then it probably wasn't. That's my stance and I'm sticking to it. -- Nat Gertler (talk) 23:05, 30 September 2024 (UTC)

It seems like the sort of sensible thing you might say. WhatamIdoing (talk) 01:10, 1 October 2024 (UTC)

I think we need to be extremely careful about how far we take that argument in terms of formulating best practice: down that road lay walled gardens and many of the exact issues that RfC is meant to combat. First off, respondents already have a significant degree of control over aligning the RfCs that they land at with their own interests and capabilities--either by self-selecting from the listings or registering in the appropriate FRS categories.

Second, and more germane, often it is precisely the editors without too much stake or pre-existing attachment to an exact issue or sub-field that you want to recruit with RfC. Because the issues resolved by RfC turn much less often on respondents having some hyper-specific understanding of arcane, subject-matter specific knowledge, and much more often on applying policy and previous community consensus on a particular editorial issue. It is extremely, extremely rare that more voices does not improve the dialogue and the synergy of the consensus outcome with broader community best practice. That's the main value of RfC; injecting more voices--and specifically voices less likely to be entrenched by prior dispute, idiosyncratic rules created by small numbers of editors that may not conform to policy or represent best practice, or even their own "expertise".

And let's remember that Wikipedia is about as susceptible to the Dunning-Kruger effect as any work space humanity has ever produced, due to it's open and anonymous nature; often these levels of expertise are self-assessed and the people most convinced of their unassailably superior grasp of the subject matter are exactly the people whose perspectives need to be diluted by a larger pool of perspective, to ease a problematic grip on the article, family of articles, or process in question. When editors arrive and cannot parse the issues because they are so technically complex, that is of course another issue, albeit relatively rare in my experience--again, most issues turn on a reading of policy vis-a-vis summarizing the conclusions of sources, not a comprehensive understanding of the field in question.

And in cases where a reasonable determination does hinge on highly field-specific knowledge, there's very little lost by an editor being pinged to a discussion, only for them to determine within a few minutes that it is outside their ability to properly come to grips with, and therefore decline to respond, or else give a bit of very guarded, limited input. As to the concern that "RFCs that require either specific prior knowledge or spending more than a few minutes reading will not appeal to most potential respondents.": a) I'm not at all convinced that is true, and b) if someone is not willing to spend more than a few minutes reading in order to catch up on the context and nuances of any RfC, they are not the ones you want responding in any event. SnowRise ^{let's rap} 13:58, 2 October 2024 (UTC)

I think it depends on the subject. Editors don't need to spend time reading sources to vote in a "Which image is better?" RFC, and those tend to attract a higher than average number of participants. Editors tend to self-select out of questions that require specific background knowledge or significant extra work, and those tend to attract a lower number of participants. WhatamIdoing (talk) 16:51, 2 October 2024 (UTC)

Electronic Intifada should be deprecated

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.

Moved to Wikipedia:Reliable sources/Noticeboard#Electronic Intifada.

---

In this article titled "German envoy admits he spread lie about 7 October mass rapes". They imply that the Germany ambassador lied about mass rapes. In reality, he only apologized for believing a story which was not corroborated correctly. it is a single story and no where he mentions "mass rapes". EI is lying openly. This should be enough for them to be deprecated given that they are used as a reliable source for A-I conflict. EI is extremely hyper partisan that it should not continue being used as a reliable source. LuffyDe (talk) 19:54, 17 October 2024 (UTC)

This seems a borderline WP:ECP violation, but there have been RFCs about EI in recent months. Remsense ‥ 论 19:57, 17 October 2024 (UTC)

@LuffyDe and Remsense: This is off-topic for this page, which is for discussing improvements to the page Wikipedia:Requests for comment. In fact, the question appears to be something that should go to WP:RSN. --Redrose64 🌹 (talk) 21:27, 17 October 2024 (UTC)

Sure, I would move it there. LuffyDe (talk) 22:34, 17 October 2024 (UTC)

The discussion above 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.

Support and suggestions for improving a draft RFC around whether advocacy groups can be used for WP:BLP or if they count as WP:SPS

User:Bluethricecreamman/SPS_RFC

This is part of a few discussions[1][2] around whether literature from advocacy groups can count as WP:SPS, especially in the context of WP:BLPSPS.

Current discussions total is about 1.4 WP:Tomats long, and I haven't had a good chance to really parse everything in the discussions out.

I'm looking to: 1) Make this RFC more succinct/formatted correctly. 2) Figure out if the proposed language should belong in the RFC to update a longstanding essay, WP:USESPS, or just drop it. 3) Trying to see how vague/abstract the question should be.

I suspect there is no RFC that would answer the discussions entirely and that debates will last longer, but I am looking for the question that provides the most information about what current community consensus is, and provides a useful framework to debate around. Bluethricecreamman (talk) 04:34, 4 November 2024 (UTC)

RfC metadiscussion

I reverted the recent addition of a sentence saying that debates about an RfC are not allowed in the RfC discussion, because it sounds like this is forbidding discussion of such things as whether the RfC is a waste of people's time because it's not timely or the question is too vague to yield useful comments or the question is biased. These are all important discussions, and we don't state any better place to have it. It's true that such discussion does not answer the request for comment, as it is not about the topic of the request, but the talk page section where those comments go seems like an appropriate place for it.

I think we should, if we don't already, advise people to try to have those discussions before creating the RfC.

If there is some other kind of debate that we want to discourage in the RfC discussion, we need to be more specific. Bryan Henderson (giraffedata) (talk) 18:19, 9 November 2024 (UTC)

This relates to an RFC on RSN, see WP:RSN#RfC on People’s Daily. There is no need for the policy change, it wouldn't be helpful to block such discussion and prior discussion is what was missing in this case. -- LCU ActivelyDisinterested «@» °∆t° 20:58, 9 November 2024 (UTC)

AFAIK the only kinds of debate that we want to discourage in RFC are behavioral problems. So long as people are following ordinary rules about participation, then whether their responses are what the OP wants to hear vs saying that it is a bad question is not really something we need to constrain.

I can imagine certain people (e.g., limited English skills, limited social/communication skills) would prefer a straight-up vote with no explanation, but that's not what RFC does. WhatamIdoing (talk) 23:10, 9 November 2024 (UTC)

I think you missed the point of the rule being proposed -- it isn't about what kinds of things are OK in any RfC discussion; it's about off-topic discussion. While there is nothing wrong with discussing flavors of ice cream in an RfC discussion, we don't want to see them in a discussion where the question is, "Is The People's Daily a reliable source"? By the same token, metadiscussion (discussion about the discussion) is off-topic; it is not under the topic on which comments were requested. Explaining a vote is not metadiscussion (but arguing about whether people should vote is). While I think metadiscussion is appropriate for RfCs in spite of it not being the comments that were requested, I do see why someone would think that discussion should go somewhere else. Bryan Henderson (giraffedata) (talk) 18:20, 10 November 2024 (UTC)

RFC signer

What is the current consensus on whether an RFC should be signed with the filer's username? I see there was a discussion about this in 2018, where most agreed that the filer should be identified. Asking this question after an RFC was left unsigned here, which although is in line with the RFC information page, is a rare occurrence and obscures accountability whenever it is required. I think the RFC information should be changed to reflect the consensus of editors. Makeandtoss (talk) 10:19, 10 November 2024 (UTC)

There was a discussion on this a couple of months ago. The opinion seemed to be that it is acceptable, even sometimes beneficial for them not to be signed, as it let’s editors come to an opinion on the question without being prejudiced by their opinion of the opener.

It’s not all that rare; about 9% are unsigned. BilledMammal (talk) 10:35, 10 November 2024 (UTC)

9% is rare. They can avoid prejudice by having the statement filed by an admin for example, not necessarily by having an empty signature. I think after all these discussions an RFC on the way to open an RFC is needed. Makeandtoss (talk) 10:50, 10 November 2024 (UTC)

Given page history is publicly accessible, I don’t understand the line "obscures accountability" - if you want to know, can’t you just check the history? BilledMammal (talk) 10:52, 10 November 2024 (UTC)

If the RfC question is posed in a perfectly neutral way, it shouldn't be possible to determine the opinion of the opener from the question. The problem comes when a non-neutral question is unsigned, not when a neutral question is unsigned. Requiring a signature will encourage the opener to be more careful in drafting the question. I could imagine an exception for committee-drafted RfCs like we have on VPP sometimes, but otherwise I think that everything that appears on any type of discussion page should indicate who put it there. An RfC on this would be interesting, but I think VPP would be the right place and it would need a discussion first. Zero^talk 11:11, 10 November 2024 (UTC)

That is why I mentioned obscure not prevent. Makeandtoss (talk) 11:14, 10 November 2024 (UTC)

All of that was mentioned in the prior discussion. Another argument against signing an RfC is just a philosophical one: The request isn't personal, so there's nothing to sign. The requester isn't a secret, it's just not part of the request. It's like the reason that you don't sign a Wikipedia article when you write one. Bryan Henderson (giraffedata) (talk) 18:47, 10 November 2024 (UTC)

90% + of people who do RFCs sign them. Selfstudier (talk) 19:04, 10 November 2024 (UTC)