Wikipedia talk:Good article nominations

MainCriteriaInstructionsNominationsBacklog drivesMentorshipDiscussionReassessmentReport

Good article nominations

This is the discussion page of the good article nominations (GAN). To ask a question or start a discussion about the good article nomination process, click the Add topic link above. Please check and see if your question may already be answered; click to show the frequently asked questions below or search the archives below.

RfC: change GA criteria to require inline citations in all cases

I propose that criterion 2(b) of the good article criteria should be changed from

all inline citations are from reliable sources, including those for direct quotations, statistics, published opinion, counterintuitive or controversial statements that are challenged or likely to be challenged, and contentious material relating to living persons—science-based articles should follow the scientific citation guidelines

to

reliable sources are cited inline. All content, except for plot summaries and that which summarizes cited content elsewhere in the article, must be cited no later than the end of the paragraph (or line if the content is not in prose).

-- Mike Christie (talk - contribs - library) 10:48, 29 May 2023 (UTC)

Support

• Support. Mike Christie (talk - contribs - library) 10:48, 29 May 2023 (UTC)
• Support, criteria should show the actual current expectations. —Kusma (talk) 10:54, 29 May 2023 (UTC)

  In my view the proposal does not mean WP:BLUE would no longer apply to GAs. If this is unclear, I would not oppose a clarifying addition. "Likely to be challenged or contentious" stopped being the standard for material needing citations years ago (both at GA and elsewhere), we should not pretend it still is descriptive of current practice. —Kusma (talk) 19:57, 7 June 2023 (UTC)

• Support This is the appropriate thing to do from both a descriptive and prescriptive perspective—these are the current standards, and so they should be. TompaDompa (talk) 12:25, 29 May 2023 (UTC)
• Support, this is a change reflecting already-existing practice. The proposed wording is a simple transference of the existing criteria at WP:DYK. CMD (talk) 13:49, 29 May 2023 (UTC)
• Support per CMD. Ajpolino (talk) 13:53, 29 May 2023 (UTC)
• Support per above. TBH I've always been a bit iffy about the plot summary aspect too, it would be nice to clarify that if plot summaries aren't cited then they must only stick to unambiguous WP:SKYISBLUE type assertions that are directly stated in the work, not interpretations of what people were thinking or motivations that aren't stated. But that's not really a matter for this RFC. Cheers  — Amakuru (talk) 13:53, 29 May 2023 (UTC)

  GA nominations are already required to comply with WP:OR and WP:WAF, which hopefully cover any potential issues here. Thebiguglyalien (talk) 14:27, 29 May 2023 (UTC)

• Support long overdue. For the vast majority of articles, the most practical and likely only way for the content to be realistically verifiable is to be cited inline. (t · c) buidhe 14:01, 29 May 2023 (UTC)
• Support as one of the editors that pushed for this RfC. The GA criteria have fallen behind current standards of what's considered high quality. Thebiguglyalien (talk) 14:21, 29 May 2023 (UTC)
• Support I think it's expected of any article that is to be deemed of quality on Wikipedia.--Wehwalt (talk) 16:12, 29 May 2023 (UTC)
• Support as I assumed this bare minimum adherence to core policies was already required. — Fourthords | =Λ= | 17:46, 29 May 2023 (UTC)
• Support - and while we're at it, I think we should do away with the "is the pope Catholic/is the sky blue" exceptions entirely. FunkMonk (talk) 18:42, 29 May 2023 (UTC)
• Support. Although unlike FunkMonk I think that keeping out the [citation needed] trolls for the night sky being black on Olbers' paradox is a good thing. Really obvious statements can be difficult to source because high-quality sources don't generally waste space on them. —David Eppstein (talk) 18:47, 29 May 2023 (UTC)
• Support. No paragraph in a GA should be without at least one citation Billsmith60 (talk) 19:58, 29 May 2023 (UTC)

  UPDATE: change to OPPOSE, due to the acute disparity this new rule would create between future GAs and those that were approved some years ago. Regards Billsmith60 (talk) 13:50, 25 June 2023 (UTC)

  @Billsmith60, what disparity would that be? As Wikipedia changes its rules, articles no longer meeting the criteria get demoted. This has been happening for a long time (twenty years ago, Featured Articles did not require inline citations; at some point, expectations changed, and most FAs not meeting this requirement have since been demoted). For GAs, the same happens; as the criteria change, articles lose their green plus. —Kusma (talk) 13:58, 25 June 2023 (UTC)

• Support, with reservations about removing the link to Wikipedia:Scientific citation guidelines (SCG). Inline citations are expected for verifiability of an article broad enough to meet criterion #3. This has no global consequences as the GA system is run by a dedicated WikiProject that can set its own criteria, so long as they do not violate core content policies. Hopefully it should be implicit that this does not override SCG. I am slightly uneasy about this being enforced retroactively just because of the amount of work this could create—look at WP:URFA/2020 and note that there are 6 GAs for every FA. But, yes, old GAs that do not have inline citations likely fail verifiability in practice and should not be listed as GAs. — Bilorv (talk) 20:05, 29 May 2023 (UTC)
• Support. A sensible codification of near universal current practice which should make expectations clearer while having minimal effect on how things work going forward. Gog the Mild (talk) 20:10, 29 May 2023 (UTC)
• Support I had initial concerns, specifically surrounding how this would apply to tables, but Kusma's point in response to Thryduulf's oppose is a good one, and one that I was coming around to myself. With the modern practice of table captions, where a single citation is used to support a whole table, that can be placed on the table citation, therefore appearing before the information it is citing, and so meeting the proposed criteria. With that covered off, I think this makes it very clear what the expectation is, and should be. Rule creep maybe, but a good one. Harrias ^{(he/him) • talk} 20:17, 29 May 2023 (UTC)
• Support for all the good reasons already mentioned. -- LCU ActivelyDisinterested ∆transmissions∆ °co-ords° 20:34, 29 May 2023 (UTC)
• Support - per Gog and Buidhe. -Ljleppan (talk) 20:52, 29 May 2023 (UTC)
• Support should be a fairly uncontroversial change, this. ~~ AirshipJungleman29 (talk) 22:24, 29 May 2023 (UTC)
• Support, and recommend that we tend to rethink what we consider to be trivial. I recently rewrote John Bullock Clark. Take a look at this pre-rewrite version. I imagine we'd just waive off all that stuff in the infobox as noncontroversial and "trivial" in a review, yet there's multiple errors there. The February 18, 1864, date is a blatant date error, and the mention of Lindley is misleading, as he actually filled the vacancy left by James S. Green as noted by sources on Clark (although technically I don't think Green ever really took his seat, but Lindley was long gone by then and the sources mention Green, not Lindley). What we think is trivial and obvious is often not so; we should be checking these things against sources. Hog Farm _Talk 23:48, 29 May 2023 (UTC)
• Support. Long overdue. Iazyges Consermonor Opus meum 01:45, 30 May 2023 (UTC)
• Support, this is already the case in practice, as it should be. The nominators/reviewers of Good Articles ought to be evaluating the source support for statements in an article anyway, so it's not like this would introduce more burden, right? JoelleJay (talk) 01:49, 30 May 2023 (UTC)
• Support, to be honest, I thought this was the case, the articles I've submitted for GA all met it, and when I've reviewed articles for GA status I've insisted on it. Will more stringent sourcing requirements mean that some (not all) GAs take longer to meet that status? Probably, and I don't see that as a problem. Mackensen (talk) 02:22, 30 May 2023 (UTC)
• Support, because this (a) matches the rule to the existing review practice, (b) makes the criteria more clear to new editors, and (c) matches the GA rule to the DYK rule for consistency.Rjjiii (talk) 08:58, 30 May 2023 (UTC)
• Support An article without proper in-line citations cannot be considered a "good" article. Chris Troutman (talk) 13:08, 30 May 2023 (UTC)
• Support, but only on the condition that we make an exception for statements that fall under WP:BLUE; even though WP:BLUE is an essay, there's a good point to be made that we don't need to cite statements that virtually everyone will agree on. For example, in an article about a building in New York City, we should not also need an additional citation if we wanted to say that the building is in the United States. Getting rid of the WP:BLUE exception may invite POV warriors and other users to challenge things that are generally widely known facts, like the fact that NYC is in the US. Otherwise, this is merely formalizing what has been unwritten practice for years, and it would harmonize the rules with that of DYK. Epicgenius (talk) 14:00, 30 May 2023 (UTC)
• Support, should have been like this for years now, and most of us have treated it as if it was; it was an embarrassment that DYK had a stricter rule than GAN. Chiswick Chap (talk) 18:00, 30 May 2023 (UTC)

  I would change to Oppose if GAN had to be of the same exacting standard as DYK Billsmith60 (talk) 10:59, 1 June 2023 (UTC)

• Support to reflect current practice. I'm not strictly attached to the particular wording, but it does seems good, and as far as I could tell, the DYKSG wording was more popular than other suggestions made in the last 3 discussions. DFlhb (talk) 18:04, 30 May 2023 (UTC)
• Support. GA standards shouldn't be below that of DYK. Olivaw-Daneel (talk) 20:26, 30 May 2023 (UTC)

  I don't agree with your comparator. DYK is a pain in the neck and more like FA standard Billsmith60 (talk) 10:57, 1 June 2023 (UTC)

• Support, as it seems to me that this has long been a standard practice. My one caveat, as others have said I see no reason for plot summaries to be excluded either, but so be it. To my mind, if we don't allow OR, citations are required. SusunW (talk) 21:12, 30 May 2023 (UTC)
• Support - unverified facts should be nowhere near a GA. We can make exceptions for WP:SKYBLUE when necessary. Anything "good" on an encyclopaedia should be verifiable, above all else. Tim O'Doherty (talk) 22:02, 30 May 2023 (UTC)
• Support - Honestly, after being around WP:GAR for a while. It's just kind of made sense that GAs should use inline citations and not just general citations. Onegreatjoke (talk) 22:51, 30 May 2023 (UTC)
• Support The person who loves reading (talk) 23:01, 30 May 2023 (UTC)
• Support as long as the amendment also takes WP:SKYBLUE into consideration where necessary (per O'Doherty et al.). Wonder if the FA team has been catching up to this development? --Slgrandson (How's my egg-throwing coleslaw?) 11:17, 31 May 2023 (UTC)
• Support. Duh. CactiStaccingCrane (talk) 12:31, 31 May 2023 (UTC)
• Support even at GA-level, this should be the norm rather than aberrative. SN54129 15:24, 31 May 2023 (UTC)
• Support - Would have thought this would have been the policy already. Seems obvious.--NØ 15:36, 31 May 2023 (UTC)
• Support: Tamzin's suggested amendment to account for WP:BLUE is a good catch, and I'd support seeing it updated, but the difference doesn't matter so much in practice. That's because this guideline, at both GA and DYK, doesn't actually require reviewers to investigate the sources to confirm that the cited content is verified. DYK practice doesn't require that reviewers do spot-checks on anything but the bits of the article that contain the facts used in the hook. Tamzin's example about a BLUE bit of the article not being verified in the source, then, doesn't actually play out in this guideline, because it doesn't care about verification. GA hasn't actually nailed down what source reviews should look like in practice, and this guideline doesn't do that either: it just says what needs to be cited. That leaves plenty of room for BLUE considerations ("the provided source backs up the substantive content, BLUESKY background is verifiable elsewhere"). theleekycauldron (talk • contribs) (she/her) 01:13, 2 June 2023 (UTC)
• Support Long overdue, and is already de facto policy among GA reviewers. Surprised that this isn't in the criteria already. JML1148 (talk | contribs) 04:37, 3 June 2023 (UTC)
• Support Since DYK requires this, and GA are supposed to be DYK-ready, this is already a requirement. Hawkeye7 (discuss) 05:56, 7 June 2023 (UTC)
• Support as the de-facto standard, as the standard the community has already drifted toward for verifiability reasons, as the standard we ought to be enforcing for verifiability reasons in any case, and the standard I have enforced as a reviewer. Contra my colleague above, though, GAs are not subject to DYK criteria, and therefore this is not already the letter of the law.Vanamonde (Talk) 18:29, 7 June 2023 (UTC)

  Addendum; I would argue that common sense still applies, and content we would never cite is still exempt per WP:BLUE; but I would also support modifying the proposal to make this clear. Vanamonde (Talk) 18:33, 7 June 2023 (UTC)

• Support this is a BLUESKY kind of vote. GA can and should be higher standard. I am not so worried about abuse of BLUESKY as it is similar situation in DYK which arguably is even more nitpicky crowd. ~ 🦝 Shushugah (he/him • talk) 21:57, 7 June 2023 (UTC)
• Support reflects existing practice and helps ensure high quality. Phlsph7 (talk) 21:46, 9 June 2023 (UTC)
• Support An article without inline citations is as good as citing a book without a page number; it's necessary for verification that we can point to specific sources that support specific sections. The concern about SKYISBLUE seems overblown. I tend to lean towards the view of if it's so obvious and important to mention surely someone reputable would have written it down? I don't think anyone is going to wikilawyer over whether, say, "Joe Biden is a homo sapien" due to lack of RS explicitly saying that as the people below seem to imply. I do think we should also include a specific caveat or at least understand, as some of the folks down below have noted, that it is fine when there is a large table with a single "source" box for including one or two citations or if there is a block quote formatted in the style of "Johnson wrote:[ref 57] Blah Blah Blah." Though I doubt most GA reviewers are going to wikilawyer around with what is currently proposed to prohibit such practices. -Indy beetle (talk) 07:00, 13 June 2023 (UTC)
• Support per CMD.Moxy- 11:24, 13 June 2023 (UTC)
• Tentative support. I think the opposing voices make some decent points as to obvious information, or, as Hut 8.5 best put it, "statements which every reader will recognise as true". But frankly, I don't think these kinds of statements come up that often, and I wonder if we're over focusing on exceptions that prove the rule. I've only participated in a couple GA discussions (two reviews/two articles), and, believe it or not, none of those articles featured a sentence like "The sky is blue." or "Manhattan is in New York." Even if the BLUESKY instances are more common than I think, surely we can footnote or simply amend the above list of exceptions? I mean, off the top of my head: Content should be supported by an adjacent citation appearing no later than the end of the paragraph. This rule does not apply to plot summaries, statements that every reader will recognize as true, or content summarizing a portion of the article (assuming that portion is adequately cited).--Jerome Frank Disciple 13:23, 14 June 2023 (UTC)
• Support I have been reviewing to this standard for years now anyway. Daniel Case (talk) 04:06, 17 June 2023 (UTC)
• Support per the reasons said above. TheCorvetteZR1^{(The Garage)} 17:29, 21 June 2023 (UTC)
• Support – obviously. No good reviewer will be passing articles that lack them anyways. This is already expected at FAC (although not formally instated, the same de facto norm applies there). No reason to hide this information from new nominators. – Aza24 (talk) 19:02, 24 June 2023 (UTC)

Oppose

• Oppose This is not a Wikipedia criteria or even an FA criteria. It's not even reasonable....it would require citing all "sky is blue" material. North8000 (talk) 12:30, 29 May 2023 (UTC)

  must be cited no later than the end of the paragraph. Isn't one citation per paragraph the de facto citation requirement nowadays anyway? New articles not following this for example are likely to get a maintenance tag. –Novem Linguae (talk) 13:56, 29 May 2023 (UTC)

  @Novem Linguae: The proposal doesn't say 1 cite per paragraph. It says that everything must be cited, e.g. sky is blue items. North8000 (talk) 21:13, 30 May 2023 (UTC)

  This is not a Wikipedia criteria or even an FA criteria Is WP:5P2 good enough for you? All articles must strive for verifiable accuracy, citing reliable, authoritative sources, especially when the topic is controversial or is about a living person. Editors' personal experiences, interpretations, or opinions do not belong on Wikipedia. Trainsandotherthings (talk) 16:15, 29 May 2023 (UTC)

  That sounds a lot more like the existing language than the proposed change. Expecting verifiability "especially when the topic is controversial" is different from expecting citations for every statement. -- Visviva (talk) 17:21, 29 May 2023 (UTC)

  Yes, per Visviva you are describing existing policy, not the proposed change which goes far beyond existing policy including requiring cites for all sky is blue items. North8000 (talk) 21:17, 30 May 2023 (UTC)

  What do you mean by "This is not a Wikipedia criterion"? When taken literally, it's bizarre – of course it's not currently a criterion, it's being proposed! jlwoodwa (talk) 06:59, 9 June 2023 (UTC)

• Oppose. This is good practice in most cases, but that is very different to requiring it in all cases. Visviva gives a good example, others are lists or tables cited to the same source, e.g. filmographies, sports and election results, several tens of identical footnotes do not make the article more verified than a sentence stating that all entries are sourced to X. Quotes of lists or bullets, etc. would also be required to have an inline citation at the end of every line, even when that would not make sense - see Universal Declaration of Human Rights for several examples. This change would also require inline citations in addition to prose citations, including midway through multi-paragraph quotes. Thryduulf (talk) 18:40, 29 May 2023 (UTC)

  I see no language in the proposal that forbids putting the citation before the content as in your example. —Kusma (talk) 19:52, 29 May 2023 (UTC)

  I read the proposed text as requiring citations after the cited material. Given that we are both reasonable people this different understanding is another reason to object to the change. It also unquestionably declares that a citation that comes at the end of a (quoted) list that covers more that the last entry in that list as insufficient. Thryduulf (talk) 21:17, 29 May 2023 (UTC)

  I think this raises an interesting point. For instance, in an FA I worked heavily on, there's a plain text listing of all the rail lines the company operates. Each section of the list begins with citations which support the bulleted text immediately following. I do agree it would be dumb to add the same citations again and again for each list item. Trainsandotherthings (talk) 23:41, 31 May 2023 (UTC)

  I think it self-evident that placing a citation ahead of the list it verifies (example from a recent article of mine) is both appropriate and not something that would be prohibited by the "no later than" phrasing. TompaDompa (talk) 23:46, 31 May 2023 (UTC)

• Oppose. This seems to be just a rule creep to me. Ruslik_Zero 19:37, 29 May 2023 (UTC)

  In addition, omission of scientific citations guidelines is concerning. Ruslik_Zero 19:41, 29 May 2023 (UTC)

  The scientific citation guidelines are weaker than the proposed change. Keeping them would essentially carve out an exception where science articles don't have to be as well cited. Thebiguglyalien (talk) 20:26, 29 May 2023 (UTC)

  Indeed: Wikipedia_talk:Scientific_citation_guidelines#Update_needed? Rjjiii (talk) 09:01, 30 May 2023 (UTC)

• Oppose. I would support if "content" were changed to "content that could reasonably be challenged". Consider a sentence like "The plaintiff claimed that the city's actions had violated the First Amendment to the United States Constitution, which protects freedom of speech". Suppose the citation, a high-quality law review article, does not explicitly say that the First Amendment protects freedom of speech, because that's so basic a claim that it often goes without saying. The proposed wording would say that this citation is insufficient, even though the claim is trivially verifiable. To me, that defies common sense. -- Tamzin^{[cetacean needed]} (she|they|xe) 20:45, 29 May 2023 (UTC)

  In my experience, the "trivial" things are either also trivial to source, or alternatively the process of finding a source reveals they are not quite so trivial after all. Ljleppan (talk) 20:55, 29 May 2023 (UTC)

  The point is that this change would institute a requirement that truly trivial and basic matters would need a citation where they do not now, and indeed where doing so would be unnatural - e.g. if an article says that something was criticised by "Multiple United States government departments." with one source that states that thing was criticised by the Bureau of Alcohol, Tobacco, Firearms and Explosives and another that says the National Institute of Corrections were unhappy with it. Neither source explicitly says that these are US Government departments, because the target audience of those sources would clearly know that. Thryduulf (talk) 21:32, 29 May 2023 (UTC)

  Why would trivial matters be in the article in the first place? Billsmith60 (talk) 13:19, 31 May 2023 (UTC)

  There is a massive difference between "trivial matters" and "trivially verifiable matters", but some things that are trivial in isolation are important in context (often to provide context). Additionally, some things are too trivial to explicitly state for a source but not for us, because of the different contexts of a general purpose global encyclopaedia and a specialist source. This proposal would require an inline citation for all of this, regardless of context or other ways to verify information (e.g. links to other articles, citations elsewhere in the article, etc.) It's far, far too blunt an instrument. Thryduulf (talk) 12:21, 1 June 2023 (UTC)

@Tamzin Re: your example, why would we be adding in an inference that isn't supported by the cited text? That just seems like OR. I also suspect that very few BLUESKY statements aren't really supported by an adjacent substantive citation; in your example, how likely is it that a source reporting an alleged 1A violation on the basis of free speech would not actually mention that reason? If it isn't referred to at all, then we shouldn't be introducing that interpretation in the first place. I struggle to think of any instances where a statement is a) necessary to include in our article; b) unsupported by whichever sourced statement it accompanies, let alone by any citations elsewhere in the article; and c) would be too difficult/seem too silly to cite. In fact, the example in BLUESKY itself would already be covered by both this proposal and existing policy.

"Humans have five digits on each hand"

This is referring to two incidents in the version history, one from 2007 with zero citations and a poorly-cited 2018 version, and in both cases the request for citation was already spurious as the content was in the lead, was expanded on later in the body, and anyway would easily have qualified for the new OR/calc policy of accepting routine reading of images as sources (there were pictures of hands in the article). Further, in the 2018 case, the "5 finger" claim actually did contain a citation at the end of the sentence: Normally humans have five digits,[failed verification] the bones of which are termed phalanges[citation to an illustrated dictionary that supports "5 fingers"], on each hand, although some people have more or fewer than five due to congenital disorders such as polydactyly or oligodactyly, or accidental or medical amputations. It was also immediately preceded by lead text cited to a dictionary that defines "finger" as "one of the five terminal parts of the hand". Even if it wasn't cited in the lead, the corresponding body text A rare anatomical variation affects 1 in 500 humans, in which the individual has more than the usual number of digits; this is known as polydactyly was undersourced, but obviously any source used to define polydactyly would necessarily also verify "5 fingers".

JoelleJay (talk) 18:57, 23 June 2023 (UTC)

• Weak oppose per Tamzin. I particularly take issue with the bolded content in this sentence All content, except for plot summaries and that which summarizes cited content elsewhere in the article, must be cited no later than the end of the paragraph (or line if the content is not in prose). If the content can reasonably be challenged, then sure, we should add a citation. But take an example like railway stations - are we supposed to require citations for simple info such as the next stations on the line? The proposed wording would require that we add an explicit citation for the next stations on the line, which are not in prose, even though this is such a trivial matter that it should not need a citation.
  I do agree, on the whole, that almost everything should be cited inline, but this really should not apply to something that should be, nominally, very easy to verify. My problem with this specific wording is that, in order to meet these new requirements, people may be encouraged to add citations that are not directly related to the topic itself. Take, for example, Palace Theatre (New York City). If I were to write that the Palace Theatre is located in Manhattan, New York City, New York, United States, would I have to find a citation saying that NYC is in the United States, even though there is a citation for the Palace Theatre being in Manhattan, NYC? The addition of a citation for the Palace Theatre being in the US would be unnecessary at best and harmful at worst, since not many citations will explicitly mention such a trivial detail as "The Palace Theatre is in the United States", even though this can be very easily verified. – Epicgenius (talk) 23:15, 29 May 2023 (UTC) Moved to Support. Epicgenius (talk) 14:00, 30 May 2023 (UTC)

  I would cite the rail stations but I wouldn't require a citation for Manhattan being in the US, to me that's WP:SKYBLUE if included in another sentence with information cited to a specific source. (t · c) buidhe 01:14, 30 May 2023 (UTC)

  Fair enough. I suppose that citing the next railway stations would be trivially easy, anyway; it just requires an additional sentence in the prose and perhaps a map or timetable. My concern was more with situations like the latter - people would be motivated to find citations for every little detail, even if the addition of such a citation is not important to readers' understanding of the subject, and that is why I slightly opposed the wording of this proposal. – Epicgenius (talk) 01:38, 30 May 2023 (UTC)

  The proposed wording is taken verbatim from WP:DYK (specifically from WP:DYKSG#D2). I don't participate at DYK often, but how frequently is this a problem there? Thebiguglyalien (talk) 02:18, 30 May 2023 (UTC)

  I can't recall it coming up there. WP:BLUE situations are generally easy to solve. If they're not stupendously BLUE (ie. Manhattan being in the United States, which is almost a matter of disambiguation), they're usually in an existing citation. If they're not, they probably aren't BLUE. CMD (talk) 03:18, 30 May 2023 (UTC)

• Oppose seems like a contradiction to WP:BLUE. This specifically says that all content should be cited online, which would include things that we would not typically cite. I'm all for stating where your information comes from and citing to specific sentences, but this is far too damaging. Also, specifically stating about plot sections isn't specific enough. There are other instances where text is not generally cited (image captions, navigational/keys, glossary meanings etc). We only don't cite plot summaries because the source of the information is the subject itself which is a bit of a redundant ref. There are bound to be other uses where the sources are just taken from the subject that aren't limited to plot summaries.

I'm also not the biggest fan of having additional requirements that are higher than that of Wikipedia as a whole. Lee Vilenski ^{(talk • contribs)} 13:38, 30 May 2023 (UTC)

This discussion pertains to nominations for "good article" status. It goes without saying that such articles are of a higher standard than "Wikipedia as a whole". I wouldn't object to examples like the First Amendment one above, where a subclause contains an "sky is blue" type assertion, but there shouldn't be standalone claims that are uncited in a GA, however "obvious" you may think them.  — Amakuru (talk) 13:49, 3 June 2023 (UTC)

So why does recently promoted GA Songbird Sings Legrand contain an uncited assertion in the first sentence? None of the citations in the article verify that Regine Velasquez is Filipino - source 2, the Philippine Daily Inquirer, states that she is "local" but (a) that is the sort of "obvious" you are arguing against, and (b) this citation comes two paragraphs after the claim. It has been explained elsewhere that you cannot rely on claims in other articles, so whether it is verified there or not is irrelevant. Thryduulf (talk) 14:13, 3 June 2023 (UTC)

• Oppose for many of the same reasons discussed above. An example of an issue that could emerge based on GAs I've written: Would TV episodes have to cite credits? That's not in line with current practice. It's reasonable to require inline citations for the types of statements outlined at WP:BURDEN, but this is already implicitly covered by criteria #2: "Verifiable with no original research", with a link to WP:V (which WP:BURDEN falls under). I would consider supporting new language more in line with those policies. RunningTiger123 (talk) 16:40, 30 May 2023 (UTC)

  If this does pass, I think "plot summaries" should be changed to "statements drawn from a work being discussed". That would cover cast lists, and also brief mentions of a work's contents, e.g. "In Genesis, God creates Adam and Eve", while also clarifying that a plot summary is not a get-out-of-jail-free card to write statements unsupported by the work's plain meaning. -- Tamzin^{[cetacean needed]} (she|they|xe) 19:53, 30 May 2023 (UTC)

  I support this, and I'll go a step farther to say that this should be updated in other project space pages as well. The proposed text was originally taken from WP:DYKSG#D2, where it says plot summaries, and WP:When to cite similarly mentions plot summaries to the exclusion of other types. Then there's also the fact that we have MOS:PLOTCITE and MOS:PLOTSOURCE, which go into more depth about this, but I don't know of any equivalent for things other than works of fiction. I've run into this issue when writing articles about nonfiction works, where I basically had to assume that I could use it as a primary source for itself in the same way. Thebiguglyalien (talk) 20:38, 30 May 2023 (UTC)

• Oppose usual practice is that some information does not need a citation, such as statements which every reader will recognise as true. This is the case for GAs as well. Hut 8.5 17:49, 30 May 2023 (UTC)

• I'd be happy with something along the lines of "supported by inline citations where appropriate" from the FA criteria. Hut 8.5 16:57, 14 June 2023 (UTC)

• Oppose as contradicting WP:BLUE and going past FA criteria. I hope whoever closes this discounts arguments that Every paragraph must have a citation as the proposal is about every fact/statement and not every paragraph. I'd appreciate someone explaining to my why this proposal is necessary although I appreciate the work put into proposing this. I also fail to see how this "matches DYK criteria" as the DYK criteria (to my knowledge) is only about the hook and not the whole article, Chipmunkdavis. — Ixtal ^{( T / C )} ⁂ Non nobis solum. 13:45, 31 May 2023 (UTC)

  Ixtal, see WP:DYKSG#D2. It's verbatim. Thebiguglyalien (talk) 13:54, 31 May 2023 (UTC)

  Thebiguglyalien, thanks. I remain unconvinced GAN rules should be made to match DYK supplementary rules, but appreciate the clarification. — Ixtal ^{( T / C )} ⁂ Non nobis solum. 14:10, 31 May 2023 (UTC)

  I'm also unconvinced. Why should I have to follow the DYK rules if I write a GA but don't want to bother with DYK? XOR'easter (talk) 14:57, 31 May 2023 (UTC)

• Oppose As a rule of thumb, I think this is a good one, but this would elevate that to a status it does not deserve. XOR'easter (talk) 14:52, 31 May 2023 (UTC)

  Amending my comment: the current version does a better job of conveying why citations matter. The proposed replacement tries to reduce good writing to a formula. I don't think that's progress. XOR'easter (talk) 17:31, 11 June 2023 (UTC)

  Or, to try saying it yet another way: "at least one at the end of each paragraph" turns out to be a reasonable density of citations in practice, because a paragraph is supposed to be one or more sentences working together to express an idea, and each idea we talk about has to come from somewhere else first. But making that the standard instead of expressing why it is often a reasonable minimum puts the focus in the wrong place. It's like saying that because smoking is bad for your lungs and smokers buy more lighters than non-smokers, the way to avoid lung cancer is to stop buying lighters. The "solution" looks at the problem the wrong way around. A standard which implies that text goes from acceptable to unacceptable because of a paragraph break may be easy to apply, in that it makes for easy box-ticking, but it's still misguided. XOR'easter (talk) 20:14, 12 June 2023 (UTC)

• Oppose, seems too strict to me. Plot summaries are an odd exclusion. As the original author of WP:BLUE I'm heartened to see it's stuck around. Stifle (talk) 15:13, 31 May 2023 (UTC)
• Oppose per Tamzin, Lee, and XOR'easter. The current wording reflects two existing major guidelines with broad community consensus; the proposed change would institute an arbitrary standard for no obvious benefit. Additionally, I think the new wording is less helpful. Giving specific advice on citing contentious material about BLP subjects, citations for direct quotes, and reference to the scientific citation guidelines is helpful for reviewers and editors. The proposed change would remove that generally applicable advice in favor of noting a special exception for plot summaries and "summariz[ations of] content cited elsewhere in the article". Leaving aside the issue of what the standard should be, the proposed change seems to make the criteria less helpful in orienting editors and reviewers to our actual policies. If all that's desired is to add a requirement that inline citations be used to cite information within the paragraph it appears, I'd rather that simply be tacked on to the existing, helpful text, instead of overwriting it with a less helpful criterion. — Wug·a·po·des 23:16, 31 May 2023 (UTC)
• Oppose, per WP:BLUE, the fact Wikipedia:Verifiability is a core content policy and Wikipedia:CITE EVERYTHING is not, et cetera. -- Maddy from Celeste (WAVEDASH) 16:41, 1 June 2023 (UTC)
• Oppose per Tamzin. I think that there are some trivially verifiable statements that we don't need explicit cited sources on, and we need not impose anything further than WP:NOR and WP:MINREF for GAs. — Red-tailed hawk _(nest) 02:12, 2 June 2023 (UTC)
• Oppose per Tamzin. Volunteer Marek 05:55, 2 June 2023 (UTC)
• Oppose Solution in search of a problem. I agree with the arguments in opposition given above by Tamzin and Lee Vilenski. —Ganesha811 (talk) 20:28, 3 June 2023 (UTC)
• Oppose, however I would support something like all citations are inline and are from reliable sources, including those for direct quotations, statistics, published opinion, counterintuitive or controversial statements that are challenged or likely to be challenged, and contentious material relating to living persons—science-based articles should follow the scientific citation guidelines. --Rschen7754 20:34, 3 June 2023 (UTC)
• Oppose as seemingly stricter than the featured-article criteria (Wikipedia:When to cite). DMacks (talk) 02:48, 5 June 2023 (UTC)
• Oppose while it's a FA candidate not GA Wikipedia:Featured article candidates/Archaeology, Anthropology, and Interstellar Communication/archive1 is an example of an article which doesn't use inlines everywhere for a reason. Jo-Jo Eumerus (talk) 05:12, 6 June 2023 (UTC)

  Would you support then the alternative wording provided by tamzin above? (If this does pass, I think "plot summaries" should be changed to "statements drawn from a work being discussed". Rjjiii (talk) 08:21, 6 June 2023 (UTC)

• Oppose per many others. To be honest, I'd drop criteria 2 and 4 entirely and replace that with something general about meeting all content policy and relevant guidelines. Leave it to the whole community to determine the V and RS guidelines, not some quality club pet rules. -- Colin°^Talk 07:51, 6 June 2023 (UTC)

  If that were the policy, how would you explain the review process to a new reviewer. What actions would you tell them to take in order to review a nomination? Rjjiii (talk) 08:19, 6 June 2023 (UTC)

  You'd just tell them to ensure that the article met the relevant policies and guidelines, perhaps with examples of which ones are relevant and what meeting them looks like in various common situations. Thryduulf (talk) 10:25, 6 June 2023 (UTC)

  Based on a conversation elsewhere, I've realised I didn't spot the "and that which summarizes cited content elsewhere in the article" tucked in after the "plot summaries". Perhaps if the proposal succeeds, then those two exceptions might be reversed since the first affects a minority of articles and the latter is relevant to all. I thought this proposal was another attempt to demand our leads be overcited. Regardless, I still oppose. I think the job of quality standard projects should be to enforce community guidelines, not create new ones and new exceptions of their own. And others have noted that the problem may include images and also other information readily drawn from the subject (e.g. song) itself, not just "plot summaries". The existing practice of not needing a citation if the fact is already cited elsewhere does not insist (as this proposal does) that one location of this fact "summarises" the other location. Merely that they cover the same fact. Is "Genre R&B" (in the infobox) a summary of "Musically, "It's a Wrap" is an R&B song". Nope. Just another way of saying the same fact. Remember too that categories are also "content" that may need to repeat a fact cited in the body. This is why we get into a mess when we try to restate P&G and it isn't this project's job. -- Colin°^Talk 13:01, 6 June 2023 (UTC)

  Why should the "quality club" not be allowed to dictate the "quality club pet rules" when they don't affect any other P&Gs or projects? JoelleJay (talk) 19:45, 7 June 2023 (UTC)

• Oppose current proposed wording (although I agree many things like this need to be fixed at GAN). This proposed wording takes the GA criteria well beyond even the FA criteria, which state:

  1c – ... supported by inline citations where appropriate

  Problems in citations and much more in the GA process need to be fixed, but this proposal is overreach, and won't solve the core problems, which are about how reviews are conducted. SandyGeorgia (Talk) 15:36, 6 June 2023 (UTC)

  Sandy, I agree that the proposed wording is stronger than what you quote from the FAC criteria. In practice, though, FA reviewers operate on the same basis that this wording proposes -- it's been years, perhaps as much as a decade, since I recall seeing a nominator quoting WP:When to cite in refusing to add a citation request from a reviewer. Mike Christie (talk - contribs - library) 15:45, 7 June 2023 (UTC)

  I don't disagree, but the way this RFC has been approached is backwards. I can't even tell what the proposal intends (probably partly related to I have never understood what a GA is ... in my way of editing, if you're writing a decent article, you cite it. Period.) Is the intent to make the GA standards the same as the FA standards? Then, what is a GA? If the intent is to clarify what the FA standards are, or to change them, then this is the wrong page for doing that. You seem to be saying the intent of the proposal is to make the GA standards the same as what you believe the FA standards to be, and if your interpretation of the FA standards is what you are stating here, then the place to change WP:WIAFA is not WT:GAN. SandyGeorgia (Talk) 15:54, 7 June 2023 (UTC)

  The proposal doesn't change GA standards to FA standards, it fixes one tiny aspect of GA to reflect the current practice. Plenty of differences still exist between FA and GA articles. JoelleJay (talk) 19:40, 7 June 2023 (UTC)

  Then it's hard to tell what it does. WRT citations, I have never understood the point of an allegedly "good" article (decent enough) not being fully supported by inline citations. I agree that should happen (although the whole point is moot as long as most reviewers aren't even checking those citations). I don't think the way this proposal is worded does the job appropriately. SandyGeorgia (Talk) 20:35, 8 June 2023 (UTC)

  Sandy, per your reference to DCGAR, the instructions have since been changed to specify that spot-checks must be performed. There's also now a page which logs recent activity at GANs, and some GA regulars look at this periodically to try to spot weak reviews. Mike Christie (talk - contribs - library) 21:12, 8 June 2023 (UTC)

• Oppose—the propose is well intentioned but fatally flawed. We'd end up with the oddity of stricter requirements for a GA than an FA. We'd also end up with issues related to requiring sourcing for content that site wide policies and guidelines do not require to be sourced. Imzadi 1979 → 23:30, 6 June 2023 (UTC)
• Oppose echoing others. GAC should not be stricter than FAC. –Fredddie™ 01:02, 7 June 2023 (UTC)
• Oppose - I see no constructive purpose or reasoning to make GA content a higher standard than A-Class or Featured, both tiers above GA. Having it stricter than either would defeat the purpose of having A-Class and Featured. The second reason is that there aren't always inline citations available. Using highway articles as an example, the respective highway authorities sometimes place the data into route logs, straight-line charts, etc. Many of these lack details for proper inline citations, which does not make them less credible. If these cannot be used, it would be impossible for highway articles to reach GAN status. I can understand the need for better sourcing, though the push by editors on sourcing, material, article content, etc., on Wikipedia has gone too far, is unrealistic, draconic, and threatens to purge information about major topics off the website altogether. This isn't Encyclopǣdia Britannica. This is supposed to be a repository of common knowledge anyone can contribute to or edit, and the ever-increasing standards make Wikipedia lose sight of its original and intended goal every time the standards are raised too high. I will not support this proposal and strongly advise anyone who reads this to do the same. — MatthewAnderson707 (talk|sandbox) 05:46, 7 June 2023 (UTC)
• Oppose. I get what the idea was, but this would impose stricter rules for GA than FA. And it also loses too much in translation. The original is correct that for a big class of things we want to see an inline citation per statement. It is not okay to bury a quotation or a controversial claim in the middle of a large paragraph and then just dump some citations at the end of the paragraph. — Preceding unsigned comment added by SMcCandlish (talk • contribs) 06:25, June 7, 2023 (UTC)
• Oppose per WP:BLUE. Editors should be encouraged to use common sense about where a citation is appropriate in a GA. —pythoncoder (talk | contribs) 19:44, 7 June 2023 (UTC)
• Oppose. I really wanted to support this, because I certainly agree with the general principle that content should be well-supported by cited sources. But I don't think that the proposed change works. If we were to implement the change, there would still be a need for commonsense editorial judgment to deal with BLUE situations. And the status quo language also requires common sense to decide that a cite is needed. Per Tamzin and others, I tried to think of amendments that would achieve the nuances that I would want, and I end up with more complicated ways of saying what the status quo language already says. --Tryptofish (talk) 22:49, 9 June 2023 (UTC)
• Oppose as written. As others have noted, this seems like a very strong "no exceptions, no excuses" guideline that doesn't weigh appropriateness. To be clear a minimum citation per paragraph is absolutely expected, but what about allowed WP:CALC tables? For non-prose content? For image captions mentioning non-controversial facts from another article which is wikilinked in the caption for a reader curious about more on this other topic? This is easily fixable, just include "appropriate" or something and encourage the reviewer / nominator to use common sense. A prose paragraph with relevant, non-trivial claims about the work under discussion needs to be cited; other stuff, use discretion. SnowFire (talk) 16:36, 10 June 2023 (UTC)
• Oppose as written. Strongly agree we need to bring criteria in line with practice, but this wording risks conflict over skyisblue kind of things needing a citation. Per SnowFire / Tamzin. —Femke 🐦 (talk) 20:10, 12 June 2023 (UTC)
• Oppose. As written this is essentially demanding a citation, or possibly multiple, for every single sentence. That's simply unreasonable. Wizardman 01:54, 13 June 2023 (UTC)
• Oppose Going beyond WP:V is excessive. Research shows that citations are rarely used by readers and, in the few cases that they are clicked through, this is typically to provide more information about the topic rather than to verify something. So, while they may be useful as an audit trail for editors and reviewers, they are not normally required by readers and so should not be mandatory in the finished article. Andrew🐉(talk) 21:44, 15 June 2023 (UTC)
• Oppose as written. This is more than is required even at FAC. Furthermore, verifying that an article meets this standard would increase the burden on reviewers to an intolerable level. ♠PMC♠ (talk) 05:01, 18 June 2023 (UTC)
• Weak oppose. Like Tamzin, I would support this if the requirement was "all content that can reasonably be challenged." I like this idea in spirit as there are GAs out there currently I think are undersourced, so I think strengthening the requirements in that regard would be good. At the same time, it's a lot of work to source an article well in the terms we currently recognize, and if we were to impose such an inflexible requirement as this it would add a portion of un-common-sensical work on top of that for anyone working to get an article to GA. Mesocarp (talk) 11:21, 21 June 2023 (UTC)
• Oppose. As a general rule of thumb I like this, but as a hard-and-fast rule I see a WP:BLUE conflict (though BLUE is only an essay most editors note its relevance and importance). Generally speaking, GA hopefuls are going to adhere to this anyway (odd cases of lists and tables and such has been discussed extensively above and below). I would support changing the current wording to something along the lines of reliable sources are cited inline. As a general rule of thumb, all content, except for plot summaries, common knowledge, and that which summarizes cited content elsewhere in the article, should either be cited no later than the end of the paragraph (or line if the content is not in prose), or once immediately preceding the content in certain cases of tables, lists, etc. Cessaune [talk] 04:39, 25 June 2023 (UTC)
• Oppose strictly per Tamzin; and as per xem would change to Support if xyr multiple proposed amendments were incorporated. --GRuban (talk) 01:01, 28 June 2023 (UTC)
• Oppose as per change of mind noted in 'Support' Billsmith60 (talk) 10:57, 28 June 2023 (UTC)

Discussion

See the discussion of this change here. The proposed wording reflects current practice in good article reviews -- experienced reviewers have required everything to be cited, with the exceptions noted above, for a long time. Mike Christie (talk - contribs - library) 10:48, 29 May 2023 (UTC)

I've posted a note at the village pump, and added it to the centralized discussion template. Any other notifications needed? Mike Christie (talk - contribs - library) 12:13, 29 May 2023 (UTC)

What does (or line if the content is not in prose) mean? Might need clarification. –Novem Linguae (talk) 13:56, 29 May 2023 (UTC)

@Novem Linguae: I take this to refer to lists, tables, image captions, infoboxes, etc., where there is content that may need a citation, but which is not organized in paragraphs. --RL0919 (talk) 16:20, 29 May 2023 (UTC)

• I'd support if there was a "grandfathering" exception for current GAs. Johnbod (talk) 14:05, 29 May 2023 (UTC)

  I wouldn't mind giving current GAs a grace period, but grandfathering is bad: people should be able to look at existing GAs as model for new GAs, and different rules are bad for that kind of things. —Kusma (talk) 14:17, 29 May 2023 (UTC)

  This is my position as well. I suggested in the WP:RFCBEFORE discussion that we don't immediately delist any article that fails this updated criterion, but that we use the standard reassessment process as we find issues just like all of the other GA articles that have fallen below the minimum requirements. I'll note that, whether it's procedurally valid or not, WP:GAR has already been delisting articles failing this citation standard for a long time, even though it's not part of the criteria.Thebiguglyalien (talk) 14:31, 29 May 2023 (UTC)

  I've been de facto applying this standard to GAs when I check for needed GARs and I'd imagine I'm not the only one doing so. Trainsandotherthings (talk) 16:16, 29 May 2023 (UTC)

• I can't quite find my way to opposing the proposal, but I share North8000's concerns. On the one hand, there is a problem to be solved here: it's not good for the project to have an opaque GA process in which candidates are judged on criteria that authors don't have fair notice of. That weakens Wikipedia by making it less welcoming to the newbies and lower-engagement editors on whom so much of the project's content depends. On the other, this change doesn't really feel like a step in the right direction, and has a distinctly rules-creepy vibe. The GAC are aspirational, but the same could be said for a lot of our policies and guidelines in practice. So this isn't really just a GA process issue: quality-conscious editors are going to look to GAC as a lodestar even if they aren't planning on getting a particular article to GA. And anything that moves the culture of Wikipedia further toward "get everything right first" and away from "give it your best shot and then sort it out collaboratively" is going to tend to further weaken and slow the project. This is a small change, but every little change like this shifts Wikipedia further away from the kind of editing culture that makes wikis an effective collaboration tool. At the same time that isn't really a problem for the GA process to solve. So I guess it's reasonable to focus on just getting the GAC to reflect actual reviewer practice. -- Visviva (talk) 17:21, 29 May 2023 (UTC)

That's an insightful overview. I was on the fence, but "So this isn't really just a GA process issue: quality-conscious editors are going to look to GAC as a lodestar even if they aren't planning on getting a particular article to GA. And anything that moves the culture of Wikipedia further toward "get everything right first" and away from "give it your best shot and then sort it out collaboratively" is going to tend to further weaken and slow the project." has tipped me to "Support". If editors are looking at GAs as good practice, then let us try to make them so; and GAN should be post "sort it out collaboratively" (IMO). Thanks for helping me to form a view. Gog the Mild (talk) 20:03, 29 May 2023 (UTC)

If people look to the GA criteria as an example to follow, and this leads them to relying upon a formula when thinking is required, then that's a bad outcome. XOR'easter (talk) 15:10, 1 June 2023 (UTC)

That's a bad outcome whatever our criteria are, though. To the extent that editors are mindlessly applying rules without doing any thinking of their own, I would rather them mindlessly apply rules that make it more likely that they are producing good content, and easier for other people to find and fix problems. I suspect on balance GAN applying an incentive for editors to be over- rather than underenthusiastic in their citation practices would be a positive change. Caeciliusinhorto (talk) 16:14, 1 June 2023 (UTC)

I want to invite those opposing this to provide an alternative update. Perhaps using some of the language in Tamzin's post? The current rule does not reflect the current practice. To keep the language of the criteria the same, would not actually keep the existing rule, but would instead keep the rule a secret from new nominators. Rjjiii (talk) 04:45, 2 June 2023 (UTC)

In which case the current practice is what needs to change - GA should not be requiring references for BLUESKY things, should not be requiring citations in the middle of quotes, inline citations for things referenced in prose, etc. The rule as currently written matches WP:V and I see no benefit to making anything more stringent than that. Thryduulf (talk) 12:31, 3 June 2023 (UTC)

GA does not require those in practice, nor would the rule change require them. CMD (talk) 12:36, 3 June 2023 (UTC)

The rule change absolutely would require them - it says All content, except for plot summaries and that which summarizes cited content elsewhere in the article, must be cited no later than the end of the paragraph (or line if the content is not in prose). (emphasis mine) there is no exception for anything to is not a plot summary or summary of other parts of the article. Unless you are arguing that the things Tamzin, me and others have pointed out are somehow not "content" then they absolutely would require inline citations no later than the end of the paragraph/line in all cases. If this is not what GA requires at the moment, then the proposed text does not represent current practice as claimed. Thryduulf (talk) 13:15, 3 June 2023 (UTC)

Every case brought up would be or is part of a portion of cited information. Ideas such as "Manhattan is in the United States" and "the First Amendment protects freedom of speech" are going to be part of larger prose that has citations. The new guideline does not mean that citations need to define every word used in the prose, or similar. We rely on editors to understand their sources enough to copyedit and explain them, and that remains true even in the new rules. As for the finding of edge cases regarding particular formatting for listes, quotes, tables, or other forms of presentation, no wording would meet all edge cases.

The GACR do not exist in isolation as a piece of legal text. Much as with all our policies and guidelines, they are read and understood within the framework of how en.wiki works as a whole, including other policies, guidelines, and practice. Every single policy and guideline on en.wiki explicitly has exceptions, and the GACR are no different. Anyone trying to wikilawyer a fail through some sort of edge case or another would be rebuked, and those raising them as general questions would create discussion that helped clarify community consensus. As Rjjiii notes, the current wording gives a misleading impression as to what is expected from the community throughout all our various content evaluation processes, including GA/GAR, particularly for those not deeply versed in wiki-understanding (which anyone who accurately references and understands WP:BLUE is). There is not going to be a perfect formulation, but the new text goes a long way towards giving a more accurate impression of what is done. CMD (talk) 14:56, 3 June 2023 (UTC)

The current wording (and WP:V) encapsulates that these are guidelines with general exceptions, the proposed wording does not. The proposed wording literally and explicitly says that "all content" must have inline citations. Lists, multi-paragraph quotes, etc. are not edge cases they are core parts of hundreds of thousands (at least) of encyclopaedia articles. You and others have made it clear that there is no exception for SKYISBLUE, etc. Except now it seems like you are saying that there can be exceptions when its convenient but not at other times - that is much less friendly to those not versed in "wiki-understanding" than what we have at present.

The current wording gives a misleading impression as to what is expected from the community throughout all our various content evaluation processes. Looking through recently-promoted GAs, as someone who is not intimately familiar with the GA process, the current wording appears to be a much better reflection of the expected standards than the new wording is. Thryduulf (talk) 19:01, 3 June 2023 (UTC)

I don't see where I or anyone else has ever said there is no exception for SKYISBLUE or etc., let alone "made it clear". I will however state that lists and quotes are generally not SKYISBLUE, and should be cited. As someone who is familiar with the GA process, the existing wording suggests a much lower standard that is actually expected. If you have seen GAs which do not cite things properly, please bring them up in a new section here or take them to GAR. CMD (talk) 02:12, 4 June 2023 (UTC)

CMD, I think you've hit the nail on the head here. I have never seen a single reviewer suggest that WP:BLUE be ignored at GA or at DYK, nor has anyone come up with such an interpretation until this RfC after several discussions about this. While I'm sure the oppose !votes all have legitimate concerns, some of these look so similar to wikilawyering that there's really no practical difference. Thebiguglyalien (talk) 19:09, 3 June 2023 (UTC)

Well, perhaps we took the wrong thing from DYK, and have overlooked the DYK solution of having 60+ supplementary rules explaining the 18 criteria explaining the 5 core requirements. We only have 10 criteria to explain 6 core requirements, bunch of slackers really. CMD (talk) 02:12, 4 June 2023 (UTC)

I have never seen a single reviewer suggest that WP:BLUE be ignored at GA or at DYK – that is literally what this RfC would do, though. "All content must be cited". -- Maddy from Celeste (WAVEDASH) 05:09, 4 June 2023 (UTC)

@Maddy from Celeste: I can't speak for everyone, but I think virtually all of the supporters (including myself) are of the view that true SKYBLUE supersedes the need for citations for that particular area, but not the rest of the content of the paragraph. I just struggle to imagine a time when there would be an entire SKYBLUE paragraph, such that no citation would be necessary for the entire paragraph. Perhaps it's just that we've become overly legalistic, on both sides. Iazyges Consermonor Opus meum 05:28, 4 June 2023 (UTC)

It's a blurry line between clarity and legalisticism. I think we should strive to have guidelines that represent the real expectations, not the expectations minus certain implicit-to-some exceptions.

To address the matter at hand, WP:BLUE is an essay that presents an interpretation of WP:V and WP:MINREF. But this is a proposal to explicitly set higher standards for GA than V and MINREF. As such, you can't rely on BLUE to justify exemptions. In fact, BLUE bases one of its arguments on the current GA criteria. Even beyond BLUE, which as an essay should not be given too much weight here, the very argument that some facts do not need inline citations has its policy basis in V and MINREF. Again, if we set higher standards than MINREF, we cannot implicitly rely on MINREF for certain unwritten exceptions. -- Maddy from Celeste (WAVEDASH) 08:34, 4 June 2023 (UTC)

We already have higher standards than MINREF. MINREF itself explicitly states "Substantially exceeding [MINREF] is a necessity for any article to be granted good or featured article (or list) status". It's also, inherently, the minimum, which seems an unusual bar for a process aiming to identify articles which are developed well beyond the minimum. CMD (talk) 10:45, 4 June 2023 (UTC)

@Chipmunkdavis, it looks like that sentence was boldly added without discussion in 2020 by an editor who no longer seems to be active. I don't agree that it's true, so I'll remove it. WhatamIdoing (talk) 20:37, 5 June 2023 (UTC)

Disagreement could change it I suppose, but it is true. It's not even GA and FA, there is furor when even a DYK goes up with poor referencing, ITNs are routinely rejected on sourcing, and OTDs are regularly removed. CMD (talk) 01:28, 6 June 2023 (UTC)

"Poor referencing" and "substantially exceeding MINREF" are not, necessarily, the same thing. It is entirely possible for an article to both substantially exceed MINREF and have poor referencing - for example every sentence having citations to half a dozen sources of varying quality. An article about a contentious BLP could easily be GA quality without exceeding MINREF by virtue of there being almost nothing that hasn't been or is not likely to be challenged. Thryduulf (talk) 08:10, 6 June 2023 (UTC)

You're quite right, I conflated two things. What I meant: The current GA criteria, just like MINREF, do not require everything to have an inline citation. Therefore, some things do not need inline citations – WP:BLUE. If we establish as a criterion that everything must have an inline citation, it is no longer true that not everything is required to have an inline citation. Hence, WP:BLUE would not be applicable anymore. It is important that WP:BLUE is an essay that explains why some things do not require inline citations; it is not a policy or guideline that could grant an exception if GA criteria specifically contradict it. -- Maddy from Celeste (WAVEDASH) 13:58, 4 June 2023 (UTC)

Even if it were a policy, GA could still require it. GA is a purely voluntary thing. We could make a GA criteria that says the letter E must never appear at the start of a sentence if we wanted to, or that there must always be an odd number of images. So long as it doesn't contradict a policy, it's free to have higher (or weirder) requirements. WhatamIdoing (talk) 20:00, 6 June 2023 (UTC)

One of these days, the community is going to have to reconcile the fact that the absolute bare minimum of what has to be cited under policy (WP:MINREF) is nowhere near the level of citation that's expected of a half way decent article (everything except common knowledge, plot summaries, and summaries of content elsewhere in the article). Thebiguglyalien (talk) 13:29, 6 June 2023 (UTC)

Editors who believe that nearly everything is Wikipedia:Likely to be challenged, and that therefore nearly everything already requires an inline citation, would not agree with you. WhatamIdoing (talk) 20:01, 6 June 2023 (UTC)

That is where we are in practice, and what the proposal is trying to address. This RfC is a case study in the difficulty of making Wikipedia processes more accessible. CMD (talk) 01:32, 7 June 2023 (UTC)

Examples from recent GAs

Here are some recent examples from reviews I've done in which the proposed wording reflects my current reviewing practice better than the current wording. These are not BLUE cases; they are cases where the uncited content doesn't fit into the categories required by the current wording.

• Talk:Carbon accounting/GA1 -- search the review for "There are uncited sentences...". The nominator initially argued that citations weren't needed, but I pushed on despite not having GACR supporting me.
• Talk:Marisa Anderson/GA1 -- the discography was uncited.
• Talk:Les Avariés/GA1 -- interesting because the nominator pushed back on the need for citations, citing WP:WHEN, and I had to make an argument that they should be cited anyway despite the GA criteria. When the nominator complied they responded "I have to admit, this actually did end up being a productive exercise, as I ended up finding some new details along the way, and identified and removed a claim that was not supported by sources and possibly incorrect."

This is three examples from the last twenty or so reviews I've done. I felt that none of my requests were supported by the current criteria, but I asked anyway, and I think the criteria should be changed to support these requests. As for BLUE issues, I agree with others above that true SKYBLUE issues are extremely rare. I don't think I've ever had a debate with a nominator over whether something was BLUE and needed to be cited, and I don't expect it to ever come up. I probably wouldn't object if, after this RfC, another was started to add some kind of wording about BLUE, but I don't think it would make any difference at all to reviewing practice. I would oppose adding such wording to this RfC, for fear of derailing a needed change. Mike Christie (talk - contribs - library) 12:01, 4 June 2023 (UTC)

Looking at your three examples:

• At Carbon accounting#Other applications the very first sentence fails the proposed criteria because it has no inline citation. It's not a SKYISBLUE by the arguments in this discussion so even if that unwritten exemption does exist (and there are different interpretations about whether it does or should) it wouldn't apply here.
• Marisa Anderson#Discography is currently only partially cited (I've not looked to see if this has changed since the GA review), but even if it were fully citied it would fail under the new wording if one citation was used to cover more than one line rather than having one per line.
• Les Avariés#Synopsis passes because of the plot summary exception.

So two of your three examples do not demonstrate that the new wording matches the current requirements. Thryduulf (talk) 14:30, 4 June 2023 (UTC)

We don't need an unwritten exemption for "GHG accounting is used in other settings, both regulatory and voluntary", as we have the written one about summarizing text. I've fixed the discography, good catch. CMD (talk) 14:52, 4 June 2023 (UTC)

Thryduulf, can you clarify? For Les Avariés, this is the version I was reviewing. I feel the current wording doesn't strictly allow me to ask for sources for the last three sentences in the "Film adaptations" section, because those statements don't fall into any of the categories listed in 2(b) of GACR. Are you saying that the current wording does allow me to ask for sources, or that asking for sources for these three should not be required for GA? I think you're saying the former, but I don't see how you're basing this on the current wording. Mike Christie (talk - contribs - library) 19:06, 4 June 2023 (UTC)

I want to know more about the expectations for citing a discography. Currently, under normal WP:V rules, the correct citation for a statement that amounts to "The Beatles recorded an album called The Beatles" is – nothing. That is the form of inline citation described in Wikipedia:Inline citation#In-text attribution, because the authoritative citation for it is the album itself. You could spam in a copy of {{Cite AV media notes}} between a pair of ref tags, but it would be redundant. It seems to me that this proposal would basically end up with reviewers insisting that ref tags be added anyway, so that the material will look like it's been provided with an inline citation. Is that what you expect? WhatamIdoing (talk) 20:11, 5 June 2023 (UTC)

In practice, the discography will also tell us things like when the album was first published and how long it is, which needs a citation. You should not cite this to just "the album" without further qualifications as there are sometimes different versions and you need to specify them clearly for verifiability. —Kusma (talk) 20:17, 5 June 2023 (UTC)

I believe it's typical for albums to have both copyright dates and information about how long it is. In the event that there were different versions with different information, I could see a need to specify, but not a need to use a pair of ref tags to do so. Consider "Album length: 23:45 (US version), 25:01 (UK version)".

With bibliographies, which are also part of "all content", different book editions are very likely to have different information, but we have never used ref tags in lists of works to specify that this item is the 2003 book with 243 pages published by HarperCollins vs the 2009 re-issue with 312 pages. WhatamIdoing (talk) 21:34, 5 June 2023 (UTC)

@Mike Christie re Les Avariés#Film adaptations, I hadn't considered that section. The proposed rules would require citations for those statements because release dates and national origin of production are not summaries of the plot. Whether the film genre and a film being a modernisation are summaries of the plot could be argued either way. The current rules do not require there to be citations, but WP:V does require them for material that has been challenged and the GA criteria do not prohibit you from challenging unsourced statements. Thryduulf (talk) 11:46, 6 June 2023 (UTC)

So do you feel that with the current wording of the GA criteria, all unsourced statements can be challenged by the reviewer and hence a source can be required for everything? With the usual BLUE etc. exceptions? That's (more or less) the outcome I would want, since that's current practice. If there's a consensus that that's the case, some wording change to clarify that that's how the review should be conducted would satisfy me. I don't know if other supporters of this RfC would agree with that, though. Mike Christie (talk - contribs - library) 11:52, 6 June 2023 (UTC)

Any unsourced statement can be challenged by anyone at any time for any (good faith) reason, and WP:V requires challenged statements to be sourced. I don't think GA rules could override that if even if that was desirable. Thryduulf (talk) 12:41, 6 June 2023 (UTC)

I think that answers half the issue the RfC is trying to address, but only half. The current wording means that if I had passed Les Avariés without challenging those statements, nobody could complain, because the GA criteria don't require me to do that. You're saying that a GA reviewer can challenge those statements, per V, and of course you're right. The RfC is trying to change the criteria so that it's clear the reviewer should challenge those statements. Do you see a better wording that would convey that? Mike Christie (talk - contribs - library) 13:37, 6 June 2023 (UTC)

Off the top of my head I can't think of a wording that encourages reviewers to challenge statements where relevant but doesn't require them to do so (or punish them for not doing so) where they aren't relevant (concise, clear wording is not my forte!). Any wording should also not be overly prescriptive about how and where a citation is provided, e.g. if there was a single source about adaptations of the work it should be permissible to cite that after every entry in the list or once at the start or end of the list (which is best will depend on context). Thryduulf (talk) 14:21, 6 June 2023 (UTC)

Then I think we've put our finger on something that is behind the disagreements above. Most experienced GA reviewers now do challenge all unsourced statements (barring BLUE etc.). You say any wording "[shouldn't] require them to do so", but the impetus for this RfC is that the supporters believe the GA criteria should be made more stringent -- that a reviewer should do so. That was the intent of the wording change. I think it's possible to oppose this RfC because one agrees with this approach but feels the wording is flawed, or because one feels the criteria should not be made more stringent in this way. (It sounds like you might oppose on both grounds.) I might be sympathetic to other wordings, but I think the change is needed. Perhaps supporters above are not changing their !votes based on the oppose arguments because the opposes are engaging with the wording rather than the fundamental question of whether the change is needed in the first place. Mike Christie (talk - contribs - library) 15:34, 6 June 2023 (UTC)

I don't think GA reviewers should challenge everything that could theoretically be challenged, rather they should apply same standard employed by the featured article criteria quoted by Sandy Georgia which strike the right balance and allow for the application of common sense and context while still requiring high standards of sourcing. The intent of the proposal is imo too strict, but wording is much stricter than even that intent (or at least the intent according to the majority of those commenting, others think the strictest possible wording is the stated intent).

For a real-world example, SECR N class is a featured article but would fail GA if the proposal passes because the N class construction history and Withdrawal sections have tables that are referenced with a single citation in the table header rather than one at the end of every line. Thryduulf (talk) 23:26, 6 June 2023 (UTC)

"High standards of sourcing" is incompatible with leaving challenge-able claims unsourced. Also, why wouldn't citations in the table header be valid under this new criteria? It doesn't say at the end of the line. It says no later, meaning you can't use WP:GENREF or anything like that. Thebiguglyalien (talk) 00:22, 7 June 2023 (UTC)

The "standard employed by the featured article criteria" is effectively that everything is sourced, which is exactly what is being proposed here. It also remains a higher standard than the proposed wording, because it requires that sources be of high quality while GA still requires sources only be reliable. CMD (talk) 01:22, 7 June 2023 (UTC)

That's perhaps not a great example as it has a number of unsourced claims in the body which should be addressed to bring it back in line with FA expectations. Trainsandotherthings (talk) 02:04, 7 June 2023 (UTC)

You could argue that it still meets the FA criteria even with a bunch of unsourced claims. And that's the problem here. On a sidenote, I'm loving the votes that essentially oppose because the FA criteria are weak and out of line with expectations, so the GA criteria should be even more so for consistency. Thebiguglyalien (talk) 02:12, 7 June 2023 (UTC)

Well... those votes are consistent with the reason GA was created. Editors wanted a lightweight way to identify decent articles – the ones that were "only" good, not the best ones. The kind of thing that most decent editors could realistically aspire to, without having to go through a nitpicky review process, or argue with a bunch of people, or know exactly how to format the citations, or anything like that.

Scope creep has been something we've had to push back on for as long as I can remember. There's always someone who wants to add just one more little thing. Back when we added most of the footnotes at Wikipedia:Good article criteria#cite note-1, we had a tool that flagged dead URLs, and multiple reviewers falsely claiming that dead URLs were banned in GAs. Then it was extra MOS pages. Then it was citation formatting. We didn't write Wikipedia:Reviewing good articles#Imposing your personal criteria and Wikipedia:What the Good article criteria are not just for the fun of it; we were addressing the problems that we were seeing with multiple reviewers trying to push GACR towards a personal FAC process. WhatamIdoing (talk) 01:24, 8 June 2023 (UTC)

I also really don't like when reviews involve personalized criteria (nitpicks about reference formatting are especially a pet peeve of mine). But meeting WP:V is just the bare minimum to not have content removed. I'd think that to say an article is "good" would require a bit more than that. Maybe it's because I wasn't around in the early days before this became the norm, but I do think that a citation for each unique claim is what's necessary before an article should have any little icon indicating quality, whether GA or FA. Not only is it impossible to check for OR without them, but it raises the question of how an editor summarized reliable sources without any reliable sources. An article with a bunch of uncited statements is not "good" or even decent. I'm not even convinced an article like that would make it through AfC, let alone GA.

But on the note of scope creep and simplifying the process, I think there's general agreement that GA needs some sort of reform to make it easier to nominate and review, especially for newer editors, and I'd love to see ideas to make that happen as well. Thebiguglyalien (talk) 02:17, 8 June 2023 (UTC)

Not only is it impossible to check for OR without them: Really? OR = claim has never been published in any reliable source, in any language, anywhere in the world.

Are you incapable of determining whether a claim like "Smoking tobacco raises the risk of lung cancer" or "Chris Celebrity won the 2022 Big Award" has ever been published in a reliable source, unless a source is presented to you on a silver platter in a little blue clicky number?

Mind the gap between "I don't feel that I should have to bother" and "It's actually impossible". WhatamIdoing (talk) 04:17, 14 June 2023 (UTC)

The current proposal is one that works towards this goal. It replaces a vague and individually interpretable criteria with something consistent that would reduce nitpicks and arguments. CMD (talk) 02:44, 8 June 2023 (UTC)

Images

In terms of what counts as "all content": Should there be an exception for images?

User:Mike Christie, I see that you recently passed several GAs that do not have citations for the images: Frederica Planta, Andrew Planta, Abishabis. The Wikipedia:Image use policy is listed in Wikipedia:List of policies#Content, so I think it's safe to say that images are content, and the proposal says it applies to "all content". How do you expect editors to cite the images in those three articles? What function are the citations supposed to serve? WhatamIdoing (talk) 20:26, 5 June 2023 (UTC)

Depends on the image, but if what it depicts is at all controversial it's a reliable source should be provided, either in the article or image description, for what it depicts. I believe that the proposal is only for text content however. If rewording the proposal to "prose content" would ameliorate your objections to it, perhaps we should consider that. (t · c) buidhe 22:01, 5 June 2023 (UTC)

"Prose content" means "not content contained in lists, tables, infoboxes, images, graphs, captions, lyrics, or poems". That's probably a bigger exemption than you intended. WhatamIdoing (talk) 20:05, 6 June 2023 (UTC)

Agreed - we don't normally expect citations for images, which I suppose amounts to believing what the Commons file says, and cites, in most cases (though of course these are sometimes wrong, just as "reliable sources" are). Sometimes they are desirable though, often mainly for the caption rather than the image itself. Since images get used all over the place, it is better to add a reference to the Commons file than to a single use of the image. Johnbod (talk) 22:06, 5 June 2023 (UTC)

My usual standard is that an image that merely illustrates something in the article text, with a caption that merely summarizes something in the article text, does not need a citation as being "that which summarizes cited content elsewhere in the article". If there is actual new information or claims in an image or its caption, they need citation. And in some cases the source of an image may need a citation for credit instead of for verification. —David Eppstein (talk) 22:36, 5 June 2023 (UTC)

A technical drawing might "summarize cited content elsewhere in the article", but a photo of a person does not, and a source that directly says "This is a true and accurate representation of what this person looked like" is generally hard to come by. WhatamIdoing (talk) 20:07, 6 June 2023 (UTC)

Compromise wording?

Is there the possibility that supporters and opposers could come to some sort of compromise wording regarding WP:BLUESKY? My support would still stand with an explicit BLUESKY exception added in. (I think it is implicit but I understand that others are afraid they would have to cite every use of {{convert}} or similar things that are universally used without citations). Any thoughts? —Kusma (talk) 14:39, 21 June 2023 (UTC)

I think any issues could easily be solved by running a second poll regarding adding in BLUESKY exception (and potentially WP:SCICITE)? Not sure whether this should be concurrent or should follow the close of the above RFC. Iazyges Consermonor Opus meum 19:31, 21 June 2023 (UTC)

Pardon my impertinence, if you like. But where is this discussion going? Clearly there is insufficient consensus after a number of weeks, hence ending and archiving it accordingly – with thanks for all the contributions made Billsmith60 (talk) 11:45, 23 June 2023 (UTC)

It will end after 30 days, and an uninvolved editor will close the discussion. If it passes, I agree a second poll adding a BLUESKY exception seems likely; if it fails, another RfC with revised wording seems likely. Mike Christie (talk - contribs - library) 12:04, 23 June 2023 (UTC)

WP:Polling is not a substitute for discussion. Consensus brought about through discussion is always prioritized over simple support and oppose votes. Thebiguglyalien (talk) 14:39, 23 June 2023 (UTC)

We don't need a "compromise" regarding BLUESKY; there isn't really disagreement about the application of BLUESKY in the above discussion, nor do I remember any hint of disagreement surrounding it in any previous discussions. What seems to be needed is wording to account for the potential misunderstandings in a future RfC, but I don't think that RfC needs to come quickly. CMD (talk) 15:28, 23 June 2023 (UTC)

How is there insufficient consensus... I count a solid 60% for straight support, and 80% if conservatively counting opposes citing only BLUESKY as supports for an amended version that explicitly accounts for BLUESKY. JoelleJay (talk) 17:51, 23 June 2023 (UTC)

What is required for a 'pass'? I was in favour but have changed my mind and will vote 'no' now. Would not a c. 40% opposition score, as part of a large turnout, indicate significant opposition? I'm not bothered, of course, if the proposal does pass. Thanks Billsmith60 (talk) 13:47, 25 June 2023 (UTC)

Closure requested at WP:CR

I've requested a close for the RfC as it's been 30 days. Mike Christie (talk - contribs - library) 11:39, 28 June 2023 (UTC)

GA review exceptions June 4–10

I could have sworn I said I wasn't going to do this regularly.

• Reviews opened by new accounts:
  • Talk:Ontario Highway 41/GA1
• Checklist reviews:
  • Talk:HMS Munster/GA1
  • Talk:Maurice Duplessis/GA1
• Isamu Imoto is set for a quickfail. I bring it up because I left this user a message about their GA nominations last week, and they are also responsible for one of the checklist reviews above. This may need further attention.
• Possible failure to closely review the article or to spotcheck sources:
  • Talk:HMS Penn (1916)/GA1
  • Talk:Pharah/GA1
  • Talk:Twenty (concert)/GA1
  • Talk:Widowmaker (Overwatch)/GA1
• Nominations that have been waiting for a reviewer for over half a year: 66

Thebiguglyalien (talk) 02:04, 11 June 2023 (UTC)

I've removed the Isamu Imoto nomination, and left a follow up message on their talkpage. CMD (talk) 03:12, 11 June 2023 (UTC)

Should we ping the people involved in the reviews and nominations? Onegreatjoke (talk) 23:19, 13 June 2023 (UTC)

Requested another review at Talk:Maurice Duplessis (t · c) buidhe 23:39, 13 June 2023 (UTC)

Are "checklist reviews" actually banned, or just abnormal? The main point is to figure out whether the criteria are met. If you can do that without putting hundreds or thousands of words on the talk page, then that's probably okay. WhatamIdoing (talk) 04:31, 14 June 2023 (UTC)

The same editor thinks this is a GA, so I don't think their review was very thorough. (t · c) buidhe 04:42, 14 June 2023 (UTC)

They're plausibly overrepresented amongst 'articles that shouldn't be GAs', but the extreme subjectivity of that label makes it tricky to track, and some of the absolute worst cases of 'this should not have passed' I've seen were far from trivial reviews. Partially this ties in with the major debate about just what kind of article suffices for GA -- the example of Isamu Imoto is a kind of article I think shouldn't pass, but articles in its length range have passed. (Most of the '200-word road GAs' have been some kind of merged since their heyday, but slightly longer very-short-articles still exist and indeed pass routinely today.) It would be interesting to try survey multiple experienced editors in this area on GAs they think shouldn't have passed and what their reviews were like; as well as getting stats for whether quickpasses are actually disproportionate, you'd get some pretty fascinating variation in what each person complains about in the first place. Vaticidalprophet 04:53, 14 June 2023 (UTC)

Actually, "don't just leave an all-positive checklist." CMD (talk) 08:39, 14 June 2023 (UTC)

It may be worth adding a note to this effect to WP:GAN/I#R3, which currently only says that reviewers should "provide a review on the review page justifying that decision", with not discussion of what an adequate review entails Caeciliusinhorto-public (talk) 13:26, 14 June 2023 (UTC)

It may also be worth not having a huge checklist commented out with the strong implication you're intended to use it as your review. Vaticidalprophet 16:38, 14 June 2023 (UTC)

God knows I loathe the preloaded template, but it was one of the proposals implemented from Wikipedia:Good Article proposal drive 2023, so clearly some people thought it was a good idea. I certainly would have no objection to removing it again. Caeciliusinhorto (talk) 17:18, 14 June 2023 (UTC)

I read over all the proposal drive pages when I came back. I certainly appreciate the vigour and some of the results, but both the divide-by-zero-error re-ranking and the checklist seem to have developed hindsight consensus-against after people got to see what they actually do in practice. Vaticidalprophet 17:39, 14 June 2023 (UTC)

Agreed. These two proposals haven't been very helpful in solving the problems that they were designed to solve, but they both severely exacerbated other problems. Thebiguglyalien (talk) 05:58, 17 June 2023 (UTC)

I'd like to hear from people who use it, but I certainly would appreciate not having to delete the preloaded text every time. Not many people participated in the discussion, so it should be possible to change this fairly easily. —Kusma (talk) 17:58, 14 June 2023 (UTC)

I quickly reviewed Widowmaker because the nominator asked me to do it on lunch break. I verified that the information was accurate to the best of my ability by checking the sources. - Cukie Gherkin (talk) 16:24, 14 June 2023 (UTC)

Which sources did you check? CMD (talk) 00:35, 15 June 2023 (UTC)

I checked all reception sources that I could see online, and I checked the sources through the Gameplay and Design sections. The Appearances details I'm aware of prior, and since they were cited to primary material about the shorts and comics, I did not read or watch them. - Cukie Gherkin (talk) 23:26, 15 June 2023 (UTC)

@Thebiguglyalien, why did my review of the Ontario Highway article get flagged as opened by a new account? I am not a new account. There must be a glitch somewhere. —Ganesha811 (talk) 22:08, 23 June 2023 (UTC)

Ganesha811 An account with about three edits opened the review and inserted nonsense into it, and then it was deleted. Notice my post was on June 11, five days before you opened your review. Thebiguglyalien (talk) 22:14, 23 June 2023 (UTC)

Ah, got it! Didn't see the dates. Thanks for clarifying. —Ganesha811 (talk) 22:16, 23 June 2023 (UTC)

GA review exceptions June 11–17

• Simultaneous GA and peer review open: Family (1976 TV series)
• Possible failure to closely review the article or to spotcheck sources
  • Talk:The Birds (Alexander McQueen collection)/GA1
  • Talk:Death of Cooper Harris/GA1
  • Talk:Hits Different/GA1
  • Talk:Ontario Highway 6/GA1
  • Talk:Shigi Qutuqu/GA1
  • Talk:Siege of Lankaran/GA2
  • Talk:Snake Ridge Fire/GA1
  • Talk:Suitcase/GA1
  • Talk:Sweet Nothing (Taylor Swift song)/GA1
• Nominations that have been waiting for a reviewer for over half a year:
  • Fakhruddin Ali Ahmed (nom)
  • Serbian Progressive Party (nom)
  • Zail Singh (nom)
  • 64 others (needs urgent attention)
• Miscellaneous shoutout: Great work by Ozzie10aaaa and Shibbolethink at Talk:COVID-19 pandemic/GA4. Now that's a review!

After a month's-worth of these reports, I think a few conclusions can be drawn:

• The default review text may be doing more harm than good. I don't know how many checklist reviews there were before, but this does seem to encourage them. We might consider replacing it with a brief note at the top of the page laying out basic instructions for the review.
• If we're going to have source spotchecks be the norm, we need to be clearer about what this entails at WP:GAN/I and maybe even on the GA review page itself. If the above reviews did spotchecks, they don't say so (or I missed it).
• We need to do something to make sure older nominations get reviewed; six months used to be an unfathomably long time to wait for a review, but now it's increasingly becoming the norm for challenging nominations. Even AfC and its nightmare backlog doesn't keep pages waiting that long.

Thebiguglyalien (talk) 02:26, 18 June 2023 (UTC)

Six months has always been a reality for some GA reviews. I once waited more than a year for one in 2018-2019.

With 600 GANs waiting for review some of them will be waiting for a long time; maybe a backlog drive is in order (willing to coordinate). (t · c) buidhe 02:53, 18 June 2023 (UTC)

"Unfathomably" was probably the wrong word. But there was only a single 6 month nomination on January 1 this year, and it ballooned almost immediately after the sort order changed to stop prioritizing the oldest nominations. Thebiguglyalien (talk) 20:35, 18 June 2023 (UTC)

GAN backlog in mid-2018

2018 was a particularly bad time for the backlog (see attached). Certainly it seems to have spiked back to that height, after major recoveries from that level. I agree with every take, so to speak; it's very clear from 'returning from a time when there was much less of a backlog', and reading over ~everything that's gone down since the beginning of the year, that all of 'auto-checklist with heavy implication to newbies you're meant to use it as your review', 'divide-by-zero-error reranking', and 'public wall of shame for people whose reviews don't explicitly mention they've done a specific thing' are contributing to widespread issues at GAN.

My hobby horse for GAN is that people don't fail enough, and in particular are too hesitant to fail articles during a review rather than as quickfails. To a substantial degree this cuts at the heart of a rarely-mentioned-but-constantly-extant dispute, which is what is a GA? I can name articles with green badges that I think do not approach good-article status, and I'm sure anyone else here can too. I suspect those lists have very little overlap and very different contentions (in particular: "can very short articles be GAs" is the greatest thread in the history of forums, locked by a moderator after 12,239 pages of heated debate). All Wikipedia-quality-assessment-processes also possess the elephant in the room that their criteria trace back to lists constructed by very few people that amount to "yeah, this sounds good, let's go with it" that have never been particularly tested to see how well they match reader opinion, and much of what we do know about reader opinion to work from implies they at least intermittently come apart. I'm looking to apply for a WMF rapid grant to do some testing on that note, but things have gone a bit pear-shaped in my life now, so I'm not sure if I'll get in under the deadline. Vaticidalprophet 23:05, 18 June 2023 (UTC)

"What is a GA?" really is essential, and the above RfC suggests that we're nowhere near answering it. I imagine I'm not the only one who's feeling less and less inclined to review when I can't even get a straight answer of what a GA review is or what purpose it's supposed to serve. Thebiguglyalien (talk) 23:31, 18 June 2023 (UTC)

@Thebiguglyalien: Regarding The default review text may be doing more harm than good, is there a way to get data on first time reviewers (per week, per month, and compared to last year or previous weeks, etc.). The goal of having a template was to make the process more approachable to a first time reviewer. If it is not having a positive impact there, and it's having any negative impact, then the next step should be to just revert back to the blank page. Only if it is having a positive impact in recruiting first-time reviewers, should we discuss and revise. Rjjiii (talk) 03:02, 18 June 2023 (UTC)

I think we need to recognize the connection between increasing standards at GA, and the increasing levels of backlog. Of course no one wants to take on long or difficult articles if they are going to be lambasted for providing a review which spectators decide is insufficiently deep, or blamed for missing errors or copyright issues. Now with spot checking requirements, reviewers are forced to spell out precisely how they read and evaluated what may be dozens of sources, lest they be accused of not spot checking (a standard that's actually more strict than FAC, which only mandates spot checks for first timers!). If the increased sourcing requirement being discussed above gains consensus, I guarantee the backlog will get even worse, because it forces the reviewer to do even more work to prove a stricter-than-FAC standard. ♠PMC♠ (talk) 03:35, 18 June 2023 (UTC)

Reviewers are only forced to spell out how they read or evaluated something like 2-3 sources, no high figure gained consensus. This does not seem higher than FAC, each of which gets a source review covering far more than 2-3 sources that is independent of the rest of the reviews. CMD (talk) 12:17, 18 June 2023 (UTC)

Source reviews at FAC are distinct from spot checks at FAC. They generally involve evaluating the reliability and quality of the sources and the formatting of the citations. Spot checks at FAC are required only for first-time nominators, otherwise they are to the discretion of the reviewer. (See the somewhat obscure Wikipedia:Guidance on source reviewing at FAC#Spot checking). And of course FAC is a collaborative effort whereas GAN puts the burden entirely on a single reviewer, so any increase in standards falls on one person. ♠PMC♠ (talk) 13:15, 18 June 2023 (UTC)

Interesting. How many FACs these days don't involve any source checks? CMD (talk) 15:49, 18 June 2023 (UTC)

Most. The only time a source check is required is for a new nominator, or a nominator who has not nominated in a long time (years, at least). Every now and then a reviewer will do some spot checks as part of their review, or because they're looking at the sources anyway, but I would guess less than 1 in 6 FACs is spot-checked. Mike Christie (talk - contribs - library) 16:25, 18 June 2023 (UTC)

Which feels like another way of saying that our content review processes give very low priority to accuracy and only pay lip service to original research issues. And this problem is made worse by the general sense that a weakness in FAC should be magnified for GAN because "FAC is higher". Then any attempt to fix the issue both meets push back and risks depleting the already stagnant reviewer base. Should be an easy fix, right? Thebiguglyalien (talk) 17:12, 18 June 2023 (UTC)

Certainly appears 2 of the last 3 explicitly had none! What an interesting quirk, it provides an aspect by where DYKN can be more thorough than FAC. Perhaps the multiple reviewer format provides a better sniff test buffer, a wisdom of the crowd insurance mechanism that DYK and GA lack? I'm not sure how that might be imported here; a non-new nominator allowance is sadly part of what led to the February large-scale GAR. CMD (talk) 17:13, 18 June 2023 (UTC)

If it makes you feel any better, I have a hunch that most DYK reviewers don't check sources either. Thebiguglyalien (talk) 17:19, 18 June 2023 (UTC)

One thing that helps at FAC is that FAC has an institutional memory, both via the coordinators and because the core group of regulars changes relatively slowly. If a nominator were to start using sources sloppily after having passed their first FAC, there's a reasonable chance that a reviewer would notice something amiss -- multiple reviewers means there are more chances for someone to see something in the article they know is incorrect, or to read a source and realize it doesn't correctly support the article text. I know that for nominators I've seen misuse sources (naming no names since I can't bring any to mind, but this has certainly happened) I would do a unofficial spotcheck, sometimes not bothering to note it in the FAC. It's not a perfect system, and if someone suggested that e.g. the FAC coords randomly nominate articles from experienced nominators for a spotcheck now and then I would probably support that. But I don't think we've had anything slip through on anything like the scale of the Doug Coldwell affair. Mike Christie (talk - contribs - library) 17:35, 18 June 2023 (UTC)

The thing about Doug Coldwell that everyone tends to gloss over is that he was a known problem for years. There are ANI threads about his issues dating back to 2007. Because he was so prolific and so difficult to deal with, most people who were aware of this simply avoided reviewing his nominations rather than attempting to pursue sanctions (not blaming anyone - I was one of the people avoiding him). It took TAOT getting fed up and taking him to ANI for self-promo issues that finally opened the floodgates.

The lesson from Coldwell is not that we need to apply a microscope to every single person at GAN/FAC under the assumption that they may be bad actors. Most people at DYK/GAN/FAC are not fabricators, self-promotors, or copyright infringers. I can count the number of times I've had to bring up a source-text integrity issue at GAN on one hand, and most of those have been honest and reasonable mistakes. The lesson from Coldwell is that we need to be willing to aggressively quickfail nominations that have serious issues, and we need to more aggressively enact topic bans for repeat offenders. We could have prevented Coldwell from metastasizing if we had had the courage to engage with the issues rather than letting him run wild because he was a prolific content creator. ♠PMC♠ (talk) 22:47, 18 June 2023 (UTC)

I agree that virtually no users are intentionally gaming GA nominations, but I do think that text-source integrity issues need to be addressed. I went over my reviews, and of the 21 reviews where I listed my spotchecks in detail, I found some form of text-source integrity issue in 12 of them (and that's not counting close paraphrasing). Two of those 12 were severe enough that they contributed to a fail. Thebiguglyalien (talk) 23:26, 18 June 2023 (UTC)

There are some experienced FA nominators where I've found text to source issues significant enough to lead to withdrawal. It can happen because of bad faith but is more likely to be accidental or simply a mistaken copyedit that changes the meaning. (t · c) buidhe 00:22, 19 June 2023 (UTC)

We don't apply yet a microscope to anyone, and I'm not seeing how we'd have that capacity even if we'd want to. As for failing, that's not a simple solution. I failed a Coldwell article, it got accepted without changes by another reviewer. As Mike notes, we don't have pass coordinators here, and presumably we have a much larger pool than FAC, so part of the issues here are trying to find mechanisms that might replicate similar engagement functions. CMD (talk) 01:29, 19 June 2023 (UTC)

I failed a Coldwell article, it got accepted without changes by another reviewer. This is exactly the kind of thing I'm talking about, though. It was well known by many people that Coldwell had issues, but people either avoided reviewing his nominations or simply let it drop when he pushed back or renominated. I'm not attempting to single you out - as I said, I'm equally culpable by inaction. But perhaps if people had pushed back earlier and with more force, Coldwell might have been removed from DYK/GA earlier and created less of a mess. Regardless, we're derailing into rehashing Coldwell when that wasn't the original point I was making.

My original point is that there is a direct connection between increasing standards for GA reviews and the increasing backlog. It is more daunting for any reviewer to take on a difficult or lengthy GAN than it has ever been. We are expecting a single reviewer to do more work than is done by multiple reviewers at FAC, and then acting mystified about the fact that we have articles waiting for over 8 months. ♠PMC♠ (talk) 05:47, 19 June 2023 (UTC)

I'm afraid I don't see how checking a few sources makes GAN more work than FACs. There's a lot of work carried out in FACs which we don't require here. While we don't want basic ticking exercises, we require very little beyond that. CMD (talk) 06:09, 19 June 2023 (UTC)

This gets into the issue that no one really knows what we require. Yeah, we have WP:GAN/I#R3, but you'll notice it's missing one key detail: what's actually supposed to go in the review beyond "basic ticking exercises". If we disallow checklist reviews, we're quite vague about what the minimum is. I tried to compile what I gathered to be the minimum, and it requires a pretty thorough reading of the article. I've also found source checks to be by far the most complicated and time consuming aspect when I review an article, which does make me sympathetic to the idea that GAN reviews might be more intimidating than FAC reviews. Thebiguglyalien (talk) 06:56, 19 June 2023 (UTC)

Chipmunkdavis, follow up: I've since done my first two FAC reviews, and I have to say that it's much easier and faster than doing a GA review. First, a significant portion of my time on a GA review involves checking criterion 2, which isn't an issue at FA. And second, I don't have to worry about looking for specific things (WP:GACN) and can just write down all of the problems as I spot them. The only way I see a GA review being this painless would be if the reviewer just ran it like a peer review, ignoring the GA criteria and skipping sourcing entirely (which I suspect is how many of the more casual GA reviewers already do it). Thebiguglyalien (talk) 02:35, 26 June 2023 (UTC)

What can make a GAN review more work than a FAC notwithstanding the latter's higher requirement is that GA requires you to conduct a full review of the whole article, whereas with FAC because there are multiple reviewers you can just do a little bit. Hawkeye7 (discuss) 07:37, 19 June 2023 (UTC)

Yes. It would be unusual, but I could conceivably hop into a FAC and say "I've only looked at Sections X and Y because that's where my own knowledge lies" or something like that. But with a GAN, the expectation is all on the reviewer, and that makes it intimidating for new people to take on (not to mention the unclear instructions and increased scrutiny of reviews). Basically, it is not easy for anyone to review difficult articles, it is not easy for new people to jump into GAN, and we have a backlog as a result. ♠PMC♠ (talk) 07:59, 19 June 2023 (UTC)

I for one would be more inclined to review were the autogenerated text removed. It's a pain and actively promotes drive-by reviews. Just point reviewers to the page that lists the templates should they wish to use one. Trainsandotherthings (talk) 03:58, 18 June 2023 (UTC)

Pinging @Tkbrett, Actualcpscm, Jerome Frank Disciple, Unlimitedlead, Infsai, The Corvette ZR1, Floydian, Aza24, AirshipJungleman29, HistoryofIran, Actualcpscm, Vami IV, Cessaune, Benmite, Heartfox, and Ippantekina: as I feel that they should about this conversation considering that their reviews and GAs were called out here. Onegreatjoke (talk) 18:52, 19 June 2023 (UTC)

I'd feel a bit odd speculating as to the review of the article I wrote, and my reviews aren't listed here, but I think they could be, so I'll respond. I didn't explicitly list my spot checks in either of my two reviews: I—and perhaps I was mistaken here—didn't understand that to be required. I just assumed the "5 spot checks" were a precondition to a "pass" in the relevant category, not something that needed to be spelled out. In fact, in Talk:Keith Raniere/GA1, I actually forgot to remove "Have yet to do a spot check" from the review (because I had, mistakenly, said it twice and only removed one instance). But if you read that review or look at the edits I made, it become very obvious I checked almost every source, not just 5 sources.

I'm not totally sure what the precise suggestions are here, but, in general:

1. In both reviews—one of which I didn't finish because it was ultimately a quick fail—I found errors thanks to spot-checking. I think spot-checking is a good practice.
2. As a very new reviewer, I found the template helpful, and I would suggest keeping it: it structures the review in what, to me, feels like a fairly logical way. (Granted, I played around with the categories when I saw fit: While I understand that there might be OR that is attributable to a reliable source—like in a WP:SYNTH situation, in practice, those categories have been very similar for me.)
3. I do think I'd probably feel weird failing an article outside of a quick fail. I hadn't thought of that until I saw it mentioned by Vaticidalprophet, but, upon reflection ... yeah, I can see that. Granted, in my second review, I ended up working on the article itself quite a bit. I read a few other reviews before doing my first GA review, and I saw that reviewers do balance fixing what they can fix themselves with asking the nominator / other editors to address certain things. I don't think I balanced that perfectly—in hindsight, my mindset was more "it's our job to get the article to the GA standard", and, short of an editor insisting on an unsubstantiated statement or refusing to do work on the article themself (and, to be clear, my experience was the opposite of that), I can't imagine that I would have failed.

--Jerome Frank Disciple 19:29, 19 June 2023 (UTC)

Regarding failing a candidate, I saw discussion recently that boiled down to this: In the course of the review, if the article is not up to standard, the reviewer is supposed to provide feedback to improve the article to GA standards. The reviewer thus provides general feedback, such as "There are some instances of OR and the article could use copyediting for grammar." From there, only a few outcomes can happen:

A: The nominator understands this feedback and makes the improvements --> Passing feels required for the reviewer, since the article has been "fixed".

B: The nominator does not understand this feedback and asks for specificity. --> The reviewer has to point out specific issues to fix, which is basically just as much work as just fixing them directly; this sucks for both the reviewer and the nominator.

I don't remember where I read about this, but it was in the context of Wikipedia and referred to as something like "draft-decline loop" or "review-fix loop". Actualcpscm (talk) 19:41, 19 June 2023 (UTC)

@Actualcpscm: You probably read it in the same place I did: Mike Christie's essay WP:FIXLOOP. TompaDompa (talk) 19:54, 19 June 2023 (UTC)

That‘s it, thanks! Actualcpscm (talk) 19:55, 19 June 2023 (UTC)

Thanks for the ping Onegreatjoke, that's very thoughtful of you. I didn't even know this kind of review-review existed.

If I'm reading correctly, the issue with my reviews at Talk:Death of Cooper Harris/GA1 and Talk:Snake Ridge Fire/GA1 is that I did not mention whether or not I carried out source spotchecks. For the record, I did, particularly those that were used to support particularly contentious claims about living persons. If I missed something, do let me know.

If there is an expectation to mention source spotchecking in GA reviews, I wasn't aware of it. Likewise, I didn't mention "I checked for OR" or "I checked the history for edit warring" under the relevant criteria (referring here to this review). WP:GAN/I#R3 mentions that a spotcheck must be performed, not necessarily mentioned.

I'm very new to GA review, so my thought process was this: I presumed that having checked the criterion would be implied by marking the criterion as a Pass; I don't need to say "This is fine" for every item that I marked as a Pass with the template. If this goes against established practices or expectations, I don't think I saw that being mentioned anywhere. This situation might have been at least partially a result of me choosing somewhat short and easy candidates for my first reviews; I suppose that most candidates require at least some fixes under every criterion.

Rjjiii and Thebiguglyalien, I also want to address your concern regarding the default review text. As I mentioned above, I'm very new to GA review. So far, I have carried out 5 reviews: Talk:German torpedo boat T2/GA1, Talk:Snake Ridge Fire/GA1, Talk:Death of Cooper Harris/GA1, talk:Vatican City at the 2022 Mediterranean Games/GA1, and Talk:Dimitri Konyshev/GA1. The first 3 passed, the last 2 were quick fails. I really struggled with the default review text. I consider myself quite tech-literate, and I have gotten quite used to the editing interface(s) of Wikimedia/Wikipedia. I'm fine with both visual and source editing, and I often use a mix of both. In my experience, there were two major issues with the default review text. Firstly, it's formatted as basically a wall of text, which makes it difficult to read and interact with in the beginning. Secondly, I wasn't quite sure where I was supposed to make what changes. I particularly struggled with the pass/fail templates, as evidenced by this lovely attempt of using the review template.

I think GA review would benefit a lot from simplifying this aspect in particular. I'd be happy to provide more detailed feedback if you'd like to hear more about my experience! Actualcpscm (talk) 19:31, 19 June 2023 (UTC)

Your thoughts on the process are correct. There are no formal expectations or requirements of any sort. If you wanted to, you could just glance at the article for thirty seconds, slap six green check marks on it, and go your merry way. That would make people upset, but there's no mechanism to stop you or to review the review. That's the problem that I was (hopefully) trying to address with this scanning of reviews, but it doesn't seem to make much of a difference. Regarding the default review text, that's a very recent addition to GA, and one that appears to be having unintended consequences. Thebiguglyalien (talk) 19:39, 19 June 2023 (UTC)

Right. I always try to be through with my reviews (see Talk:Charles Edward Stuart/GA1 for an extreme example), but I was not aware that there was any expectation to explicitly declare OR/spotchecking-related verification. I thought that checking the green check marks was already a sign of confirmation that one went through those steps; if this is something that I need to start doing in my reviews, please let me know. Cheers, Unlimitedlead (talk) 21:27, 19 June 2023 (UTC)

Wow, what a thread. To be honest, I didn't understand most of the situations brought up here, but I read it all, and I can agree that GAN review should be consisting of more people than one reviewer. Like many pointed out before, it's easier to spot some sort of mistake or error in article by looking through it with multiple eyes, rather just one pair. Even though I never reviewed a single article in my life, and I probably won't any time soon, I think it would be a neat change, and maybe encourage more people to participate and help. infsai (talkie? UwU) 22:40, 19 June 2023 (UTC)

The article by me that's been brought up on discussion by me is "Hits Different", and I can see why. I was caught off guard when I applied to all of the suggestions of Unlimitedlead to get an immediate pass, even though I haven't fixed something they pointed out when the article passed. I changed it as soon as I could, and I'm still not sure if it's good, because review is over, and probably noone would bother to check it. If I'm being honest, I had the best experience with Kyle Peake, who reviewed half of my GA entries (3 passing, 1 failing), and it was a longer process taking up weeks (or months like it was with "NDA"), and he split review into multiple sections, checking each one from the article. I had a room to discuss with him and come to conclusions, nothing was rushed. And he was really forgiving, because with me constantly changing the content of "NDA", the review should result in failure, and I should come back when i could present a final form of the article (at least that's my opinion by looking at it retroactively, but still huge thanks to you Kyle!). infsai (talkie? UwU) 22:40, 19 June 2023 (UTC)

This is very helpful feedback. As a general note, the GAN reviews are not only a tool for the current editor, but also a resource for future editors to see when and why an article became a GA. It's helpful to see the thoughts that went into the decision. The spotchecking is something we look for because it's easy to note (eg. checked source X and it matched the content), and has in the past been skipped over by reviewers leading to problems. CMD (talk) 00:56, 20 June 2023 (UTC)

I do sort of wonder how often past reviews are used as a tool. The only time I ever check past reviews is to make sure the reviewer actually reviewed the article. And it seems a recurring issue is that in addition to the GA criteria not matching common practices (as is discussed in the ongoing RfC), we also have a problem with common reviewing practices not being described in the instructions. I've tried to compile an expanded set of instructions in my userspace that's closer to unspoken expectations, but you can't really enforce unspoken expectations. Thebiguglyalien (talk) 02:57, 20 June 2023 (UTC)

I feel those instructions go beyond basic expectations, but it is useful to have the example of the source check clearly laid out. CMD (talk) 03:26, 20 June 2023 (UTC)

Thanks for the belated ping. If one looks at my review, it is very clear that I only commented on issues I found with the article; there is no requirement to use one of the many different GA review templates. I did spotchecks and looked closely at the sources (I have done 100+ source reviews at FAC&FLC so I am no stranger to their importance) but found no issues, so alas I did not note any issues.

Very Wikipedia-esque an editor to call out 10+ other editors though not alert them; makes it appear that the OP doesn't want their input, just wants to prove a point. I'm sure this strategy is great in the communication and general improvement of content-reviewing practices. Or perhaps it is a "Possible failure" in that regard? Aza24 (talk) 05:55, 20 June 2023 (UTC)

If one looks at my review, it is very clear that I only commented on issues I found with the article - I'm going to chime in even though the original post didn't mention any of my reviews, but I generally do the same thing. My reviews tend to be similar to something like Talk:The Birds (Alexander McQueen collection)/GA1 or Talk:Shigi Qutuqu/GA1). If the ref spotchecks don't reveal any issues, then I usually am only going to say that I didn't find any issues - at most, I might mention the footnote numbers that I checked. This doesn't mean that no spotcheck was done; it just means that no issues were found in the spotcheck. – Epicgenius (talk) 17:40, 23 June 2023 (UTC)

Yeah, looking at the review itself, it does seem a bit lazy, but I was rushed for time and couldn't be bothered to properly format and all that. I can assure you that my fact-checking and analysis of the content was up to par, though it definitely doesn't look like that, and my assurance means little to nothing. I can go back and re-review the entire thing if necessary and make it look a bit better (referring to the possible failure to spotcheck and closely review the article mentioned above), or someone else can do it. Sorry, my mistake. It won't ever happen again. Cessaune [talk] 15:17, 24 June 2023 (UTC)

Removing nominations by retired nominator

Floydian states that they have retired (see discussion here). They still have two nominations listed. Would it be appropriate to open and then immediately close reviews for those two articles, or should they be removed from the list in another way? —Ganesha811 (talk) 22:10, 23 June 2023 (UTC)

I would just remove the GAN template from the talk page. (t · c) buidhe 23:42, 23 June 2023 (UTC)

Done. Thanks for the guidance, wasn't sure if that would mess anything up with the bot. —Ganesha811 (talk) 00:03, 24 June 2023 (UTC)

Sorry, I have covid. Would someone please take Scripto strike?

I never even got started. Went traveling, had a 16 hour layover, caught covid. I may be better in 1 week, or at least that's what I'm told. Right now, however, I feel very unable to edit.  § Lingzhi (talk|check refs) 21:43, 24 June 2023 (UTC)

1964–1965 Scripto strike (nom) is the link for anyone interested. Thebiguglyalien (talk) 02:19, 26 June 2023 (UTC)

Can I get a quick check on...

...this GA review? It was quite quick by what looks like someone new to the GA process. Nothing specific jumps out to me, just the tempo. Jo-Jo Eumerus (talk) 05:27, 26 June 2023 (UTC)

The timing is confusing, a five minute review, but it starts with "This article looks pretty good at a cursory glance, can't wait to get into it". Perhaps it might have been in draft for a bit. A couple of points where I think policy isn't fully understood (eg. image licencing), and a somewhat expansive claim around inline source integrity, but there's a bit of commentary in the review and I don't see any glaring issues with your article the review missed. CMD (talk) 05:46, 26 June 2023 (UTC)

I notice that the reviewer is not extended confirmed, which should probably be encouraged if not a requirement for reviewing. Thebiguglyalien (talk) 05:50, 26 June 2023 (UTC)

Yes, it looks as if they fortunately or wisely picked a particularly tidy article to review. We are torn between wanting GA to be an "intentionally light" process, not bureaucratic, and to achieve a substantially larger throughput of articles than is possible at FAC, with less effort on the part of both nominator and reviewer — and, it seems, wanting GA to be a mark of high quality, a formal review process, with well-defined criteria and some sort of oversight to ensure full and correct application of agreed standards. Well, these two things are in opposition to each other; and in case anyone has been holidaying on Betelgeuse lately, the wait in the GAN queue is getting steadily longer, so that a 6 month delay is not uncommon. If we don't want to train new reviewers, and we don't want to insist on Quid Pro Quo to force nominators to serve as reviewers (what a lot of things we want and don't want!) we need some other way to attract people into the system. At the moment, it's heading for collapse. Indeed, compared to how it used to flow along, it has seized up already. Chiswick Chap (talk) 08:35, 26 June 2023 (UTC)

Yup. I said as much above somewhere. This increased scrutiny of reviews is not particularly encouraging or welcoming to new reviewers. I especially disagree with the now-common practice of bringing reviews here rather than discussing them with the reviewer first. If I were an inexperienced user looking to get into GAN and I saw this talk page, I would be seriously intimidated about even trying. We should be thanking interested newbies and showing them the ropes where they have been imperfect, so they can feel confident about coming back to do more reviews. ♠PMC♠ (talk) 08:52, 26 June 2023 (UTC)

I can only speak for myself, but I think an increase in waiting times is definitely preferable to a decrease in standards. I don't really see the point of having a formal quality control process if it's not going to mean much to pass it. TompaDompa (talk) 16:59, 26 June 2023 (UTC)

The whole point of GA as a process is that it doesn't take much to pass it! It's not supposed to be FA! If we're going to hold GANs to the same level as we hold FAs, we might as well can the process and all move over to FAC. Something which I think gets forgotten at GAN lately is that nominators and reviewers alike are hobbyists doing this for fun. Yes, we want to have as few errors and issues as possible, but perfection across every GA is impossible, and we need to accept that rather than making this process damn-near impossible trying to pursue a level of perfection that even professionals don't achieve. ♠PMC♠ (talk) 23:26, 26 June 2023 (UTC)

I'm aware that it's supposed to be broadly attainable for most articles and somewhat experienced editors. What I'm saying is that it's also supposed to be a mark of quality. If it doesn't function reliably as such, the entire process is a complete waste of volunteer time, and we could scrap it in favour of WikiProject quality assessments (B-class and so on) without a formal review (and WP:Peer review for article improvement, I suppose—though GAN is designed to be a reviewing process, not an improvement process). We don't need to (and shouldn't) hold GANs to the same standards as FACs (and we don't), but we do need to hold them to some kind of standard. We can't let it be a crapshoot whether a WP:Good article is actually up to the standards we expect from a decent-quality article, because if that happens the process is not conferring any benefit. TompaDompa (talk) 23:59, 26 June 2023 (UTC)

I half agree, but I think it's worth remembering that GA has never been a guarantee of quality and yet it hasn't been a waste of time. There have always been weak reviews, and articles passed which clearly should not have, but we've also had many editors gain useful experience from nominating and reviewing. PMC is right that more negative feedback to apparently weak reviews may drive some reviewers away, and Chiswick Chap is right that there's an unresolvable tension between the need for a quick unsupervised process and the desire for a guarantee of quality. It's also always been true that without reviewers willing to review more than they nominate, the process would have collapsed long ago. In that sense nothing has changed; as Thebiguglyalien says below, the increase in the size of the backlog has been gradual, as it always is. The oldest reviews are now much older than would have been the case a year ago, and that's obviously because of the new sort order, but if in fact the number of nominations and reviews is about the same as ever, that has to mean that some other nominations are getting reviewed more quickly. I wish I could make confident statements about the effect the sort order has had on GAN, but I haven't been able to come up with metrics or statistics that I have confidence on the interpretation of. Mike Christie (talk - contribs - library) 00:21, 27 June 2023 (UTC)

The problem is that it's still not entirely clear how we're supposed to respond to reviews if we're not happy with the quality. My first ever GA got a checklist review, so I made a post to this talk page, which largely went ignored, so I posted it at reassessment (back when it was a months-long process) and it was confirmed as a GA there. We should probably write something to the effect of "if you think a review inappropriately promoted an article, list the article at WP:GAR". That might at least temper the in-fighting that's been developing here. A more challenging problem is getting everyone on the same page about how thorough a GA review should be, or if there should be any thoroughness requirement at all. Right now it seems like there's an unspoken one, but as with most unspoken rules, most people either don't know about it or disagree on what it is. Thebiguglyalien (talk) 00:33, 27 June 2023 (UTC)

I can only speak for myself, but as long as this doesn't become a frequent occurrence I don't see why we couldn't use GAR for this purpose. Trainsandotherthings (talk) 02:06, 27 June 2023 (UTC)

(EC) Really? Straight to GAR, even if there's nothing actually wrong with the article? How about talking to the reviewer? How about explaining what seems to be insufficient and encouraging improvement? Sample message: "Hello, thank you for your review! I'm not sure if you have much experience with GA reviews but usually reviewers are looking for some comments about the article. Do you think you could take a look and post some more detailed thoughts? Reviews like [some example] and [some other example] might give you an idea of the kind of feedback we're looking for." Tweak as necessary.

That being said, it seems to me that if an article meets the criteria as-is, there is nothing wrong with passing it with minimal feedback. Why do we expect reviewers to leave detailed criticism for an article if it is already well-written, broad, and properly sourced? Why should we force someone to come up with nitpicks just to make the review look more thorough? (I'm not dunking on people who want to review like that, by the way - I do all my reviews FAC-style, but that's my personal preference). ♠PMC♠ (talk) 02:07, 27 June 2023 (UTC)

All I'd expect personally is a sentence or two. "I checked against criterion XYZ and found no problems." That's it. We can't read people's minds so we don't know what they do unless they say so. Trainsandotherthings (talk) 02:11, 27 June 2023 (UTC)

Ironically I made a proposal at the feedback drive that might have handled the question about the thoroughness of reviews by leaving it up to the nominator to request either full PR-style review or not. It got shot down in flames because people argued that it's up to the reviewer to decide what kind of review they want to do, not the nominator. But obviously it isn't up to the reviewer, or we wouldn't be having these discussions all the damn time. ♠PMC♠ (talk) 02:28, 27 June 2023 (UTC)

Premeditated Chaos I'll point out that the increase in the backlog has been fairly slow. It has been gradually climbing up, but the number of reviews is pretty close to the number of nominations. The problem is that the sort order on the GAN page was changed in January so that newer nominations now cut in line ahead of older nominations. This means that if you have a nomination where it might be hard to find a reviewer already, then it's made even harder because it will sit in the middle of the queue indefinitely. So now the easy ones have even shorter wait times, while the harder ones have even longer wait times. Thebiguglyalien (talk) 18:01, 26 June 2023 (UTC)

Obviously the sort order had something to do with the recent upswing, but our tendency to jump on anyone who fucks up even a little rather than trying to show them the ropes is not helping to attract people to reviewing in any way. And if we're driving away potential reviewers, we're certainly not decreasing the backlog, are we? ♠PMC♠ (talk) 23:13, 26 June 2023 (UTC)

I think we need to strongly promote mentors, and as PMC says not come on people like a pile of bricks, as tempting as that may be. We should also reassess some of the changes made in the past year and see if any need to be modified. Trainsandotherthings (talk) 02:05, 27 June 2023 (UTC)

New to GAN but made my way to this talk page somehow, so I'd like to state a few things (After reading some of the convos above and most of this one)

FIRST: The template is invaluable to new reviewers such as myself, and I think by removing it the (already limited) flow of new reviewers will collapse entirely.

SECOND: A Quid-Pro-Quo system of 1:1 seems like a great idea to me, and I can not for the life of me understand why one hasn't been implemented. Maybe something to DYK, where after your first X # of nominations you have to start reviewing at least one for each additional nomination? Not to name names, but just scrolling through the nomination list there are a *lot* of nominators that don't have anything close to a 1:1 ratio, or even a .5 reviews for each nomination ratio. Or even a .2 reviews for each nomination ratio! It is frankly ridiculous. This is a surefire way to reduce the backlog, though it may result in less good articles being produced.

THIRD: The lightweight standard of GA's makes most sense to be. It's not supposed to be an FA, it's supposed to be a GA. To quote WP:ASSESS: "Useful to nearly all readers, with no obvious problems; approaching (but not equaling) the quality of a professional encyclopedia." Alexcs114 :) 14:10, 29 June 2023 (UTC)

The template is very recent, and was implemented at the same time the reviewer flow collapsed for other reasons. It's probably a net negative, because people who actually use it keep getting come down on like a ton of bricks for "following what they reasonably assumed to be the instructions". GA templates were quite heavily signposted before one was officially added, so it was easy to use them as a new reviewer; I did for my first few reviews before growing out of the structure.

QPQ has never had formal consensus at GAN, specifically because GA reviews are so unlike DYK reviews. At DYK you have to check about four things, and people still get them wrong often enough to have a thriving ERRORS cottage industry. GA is significantly more complex -- even the 'lower standard' version of it (which I don't subscribe to, but what I do subscribe to is fairly different to what most 'higher standard' people seem to) needs way more, and formally implementing a requirement is doable only if GA is outright reduced to around DYK level. This would certainly result in fewer good articles being produced, regardless of what exactly it did to the number of articles with green plus signs.

I've always found the 'paper encyclopedia' quote interesting. WP:NOTPAPER is the highest law of all in ways a lot of people never really notice, simply because we barely even interact with paper encyclopedias these days. Pick one up and go through the articles; almost all of them, as we'd call it, are stubs. Vaticidalprophet 18:31, 29 June 2023 (UTC)

(Replying solely to the third point) Then it would seem to me that WP:NOTPAPER & WP:ASSESS contradict, as almost all GA articles and 99% of FA articles contain more valuable information than their corresponding entry (or lack thereof!) in a "professional encyclopedia" such as Brittanica (Print version now defunct) or World Book. Note however of course that quantity =/ quality, though I think nearly all of our "GAs" or above are of a higher quality than Brittanica or World Book. Alexcs114 :) 20:18, 29 June 2023 (UTC)

Yes, they do. NOTPAPER is policy, and ASSESS was written many years ago and has been a pretty marginal page since. That's not to say the article assessment system isn't important -- well, I consider it important. But I notice my considering it important is fairly unusual, especially for non-peer-reviewed levels (B and below). The specific wording on ASSESS isn't really reflective of how people treat the categories anymore, and hasn't been since about the turn of the previous decade. Vaticidalprophet 09:05, 30 June 2023 (UTC)

Interesting, how can we go about getting that updated then? Alexcs114 :) 13:35, 30 June 2023 (UTC)

May I ask as to what about the "tempo" sticks out to you

-The reviewer, DimensionalFusion (talk) 11:11, 26 June 2023 (UTC)

Probably the fact that the initial review begins "This article looks pretty good at a cursory glance, can't wait to get into it" but marked every criterion as passed, and then you completed the review only seven minutes later. As a couple of people discuss above, there's some tension between the "GA reviewing should be a lightweight process" and "GA reviews should be a reliable mark of quality" schools of thought, but many people who tend towards the latter view look askance at a review which doesn't pick up on any issues in an article, even if it does meet the letter of all of the GA criteria.

Jo-Jo Eumerus has plenty of experience writing GAs, and a quick glance through the article doesn't show me any glaring issues, so I expect that the article is indeed GA standard and you are right to promote – but reviews which appear to have been conducted very quickly tend to raise eyebrows, even if you in fact spent some time checking everything before you opened the review. Caeciliusinhorto-public (talk) 11:27, 26 June 2023 (UTC)

Yeah, it seemed to me that the entire review was quick. But it seems like most people here don't see any issues with this particular review, so I guess I was imagining things. And it's true, folks can be reviewing a nomination before officially initiating the review, which won't show up as a timestamp. Jo-Jo Eumerus (talk) 18:14, 26 June 2023 (UTC)

Yeah. I see some issues have come up because I reviewed while in draftspace instead of publishing first DimensionalFusion (talk) 20:14, 26 June 2023 (UTC)

That doesn't explain the seven-minute gap between initial placeholder and final review. It should take you far longer than that just to spot-check the references. —David Eppstein (talk) 21:04, 26 June 2023 (UTC)

It was in draft for much longer than seven minutes DimensionalFusion (talk) 21:55, 26 June 2023 (UTC)

Thanks for the review. I don't think there was a significant issue, just a mismatch in available information. Your "This article looks pretty good at a cursory glance, can't wait to get into it" note was a helpful indicator in this regard. Best, CMD (talk) 02:59, 27 June 2023 (UTC)

Yea, Johannes Gensfleish.../ 2A02:2F08:B60C:7400:9C5:8EF2:BBCE:9477 (talk) 11:41, 30 June 2023 (UTC)