Jump to content

Wikipedia talk:Good article nominations/Archive 28

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 25Archive 26Archive 27Archive 28Archive 29Archive 30→Archive 33

Evaluating copyvio and plagiarism in GAN

Criterion 2d requires that any article contains no copyright violations or plagiarism. In this aspect of GA reviews, I think it's fair to say that we have failed at a fundamental level. WP:DCGAR has been a debacle. But it's only a symptom of a larger problem and likely only a small fraction of the total number of GAs passed with plagiarism.

I'd like to make two observations. First, hundreds of good article reviews were passed without any meaningful evaluation of copyright or plagiarism issues. Second, the fact that it took hundreds before a problem was noticed indicates that nearly every GA review on Wikipedia either fails to adequately consider copyright violations and plagiarism or is willing to overlook it. Subpar reviews and subpar good articles will continue to be the norm until we can answer some difficult questions:

  • What steps should reviewers take to evaluate for plagiarism? You can say "spotchecks", but that answer has little practical use for a reviewer trying to figure out where to start. And don't say WP:EARWIG; that only discovers the most surface-level copy-and-paste edits. It probably doesn't even detect 1% of all plagiarism. There needs to be a clear and accessible process for checking plagiarism that is made clearly visible to all GA reviewers.
  • How should repeat offenders be handled? The problem with plagiarism is that it's rarely a one-time thing. If substantial plagiarism is suspected in an article, how is a reviewer to respond? The current standard appears to be either to have them fix it or to quickfail them, and then just hope it doesn't happen again.
  • How should insufficient reviews be handled? There are far too many reviews that just list a few prose issues to fix without any meaningful input or analysis or focus more on formatting than content. Go ahead and randomly check any three reviews at Wikipedia:Good article reassessment/February 2023/Doug Coldwell GA list. Odds are that none of them have any meaningful analysis of sourcing or plagiarism. Between nearly 100 reviewers, most of whom are talented, well-established editors, almost none of them put any effort into evaluating sourcing and plagiarism. Even one review without such an analysis is a problem. How can we make it clear to reviewers that this is not optional? How should we respond to reviewers that fail to evaluate for plagiarism? This isn't meant to point blame at anyone, but it's clearly a problem that needs to be solved as soon as possible.

Until these questions are answered, we should assume that most GAs are being passed without being checked for plagiarism, and that an unacceptable number of GAs are passing with significant plagiarism issues. Thebiguglyalien (talk) 03:20, 9 March 2023 (UTC)

And don't say WP:EARWIG; that only discovers the most surface-level copy-and-paste edits - It also fails to differentiate between a Wikipedia article copying a source and a source copying Wikipedia. I've both seen it happen to others and had it happen to me that earwig raises a false flag and then a erroneous accusation is made.
The difficulty with criterion 2d is that a decent source evaluation requires subject-matter knowledge. Either you already have it, or you'll have it by the time you're done doing it. To properly check for plagiarism, you'd have to acquire the source yourself – easy for online sources that don't require a subscription, but not for offline sources that require either a purchase or access to a library containing a copy of it.
This is less demanding at FAC because the labour is divided among editors. One editor does a source review, another the image review, and three will comb through the prose to tighten and clarify it. In a GA review, one editor has to complete all those tasks, albeit with a lower threshold. It is still critical, and for good reason, but how you ensure it, I'm not sure. Mr rnddude (talk) 03:41, 9 March 2023 (UTC)
  • Part of the solution I think is that the low probability of catching it must be made up by an inversely proportional punishment: I doubt Doug maliciously violated copyright policies. He was simply unaware and unwilling to follow them. A vast percent of his achievements have been wiped, with nearly all of his GAs stripped of their status, and many of his articles WP:PDEL'd. Having a policy that significant copyright violators will be mass delisted and PDEL'd might help to reduce occurrence going forward. Other than this, any policies we make rely on the good faith and effort of individual reviewers, and only God above knows how often people will adhere to such. Iazyges Consermonor Opus meum 04:22, 9 March 2023 (UTC)
    My main argument here is that we do know how often people will adhere to such, and the answer is almost never. Thebiguglyalien (talk) 04:36, 9 March 2023 (UTC)
    Exactly; putting a warning out that anyone who violates copyright policy will have their work wiped from history shifts the onus even further toward the most interested party (the nominator) and away from reviewers, who may or may not (and as you say, very likely will not) follow the rules. If the pre-Doug "ah well, you'll get 'em next time champ" standard is allowed to continue, further issues will fester. Aggressive pursuit of consistent copyright violaters is our most useful enforcement tool. Other than this, the only real method I can think of is perhaps a random selection of GAN reviews be quality tested, but that is likely to be insufficient and wildly unpopular. Iazyges Consermonor Opus meum 04:46, 9 March 2023 (UTC)
    What do we do when there are truly abysmal reviews, like quickpasses or serious WP:CIR reviews? I don't think it's unreasonable to treat reviews the same way when they fail to evaluate the sourcing, especially given the consensus that source checks are required per the proposal drive. Thebiguglyalien (talk) 05:02, 9 March 2023 (UTC)
    Truly horrible reviews are often either reopened for second opinion or just G6'd with the article going back in the queue. Hog Farm Talk 05:05, 9 March 2023 (UTC)
    We should probably ask for quickpasses to be redone. However, specific examples of such passes were discussed last month, and proposals to have them redone received little support. CMD (talk) 05:31, 9 March 2023 (UTC)
    We need to start topic-banning people from GAN/DYK the first time they're caught committing plagiarism. If you can't nominate articles, the incentive to plagiarize is significantly lowered. ♠PMC♠ (talk) 05:15, 9 March 2023 (UTC)
  • Could you explain how spotchecks provide little practical use for a reviewer figuring out how to start? How else would they start? What else is needed? CMD (talk) 05:33, 9 March 2023 (UTC)
    It doesn't tell the reviewer what they're actually supposed to do once they find the source. For some new reviewers, I suspect that spotchecks consist entirely of confirming that the source exists and in some way talks about the topic. At what point is a reviewer told they're supposed to check source-text integrity, let alone how to evaluate for WP:SYNTH and WP:CLOP? Thebiguglyalien (talk) 06:33, 9 March 2023 (UTC)
    The first idea that comes to mind then is adding an explanation of spot-checks to WP:RGA. It could cover copyright and text-source integrity (including OR). We could adapt text from Wikipedia:Plagiarism and Wikipedia:Close paraphrasing, and/or just point to them. CMD (talk) 06:57, 9 March 2023 (UTC)
    I think we need something like a friendly "reviewer's lounge" where especially new reviewers are invited to ask questions and can obtain informal advice on things like close paraphrasing. We don't need more help pages, we need more help people. —Kusma (talk) 07:04, 9 March 2023 (UTC)
    We don't need more help pages, we need more help people. Can we pin this to every talk page? Thebiguglyalien (talk) 16:55, 9 March 2023 (UTC)

Last year I realized I should be doing spotchecks, and I went back over my last hundred or so GA promotions and did spotchecks for them, and recorded the results. A byproduct was that I now have statistics from a sample of 109 to say how many errors get past GA reviews if no spotchecks are done. Of those 109, six were by Doug Coldwell; 67 had no issues; another nine had very minor issues (one word needed to be changed or something similar); five had no or almost no sources I could access; 21 had at least one issue that had to be resolved by the nominator; and one I had to GAR and ended up delisting. Mike Christie (talk - contribs - library) 13:41, 9 March 2023 (UTC)

Death, taxes, and Mike Christie making our lives easier with the relevant GA stats. Thebiguglyalien (talk) 16:54, 9 March 2023 (UTC)
  • I agree we should be more strict. We often let copyright problems simmer much too long. The detection of copyright violation is hidden on the GAN page, and not on the nominator's user talk. As such, it's very difficult to detect a pattern of copyvio. There are no signals to other editors to check for plagiarism and explain / warn as appropriate. I once reviewed one of Doug's articles, and naïvely thought I was giving a strong signal when I quick failed it. But of course, that wasn't really visible to other people.
    1. Maybe suggest in the instructions that one should give a (user talk) warning to the nominator for plagiarism
    2. "We need to start topic-banning people from GAN/DYK the first time they're caught committing plagiarism" -> Not decided on this. GAN and DYK are very imperfect in their detection of copyvio, but pushing people away from the process might make it even more difficult to detect copyvio. On the other hand, having a topic ban does give this signal to other people to doublecheck.
    3. Should we track the ratio of successful to failed nominations? Doc managed to get hundreds of articles through the process, but I guess about a hundred also failed. When the ratio is that skewed, there is certainly a bigger problem at hand. Tracking this may also help us find other types of problems with frequent GA nominators.
    4. Wider than GAN, we may need to track when people rack up their second/third plagiarism warning, so that we can manually sample their contributions for problems at an early stage. Hopefully with some automated tool. Too many of the CCI pages are exceedingly long with easy-to-spot copyvio. Often these people have racked up multiple warnings over a longer period of time before a CCI is started. —Femke 🐦 (talk) 18:12, 9 March 2023 (UTC)

My feeling is that DYK and GAN are only a small sample of the Wikipedia content, and a sample that is in both cases somewhat biased towards better-quality content. If we're catching so much copyvio here (and I do regularly catch copyvio in DYK nominations) then it's a problem, but it's a symptom of a bigger problem that far too much copyvio is getting through Wikipedia's general mechanisms for detecting and preventing it. Ideally, we wouldn't have copyvio at DYK and GAN because the problems would have been spotted much earlier. —David Eppstein (talk) 18:52, 9 March 2023 (UTC)

To which I ask: what general mechanisms for detecting and preventing it? Are there such mechanisms for detection beyond "hope that someone spots it"? Thebiguglyalien (talk) 19:07, 9 March 2023 (UTC)
This should be part of New Page Patrol (and often is), but of course that doesn't address problems added to existing articles. —David Eppstein (talk) 19:30, 9 March 2023 (UTC)
There's no silver bullet before we get to CCI, but edit filters/tags could be used to highlight edits adding a lot of content so RC patrollers can have a look. —Kusma (talk) 19:37, 9 March 2023 (UTC)

On this page I think our main concern needs to be how reviewers can be alerted to problems that go beyond the single article they're looking at. One option would be to add pass/fail stats to the review page -- that is, once the review page is created, ChristieBot could add a line somewhere near the top that looked like this: "Nominator's prior GAs: 287 nominations: 245 listed/42 not listed." (Those are Doug's numbers.) The numbers would be links, and clicking them would bring up a page with links to the passes and fails. For an experienced reviewer, particularly one who like me was naïve about spotchecking, alarm bells should go off with that "42 not listed", and they could take a look at the fail reasons. That would minimize the work needed by the reviewer. Mike Christie (talk - contribs - library) 19:51, 9 March 2023 (UTC)

Just an idea: would it make sense to separate "general" failures from "plagiarism" ones or is that too much of a black mark? —Kusma (talk) 20:01, 9 March 2023 (UTC)
I worry that this would create an even bigger stigma against failing a review. I recently quickfailed a review and almost immediately got an angry message on my talk page. I can only imagine the indignation from nominators and the hesitancy to fail from reviewers if failing puts it on an editor's "permanent record". Thebiguglyalien (talk) 20:07, 9 March 2023 (UTC)
I'm sure it would create more resistance, and yes, marking a failure as a copyvio/spotcheck/plagiarism fail would be seen as inflammatory. I think we can't have it both ways: if we want reviewers to know about prior reasons for concern about a nominator's articles, then they have to be recorded in some way. If we don't want scarlet letters, nominators won't have a way to find out about potential problems. I can see good arguments in both directions. Mike Christie (talk - contribs - library) 11:35, 10 March 2023 (UTC)
The DCGar that I failed was failed because of text-source integrity issues. I didn't detect copyvio, but of course it is more difficult if the sources don't match the text. If we had a copyvio marker system, I would have marked it as a non-copyvio fail, which might come off as suggesting it was fine when in reality I wasn't able to check for it. CMD (talk) 12:25, 10 March 2023 (UTC)
I guess instead of mentioning failed reviews on the main GAN page, we could have an accessible list of all reviews belonging to a given nominator. —Kusma (talk) 12:34, 10 March 2023 (UTC)

God this is a mess. I'm going back through old GA and FA reviews, and practically none of them are checking for source-text integrity or original research. And these sub-par reviews are coming from experienced, well-established editors. The most frustrating are ones where reviewers are creating a section for checking sourcing, and then their advice is along the lines of "link to this publisher" and "use em dashes". Besides having nothing to do with the GA criteria, reviewers like this are a huge obstacle in finding sourcing issues. Using the 15 GA reviews I've received as a sample, only a few did any sort of spotcheck, and only one checked for close paraphrasing. We desperately need some form of intervention for GA reviewers, because even really experienced editors are botching this very badly. It's no wonder an editor was able to get a green plus on hundreds of broken articles. I would be surprised if there weren't multiple Coldwell-like editors on the GAN page right now. Thebiguglyalien (talk) 16:29, 10 March 2023 (UTC)

We also need to keep in mind that not every reviewer will explicitly state they checked source-text integrity if it was done informally and no issues were noted. I started checking a few refs on pretty much ever nomination I reviewed several months ago, but haven't always documented that I did so if there weren't any issues. Hog Farm Talk 16:32, 10 March 2023 (UTC)
Given we now have consensus for spot-checks, it should hopefully be a different situation going forward. Are the reviews you are checking general GA reviews, or GA drive reviews? CMD (talk) 16:37, 10 March 2023 (UTC)
(ec) For FAC, the rule is that first-time nominators or nominators who've not had a FAC in years get a spotcheck, which is separate from a source review. See Wikipedia:Featured article candidates/Ignace Tonené/archive1 for one I just finished that failed, and Wikipedia:Featured article candidates/Airport Central railway station/archive1 for one that just passed. If the FAC review just says "source review" then they're not doing these checks automatically, though occasionally a problem will emerge as part of a review and a reviewer will recommend to the coords that a spotcheck be done -- I've certainly done that on occasion. I'm sure nothing on the scale of Doug Coldwell's problems has made it through FAC, though I wouldn't be surprised if some problems get through. Perhaps FAC should do spotchecks more frequently, but that's a topic for WT:FAC.
For GAN, I've taken to using the same format for my reviews as in those FACs -- listing exactly what I've checked. See Talk:Víctor Hugo Zamora/GA1 for an example. How about requiring that a GAN review, even if it lists nothing else, lists the citations that were spotchecked? I have very occasionally passed a GAN without finding anything at all to complain about, and I'd object to having to list all the reasons why I thought the article passed GACR, but listing the spotchecked citations would be quick and would be a prompt to reviewers to actually do the check. It would be somewhat disruptive though -- I imagine a lot of experienced reviewers would get annoyed if we made such a change and they missed the discussion and get nagged about their reviews because they didn't list the checks. Mike Christie (talk - contribs - library) 16:42, 10 March 2023 (UTC)
Looking at the sources is necessary for more than text-source integrity and plagiarism/close paraphrasing checks; especially NPOV and broadness are impossible to check without reading something different from the article under review. I try to make my reviews show that I have read some sources, even if I don't always mention all of them. I guess what I'm trying to say is "read the sources if you can, don't just do spotchecks for copyright reasons". —Kusma (talk) 16:50, 10 March 2023 (UTC)
Here's another example of what I think a thorough source review should look like: Talk:El Palo Alto/GA1. This was a passing review (as I expected going into the review) and didn't find any close paraphrasing but nevertheless did find quite a few (easily fixable) article-source integrity issues. On the other hand, it took a couple hours just for the source checking in that review; do we expect GA reviewers to put that much time into each review? —David Eppstein (talk) 17:01, 10 March 2023 (UTC)
David, it's very worthy, but it's more than halfway towards FAC, i.e. it (and other such changes) are greatly increasing the effort needed at GAN, and heaven knows that reviewers are hard enough to find already. A straightforward Turnitin-style plagiarism check would catch the Coldwells of this world, and many reviewers have for years now run such a thing already. Chiswick Chap (talk) 09:07, 13 March 2023 (UTC)
@Chiswick Chap I have not done a good article review for a long time as I don’t enjoy them but by “A straightforward Turnitin-style plagiarism check“ do you mean the Earwig tool which is well linked and easy when I do DYK reviews? If not please could you link such a straightforward tool in a conspicuous place for GA and DYK reviewers Chidgk1 (talk) 07:08, 14 March 2023 (UTC)
Yes, Earwig has a Turnitin option. Obviously if it gives you a copyvio "hit", you must then also check that they didn't copy from Wikipedia, which often happens. Chiswick Chap (talk) 08:35, 14 March 2023 (UTC)
@Chiswick Chap Thanks I never noticed that option before. Do you know why it is not ticked by default? Chidgk1 (talk) 18:46, 26 March 2023 (UTC)
I think before the Coldwelling, most reviewers assumed that just about anyone taking things to GAN would be just as careful about copyvio as themselves. We have since learned that is not at all the case. I was one of those who did not do comprehensive copyvio checks pre-Coldwell, but now I do them even on articles written by people I am all but certain will not have copyright issues, just to be safe. We will always have issues with copyvio so long as Wikipedia is the encyclopedia anyone can edit. Many people do not even understand what copyright means and what plagiarism is. One solution might be to perform more rigorous copyvio checks for first-time nominators, but that of course means more work for reviewers. Trainsandotherthings (talk) 19:39, 10 March 2023 (UTC)
I've been watching this discussion for several days and as is my usual bent, not commenting. My take on the discussion is quite simply that the "lightweight" review advocated for the process is simply insufficient to catch issues like this. In the real world, writers, proof-readers, fact checkers, and editors are completely different jobs. We here on WP expect a single amateur to be able to do all of those jobs, and that just boggles my mind. In the previous discussion, I explained that I struggle with reviews because I am meticulous (probably overly so). I verify every source I can access and check copyvios as well as I can. It's a hard job and I just do not understand how it can be done otherwise. How can we advocate it being a lightweight process and at the same time be having this discussion? There seems to be a disconnect between expectation and reality, if we want people to actually verify the information it can't be about checking boxes. Just my 2 cents. SusunW (talk) 23:29, 10 March 2023 (UTC)
I agree. I am quite unhappy with lightweight box-checking reviews (for example), and with the advance of reviewing templates that encourage this mentality. —David Eppstein (talk) 23:58, 10 March 2023 (UTC)
While we don't require doing too much more in a review, box-checking reviews are not complete per WP:RGA. CMD (talk) 04:06, 11 March 2023 (UTC)
CMD, "checking boxes" is a euphemism for lightweight reviews. I thought what I meant was obvious. We can't want an in-depth verification and review of copyrighted material and say it's a lightweight process. SusunW (talk) 14:25, 11 March 2023 (UTC)
Yes, because WP:AGF is a fundamental Wikipedia guideline. If you expect each GA reviewer to perform what seems to be the equivalent of an entire FA first-timer review, a task which takes at least five people and probably around a month, you might as well just consign the entire process to the dustbin, right next to PR. ~~ AirshipJungleman29 (talk) 00:08, 11 March 2023 (UTC)
Quite. And GAN was intentionally a light process, assuming collegiality as well as good faith. User:AirshipJungleman29 is right to say that we must not make GA into a one-person FAC (the 5 man-months' of work would take 5 calendar months, then ...). Joking apart, the response to one Coldwell is not to lock the entire process solid for everyone: "Hard cases make bad law", as the legal profession has it. The question we should bear in mind is "How do we intend GAN to differ from FAC?" – and the answer has to be that the process is at least five times simpler, if there's 1/5 as many people to do the work and 1/4 of the time available (ok, that's at least 20 x simpler, then). Chiswick Chap (talk) 08:46, 14 March 2023 (UTC)
I think we fare better overall with a collegial and collaborative process than with one that assumes bad faith unless spotchecks have been passed. People need to enjoy working here. Of course we don't want hundreds of copyvio GAs, but we need to be careful not to make GA reviews so unpleasant that people don't want to do them anymore. (Think of RfA: it does an excellent job at keeping underqualified people out, but an extremely poor job at producing new admins). —Kusma (talk) 09:49, 14 March 2023 (UTC)
I think something else we're forgetting is that Coldwell was well-known to have issues. Like, for years. I deliberately avoided reviewing his work because I had heard about his GANs being bad and him being downright obstructive to work with. And yet it took years for anything to be done because there's no real precedent for removing people from GAN, and it's very difficult to get a "good content creator" sanctioned at ANI because you will be painted as the evil jealous villain. Even with every damn thing we knew about Coldwell's issues, people at the ANI were still arguing that he was doing good work.
Rather than assuming everyone is a bad actor engaging in CV/close para who needs to be checked with a fine-toothed comb, it would be better to make it clear that anyone found to be engaging in significant CV/close para is grounds for being TBANned from GAN, period, appealable only by demonstrating the ability to write without CV. ♠PMC♠ (talk) 00:31, 11 March 2023 (UTC)
This is a good point. I also avoided DCGANs following a poor experience reviewing one article, however I'm not too sure where I could have taken that. Wondering about precedent, there have been restrictions to individual editors for AFC and DYK, so we could craft such a proposal. It would have to go through AN/I though, with all that entails. CMD (talk) 04:10, 11 March 2023 (UTC)
I had no interactions with Doug that I'm aware of, but I will note that in every discussion that came up initially, a huge amount of people were painted as villains, including for reasons such as "having the word train" in your username, and Doug openly accused others of jealousy. Iazyges Consermonor Opus meum 08:37, 13 March 2023 (UTC)
I am not suggesting AGF should be thrown out the window. What I am suggesting is essentially Trust, but verify. We assume good faith barring reason to assume otherwise, but still verify. That means performing basic spot-checks on all nominations. It is not a huge ask to ask reviewers to check 2 or 3 sources quickly for close paraphrasing or failed verification; I'm not contemplating anything like a FAC source review. The time when we could use AGF as reason to assume all nominations were free of copyright issues is behind us. I am well aware that reviewer time is our most limited resource at GAN, and don't want to waste it. But even worse is having reviewers spend time on nominations which are only later found to contain serious issues. Trainsandotherthings (talk) 13:30, 14 March 2023 (UTC)

Just came over to read this discussion, several thoughts:

  1. See User talk:SandyGeorgia#Copyright: Ideas so far; once I get caught up, I hope to better organize that and start an ideas discussion somewhere. Not quite ready yet as we are still uncovering issues at the talk page of WP:DCGAR. A centralized discussion of the various ideas for improving the copyvio situation should be had in about another month, once we've all hashed out preliminary ideas in various places.
  2. An early priority should be for someone to write up somewhere all of the shortcomings of Earwig; we still have people running Earwig on DC articles and pronouncing them clean. There is a very weak understanding of what Earwig can and can't detect.
  3. As GA folks discuss how the process can better detect copyvio earlier in GAs, please keep in mind that copyvio is but one subset of the serious policy problems in DC work, and every kind of policy problem made it through GAN in shocking amounts. Original research, failed verification, non-reliable sources, notability not met-- the works. I have now been through, what, a couple hundred maybe, DC articles, and a good number of the GAs, and there are blatantly non-reliable sources throughout that were not even queried by GA reviewers. Please do not think of the problem here as only related to copyvio.
  4. I believe (but I may be repeating myself :) the answer to the many queries above about where one might have taken the problems they noticed sooner so that others, too, would have realized, is obvious. GAN needs Coords. It would be their job to notice trends, they would be a place others could take concerns, they would theoretically notice when one reviewer is consistently pushing another editor's GAs up the line without adequate review, and so on. A GA continues to be only as good as the one reviewer who pushed the button, and not all reviewers are Ealdgyth. (Not watchlisting this page, so pls ping if a response from me on something is needed.).

That's all I've got for now, SandyGeorgia (Talk) 14:15, 16 March 2023 (UTC)

On the Earwig thing: I thought it was well-known that Earwig can detect blatant copy-and-paste plagiarism but that it doesn't help much with WP:Close paraphrasing since it detects identical sequences of words but cannot account for synonyms, minor restructuring, altered grammar, and the such. Likewise, I thought it was well-understood that Earwig cannot tell if Wikipedia has copied a source or if the source has copied Wikipedia, and that it cannot tell whether it's a problem that there is verbatim overlap between Wikipedia and one of the sources. I have two examples that I like using:
  1. Earwig flags Wikipedia's article on Robinson Crusoe as overlapping significantly with Encyclopædia Britannica's article on the same because both contain the full 68-word title. That's of course not a problem—the article wouldn't be complete without the full title, and it obviously cannot be rephrased since it would no longer be the actual title.
  2. Earwig cannot tell that To go boldly where no one has gone before is a WP:Close paraphrasing of To boldly go where no man has gone before. Swapping the order of the second and third word and replacing "man" with "one" has made the sequence of words non-identical, even though the former is very obviously a minimally rephrased version of the latter.
I think these examples could be useful in helping people understand the limitations of Earwig. TompaDompa (talk) 14:33, 17 March 2023 (UTC)
None of those Earwig shortcomings should be assumed to be "well known", even among highly experienced editors, even admins. In working on WP:DCGAR, I have come across numerous experienced editors claiming a DC article was fine because it cleared Earwig. SandyGeorgia (Talk) 14:52, 22 March 2023 (UTC)
I also know that Earwig doesn't check books and PDFs properly. So, google books, archive.org materials, newspaper.com clips, etc. aren't checked properly by it. Maybe it has to do with whether they are OCR encrypted? But, as we all recognize I am not technically inclined, I don't know why it doesn't, I just know that it doesn't. Also, clear to me is that it doesn't check translations, so one can verbatim translate and it won't catch that either. SusunW (talk) 15:13, 22 March 2023 (UTC)

Concerns with a review

User:Vdfoxcc's first action came at 23:47 UTC yesterday, about half an hour before I write this, when they commented at Talk:Unit fraction/GA1. Since then, and in approximately ten minutes, they have reviewed and attempted to pass Talk:Hogwarts Legacy/GA1 with no comments. Thoughts? ~~ AirshipJungleman29 (talk) 00:15, 29 March 2023 (UTC)

I conducted a thorough review and it passed the requirements Vdfoxcc (talk) 00:19, 29 March 2023 (UTC)
There's no review currently on the review page. -- ZooBlazertalk 00:25, 29 March 2023 (UTC)
My review speaks for itself Vdfoxcc (talk) 00:26, 29 March 2023 (UTC)
It ... certainly does, yes. Big fan of this rather grandiose edit summary, by the way. ~~ AirshipJungleman29 (talk) 00:29, 29 March 2023 (UTC)
This appears to be either a CIR case, quite enthusiastic trolling, or some mixture of both. Either way, the "review" should be reverted. Trainsandotherthings (talk) 00:31, 29 March 2023 (UTC)
Can I not review the article ? There are no pre-reqs listed on the GA article Vdfoxcc (talk) 00:33, 29 March 2023 (UTC)
Right, that's them blocked. Is CSDing the review the normal procedure? ~~ AirshipJungleman29 (talk) 00:36, 29 March 2023 (UTC)
I'd say so, yes. Trainsandotherthings (talk) 00:36, 29 March 2023 (UTC)
I've indeffed for clearly NOTHERE trolling. Feel free to revert things back to before the shenanigans. ScottishFinnishRadish (talk) 00:36, 29 March 2023 (UTC)
I've fixed the Hogwarts Legacy talkpage and cleared the nominator's usertalkpage. CMD (talk) 01:33, 29 March 2023 (UTC)

Excerpts in GAs

A large proportion of Energy in Turkey, a 2021 GA, consists of excerpts from other related articles. Excerpts are not mentioned in the GA criteria, but I do feel that their inclusion might compromise stability and cause issues such as those outlined at H:TRANSDRAWBACKS (unfortunate name for a shortcut btw). What do people think about the inclusion of excerpts in GAs? ~~ AirshipJungleman29 (talk) 14:20, 29 March 2023 (UTC)

Pinging the page creator Chidgk1 for their input. ~~ AirshipJungleman29 (talk) 16:43, 29 March 2023 (UTC)
I think the usefullness of not having to keep multiple articles up to date outweighs the drawbacks unless the subject is very static such as something historical Chidgk1 (talk) 13:59, 30 March 2023 (UTC)
I think you're probably correct that they tend to make maintenance difficult; on the occasions when I've come across them they have a bad habit of not being quite what is wanted as a topic summary. Unless it is *certain* that the original and the transcluded copy *cannot and must not diverge*, then they are basically undesirable. Their presence does not mean that the GA was wrongly awarded or that a GAR is needed, but it might be as well to see if they would be better Subst:ed rather than transcluded, and then perhaps edited to fit their new context better. Chiswick Chap (talk) 14:26, 29 March 2023 (UTC)
I tend to think if you are transcluding anything from another article, it's because it's going to change a lot. I don't see summaries of items of a full topic like this to be suitable. Lee Vilenski (talk • contribs) 15:16, 29 March 2023 (UTC)
One could interpret the stability criterion as precluding large amounts of the article being excerpts, and regardless it's poor practice. If most of an article is excerpts, you have to wonder 1. if it should be an article, and 2. if it could really be considered a good article at all. When we think of a GA, we picture something someone created or improved, not stitched together from other articles. As mentioned, this isn't anywhere in the GA criteria, but I think it absolutely goes against the spirit of what a GA should be. Trainsandotherthings (talk) 23:03, 29 March 2023 (UTC)
As in previous discussions, I am generally against excerpts. If nothing else, it makes the GA template's record of the article at the time of the review almost meaningless. I would recommend all excerpts be substituted, although as per Lee it is very rare that an excerpt will be well-written for its new location. CMD (talk) 01:02, 30 March 2023 (UTC)
I think more to the point is that if the information you are transcluding is stable, then there's no reason to transclude it (as it won't change rapidly). If it needs to be transcluded, it isn't stable. I don't see how something that is transcluded in this way can meet the criteria. Lee Vilenski (talk • contribs) 14:03, 30 March 2023 (UTC)
The real world is not stable - Energy in Turkey or any of the excerpted articles might need changing again in a couple of months after the election Chidgk1 (talk) 14:09, 30 March 2023 (UTC)
A couple of months is NOT a rapidly changing event. We shouldn't be transcludng just because something changes periodically. Lee Vilenski (talk • contribs) 14:14, 30 March 2023 (UTC)
@Lee Vilenski It is unreasonable to expect me to be copying and pasting between multiple articles every time something changes and even if I did it is more likely I would get something wrong Chidgk1 (talk) 14:19, 30 March 2023 (UTC)
Why exactly would you copy and paste between articles in this way though? Isn't the main article supposed to be a summary of the child article? If something is super changeable, perhaps the item should point to the child article and give an overview. Lee Vilenski (talk • contribs) 14:26, 30 March 2023 (UTC)

Two GAN error pages to watch

I'd like to request that any GAN regulars who are willing to help with occasional clean up watchlist two error pages: User talk:ChristieBot/GAN errors and User talk:ChristieBot/Incompletely moved pages. Both are currently showing errors, which I haven't fixed, partly so I could post this message explaining them. The first one is complaining because the deletion of a GAN page has left the talk page in an incorrect status -- it says a review is under way, but it isn't, and the bot doesn't know what caused the problem. The solution is to remove the "onreview" status from the talk page. The other page lists moves that left behind GAN sub-pages. The first bullet shows the move that was done; the indented bullet shows the potentially problematic un-moved GAN page. The fix is to move the subpage to match the parent. If the parent has been moved back, as happened with 2013 Hattiesburg, Mississippi, tornado, for example, then of course the move shouldn't be made. I know that CMD watches these pages and fixes errors; I do some fixes myself, but was hoping to recruit a couple more folks to do the same. The first kind of error is fairly rare -- every few days at most. The second kind shows up every day or so, and isn't at all urgent; it's a backlog job. The second one is helpful because fixing these makes the GA stats reporting more accurate. Thanks for any help. Mike Christie (talk - contribs - library) 18:41, 30 March 2023 (UTC)

Shushugah fixed the incomplete moves; thanks! When the moves are done the page or section can be blanked ready for the next errors; I don't think we need a log other than the page history. Mike Christie (talk - contribs - library) 19:04, 30 March 2023 (UTC)
@Mike Christie how careful should I be when a redirect was made, and then reversed? I personally don't think the GA link matters so much as long as it exists and a user can reasonably find the discussion on either old/new target names' subpage? ~ 🦝 Shushugah (he/him • talk) 19:06, 30 March 2023 (UTC)
The worst problems are when the move is because a dab page is being put in place, so if you don't do the move the GA subpage ends up under the dab page rather than a redirect. You're right that if a redirect is in place a human (and sometimes a bot) can figure it out, but what often happens is that after a move another change is made -- the "main" article under a dab is usurped, or something like that -- and the subpage gets harder to connect to the original page it was a review of. If they always move together then this never happens. When I did the mass clean up back in January I think out of several thousand un-moved subpages something like four hundred or so were sufficiently non-obvious I had to do them by hand, and something like fifty or a hundred of them took some real investigation to decode the history. So it's not always necessary, but I think it's safer just to move all of them. Mike Christie (talk - contribs - library) 19:13, 30 March 2023 (UTC)
@Mike Christie: currently the errors being shown on User talk:ChristieBot/Incompletely moved pages are because the article has been deleted – can the review pages just be CSDed as WP:G8 or is there a reason to keep them? Caeciliusinhorto-public (talk) 09:45, 31 March 2023 (UTC)
That's interesting - deletions don't normally show up there. I think it's because it was recreated from a draft and then deleted again; I'll have to see, and stop those from appearing. I'd prefer the review pages and the talk pages of a deleted page to stay -- the talk page is tagged with {{G8-exempt}} to stop it from being deleted. Sandy's been adding the tag as she goes through the myriad Doug Caldwell GAs, when she ends up flagging an article for CP deletion. Mike Christie (talk - contribs - library) 10:11, 31 March 2023 (UTC)
Okay - I've tagged the review pages as G8 exempt but will leave them on the ChristieBot list for now - there's no point taking them off if the bot's just going to try to relist them... Caeciliusinhorto-public (talk) 10:40, 31 March 2023 (UTC)
I'll remove them in a moment. The bot only looks at the last 24 hours of activity when it runs, once a day, so it shouldn't list them again. I don't think the GAN pages are at risk of being deleted; they have a parent page (which is exempt) so they're not eligible for deletion as the child of a deleted page, but it's certainly harmless to tag them. Mike Christie (talk - contribs - library) 11:00, 31 March 2023 (UTC)

Implementing changes to GA sourcing requirements

Previous discussion "Should everything be cited?" was archived; my understanding is that there was consensus to require that everything (save the usual exceptions) be cited inline, which the current WP:GACR do not make clear, and which WP:GACN currently contradicts (it allows WP:GENREFs).

Shall we discuss implementation? Seems that the most popular ideas were to add something inspired by SG?inline citations, or to imitate the WP:FARC wording and link to WP:WTC. DFlhb (talk) 18:54, 29 March 2023 (UTC)

I think it would be best to declare consensus from a formal RFC with official closure, rather than a talk page discussion with uncertain ending, before discussing implementation. For the record, I do think it should be implemented, I'd just like to see it done properly. ~~ AirshipJungleman29 (talk) 19:00, 29 March 2023 (UTC)
Fine by me; how should that RfC be phrased? How about:
Should the WP:GACR require that everything be cited inline, save for the usual exceptions?
DFlhb (talk) 19:04, 29 March 2023 (UTC)
Yes, inline, and using <ref>...</ref> tags. I'd love to add "and in a citation template" but that remains a bridge too far, probably. Chiswick Chap (talk) 19:02, 29 March 2023 (UTC)
No, not "using ref tags". And definitely not "in a citation template". List-defined references with the references defined using the {{r}} template rather than ref tags are completely within our guidelines. For that matter so are {{ran}}/{{rma}} formatting for footnotes and the things they go to, although I think using that for the main references of an article would be pretty unusual. And GA requirements have historically not even required that references be formatted consistently with each other, let alone that the formatting be done with a citation template. There are some citations that citation templates are bad at handling, and some editors with strong preferences for avoiding citation formats deliberately (for one reason, so that they do not get filled with cruft useless id numbers by bots). These formatting consistency issues are a completely separate issue from the existence of adequate inline sources and should not be conflated with it. If we want to tighten up our rules for reference formatting (and I don't see why this is a problem in need of a fix) that should be a separate discussion. —David Eppstein (talk) 19:52, 29 March 2023 (UTC)
I'm thinking of obsolete rubbish like inline parentheses "(Bloggs & Bloggs 1967)" with no tags or templates of any sort. Chiswick Chap (talk) 20:04, 29 March 2023 (UTC)
That format is deprecated for citations. It is also not "obsolete rubbish", at least not in the world beyond Wikipedia: there are many well-established academic journals that demand this style. But I am EXTREMELY wary of attempts to reiterate its deprecation everywhere citations are mentioned, because it is very easy for it to leak over into attempts to legislate the ways in which article text can refer to people who published things. If it's a parenthetical citation in the format you give, that could be removed from a sentence without affecting the grammar of the remaining sentence, then it's in the deprecated format and could reasonably be converted to a footnote. If it's a hypothetical sentence like "Asimov (1942) first codified the Three Laws of Robotics", and we're going to use the deprecation of footnotes as a bludgeon to insist that that exact format of the sentence subject is wrong and must be replaced by "Asimov (in 1942)..." or "In 1942, Asimov..." or the like, then that's not a citation and that's going well beyond what the deprecation of a certain reference format should be used to do. You may well handle this distinction appropriately but I don't trust inexperienced reviewers to do so and I don't want to encourage them by putting this into the GA rules. Besides, again, that's a formatting issue, and completely irrelevant to the separate issue of requiring claims to have citations. —David Eppstein (talk) 20:14, 29 March 2023 (UTC)
Interpreting that strictly, you could argue the FA Providence and Worcester Railroad fails that requirement by using Sfn in places. This needs to be thought out further. Trainsandotherthings (talk) 23:05, 29 March 2023 (UTC)
I'm a fan of the rule DYK uses, where at least one citation per paragraph is expected. That doesn't have to be exactly what we do, but I think we can all agree anything less than that would not meet the GA criteria. I also agree that if we're going to make a formal change an RfC is called for. In particular you'd have to see what the community consensus is on general references - I personally never use them, but I think some will argue they are fine for short articles, even for GAs. Trainsandotherthings (talk) 22:51, 29 March 2023 (UTC)
Personally I would prefer a wording similar to FAC requirements but allowing general references for articles under 500 words. I don't feel strongly on the latter point though. (t · c) buidhe 23:35, 29 March 2023 (UTC)
One per paragraph is an outdated and somewhat misleading guideline. What matters is the text is sourced inline in a way that maintains the text–source relationship, be it through cite templates, list-defined references, sfn, etc. CMD (talk) 01:06, 30 March 2023 (UTC)
One thing I never really got is how we're supposed to read a "citation at the end of each paragraph" type article. Are we expected to assume that a citation supports the entire paragraph? Or is it just a mess of "here's a citation that has something to do with one of the ideas somewhere in this paragraph"? In-line citations don't exactly lend themselves to saying what specifically it's supposed to support unless they're at the end of every sentence (and even that can still be vague). Thebiguglyalien (talk) 01:12, 30 March 2023 (UTC)
Theoretically yes, I believe the idea is that it supports the whole paragraph. It feels like the sort of convention that would emerge while converting general sources to in-line sources, back in the early days. In-line citations should lend themselves to specifically supporting the relevant text, if they don't then that's an issue to raise in the GAN. CMD (talk) 01:36, 30 March 2023 (UTC)
In my ideal world, every citation in the body text would support all of the text from the previous citation. If there's a single citation at the end of each paragraph, therefore, everything in the paragraph should be verifiable from that one citation. (The "citation at the end of a paragraph supports some but not all of that paragraph" is not unique to that formatting; it happens in every system except for "explicit inline ref after every discrete claim of fact" which leads to unreadable stuff like "Julius Caesar (12 July 100 BC[1] – 15 March 44 BC[1]) was an ancient Roman[1] general[1] and statesman.[1]"!) Caeciliusinhorto-public (talk) 10:55, 30 March 2023 (UTC)
Your ideal world is the current in-line citation system, baring BLUE exceptions. CMD (talk) 15:35, 30 March 2023 (UTC)
Well, my ideal world is how the system currently works in theory; in my ideal world it would also work in practice! Caeciliusinhorto-public (talk) 15:39, 30 March 2023 (UTC)
This is exactly the sort of thing that I would hope GA reviewers are checking: when we have a citation following multiple claims, does that citation support all of the claims (as it should, in the ideal world) or only some of them? We should go farther than the typical DYK-level "are there footnotes" check, to instead look at "do the footnotes do what they are supposed to do". —David Eppstein (talk) 19:28, 30 March 2023 (UTC)
  • Before the discussion loses steam: please comment on what you think of my proposed RfC wording suggested above, whether you think such an RfC should be held, and where you think it should be held (here?). We've gone a little off topic here, but I believe this proposal matches current accepted GAR practice, and shouldn't be too controversial. DFlhb (talk) 21:13, 31 March 2023 (UTC)
    I'm not opposed to that wording for a rfc. My preference would be to allow general references to articles under 500 words , but that could be a separate rfc. (t · c) buidhe 23:13, 31 March 2023 (UTC)

User talk template for drive by GA nominators?

Has anyone looked into making a user talk template for editors that nominate without substantial contributions to an article? I'm noticing more of these nominations, and it would be helpful if we had a template that could point these editors in the right direction without being bitey. Of course, it would also be helpful if we had a way to detect these quickly, but I know there are some difficulties around what exactly such a system would look for (edit count, bytes added, etc). Thebiguglyalien (talk) 03:37, 31 March 2023 (UTC)

Edit count and contribution count to an article is certainly something the bot can find, but those are too slow to do every twenty minutes. However, the bot could find that information the first time it sees the nomination, and could then post a note to this talk page, saying something like "New nomination Example was nominated by user User:Example 12345 who has X total edits and Y edits to the article", for whatever values of X and Y we decide should trigger the message. Mike Christie (talk - contribs - library) 12:16, 31 March 2023 (UTC)
I understood that an editor that has not substantially edited the article can nominate it anyway, as long as first he asks to confirm that nobody has a problem with that. If the main writers give it the OK, or do not answer after a reasonable time, then he's free to nominate it if he considers it is in good shape (and of course, he would be the one to answer any concerns that may take place during the review, as any regular nominator). I'm not sure if a bot can detect if a user has done that step. Cambalachero (talk) 13:43, 31 March 2023 (UTC)
Cambalachero, I believe that the process you describe was ended with a recent GA policy change. Hog Farm Talk 13:49, 31 March 2023 (UTC)
I think Mike's idea is very good to have a message stating what the details are on nomination. As a reviewer, you can then tell if someone has done the suitable steps to let major contributers know (or not). Lee Vilenski (talk • contribs) 13:51, 31 March 2023 (UTC)
How about X = 500 and Y = 5? Either would trigger the message. And the message would have to be on this page, I think, unless someone has a better idea. Mike Christie (talk - contribs - library) 02:44, 1 April 2023 (UTC)
This feels like the type of thing that would be recorded at Wikipedia:Good article nominations/Report. Though I don't know whether that would be more efficient or less efficient. Thebiguglyalien (talk) 02:56, 1 April 2023 (UTC)

Template in auto-filled review page leads to perfunctory reviews

I've noticed a significant increase lately in the number of "reviews" of Good Articles that are just perfunctory box-ticking, without any evidence of in-depth checking. Examples: my most recent two GA passes, Talk:Unit fraction/GA1 and Talk:Beckman–Quarles theorem/GA1. This started happening more or less at the same time that the "create review" link on GAN started populating the review page with a review template with little checkboxes for the reviewers to tick. My strong suspicion is that, by making this template so prominent, we are strongly encouraging the reviewers to tick checkboxes and do little or nothing beyond that. Is that the level of review we want for GA? Especially given the recent Coldwell debacle? Could we maybe revisit this decision to push template-filling in place of written reviewing? —David Eppstein (talk) 21:42, 1 April 2023 (UTC)

I've actually found the autofilled templates to be annoying, it actually makes a review take longer to complete since all the comments get in the way of trying to actually add my comments and filling out parameters. It's enough to have the "GA toolbox" in my opinion, which links to the instructions page and the list of review templates. Trainsandotherthings (talk) 21:46, 1 April 2023 (UTC)
I just delete the whole template before starting the actual review. —David Eppstein (talk) 22:35, 1 April 2023 (UTC)
I fully agree; most bad reviews recently seem to come from the template. A far less important issue is the slight annoyance of having to remove the template every time I start a review, as I prefer to structure my reviews differently (while still filling out a checklist to ensure I really check all of the criteria). —Kusma (talk) 21:48, 1 April 2023 (UTC)
Pinging @Brachy0008 and The person who loves reading: to have their opinion on this discussion. Onegreatjoke (talk) 23:40, 1 April 2023 (UTC)
If you must. I was hoping to avoid pointing fingers. I have good faith that those reviewers thought they were doing it right but simply did not have enough guidance on what a better review would look like. (In case they come here and want a suggestion: Kusma's Talk:Dehn invariant/GA1 was much more the sort of thing I would like to see in a GA review. Critical where it needed to be, and with suggestions that led to real improvements in the article over the course of the review.) —David Eppstein (talk) 00:26, 2 April 2023 (UTC)
This is the crux of the issue. Every step of the GA process is hampered by opaque processes and instructions, making it very difficult for less experienced editors to get involved. We can make all sorts of cosmetic changes, but it won't help anything unless we make the process simpler. We need a guide for new reviewers that says "here's what you check, here's how you check for it, and here's what it should look like when you're done". Obviously something like this wouldn't be binding or required, but it would make all the difference for a new reviewer. Thebiguglyalien (talk) 01:19, 2 April 2023 (UTC)
What. I was just enjoying April Fools! Brachy08 (Never Gonna Give You Up, Never Gonna Let You Down) 06:18, 2 April 2023 (UTC)
Thank you for ping me. Sorry about reviewing badly; I can't actually find mistakes in this article, so I can only pass it. Also, I check some similar mathematics articles (good articles) before reviewing, and find that it's very similar to this article. I checked some citations (I can't open some of them), and they are all reliable. I thought that I only need to check each box and give a short comment at that time, and I look at some reviews from more experienced users. Then, I find that I need to write lots of comments to improve this article before deciding whether to pass or fail or on hold. I agree because this misleads me at the first time I review an article (this article). The person who loves reading (talk) 01:31, 2 April 2023 (UTC)
I am strongly in favor of keeping the auto-generated template. It allows us to simplify our instructions and even add more instructions in the preloaded template. Removing it is easy enough for experienced editors as Kusma mentioned. We need a better way to ping/alert new reviews in good faith, just like it happened here. It's a lot easier to give feedback on an incomplete article than it is on a basically already Good Article, so I even empathize with @The person who loves reading. ~ 🦝 Shushugah (he/him • talk) 02:23, 2 April 2023 (UTC)
What the template doesn't do a good job at is making people engage with the article content and show that they have done so. If you look at the second review mentioned above, there is nothing specific to the article in that review. Even the praise is unspecific. It should be possible to infer from a GA review that the reviewer has read the article. —Kusma (talk) 06:53, 2 April 2023 (UTC)
@The person who loves reading: To be clear, all you strictly speaking need to do is make sure that the article meets all the WP:Good article criteria before passing it. That includes checking that all the sources are reliable, checking a sample of the sources for verifiability and plagiarism, and so on. If you have done all of that, you have reviewed the article properly. More in-depth comments are not mandatory, but they are best practices since they make it easier to tell whether you have done the things you are supposed to when reviewing a WP:Good article nomination. Explanatory comments that outline your findings become evidence of in-depth checking, to borrow a phrase from above. TompaDompa (talk) 03:15, 2 April 2023 (UTC)
Thank you very much for explaining this! The person who loves reading (talk) 03:25, 2 April 2023 (UTC)
What even is a "bad review"? If you take up an article of FA-standard at GAN, are you expected to find improvements? I think in some cases, box-ticking is perfectly acceptable, and I don't really understand why people need comments on articles that are very clearly GA-standard. If it meets the core set of editorial standards that are the criteria, why faff about with improvements that aren't needed? ~~ AirshipJungleman29 (talk) 11:14, 2 April 2023 (UTC)
A review that does not include any evidence that the reviewer has read the article certainly is a bad review. —Kusma (talk) 11:25, 2 April 2023 (UTC)
Box checking is specifically against our guidelines. I note that neither review in question explained their spotchecks, one mentioned there was spotchecks but the review doesn't say what the spotchecks looked at or what they found. CMD (talk) 11:55, 2 April 2023 (UTC)
Fair enough. ~~ AirshipJungleman29 (talk) 12:06, 2 April 2023 (UTC)
You can also give feedback without suggesting improvements. For example, "I really liked the background section, it helped me understand the context although I had never heard of the Mongol-Russian wars before". Once I said something like "This section reads a bit dry"; I would have passed the article unchanged but the nominator still went ahead and made it less boring. You know, just have a friendly chat with the author about the article. —Kusma (talk) 14:04, 2 April 2023 (UTC)

Seeking review of first review

WP:GAN/I#R3 recommends that users doing a GA review for the first time ask a good article mentor to review their review. I am looking for someone to review my review of 1912–1913 Little Falls textile strike. Picking someone at random from Wikipedia:Good article help/mentor seems like an ineffective way to do that, so instead I am posting here to find whoever might be available to review my review. Apocheir (talk) 17:43, 2 April 2023 (UTC)

Hi @Apocheir:, I will take a look today. Trainsandotherthings (talk) 18:05, 2 April 2023 (UTC)
From a basic overview, the only thing I see missing is a check for copyvio. I don't expect you'll find any, as the nominator is quite experienced, but especially following Wikipedia:Good article reassessment/February 2023, it is crucial to perform some basic spotchecks. I do see some sources are paywalled or offline, which may complicate this, but a number are freely accessible such as the International Socialist Review articles. All that is needed are a few spotchecks, in other words checking that the sources back up claims they are cited to, and the Wikipedia article does not copy or too closely paraphrase sources. Typically a quick check of the article via Earwig's Copyvio Detector is also completed. Trainsandotherthings (talk) 18:12, 2 April 2023 (UTC)
Agree with Trainsandotherthings, the review seems to be quite good, with comments showing thought on prose quality, neutrality, and article structure. All that's needed is to check a few sources yourself to make sure there is no copyvio, and that the sources actually support the cited information. Frankly, given the stuff about chatgpt emerging, we should consider such a check would also helpfully show if the sources actually exist... CMD (talk) 23:50, 2 April 2023 (UTC)

Question about splitting an article

Hi. Hope this isn't a frivolous question. I haven't done GANs in years, and am not very well-versed in current prevailing opinions etc. So, the article is Mi Shebeirach. It's an ancient Jewish prayer/song. There is a contemporary version written relatively recently that apparently is massively popular in Reformed synagogues (and perhaps others, IDK). It's to the point where I saw one source that said it is sung in most if not all almost every week. [Crap, I lost the source, I suck.] My point is, discussion of the modern version seems to be large enough in the current GAN that the article could be split into Mi Shebeirach and Mi Shebeirach (Debbie Friedman song). The latter might even take up more space than the former, haven't checked yet... I haven't checked WP:CLOP yet, but if that's OK, this article is almost certainly "Pass GA", IMHO. So... splitting... what should I do? Tks in advance § Lingzhi.Renascence (talk) 00:37, 5 April 2023 (UTC)

It's quite a short article, what benefit would a split have? CMD (talk) 00:45, 5 April 2023 (UTC)
Perhaps none. It just seems odd to me that the modern version seems to almost overwhelm the article. Plus... perhaps in a nebulous way... if it were split, the article about the new song could get its own GA, kudos to the (sadly, deceased) songwriter, etc. § Lingzhi.Renascence (talk) 00:49, 5 April 2023 (UTC)
[Creator/nominator] I'm undecided myself on whether to split, but I think one thing to consider here is that, a bit counterintuitively, the original Mi Shebeirach isn't very important (relatively speaking). Millions of Jews hear it every Saturday, yes, but it's not really one of the "greatest hits". Anecdotally, I've talked to a number of Jews from both liberal and Orthodox backgrounds about this, and none had any familiarity with any Mi Shebeirach other than the one for healing, and in that regard almost all were primarily familiar with Friedman's version. I'm pretty confident this isn't just some bias on my part, based on the further evidence that 1) none of our Jewish liturgy editors—who tend to write from an Orthodox perspective—had seen fit to create this article till I did and 2) there are literally three academic papers ever published on any Mi Shebeirach other than the one for healing. (None of which are cited in the article, because they're in Hebrew and not available online, even for money... If I ever take this to FAC I guess I'll have to track them down and beg someone who speaks better Hebrew than I, but for now I've cited some sources that cite them.) Anyways, that's not arguing for any outcome, because like I said I'm still undecided, but just wanted to be clear on the significance (or lack thereof) of the older prayer. -- Tamzin[cetacean needed] (she|they|xe) 01:43, 5 April 2023 (UTC)
All things considered, perhaps it's best just to proceed as is. Tks  § Lingzhi.Renascence (talk) 05:31, 5 April 2023 (UTC)

Implementation of proposal 2: spotchecks

I'm talking to a new nominator about possibly starting to review, and after looking at the proposal drive he asked why proposal 2 was not reflected in the criteria or instructions. It's not mentioned in the default template either. Wasn't the intention of that proposal that some sentence such as "spot check a sample of the sources in the article for source-text integrity" should show up somewhere? I don't see anything like that at either WP:GAN/I or in WP:GACR. Mike Christie (talk - contribs - library) 10:57, 29 March 2023 (UTC)

What's the bar for "significant contributor"?

Per this change, drive-by nominations are not permitted any more. I was about to review AC/DC; the nominator, Vaughan J., has contributed less than 2% of the article, and has a total of 66 edits to it. Few of those edits are substantial. They did not post to the article talk page before nominating the article. Would this be considered a drive-by nomination? It seems borderline to me. Mike Christie (talk - contribs - library) 15:41, 3 April 2023 (UTC)

Similarly, Neighbours: The Finale was nominated by U-Mos, who has 3% of the text in the article, though it's clearer in this case that they have edited it frequently. Still, the main editor is obviously Raintheone, who contributed most of the text and has been recently active. I'm not clear how to judge if these are "drive-by" or not. Mike Christie (talk - contribs - library) 15:48, 3 April 2023 (UTC)
Percentage of text written seems a bit of a blunt guide, seeing as my contribution to that article is to thoroughly copyedit/check amend sources etc. after its move to the mainspace (having been collaboratively drafted by two other editors). The guidance does not require GA nominators to be the single most significant contributor, at any rate. U-Mos (talk) 18:07, 3 April 2023 (UTC)
I would say that any borderline nominations, where the authorship is either less than 10% or not one of the top five editors, should be preceded with a post on the talk page. In any case, the AC/DC article is definitely not GA standard. ~~ AirshipJungleman29 (talk) 15:56, 3 April 2023 (UTC)
I can't find any real discussion on this in the archives, but in the case of AC/DC, Vaughn J. has not contributed any actual content to the article that I can find; they have added a couple of sources and done some minor wording tweaks. I would not consider that a signficant contribution - especially as, with the exception of No-Bullet, all of the editors with more contributions have been active on wikipedia in the last few months. 66 edits to the article isn't "drive-by", but nor have they reached the point of "significant contributor". Caeciliusinhorto-public (talk) 16:08, 3 April 2023 (UTC)
I think that, like most things, the borderline cases are going to come down to a judgement call. There's no magic number of edits or contribution percentage where a person becomes a significant contributor. In the cases of AC/DC and Neighbours: The Finale, I would argue that neither amounts to "significant" contribution. Thebiguglyalien (talk) 16:18, 3 April 2023 (UTC)
So just to be clear I'm doing the right thing (since I haven't reviewed a drive-by before) I should now quickfail each of those on the basis that the nomination contravenes those instructions? Mike Christie (talk - contribs - library) 16:58, 3 April 2023 (UTC)
When I came across one, I reverted the nomination on the article talk page, and then left a notice on the nominators talk page: like so. Harrias (he/him) • talk 18:14, 3 April 2023 (UTC)
That one is a good example of a genuine drive-by nomination. I did not see any contributions by the nominator in the last 500 edits to the article. (If that were in doubt, though, it could also easily have been a quick fail, on the basis that the article still has a big cleanup banner from 2022.) —David Eppstein (talk) 19:58, 3 April 2023 (UTC)
If you are in doubt whether it is drive-by, just assume that it isn't. Better to assess the article on its merits than on technicalities. —Kusma (talk) 19:38, 3 April 2023 (UTC)

To me, having 19 of the 50 most recent edits (as is the case for Vaughan J. in AC/DC) is above the bar for being a drive-by contributor. Drive-by contributors are the ones who make one or two copyedits that could arguably be marked as minor and then nominate. Anything more than that can be taken seriously. Otherwise, there would be no way to bring old well-established and mostly-in-good-shape articles to GA, because if we just measured contributor percentages it could easily be the case that nobody (or nobody still active) has more than single-digit percentages. Percentage is a bad measure of contribution, anyway, because a lot of the effort in bringing something from B-class to GA-class is the sort of polishing that does not significantly affect the authorship contribution percentages of the article, but still can be a lot of work. —David Eppstein (talk) 17:54, 3 April 2023 (UTC)

If someone has just done mainly minor wording alterations how can they attest that the article is free of original research, failed verification issues or copyvio? If they have encountered a mostly good article and done due diligence to ensure it's as good as it looks, then they should make a talk page post explaining this. (t · c) buidhe 18:17, 3 April 2023 (UTC)
If the writing is already good, but the article uses general references instead of inline citations, what is wrong with someone adding inline citations and then nominating? They will have very little "contribution percentage" but will know the article and its sources well. —Kusma (talk) 19:36, 3 April 2023 (UTC)
What you've described is a significant contribution. The editor had to check the sources and verify which of them matched which content. In contrast, copy editing or altering a small percentage of the article's overall number of citations, is not a significant contribution. (t · c) buidhe 20:03, 3 April 2023 (UTC)
You can't always tell though if people checked. I usually don't tell in my edit summaries I've checked a source unless there is an issue with INTEGRITY. You would only see me making copy edits when I check sources in a decent article. I'd prefer to assume they have when they've made a lot of copyedits. —Femke 🐦 (talk) 20:14, 3 April 2023 (UTC)
It's not a reasonable assumption because in my experience most copy edits do not involve source checking. For example, GOCE does not usually check. (t · c) buidhe 21:06, 3 April 2023 (UTC)
I think it would be unreasonable to quickfail a nomination by a good copy editor just because you have not seen them touch the sourcing (and you should check the sources either way). If you are afraid of wasting your time, just find something else to review. —Kusma (talk) 17:54, 4 April 2023 (UTC)
So many people in this discussion are talking about quickfailing these nominations. You don't have to quickfail the article to decline a drive-by nomination. It's better to just remove the nomination template from the article talk page and leave the nominator a message on their talk page. At worst, if they do turn out to have more involvement than you think, they can just revert your removal and no harm is done. ♠PMC♠ (talk) 18:14, 4 April 2023 (UTC)
I think I was the first to mention quick-failing, and you're right that it wouldn't be the right answer; my mistake. But to the point I was hoping to get an answer to: are you suggesting that if, as a potential reviewer, one considers a nomination to be a drive-by nomination, the right approach is to remove the nomination from the talk page and leave a talk page note explaining why? And that there is no more definite way to decide if something is a drive-by nomination than by qualitative assessments such as "significant contributions"? Mike Christie (talk - contribs - library) 18:22, 4 April 2023 (UTC)
If the article is a quickfail for other reasons, it may be good to quickfail based on those to leave a record for other potential nominators that there is a lot of work to be done. But yes, for good faith nominations by clueless newbies, reverting plus a gentle talk page notice may be the best course of action. Overall, reviewing an article although the nominator is "unworthy" strikes me as less of a problem than mistakenly rejecting a nomination. —Kusma (talk) 18:35, 4 April 2023 (UTC)
The rules do not allow (or even suggest the possibility) of a note on the talk page. "Nominators must have contributed significantly to the article" although this is contradicted with "If you are not a significant contributor to the article, you must secure the assent of the significant contributors before nominating." As David Eppstein notes, this prevents the nomination of worthy articles where the original contributors are no longer active. I have in the past nominated many such articles, most recently Tailhook scandal. Making large numbers of articles ineligible for GA is a blow to the entire process. Hawkeye7 (discuss) 20:25, 3 April 2023 (UTC)
The previous rule was that a nominator who was not a significant contributor to the article should ask the major contributors before nominating; my understanding was that the new rule was intended to remove this option and require a nominator be a significant contributor. If that is truly what was intended by the rule change, then presumably If you are not a significant contributor to the article, you must secure the assent of the significant contributors before nominating ought be removed from the instructions – though like Hawkeye I am not convinced that is actually desirable! Caeciliusinhorto (talk) 20:53, 3 April 2023 (UTC)

The original proposal received strong support; I didn't !vote or comment there partly because I hadn't run into this sort of nomination and didn't want to guess at the right way to handle them. I'm not comfortable failing either AC/DC or Neighbours: The Finale without a clearer definition of what constitutes "drive by", and I don't want to review them if they are not nominated by editors who are familiar with the sources and the article. Based on U-Mos's comments above, I'm going to go ahead and review the Neighbours article as they say they were more involved with the article than it appears at first glance; I'll have to let someone else, or the consensus here, decide what to do with AC/DC. Mike Christie (talk - contribs - library) 21:11, 3 April 2023 (UTC)

Having re-read that proposal, to my mind part of the issue is that drive-by nominator/significant contributor is not a useful binary. I would consider there to be a fairly large intermediate area between them where someone is deliberately editing a particular article, but they are not a significant contributor. For example, an editor who watchlists an article and has a history of reverting vandalism going back years may not be a significant contributor, but they clearly have a degree of investment in that article which distinguishes them from the drive-by nominator. Caeciliusinhorto (talk) 21:39, 3 April 2023 (UTC)
I think it should have been common courtesy for User:U-Mos to notify myself and User:Raintheone about nominating Neighbours: The Finale at GAN. We worked together on the article and are the two most significant contributors to it. Personally, I don't think it was ready, but we shall see. - JuneGloom07 Talk 18:13, 7 April 2023 (UTC)
JuneGloom07, I think the article is not far off a fail and have left notes at the GA. Please comment there if you have time. Mike Christie (talk - contribs - library) 18:54, 7 April 2023 (UTC)

FYI for anyone watching this thread, I've reverted the GA nomination of AC/DC, with a link to this thread. Mike Christie (talk - contribs - library) 18:59, 7 April 2023 (UTC)

Not every nomination of any TV series like Succession?

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Hi, not every nomination of any TV series like Succession and/or episodes of The Last of Us? Since these are unable to cover the nominations. CastJared (talk) 08:40, 8 April 2023 (UTC)

Hi CastJared, it is not clear exactly what you are asking. In general, most if not all articles can become Good articles, however you should not nominate an article that you did not develop yourself. Even if you did write them yourself, it is extremely odd to nominate a few dozen in the span of minutes. CMD (talk) 08:52, 8 April 2023 (UTC)
To add to that, an important exception to the general observation that most if not all articles can become Good articles is list articles (see WP:What is a good article?). So articles like List of awards and nominations received by Zendaya should not be nominated for WP:Good article status, as they are ineligible. TompaDompa (talk) 10:06, 8 April 2023 (UTC)
Example: Any Succession related episodes, are set to be nominated. Soon, an editor will pull it's nominations out. CastJared (talk) 11:40, 8 April 2023 (UTC)
For context, see their talk. The issue is not just about being able to respond to feedback on all, but also as has been stated in edit summaries and your talk, your lack of contribution to the articles as required. CastJared nominated many articles for good article status and was reverted on all, eight episodes of The Last of Us (TV series) plus the series itself, House of the Dragon, Zendaya, 95th Academy Awards, Westworld (TV series), Succession (TV series), List of awards and nominations received by Zendaya, List of awards and nominations received by The White Lotus, List of awards and nominations received by Succession. Indagate (talk) 08:53, 8 April 2023 (UTC)
So not all of them are nominated? CastJared (talk) 09:30, 8 April 2023 (UTC)
All reverted so not nominated. Indagate (talk) 09:35, 8 April 2023 (UTC)
Another example: Remaining The Last of Us are prepared to be nominated. After that, an editor pulls it's nominations away and find out by discussing this. CastJared (talk) 11:52, 8 April 2023 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Is there a need for a backlog drive

540 nominations waiting for review, quite a few from more than six months ago. Is this the sort of time we start thinking about the tactical nuke option? I know that TAOT has in the past been very disparaging about the amount of manual work needed; I wonder if it would be possible to automate some parts of the process, if Mike was willing. But first and foremost, would it be useful? ~~ AirshipJungleman29 (talk) 12:12, 2 April 2023 (UTC)

I think it's useful. I review more during a backlog drive (even if I don't sign up, as I don't like admin). More importantly, as backlog drives contain checks on review quality, this will be an opportunity to make sure everybody is aware of the new spot check requirements. —Femke 🐦 (talk) 12:16, 2 April 2023 (UTC)
I'd support a backlog drive. But "checking the review quality" of all reviews is what didn't work so well in the last two drives. We can't have the coordinators check all reviews (too many reviews, not enough coordinators). Either the checking needs to be very limited or it should be crowdsourced. —Kusma (talk) 16:34, 2 April 2023 (UTC)
I'm not opposed to a drive but I 100% will not be a coordinator. Not repeating that mess again. Trainsandotherthings (talk) 16:35, 2 April 2023 (UTC)
I think either crowdsourcing or sampling reviews is the realistic way forward. —Femke 🐦 (talk) 16:40, 2 April 2023 (UTC)
I’d support a backlog drive but would expressly skip the “awards”/review of the review phases, that's a tangential goal. They take so long as is and instead I see the value in just seeing many people motivated to clear the backlog. ~ 🦝 Shushugah (he/him • talk) 16:52, 2 April 2023 (UTC)
There has to be some element of review checking, because during a drive there is more risk of people focussing on quantity rather than quality, so some checking needs to take place to try and mitigate against that. But I've always been in favour of a sampling, similar to the spot-checks we use for source checking: in general, check a few reviews from each participant – if they are fine, AGF on the rest; if any concerns are raised, spot-check a few more from that participant. Harrias (he/him) • talk 18:58, 2 April 2023 (UTC)
I could see some sort of combination between automation and spot-checking. An automated tool that automatically flags passed reviews with less than X words would probably be valuable regardless. Thebiguglyalien (talk) 21:09, 2 April 2023 (UTC)
Shushugah I was thinking of automating that side of the drive; Mike Christie, would that be possible? With Wikipedia:WikiProject Good articles/GAN Backlog Drives/June 2022 for reference, I'm thinking that if the review (or user section) is ticked, ChristieBot calculates total score and which barnstar to give? ~~ AirshipJungleman29 (talk) 21:20, 2 April 2023 (UTC)
I haven't participated in a backlog drive, so I'm not sure I'm following along here. ChristieBot already knows what reviews are started/open/closed, and who the reviewer is. I think it would be relatively easy to post a list page somewhere listing reviews under X words long. Might be better to have two lists -- one for reviews using the template, one for reviews not using the template -- in order to avoid an unfilled template from making it appear to be a long review. What do you mean by "if the review (or user section) is ticked"? And what calculation? Mike Christie (talk - contribs - library) 21:40, 2 April 2023 (UTC)
Well, as above, people are discussing whether the previous coordinator method (checking every review for quality individually) could be slimmed down to maybe a sampling, after which the user's contributions to the backlog drive would be ticked as sufficient. Then comes the drudgery of the thing—points are added up (one point for a review, 1.5 for an old nomination) and barnstars are awarded based on contributions. If these purely mathematical calculations could be automated, it would make life for the backlog drive coords significantly easier. ~~ AirshipJungleman29 (talk) 22:58, 2 April 2023 (UTC)
I believe previous coordinators didn't check every review, just a sample of each reviewers. I don't think automating review selection will make much difference (it might even be detrimental?). The point scoring and barnstar awarding though, is an area that could be automated, assuming the automation coding doesn't take up even more time... CMD (talk) 23:46, 2 April 2023 (UTC)
I'm pretty sure I could write something that took a list of users and calculated the point count for them; I know the age of the nomination when it was picked up. That could be the list of all users signed up for the backlog drive, rather than just the ones that had been marked as sufficiently checked, but either way so long as there's a list somewhere on-wiki it should be easy. Not so sure I want to take on posting the barnstars on talk pages. For either task, we might need to go back to the BAG page -- the current BRFA is here; it might be considered to cover the first task, but posting barnstars seems beyond the scope of that. At a minimum I could run a query and post the results manually every day or two. Mike Christie (talk - contribs - library) 01:20, 3 April 2023 (UTC)
That would probably help, although if I remember correctly barnstars are all awarded at the end, rather than one by one. ~~ AirshipJungleman29 (talk) 01:24, 3 April 2023 (UTC)
If calculation and posting barnstars is the non-desirable task, I'd be happy to do that part. Can also help with checking reviews. Firefangledfeathers (talk / contribs) 01:28, 3 April 2023 (UTC)
Chipmunkdavis every review was checked at Wikipedia:WikiProject Good articles/GAN Backlog Drives/June 2022, the most recent one and the only one I can attest to. ~~ AirshipJungleman29 (talk) 01:25, 3 April 2023 (UTC)
That's asking the checkers for too much, the model we should look to is Wikipedia:WikiProject Good articles/GAN Backlog Drives/July 2021. CMD (talk) 01:46, 3 April 2023 (UTC)
I think that would be preferable, yes. ~~ AirshipJungleman29 (talk) 14:36, 3 April 2023 (UTC)
Im all for a new backlog drive, its about time we shut down the Proposal Drive Feedback tab. 🏵️Etrius ( Us) 03:17, 4 April 2023 (UTC)
Since this was first posted a week ago, the number of unreviewed nominations has jumped from 540 to 575. So it seems that the answer is yes. But I also think it would be a good idea to rework the instructions and process a little bit before this happens. Namely, we need to provide clear instructions for inexperienced reviewers to confidently spotcheck sources, and we need to decide whether we're going to implement more rigorous citation requirements (rather than the current system which only requires minimal citations). Thebiguglyalien (talk) 16:08, 9 April 2023 (UTC)
Adding the citation req to the GAN drive would be a good way to ease in a rule change for which there appears to be consensus. (t · c) buidhe 16:19, 9 April 2023 (UTC)

May I suggest that people participating in their first drive should have their first completed review or two checked as quickly as possible by coordinators so that they can get feedback on the quality of their reviews—perhaps we should ask them to self-identify during signup so the coordinators know who to prioritize? All reviewers should probably have at least one review checked, but helping to guide new ones promptly will be important.

Another potential issue: back when LegoBot was creating the WP:GAN, it stopped being able to include all the nominations after a certain point once we hit around 700 total nominations on the page on 28 March 2020 due to the transclusion page-size limits. Mike Christie, is that likely to be an issue given the way your page is constructed, or are we safe from that possibility? We're only around 40 nominations away from that size. BlueMoonset (talk) 14:53, 9 April 2023 (UTC)

I think it's fine -- the change to the GANentry templates seems to have increased the maximum size. I just took a look at the size and we're at 854K; the limit is well over twice that. Mike Christie (talk - contribs - library) 15:45, 9 April 2023 (UTC)
@Mike Christie Is there an easy way to note/self identify how many reviews/reviewed articles are associated with each editor though a transclusion? Similar to the GAN feed, this would help coordinators identify newbies early on. Not necessary at all if a huge hassle ~ 🦝 Shushugah (he/him • talk) 16:11, 9 April 2023 (UTC)
I'm not sure what you're asking for -- there's this recent conversation which didn't reach a consensus. Is that the sort of thing you're thinking of? Mike Christie (talk - contribs - library) 10:36, 10 April 2023 (UTC)

Closing a Good Article Nomination

Can somebody help me out. I need to close a GAN, but I can't locate the version number that I should be putting in under oldid, here, {{FailedGA|~~~~~|topic=|page=|oldid=}} Can somebody advise. Many thanks. KJP1 (talk) 09:42, 10 April 2023 (UTC)

This is on the list of enhancements for ChristieBot to do automatically, but for now it has to be done manually. What I do is just go to the history of the article and pull the id from the URL of the last page in the history. Mike Christie (talk - contribs - library) 10:31, 10 April 2023 (UTC)
Mike - that did indeed work. Many thanks. KJP1 (talk) 10:39, 10 April 2023 (UTC)
There's also a script that does the entire closing process for you at User:Novem Linguae/Scripts/GANReviewTool. Thebiguglyalien (talk) 15:22, 10 April 2023 (UTC)

Backup reviewer needed

During the course of reviewing M1841 12-pounder howitzer, I ended up adding a paragraph of content using a source I could access and then nominator couldn't. As a result, I'm not longer independent from the article enough to pass the review. I would very much appreciate getting a backup reviewer to make a pass over the article and then pass or ask for further work as appropriate. Hog Farm Talk 23:03, 10 April 2023 (UTC)

I'm on it. (t · c) buidhe 23:33, 10 April 2023 (UTC)
(edit conflict) Using the original edition (available here, incidentally published 1984 not 2005), I've verified citations 37 (p. 109), 38 {p. 148), and 39 (p. 167). I'm willing to AGF on 36, as it's late and I may just have missed it, and allow Hog Farm to make a decision on the nomination. ~~ AirshipJungleman29 (talk) 23:39, 10 April 2023 (UTC)

Reviewer blocked

Shawn Teller has been blocked as a sockpuppet of Architect 134, a long-term abuser/troll. There are ten GA reviews that need another look—there were some serious concerns about them even before the block, so the fact that all the reviews were in bad faith probably means they all need to be re-reviewed, in my opinion, even though that's not something we normally do. Extraordinary Writ (talk) 18:40, 5 April 2023 (UTC)

One option would be to list them at WP:GAR, but at the same time that risks clogging an already overworked area. More importantly, this is just more evidence that we need some mechanism to detect this sort of thing. Who reviews the reviewers? Thebiguglyalien (talk) 18:55, 5 April 2023 (UTC)
As one of the coords, I think we could handle it. Ideally we'd get engagement from the nominators and other editors, hopefully none would actually be delists. But they do need to be re-reviewed (not Capri-Sun though, as a second review was done). Trainsandotherthings (talk) 22:23, 5 April 2023 (UTC)
We the community review the reviewers, however that requires consensus of some kind. Shawn Teller's GANs are the first 8 of Wikipedia talk:Good article nominations/Archive 27#February 2023 raised reviews (everything up to Eurovision Song Contest 1999). CMD (talk) 23:48, 5 April 2023 (UTC)

JoelleJay noted at AN/I that the reviews were likely done by ChatGPT (thanks also to XOR'easter for posting at WT:GAR), which is why they sound so asinine. Happily, as Shawn Teller was a sock the GAN pages can be tagged for WP:G5. My suggested solution is to do this, and restore the original GAN tags on each article talkpage (and drop an explanation to the nominators). If there are no objections, I intend to do this in the next few days. This is an indicative sign that we need to become more firm on reversing checklist reviews. CMD (talk) 08:49, 8 April 2023 (UTC)

CMD, having read the contents of one of Shawn Teller's GA reviews after JoelleJay's list of egregious comments, I don't think any of the reviews can be trusted. I agree that a fresh review via GAN is warranted for each (with the exception of Capri-Sun, which already has a subsequent review), and the GA status reverted. It's unfortunate for the original nominators, but perhaps if the original nomination date is restored along with the GA nominee template, they won't have to wait too terribly long for a new review. BlueMoonset (talk) 14:28, 9 April 2023 (UTC)
Having looked at only one of these (Talk:Ontario Highway 11/GA1) I am in complete agreement that these should be G5 deleted and the nominations restored. —David Eppstein (talk) 18:55, 9 April 2023 (UTC)
I've done that for the Ontario Highway, seeing that also the nominator agreed. The others are probably deserving of the same treatment, but should be checked individually anyway. —Kusma (talk) 19:02, 9 April 2023 (UTC)
Unfortunately the tags were removed by Why? I Ask, so this process has been blocked. CMD (talk) 14:02, 11 April 2023 (UTC)
I don't believe that tag removal was appropriate. These were not real reviews that could give editors any guidance on improving the articles. Trainsandotherthings (talk) 14:05, 11 April 2023 (UTC)
I agree. A mass nomination at WP:MFD is probably the simplest next step. Mike Christie (talk - contribs - library) 14:15, 11 April 2023 (UTC)
They were not good reviews at all; but they are all currently good articles. Until they are delisted, the reviews should stay up. And if delisted, should it even say "former good article" if there is no review to link? Why? I Ask (talk) 14:26, 11 April 2023 (UTC)
This was an unwise move. You're now responsible for the content of those reviews (if you re-instate a blocked editors edits you take responsibility for them per WP:PROXYING), which by your own admission are mediocre. The reviews were to be deleted, and the articles returned to the nomination pool. I suggest you self-revert. It's a strangely bold move for an editor with limited experience in the area having a single pending GA nomination and no reviewing experience. Mr rnddude (talk) 14:35, 11 April 2023 (UTC)
Wow, what a bad argument. I hadn't known this discussion existed until I was pinged, but do follow G5 deletions. But from my own wiki-experience, if something is reliant on another page (i.e., a good article tag), then it should not be deleted until the other page is fixed. Why? I Ask (talk) 14:40, 11 April 2023 (UTC)
@Why? I Ask, your logic is sound but we appear to disagree on the right outcome. If you hadn't removed those tags, the reviews would have been deleted and the articles immediately delisted without the need for a GA reassessment. Your removal of the tabs forces more bureaucracy into a process that seems certain to have the same result. Mike Christie (talk - contribs - library) 14:37, 11 April 2023 (UTC)
Yes, I didn't know this discussion was underway, and have self-reverted. Does deleting the review automatically delist it, though? Cheers! Why? I Ask (talk) 14:42, 11 April 2023 (UTC)
No, but the GAR coordinators here and the other regulars are watching these reviews and can handle the removals. I'm the operator of ChristieBot which handles the automation of GAs and will have to check to see if it copes automatically, but I think it will. Thanks for self-reverting. Mike Christie (talk - contribs - library) 14:54, 11 April 2023 (UTC)
GAN templates restored, nominators notified, GA lists fixed, GA icons removed. If there's anything else let me know, it's all up to ChristieBot now. CMD (talk) 01:55, 12 April 2023 (UTC)
Thank you for doing this! Now let us quickly re-review these articles. —Kusma (talk) 07:08, 12 April 2023 (UTC)
Appears Christiebot got to it before my above comment! CMD (talk) 01:59, 12 April 2023 (UTC)

As non-admins will lose visibility, I thought it worth copying over some examples of what ChatGPT writes. 2022 South Lanarkshire Council election has "This article inspired me to pursue further research into South Lanarkshire Council elections of years past", and "Easily one of the top ten articles I have read, not only on Wikipedia but including other encyclopaedias as well." Proposed South Shore Line station in South Bend has "The prose is magnificent and fosters a vivid understanding of the proposed new South Shore Line station in South Bend in a manner befitting its unique intricacies. The use of literary devices depicts the subject in a manner befitting the depth of its coverage." 2014 NCAA Division I women's basketball championship game has "Thanks to this article, I have an in depth understanding of the 2014 NCAA Division I women’s basketball championship game. This article breaks down multiple sophisticated concepts into something that won’t just lead to understanding, but also growing." If I ever feel down, I might ask ChatGPT to review some of my articles. CMD (talk) 14:19, 11 April 2023 (UTC)

I agree. And then repeated recourse to AI for reviews would count as a form of ersatz m*sturb*tion. § Lingzhi (talk) 14:26, 11 April 2023 (UTC)

Increase in 6 month nominations

For the last few years, GA reviews have generally not lasted in the queue longer than four months or so, and it was rare to have a six-month-old nomination at all. But since the sort order was changed a few months ago to ignore wait times, there's been a dramatic increase in wait times for the longest nominations. As I write this, there are now 21 unreviewed nominations that are at least six months old, and it's gradually trending upward. 13 more nominations are set to become six months old in within the next week. Ideally, this number should always be zero. Some sort of change needs to occur to prevent these massive wait times that aggravate the backlog, whether it be in how we present pending nominations or a more systematic way to get older nominations reviewed.

In the meantime, I highly encourage everyone to try to grab some of the outstanding 6+ month nominations so we can get that number back to zero:

Also, shoutout to editors like AirshipJungleman29 and Onegreatjoke (among others) that have already been working on the 6+ month backlog. Thebiguglyalien (talk) 02:13, 23 March 2023 (UTC)

Ah, a list the eye can comprehend, oldest. Maybe those should be at the top ;-} --- I'll take one now. Chiswick Chap (talk) 14:45, 24 March 2023 (UTC)
Honestly, I'm starting to think the sort order change was a good idea with bad unintended consequences. I'd probably support going back to the old system. Trainsandotherthings (talk) 13:53, 23 March 2023 (UTC)
As the nominator of 4 of these (and also 2 other 6+ month noms that were recently passed this week), I think this is directly related to the sort order and decreased visibility for older noms. I also probably shouldn't have inundated the queue with so many similar articles, but I had the time back then! The recent changes did prompt me to start reviewing GAs, which I had never done before, but I quickly lost interest since they're based on a ratio (# reviews/# GAs). The top sorted articles are always newer editors to the process since it's so easy to have a good ratio with so few noms and reviews. At this point, I could review 10 more nominations and likely not even get my noms to move up in the list. I wonder if the ratio method should be reconsidered and the top sorted articles could be based solely on number of GA reviews completed? I'm nowhere near the top of that list either, but it seems more respective of effort. It would on the other hand push newer editors to the bottom of the list, which I think would delay their noms instead and not really set them up for continued interest in the process. I'd agree with Trainsandotherthings that this doesn't seem to be working as intended. There are other ways to nudge people to start reviewing. This sort order was supposed to entice people to review noms, but it doesn't appear to keep that momentum. Grk1011 (talk) 13:59, 23 March 2023 (UTC)
We've talked about restricting both the GA and review parts of the ratio to some recent time period -- six months, or a year, perhaps. Would that be better? Mike Christie (talk - contribs - library) 14:07, 23 March 2023 (UTC)
I don't think it's the same thing. Windowing the dates in the sort order changes how recently you have to be active to make your nomination more visible, but doesn't cause the older nominations to float to the top. —David Eppstein (talk) 16:33, 23 March 2023 (UTC)
Grk1011, rereading your comments, I have another question. You say the change in sort order prompted you to start reviewing, but since that didn't get you to the top of the list you gave up. I don't want to pick on you, but I'd like to understand that a little better. Not everyone has the time to do reviews, and there's no requirement to review GAs, but we all understand that GA would have an instant turnaround if every nominator reviewed an article every time they nominated one. What would get you to do more reviews? A lot of the people who participated in the discussions about reviewing are editors who already review more than they nominate. What do you think would encourage editors with more GAs than reviews to start reviewing more? Mike Christie (talk - contribs - library) 19:15, 23 March 2023 (UTC)
Mike Christie I'm a bit of a box-checker in that I hate having things pending for a while. I like to improve an article as quickly as possible and get it over with. My "work list" is rather arbitrary; I randomly feel inspired to completely redo an article with the intent of bringing it to GA. For me, GA is an appropriate stopping point for articles about the subjects I focus on. Right now, given my work load in real life, I'm not too bothered by the delay in having my nominations reviewed, but it definitely has discouraged me from bringing any additional nominations here. Not only that, but I've more or less stopped feeling encouraged to do comprehensive edits to get other articles ready, instead spending my time on organizational or consistency edits across similar articles or updating recent events in their chronology. I'm not really a half-in type of person. I've thought about what articles I could review from the nomination queue, but I keep going down the path of things like "ooo I could review all of Floydian's 11 highway noms and what a personal accomplishment that would be!". I just don't have the time right now and after thinking of goals like that, one-offs seem boring or futile. They also take me away from the articles I actually do want to edit, but am currently not because of the lack of the task-completion feeling. Honestly, I'm not sure what would get me to naturally start reviewing more, but I hope you can at least understand my mindset. Grk1011 (talk) 13:18, 25 March 2023 (UTC)
I've begun to think the issue with the sort order is that we were aiming to prioritize nominations by frequent reviewers, but instead the system just pushes all of the nominations by editors who have had no prior involvement with the GA process to the top. Take, for instance, the warfare list as of this point in time, where the top 6 nominations combine for 1 review and 0 GAs, while 7-12 combine for 1,854 reviews to 1,128 GAs, 13-18 combine for 5,298 reviews to 4,962 GAs, with 19-25 being 94 reviews to 236 reviews. It's nice to put new contributors here as priorities, but it's just resulting in burying the nominations this was suppose to highlight. And if the editors with good review to nom ratios are getting buried, imagine what's happening to those with poorer ratios. Hog Farm Talk 15:41, 23 March 2023 (UTC)
I agree it should probably be revisited. I've just started reviewing GAs again, and am trying to focus on new nominators, for a couple of reasons. One is that I do think they are a priority -- whether they are experienced editors or new to Wikipedia I don't want newcomers to GAN to immediately learn that you have to wait 3-6 months for a review; I don't want that to be the norm. Another is that I think this group is where new reviewers are most likely to come from, and encouraging them by reviewing their work quickly seems like it might be a good way to bring editors into the GAN community. I also wonder if things might look different if the current backlog of nominations by new nominators were cleared; that is, that the situation we're seeing is partly because the list is in a transitional state.
I think there should be a prioritization benefit for frequent reviewers. If the current sort order isn't working, another approach might work better, but I'm not yet convinced that we'd better off going back to just the age of the review. Mike Christie (talk - contribs - library) 16:13, 23 March 2023 (UTC)
I also think we should play with this a bit and try to see what works. Basically, I think waiting time should play a role. Let's say that every user gets points for waiting time based on their review quotient, and we then sort by points. If users with a bad review quotient get 1 point per day and users with good review quotient (or newbies) get 5 points per day, we might get a more reasonable outcome. —Kusma (talk) 16:47, 23 March 2023 (UTC)
I like that idea. —Femke 🐦 (talk) 17:10, 23 March 2023 (UTC)
I worry that every change we make off the simple "oldest first" formula will only make the listings seem more arcane and complicated. ♠PMC♠ (talk) 18:51, 23 March 2023 (UTC)
As the listings are done by bot, the complicated formula is behind the scenes. In my idea, new noms go to the bottom and are easy to find, they just rise at different speeds as a carrot for reviewers. If we get the point scores right, I hope this leads to a usable compromise system. —Kusma (talk) 19:14, 23 March 2023 (UTC)
I would also support a return to the old system. I think the reasoning behind the change is good in theory but flawed in practice: there is little stick and even less carrot. It turns out that people don't care if their nominations go into a chaotic mess in the middle of a list, because there will always be a number of nominations by people who haven't put anything into the process at the top. As far as I can see, the new system doesn't really encourage anyone to review, so it's a failure with respect to its foremost goal. ~~ AirshipJungleman29 (talk) 16:49, 23 March 2023 (UTC)
Yes, the current ordering might as well be: nominations from people who have never successfully written a GA first, everybody else second. The ordering within those two groups is much less significant than the fact that the first group is first. —David Eppstein (talk) 01:02, 24 March 2023 (UTC)
Pinging those who participated in the first sort order discussion at WP:GAPD23: @Rjjiii, Lee Vilenski, Olivaw-Daneel, Buidhe, Eddie891, Sparkl, Shearonink, Epicgenius, Sammi Brie, Unexpectedlydian, Czar, Chiswick Chap, Iazyges, JPxG, Ganesha811, Lord Roem, and Chipmunkdavis: ~~ AirshipJungleman29 (talk) 13:36, 24 March 2023 (UTC)
I'd favor retooling the formula like Kusma discusses above and trialling that before reverting to the old format. Eddie891 Talk Work 13:42, 24 March 2023 (UTC)
Yeah, I think the idea behind changing the sort order was well-intentioned. But, as it turns out, if one group is prioritized then that means someone else has to go to the back of the line. Perhaps we should try Kusma's idea, giving "points" based on noms' review ratios. – Epicgenius (talk) 13:44, 24 March 2023 (UTC)
I am certain the intentions were good; I agree it doesn't seem to be working terribly well; and for all I know Kusma's approach might work – or perhaps it won't, and we'd be better off with the old system. Chiswick Chap (talk) 13:59, 24 March 2023 (UTC)
If it's not working, it's not working. I don't have any issue with changing it Rjjiii (talk) 14:41, 24 March 2023 (UTC)
I wonder, is it possible to order nominations by how many reviews they did and not by a ratio of GAs to reviews. Like for example, the people who've reviewed the most GAs have their noms at the top while the ones who've reviewed the least go to the bottom. This could possibly incentive more people to review articles as they would desire to go higher up the nomination list by reviewing articles. Though that may just be some wishful thinking. Onegreatjoke (talk) 13:48, 24 March 2023 (UTC)
That too might work, and could certainly be worth a try. I guess the question is what do we want to achieve, really, and following on from that, whether we really think we want to prioritize in some way. If we do, we might ask ourselves whether submitting a GAN is a cost or a benefit to the system. If it's thought to be a cost, then the current system of dividing by the number of GANs someone has made makes some sort of sense (even if it doesn't seem to be working as a method); if it's actually a benefit – readers get more good stuff to read, the encyclopedia gets more good stuff to show, and reviewers get a necessary input so they have something to work on (grist to their mill), then the current system is actively counterproductive. Perhaps we should be ADDING nominations to reviews, not dividing the one by the other...... Chiswick Chap (talk) 13:59, 24 March 2023 (UTC)
Personally, I'd say having the ratios is enough - I'd like it to be sortable on preference, rather than a list of what a specific set of users would like to state. Lee Vilenski (talk • contribs) 14:06, 24 March 2023 (UTC)
Among the bad effects of the ratios are that all the GANs by any one editor get lumped indigestibly together; this in turn makes any new ones (or the oldest ones) hard to spot in the mass. Chiswick Chap (talk) 14:43, 24 March 2023 (UTC)
I agree with this comment. If someone nominates one article in December, and another in March, it doesn't make much sense for them to be in the exact same place in the list. Historically we have long considered it important to review older nominations first, and I'm finding the sort order change makes this harder, not easier. Trainsandotherthings (talk) 14:45, 24 March 2023 (UTC)

A couple of other ways to think about this occur to me.

  • A sortable table instead of a list might please everyone, but might be uglier.
  • We could add icons, colours, or mark certain nominations in some other way, instead of changing how they sort -- e.g. a "NEW NOMINATOR" icon of some kind, if we want to mark nominations by new nominators.
  • One thing that would reduce the number of "0 GAs" at the top of the lists would be to change the metric to count the reviews received, not just passed GAs. For example, I just reviewed Gdańsk, which was top of the list, but which had recently failed a GA and still had most of the same errors.
  • There aren't that many of these new nominators clogging up the top of the lists - about 18 or 20, at the moment. That represents three or four days of reviewing throughput. If after two months of the new sort order, we still have them sitting on top of the lists, does that mean the sort order doesn't have that much impact on what gets reviewed? Would we still have the same concerns if every one of those was under review?
  • There seem to be two separate issues: do new nominators deserve to be top of the list, and does the sort order work for everyone else. Some of the comments such as Grk1011's above are solely about the second point. I think we need to keep these separate in the discussions.
  • Putting in the sort order was an attempt to change behaviour by changing the algorithm, and seeing if the resulting order looked more appropriate. Can we change this around and agree on which nominations deserve to be top of the list, then figure out what the means the algorithm should be? For example, in the Transport section, here are some nominations, reduced to just the nominator, the R and G stats, and the nom date. These are in the order currently listed. I've skipped a few, as this is just for discussion purposes but the data is real.
    1. IP nominator, 0 reviews, 0 GAs, 10 Feb 2023
    2. Sturmvogel 66, 883 R, 827 G, 23 Feb 2023
    3. John M Wolfson, 9 R, 9 G, 1 Mar 2023
    4. Trainsandotherthings, 20 R, 25 G, 21 Feb 2023
    5. X750, 3 R, 4 G, 11 Mar 2023
    6. HoHo3143, 2 R, 3 G, 6 Jan 2023 (and a dozen others more recent)
    7. SounderBruce, 74 R, 189 G, 2 Mar 2023
    8. Floydian, 9 R, 101 G, 4 Oct 2022 (and a dozen others)
    9. NotOrrio, 0 R, 1 G, 6 Feb 2023
    What order do we think these should really be in? Looking at this I think NotOrrio doesn't deserve to be at the bottom, and I think Floydian and SounderBruce are unfairly doomed to stay at the bottom even if they do a dozen reviews in the next two weeks. I think the IP nominator shouldn't be above Sturmvogel 66. Other than that I think the order is mostly OK. If we can agree on how this list should be sorted perhaps we can work backwards from that to something the bot can do.

-- Mike Christie (talk - contribs - library) 14:10, 24 March 2023 (UTC)

  • (edit conflict) I opposed Prop 9 because I didn't think this would be effective, but it seems to be accomplishing what it set out to do: changing default behavior away from reviewing the oldest noms. For it to be used as a priority worklist, the sort order needs to be apparent to the viewer whereas now it appears haphazard (i.e., no easily distinguished reason for what goes to the top). My suggestion would be to revert to first-in, first-out date order and to more promiently display alternative sort orders to encourage editors to work from different views of the list. (Or, if those alternative sort orders were maintained, pull the "top five" for top sorts as a discovery page: five oldest noms by reviewers with good ratio, five oldest noms by new editors, five noms in need of second opinion, two oldest noms in each category.)
Changing the default sort order is bikeshedding because it only seeks to affect the default editor behavior in a very small pool of existing GA reviewers. The point of the GA reforms was to expand the pool: to make GA as an overall process more accessible and easier to complete. I think there are clearer ways to do that (I proposed several) but I also don't think the community wants it enough to try the bigger changes that would be needed. czar 14:22, 24 March 2023 (UTC)
A sortable table seems like the best solution to me. That way prospective reviewers can sort by whatever criteria they desire, be it review age or nominators' review to GA ratios. Trainsandotherthings (talk) 14:36, 24 March 2023 (UTC)
  • During the proposal I wrote that I did not oppose as I did not think that the sort order would have a large impact. If it is having a large impact, I would reconsider, and I'd like to know if anyone has any stats for the usual number of 6 month old noms. That said, there was a strong consensus for this change, and per the logic of the consensus surely we would expect an increase in older unreviewed nominations? Switching from age was somewhat the point of the change. CMD (talk) 14:42, 24 March 2023 (UTC)
Something that is sortable would be useful. I personally want to know how many GANs & how many Reviews an editor has done, it's important to me but I do think being able to sort on a variety of topics - oldest 5 GANS in a particular subject, oldest 5 GANs overall, best ratios of noms to Reviews for a GAN'er and a GAR-editor, newest Nominators efforts, whatever... - is important to various editors. Maybe what I want isn't applicable to another editor and that's ok. And if a sortable table doesn't work? well, at least something was tried. We can always go on and try something else. Shearonink (talk) 01:45, 25 March 2023 (UTC)
Update: despite the reviews that were taken above, the number of nominations waiting for at least six months has increased to 25. We also have a 7+ month nomination now, with a few more in the above list set to hit seven months in the coming days. Thebiguglyalien (talk) 01:57, 28 March 2023 (UTC)
How long before we have to use the tactical nuke? ~~ AirshipJungleman29 (talk) 02:25, 28 March 2023 (UTC)
I'm not happy about these old nominations, but as CMD said it's what we should expect given the sort order change. A change away from sorting by age can only increase the age of the oldest nominations because it will divert reviewing to other nominations -- more deserving nominations, in theory. The question is "what does 'most deserving' mean?". Switching back to sorting by age would mean we think a six-month-old nomination by someone with 70 GAs and no reviews is more deserving than a five-month-old nomination by someone with 10 GAs and 40 reviews.
Over the last week or two I've been reviewing only articles by nominators with no GAs. There are few enough of these that I can probably clear them myself in a month or so, but if a couple of other reviewers want to chip in and review some of these, it would eliminate the "two classes" that David mentions above very quickly. I'm also taking the opportunity to ask the experienced editors among those new nominators if they'd be interested in doing GA reviews, but I've had no takers yet.
As for sorting, the bot can definitely do that, and in fact six months ago that was my original proposal for a change to the page -- see User:ChristieBot/SortableGANoms for a version from December. There are some downsides -- we have information spread across multiple lines right now which would make a table ugly. But that's certainly an option. Mike Christie (talk - contribs - library) 10:19, 28 March 2023 (UTC)
I don't see "deserving" as being relevant. Our goal is to bring articles to a certain standard, and anyone that engages with GA, whether as a nominator or a reviewer, should be assumed to be contributing to that goal. Much in the same way editors with higher edit counts or older accounts don't have seniority, editors don't have seniority just because they've written or reviewed a certain number of articles. Thebiguglyalien (talk) 16:04, 28 March 2023 (UTC)
As an honest personal aside - Sometimes. *Sometimes*, when I have time for doing a GAN workup and I look over the GANs the loooooong candidates overwhelm the amount of time I can devote to doing it, some of the GANs are just kind of enormous y'know? The amount of time I can honestly devote to a workup, I just can't do them justice. And sometimes the subject matter is something I have little interest or expertise in. Does this possibly ding great candidates? Yeah, yeah, I admit it does but time is precious to me at the moment and I have to be cognizant about how I spend it. Shearonink (talk) 19:34, 28 March 2023 (UTC)
Re I don't see "deserving" as being relevant: I know what you mean, but given that we're talking about a scarce resource, any sort order that puts nomination X above nomination Y implies we think X is more deserving than Y. The only way to avoid that would to randomize the list on every refresh. Sorting by age has an implied valuation, just as the current sort order does. I'm not arguing that we should keep the current order, just that age ordering is not neutral either. Mike Christie (talk - contribs - library) 21:10, 28 March 2023 (UTC)
Different reviewers may have different priorities for what to select. Some may wish to prioritize new participants in the process. Some may wish to prioritize participants with a good review-to-nomination ratio. Some may wish to prioritize older nominations. Some may wish to prioritize newer but unready nominations, to spare their nominators a long wait for a quick fail. The original date-ordered listing allowed reviewers to find both older and newer nominations easily, but maybe not to pick out the ones by newer participants or the ones with many reviews. A sortable list could potentially accomodate more of these needs, if the sort keys are chosen well. —David Eppstein (talk) 21:43, 28 March 2023 (UTC)
Considering that it's the standard order for every other talk or discussion page on the project, I think it's fairly clear that general consensus believes age ordering is the most neutral. ~~ AirshipJungleman29 (talk) 21:45, 28 March 2023 (UTC)
There was a specifically "strong support" for the current sort order at Wikipedia talk:Good Article proposal drive 2023#Proposal 9: Change sort order of GAN page to prioritize frequent reviewers, only a few months ago. CMD (talk) 22:36, 28 March 2023 (UTC)
Yes, I supported that, with the logic "Prioritisation of those who give back to the community is a good idea"—i.e. I thought dropping the conventionally used sort order in favour of a less-neutral formula would benefit the project. The "strong support" was for the idea, not because I and others believed it was neutral. I now think that the initiative was misguided. ~~ AirshipJungleman29 (talk) 23:29, 28 March 2023 (UTC)
This sums up my position - I thought changing the sort order was a good idea and would encourage reviewers to review in order to move their reviews closer to the top. I see now that it has not done so, likely because the reasoning behind the sorting is opaque and people don't realize the connection between sort order and reviewing. I don't believe any twiddling with the sort algorithm will make this less opaque, and I no longer believe the revised sorting is a good idea. ♠PMC♠ (talk) 02:14, 29 March 2023 (UTC)
I agree that the sort order change is not working. I believe the article's number of views should somehow be taken into consideration. Currently, the Battle of Shiloh has been waiting since October. It gets about 30,000 views in 30 days, and is a level-5 vital article that was demoted to C-Class. I have redone the entire article, and believe it is Good Article quality. It had an informal Peer Review which has been saved, but that eliminates one of the few people willing to review American Civil War articles. I have only done two reviews, so the article will always be at the bottom of the list. Nobody wants to review an article that has 262 citations, even if it is important. The current procedure incentivizes me to find short articles to review or write, and makes it unlikely that I will get involved with long or difficult articles—even if they have a high importance and a high number of views. TwoScars (talk) 16:32, 1 April 2023 (UTC)
Sort order isn't going to change people's reluctance to sign up for reviewing long articles with many citations. That's always been a problem no matter what order the articles are presented in. ♠PMC♠ (talk) 17:07, 1 April 2023 (UTC)
I feel that the current sort order (by ratio) is better than the old one. I think I put in two reviews that I would not have done with the old system, which I found demoralising. The ratio is an intelligent system. It encourages 1st time nominators to start participating in the process and penalises nominators who do not review. But this is not enough. We need to reward reviewers in proportion to the work they do. A fail is less worth than a pass. A long article is more work than a short one. In last resort, age should probably also be taken into consideration so that we do not condemn some nominations to wait eternally and so that a failed nomination does not directly jump back to its previous rank. It might also help if nominations that are not ready are failed rather than improved in longish reviews. Perhaps GA instructions should say something about this. I usually tend to pass even quite unready nominations in reviews that exceed the recommended 7 days by far. Perhaps I should rather do another review during that time. There should be a sentence in the introduction to the list that explains the sorting. I would however also limit the number of nominations a wikipidian can have in the queue. This should be related to the users GAs. Beginners can have only one nomination in the system. Successful GA nominators are allowed more. That might cut back on mass production of nominations that are not ready for review. I feel that GA nominations must make the work of the reviewer as easy as possible. URLs of sources or even of pages in the sources should be given whenever possible so that reviewers do not need to lose their time searching for them. Thanks, Johannes Schade (talk) 21:17, 1 April 2023 (UTC)
It encourages 1st time nominators to start participating in the process and penalises nominators who do not review. This was the intention, the discussion above indicates that this has not actually been the result. ♠PMC♠ (talk) 01:27, 2 April 2023 (UTC)

(←) I'm not clear on what the new sort order is. My suggestion would be two-tier: 1) Sort by date added, oldest--> newest 2) sort by number of nominator GAs (not reviews), lowest --> highest. So the oldest nom with the fewest GAs under nominator's belt gets priority. That would encourage participation. Participation engenders collegial goodwill; goodwill encourages reviews. As for "no one wants to review long article with many cites", maybe a separate transcluded section with top 5 articles sorted by length, longest --> smallest, as a challenge to the daring § Lingzhi.Renascence (talk) 04:39, 6 April 2023 (UTC)

The new sort order is by the quotient (number of reviews)/(number of GAs). I don't understand your "two-tier" suggestion. We sort by age, and then we sort the one in a million case of equal age by GA count? There is no practical difference to just sorting by age. But perhaps I misunderstand. —Kusma (talk) 10:57, 6 April 2023 (UTC)
Mine is, sort order is (ranking descending by age, oldest ==1, second oldest==2, etc.) PLUS (ranking ascending by number of GAs, lowest == 0, next lowest == 1, etc.).. then score like golf: lowest score wins. § Lingzhi.Renascence (talk) 11:23, 6 April 2023 (UTC)
Ah, I see. I made a similar proposal, just multiplying with a rank-based multiplier instead of adding the ranks. —Kusma (talk) 11:50, 6 April 2023 (UTC)
Great minds think alike, and so do ours... § Lingzhi.Renascence (talk) 11:54, 6 April 2023 (UTC)
I'll give an alternate take on this, since while this discussion didn't arrive at a consensus, it did have a general feeling. I've been meaning for years to do a review for every GAN I nominated- it only seems fair! And yet, when the day came to sort nominations by R/N percentage... I was about 3/4ths down through the list for my section, at 35:100. So, to catch back up, I'll need to do dozens of reviews, and until then I'll be buried underneath other nominations. And fair enough! I have a multi-year backlog to work through, and it seems fair that I don't get special consideration on not doing enough reviews just because I've done a lot of nominations. So, the end result is that I've done more reviews in 2023 then I have in years, which I think is a good result. Maybe it doesn't sting as much because my section doesn't seem to exclusively get reviewed in sorted order- so I do still get reviewed eventually- but I don't think the current sorting is all bad. --PresN 00:50, 10 April 2023 (UTC)
I realise I am slightly late to the party on this too, but I thought I would add my thoughts to this anyway. I can see the merit in the introduction of the current ordering system but also it has become apparent that much of the intention behind its creation appears to now have failed to live up to its intention. However, on a positive note, and speaking from a personal viewpoint, the new ordering system has incentivised me to start reviewing GA nominations while waiting for my own to be reviewed, which was something I had intented to do for a long time but never actually got round to doing previously. However I can only speak for myself, potentially as compared to other editors my number of GA nominations passed is much lower and therefore the task of improving my ratio seemed a lot less daunting. I do believe there is still merit in sorting based on GA review to nomination activity, however adding in an age component would also make sense; I myself have noticed the struggle that can occur to pick out which articles have been languishing to be reviewed for a long time. Sims2aholic8 (talk) 15:50, 12 April 2023 (UTC)

Review counting bug

There's an intermittent bug causing some reviews not to be properly counted; I'm not sure why. I'm about to go out of town for a couple of days and probably won't be able to look at this seriously till next Thursday. I should be able to fix whatever it is then, and any uncredited reviews will be added to the stats then. Mike Christie (talk - contribs - library) 15:35, 8 April 2023 (UTC)

This has been fixed and the GA stats page should now be up to date again. Mike Christie (talk - contribs - library) 16:05, 13 April 2023 (UTC)

Should we make the nominator of GANs clearer?

Especially if a nomination is a quick fail, there's no easy way to see who the nominator was without digging through the article history. Neither the {{Article history}} template nor the /GA subpage currently mentions the nominator name.

Any interest in programming some of the tools or templates to stamp the nominator's name in one of those places ({{Article history}} or the /GA subpage)?

Example: Talk:All your base are belong to us/GA1

If not, no worries. Thanks. –Novem Linguae (talk) 21:43, 10 April 2023 (UTC)

I've noticed this as well. It can be a bit annoying sometimes, especially when trying to look at older GA nominations where there isn't a dedicated GA subpage. Thebiguglyalien (talk) 21:46, 10 April 2023 (UTC)
The question I see is whether it is desirable to have the nominator featured prominently and permanently, especially for fails. It is not exactly rocket science for nominations from the recent couple of years, you find out when the review was closed as a fail and look at the previous version of the talk page to see the nominator featured in the {{GA nominee}} template at the top. —Kusma (talk) 21:54, 10 April 2023 (UTC)
This was a problem when I went trawling through article histories to build the historical GA statistics. I would like to add a "nominator=" optional parameter to FailedGA, and an "actionNnominator=" parameter to articleHistory. I proposed this, I think at the articlehistory talk page and at VPM, and got silence at one place and mild opposition at the other. I still think it would be a good idea. Kusma, it's mostly useful for situations where the talk page history doesn't easily give you that information -- for example because the relevant revisions have been deleted. Mike Christie (talk - contribs - library) 01:24, 11 April 2023 (UTC)
I would be wary of having it on the talkpage, per Kusma. I wouldn't mind it on the GAN page itself, but perhaps there is a subtle way to include it, such as creating a link to the nomination diff. CMD (talk) 02:05, 11 April 2023 (UTC)
If the nominator gets mentioned on the review page, we should also have the comments field from the nomination copied there, as this is where information about collaborative nominations is usually stored. —Kusma (talk) 05:40, 11 April 2023 (UTC)
This would be great on the GAN page, next to or above the reviewer. I've often spent too much time finding this info too. —Femke 🐦 (talk) 16:35, 13 April 2023 (UTC)

Should we explicitly ban AI reviewing?

There has been discussion above about a reviewer being blocked, who seemingly used ChatGPT or a similar tool to create his reviews, with sometimes funny results ("I now understand that Ontario Highway 11 is a paved road that vehicles use to travel."), but which generated a large headache to clean up. It seems prudent to explicitly ban using AI for reviewing, by means of adding a short line somewhere in the instructions along the lines of "thou shalt not use AI", but I wanted to gather consensus before going forward with such a change. Iazyges Consermonor Opus meum 13:43, 13 April 2023 (UTC)

Seems like a sadly necessary (Sadnecessary) rule addition. Trainsandotherthings (talk) 15:36, 13 April 2023 (UTC)
Yes, we probably need to. If we can tell... Chiswick Chap (talk) 16:13, 13 April 2023 (UTC)
It's common sense not to do that right? It's instruction creep to mention "don't use AI" in dozens of policies. Instructions are probably overly long for new users as is. They instruction, especially around spot checks, already preclude this. —Femke 🐦 (talk) 16:33, 13 April 2023 (UTC)
I wonder if this might be something that's better handled at a sitewide level. I get the impression that the community is leaning against LLM content being welcome in any namespace. I'm worried that explicitly listing everything not to do on a review will become a WP:CREEP issue. Thebiguglyalien (talk) 16:34, 13 April 2023 (UTC)
I think we should explicitly say here that we don't accept AI reviews. These are a much bigger problem than people using AI to copyedit. If a sitewide AI ban happens, we can go and remove any redundant instructions at that point. —Kusma (talk) 17:17, 13 April 2023 (UTC)
It should go without saying and it feels like a bit of both WP:CREEP and WP:BEANS to add to the instructions, but on the other hand this is apparently already happening... I agree that it would be preferable to address this sitewide. TompaDompa (talk) 17:13, 13 April 2023 (UTC)
Ppl who are using llms are probably not the same ones as read the directions. The main benefit of making it explicit that I can see is to make it easier to overturn bad reviews. (t · c) buidhe 17:40, 13 April 2023 (UTC)
Fair point. TompaDompa (talk) 17:46, 13 April 2023 (UTC)
Instinctively I lean more towards not bothering because it's site-wide good practice not to and we don't need a rule here. A ChatGPT review would be poor surely anyway, they can't spotcheck, and if they write prose that looks like a spotcheck it would presumably be nonsensical at a quick glance at the source. It was pretty blatant in the sock reviews that the reviews were terrible. CMD (talk) 23:35, 13 April 2023 (UTC)
Yeah, I'm inclined to CMD's position here. Bad reviews are bad reviews, LLM-produced or not. Let's be honest, someone who's not prepared to put in the effort of actually reviewing the article themselves isn't the sort of person who'd read the instructions thorougly. ~~ AirshipJungleman29 (talk) 23:49, 13 April 2023 (UTC)

SyntheticSystems again

There was a recent conversation about the revews done by SyntheticSystems; it's here in the WT:GAN archives. Their reviews have generally been very skimpy -- examples are Talk:Magdalena Cajías/GA1, Talk:Schooling and the Struggle for Public Life/GA1, and Talk:Yella Hertzka/GA1, which was the subject of the previous conversation about their work. There wasn't exactly a consensus at the previous discussion, but there was agreement that the rules don't forbid reviews that don't say very much.

SyntheticSystems has a total of 11 reviews, and has nominated 5 articles at GAN, with a total of only 494 edits. These are not drive-by nominations; for the ones I've checked they are the main contributor. The low edit count against this activity made me wonder about the quality of their work. I noticed one of the articles, Verrado High School, was nominated at FAC; the nomination is here. Some of what I read there makes me think this is a WP:CIR issue -- they went to FAC because they didn't want to wait for months at GAN, and wouldn't withdraw despite being told the article was not up to standard; they attempted to make fixes but unsatisfactorily ("Fixed this. -- SyntheticSystems. / Not really. AirshipJungleman29"); they were given an example of an error by reviewer Nick-D ("The statement that "Initial plans for the school failed to consider acoustics within the auditorium" isn't supported by the source") and left that text in the article (and it's still in the article now); to the next comment from Nick they responded "I don't really know how to fix it".

I don't know what the right answer is for a user whose prose and editing skills are not good enough for GA quality. I don't mind occasionally reviewing a nomination from a user like this, but I don't think it helps the encyclopedia to have five nominations from a user like this sitting in the queue. When I review an article, I want to feel that the nominator is able to take the advice and feedback and use it to make the article better. I don't have that confidence in this case, and that also makes me think that the 11 articles they've reviewed, though they technically may have met the criteria we set at WP:GAN/I, were probably not reviewed to the standard we want to see. I think we should consider banning them from reviewing and from having more than one nomination at a time. Mike Christie (talk - contribs - library) 13:10, 13 April 2023 (UTC)

FYI, it seems like they've indicated that they've quit Wikipedia on their talk page (#Inactivity section). ~~ AirshipJungleman29 (talk) 13:24, 13 April 2023 (UTC)
My experience with them has indicated major CIR problems (in addition to what Mike has highlighted above, just see my previous comments on their talk page, or sending in ineligible DYKNs multiple times, or trying to get their own nominations reviewed by doing drive-by reviews on other nominations in the education category). This user just ignores people trying to help them or point out issues. Since they've apparently quit, we might as well pull all their nominations. Trainsandotherthings (talk) 15:34, 13 April 2023 (UTC)
You're literally the one who drove me off this website. Every single thing you've sent has been so incredibly rude. You act like I've been on this website for years and know every single rule and then accuse me of not listening. And then you try to do this. It's so insanely obvious that you have zero regard for new users. Why do you act like this? Are you really going to tell me that "Please do better" isn't passive aggressive? And then in that, you automatically assume that it's intentional. Some conspiracy, like I'm sending in did you knows on purpose! To waste everyone's time! And I'm planning my next move, reviewing good articles!
There is no reason for you to act like this. I just can't even comprehend acting rude to someone who doesn't know the rules. That's not right, dude. I came in here because I thought it was a fun thing to do and I was bored. And I actually found someone who helped me out. Makes it all the more disappointing that you have to flex that you've been in here for 2 years and act better than everyone else. Not cool. And you lack the ability to see that this your impact on me. You're the one who hasn't been listening.
I'm not gonna lie, I make mistakes. But instead of trying to help me, you just do the easy thing of attacking me for making mistakes. You can't seriously expect people to go along with that. And then you lie and say I'm inactive? I literally said I didn't want to review it because I just thought it was way too much and I kept telling that editor stuff to improve it and it still wasn't good enough so I just left it. Way to twist my words dude.
I guess you win. If this is how most of the editors on here act, I'm not going to give them any more satisfaction. SyntheticSystems (talk) 17:51, 13 April 2023 (UTC)
The comments on SyntheticSystems' talk page by Trainsandotherthings seem completely reasonable to me. I'm going to go ahead and remove the nominations SyntheticSystems has made; from the above it seems unlikely they will review or nominate again so we can consider this closed. Mike Christie (talk - contribs - library) 19:05, 13 April 2023 (UTC)
Sorry, you can't just claim "I don't know the rules!" as a get out of jail free card, after they've been explained to you multiple times. You've basically just proven my point with this aggressive and utterly false response. I have not once attacked or been rude to you, you are just utterly incapable of handling any sort of criticism or advice. Do you think this is some sort of game? You twisted my words, I was not "flexing" how long I'd been editing, I was emphasizing I was relatively new to Wikipedia myself! Trainsandotherthings (talk) 12:58, 14 April 2023 (UTC)

I removed Proposal drive from banner

Revert if you strongly feel otherwise, but I think we've run its course at Wikipedia:Good Article proposal drive 2023/Feedback. May be useful to link this elsewhere. Generally, I think we have too many tabs as is. Instructions and Criteria can both be trimmed/combined further together imho. ~ 🦝 Shushugah (he/him • talk) 01:39, 14 April 2023 (UTC)

I believe there's been talk of removing the proposal drive tab to make room for a backlog drive soon. I'd also support combining instructions and criteria (as well as WP:RGA and WP:GACN) into a single page. Thebiguglyalien (talk) 01:43, 14 April 2023 (UTC)
I wouldn't combine criteria and GAN instructions, as the criteria are also used for GARs, and nor would I combine WP:GACN into anything, as it's an essay. I would however merge WP:RGA and WP:GAI, as the former contains detail the latter should really have. ~~ AirshipJungleman29 (talk) 15:20, 14 April 2023 (UTC)

Proposal to Split Sports and Recreation into Seven Sub-Topics

See also: User_talk:ChristieBot § Feature Request

The sub-topic "Sports and recreation" is 83 nominations long as of this writing and has historically been the longest sub-topic at WP:GAN. I propose splitting this sub-topic into seven more refined sub-topics, based on the explanation at WP:GAN#SPORT of all the different article that fall under this sub-topic. Thus, "Sports and recreation" would look similar to "Natural Sciences", i.e. a larger grouping of specific sub-topics. I propose the following sub-topics, again based on WP:GAN#SPORT:

  • Sports and recreation
    • Football
    • Baseball
    • Basketball
    • Cricket
    • Hockey
    • Pro Wrestling
    • Other sports
    • Recreation

Obviously, the biggest benefit would be better categorization for editors who are interested in specific article topics. Interested to hear everyone's thoughts. « Gonzo fan2007 (talk) @ 22:02, 3 April 2023 (UTC)

It is a bit silly that this section isn't meaningfully divided into subsections. Is this list representative of the sports that are most common here? "Football" doesn't really work since there are two incredibly popular and completely unrelated sports by that name. "Association football" and "American football" would probably both need to be subsections in a list like this. Thebiguglyalien (talk) 22:20, 3 April 2023 (UTC)
I think seven subtopics is probably overkill. How about the three sports that are most frequently in GAN (possibly association football, American football, and basketball), "other sports", and "recreation"? (t · c) buidhe 22:20, 3 April 2023 (UTC)
Wikipedia:Good articles/Sports and recreation breaks it down like this:
  • Football (association, American, Canadian, Australian, and rugby)
  • Baseball
  • Basketball
  • Cricket
  • Hockey (field and ice)
  • Motorsport
  • Pro wrestling
  • Recreation
  • Multi-sport event
  • Other sports
I support using that. – Muboshgu (talk) 22:40, 3 April 2023 (UTC)
If there are 83 GANs divided between these categories, probably some will be empty or nearly empty most of the time, which will unnecessarily clutter the GAN page. (t · c) buidhe 01:59, 4 April 2023 (UTC)
Maybe I was too aggressive with providing a recommended list. I think my main proposal is to split it up some way to make it easier to filter out. However that break down ends up being would be fine with me. Thanks Muboshgu for catching my error and tightening up the list! « Gonzo fan2007 (talk) @ 02:14, 4 April 2023 (UTC)
Muboshgu didn't "tighten up" the list, they added an additional three categories that are not on your list. (t · c) buidhe 02:21, 4 April 2023 (UTC)
Assuming Wikipedia:Good articles/Sports and recreation GAs roughly reflect nomination ratios, then significant mileage could be had by just splitting out Football (footballs?). CMD (talk) 09:55, 4 April 2023 (UTC)
"Football" (Association, American and others), "Other sports", and "Recreation" would work for me. Great idea by the way Gonzo_fan2007 ~~ AirshipJungleman29 (talk) 10:33, 4 April 2023 (UTC)
I've brought this up before, but cue sports currently outweighs cricket for GAs (186 to 148). My big worry with splitting the nominations list, is that people who are say, into football would never get introduced to other articles in other lists (there's also some cross-over, plenty of pro wrestlers have American footballing history for example). Lee Vilenski (talk • contribs) 10:37, 4 April 2023 (UTC)
That's a good point about cross-pollination. If people are wanting to focus just on one sport, are GANs not popping up on Wikiproject article alerts? CMD (talk) 11:05, 4 April 2023 (UTC)
Oppose such heavy stratification, per Lee Vilenski: "My big worry with splitting the nominations list, is that people who are say, into football would never get introduced to other articles in other lists". There is a lot of shared knowledge across sports articles, so I think keeping them together makes sense. I do think splitting Sports and Recreation off, the latter for for theme parks, board games etc. would make sense. By rough count, that would only split off about 10 articles right now, but it would be a start, and hopefully a less controversial first step? Harrias (he/him) • talk 11:43, 4 April 2023 (UTC)
@Chiswick Chap: If sports and recreation were split, your yoga nominations would end up in recreation based on current categorisation. Would this bother you, would you be worried you'd be less likely to get attention for those reviews, or happier that they would be in a shorter list where they might get noticed? Harrias (he/him) • talk 11:46, 4 April 2023 (UTC)
Always hard to say; it might be better with a shorter and more specialised list and viewers more likely to be interested. I can tell you with certainty that I'm quite unqualified to review most "Sports" articles. Chiswick Chap (talk) 11:50, 4 April 2023 (UTC)
I think it's likely that especially with the current format, your yoga articles are getting lost in a maelstrom of football and basketball articles ~~ AirshipJungleman29 (talk) 12:11, 4 April 2023 (UTC)
* Support (two-way split to Sports/Recreation), ok, happy to go along with that and let's see what happens. Chiswick Chap (talk) 13:58, 6 April 2023 (UTC)
General comment I appreciate what everyone has said above. I am looking at this though as "knowing what we know now, if we started over, how would we split all the topics up". And with Sports and Recreation consistently pulling 10 to 15% (currently 13%) of total nominations, it just seems logical to reorganize the topic. Honestly, even splitting it off to two topics, "Sports" and "Recreation", as Harrias proposed, would help. In that light, I think AirshipJungleman29's proposal would be a good start and I would support that. « Gonzo fan2007 (talk) @ 14:33, 4 April 2023 (UTC)
I was just about to suggest that the obvious split would be to create "sports" and "recreation" when I saw your comment here. Trainsandotherthings (talk) 14:34, 4 April 2023 (UTC)
I also agree that we should instead split these GANs into "sports" and "recreation". That way, topics like parks and roller coasters don't get lumped in with sporting events and games. – Epicgenius (talk) 15:10, 4 April 2023 (UTC)
Unless there are objections I'll work on this approach first, since we can always split further. A simple split like this is probably a good way to start anyway so I can figure out what's needed. I have house guests and a trip planned so it's likely to be the middle of the month before I can make the change. Mike Christie (talk - contribs - library) 15:34, 4 April 2023 (UTC)
I support just splitting into sports and recreation. I think that's a good start for now. -- ZooBlazertalk 18:49, 4 April 2023 (UTC)
Support Splitting it into just sports and recreation. Having 7 subtopics for it is far too much IMO. Onegreatjoke (talk) 22:45, 9 April 2023 (UTC)
This is done, and it turned out to be fairly easy; if more splitting is needed I think it can be done quickly. Mike Christie (talk - contribs - library) 22:01, 13 April 2023 (UTC)
Mike Christie, looks like the "Reassessment" part of Sports and recreation needs to be updated. Currently Acrisure Stadium is showing as a Recreation Reassessment. « Gonzo fan2007 (talk) @ 17:19, 14 April 2023 (UTC)
That's not going to be easy to fix. The problem is that I get the topic from the historical GA database I built, which in this case probably pulled it from the old GAN page revision, here. Back then "Everyday life" was a topic, so that's what I've got recorded for the GAN. "Everyday life" is now a valid alternate name for a GAN nomination, and until yesterday it would have put the nomination in the "Sport and recreation" subtopic, but yesterday when I split them I had to decide where to put each alternate name, and I thought it made more sense to put that one under "Recreation". I can change it to point to "Sports", but video games were part of the "Everyday life" topic back then so those are going to go there. I think the best answer would be to add a subtopic parameter to {{GAR/link}}. It could be ignored except in cases like this where we need to override a historical topic mismatch. I'd still have to parse the template to get it but that shouldn't be too hard. Mike Christie (talk - contribs - library) 17:49, 14 April 2023 (UTC)
Mike Christie, is there a reason that reassessments need to be categorized under their respective sections anyways? Why not just add one catch-all "GAN Reassessments" section at the end of WP:GAN? This would delete 10 existing sections in the page (shortening up the TOC) and the new catch-all category would only have 17 pages on it. « Gonzo fan2007 (talk) @ 18:04, 14 April 2023 (UTC)
Sure, that would be easy to do. I think we'd have to get an OK from at least the @GAR coordinators: coords. Mike Christie (talk - contribs - library) 18:45, 14 April 2023 (UTC)
Assuming we don't suddenly have 50 GARs going on at once (which I think is quite unlikely) I don't see why not. Trainsandotherthings (talk) 00:34, 15 April 2023 (UTC)
I often look through the Mathematics and Literature sections to see if there is anything interesting to review. If GARs are listed in these sections, I will notice them. If they are all lumped together at the bottom, they could just as well be on a separate page, as I won't look at them. —Kusma (talk) 18:48, 14 April 2023 (UTC)

Possible GA error?

I noticed on Talk:Camille Cosby that the GA review is dated to 21 December 2023. Was there an error with the bot? Unlimitedlead (talk) 01:55, 16 April 2023 (UTC)

This edit fixed it. I guess the GA template doesn't like a comma in the date, and assumes the month and day refer to the current year. Mike Christie (talk - contribs - library) 02:00, 16 April 2023 (UTC)
Ah, thanks for clearing that up. Unlimitedlead (talk) 02:02, 16 April 2023 (UTC)

Premature GA nom

Could someone please look at Ludington Historian (talk · contribs · deleted contribs · page moves · block user · block log) in relation to WP:DCGAR and a recent GA nomination, following on this twice reverted most curious edit? (Have I mentioned before that GAN needs Coords????). It looks like WP:G3 removal would apply. SandyGeorgia (Talk) 08:08, 18 April 2023 (UTC)

Before I saw this I removed the GA nomination as not in compliance with the GAN instructions -- I would think you're the "significant contributor" that the instructions say should be consulted, Sandy, since you have the most edits of any unblocked editor. Mike Christie (talk - contribs - library) 09:51, 18 April 2023 (UTC)
Thanks, Mike; I wasn't sure how to proceed on this one, but AN/I may be the next stop. SandyGeorgia (Talk) 10:01, 18 April 2023 (UTC)
Speaking to the more general issue to "significant contributors", personally, I don't think that Sandy's contribution to this article would really necessitate consultation. Doug, if he were active, and Janweh64, but none of the other editors can really be said to be particularly significant. Personally, I'd be content with a nomination of such an article with a note that the most significant contributor(s) could not be consulted as they were now blocked, and a "notice of nomination" style post on the talk page, pinging Janweh64, with maybe a week left for responses. Harrias (he/him) • talk 13:34, 18 April 2023 (UTC)
I don't know how (and didn't even know that) significant is measured at GAN, but in case it's any help, I can say that there is unlikely to be any other active editor on Wikipedia who (now, that is, post-DCGAR) knows the Ludington sources as I do. That article is not even close to broad in coverage after the DCGAR issues/cleanse. (Without mentioning the poor writing still left over.) I'm further curioius, though Harrias, as to how any one can see a viable GAN there.
More relevant to this case was how to remove a GAN from an editor whose previous edits to the article appear as blatant but sneaky vandalism. It makes little sense to go through the motions of having a quickfail added to articlehistory. My suggestion is that Coords (thank you Mike for jumping in) would quickly remove such a spurious nom. SandyGeorgia (Talk) 15:27, 18 April 2023 (UTC)
"I'm further curioius, though Harrias, as to how any one can see a viable GAN there." – I have no idea; I wasn't commenting on that, just the concept of "significant contributors", hence saying "Speaking to the more general issue to "significant contributors".." With reference to judging what a significant contribution is, I was using the crude tool provided here. Your content contribution is less than 10%, although your total edits is around 20%. And yes, other things are much more relevant to the specific case. Again, that's why I specified I was talking about the general case. Harrias (he/him) • talk 16:28, 18 April 2023 (UTC)
Got it, thx! SandyGeorgia (Talk) 17:57, 18 April 2023 (UTC)
I don't want to immediately assume bad faith, but I'm sure we all see the Michigan-residing elephant in the room here? Trainsandotherthings (talk) 15:45, 18 April 2023 (UTC)
I thought of that but I don't think so in this case; seems more trollish, with the sneaky vandalism. I think with Doug it was a CIR issue, not vandalism. Mike Christie (talk - contribs - library) 16:55, 18 April 2023 (UTC)
Doug has now been community banned and had two likely sockpuppets/meatpuppets blocked. Not to assume either, but it's not impossible that he has lost any desire to contribute in good-faith, but simultaneously refuses to leave. Iazyges Consermonor Opus meum 16:58, 18 April 2023 (UTC)
TAOT, I don't suspect that to be the case ... but something (else) is wonky. SandyGeorgia (Talk) 17:58, 18 April 2023 (UTC)
I don't think it's Doug, but I do think it's related to him, and potentially a troll. Trainsandotherthings (talk) 18:53, 18 April 2023 (UTC)
I've blocked the user - their addition to James Ludington was a blatant hoax, and they are likely a sock of various accounts that were doing vandalism/trolling related to the Doug Coldwell ANI. Spicy (talk) 19:41, 18 April 2023 (UTC)
Thank you, Spicy! SandyGeorgia (Talk) 20:26, 18 April 2023 (UTC)

I mentioned this before...

I mentioned this before. It may be a relatively easy thing to do and is probably non-controversial. It's also related to the idea of "find the easy articles for newcomers". I'd like to see an updated list of top 10 or 15 hard articles. here "hard" is defined as some ratio of most cites and most readable text. Or just pick the top 6 of each of those stats, make a list of 12, and call them the "dirty dozen" (wait, that sounds negative...). The idea of course is to make it easy for people who like to help out by washing the worst dishes. § Lingzhi (talk) 14:02, 19 April 2023 (UTC)

I liked this idea when you first suggested it. I think something like this is a relatively inexpensive way to give a little push toward those more difficult reviews. Thebiguglyalien (talk) 14:36, 19 April 2023 (UTC)

A-class

There is some discussion at Wikipedia talk:Content assessment about what to do with the A-class quality assessment, and I thought I would gauge people's opinions here. There is confusion about whether A-class is or should be higher than GA, lower than GA or about the same. Is it useful to have a classification between GA and FA? Would it more useful to have a classification between B and GA? Or is A class unnecessary? — Martin (MSGJ · talk) 13:16, 17 April 2023 (UTC)

Part of me says it should be "Between B and GA", as we all want a high percentage of articles to be "decent" as GAs are supposed to be; and with the increasing pressure on GAs to be almost FA-perfect, and the GA queue seemingly getting longer and longer, it may be that a not-quite-GA would be useful. I think that trying to split the hair's breadth between GA and FA is becoming impossible, if it was ever really doable, for the same reason, that GA has been creeping FA-wards, with reviewers starting to check all refs and demanding that all claims be reffed, for example (something we're all coming to expect, but which was once explicitly not part of the criteria). But what the non-quite-GA should be, if it's to be "decent but simpler to do" might be problematic.
Perhaps the question that should exercise minds is "What percentage of Wikipedia do we want to consist of 'decent' articles?" If the answer is "say, 25%, and of the rest, most are clean stubs or very short 'Start' articles", then with the GA+FA+FL total today of around 1/139 = 0.7%, we have a very low baseline to start improving from. Right now, we have a few 'decent' articles, a pile of what a distinguished editor memorably called "shitty stubs", and a large quantity of what are politely called "C-class" articles, i.e. quite long, with several citations, and a large amount of what might well be original research, guesswork, garble, or vaguely-attributed text. If an "A-class" is to impinge on this enormous problem, then it must be easy to apply, meaning simple to edit and quick to review. Chiswick Chap (talk) 16:24, 17 April 2023 (UTC)
A class is higher than GA because it is usually only awarded after a multi person peer review by the MILHIST project. The designation is not useful for projects with less manpower. We generally have too fine assessment categories. Merging Start and C would probably not destroy much useful information, given also that many ratings are 10 years out of date and we have no mechanism to systematically update them. —Kusma (talk) 18:14, 17 April 2023 (UTC)
Yep. (t · c) buidhe 19:55, 17 April 2023 (UTC)
I think we have two separate threads here: the "should be" and "is today". They aren't at odds with each other, just different topics. Chiswick Chap (talk) 20:06, 17 April 2023 (UTC)
Yes, you are right. I only talked about what is, not what should be. I personally don't think A-Class is needed, but that is perhaps mostly because I haven't had any real interaction with A-Class reviews and A-Class articles in the last 15 years. There were various useful thoughts related to assessment issues over at Iritalk a while back, see User_talk:Iridescent/Archive_50#Content_Assessment. —Kusma (talk) 08:25, 18 April 2023 (UTC)
GA is one reviewer, MILHIST A-class is multiple reviewers. On that basis I'd say A-class is "higher", but I don't think it interacts with the GA process either way. CMD (talk) 02:34, 18 April 2023 (UTC)
There aren't very many A-class projects about. personally, I've always thought we should change the name of what the A-class is to ease confusion on this. I think people generally understand that GA comes after B, so the A class can be a bit bewildering. Lee Vilenski (talk • contribs) 09:41, 20 April 2023 (UTC)
One thing that doesn't help is that A-Class articles can be GAs and display a GA icon (like Arado E.381) or not be GAs and not display a GA icon (like Benjamin Franklin Tilley). I don't think we should have A-Class icons on the page as A-Class articles are so rare, but the situation is not ideal. —Kusma (talk) 09:47, 20 April 2023 (UTC)

Concerns about a reviewer

I've recently come across the reviews of User:UnidentifiedX, a relatively unexperienced editor (less than 500 edits total, 68 edits in mainspace mostly on one Start-class article, and no articles created) who has performed multiple hasty GARs in the last several months. The list on their talk page no longer links to articles, but can be seen in history, including odd claims about some being "doubly checked" by other editors. A review of Isles of Scilly they had conducted was recently delisted, but several are still up and at least one is in very poor shape.

I left a note on their talk yesterday where I explain my concerns in detail and use specific examples to give them a chance to respond before I request any reassessments. I have yet to hear back (they have not been active since March) and was wondering if others could chime in regarding the best course of action. In short, they tend to use a GAR criteria template as a checklist and routinely pass articles with very little feedback. As I explained in my message to the editor, several of their reviews were of content written by highly experienced editors where their minimal intervention would not be as much of an issue; however, their hasty review process seems to have been applied to every GAN, which is serious cause for concern.

I think the reviewer, who says they enjoy editing and expanding articles, as well as conducting GA reviews. I am familiar with what a good article is and what they should look like, so if you are ever in need of assistance contact me via my talk page may have good intentions. Nonetheless, it's a pretty unfortunate situation, particularly for editors without much experience whose articles may have well been erroneously promoted. I am afraid that all their GA reviews warrant additional scrutiny. Ppt91talk 19:04, 21 April 2023 (UTC)

I would suggest listing them at good article reassessment. With that said, "checklist" reviews are completely acceptable under the good article instructions as they're currently written. The actual guideline that governs GA reviews isn't much better. Thebiguglyalien (talk) 19:09, 21 April 2023 (UTC)
@Thebiguglyalien Yes, but the question is whether all of them should be reassessed. Using the checklist itself isn't really the issue as much as their poor understanding of the criteria and the haphazard manner in which the checklist is used. In one instance, which I described on their talk page, reliability of sources was left "Undetermined" and the article still passed. This is why I am concerned each of their reviews might need to be reassessed. Ppt91talk 19:15, 21 April 2023 (UTC)
A checklist review is not acceptable under the good article guidelines, the guidelines say "don't just leave an all-positive checklist". CMD (talk) 01:52, 22 April 2023 (UTC)
@Chipmunkdavis Thanks for pointing that out. Would you recommend reassessing all articles reviewed by this editor? I'm going to go ahead and request reassessment of the one I found to be most problematic for now, but am not sure about the rest of them. Ppt91talk 21:22, 22 April 2023 (UTC)
Sometimes I wonder whether we do need coordinators who check the reviews before the bot hands out the green plus. Perhaps the first two/three/five reviews of any new user should be put into a holding area by ChristieBot, and only released after an OK by a more experienced editor.
Slightly less workload intensive, we could have a spot on the main GAN page where we list links to the most recent reviews by inexperienced reviewers, making it easy for others to see them and gently intervene if necessary.
The bad thing about either of these ideas is that they might make it harder for us to recruit new reviewers. —Kusma (talk) 19:59, 21 April 2023 (UTC)
Yeah, I'm becoming increasingly convinced that GA coordinators are necessary in some form. The idea that any editor can say "yes, this is a good article" might be good in theory, but it doesn't work in practice without some sort of oversight. Our current method of oversight is a game of whack-a-mole, and we're losing. Thebiguglyalien (talk) 21:19, 21 April 2023 (UTC)
The idea of reviews by editors with little reviewing experience being put in a queue to be "review assessed", or at least visibly flagged as needing a spot check, was floated not that long ago and got support. Unfortunately it was not implemented. Unless a nominator flags a questionable review, or someone happens to find it by chance, it's tricky to spot. Having better visibility of reviews by newer reviewers would at least make them less likely to fall through the net. Bungle (talk • contribs) 21:25, 21 April 2023 (UTC)
The bot could post a thread on this page for new reviewers -- e.g. when the review is first created. Then any GA regulars who want to can watch that review page and keep an eye on the review. How does that sound? Mike Christie (talk - contribs - library) 21:29, 21 April 2023 (UTC)
@Mike Christie I like that idea. Even though I'd be right at the "experienced" threshold in regard to my numbers (I am currently at my 7th review), I think that would be very helpful. Do others, especially those with more experience, agree? Ppt91talk 21:11, 22 April 2023 (UTC)
I'm not 100% sure. Say an user reviews 6 GANs, and the first 5 are flagged as needing reviewed by an another editor. Upon checking, let's say the 5 flagged were inadequate reviews. Would the 6th review require checking as well? Perhaps there could be a to-check list where new reviewers with all of their reviews are listed, whether they are flagged or not. Then, if the reviewer has been making proper reviews and approved by another experienced editor, the new reviewer could be removed from the to-check list. I don't have a preference with the amount of reviews flagged to be checked. MrLinkinPark333 (talk) 22:13, 22 April 2023 (UTC)

Some thoughts from reviewing nominations from nominators with no GAs

Since I'm one of the people who supported the idea of putting new nominators at the top of the subsection lists, I've been mostly reviewing those nominations for the last month or so, and I have a couple of thoughts I wanted to share.

  • The failure rate is *much* higher than I expected -- close to 50%. That's been sufficiently depressing that I've deliberately gone out and picked a couple of articles by experienced nominators to review, to reassure myself that I'm not being unreasonable.
  • There have been some very rewarding reviews of nominations by new nominators. In a couple of cases I've suggested to the nominator that they consider reviewing, and at least one has started doing so, so that feels like a win.
  • I am astonished by the number of nominations there are from people who have never reviewed. At the moment there are well over six hundred nominations; 195 of those are from nominators who have never reviewed. Adding in nominators who have only reviewed one article brings the number to 229 -- well over a third of GAN. This seems undesirable to me; it means we have a small community of regular reviewers trying to service an essentially infinite demand.
  • Most of the failures have been well received by the nominator; I've only had a couple of cases in which the nominator felt the fail was unjustified.
  • The failures seem to break down into three groups:
    • Obvious immediate fails -- the article isn't close to GA and only the most superficial scan gives immediate reasons for a fail. There weren't that many of these.
    • Failures because of the spotcheck or other source related issues. Most were in this category. I have started doing the spotcheck first, instead of last, if I suspect that there are going to be source problems, because if it's a bad fail it saves time -- there's no point continuing with the review if the article is going to have to be rewritten to deal with source issues.
    • Miscellaneous. In a couple of cases I failed an article because it was essentially a driveby and I didn't notice that -- in one case the main contributors both said it shouldn't have been nominated; in another case the nominator made a valiant effort but there were still sourcing issues. There have also been prose problems in some cases that were bad enough to lead to a quickfail, though these were often combined with source issues.

One of the reasons I think it's important to review nominations from new nominators quickly is that we want to get editors involved with the GA process; we don't want their first experience to be a six month (or longer) wait. Robminchin is the nominator who is now reviewing an article, and I'd be interested to know his thoughts on that. But what about nominators who fail their first, or second, or third nomination? Do they stay at the top of the list? For example, we have some editors with zero GAs and zero reviews and multiple nominations. I've just started reviewing one of their articles. If that were to fail, that nominator's nominations would still be at the top of their subsection list. Is that what we want?

Comments that have been made here about the new sort order over the last couple of months include:

  • Some want to return to the old sort order -- strictly by date.
  • Some like the new sort order and want to stick with it.
  • Some have suggested sorting by a combination of age and the new sort order, perhaps by adding together weightings from both.
  • A couple of people have commented that the new sort order has prompted them to review more, which was one of the effects we were hoping for.
  • A presentation that allows everyone to choose their own preferred sort order has been suggested -- it would probably have to be a sortable table inside each subsection.

Personally I think there have been benefits from the new sort order but all the points made seem reasonable to me. I don't want to go straight to an RfC about how to change the sort order again, but I was hoping some discussion of pros and cons would be a useful preliminary to any discussion. Mike Christie (talk - contribs - library) 14:47, 17 April 2023 (UTC)

I think one thing we need to address is that writing content and reviewing it are two different skillsets and two different areas of interest. There will be editors that enjoy both and are proficient in both, but it's not a guarantee. Another thing is that source checks are an absolute pain in the ass, and they're probably the main reason that I don't review as much as I could. Thebiguglyalien (talk) 15:20, 17 April 2023 (UTC)
Source reviews are a complete pain in the ass, and they are also probably the single most important part of the review. Like Mike Christie, I've been doing this first for a while now, because issues with the sourcing can be much more terminal to a review than most prose issues.
Regards sorting, I favour the combined approach; I think there needs to be some age-related factor to progress articles up the list, but I think some priority should still be given to new nominators and frequent reviewers. Harrias (he/him) • talk 15:41, 17 April 2023 (UTC)
Personally, I think passive actions, like cosmetic adjustments to sort order and adding a GA/review count, are never going to do much. At the end of the day, GA reviewing just isn't interesting. Backlog drives, being intrinsically active and rewarding, are the only thing that consistently bring the waiting list down. A month and a half ago, I suggested to an experienced editor that they try to review more; while they responded somewhat positively, they've nominated ten articles since then and reviewed none.
I know that consensus is generally against any sort of quid pro quo, but I can't help but feel that a limited/leveled enforced system would be beneficial. Say, your GA/review ratio is more than 2 (with a minimum of say 30 GAs), you have to review two articles per nomination, and so on. I know that many people would likely be against this because they believe it'll lead to bad reviewing, but if you already have 30 GAs you're probably less willing to jeopardize the work you've already done. ~~ AirshipJungleman29 (talk) 18:06, 17 April 2023 (UTC)
I agree as far as cosmetic changes. We need to make the reviewing process easier at a fundamental level. Anecdotally, I can say that I feel more comfortable reviewing WP:DYK posts. Obviously there's less expected of those, but I think what really makes the difference is the existence of WP:DYKR. A simple checklist style reviewing guide goes a long way in keeping the process straightforward and amateur-friendly. This relates to what I've been saying for a while, in that I think we need to rework WP:GANI, WP:GACR, and WP:RGA into something more accessible. RGA is the "official" reviewing guideline, but it's tucked away in a corner and hasn't really been substantially edited in years. I'm pretty confident in saying that I'd review more often if there were a more standardized process for source checks rather than trying to figure it out myself each time I do a new review.
In terms of short term solutions to alleviate the backlog, I think a backlog drive would be incredibly helpful at this point. It seems there was broad support for one up above, but it's hung up on the matter of who's actually going to set it up. Thebiguglyalien (talk) 19:24, 17 April 2023 (UTC)
As seen from the proposal drive, some editors don't give a damn about the backlog, they just want to do nominations and don't see anything wrong with not reviewing, to them that's someone else's problem. I choose not to review any nominations by those who refuse to review, and invite other reviewers to join me. Yes, we can do more to make reviewing less time-consuming and tedious, but at the end of the day those of us who do review should focus our attention on nominators who do review or are new, in my opinion. Backlog drives are just a band-aid, the backlog will only keep growing unless we achieve structural changes to both the review process and reviewer mentality. Trainsandotherthings (talk) 20:20, 17 April 2023 (UTC)
I agree that backlog drives are a band-aid. I wouldn't go so far as to oppose a backlog drive, but I don't think the GA process will be healthy until they're not needed. Mike Christie (talk - contribs - library) 20:34, 17 April 2023 (UTC)
And IMO backlog drives create a perverse incentive to lower review quality and increase review speed. For instance, how many of the Coldwell GAs came from GA drives? (And I'm not saying that I'm innocent of any role in the Coldwell situation; I've significantly tightened my GA reviewing standards as a result). With backlog drives in the past, we've either almost encouraged shoddy reviewing, or delegated policing reviews to an overworked set of coords (often just TAOT). There must be a better way. Hog Farm Talk 21:13, 17 April 2023 (UTC)
Is there any incentive type system that this doesn't apply to? I'd be far more worried about shoddy reviews with a quid pro quo setup than I would with a drive that's being monitored. Heck, are we doing anything to police shoddy reviews now? Thebiguglyalien (talk) 22:01, 17 April 2023 (UTC)
As a backlog drive coordinator I introduced stronger review requirements and an oversight system, which does not exist outside of drives. That said the ideal solution is not to need drives. (t · c) buidhe 23:49, 17 April 2023 (UTC)
@Mike Christie Thank you for writing this. I like the current grouping order, although I usually scroll through the entire list in my area of interest anyway, so the order issue isn't as pressing to me. What I see as a more urgent problem, however, is the ratio of reviews/reviewed you mention. I take the suggestion of 2 reviews per each nomination to heart as an important part of giving back to the community (with limited experience compared to other reviewers here)--maybe it's that I have more confidence to actively review by staying within my comfort zone or maybe because the very activity of reciprocal reviews is part and parcel of academia--but I also cannot blame those who are turned off by the potential stress associated with GAR activity and by its often arbitrary nature; just recently I saw an article nominated by an experienced editor failed by another experienced editor on formatting grounds that felt, well, like a very easy fix.
My reviewing approach is to try to work with the nominator to get the article through and I don't mind spending a significant amount of time doing so. Of course, that's not always possible and I have my shortcomings, but I try because the state of WP:WPVA isn't great and needs both committed editors and committed reviewers. At the same time, I can see why people would want to avoid participating in the process while still being prolific GA authors. They nominate multiple articles hoping to find some interest and then are actively participating in the review process; in other words, it's still very much net positive as far as the readers are concerned. Conversely, I've encountered valuable contributors who avoid GAR altogether finding it too overwhelming and preferring to stick to C/B territory without the extra hassle.
The DYK system, I think, works because of the limited scope of the task compared to GAR. Requiring a review for a nomination might actually end up alienating a number of current GA authors and new authors who already find the process to be daunting. Perhaps one solution to this problem is some kind of author/reviewer retention program or a "school" akin to the kind we have for NPP with an aim of making the environment more welcoming and appealing. That could come with a GA Reviewer topicon or some other kind of incentive or form or recognition that would help attract future editors (not to encourage hat collecting, it was just the first idea that came to my mind in terms of potential editor recognition tools). Does anyone feel like some kind of retention program and/or academy could be a potentially helpful idea? Just my 2c. Ppt91talk 23:10, 17 April 2023 (UTC)
The "school" is probably the only reason that I haven't tried my hand at NPP. I'm not sure if I would enjoy doing NPP work. It's something I'd consider doing to get a feel for it, but I'm not going to commit to a whole process for something that odds are I won't find interesting. To my knowledge, a system like that exists to intentionally slow editors down rather than bring in more, because places like NPP have much higher stakes associated with bad reviews. Thebiguglyalien (talk) 23:36, 17 April 2023 (UTC)
Just a heads up, you don't need to pass NPP school to get NPP. I didn't, I just applied after being recommended by an existing NPPer and was granted the permission. Trainsandotherthings (talk) 23:43, 17 April 2023 (UTC)
@Thebiguglyalien I see your point, I guess it seemed like the best analogy to illustrate the idea, but I am not suggesting to replicate the model. In fact, I don't find the NPP school structure and "graduation" process appealing at all. I was more so thinking that the existence of some kind of "academy" (that name seems more appropriate) would simply make GAR a more appealing subject for new editors and also create an incentive for ambitious editors. It all boils down to the retention question in my opinion. Creating a better and more stimulating sense of community among reviewers and writers would certainly help achieve that. Ppt91talk 23:49, 17 April 2023 (UTC)
There does need to be something. I'm pretty sure there was support for some sort of mentorship for new reviewers, but it doesn't look like anything has come of that. Maybe some sort of obvious "Want to review? Click here!" that leads to a dedicated landing page specifically for new reviewers. Basically something that would make it as painless as possible to get started. Ideally, this would include an easy way to get help without having to reach out to a specific editor. Maybe it would take the form of a list where you add your name and areas of interest, or maybe a dedicated talk page for "reviewer help" questions.
One possible mentorship strategy I haven't seen mentioned yet would be matching new reviewers to easy reviews. Experienced reviewers can (usually) tell at a glance what reviews will be easy or hard based on length, sourcing, prose, etc. Another thing is that there was discussion about a talk page notice for people that nominate a lot and might be interested in reviewing; I think we can retool this to provide a general "invite/welcome to reviewing" that includes helpful links that are enough to get a new reviewer started. Thebiguglyalien (talk) 00:13, 18 April 2023 (UTC)
I like this kind of idea matching new reviewers to easy reviews and an invite; tell people interested in reviewing for the first time to ask about it, and someone will point them to a review to handle and that user will be their point guy? Could definitely work. Kingsif (talk) 21:16, 18 April 2023 (UTC)

(←) There are 8 million Welcome templates. Many of them include "You can contact me if you have questions." Find maybe two or at most three that look good, change a few words to invite people to be GA reviewers. Then template them. § Lingzhi (talk) 02:09, 18 April 2023 (UTC)

I think it's great news that there is incentive for source checks to be done early on in the process. They are the most crucial part of a Wikipedia article, and among the hardest to fix if problematic, so it makes a lot of sense to front-load them in a GAN. Regarding list order, I don't have an issue for those without GAs to be at the top of the list, I would suggest editors unfamiliar with GA are not those you'd like to incentivize to review. CMD (talk) 02:41, 18 April 2023 (UTC)
Speaking of which, see Agriculture in Turkey, which was nominated by Chidgk1, and picked up for review by TheTrueSauce with their fourth edit. CaptainEek has stepped in and offered to review but I wonder if it would be better to delete the review page as out-of-process, though TheTrueSauce is right to say (on their talk page) that we have no rule against this. And I agree with CaptainEek that nominating something that isn't ready because you expect it to be in the queue a long time isn't the right approach. Mike Christie (talk - contribs - library) 09:49, 18 April 2023 (UTC)
Gatekeeping Wikipedia, more then it already is, would be contrary to the principles of the site, which aim to provide open and collaborative information to the public. TheTrueSauce (talk) 12:17, 18 April 2023 (UTC)
And as part of that collaboration, the GA process aims to provide articles reviewed to a relatively high level by editors who are qualified by their previous actions to make these judgements. ~~ AirshipJungleman29 (talk) 12:34, 18 April 2023 (UTC)
It sounds like Chidgk1 feels it is not ready anyway, so deleting and removing the template seems reasonable. CMD (talk) 13:03, 18 April 2023 (UTC)
My thoughts on newbie GA reviewing align with Remy the rat, from Ratatouille: "Anyone can GA review". That doesn't mean "everyone can GA review, but rather that a great GA reviewer can come from anywhere." It is not impossible for a new editor to be a good GA reviewer, though it is certainly unlikely. So we have no absolute proscription on the practice. But it is also generally inadvisable for a new user to take up a GA review. However, GA reviewers do have to start somewhere. My first GA review was...a definite learning experience. I thought you either passed or failed right away. I didn't realize it was a collaborative advice giving process aimed at getting the article to a pass. It wasn't my first edits, but it was in my first 100: Talk:Metallic Metals Act/GA1. Now, I expressed a willingness to learn, and an understanding that I might have been wrong, and some helpful folks helped me out in the process, and I've now gone on to do many more GA reviews and write several of my own. The lesson I learned is that I think the reviewing criteria don't do a good enough job of explaining what the process is really like. I'll ruminate on how we could improve our documentation. But beyond that, I'm not honestly a fan of putting some hard restriction on GA reviewers. I think that would have kept me from advancing as an editor. Perhaps the better solution would be that first time reviewers should have their work checked by an experienced reviewer, with an eye towards being helpful and mentoring? CaptainEek Edits Ho Cap'n!⚓ 19:55, 18 April 2023 (UTC)
I agree with most of what you say. It would be easy to flag or list reviews by new reviewers, if we want to. Who would review them though? We could keep a community list (e.g. in a box on this page) of "unvalidated first-time reviews", for any experienced reviewer to remove after validating. Mike Christie (talk - contribs - library) 20:54, 18 April 2023 (UTC)
nominating something that isn't ready because you expect it to be in the queue a long time isn't the right approach – Unfortunately, it's one that we're incentivizing. Thebiguglyalien (talk) 20:00, 18 April 2023 (UTC)
IMO reducing this incentive is a significant advantage of the sort order reorganization. (t · c) buidhe 21:07, 18 April 2023 (UTC)
Unless you're a new nominator, the reorganization incentivizes it even more. The longer end of the wait time range has been increasing significantly since the change. Thebiguglyalien (talk) 21:11, 18 April 2023 (UTC)
FWIW TheTrueSauce has been blocked as a sockpuppet. —David Eppstein (talk) 20:50, 23 April 2023 (UTC)
The usual unhelpfulness about not tagging socks at play there, but it seems a reasonable assumption that this is the same LTA as the chatgpt reviews. Perhaps we should be triaging more actively, or restricting to a certain edit count. CMD (talk) 01:43, 24 April 2023 (UTC)

February 2023 reviews - Part 2

Looking at @Chipmunkdavis:'s list at Wikipedia talk:Good article nominations/Archive 27#February 2023 raised reviews, there are several articles that need to be rechecked for one of the following reasons:

MrLinkinPark333 (talk) 20:08, 22 April 2023 (UTC)

Further evidence that we need GA coordinators/patrolers to check off reviews as they go through and to make sure rechecks these get addressed every time. Thebiguglyalien (talk) 20:59, 22 April 2023 (UTC)
I can probably pick up the Eurovision one, I've done some related GANRs before. Kingsif (talk) 23:33, 22 April 2023 (UTC)
Awesome! There's quite a few Eurovision ones in the backlog (both individual years & individual countries). MrLinkinPark333 (talk) 23:39, 22 April 2023 (UTC)
I'll drop this reminder to anyone reading that technically my GANR request page is still open - if you think any open nom fits my interests, feel welcome to recommend I review it! Kingsif (talk) 23:41, 22 April 2023 (UTC)
@MrLinkinPark333: Probably not how you want to end a 9-month wait, but now I'm concerned over the sourcing quality of all the quite a few similar GANs you mention - are any on review (if you know) and, if not, can I have a stab at just source checking them (and, if poor, suggesting withdrawal?) Kingsif (talk) 02:48, 23 April 2023 (UTC)
@Kingsif: I was mentioning them in general as you mentioned you've done Eurovision ones before. Most of them are available at Wikipedia:Good_article_nominations#Other_music_articles and at the bottom of the section if you are interested. As for quickfailing the 9 month one, I recently quickfailed a 6 month one for a lot of uncited parts. So, it's not just you who've quickfailed a super old one. MrLinkinPark333 (talk) 02:57, 23 April 2023 (UTC)
I find myself more and more convinced Sandy was right about needing coordinators. I was hoping we could keep things less bureaucratic at GAN but it's quite clear more oversight is needed. I don't really see any alternatives. Trainsandotherthings (talk) 03:32, 23 April 2023 (UTC)
Agreed. Iazyges Consermonor Opus meum 17:21, 24 April 2023 (UTC)
Some sort of oversight, definitely. Can't have Mike doing it on his own forever. ~~ AirshipJungleman29 (talk) 17:36, 24 April 2023 (UTC)

Review advice needed - Crocheron–McDowall House

@Morogris: has nominated this for GA. Beautiful property that has had a lot of editors over the years. It's already NRHP. On one of my own sandboxes, I started a preliminary review and read-through via the NRHP nomination form. Much of the text on the article is lifted in hunks from the nomination form. You know ... take a hunk of text and flip around the words/phrases in the sentence, etc. And that's just the history section. It's a waste of time to go any further with the review, but I'm wondering if the nominator should withdraw it, or what should I do? This is one of those cases where I don't think the WP:GOCE would be an option, because I'm not aware that they check sourcing, just the prose. Please advise how I proceed on this. Thanks. — Maile (talk) 18:03, 22 April 2023 (UTC)

P.S. also, I'm thinking we should perhaps revert that article back to its state when Nv8200pa created the page. I don't want to delete the article, but I also don't want to nitpick through several years of text to take out each plagiarism. — Maile (talk) 18:13, 22 April 2023 (UTC)
When I run into a nomination like this, I usually fail it with a list of the examples. I've failed two just today for this reason: Talk:Homi J. Bhabha/GA1 and Talk:Death of Patrick Cronin/GA1. Since you haven't started the review you could ask the nominator if they'd like to withdraw, but to be honest it doesn't hurt to have the GA review page exist so that future reviewers know it's something they should be checking. Mike Christie (talk - contribs - library)
Ah, yes. That's good advice. Even better to have an existing failed review page. I have a few little notes already, so maybe I'll do that. — Maile (talk) 18:25, 22 April 2023 (UTC)
Mike_Christie I went through the template to reject the nomination. It's been years since I've reviewed at GA, so I hope someone else will note whether or not I correctly closed the nomination. Too long ago for me to remember the steps. Thanks. — Maile (talk) 21:14, 22 April 2023 (UTC)
Looks like you did everything right, though I did just correct a typo in the topic. That makes no difference to the process, though; the bot's removed it from the list. Mike Christie (talk - contribs - library) 22:18, 22 April 2023 (UTC)
Thanks. I also just added a few comments on the template for the editor, to encourage them to keep going at this. It has potential for GA. — Maile (talk) 22:54, 22 April 2023 (UTC)
I've commented on my userpage to address this. I'm trying to do a text-by-text comparison but this won't pull in Earwig due to image/PDF issues not being captured. Looking for a workaround. This already went through a GOCE and I'm happy to give it another trim, just want to make sure I'm not missing any phrasing matches. Morogris (✉ • ✎) 19:15, 24 April 2023 (UTC)

Second opinion requested

Above everything else, I would like to thank Kingsif for his time in reviewing Eurovision Song Contest 1999, however unfortunately I am not satisfied with how the review process for this article has played out since it was nominated back in July 2022. I am generally fine with waiting a while for articles I have nominated to get reviewed for GA; I understand Eurovision Song Contest related subjects are not everyone's cup of tea, and there have been a lot of related nominations from different editors over the past couple of years, sometime consecutively. However I do not believe that after 9 months of waiting for a review, including a false start with a review from a sockpuppet using ChatGPT, that this article deserves to be quick-failed. Of course I will concede that there may be issues with the article, as every article has issues that need worked on and there will always be points for improvements that come from a GA review, however I believe that Kingsif has been unduly harsh in quick-failing this nomination without any attempt to work with me to improve the article in any way. With this in mind, I respectfully request a second opinion on the review/article, as I do not believe this article to be so irredeemable that it deserves to have its GA nomination quick-failed. Sims2aholic8 (talk) 07:55, 23 April 2023 (UTC)

While I understand your frustration, source-text integrity is pretty much the single most important thing we can judge an article on. Everything else is irrelevant if the article fails to accurately reflect its sourcing. Furthermore, articles are expected to be in a reasonably close state to meeting the GACR when they are nominated, and you have (through no fault of your own) had plenty of time to review the article and ensure that it meets the GACR. With that many sourcing issues just from a single spot check, I don't think it's unreasonable for Kingsif to have concluded that the article "is a long way from meeting any one of the six good article criteria" and issue a quickfail. ♠PMC♠ (talk) 09:15, 23 April 2023 (UTC)
I think if I had were to review an article and find that many issues with a spotcheck I would probably have failed the article as well. I agree it seems cruel, particularly with the bad luck the article had with the sockpuppet review, but I think the reviewer did the right thing. I am sometimes more willing to leave a review open with a long list of problems if the article has been waiting a long time, but when spotchecks are the problem it's very time consuming to fix and I would usually not delay the fail. A separate point: the review mentions problems with citation formatting that go beyond what GACR requires. Of course those are good things to fix, but I think it's a good idea when reviewing to make sure the nominator knows which issues are required for GA and which are not. Mike Christie (talk - contribs - library) 10:59, 23 April 2023 (UTC)
Thank you for bringing this up, and thanks everyone for their input. I replied at the review page before seeing this, but I will say here that I believe when I mention things that are nice-to-haves, rather than GA criteria requirements, I did write something like "you could do this to improve"; do people have opinions on if this is clear enough (that the suggestion is not required), or should I explicitly say so? (@Sims2aholic8 and Mike Christie:) Kingsif (talk) 20:44, 24 April 2023 (UTC)
It seemed clear to me that you failed on the basis of the sourcing issues, rather than any of the ancillary commentary. ♠PMC♠ (talk) 20:56, 24 April 2023 (UTC)

Change in nominations by new nominators

When the new sort order was implemented, three months ago, one of the complaints was that there were a great many new nominators listed at the top of every subtopic, meaning that those nominators who had done plenty of reviews, and whom the new sort order was supposed to reward, were in fact pushed well down the list. At that time there were 114 nominations by nominators who had never had a GA promoted, and only 13 of those nominations were under review, leaving 101 in the queue waiting for a reviewer. Today there are 57 nominations by nominators who've never had a GA promoted, and 30 of those are under review. As a result, most of GAN's 30 or so subsections no longer have unreviewed nominations by new nominators at the top. This is something I hoped would happen. New nominators are now getting much faster turnaround time, which is good because their nominations frequently fail, and it's a lot easier to accept a failed nomination if you only nominated it a few days ago. I hope we'll see the number of not-on-review nominations by new nominators drop to single digits soon. I'm aware this doesn't address all the problems with the new sort order that some have raised, but I think it's a real benefit. Mike Christie (talk - contribs - library) 11:06, 25 April 2023 (UTC)

Is chess a sport or a recreation?

I moved a few nominations to the new "Recreation" subtopic, including Vera Menchik; the nominator (Sportsfan77777) has just put it back into "Sports". That's fine with me, but currently I have "Recreation" defined as "... board/card/role-playing games, chess, poker, toys, zoos, public parks, and amusements". If nobody objects I'll change the description to put chess in with the other sports. Any other changes needed to the definition? And while I'm starting a new topic, do we want to split the enormous "Sports" subsection into e.g. "American football and Association football" and "Other sports"? Mike Christie (talk - contribs - library) 23:59, 13 April 2023 (UTC)

Maybe basketball as well as there's currently 15 of them. MrLinkinPark333 (talk) 00:24, 14 April 2023 (UTC)
(edit conflict) The International Olympic Committee regards chess as a sport, so it should probably go there. The further splitting of the sports section was I think rejected above, if I read the discussion clearly (I do think splitting it would be beneficial, but whatever). ~~ AirshipJungleman29 (talk) 00:25, 14 April 2023 (UTC)
The current definition is fine and in line with the majority of English-language definitions. Following the above discussion I lean against pulling out individual sports, due to that limiting the number of potential reviewers who might scroll through it. Shorter lists exist in article alerts, eg. Wikipedia:WikiProject Football/Article alerts. CMD (talk) 00:27, 14 April 2023 (UTC)
Yes, chess works fine as a recreation, as defined. Of course, even yoga can be a sport if you don't think "competitive yoga" is an oxymoron. Yoga is fine as a recreation too. Chiswick Chap (talk) 13:15, 14 April 2023 (UTC)
If limiting the number of potential reviewers who might scroll through was an actual problem, we'd just have one big list of all 600+ nominations, no? ~~ AirshipJungleman29 (talk) 00:34, 14 April 2023 (UTC)
It was the point made be Lee Vilenski above. A balance between having specific slots for everything, and having a giant list. Someone who knows a lot about basketball probably has a good sports background to review an association football article, whereas I wouldn't expect the same general knowledge to transfer as easily to chemical processes or short-form literature. CMD (talk) 01:21, 14 April 2023 (UTC)
Chess==Sport; split==Support § Lingzhi (talk) 00:28, 14 April 2023 (UTC)
Move chess to video games. That's what you get for trying to start the "is chess a sport" debate.[sarcasm] Thebiguglyalien (talk) 01:45, 14 April 2023 (UTC)
Sportsfan77777 has twice reverted to "Sports" for their chess article, so I've just switched the definition -- it should take effect in a few minutes unless I screwed something up. I don't think there's a consensus here yet but I'm fine with going with the opinion of the nominator of the only chess GAN until a consensus does emerge. Mike Christie (talk - contribs - library) 13:21, 14 April 2023 (UTC)
Chess is one of those things which could be reasonably classified under either sport or recreation; I wouldn't quibble with a nominator's choice as to which to put it under. Personally I would consider some chess articles to be more on the sport side of things (serious competitive players such as Magnus Carlsen or Garry Kasparov; major tournaments like Tata Steel Chess Tournament and the Sinquefield Cup) and others to be closer to recreation (chess players who are best known as Internet Chess Personalities like Alexandra Botez and Levy Rozman; nonserious competitions among amateur players like Pogchamps). Caeciliusinhorto-public (talk) 14:52, 14 April 2023 (UTC)
Sportsfan77777, my bad on Talk:Vera Menchik. I didn't realize a discussion had already started on it and it had been reverted already. I don't have a strong opinion either way. « Gonzo fan2007 (talk) @ 17:16, 14 April 2023 (UTC)
As the reviewer of the recent successful GA nomination for "Garry Kasparov", I'd have no objecton to chess moving to "Recreation' in the interests of keepkng things more streamlined. Would it be helpful, or cause a row, to define 'Sports' as only those having an athletic or physical content? Billsmith60 (talk) 19:39, 19 April 2023 (UTC)
Well, no, as that's not what sports are defined as. Sports are specifically games that are done competitively. A walk in a park is athletic, but it's not a sport.
Chess is indeed classified as a sport, so I don't know why we'd suggest it wasn't one. Lee Vilenski (talk • contribs) 19:45, 19 April 2023 (UTC)
It's an Olympic recognised sport, believe it or not. Hawkeye7 (discuss) 21:25, 19 April 2023 (UTC)
If we're going to roll up everything with a competitive aspect into sports, then having the sports/recreation split here seems entirely pointless. (The original issue was the section was 83 long total, Sports is currently 80.) CMD (talk) 05:34, 20 April 2023 (UTC)
Chess is definitely a sport. Cheers  — Amakuru (talk) 10:21, 25 April 2023 (UTC)
Then let's revert the split. CMD (talk) 11:38, 25 April 2023 (UTC)

More scrutiny of reviews

The "one-reviewer decides" system of GAN is great because it is a very lightweight process, but it means that bad reviews (or even non-reviews) often do not get noticed by anyone but the nominator, who might have a slight conflict of interest whether to report the sloppy reviewing. If we see this as a loophole in the process that needs fixing, the standard answer is to have coordinators formally award the GA status after checking the review. Personally, I am not particularly keen on having another layer of process, so I would like for us to consider alternatives. One reason why I rarely look at reviews is that they are not collected and linked to very prominently; the only way to find them (other than the article talk page) is via the "discuss review" links at GAN (hidden in terribly long lists that I usually look through when I am after something else). For recently closed reviews, it is worse: the only way to find them is a few clocks away from Wikipedia:Good articles/recent. I would like to brainstorm some ways to make reviews much more visible and public. Here are some things we could do (not mutually exclusive):

  • Idea A Make lists of links to open reviews that include the reviewer and nominator data as on WP:GAN, so we can easily look for reviews by unexperienced reviewers.
  • Idea B Make lists of links to recently closed reviews to invite scrutiny
  • Idea C Make daily GA pages that transclude all GA reviews that have been opened on a given day, and encourage people to read the entire page (perhaps by linking to that page when notifying them of the GA review).
  • Idea D Have a dedicated group of volunteers ("coordinators"?) who look through such lists and remove links/transclusions of things that are not problematic.
  • Idea E Post all reviews opened by newbie reviewers in some place and invite a group of experienced reviewers to welcome the newbies and provide constructive feedback

As you can see, I don't know exactly what we should do, but perhaps it is worth thinking about ways to improve review quality without adding a new layer of process, ideally in a way that minimises any assumptions of bad faith and results in better onboarding of new reviewers. Happy to hear better ideas. —Kusma (talk) 12:48, 23 April 2023 (UTC)

ChristieBot keeps an internal audit trail of all these things and it would be a simple matter to create some or all of these as 90-day log pages in table form. Then whoever wanted to watch any of these pages could do so. A blank column could be added to record "checked" or other notes. Mike Christie (talk - contribs - library) 13:15, 23 April 2023 (UTC)
It is great to hear that these ideas are not too hard to implement from a technical point of view. The hard question, of course, is whether any of these ideas would help to make GA reviews better while maintaining GA as a lightweight process. —Kusma (talk) 16:13, 23 April 2023 (UTC)
This, which seems similar to idea D above, is most preferable imo. Same sort of deal as WP:URFA/2020. ~~ AirshipJungleman29 (talk) 18:29, 23 April 2023 (UTC)
I would support any of these, depending on what's determined to potentially be the most helpful. Thebiguglyalien (talk) 16:49, 23 April 2023 (UTC)
I think having experienced users help new reviewers with the GAN process would be great (Idea E). However, I think that this co-review should happen when the review is open. Therefore, any issues with reviewing could be addressed earlier. This could help prevent having inadequate reviews reverted and sending these articles back to GAN. On the other hand, I think reviews that were conducted and closed by new reviewers should also be re-checked by experienced users. If these closed reviews are found to be inadequate, then they could be reverted. MrLinkinPark333 (talk) 17:52, 23 April 2023 (UTC)
Would it be worthwhile to have a bot-generated list of all reviews opened each month or week? I would not be opposed to any of the ideas here, though at least GA coords would probably require a RfC or similar to implement. Trainsandotherthings (talk) 20:01, 24 April 2023 (UTC)

Here's what such a list could look like. This is a list of all reviews opened this month where the reviewer has less than 3 completed reviews. Mike Christie (talk - contribs - library) 21:00, 24 April 2023 (UTC)

That's a nice looking table! I was wondering about users who started out with zero reviews and had 3 or more reviews by the end of the month. Would their first two reviews be covered in the table? MrLinkinPark333 (talk) 02:18, 25 April 2023 (UTC)
At the moment I don't store what the stats were in the past, but the data is available so I could probably do a retroactive end of month report that accurately showed reviews that were first or second for that reviewer, regardless of how many they had at the time of the report. However, it might be easier to just run the query daily and append to a log. Mike Christie (talk - contribs - library) 02:54, 25 April 2023 (UTC)
If a daily run is easier, then that works for me :) MrLinkinPark333 (talk) 03:01, 25 April 2023 (UTC)
I guess I'd like to see a consensus here that this would be a good thing to do - not just to make sure someone will be making use of it, but also because I recall at least one person (I think it was CMD) saying they thought this approach might be BITEy. Mike Christie (talk - contribs - library) 03:23, 25 April 2023 (UTC)
I'm going to make a bold proposal here and say that we should also be checking the work of more established reviewers. I've seen usernames that I recognize doing bare minimum reviews and making no mention of referencing. A common pitfall I see is reviewers evaluating criterion 2 by suggesting cosmetic changes to references (such as formatting ISBNs or linking publishers) rather than evaluating text-source integrity. I suppose I'm arguing in favor of Idea B and probably Idea D. Thebiguglyalien (talk) 01:16, 26 April 2023 (UTC)
April reviews by new reviewers
title                             nominator       reviewer                     review_started_ts   subtopic                       num_reviews
Talk:Beckman–Quarles theorem/GA1  David Eppstein   The person who loves reading 2023-04-01 16:03:39 Mathematics and mathematicians 2
Talk:Annunciation Seminary (Pengzhou)/GA1 TheLonelyPather The person who loves reading 2023-04-02 19:50:45 Art and architecture           2
Talk:Shōnen manga/GA1 Morgan695       Cessaune                     2023-04-04 17:01:58 Language and literature         2
Talk:Island of Lost Souls (1932 film)/GA1 Andrzejbanas     Eiga-Kevin2                   2023-04-11 03:30:54 Film                           1
Talk:St. Francis Cathedral, Xi'an/GA1 TheLonelyPather Jacoblee628                   2023-04-13 00:59:15 Art and architecture           1
Talk:Pizza Tower/GA1                  TheJoebro64     204.132.81.138               2023-04-13 14:03:31 Video games                     1
Talk:Battle of Carthago Nova (209 BC)/GA1 Gog the Mild     Ifly6                         2023-04-14 23:52:51 Warfare                         1
Talk:New College, Oxford/GA1 Chiswick Chap   Robminchin                   2023-04-15 14:37:35 Education                       2
Talk:Agriculture in Turkey/GA1 Chidgk1         TheTrueSauce                 2023-04-17 18:29:31 Agriculture, food and drink     2
Talk:Logical reasoning/GA1       Phlsph7         Non-pegasus                   2023-04-18 23:54:05 Philosophy and religion         1
Talk:2021 Nabisco strike/GA1 JJonahJackalope HadesTTW                     2023-04-20 17:00:28 World history                   1
Talk:Rook's graph/GA1        David Eppstein   TheTrueSauce                 2023-04-21 13:33:28 Mathematics and mathematicians 2
  • I do agree that the "one reviewer" system is a bit hit and miss at the moment, and not just for newer reviewers, but for veterans too. I don't think the criteria are being consistently applied. Take Friedrichshafen FF.1 for example, a recent GA I just happened across. This is not to criticise the nominator or the reviewer in particular, but personally if I were doing a GA review on that article I'd reject it as not being broad in its coverage. Unless what's written there is really the sum total of what can be said about that plane (I haven't specifically checked the sources in detail) but superficially the article looks more like start class than B class to me... Cheers  — Amakuru (talk) 10:26, 25 April 2023 (UTC)
    • I absolutely agree that there's a huge variation in how thoroughly reviewers review articles, but without looking into the sources in depth I think it's entirely plausible that this is in fact all there is to say about a century-old aircraft of which only one prototype was constructed, and the reviewer does seem to have looked for additional sources. Unless I could demonstrate that the article was missing something, I would be reluctant to fail this on broadness of coverage. Caeciliusinhorto-public (talk) 11:27, 25 April 2023 (UTC)
      It's consistently been a bugbear of mine that we have no actual standard for what "broad in its coverage" means. I tried awhile back to get one, but no consensus arose and I got frustrated with people constantly misreading my argument, so I dropped it. ♠PMC♠ (talk) 00:40, 26 April 2023 (UTC)
      When I tried to address an obvious broadness issue in USA for Africa, I settled on the despite being discussed significantly in reliable sources line in Wikipedia:What the Good article criteria are not. That is probably the best standard we have at the moment even if it is not stated directly. Username6892 03:34, 26 April 2023 (UTC)
      I agree that criteria 3 could be better defined but I'm unsure if there is any realistic way to go about that. Some editors have ad-block, newspaper subscriptions, and university resource access. I've run into a few instances during a review where the nominator cannot access the same information I can, and vice-versa. I get that "Broad in its coverage" isn't exactly perfect, but I fail to see an alternative that doesn't gatekeep GA or come out equally vague. Wikipedia:Out of scope and Wikipedia:What the Good article criteria are not give a roundabout answer, including Not noticing that a major aspect is completely omitted from the article, despite being discussed significantly in reliable sources. We could realistically squabble the meanings of "significant" and "major aspect" till the Heat death of the universe. In an attempt to avoid a discussion every time a GA is closed, perhaps Idea B is a best fit. 🏵️Etrius ( Us) 20:53, 26 April 2023 (UTC)
      Yes, even FAs, with their stricter comprehesiveness requirement, can be short and undeveloped if there's nothing more to say about the article topic. (t · c) buidhe 01:19, 26 April 2023 (UTC)
Agreed, I don't know if I'd consider that article or that review to be as through as we'd expect for GA. What really grabs my attention though is that both the reviewer and the nominator are hiding the review from the talk page. Is that common practice? Thebiguglyalien (talk) 00:37, 26 April 2023 (UTC)
Yes, I've seen that a few times -- if article history is in place it provides a link, and the GA template lets people know there's a review there if they want to. If there's no a-h template I wouldn't remove it myself but I think it's harmless to do so. Mike Christie (talk - contribs - library) 02:09, 26 April 2023 (UTC)
@Amakuru: Having reviewed a few of Sturmvogel 66's Friedrichshafen articles, I can vouch for their apparent 'broad-ness'. Whether each prototype warrants its own article, or should be contained in a list-class article for the manufacturer is perhaps a different issue, but AfD and GA are different processes, and there is nothing in the GA criteria that requires an article to be notable! Harrias (he/him) • talk 21:06, 26 April 2023 (UTC)
@Harrias: thanks for that, that's my bad then, apologies.  — Amakuru (talk) 17:21, 27 April 2023 (UTC)

AI again

I can across this at Peer review. I know the issue of AI-generated reviews has been discussed above. Was any final position on them reached? As an aside, given the reviewer's remarkable speed in article creation, including at least one that has come here and passed, I'm guessing they are also using GPT-4 for that. Has Wikipedia got a stance on AI for article generation? It must have been discussed. I've let the reviewer know that I have brought the issue here. KJP1 (talk) 05:12, 20 April 2023 (UTC)

The discussion regarding AI-generated content is taking place at WT:Large language models. For what it's worth, I ran a few of their created articles at random through an AI detector (which are fairly reliable but not perfect), and it didn't find anything. Thebiguglyalien (talk) 05:38, 20 April 2023 (UTC)
Many thanks. I admit that the capacity/ability of AI to create articles is way beyond my limited brain, but five articles in nine minutes, and 82 in a single day! KJP1 (talk) 05:48, 20 April 2023 (UTC)
Our discussion was on AI reviews, AI articles are a different, beyond GAN issue. I did one spotcheck on the Verrucariaceae article linked: the sentence "Servít's classification was not widely adopted by future authors, and according to Cécile Gueidan and colleagues, it was the lack of clear morphological characteristics in the Verrucariaceae that hindered future proposals for changes in classification." is sourced to [1] "Servít (1953) revised the generic classification of the Verrucariaceae...but other authors rarely adopted this new system. The scarcity of additional morphological characters available for proposing a new generic system prevented the proposal of important changes in the classification. It is only recently, with the development of molecular techniques, that a new set of data became available for assessing the generic circumscription and the relevance of morphological characters in classifying species from this family." I get a feeling it's not quite right, and also not the most distant paraphrasing. However, I could see my questions being raised by text made by either a human or an AI; it's certainly not the blatantly obvious AI nonsense the previous reviews discussed. CMD (talk) 06:52, 20 April 2023 (UTC)
I checked one of the new contributions, Plectocarpon galapagoense, for which I have access to the main source reference. All of the claims in the new article appear to match those in the source, translated into less-technical English and rearranged into a more article-like overall structure. If it is AI-generated, it is a high-quality and source-based AI generation, not the other kind we have seen that makes up things out of thin air. —David Eppstein (talk) 07:33, 20 April 2023 (UTC)
@KJP1 It would be WP:MEATBOT which treats bot-like editing. But I believe there will be no consensus that policy will be applied except one actively states they are using AI for content creation. I have raised the policy to regulate the masscreation of articles at the Village Pump (policy), but there I couldn't find consensus for a regulation of masscreation.Paradise Chronicle (talk) 21:41, 20 April 2023 (UTC)

This is an example of a ChatGPT-assisted review, and probably more relevant for discussion in this forum. @Etriusus: kindly agreed (beforehand) to be a guinea pig for this experiment. The review goes into more rigour than expected for a GAN, but Etriusus and I have reviewed each other's GANs before, and I know they're ok with extra article-improving commentary. I wanted to see if the chatbot would be helpful in reviewing, and I think it is, if you ask it the right questions. Esculenta (talk) 12:03, 20 April 2023 (UTC)

It looks like there were a lot of bad suggestions, given Etriusus provided a lot of not dones in response to the AI comments. Contrast with your human comments, which were all done. For a less experienced author, and especially for a less experienced reviewer, this seems like it would be highly problematic. In particular, it seems its comments in relation to GACR3 were particularly poor, although it seems to do better with GACR1. These both make some sense, given how ChatGPT is supposedly a word predictor. I don't see any comments on GACR4, which leaves me to wonder if nothing came up due to the topic or if it didn't understand the instructions. CMD (talk) 15:39, 20 April 2023 (UTC)
I agree re: bad suggestions. I don't have any problems with using AI as a starting point, but some of the AI comments, upon further inspection, point out things that aren't actually issues. For example, the AI's comment that The article does not consistently refer to the Ionia Volcano by a single name throughout the text, using "Ionia Volcano," "Burnt Bluff," and "Volcano Hill" interchangeably. was marked as "not done" because the alternate terms were only used in the lead sentence. Though, given that Esculenta used GPT-4, I don't think it has the same word-prediction problems that GPT-3 or older language models have. – Epicgenius (talk) 18:33, 20 April 2023 (UTC)
I'll repeat what I said in the review. The criteria 1 suggestions were of decent quality while the criteria 3 suggestions weren't as good. The program was keen on asking for more information about a relatively obscure topic that, frankly, didn't have the sourcing necessary to fill the gaps. This isn't inherently a bad thing, I personally believe it could be quite useful at GCE or Peer Review, but GTP-4 is still operating a bit outside of GA criteria. This could be useful for more broad topics such as Vital articles. My best suggestion is to continue to tweak the starting parameters, GTP-4 shows promise but is still missing the nuance that a human eye can provide. I have no issue with the review overall, it was a nice change of pace and a bit of fun to try out. 🏵️Etrius ( Us) 19:19, 20 April 2023 (UTC)
If you want to use ChatGPT while reviewing, it would be much more useful to take the chatbot suggestions and then use common sense to see which of them are actually applicable to the article. The example given by Epicgenius is something that you could just have removed yourself. There was also no need to include the AI's comments about Ionia, Nebraska or to ask for a dedicated section about religious significance. Please don't cause extra work for article authors by making them say no to obviously stupid suggestions. If you do not want to filter the ChatGPT review, you should not use ChatGPT while reviewing. —Kusma (talk) 18:49, 20 April 2023 (UTC)
I believe we have no other choice than supporting AI on wikipedia, but finding the editors who use AI correctly. I believe Esculenta will be able to assist us in using it in a useful manner.~~~ Paradise Chronicle (talk) 22:02, 20 April 2023 (UTC)
That may very well be true. But their unwillingness to answer my question as to whether they had used AI for article creation, and their subsequent repeated blanking of their Talkpage, raises a question mark for me. KJP1 (talk) 22:10, 20 April 2023 (UTC)
This is also something I do not understand. I wanted Esculenta to be my hero, wikipedias example to follow, but so far they didn't apply for permission at the WP:BRFA. Maybe they also intended to and I didn't notice it, yet, who knows. But I believe AI should enter in semi-automated (bot-like) editing per WP:MEATBOT. Paradise Chronicle (talk) 22:29, 20 April 2023 (UTC)
This certainly isn't good. There's ongoing discussion about whether AI can be used, it seems deeply unhelpful for the use of AI to be hidden. I would like to suggest as a formal policy that we advise against using AI to review articles, and would not be opposed to stronger wording. I don't think we can do much about AI-generated articles here, however if Wikipedia:Large language models becomes a guideline we can adjust to keep in line with it. CMD (talk) 01:07, 21 April 2023 (UTC)
Yep, I appreciate there's a difference between AI for GAR, the focus here, and AI for creation, a much bigger issue. On the former, I'd also support strengthening the guidance/policy. At the very least, its use should be transparent, which, to be fair to Esculenta, they were. On the latter, I'll raise it at another forum. KJP1 (talk) 06:45, 21 April 2023 (UTC)
If someone used a chatbot to review an article I wrote and nominated for GA, I would be furious. I put in the effort, and so should the reviewer. I want feedback from a human being, not a blurry JPEG of the Internet. On general principle, I cannot take seriously any suggestions made by a reviewer who even uses this technology as a crutch. XOR'easter (talk) 14:21, 21 April 2023 (UTC)
I agree. As well-intentioned as it may be in terms of facilitating the process, it's entirely misguided. I am not against using ChatGPT for minor copy editing (similarly to Grammarly and other similar software) or for suggesting a synonym for word or assist with better language precision (similarly to an online thesaurus). However, using on it as one's main reviewing tool defeats the whole purpose of human interaction and critical intellectual exchange. Ppt91talk 18:34, 21 April 2023 (UTC)
I don't think there's anything wrong with feeding these programs text and then asking them to critique that text. The problem is that a human editor needs to then go through those recommendations and see if they hold water (and ideally, add their own comments, too). Sometimes, these programs are great at figuring out logical or grammatical issues in articles; they're also pretty good at recommending lede summaries and things like that. Other times, they just say bonkers stuff. AI is here to stay, so we shouldn't just brush it off, but we do need to be transparent and smart with how we use it.--Gen. Quon[Talk] 18:44, 27 April 2023 (UTC)

Why is the information at the top of WP:GAN duplicated

^Title. You have one set (the stuff at Wikipedia:Good article nominations/guidelines above), and then under the alternative list, another set. I'm fairly sure I tried to remove the bottom one, but ChristieBot reverted me for whatever reason. ~~ AirshipJungleman29 (talk) 01:25, 28 April 2023 (UTC)

ChristieBot puts in some of that text, and also transcludes the guidelines subpage, so this could be fixed by either removing it from the guidelines (which anyone can do) or removing it from the text ChristieBot is configured to write. I can do the latter if that's the text that should be deleted -- or I can make more of the text a transclusion, like the guidelines page, so anyone can edit it. Mike Christie (talk - contribs - library) 02:02, 28 April 2023 (UTC)
With the best will in the world, no-one really reads the stuff under the big banners, so I think what ChristieBot puts in should be removed. ~~ AirshipJungleman29 (talk) 02:08, 28 April 2023 (UTC)
Works for me. I'll remove the duplicated text, probably tomorrow. Mike Christie (talk - contribs - library) 02:39, 28 April 2023 (UTC)
Done; it should take effect in a few minutes. FYI, it looks like the reason it's duplicated is this edit, which was presumably intended to move that text above the banners, where as you say it's more likely to be read. Since ChristieBot works from boilerplate text, if we need to move text into or out of the guidelines subpage, we do need to update ChristieBot at the same time. Mike Christie (talk - contribs - library) 10:40, 28 April 2023 (UTC)
I don't actually remember doing that ! Dear me. ~~ AirshipJungleman29 (talk) 11:09, 28 April 2023 (UTC)

Co-nominators

There is currently no method in the GAN template to address co-noms. Would it be technically feasible to add a |nominator= field to the GAN template so co-noms can be credited to both nominators? QuicoleJR (talk) 14:19, 28 April 2023 (UTC)

For anyone answering this question, how should the ordering at Wikipedia:Good article nominations be affected by co-nominators? ~un6892 tc 19:33, 28 April 2023 (UTC)
I would say order it by the more experienced nominator. But any way works. QuicoleJR (talk) 19:37, 28 April 2023 (UTC)

Personal Life

Should we list more things in the personal life section, as much of the personal information seems to be hinted at throughout the article, but distinctly lacking in the section Louister41 (talk) 02:16, 1 May 2023 (UTC)

I don't think it is a standard practice to create a "personal life" section. There are some people who have a public persona and a private one, but the great majority of people do not have separate lives. Personal details give context, thus are often written chronologically within the article. If on the other hand, the person is, say an entertainer, it might make sense to separate out the personal section and group it all there. It's a judgment call on what works better for a particular subject. SusunW (talk) 04:36, 1 May 2023 (UTC)
Did you mean to post this on an article talk page, Louister41? ~~ AirshipJungleman29 (talk) 10:42, 1 May 2023 (UTC)
It’s currently exam season for me, but I would’ve done some research and maybe.
I guess we have different approaches when it comes to writing the personal section 198.168.103.75 (talk) 12:39, 1 May 2023 (UTC)

Question

Hoping for a quick response here: Is there any requirement to review a GAN? I know I am not the most knowledgeable in prose quality, is that OK? QuicoleJR (talk) 20:54, 28 April 2023 (UTC)

You can always ask for a second opinion if you are unsure about one aspect. (t · c) buidhe 21:14, 28 April 2023 (UTC)
Per the top of Wikipedia:Good article nominations, any registered user can review as long as they are knowledgeable and experienced enough. Given your lack of confidence in your knowledge of prose quality, I would recommend following the instructions for reviewers, particularly familiarizing yourself with the good article criteria given that is the standard by which you will be reviewing articles. ~un6892 tc 21:15, 28 April 2023 (UTC)
OK! QuicoleJR (talk) 15:01, 1 May 2023 (UTC)

Asking for a second opinion

Hello everyone! If possible, I just wanted to ask for a second opinion on this article I'm currently reviewing, please.

I noticed that @Thebiguglyalien: and @Mike Christie: previously took over nominations by the same user, but any kind of input by experts of Japanese cinema (someone like @Nihonjoe:, perhaps?), or cinema in general, would be highly appreciated as well!

Oltrepier (talk) 20:28, 1 May 2023 (UTC)

Is there still a one nom per person rule?

There are plenty of nominations in here, several nominators have more that one nom active. Is this still against the rules? Thanks, QuicoleJR (talk) 17:56, 2 May 2023 (UTC)

There has never been a one nom per person rule at GA. That's a FAC rule only. ♠PMC♠ (talk) 18:10, 2 May 2023 (UTC)
AFAIK, this has never been a rule at GA; only FAC. Caeciliusinhorto (talk) 18:12, 2 May 2023 (UTC)
Ok. QuicoleJR (talk) 18:13, 2 May 2023 (UTC)
It's not even strictly speaking true at FAC, where you can (technically) have two. We currently have a rule that if you have over 20, the remainder are hidden until your other ones are processed. Lee Vilenski (talk • contribs) 18:48, 2 May 2023 (UTC)
OK. QuicoleJR (talk) 18:52, 2 May 2023 (UTC)

Probably spurious IP edit review start

An article I nominated just had its review started by an IP editor whose edit summary and edit history don't suggest that a serious review is forthcoming; also, WP:GAN/I#R2 says that one has to be registered and signed in to review a nomination. Can an admin un-start that review? Thank you! -Bryan Rutherford (talk) 22:19, 2 May 2023 (UTC)

Standard practice is to G6 such review pages. This will return the nomination back into the pool. Trainsandotherthings (talk) 22:55, 2 May 2023 (UTC)
Alright, I've posted a G6 template. Thank you! -Bryan Rutherford (talk) 00:22, 3 May 2023 (UTC)

Help at caracal

In the process of making some gnome-edits, I found that, while only one GAN was listed at Talk:Caracal, it'd actually been nominated twice- one in 2015, and again in 2016. I can't figure out what the outcome of the 2015 nomination was, though, and the article history template is unhappy without out. Could someone more knowledgeable about GANs and the inner workings of this area please help me out? SilverTiger12 (talk) 00:47, 5 May 2023 (UTC)

The first was a pass, and I've added the info to the talk page template. But it looks like it was wiped with this edit a month later, after copyright violations were discovered. Unless it went through a formal delisting, I'm not sure how that would be represented in the template. Thebiguglyalien (talk) 00:54, 5 May 2023 (UTC)
It doesn't look like it was properly GAR'd, just had the status removed. Maybe we could make a reassessment page and copy those comments over and treat it as one. ♠PMC♠ (talk) 00:57, 5 May 2023 (UTC)
That was quite a while ago, so I'm not sure if it is worth it. I was just baffled by the rather scant 2015 review and doubted whether it had really passed. SilverTiger12 (talk) 00:58, 5 May 2023 (UTC)
Scant reviews are quite normal at GA, then and now. Thorough reviews are recommended, but plenty of reviewers just throw a few words in front of each criterion and say it passes. A few get reverted if the nominator makes a stink about it, but for the most part there's no mechanism to prevent it. Thebiguglyalien (talk) 01:09, 5 May 2023 (UTC)

Charles III

Noticed a very cursory review just passed at Charles III, don’t have time to have a look myself. ~~ AirshipJungleman29 (talk) 15:18, 2 May 2023 (UTC)

Hmm. The reviewer's comments would be very cursory if they were passing a 9000-word article as GA in this state, but they... didn't. They made their initial comments yesterday in this edit and haven't edited since; the article nominator appears to have then updated the talkpage as if the article had passed here, less than 24 hours later. That should probably be reverted, and @RonaldDuncan: be given the chance to complete the review. Caeciliusinhorto-public (talk) 15:35, 2 May 2023 (UTC)
That should be reverted immediately. It's a hard rule that you don't promote your own nominated article to GA. Trainsandotherthings (talk) 16:02, 2 May 2023 (UTC)
I have reverted the GA-related edits to the talk page. —Kusma (talk) 16:14, 2 May 2023 (UTC)
I see that almost all items are marked as passing the criteria. --Gerda Arendt (talk) 15:38, 2 May 2023 (UTC)
Including "7. Overall: Pass/fail: (+)", which makes the nominator's confusion as to whether it's been passed or not highly understandable. 109.etc (talk) 23:28, 2 May 2023 (UTC)
Should this perhaps be relisted as "requiring review", or would that be a slight abuse of notation due to this one potentially still being in progress? On the off-chance that gets more eyeballs on it for a possible second review. 109.etc (talk) 23:26, 2 May 2023 (UTC)

The BLP-in-question, is currently going through an RFC. When said-RFC is concluded, the BLP will be going through another RFC. Recommend putting GA review on hold. GoodDay (talk) 00:07, 3 May 2023 (UTC)

This kerfuffle and its associated discussion simply gives me one more reason to believe that the GA process is not working. WP:GAN/I gives virtually no practical advice in what a review should look like or what you have to do while reviewing. Then WP:RGA is the actual guideline that governs GA reviews, but it may as well not exist for all the good it does. It wouldn't be an issue if this was an isolated thing, but we're reversing a few of these every week, and those are just the ones that we catch. Thebiguglyalien (talk) 00:27, 3 May 2023 (UTC)
All proposals for strong advice on what reviews should look like come up against stiff opposition relating to the freedom to review in a style a reviewer seems fit. (In this sense it is a very typical current Wikipedia issue, emerging from the tension of the freewheeling AGF amateur model coming up against increasing expectations of quality. There is a fundamental mismatch between what the process is, and what the process is seen to be.) This is at some level tied to the base practice of GAN as an individual-level review (and good reviews among experienced editors being somewhat of a I know it when I see it affair), and changing that would be a significant overhaul. There might be other way to ameliorate it, such as restricting to a certain # edits, random audits, or a clearer process for review evaluation, but they haven't had much discussion. No process will ever catch or prevent every insufficient review, so it's a question of how to balance improvement with costs.
On the Charles III article, it was nominated on 19 April amid very busy talk-page discussion. (Spirited discussion continues at Talk:Charles III#GA review review.) Talkpage activity is matched by significant editing activity; the article was edited 428 times in April, and so far is on 51 edits for May. My personal view is that this is a clear-cut stability quickfail. CMD (talk) 01:45, 3 May 2023 (UTC)
That's partly because it's a somewhat topical topic, but also because there's a fair bit of bickering about content. Not a full-scale POV-war by any means, so much as a series of stylistic skirmishes, but I guess it's not why or how much, it's how many.
On the process, if I might be forgiven for chipping in my own hot take without a huge amount of familiarity, might there be some merit to a "see one, do one, teach one" system? When someone does their first GA review, make it standard practice not to give it final approval until a more experience reviewer has given it the once-over. Ideally one would involve the "mentor" at any earlier stage to give more interactive advice, but the logistics of that might be harder. 109.etc (talk) 02:06, 3 May 2023 (UTC)
Mentors and checking reviews are both common suggestions that have reasonable support whenever they're proposed, but there doesn't seem to be much interest in actually implementing them. Thebiguglyalien (talk) 02:16, 3 May 2023 (UTC)
Well, sadly limited in its originality as it is then, this suggestion is perhaps at the lower end of implementation cost, as it's in effect the intersection of those two. One checking exercise per new reviewer. 109.etc (talk) 21:16, 4 May 2023 (UTC)
I think relative stability is a better scale for extremely-high visibility articles such as this, CMD. On that note, I've picked up the review as a 2O, but I've indicated at least a week's pause for the coronation and aftermath. ~~ AirshipJungleman29 (talk) 22:25, 4 May 2023 (UTC)
Sure, and the article is relatively unstable. It usually gets under 100 edits a month. CMD (talk) 00:06, 5 May 2023 (UTC)
Be a bit of a shame if yet another venue for article-improvement eyes-on exercises were to rule itself out on box-checking grounds. I empathise with the difficulties of trying to review a moving target, but the disputes in the case are at least textually limited and stylistic/presentational, rather that wholesale flips between two POV versions. 109.etc (talk) 02:50, 5 May 2023 (UTC)
The wholesale flip between two POV versions is a strawman. Such flips are very rarely an issue for GAN, and would in any case be far more an issue for neutrality than for stability. The Charles III article has had over 500 edits made since the start of April. It seems a great stretch that these 500+ edits are textually limited. This is a discussion about a poor quality review; having a stable article is a necessity for the base quality of any review. CMD (talk) 03:29, 5 May 2023 (UTC)
I'm not in any way disputing the inadequacy -- or to be generous, lack of completion -- of the existing review. Perhaps the reviewer intended to return to it to flesh it out, but was unavoidably detained from doing so. I'm simply responding to your other comments. Of course not all of them are textually limited, but the most extensive ones, especially the trimming work done by WP:CIII seems to have been very largely uncontroversial. If for a GAN to be considered an article has to to be "stable" to the point of inertia for several months, better to make that clear in the process page. @AirshipJungleman29, your offer to do a 2O is greatly appreciated, and no worries whatsoever about waiting a week or however long seems appropriate. 109.etc (talk) 18:26, 5 May 2023 (UTC)
Jumping from "over 500 edits made since the start of April" to "inertia for several months" right after being accused of making a strawman is... bold. Thebiguglyalien (talk) 18:53, 5 May 2023 (UTC)
So the standard is what? On either matter? It'd perhaps be "bold" if the first characterisation were at all apt, the better to serve as a bludgeon to mute all further inquiry on any other point. 109.etc (talk) 19:27, 5 May 2023 (UTC)

Hello everyone! I was going to finish my GA review for Eiji Tsuburaya and promote it, when I noticed, from a quick look at the article's info, that one of the biographers quoted in the bibliography, August Ragone, had done a few edits by himself on the page between 2008 and 2016. The problem is, he likely forgot (?) to declare conflict of interest, having published a full biography about Tsuburaya in 2014.

Obviously, the nominator isn't to blame for this, and he's become by far the main contributor to the article anyway. Still, I wanted to be as transparent as possible and ask if I can still go ahead with the promotion.

[On a side note, speaking of transparency, I've inadvertedly become the second main contributor to Tsuburaya's page, but my own edits are mostly corrections to prose and links, so I hope they don't represent another problem.]

Oltrepier (talk) 09:34, 6 May 2023 (UTC)

Unless the content added by Ragone is somehow in violation of the GACR (ie it's POV, peacocky, a copyright violation, or otherwise unacceptable), it's not a barrier to promotion. At this point his authorship is only 0.3%, so I wouldn't worry about it. ♠PMC♠ (talk) 09:51, 6 May 2023 (UTC)
@Premeditated Chaos Right, thank you for clarifying. Just to be sure, I've quickly dug deeper, and I think there isn't anything too problematic to report: I suppose the criteria were significantly different back in 2010, anyway...
By the way, the biography I've mentioned above is actually a re-edition of Ragone's original 2007 book.
So, should I proceed? Oltrepier (talk) 10:21, 6 May 2023 (UTC)
Why would it matter what the criteria were like in 2010? Like I said, it's not a barrier to promotion considering the extent to which Rangone's edits have been rewritten. ♠PMC♠ (talk) 10:26, 6 May 2023 (UTC)
@Premeditated Chaos Sorry, I expressed myself badly...
Okay then, thank you for your help! Oltrepier (talk) 14:01, 6 May 2023 (UTC)

Backlog drive

Is it time to do another backlog drive? Therapyisgood (talk) 02:59, 7 May 2023 (UTC)

Wikipedia talk:Good article nominations/Archive 28#Is there a need for a backlog drive. Trainsandotherthings (talk) 03:03, 7 May 2023 (UTC)
It seems that the general consensus leans in favor of a drive, but no one is willing to organize it (which is understandable, especially for you after your experience with the last one). Thebiguglyalien (talk) 23:24, 7 May 2023 (UTC)

Would a reviewing guide be helpful?

I've mentioned this before, but I think it warrants its own discussion. Over at WP:DYK, there's a reviewing guide which I found really helpful when I started reviewing DYK noms, and I still use it as a checklist when reviewing. Would there be any benefit of creating a GA reviewing guide in this fashion? If such a guide was created, would we be able to get it in front of new reviewers? The current GA instructions are good for procedural stuff, but they don't help with the actual reviewing itself. Obviously a GA guide would be a bit longer than the DYK guide, but I think with a little organization it could be an asset for new reviewers. Thebiguglyalien (talk) 23:22, 7 May 2023 (UTC)

There is a review guide linked from the instructions page though it does seem that a ton of people don't notice it. ~UN6892 tc 23:39, 7 May 2023 (UTC)
This suffers from the same issues as the instructions page where it's bogged down in procedural detail without providing any helpful information on actually reviewing. It's not a checklist or step-by-step type guide. Thebiguglyalien (talk) 23:51, 7 May 2023 (UTC)
honestly that sounds like a very good idea. For how confusing GA reviewing is, there's a lack of a guide that could help people. Onegreatjoke (talk) 01:34, 8 May 2023 (UTC)

I've nominated two articles, but I think in the wrong category.

I've nominated two articles for GA review, but I think I put them in the wrong category: Charles Dewey Day and Michael O'Sullivan. I put them both into "Politics and Government", but I think a better place would be "World History - Politicians". Can the categories be changed in the two nominations? Thanks! Mr Serjeant Buzfuz (talk) 00:50, 8 May 2023 (UTC)

Mr Serjeant Buzfuz You can do that by editing the GAN template on the article talk page. Bot will update the GAN page accordingly. (t · c) buidhe 00:52, 8 May 2023 (UTC)
Done! Thanks. I thought that once I'd done the nomination, i couldn't change it. Mr Serjeant Buzfuz (talk) 02:41, 8 May 2023 (UTC)
Well, I evidently screwed up somehow, because now the two articles are listed as "Miscellaneous" at the very bottom of the page. Could someone take pity on my and put them into the "World history" category? I see articles about several other politicians are listed there. Thank you, signed "Befuddled" Mr Serjeant Buzfuz (talk) 23:14, 8 May 2023 (UTC)
See this section of the instructions. In fact there are other things you can put in that field, but they are just aliases for these codes. If you don't use one of these strings, the bot puts the article in "Miscellaneous", which is what's happened here. Mike Christie (talk - contribs - library) 23:59, 8 May 2023 (UTC)
Yes, I've read those instructions several times over the past two days and I obviously don't understand them, since I keep screwing up. I guess I'm just too stupid to even nominate an article, so I've deleted both nominations. Mr Serjeant Buzfuz (talk) 01:42, 9 May 2023 (UTC)
Please don't say that! I've fixed the templates. A simple mixup doesn't mean you're stupid or that your content work isn't good. On a skim, both articles certainly look well-done, so I'd hate to see them not be nominated just because of technical issues. (Feel free to revert if you really don't want to go ahead.) ♠PMC♠ (talk) 03:05, 9 May 2023 (UTC)

MOS:LAYOUT and {{sfn}}

Do the good article criteria via MOS:LAYOUT require moving all references into the bibliography and prohibit having bibliographic references in the reference list when {{sfn}} and {{harvnb}} are used? (The question emerges from Talk:Tiberius Gracchus/GA1.) The layout guidelines for references say This section, or series of sections, may contain any or all of the following: ... Full citations to sources, if short citations are used in the footnotes which implies for me the opposite. Ifly6 (talk) 14:54, 9 May 2023 (UTC)

@Ifly6 and UndercoverClassicist: This discussion is irrelevant for GA status (see the appropriate section of WP:GACN: "If you are able to figure out what the source is, that's a good enough citation for GA."). The most important part of the relevant MOS:LAYOUT section are the words any and all: anything is good, as long as it clearly identifies the sources. Consistency of reference format is not a GA issue. ~~ AirshipJungleman29 (talk) 15:05, 9 May 2023 (UTC)
The question-mark is whether we have a list of all references (sources of information), presented in accordance with the layout style guideline. given that that list is split between the References section (which has all the sources, but mostly only as author-date-page) and the Bibliography (which has most, but not all, of the sources). Per what you've said above, I'm satisfied enough that I can identify the source for each citation, so I've now passed that criterion. UndercoverClassicist (talk) 15:29, 9 May 2023 (UTC)
No specific reference format is required as long as it's reasonably consistent. I don't think that article's references are a GA issue. (t · c) buidhe 20:57, 9 May 2023 (UTC)

2nd opinion requested for 1971 Moroccan coup d'état attempt

Does anyone have time to take a look at 1971 Moroccan coup d'état attempt? I am reviewing it, and I have concerns about the prose. The nominator, NAADAAN, is not a native speaker and has worked with a friend to improve the prose, but I think it's still fragmented and PROSELINE in places. I've done some copyediting to try to improve it, but as I'm not familiar with the sources there's a limit to what I can do. I'd appreciate another opinion on whether the prose in its current state is a fail or pass for GA. Thanks. Mike Christie (talk - contribs - library) 15:09, 11 May 2023 (UTC)

In my opinion, most of the prose is fine, but the Background section could do with some improvements. Parts of it read quite stilted, as they are a series of short, simple sentences listed one after each other, almost like bullet points. Harrias (he/him) • talk 15:48, 11 May 2023 (UTC)
That said, the only GA criteria relating to prose quality says "is clear, concise, and understandable to an appropriately broad audience; spelling and grammar are correct", and I can't see that the article fails that. Harrias (he/him) • talk 15:51, 11 May 2023 (UTC)
KJP1 has agreed to do a copyedit, which I think will probably resolve the prose issues. Yes, that's why I haven't failed it -- I think it's not technically a fail on the criteria, but in practice GA reviewers tend to ask for better prose than this. Mike Christie (talk - contribs - library) 15:53, 11 May 2023 (UTC)

Timeline of the Eureka Rebellion

A new nominator has nominated this for GA, but I think it clearly falls under the scope of FL rather than GA (and isn't up to snuff in any case). What's the modus operandi in this instance, revert the nomination with a note on the talk page, or quick fail? Harrias (he/him) • talk 15:07, 12 May 2023 (UTC)

I'll take care of it, by providing some feedback and then failing it. Courcelles (talk) 15:14, 12 May 2023 (UTC)

Unimplemented proposals

Of the proposals approved in the GA proposal drive, these have not been fully implemented:

  • We nominally accepted "Make spot checking a requirement" and added it to the instructions, but we still allow reviews to do without them.
  • Wikipedia:Exceptional reviews was approved, but then died before it began.
  • Revitalization/reform of the mentorship program was approved, but no implementation was attempted.
  • To my knowledge, invitations to review have not been created or sent.
  • GA status has not been made more prominent in main space (such as GA icons in mobile view).
  • GA by month categories have not been created.
  • Wikipedia:Former good articles has not been created.

There have also been other proposals since the drive that gained some level of support on this talk page but have not been implemented:

  • Increase citation standards to match those of DYK (got to the RfC drafting stage before going stale).
  • Make an auto-updating report of potential drive-by nominations.
  • Some way to catch, address, or prevent checklist reviews.
  • Run a backlog drive, possibly with additional automation.
  • Add the nominator's name to the GA review page or the article's talk page.
  • Make an auto-updating report of open and/or recently closed reviews so that they can be checked.

I thought it might be helpful to list these in one place in case there are any we want to revisit. Thebiguglyalien (talk) 21:08, 13 May 2023 (UTC)

I looked into implementing the GA by month and former GA categories and didn't see a reliable way to do it, but if we can come up with an algorithm, I can probably get ChristieBot to create them and maintain them. I would still support adding each GA nominator's name somewhere, for traceability, though the {{GA}} and {{article history}} templates seem the natural place to me (which would also allow for conominators to be recorded). I can implement the reports to catch drive-by nominations and/or recent reviews if there's consensus to do so, but I don't think there was. Mike Christie (talk - contribs - library) 10:30, 14 May 2023 (UTC)

Sortable GAN list now available

I've finally got around to making User:ChristieBot/SortableGANoms available; it contains all active nominations in a single table and will be updated every 20 minutes, like the main GAN page. I am also planning to do a version of this with sortable tables inside each subtopic, so that reviewers who only want to see the nominations in a given subtopic can find the article they're interested in. These are currently in the bot's user space, but I could move them to somewhere in the GA namespace if there's support for that. Mike Christie (talk - contribs - library) 14:56, 12 May 2023 (UTC)

I like it! I wonder if there's a way to make it clearer what's what without cluttering it; as I'm scrolling down, I have to remind myself what number corresponds to what metric. And this might just be a personal preference, but I would also replace the abbreviations in the status column with full phrases (e.g. replacing "H" with "On hold"). The column is already wide enough that I don't think it would significantly shift the layout. Thebiguglyalien (talk) 15:14, 12 May 2023 (UTC)
I changed the status column; I agree there's room. Not sure what to do about the column headings -- there are Javascript tables that keep the headers in view as you scroll, but I don't think Mediawiki has anything like that. Mike Christie (talk - contribs - library) 15:55, 12 May 2023 (UTC)
Good. I think the subtopics-with-tables version will be the most useful. If they aren't in the GA namespace then maybe we could at least have a link to them from there. Chiswick Chap (talk) 11:35, 13 May 2023 (UTC)
Mike Christie, I don't know how practical this would be, but would it be worth exploring something similar to sort open reviews (and maybe recently closed ones) based on the reviewer's stats rather than the nominator's? It might be helpful as a quick way to compare reviews based on how long they've been open and how experienced the reviewer is. Hopefully something like that could encourage supporting new reviewers as a more common practice instead of leaving them to figure it out themselves. Thebiguglyalien (talk) 05:26, 14 May 2023 (UTC)
Definitely doable if there's consensus, but I think one or two people expressed concerns about biting new reviewers if we start reviewing their reviews. Mike Christie (talk - contribs - library) 10:32, 14 May 2023 (UTC)
It seems like there was consensus in favor of increased assistance for new reviewers at Wikipedia talk:Good article nominations/Archive 28#More scrutiny of reviews. And I'll reiterate what I said there: I think it would be helpful if there were a list of all recently closed reviews to make sure they're sufficient. Ideally such a list would also flag reviews that should be double-checked (e.g. new reviewer, short review, quickpass). Right now the only place to scrutinize recently closed reviews is at the page history of WP:GAN. I've checked there a few times to look at samples of what's getting through, and quickpasses are regularly getting through. I also think it's helpful to give new reviewers some guidance when they're reviewing, like I did here. It's already possible to check WP:GAN for open reviews by new reviewers, but it's inconvenient. Thebiguglyalien (talk) 00:02, 16 May 2023 (UTC)
OK, I'll put something together, probably this coming weekend, unless objections emerge. Mike Christie (talk - contribs - library) 10:07, 16 May 2023 (UTC)

Script to highlight reading difficulty

I wanted to let you know that I wrote a script to highlight sentences by their reading difficulty with different colors. The script is found at User:Phlsph7/Readability.js and the documentation is at User:Phlsph7/Readability. It goes through articles sentence by sentence. Difficult sentences are colored red and easy sentences are colored green. The script also shows the readability score of the article as a whole at the top. It includes a list of sentences ordered by lowest readability to help identify where the most attention may be needed.

The script measures readability using the Flesch reading ease score. It is a very simple measure that only considers two factors: words per sentence and syllables per word. According to it, texts with long sentences and long words have low readability. This measure is very superficial and often does not reflect the actual difficulty of the text. For this reason, the script should only be used as a rough guide for potential improvements. It cannot replace human judgment. For example, some articles get low scores because their topic requires a lot of long technical terms. In this case, the point is usually not to replace long and precise technical terms with short and vague non-technical terms. It is often better to ensure that the technical terms are properly defined, even if they reduce the readability score.

Regarding the GA process, the script is probably most useful to nominators. It can help them identify potential problems in relation to criterion 1a, i.e. that the article is well-written. I'm more skeptical about reviewers using it because of all the limitations explained above. I hope it is obvious that this is in no way a quick shortcut for judging whether articles pass or fail criterion 1a.

I hope to get some feedback on potential problems, how the script may be improved, and how to discourage misuse. Phlsph7 (talk) 09:14, 13 May 2023 (UTC)

See also the discussion at WT:FAC. Phlsph7 (talk) 17:34, 15 May 2023 (UTC)
A two-clause sentence being deemed moderately difficult makes me think that it's not even got much use as a rough guide; I can't imagine any native reader finding those hard. Kids included... I wonder if the test being developed in the '70s in the US (when, perhaps, people not in a professional environment may not read much or have learned to read - they would have grown up during a war) means it assumes too little proficiency as an average. In short, I fear for civilisation if you need a post-graduate in order to read that extract. Of course, equating the "well-written" criterion even roughly to "a child could read it" isn't really what I think we're going for. Kingsif (talk) 20:50, 15 May 2023 (UTC)
Thanks for the feedback! The Flesch reading ease score is very superficial and can easily lead to false conclusions. The main purpose of the script is not to tell editors which sentences are bad. But it can help them by pointing their attention at potentially problematic cases. For example, if a section is all red from start to end then the chances of finding one or two overly long and convoluted sentences are high. But many red sentences are perfectly fine and some green sentences are incomprehensible. The script has its uses but it is in no way meant to replace the judgment of editors. Phlsph7 (talk) 08:04, 16 May 2023 (UTC)
If many red sentences are fine and many green sentences are incomprehensible, what is the value of the tool compared to simply reading the text? ♠PMC♠ (talk) 08:27, 16 May 2023 (UTC)
The probability that a red sentence is overly long and convoluted is significantly higher than the probability that a green sentence is overly long and convoluted. Compared to simply reading the text, the script can make it easier and faster to find them. Phlsph7 (talk) 08:36, 16 May 2023 (UTC)
Playing with the script briefly, it seems to me that Flesch reading-ease is an even less useful metric than I had previously thought. Three example sentences from Corinna (an article where I am the primary contributor, and consequently largely to blame for any infelicities in the writing):
  1. Corinna's poetry is almost entirely concerned with myth.
  2. Her "Orestes" is possibly an exception to her focus on Boeotian legends.
  3. About forty fragments of Corinna's poetry survive, more than any ancient woman poet except for Sappho, though no complete poems of hers are known.
The first two have virtually identical sentence structures, and both are fairly short. The first gets a score of 18.94 and is highlighted in red, the second a score of 46.61 and is highlighted in yellow. Neither are, to my eye, "college" or "college graduate" level sentences. The third scores best (52.05) on the Flesch test, despite being twice as long and having a subordinate clause to keep track of. Even if Flesch reading-ease is a useful metric for a longish passage, it doesn't look to me as though the sentence-by-sentence breakdown is super helpful. The erraticness of the ratings doesn't suggest to me that it would, in fact, be faster to find overly convoluted sentences using Flesch scores than it would be by just reading the article carefully, at least for anyone who is a good enough writer that I would trust them to interpret these scores properly! Caeciliusinhorto-public (talk) 11:45, 16 May 2023 (UTC)
Thanks for this analysis. You are right, this kind of sentence-by-sentence comparison by exact score is often not very helpful. I'm sure there are many more examples in the article Corinna for which the score is not helpful. Here are some ways how the script can be useful:
The list of sentences ordered by lowest score for Corinna has the following 4 sentences on top:
  1. According to this theory, when she was rediscovered and popularised in the Hellenistic period her poetry would have been re-spelled into contemporary Boeotian orthography, as her original fifth-century orthography was too unfamiliar to a third-century audience.
  2. Her poetry often reworks mythological tradition – according to Derek Collins, "the most distinctive feature of Corinna's poetry is her mythological innovation" – frequently including details which are otherwise unknown.
  3. The three most substantial fragments are preserved on pieces of papyrus discovered in Hermopolis and Oxyrhynchus in Egypt, dating to the second century AD; many of the shorter fragments survive in citations by grammarians interested in Corinna's Boeotian dialect.
  4. If Corinna was a contemporary of Pindar, this use of the local vernacular as a literary language is archaic – though the earlier poets Alcman and Stesichorus wrote in literary dialects based on their own vernaculars, the fifth-century choral poets Pindar and Bacchylides both wrote in Doric despite it not being their local dialect.
I think each one of them could benefit from some copyediting.
Looking at the colors, the following paragraphs could be checked:
  • section "Life": 3rd paragraph
  • section "Poetry": 2nd paragraph
  • section "Reception": 2nd paragraph
It doesn't mean that there are any problems with them. But if you have limited time and don't want to check the whole article, you should probably check them first. Phlsph7 (talk) 12:08, 16 May 2023 (UTC)
I'm not going to say that none of these sentences or sections could be improved, because I'm sure they could. But really what this tells me is that the parts of an article which are trying to explain complex ideas, use long proper nouns and precise technical terms do badly on the Flesch reading ease score. The easiest way to make any of these sentences or sections score better on the Flesch reading ease score is to either get rid of details or be imprecise. There comes a point where trying to explain technical concepts in simple terms is more confusing than using the correct technical terms that people are familiar with and know the meaning of.
To be fair, I think your documentation does do a relatively good job of pointing out some of these limitations, and I would hope that the kinds of people who read this talkpage and install user scripts to help their editing have the common sense required to not make text worse in the name of improving some sort of arbitrary metric. I just don't know how helpful I will find the script. (I've also found that calculating readability scores for particular sentences by hand I frequently get signficantly different results than the script does, which I suppose speaks to computers' difficulties in counting syllables in English!) Caeciliusinhorto-public (talk) 14:35, 16 May 2023 (UTC)
You are right, there are some topics that will generally do bad at the Flesch reading ease score. And trying to get rid of long technical terms for the sole sake of increasing the reading score is a good example of how the script could be misused. I hope this won't be too common since most script users are experienced editors who are aware of such issues.
There are various other websites that calculate the Flesch reading ease score. Their results often differ from each other and from my script. The syllable count is one important factor for this. It's relatively easy to get a rough syllable estimate, which is enough for many purposes. But it's very difficult to get a precise count since there are many different cases to be considered in the English language. One example is the word "business": the script thinks the "i" is pronounced and counts 3 syllables. If you want to test the script, you can put a single sentence in your sandbox. You can compare the syllable count and word count from the overview table with your manual count. Phlsph7 (talk) 15:18, 16 May 2023 (UTC)

Talk:Lady Dimitrescu/GA1

From Mike Christie's log above, I came across this. Seems very cursory to me, with no actual 'review' present. It does reference a user talk page for the issues raised, but even there the content seems minimal, and this seems like poor practice generally. Any thoughts? (Courtesy pings for the nominator and reviewer: Kung Fu Man / GlatorNator.) Harrias (he/him) • talk 12:31, 17 May 2023 (UTC)

I gave feedback on his talk page prior to the review, granted it was only the second GAN I'd done in all these years, but he'd fixed the issues accordingly and it just needed a light copyedit. If this were FAC I could understand harsher scrutiny towards an article's content but if everything lines up it should be fine for GA afterwards, shouldn't it?--Kung Fu Man (talk) 12:48, 17 May 2023 (UTC)
Did you spotcheck any sources to make sure they supported the content without copyvio? If so it would be great if you could note those on the GAN page. CMD (talk) 14:12, 17 May 2023 (UTC)
[2] Spotcheck shows no copyright issue. Thou, it seems like this article [3] copied Wikipedia. GlatorNator (ᴛ) 15:14, 17 May 2023 (UTC)
That is not a spotcheck, a spotcheck requires actually checking a source. CMD (talk) 15:44, 17 May 2023 (UTC)
I see, he responded below to Caeciliusinhorto anyway. GlatorNator (ᴛ) 20:03, 17 May 2023 (UTC)
I think there's probably worse than this review. I believe KFM already know mostly about the writing style and sourcing of video game articles. All of the sources in the article are reliable per WP:VG/Sources, In addition, half or most of the content was also written by an experienced editor, Haleth. GlatorNator (ᴛ) 13:16, 17 May 2023 (UTC)
It looks like some of the review was conducted at User talk:GlatorNator#Thoughts on Lady D:. In general it's probably bad practice to do that – keeping the review on the review page makes it easier for future editors to check – but those comments do at least suggest the reviewer at minimum checked criteria 1, 2, and 3 to at least some degree. I agree that the review could be better (I'd like to see some evidence of e.g. spotchecking source-text integrity, checking source reliability, and checking for copyright issues) but I've seen more cursory reviews! Caeciliusinhorto-public (talk) 14:10, 17 May 2023 (UTC)
Yeah I do apologize, I've had reviews of my GANs that were either done in a similar or just no commentary so hadn't realized things had gotten a lot more intensive since my return. That said did check the sourcing to make sure they were from reliable outlets and used appropriately, said what they were supposed to, and the text was not directly copied from a source. Special attention was given to the development and reception sections primarily (moreso the latter) while doing so.--Kung Fu Man (talk) 19:54, 17 May 2023 (UTC)
@Kung Fu Man: Could you please include such detail, and perhaps the [X] number at time of review, on the GAN pages rather than simply saying a spotcheck was performed? It ensures that those looking at the GAN in the future know what was checked. CMD (talk) 01:50, 18 May 2023 (UTC)
@Chipmunkdavis: How's that looking, does it fit the style you're looking for?--Kung Fu Man (talk) 02:10, 18 May 2023 (UTC)
Looks great to me. CMD (talk) 02:49, 18 May 2023 (UTC)

Incomplete GA nomination

It has been a month since one of my GA nominations was picked up, and yet the reviewer hasn't proceeded with the review or made any edits since then. What should be done here? Bneu2013 (talk) 19:46, 17 May 2023 (UTC)

What is the article? - Cukie Gherkin (talk) 20:10, 17 May 2023 (UTC)
Presumably Talk:Tennessee State Route 397/GA1. Mike Christie (talk - contribs - library) 20:11, 17 May 2023 (UTC)
Unfortunately this is a pattern with this particular reviewer... For future reference, if someone keeps you waiting more than a week or two raise the issue here so someone else can take over. Trainsandotherthings (talk) 20:31, 17 May 2023 (UTC)
Thanks. I'd like for someone else to take over. I'd like to get this review through with quickly. Bneu2013 (talk) 00:04, 18 May 2023 (UTC)
I can review this weekend if nobody else has gotten around to it by then. Trainsandotherthings (talk) 19:29, 18 May 2023 (UTC)
Yes, that is it. Bneu2013 (talk) 00:04, 18 May 2023 (UTC)

Invitations to review - message brainstorm

Per GA proposal drive, consensus was achieved to send a message to editors who have 2-5 successful GANs, inviting them to review articles. I suggest that a general message be drafted that can be posted to an editor's talk page, and be sent to those who either have a reviewing-to-GAN ratio lower than 1:1, or have 0 reviews. Here's a draft below that we can use as a starting point:

Thank you for nominating articles for good article status. GAN is experiencing a backlog and needs editors like you to help clear it. Without reviewers, articles remain nominated for months without feedback, discouraging others from joining this process and possibly causing them to leave Wikipedia in frustration or disappointment. We noticed that you have nominated more articles than you have reviewed, contributing to this backlog; as an experienced GA writer, we hope you will review some articles and help editors create excellent content. Instructions on how to review are found here. If you have any questions or concerns, please go to the GAN talk page. Thanks and happy reviewing!

It would also be great if someone (GAN co-ords?) follow up a month after this message is sent to those who chose not to review and attempt to get feedback on why they are not reviewing. Thoughts? Z1720 (talk) 14:56, 15 May 2023 (UTC)

We should be careful not to use any language that feels like it's guilting people into reviewing. That's more likely to create resentment than it is to create frequent reviewers. I think it should just be a friendly "you might be interested" type invitation, similar to those used by WikiProjects, and then maybe a sentence toward the end about the backlog problem and how much their help would be appreciated. Thebiguglyalien (talk) 15:28, 15 May 2023 (UTC)
Softer version below, with the section about nominating articles removed. Editors can decide which approach GAN wants to take:
Thank you for nominating articles for good article status. GAN is experiencing a backlog and needs editors like you to help clear it. Without reviewers, articles remain nominated for months without feedback, discouraging others from joining this process and possibly causing them to leave Wikipedia in frustration or disappointment. As an experienced GA writer, we hope you will review some articles and help editors create excellent content. Instructions on how to review are found here. If you have any questions or concerns, please go to the GAN talk page. Thanks and happy reviewing! Z1720 (talk) 17:08, 15 May 2023 (UTC)
Edit this post. I removed the phrase discouraging others from joining this process and possibly causing them to leave Wikipedia in frustration or disappointment and made a rough draft template (User:Rjjiii/Invitation). Everything below is via template:
Thank you for nominating 11 articles for good article status. GAN is experiencing a backlog and needs editors like you to help clear it. Without reviewers, articles remain nominated for months without feedback. As an experienced GA writer, we hope you will review some articles and help editors create excellent content. Instructions on how to review are found here. If you have any questions or concerns, please go to the GAN talk page. Thanks and happy reviewing! Rjjiii (talk) 03:23, 16 May 2023 (UTC)
Certainly getting there. I wonder if we shouldn't simply say something like that on the GA1 (GAx) page whenever a GAN passes? I suspect we're getting close to the point where we should simply begin a QPQ, articles are languishing in the queue for 6 months at a time at the moment. Chiswick Chap (talk) 13:52, 17 May 2023 (UTC)
@Chiswick Chap: I think that the review ask should be placed on the nominator's talk page rather than on the GAX because I think it's more likely to be noticed there. I would also like to send the message at the nomination, as a message is already posted on the talk page when a review is complete. However, if the ask to review articles is posted on the GAX, I will not be bothered. Z1720 (talk) 00:26, 19 May 2023 (UTC)
Yeah, maybe we should go with a talk page message if someone nominates when they already have two successful GAs and zero reviews. Then maybe one more at a higher number like ~10 GAs with zero reviews. Thebiguglyalien (talk) 01:05, 19 May 2023 (UTC)

Currently WP:GACR says (footnote 3): "Dead links are considered verifiable only if the link is not a bare url. Using consistent formatting or including every element of the bibliographic material is not required, although, in practice, enough information must be supplied so that the reviewer is able to identify the source." I take this to mean that as a reviewer I can't complain about a dead link in the sources just because it's dead, unless it's just a URL. I can complain about a dead link if I happen to spotcheck it and can't because the link is dead, but not just because it's dead, so long as it's formatted as a web citation.

I propose changing the footnote to read "Dead links in the sourcing are acceptable only if they are not the only source of the data. If the citation is to an offline source such as a book, or the information can be verified via an archive of the URL, there is no requirement to change a dead link. Using consistent formatting or including every element of the bibliographic material is not required, although, in practice, enough information must be supplied so that the reviewer is able to identify the source." Mike Christie (talk - contribs - library) 10:32, 13 May 2023 (UTC)

Seems reasonable to me. Ultimately, we need enough information for verifiability. Harrias (he/him) • talk 11:30, 13 May 2023 (UTC)
This would move us a bit further ahead of what I read as the general consensus around WP:V. The GAN process shouldn't be a leader in this regard. CMD (talk) 13:06, 13 May 2023 (UTC)
Sounds reasonable. A dead online source with no means to verify what it used to say just isn't a source. —Kusma (talk) 15:40, 13 May 2023 (UTC)
I disagree with this mentality - how is an online source that has no current backup any less readable than a book that is unreadable. This would need to be a sitewide debate on whether deadlinks are suitable to be used, rather than a discussion bound to GAN. Lee Vilenski (talk • contribs) 16:38, 13 May 2023 (UTC)
A book that isn't in any library worldwide also isn't a suitable source (say, a Lost literary work like Aristotle's book on comedy). A book that is provably accessible in some library in Romania is fine. —Kusma (talk) 16:55, 13 May 2023 (UTC)
Lee, CMD, what about if it's not just the URL but the whole site that's inaccessible, meaning it can't be assessed for reliability? That's what prompted this suggestion. I was doing a review and didn't recognize a couple of websites, and found that for one of them there were no traces of the site left at all, either live or in the archive, so I had no way to tell if it was a fansite, or a blog, or whatever. I can see that if the page is down but the site is up one can still determine if the site meets reliability standards, but otherwise I think it's reasonable to require evidence the site is reliable. Mike Christie (talk - contribs - library) 10:36, 14 May 2023 (UTC)
I think it depends on the likely context. There aren't many websites around that have zero pages archived by wayback machine.
Here's an example of what's happened to me. The Euro Tour decided to change their website a few years back, and they completely took down the old one, with the idea to merge the info from old events - which never happened. Trying to cite a tournament draw, which was cited to this page is now impossible, as the particular URL is inaccessible and not archived. Should that information now be deemed not suitable at GAN? I get that contentious information might need viewable information, but things on the internet can be changed (that's why access-dates are so important). If there are no archived URLs at all from a website, unless we know anything about the organisation - it's difficult to say it's reliable. I doubt any reliable source would have no way to verify this though. Lee Vilenski (talk • contribs) 11:22, 14 May 2023 (UTC)

This is a matter for WP:RSN, not for GAN. If a source is reliable, according to consensus at RSN, then it is reliable. If not, then not. We already have rules requiring all sources to be reliable. We should not be attempting to make our own local rules about what is reliable or not. —David Eppstein (talk) 16:35, 14 May 2023 (UTC)

I don't really care how reliable a no longer extant source used to be. Anything based on such a source is by definition unverifiable. Does RSN deal with WP:V violations? —Kusma (talk) 16:41, 14 May 2023 (UTC)
Currently, nothing at WP:V suggests that deadlinks are considered to not be verifiable (as weird as that sounds). It currently says in the lede about "previously published", which doesn't necessarily mean still accessible. I suspect this should be discussed at WT:V as this isn't a GAN only thing. Lee Vilenski (talk • contribs) 17:39, 14 May 2023 (UTC)
WP:DEADREF item 6 says material cited to dead links is considered unverified after 24 months. —Kusma (talk) 18:24, 14 May 2023 (UTC)
Oh, I'd never seen that. Perhaps that's all we need to retain if it's been discussed previously. I can see 24 months being a suitable time period to know if something is retrievable. Lee Vilenski (talk • contribs) 18:34, 14 May 2023 (UTC)
I agree with David that we can't change RSN by a conversation here. What I meant to suggest was only a change in the GA criteria. The FA criteria, for example, require no deadlinks; all claims have to be verifiable at the time of the review, which in practice means no deadlinks are allowed at FAC. I have no problem assuming good faith on the original addition of a source that's now dead, but if I'm a reviewer, I can't review a deadlink. If it's valid to request a dead link be addressed if I try to spotcheck it, doesn't it make sense to ask for any other dead link to be replaced too? Mike Christie (talk - contribs - library) 21:05, 14 May 2023 (UTC)
So you want to be able to request replacement of deadlinks, WITHOUT trying to spotcheck them? What would be gained by adding such a rule, in exchange for the unnecessary WP:CREEP? How long does it take to spotcheck a deadlink and verify that it is a deadlink? If it's already tagged as a deadlink and you don't want to spotcheck it, doesn't it already fall under the criterion of not having valid cleanup tags? —David Eppstein (talk) 22:09, 14 May 2023 (UTC)
What would be gained, I thought, was that I as a reviewer would have more faith in the quality of the article. I don't think this is important enough to push it if it seems to be a creeping increase in the GA standards; I do want to keep GA reviews lightweight. It just bothers me, as a reviewer, that I am signing off on the quality of the article, when there's something about it I can't look at even if I want to. Mike Christie (talk - contribs - library) 22:17, 14 May 2023 (UTC)
But why do you think you have to sign off, under the existing rules? If you can't spot-check the reference you can't verify criterion 2(c) and 2(d). —David Eppstein (talk) 22:45, 14 May 2023 (UTC)
Sorry, I don't follow -- are you saying that if I happen to notice a deadlink I can fail a GAN under 2(c) and (d)? Or just that I can do that if I happen to pick that citation to spotcheck? Mike Christie (talk - contribs - library) 23:06, 14 May 2023 (UTC)
I do not understand the distinction you are trying to make between "notice" and "spotcheck". What do you do differently to notice that something is a deadlink, versus attempting to spotcheck that reference and determining it to be a deadlink? —David Eppstein (talk) 20:23, 15 May 2023 (UTC)
For reliability of sources, I click through to every site I don't recognize and try to determine if it's reliable. At that point I will notice if it's dead and won't be able to evaluate it. For spotchecks, I pick a few footnotes at random and check source-text integrity. If I check something for reliability I don't necessarily pick it for a spotcheck. Mike Christie (talk - contribs - library) 21:21, 15 May 2023 (UTC)

This discussion confuses me. Many Good Article criteria require things beyond what our polices and guidelines sets as minimums so why couldn't deadlinks be such a place? And why would the Reliable Sources Noticeboard have to weigh in on this? That is a noticeboard for This page is for posting questions regarding whether particular sources are reliable in context. and we're talking about a generalized context for all kinds of sources. In terms of the merits of this proposal I have more ambigous feelings, but the discussion about whether it can be done and whether RSN would have to weigh in have just left me so puzzled I'm not even going to give my thoughts on the merits. Barkeep49 (talk) 20:57, 15 May 2023 (UTC)

I'm with Barkeep: obviously we can choose to make the GA criteria stricter than the absolute minimum required by policy. Whether or not we should in this case is a question that reasonable people might differ on, but the GA criteria are absolutely something for people involved in the GA process to determine. If the consensus is that criterion 2 should simply be "complies with the letter of WP:V" that's one thing, but it seems silly to say "WP:V allows deadlinks, therefore we shouldn't even discuss whether the GA criteria should forbid them". Caeciliusinhorto-public (talk) 11:12, 16 May 2023 (UTC)
I don't think anyone has suggested quashing discussion here. That said, we have had previous discussions on spotchecks which were not even able to come up with consensus on a minimum number. Forbidding deadlinks would presumably require all web links be checked. Less onerous individually than a spotcheck I suppose, but still a substantial task. (Unless it's automated I suppose.) CMD (talk) 14:56, 16 May 2023 (UTC)
Yes, I hadn't thought about the fact that for some types of article (e.g. music) one can have hundreds of links. What I'd like to be able to do, as a reviewer, is ask for a deadlink to be fixed at my discretion. I already do this for spotchecks, but GACR doesn't authorize me to do it just so I can check the reliability of the source. I'm OK with leaving deadlinks in the article, per V and RS, if the source is something I know is reliable. E.g. for music, a cite to Consequence of Sound that has a deadlink is fine because I already know that site is reliable. This is an example of a link I'd like to complain about in a review I'm currently doing. I've actually noted in the review that the link is dead. I don't think it should be the reviewer's responsibility to ferret around in archive.org and try to find an archived page that matches what was cited, in order to determine if the source is reliable. Mike Christie (talk - contribs - library) 15:53, 16 May 2023 (UTC)
If you're unsure I don't think there's any harm in asking, perhaps the nominator might have more information. Whether such specific cases can be generally written into the GACR is a trickier question. Individual cases will depend on what the link is, what it is being used for, and so on. My view on your specific link is that it appears to be a primary source, likely non-independent, and unlikely to have had much detail. Depending on what it's citing, I feel you could query on those grounds. CMD (talk) 16:02, 16 May 2023 (UTC)
  • Support though I think a reasonable interpretation of the second GA criterion is that a reviewer can verify the work (i.e. at the time of review; not at some theoretical past time). If a source is truly lost to time, the article is not verifiable. If that interpretation of the second criterion doesn't carry the day, I support adding a note to the GA criteria to that effect. Ajpolino (talk) 21:27, 18 May 2023 (UTC)
  • I do have a bit of concern here - if this passes, I'm worried that we'll see a swamp of GARs based on older GAs with deadlink issues. While it is a verifiability concern, I also don't think we should be sending large quantities of GAs to GAR at the same time based solely on deadlink issues. Patience will be needed to bring everything up to code if we choose this route. Hog Farm Talk 02:56, 19 May 2023 (UTC)
    After the comments above, I think I would be OK with leaving the criteria as they are. What I now understand I can do in cases where there is a deadlink is ask the nominator to demonstrate that the link is reliable. If they can do so without resolving the dead link, that's fine; if not; they'll have to update the link. That was mostly what I was concerned about. Mike Christie (talk - contribs - library) 03:14, 19 May 2023 (UTC)

Problematic reviewer

Last week, User:ErnestKrause took on the review for Talk:Federalist No. 2/GA1. Their first action was to stop the review because a few words weren't capitalized. After I fixed this, they didn't review the article, instead using the review to suggest that several articles about different works in a series should be merged because they're often talked about at the same time. After I challenged this, they began reviewing a different article that I've never edited and seemed to expect that I improve that article before this one be promoted. I declined. In response to that, they not only quickfailed the article I nominated, but two of my other nominations (Talk:Federalist No. 3/GA1 and Talk:Federalist No. 4/GA1). Apparently, they took issue with my promoting them from Start class to B class because I was "involved", coming up with some non-existent policy that someone can't set an article they've written as B class. They said that these articles still need work and quickfailed them, insisting that they can not go past start class unless they are merged. Thebiguglyalien (talk) 00:02, 17 May 2023 (UTC)

Wow. This is total nonsense and ErnestKrause is extremely out of line. News flash - nobody gives a damn if you assess your own article as B class. It's stupid pointless bureaucracy to assert otherwise and everyone knows this. And if they're being promoted to GA then it doesn't even freaking MATTER what they were rated before!!! The copy-pasted fail messages that have nothing to do with the articles they're purportedly GA reviews of is insanity. Calling articles of 1566 words and 1322 words "still being start/stub articles" is total nonsense and makes me question if ErnestKrause should be reviewing anything at this point. All of these reviews need vacating and I would probably support sanctions against ErnestKrause for this showing of extremely poor judgement and behavior. This is one of the dumbest hills I've ever seen anyone choose to die on in my ~2 years on this website. Trainsandotherthings (talk) 01:01, 17 May 2023 (UTC)
I have to agree that this is poor form on ErnestKrause's part, and I would suggest that he self-revert. ♠PMC♠ (talk) 01:06, 17 May 2023 (UTC)
I have contacted ErnestKrause after looking at the reviews and articles in question. As a general point, GAN does not assess notability. It might consider article structure, but this is difficult to do if extending across multiple articles, so a wider venue is needed for that. While the reviews have their flaws, very little discussion has happened so far. It's a bit much to jump straight to sanctions. CMD (talk) 01:31, 17 May 2023 (UTC)
This was egregious and I would 100% support sanctions unless ErnestKrause self-reverts. Trainsandotherthings (talk) 01:35, 17 May 2023 (UTC)
Stub articles? Poor lead sections? Improper nominations? "Procedural quickfails" based on utter nonsense? ErnestKrause needs a {{whale}} to the talk page at the very least. ~~ AirshipJungleman29 (talk) 01:58, 17 May 2023 (UTC)
My edit was made in agreement with two other editors, Z1720 and Cecropia. User:Z1720 has listed a GAR for problems with the Federalist Papers article. I agree with Z1720 about the problems with the main page for the Federalist Papers and its sibling articles. When I asked BigAlien to look at this main article for purposes of his sibling article nomination he refused to look at it, apparently stating that the article was not the one he nominated and that therefor he would refuse to look at it. User:Z1720 has plainly stated his reasons for seeking a GAR. My agreement is with Z1720 for listing it and bringing to everyone's attention the problems causing the GAR for the main article and its relation to its close sibling articles.
After BigAlien refused to look at the main article for the Federalist Papers, I then asked him to look at the relevant comments from Cecropia on the Talk page for the Federalist Papers regarding useful suggestions for the organization of the sibling articles for the Federalist Papers, and I'm in agreement with Cecropia's comments. BigAlien again refused to look at that Talk page and Cecropia's comments even though I asked him twice to look at the comments specifically about Federalist Paper #2. BigAlien refused again apparently for the same reason that it was a different page than the specific one which he nominated and therefore he seems to believe he has no need to look at those closely related sibling articles. My agreement is with the comments and outline presented by Cecropia about the better organization of the Federalist Papers and the closely related sibling articles, which BigAlien refused to look at during the peer review. This is the Federalist Papers table as presented by Cecropia which I'm in agreement with:

The organization of the 85 Federalist Papers from User:Cecropia on 21 June 2004:

Cecropia's list of the organization of the 85 Federalist Papers.
1General Introduction
2-7Concerning Dangers from Foreign Force and Influence
8The Consequences of Hostilities Between the States
9-10The Union as a Safeguard Against Domestic Faction and Insurrection
11The Utility of the Union in Respect to Commercial Relations and a Navy
12The Utility of the Union in Respect to Revenue
13Advantage of the Union in Respect to Economy in Government
14Objections to the Proposed Constitution from Extent of Territory Answered
15-20The Insufficiency of the Present Confederation to Preserve the Union
21-22Other Defects of the Present Confederation
23The Necessity of a Government as Energetic as the One Proposed to the Preservation of the Union
24-25The Powers Necessary to the Common Defense Further Considered
26-28The Idea of Restraining the Legislative Authority in Regard to the Common Defense Considered
29Concerning the Militia
30-36Concerning the General Power of Taxation
37Concerning the Difficulties of the Convention in Devising a Proper Form of Government
38The Same Subject Continued, and the Incoherence of the Objections to the New Plan Exposed
39The Conformity of the Plan to Republican Principles
40The Powers of the Convention to Form a Mixed Government Examined and Sustained
41-43General View of the Powers Conferred by the Constitution
44Restrictions on the Authority of the Several States
45The Alleged Danger From the Powers of the Union to the State Governments Considered
46The Influence of the State and Federal Governments Compared
47The Particular Structure of the New Government and the Distribution of Power Among Its Different Parts
48These Departments Should Not Be So Far Separated as to Have No Constitutional Control Over Each Other
49Method of Guarding Against the Encroachments of Any One Department of Government by Appealing to the People Through a Convention
50Periodic Appeals to the People Considered
51The Structure of the Government Must Furnish the Proper Checks and Balances Between the Different Departments
52-53The House of Representatives
54The Apportionment of Members Among the States
55-56The Total Number of the House of Representatives
57The Alleged Tendency of the Plan to Elevate the Few at the Expense of the Many Considered in Connection with Representation
58Objection that the Number of Members Will Not Be Augmented as the Progress of Population Demands Considered
59-61Concerning the Power of Congress to Regulate the Election of Members
62-63The Senate
64-65The Powers of the Senate
66Objections to the Power of the Senate To Set as a Court for Impeachments Further Considered
67-77The Executive Department
78-83The Judiciary Department
84Certain General and Miscellaneous Objections to the Constitution Considered and Answered
85Concluding Remarks

This table correctly shows that many of the individual Federalist Papers are very closely related; in the case of Federalist#2, it is listed as part of a full sequence of essays written by John Jay, with the understanding that these papers should be grouped together. My attempt to get BigAlien to look at this was again refused by BigAlien. Given these multiple and repeated refusals by BigAlien to even look at the useful edits made by Z1720 and Cecropia the article appeared to be making no progress toward GAN and I then closed the assessment due to non-participation by BigAlien and the other reasons given. The article would benefit significantly if a knowledgeable editor could combine these articles as properly suggested by Cecropia; when that is done then it can be re-nominated as being in agreement with Z1720 and Cecropia's useful comments. I'm in agreement with both Z1720 and Cecropia, which BigAlien has chosen to fully ignore. ErnestKrause (talk) 15:33, 17 May 2023 (UTC)

@Z1720 and Cecropia:, do you mind confirming that these quickfails were done with your agreement? For myself, I cannot fathom how opening a GAR on a separate article or outlining a possible layout nineteen years ago could possibly consist of "agreement", but perhaps EK possesses some sort of telepathy. ~~ AirshipJungleman29 (talk) 15:41, 17 May 2023 (UTC)
The conversation at Talk:Federalist No. 2/GA1 is quite short. I don't interpret it as the series of refusals it is being presented as above. It is unclear in that discussion that you were referring to the various items specified above, so I don't think there was refusal so much as the message not being conveyed. I posted on your talkpage about the limited considerations of the GACR compared to wider content debates, did you have a chance to read that? CMD (talk) 15:44, 17 May 2023 (UTC)
@AirshipJungleman29: I'll address a couple of ErnestKrause's comments below, with quotations from EK's comments above:
  • "My edit was made in agreement with two other editors, Z1720 and Cecropia." No, I did not agree to quickfail the nomination and I have not evaluated (nor have I read) those articles.
  • "User:Z1720 has listed a GAR for problems with the Federalist Papers article. I agree with Z1720 about the problems with the main page for the Federalist Papers and its sibling articles." Yes, there are problems with the Federalist Papers article and I listed it for GAR. No, I did not state any concerns with its sibling articles because I have not evaluated them.
  • "When I asked BigAlien to look at this main article for purposes of his sibling article nomination he refused to look at it, apparently stating that the article was not the one he nominated and that therefor he would refuse to look at it." BigAlien does not need to improve the main article before the daughter articles can pass GAN. There is nothing in the GA criteria that states that parent articles need to meet any quality for an article to be a GAN.
  • "My agreement is with Z1720 for listing it and bringing to everyone's attention the problems causing the GAR for the main article and its relation to its close sibling articles." When I nominated Federalist Papers for GAR, I did not know about the sibling/daughter article GANs. My concerns about the article are unrelated to the status of the daughter articles; instead, my outlined concerns were the FP's lack of citations, bloated sections, and lack of comprehensiveness.
  • "Given these multiple and repeated refusals by BigAlien to even look at the useful edits made by Z1720" I do not think I have edited any of the mentioned articles, other than initiate the GAR, and I am too busy in real life to undertake this task.
  • I don't think the GANs should have been quick-failed. Rather, if there is disagreement on if the articles should be merged, a WP:MERGE discussion can be initiated. Once that has been settled, then the nomination can be quick failed (if the articles are merged) or proceed (if the articles are not merged). I have no comment on whether the articles should be merged. The nominations should not be quick failed because of BigAlien's refusal to work on the Federalist Papers article because that is not the article that is nominated.
Those are my responses. Z1720 (talk) 16:03, 17 May 2023 (UTC)
Thank you for your precision, Z1720. While we wait to see if Cecropia will respond, do you have anything to say in response to Z1720's comments, ErnestKrause? Possibilities include: admissions of error, strikes of false statements, revertions of failed GANs, or perhaps a refusal to do any of them. ~~ AirshipJungleman29 (talk) 16:19, 17 May 2023 (UTC)
Oh, this line of argumentation again. I've seen it at Talk:Russian invasion of Ukraine. I refrained from commenting this morning, but this response merits refutation.
My edit was made in agreement with two other editors, Z1720 and Cecropia is nonsense. Cecropia posted a list to Talk:The Federalist Papers#Table of Contents twenty years ago. That list was not created by Cecropia. It was copied from a list article that is now a redirect[4] that was incorporated into the The Federalist Papers article eleven years ago and has retained the same form since that time. Z1720 posted a GAR for the The Federalist Papers article because it contains numerous unreferenced claims, an overlong lede, and fails to meet GA criteria for broad coverage. They made no comment about any of the sibling articles or their nominations. As any editor will be able to discern. The cited comments have nought to do with the three GANs. Thebiguglyalien has no reason to address either of these comments, because they are unrelated to his nominations.
This is a pattern for Ernest Krause in a dispute. Claim that their disputed edit has the support of editors, who usually have no awareness of it. A recent example of this tactic is found in in this discussion, during which I issued a 3RR warning. They claimed to be supporting several editors, with those editors then rejecting the claims.
The minimum action to be taken here is for those nominations be re-instated as if no review had taken place (because no adequate review has taken place). Mr rnddude (talk) 17:22, 17 May 2023 (UTC)
I hate to say it, but seeing as this sort of disruptive editing has clearly happened on more than one occasion, is a visit to the drama swamp necessary for some sort of sanctions? ~~ AirshipJungleman29 (talk) 17:45, 17 May 2023 (UTC)
I think so. This is ridiculous behavior and clearly disruptive. ErnestKrause seems to believe they can bend other editors to their will and acts out (and lies) when that doesn't work. This behavior is not compatible with a collaborative project. Trainsandotherthings (talk) 20:29, 17 May 2023 (UTC)
I agree with everyone else that none of these appear to be valid quickfails. It's hard to work out exactly what grounds the quickfail is being made under, but I presume it must be the first, It is a long way from meeting any one of the six good article criteria, as the others are all clearly inapplicable. The only substantive comments on the article seems to be that they have poor lede sections, and very rudimentary contents barely covering material being useful; I don't see that the leads are "a long way" from complying with MOS:LEAD, or that there are sufficient obvious omissions from the article that a quickfail on grounds of broadness of coverage can be valid.
I'm not sure what the best course of action is going forward: ideally I think ErnestKrause would self-revert and we could delete the invalid review pages. Caeciliusinhorto-public (talk) 10:51, 18 May 2023 (UTC)
I'm carefully reading your comment, Caecilius, and I'm not sure that your ideal self-revert would work due to the interedits done by other editors and the way that Wikipedia scripts work, given their limited ability to reset location setting in the GAN queues and the page number tallies used by internal Wikipedia scripts. If you are asking instead for the full report of BigAlien's refusal to answer edit requests which I made, then I need to state that my belief is that it would be more beneficial for Wikipedia for me to endorse fully the suggestion offered by another editor above that a WP:Merge discussion be initiated prior to any re-nomination of the articles. The difficulty remains that BigAlien has fully refused to discuss this issue regarding the strong relation between these sibling articles which are very closely related to each other. For example, during the peer review for the Wikipedia Rolling Stones article last year, I pointed out to the editor that the edits should be consistent with the Wikipedia sibling article for Mick Jagger; that editor thanked me for the observation and added a half-a-dozen to a dozen edits to the Rolling Stones article which was eventually successfully promoted. BigAliens idea of fully isolating sibling articles from one another because they are separate articles does not appear to be the best option.
BigAlien has also stated that he believes that cutting up the individual Federalist Papers at the joints is the only approach that he likes because some other editor long ago decided to simply break up the Federalist Papers at the joints into separate articles; BigAlien's view is made even though the Founding Fathers themselves indicated that the individual Papers were in numerous cases (as presented by User:Cecropia) fully associated in sequential order, the one Paper with its following Paper. BigAlien has rejected this viewpoint indicated that he is opposed to seeing any relationship between these sibling articles because they are separate nominations. My support is for the suggestion for someone/anyone to start the WP:Merge discussion mentioned above by another editor as a useful option and I'll support them in doing this. I'm conscious of the fact that there of six of you who appear to love all the edits from BigAlien under any circumstances, however, I think that Z1720's suggestion to start the WP:Merge discussion should be supported. I'm also fully supporting Z1720's published request to review the main article for the Federalist Papers for possible delisting at this time; this opinion of Z1720's is useful for someone/anyone to initiate the WP:Merge discussions mentioned above. Once that's done, then the new nominations can be renominated by any editor and all the Wikipedia GAN scripts will be reset automatically. ErnestKrause (talk) 15:32, 18 May 2023 (UTC)
Quit the bullshit and the walls of text. You're completely out of line, attempting to force through your personal preferences again unanimous opposition, and if you won't respect consensus we will be going straight to ANI. You've wasted enough of our time. Trainsandotherthings (talk) 19:02, 18 May 2023 (UTC)
I've had enough of the WP:SEALIONing, and am preparing to jump into the swamp Trainsandotherthings. Welp, great minds think alike, as they say... ~~ AirshipJungleman29 (talk) 19:57, 18 May 2023 (UTC)

Per the above I've returned those three GANs to the queue. I closed the old reviews and pulled them from the talk pages. I iterated the GAN templates so hopefully everything should be in order now. Thebiguglyalien sorry for the slow action. ErnestKrause the purpose of a GA review is to review an article against the Wikipedia:Good article criteria. Your greater vision for a suite of articles is outside the scope of a GA review, as is your sense that a different article on a similar topic has elements you wish were incorporated into a given article. If you feel your opinions on a topic are too strong to allow you to review an article against the criteria, I'll remind you that you don't have to review GANs on that topic (or any topic). You're always welcome to voice your opinion at talk (e.g. "I truly feel this entire suite of articles would be best covered as a single article at Parent topic"). But opening GA reviews, expounding your greater theory for how they should be, then failing them for reasons outside the GA criteria is disruptive, and wastes all of our time. If you can't play within the rules/norms of the GA process, I suggest you avoid it. Ajpolino (talk) 20:58, 18 May 2023 (UTC)

Also page ratings can be changed by anyone. Folks aren't watching most of these articles (and most WikiProjects are moribund), so typically the editor who updates the article updates the article rating on the project templates. Ajpolino (talk) 21:02, 18 May 2023 (UTC)
Thanks. They've been waiting a few months for a review, so a day out of the queue won't make much of a difference. I'm just hoping that the fracas around them doesn't ward off potential future reviewers. Thebiguglyalien (talk) 22:32, 18 May 2023 (UTC)
Once the ANI discussion concludes, I'm happy to delete the old reviews if you prefer. But I doubt folks will be much put off by them... Maybe it'll help you attract a pity review :D Ajpolino (talk) 03:15, 19 May 2023 (UTC)

Concerns about a review

So I have some concerns about a review that was done. The review in question was done for the article 2022 South Ayrshire Council election and it was done by the reviewer @Asheiou:. What concerned me about the review and article (aside from it being a simple checklist review) was that the article used multiple Daily Record sources, which are related to the possibly unreliable source of the Daily Mirror, while the candidates section (also using the Daily Record) has some notable verification concerns. Particularly with how "but the Greens did not contest the election as they did five years previous. Both the Scottish Family Party and the Alba Party fielded their first ever candidates in a South Ayrshire election." is not noticeably said in the source. Despite this, none of this was brought up in the review and the review passed with no issue. Thoughts on this? (do note that Asheiou has reviewed some other articles that could also need checking.) Onegreatjoke (talk) 02:47, 19 May 2023 (UTC)

I don't think we can expect reviewers to understand the ins and outs of UK broadsheets. Regarding the sourcing issue, a single unfound text-source integrity issue a cause for great concern on its own. That said, I do note that Talk:2022 South Ayrshire Council election/GA1 does not contain any evidence of source spotchecks for text-source integrity or copyvio, so perhaps that was lacking during the review. CMD (talk) 03:18, 19 May 2023 (UTC)
As the editor who nominated the article, I had hoped wee things that I had missed like this would have been spotted in the review. It was based on a previous GA that I had nominated and had promoted successfully (2022 East Ayrshire Council election, and yes I understand that doesn't automatically make this article a good article). What I had meant to do with that section was provide a reference with candidates from 2017 and from 2022 so that all of the statements were verifiable if not explicitly written in the source or paraphrased from what I'd written. The 2017 source had been missed out from this article but I've fixed that. I'm going to now find a source which lists the Scottish Family Party so that is not missing as well, I hadn't realised the source listed every party but one. Any other feedback on the article is welcome.
As an aside, the Record website is a bit complicated as the majority of the stories are posted by local newspapers who no longer have an independent website. These articles have been written by staff at the Ayrshire Post. I find the style of writing is often too tabloid like for a local paper but they are reliable. Stevie fae Scotland (talk) 07:57, 19 May 2023 (UTC)
I apologise if I did not put the care into this review that I should have done. I did do source spotchecks and checked for copyvio, but I forgot to write that down on my review. As Stevie mentioned, something being published through the Daily Record does not automatically discount the information inside of it, especially when, as he's mentioned, they're essentially just a web mirror of an actual local newspaper. -Asheiou (they/them • talk) 13:30, 19 May 2023 (UTC)
I think you might have misunderstood me. The Daily Record does not mirror another website. In Scotland, Trinity Mirror (the parent company) did away with a lot of the local newspaper websites and instead publishes the content on the Daily Record website (NB- it may not have been Trinity Mirror who decided to do this as they acquired Scottish and Universal at some point in the last 20 years so the websites may have been moved before the acquisition). In practice, this means the website for newspapers like the Ayrshire Post, Hamilton Advertiser, Stirling Observer etc is the Daily Record's website. Stevie fae Scotland (talk) 14:08, 19 May 2023 (UTC)
By "mirror" here, I meant "the place print articles are put online". They're "mirrored" from the newspaper. Apologies for the confusion. -Asheiou (they/them • talk) 15:10, 19 May 2023 (UTC)
In future I will take more care to thoroughly explain every stage of my review, including copyvio searching and the like. Thank you for bringing this to my attention @Chipmunkdavis. -Asheiou (they/them • talk) 00:42, 21 May 2023 (UTC)

Reviewing guide

I've written a draft for a reviewing guide at User:Thebiguglyalien/Good article reviewing guide. Current attempts to provide instructions for reviewers at WP:GAI and WP:RGA have a serious omission in that they don't actually describe the reviewing process itself. RGA is also quite outdated, and it would probably be better suited as a guideline to govern GA rather than as an instruction manual. The guide I've written draws from GAI, RGA, and WP:GACN, and it provides additional detail on each thing that needs to be checked so that this information is in one place. It also includes links to guideline and advice pages for each thing that needs to be checked so reviewers can get further information if needed. I'd love to get some feedback on it and what it might need to be useful for new (or experienced) reviewers. It's in my userspace right now because it's still a rough draft, but anyone can feel free to edit it. Thebiguglyalien (talk) 02:14, 18 May 2023 (UTC)

You're awesome! I see that you advise inline citations. I know this is the norm (and verifying general references quickly become unrealistic as an article grows in size or complexity). Did we ever officially update the the policy to reflect the practice? I think DFlhb was going to run an rfc to change it? We should go ahead and do that; get the criteria and reviewing practices on the same page.Rjjiii (talk) 01:56, 19 May 2023 (UTC)
My wording might be a little more insistent on inline citations than the instructions page, but I was careful not to say "every claim must have an inline citation unless it's one of these exceptions". As of now, inline citations are only required for a few specific circumstances, which are listed at the instructions page and on the guide that I wrote. There was general support for changing the criteria to stricter citation standards among those who participated, but there was never a formal RfC or anything of that nature. I'd much prefer that stricter standard though, especially since it's something of an unspoken standard already. Thebiguglyalien (talk) 02:06, 19 May 2023 (UTC)
I think DFlhb was going to run an rfc to change it? I wanted to, but when I posted an RfC question for workshopping, no one responded, so I assumed there was insufficient interest. (Previous discussions here and here.) DFlhb (talk) 19:20, 21 May 2023 (UTC)
There's not much interest in any sort of reform or improvement at GA. Somewhere higher up on this page I compiled a list of 13 different proposals (including this one) that got some level of support in the last few months but were never implemented. Thebiguglyalien (talk) 20:03, 21 May 2023 (UTC)

One thing I'd like to add to the guide is a few examples of model reviews that new reviewers might find helpful. Did we ever figure out what sort of reviews might be good for that? Thebiguglyalien (talk) 22:35, 18 May 2023 (UTC)

Talk page monitoring

I have just discovered that the bot does not monitor subtopic=Sports for updates, but does monitor subtopic=Sports and recreation for updates to short description.-TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 18:50, 21 May 2023 (UTC)

Tony, I won't have time to look at the code for an hour or two, but that doesn't sound right to me. If I recall correctly, the bot picks up the short description from the GAN template on the talk page. If you update that template, the bot will pick it up, regardless of subtopic. If you update the article's short description on the article itself, the bot pays no attention. Let me know if that's not how it's working. Mike Christie (talk - contribs - library) 19:11, 21 May 2023 (UTC)
Let me clarify. You can nominate an article with either parameter. However, after you have nominated it, if you revise the description later, it does not update on the GAN page with the Sports subtopic. I had to do this edit to prompt the update.-TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 16:25, 22 May 2023 (UTC)
@Mike Christie:-TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 16:48, 22 May 2023 (UTC)
I think that's just a matter of changing the subtopic, TonyTheTiger. You'll probably find that if you change the short description on the article again, the bot won't pick it up until you change the GA template on the talk page. This is the same for every article at GAN. ~~ AirshipJungleman29 (talk) 16:49, 22 May 2023 (UTC)
I just had a look at the code and there is one difference which may be what Tony is referring to. If you change any parameter of the GAN nomination template except the short description, the bot will pick it up, and update everything that's changed, including the short description. If you just change the short description, the bot doesn't notice. It's a bug, albeit a minor one; I'll put it on my list. Mike Christie (talk - contribs - library) 20:33, 22 May 2023 (UTC)
You might not have assessed the logic correctly. In the past, when I was using subtopic=Sports and recreation, I could just change the short description and it would update. Now, when I use subtopic=Sports, the logic is like you say.-TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 15:04, 23 May 2023 (UTC)
Interesting; thanks, Tony. I will see if I can figure out what that's happening. Mike Christie (talk - contribs - library) 15:06, 23 May 2023 (UTC)

Deleted user

I just had to fix a bug in ChristieBot caused by a user being renamed. The nomination is for Liu Ji'en; the nominator is (or was) user Mucube. The bug is that the bot tries to look up the last edit date but there are no edits for the original user name. The bot records the error at User talk:ChristieBot/GAN errors. Because I can't reliably determine the GAs and reviews for a non-existent user, the bot sets them to 0 and 999 respectively. The username is of course a red link.

When this happens, it usually just means that either the nomination template should be updated with the nominator's new name. In this case though the user is invoking their right to vanish so I would assume we just delete the nomination from the article talk page. Mike Christie (talk - contribs - library) 12:43, 18 May 2023 (UTC)

I've removed the GAN. CMD (talk) 14:21, 18 May 2023 (UTC)
And it's been reverted by the Deleted user? I'm not sure what do to here, but I'll ask them if they're active or not. CMD (talk) 10:44, 20 May 2023 (UTC)
They've changed their name back to Mucube so the problem has gone away. The new name was invalid per WP:MISLEADNAME and presumably someone told them so. Mike Christie (talk - contribs - library) 10:21, 23 May 2023 (UTC)
Well, I asked them to change it; no idea if that's why they got their old name again and I cannot explain their general behaviour, but it appears to be an okay GAN now and hopefully ChristieBot will work as the username is the same now. CMD (talk) 15:29, 23 May 2023 (UTC)

Recently promoted good articles

Is there a place where I can view recently promoted good articles? I've looked around, but found nothing. ––FormalDude (talk) 14:40, 25 May 2023 (UTC)

User:ChristieBot/Recent GAN activity shows a weeks worth of activity; it includes passes, fails, nominations, and review starts. Mike Christie (talk - contribs - library) 14:43, 25 May 2023 (UTC)
Perfect, thank you! ––FormalDude (talk) 15:04, 25 May 2023 (UTC)
The most recent ones are at Wikipedia:Good articles/recent. CMD (talk) 15:05, 25 May 2023 (UTC)
May be helpful to offer one of these links at WP:Good article nominations/Report. ––FormalDude (talk) 15:34, 25 May 2023 (UTC)
Oh apparently I'm a moron, it's listed at WP:Good articles. ––FormalDude (talk) 15:48, 25 May 2023 (UTC)

Bellona's Husband: A Romance

Bellona's Husband: A Romance was created yesterday, and made into a GA an hour later. I objected to this, as the article isn't GA -worthy in my opinion, at Talk:Bellona's Husband: A Romance#Objection to GA pass. First people said that I should open a GAR then, but these are for articles which no longer meet the GA criteria, not for incorrect promotions. I was then pointed to this page. Would it in situations like this be best to just undo the promotion and reopen the review, per WP:BRD (while stating my reasons for this action of course)? Or is this a GA after all and are my objections unjustified? Fram (talk) 07:25, 25 May 2023 (UTC)

Regarding review quality, while we don't have hard and fast rules my observation of recent practice (of which I am involved) is that for very poor recent reviews (recent being a matter of art) we usually invalidate the GAN and restore the nomination to its original date if it is raised here. As for the broader question of disputes over whether newly promoted articles are GAs or not, that is an unusual situation, I would suggest it's usually discussed at either the nomination page or at GAR. In this particular case, I think returning it to the pool for a proper review is a sensible outcome. CMD (talk) 07:48, 25 May 2023 (UTC)
Nominator here. I'll restate for the record the question I asked at the article's talk page: you say that article does not cover the major aspects of the subject matter. Would you say that's because (1) these major aspects are mentioned in the cited sources but not the article, (2) the major aspects are mentioned in sources that are not cited but should have been, (3) the major aspects are not mentioned in any available sources and the article is consequently fundamentally ineligible for WP:Good article status, or (4) some other reason I haven't thought of? TompaDompa (talk) 07:51, 25 May 2023 (UTC)
You wrote the article and nominated it for GA, shouldn't you be the one who knows this? Fram (talk) 08:17, 25 May 2023 (UTC)
You're missing my point. Of course I know what I think, but I want to know what you think because you asserted that the article does not cover the major aspects of the topic and I don't understand how you came to that conclusion. TompaDompa (talk) 09:17, 25 May 2023 (UTC)
At the time it was promoted to GA (and when I wrote my talk page comment), the page looked like this. It had a synopsis and a reception section, but it didn't even have in the body any of the info found in the infobox (which normally should already be sufficient to not promote it to GA, an infobox should summarie key points from the article at a glance): it also had no information at all about the place of this work in the life and bibliography of the author, or any of the other aspects I raised. I am rather baffled by this line of questioning, was there anything I raised in my comments which was actually present in the article, or which you believe doesn't belong in a comprehensive article about a book? I mean, "I don't understand how you came to that conclusion", er, by reading the article? How else? Fram (talk) 09:39, 25 May 2023 (UTC)
Well, what you said was that it did not cover the major aspects of the topic. As per WP:GACRNOT, Requiring the inclusion of information that is not known or addressed by reliable sources is a mistake to avoid when reviewing nominations. Have you located the information you think is missing in sources on the topic? TompaDompa (talk) 09:55, 25 May 2023 (UTC)
So there was e.g. no source discussing his other works, placing this in his oeuvre?[5]. No short background on his career, his use of this pseudonym[6], nothing about the publisher (which was in the infobox, not in the text). If there isn't a lot, fine, but absolutely, completely nothing? Fram (talk) 10:20, 25 May 2023 (UTC)
GAs aren't required to be comprehensive, and the broadness criteria is somewhat poorly-defined. The footnote in criteria 3a explicitly allows for "articles that do not cover every major fact or detail". ♠PMC♠ (talk) 10:32, 25 May 2023 (UTC)
How can something be a good article on a subject if it doesn't even cover every major fact? With that footnote, basically every criticism can be countered, just make sure that the few things you do cover are well-sourced and copyright-free and voila, a GA? That makes the GA badge relly meaningless (it's harder to meet the Wikipedia:Content assessment/B-Class criteria than to be a GA apparently?). Fram (talk) 15:10, 25 May 2023 (UTC)
Sometimes there are gaps in the coverage of a topic in reliable sources. If there are no reliable sources discussing an aspect of a topic, even if it seems very important, we can't really do anything about that. It becomes an issue when there is coverage of major aspects that the article is neglecting. To some extent we assume good faith that nominators are trying to give us comprehensive articles, but it's certainly acceptable to raise the question if you are concerned the article fails the comprehensiveness requirements. If you can for example point to a source that covers something important that the article is ignoring, then you have a problem that needs to be addressed.
I ran into something like this with one of my FACs, where an editor was concerned there was minimal coverage of the construction of the Providence and Worcester Railroad. I was able to respond to those concerns by performing a thorough check of reliable sources and confirming there simply wasn't much coverage of that aspect, and a book written on the railroad said construction was "both swift and uneventful". I do agree that things in the infobox should in most cases be included in the body, and for the few exceptions those need to be cited directly in the infobox. I haven't taken a thorough look at the article, but from a brief glance I would also raise comprehensiveness concerns were I the reviewer. I would certainly expect some sort of background, even if just a few sentences. Trainsandotherthings (talk) 15:32, 25 May 2023 (UTC)
(edit conflict) A source on the author mentioning this book as part of his biography is to my eye not the same thing as a source on the book discussing its place in the author's oeuvre (there is no comparison to the other works, for instance), but I suppose that's an issue where reasonable people can disagree. The article in question is about the book, not the author. WP:GACR 3b specifically requires that the article stays focused on the topic without going into unnecessary detail. Deciding what's straying off-topic and what's providing relevant context is, of course, a judgment call. TompaDompa (talk) 10:39, 25 May 2023 (UTC)
A few sentences giving the background on the author and how he came to write this book would absolutely not fall afoul of 3b. That criterion doesn't mean ignore anything that isn't 100% exclusively about the article's subject itself. Trainsandotherthings (talk) 15:35, 25 May 2023 (UTC)
And I'm not saying it does mean that. There's certainly a case that "Name of Author was a Nationality X author who was born in Example year. Before Current article, their most recent work was Previous work." provides context. There's also a case that it's padding with loosely relevant details, especially if sources on the topic of the article at hand don't mention them. Like, is this addition an improvement? I honestly don't think it is, I think it's a digression. TompaDompa (talk) 16:07, 25 May 2023 (UTC)
It seems fine to me. Trainsandotherthings (talk) 16:34, 25 May 2023 (UTC)
I'm not disputing that it's acceptable. I'm disputing that it's necessary, which is the question that brought us here in the first place. I'm also saying that it's not necessarily an improvement—this is a matter of preference, but I personally view this as needless duplication of information that properly belongs at the article about the author. This is one of the major benefits of not being a paper encyclopedia: readers can just click on a link to another article to get additional information on a related topic that is not strictly relevant to the topic of the article they are currently reading. TompaDompa (talk) 17:10, 25 May 2023 (UTC)
Suit yourself. It's not my GAN. Trainsandotherthings (talk) 18:29, 25 May 2023 (UTC)

I take issue with a WP:GAN review being reverted because an editor raises objections to the article—not the review—that (arguably) go beyond the requirements of the WP:Good article criteria. That is pretty disruptive to the process. TompaDompa (talk) 18:47, 25 May 2023 (UTC)

I wasn´t the only one to see issues with the GA review though. Fram (talk) 21:12, 25 May 2023 (UTC)
Had your objection been that the review was flawed, we would not be having this argument. You said the article did not cover the major aspects, and you listed the aspects you thought were missing (translations, inspirations, and so on). Did you check the sources to see that those were indeed major aspects before doing so? TompaDompa (talk) 21:19, 25 May 2023 (UTC)
I hope it can be agreed that the review was not really a review, and that the article can be moved back to the GAN list for that reason. As for broadness, that is as noted an editorial call, and would likely better be more productively discussed at an appropriate Wikiproject or similar forum rather than here. CMD (talk) 00:59, 26 May 2023 (UTC)

Scrutinizing recent GA activity

To get an idea of what's turning up at User:ChristieBot/Recent GAN activity, I checked last week's activity for a few things:

I don't expect to post something like this every week, but I think it's worth checking these things more carefully if we're going to have any sort of quality control at GA. Thebiguglyalien (talk) 05:52, 28 May 2023 (UTC)

Potential RfC to increase citation requirements

There's recently been support for increasing the citation requirements at GA. DFlhb has proposed the following as an RfC:

Should the WP:GACR require that everything be cited inline, save for the usual exceptions?

Previous discussions here and here. I would also suggest a note that this would not mean all GAs failing this change would immediately be delisted, but that it would be applied going forward while previous GAs would be fixed or delisted gradually, just as they are for any other criterion. Are there any other thoughts or suggestions before such an RfC is created? Thebiguglyalien (talk) 22:17, 22 May 2023 (UTC)

How do you define everything? Every paragraph? Every sentence? Every word? Every sentence is absolutely overkill, not even FA is that strict... there's no point in having a 5 sentence paragraph supported by one source, and citing the same source 5 times in a row at the end of each sentence. Trainsandotherthings (talk) 23:49, 22 May 2023 (UTC)
I would consider "everything" to mean at least one citation per paragraph, at the end of the paragraph, with the assumption that all material is cited by the references; some may choose to do two cites that collectively cover an entire paragraph, or, if the material covered is interspersed, they may choose to break it up into separate cites throughout. But, basically, no prose text without some form of citation. I think this meaning is somewhat standard, but if that definition needs to be made more clear I have no opposition. Iazyges Consermonor Opus meum 23:53, 22 May 2023 (UTC)
The idea as I understand it is that every claim is associated with an inline citation. If there's one citation at the end of the paragraph, then it should support everything in that paragraph. It doesn't matter if it's every paragraph or every sentence, as long as the citation supports everything behind it. Similarly, the lead usually doesn't need citations because it corresponds to cited text in the body. Is there a better wording that would help clarify this? Thebiguglyalien (talk) 00:13, 23 May 2023 (UTC)
The wording for the same sort of rule in WP:DYKSG is: "The article should, in general, use inline cited sources. 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)." —David Eppstein (talk) 00:43, 23 May 2023 (UTC)
"Everything" is indeed imprecise. I like the DYKSG rule, both accurate and well-written.
Note that my RfC question wasn't meant to be text we add to the WP:GACR, but meant to determine whether there's consensus for that kind of addition, at AirshipJungleman29's suggestion here. I didn't personally think an RfC was needed. DFlhb (talk) 02:48, 23 May 2023 (UTC)
Do we need an RfC? This simply codifies long-standing existing practice. CMD (talk) 01:41, 23 May 2023 (UTC)
I agree. If there are objections we can hold an RfC; if there are none after a couple of days I think it would be fine to change GACR to use something like the wording David quotes. Mike Christie (talk - contribs - library) 01:52, 23 May 2023 (UTC)
An RfC would be advisable if we wanted to make a substantial change to WP:RGA, which is a guideline. It would also establish better grounds to say that there's consensus for this standard. Per WP:CONLEVEL, such a change would otherwise have no more weight than an essay. Thebiguglyalien (talk) 02:01, 23 May 2023 (UTC)
Perhaps I'm simply reading CONLEVEL wrong, but it seems that just a large-scale discussion of Good Article talk page watchers would suffice, hardly counting as being a limited group of editors, in my mind, and satisfying editors often propose substantive changes on the talk page first to permit discussion before implementing the change. I'm generally of the opinion that, if agreement is strong among those that show up in this discussion, the change can be made. Especially since the guideline applies directly and immediately to "us", and few (if any) other regions. I'm not opposed to an RFC if others are of the mind that it is necessary, however. Iazyges Consermonor Opus meum 02:37, 23 May 2023 (UTC)
It's not that we can't implement it. We just have to make sure we're not contradicting sitewide consensus at any point, that's all. Groups like this are exactly what it's about; the example it gives of a "limited group of editors" is a WikiProject. Thebiguglyalien (talk) 02:58, 23 May 2023 (UTC)
And it should be remembered that the GA process is technically a WikiProject which has expanded into being official Wikipedia policy. An RfC is definitely required per CONLEVEL. ~~ AirshipJungleman29 (talk) 11:44, 23 May 2023 (UTC)
Ah, I suppose I had forgotten that; definitely agree we need an RFC, then. Iazyges Consermonor Opus meum 14:54, 23 May 2023 (UTC)
I think WP:DYKSG#D2's 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)., brought up by David Eppstein above, is pretty good. My only issue with it is that it might be read as prohibiting citing an entire column of a table (or for that matter an entire table) at the top—though I suppose "no later than" should cover that. TompaDompa (talk) 02:43, 23 May 2023 (UTC)
I do not think we need to reinvent the wheel here. No wording will capture every edge case, and the existing wording from elsewhere works. We could also link to Wikipedia:When to cite, which clarifies the much vaguer FACR. CMD (talk) 03:03, 23 May 2023 (UTC)

I'm not convinced WP:CONLEVEL applies to the proposed wording change; we're talking about changing a definition that applies to GAs, and WT:GAN seems a sensible place for that discussion. FACR changes are certainly discussed primarily at WT:FAC. However, in the interests of making this happen, DFlhb, can I suggest you post RfC wording based on the discussion above? Mike Christie (talk - contribs - library) 01:07, 25 May 2023 (UTC)

New proposed RfC wording:
Should the GA-class citation requirements be increased by replacing GA-class criterion 2.b. (citations), with the following, taken from WP:DYKSG:
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).
Second sentence is intact; but I modified the first sentence to mention sourcing reliability and link to WP:RS, like the current 2.b does. Can just propose the unchanged DYKSG if people prefer. DFlhb (talk) 07:36, 25 May 2023 (UTC)
I like the consistency with DYK. It'll make things smoother and simpler if people take the DYK > GA [ > A ] > FA path. You have my vote in advance, Rjjiii (talk) 06:58, 27 May 2023 (UTC)
I'm happy with this wording. Have we come to a conclusion on whether to hold an RfC? Thebiguglyalien (talk) 19:09, 28 May 2023 (UTC)

I was considering just making the change, but looking at the wording again I have another question. We are proposing changing

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).

The list of things that should be cited is removed because now we require everything to be cited; that seems to have consensus. How about the mention of the scientific citation guidelines? See the example on aldol reaction given there; the second paragraph is not cited as the new wording would require. I have not written articles that use this citation approach myself but I believe it has its proponents. Should it be retained? Mike Christie (talk - contribs - library) 22:12, 28 May 2023 (UTC)

I think it's reasonable for GA to have higher standards than individual subject guidelines. It's also my understanding that the community is much more pro-citation than it was some years ago, so I couldn't say whether that guideline reflects current practice. I personally would not feel comfortable passing a GA with an unsourced paragraph like that. Thebiguglyalien (talk) 22:25, 28 May 2023 (UTC)
I think we can lose the link without a problem. I agree with Thebiguglyalien that it might no longer reflect current practice/consensus, but either way it's ok for us to have a higher standard and to lose the link. Ajpolino (talk) 22:42, 28 May 2023 (UTC)
After thinking about it some more I agree, but I also think removing that option means it would be best to have an advertised RfC. I'll post it using the wording above, probably tomorrow, unless someone beats me to it. Mike Christie (talk - contribs - library) 23:43, 28 May 2023 (UTC)
As someone whose GA nominations are often quite technical, I'm not convinced that the "general reference" method described in Wikipedia:Scientific citation guidelines documents current standard practice, and I'm happy enough for it to be lost. I do think that WP:CALC is important, but that's part of Wikipedia:No original research, so it wouldn't be affected by this. —David Eppstein (talk) 01:52, 29 May 2023 (UTC)

When a GA's talk page is being archived, what is supposed to happen to the (transcluded) GA Review?

Case in point: The article Mary Surratt with its Talk:Mary Surratt which contains the Talk:Mary Surratt/GA2 which is transcluded onto the page as {{Talk:Mary Surratt/GA2}}. In my opinion, the talk page doesn't get enough activity to justify placing an archiving-bot coding onto it so I have been manually archiving its content. So, my question is, O Wise & Experienced GA/GAN folk, what am I supposed to do with the "GA2" code? I was going to remove it since linkage to the GA Review is already in the talk page headers but thought I should ask here first to make sure. Thanks, Shearonink (talk) 17:26, 29 May 2023 (UTC)

When I have to archive a talk page manually, I delete the transclusion -- it's still linked from the GA template at the top of the page after all. Mike Christie (talk - contribs - library) 17:37, 29 May 2023 (UTC)
Ok, cool. Thanks - that's exactly what I needed to know. Shearonink (talk) 17:52, 29 May 2023 (UTC)

Draft log page for recent GAN reviews

Per a conversation higher up the page, I've just created User:ChristieBot/RecentGANActivity, which shows the last seven days of passes, fails, and review starts, along with review count, GA count, and edit count numbers for both nominator and reviewer. If there are no objections, I'll set this to refresh once a day (or more often if there's a desire for that). Like User:ChristieBot/SortableGANoms I can move this under the GA project space if we want to do that. Mike Christie (talk - contribs - library) 12:07, 17 May 2023 (UTC)

Looks useful. Would it be possible to make the 'Title' column link to the GA review, and would it be possible to add a column for review character length? That might help to highlight any more likely issues. (Obviously review length is far from the be all and end all, but...) Harrias (he/him) • talk 12:33, 17 May 2023 (UTC)
Yes to both; I'll do that tonight or tomorrow. Mike Christie (talk - contribs - library) 12:44, 17 May 2023 (UTC)
Done. For the review length, I've attempted to give the number characters added by the reviewer and nominator; that includes subtract the length of the review template if it's been used. If a reviewer deletes part of the review template this could result in negative lengths, but I think it's more useful to give the length of added text than length of the overall review. Mike Christie (talk - contribs - library) 22:21, 17 May 2023 (UTC)
A brief look suggests to me that passes with a size under 2,000 warrant scrutiny. Currently, that includes:
I suspect this is pretty normal for what gets through GA in a given week, and that it's gone on for years without being caught. Interestingly, I don't find first time reviewers to have a strong correlation with bad reviews, though the sample is admittedly small. But as I said earlier, I think some of the value in a tool like this is the ability to help new reviewers earlier in the process. Thebiguglyalien (talk) 23:35, 17 May 2023 (UTC)
I've contacted the author of Talk:Centipede/GA2 noting they can reach out for advice here, I don't think more is needed there. Talk:Visa policy of Azerbaijan/GA1 gives me AI vibes, although it has typos so that suggests it's not AI. CMD (talk) 01:54, 18 May 2023 (UTC)
I think it's also interesting to see who the reviewers are for particular topics. Is there anything in place to ensure nominees in one niche area aren't always reviewed by the same editor? JoelleJay (talk) 22:53, 31 May 2023 (UTC)
JoelleJay: There is no such mechanism. Given that there are way more nominations than there are reviewers, I'd be surprised if this wasn't the case for a few subjects. My (purely anecdotal) impression is that this is especially common with military history and video games. Thebiguglyalien (talk) 03:16, 1 June 2023 (UTC)

I've automated this so it will run once a day at about 5:00 p.m. east coast time. I moved the page to User:ChristieBot/Recent GAN activity. One thing that is not included is new nominations, which might be of interest if someone wants to check nominations by inexperienced users. I can add those easily if there's interest in doing so. Mike Christie (talk - contribs - library) 11:22, 19 May 2023 (UTC)

It might be a good idea to find a place for User:ChristieBot/SortableGANoms and User:ChristieBot/Recent GAN activity so they're still easily accessible after this gets archived. Given how spread out some of the GA stuff is, a "useful links" box somewhere might be good, if only to make it accessible for people working on the behind the scenes aspects of GA. Thebiguglyalien (talk) 02:31, 25 May 2023 (UTC)
Seems a good idea to me. Other possible pages that could be added to that box are User:GA bot/Stats, User talk:ChristieBot/Bug messages, and User talk:ChristieBot/GAN errors. And perhaps a note that people can request GA reports for their own userid -- e.g. User:Mike Christie/GA nomination lists/Hawkeye7. Mike Christie (talk - contribs - library) 13:27, 25 May 2023 (UTC)

Notice: Women in Green editathon

Women in Green is currently hosting an editathon for the month of June. Editors are encouraged to join the event to nominate and review articles about women and women's works. There's a barnstar in it for participants. Thebiguglyalien (talk) 05:51, 2 June 2023 (UTC)

Julius Schaub: Reassessment Request

I've added a note at this article's Talk page that it should/might be delisted, if others would kindly have a look: Julius_Schaub, Billsmith60 (talk) 18:00, 3 June 2023 (UTC)

Need reviewer to close review

Hi, editor (Shibbolethink) has been reviewing Talk:COVID-19 pandemic/GA4 , I've answered about 224 comments, however Shibbolethink stopped editing here [7] (actually he doesn't seem to have edited anywhere). I left him a message [8] and sent an email, however I haven't gotten a response to either. It seems he may have finished the review or be very close to it, I'd appreciate anyones help, thank you--Ozzie10aaaa (talk) 23:29, 4 June 2023 (UTC)

It's only been two days. There are all sorts of reasons he might not have time to work on Wikipedia stuff right now. Personally, I usually give it a week of no activity before I check in. It looks like you've both done good work on it though, and if the review is left open for a while longer, then a new reviewer can give it a look. Thebiguglyalien (talk) 23:41, 4 June 2023 (UTC)
(edit conflict) I would simply wait, for now. Just over 48 hours have passed since their last edit, which isn't that long. This could very well be due to a computer breakdown, for instance. TompaDompa (talk) 23:45, 4 June 2023 (UTC)
ok, thank you--Ozzie10aaaa (talk) 23:49, 4 June 2023 (UTC) (BTW now its 3 days [9])--Ozzie10aaaa (talk) 12:27, 5 June 2023 (UTC)
Seems to have returned sorry for post, thank you--Ozzie10aaaa (talk) 16:45, 5 June 2023 (UTC)

GA review exceptions report May 28 – June 3

Hopefully we can come up with some sort of flagging system so this doesn't become a regular thing:

Thebiguglyalien (talk) 16:49, 4 June 2023 (UTC)

Talk:Howard Florey/GA1 has been deleted by Yamla per WP:G5. TompaDompa (talk) 17:08, 4 June 2023 (UTC)
I have undone the promotion, restored the nomination, and notified the DYK nominator. TompaDompa (talk) 17:17, 4 June 2023 (UTC)
I've removed Battle of Longewala, no point running something borderline through given it is but a quick fail anyway due to substantial unsourced text. CMD (talk) 01:55, 5 June 2023 (UTC)
I'm the nominator for Snyder's Bluff, and would be perfectly willing to provide quotes for a spot-check if desired. I'm fairly busy IRL right now, so there might be a couple days' delay though depending on when requested. Hog Farm Talk 02:35, 6 June 2023 (UTC)
@Hog Farm: I spotchecked the two online sources, [10] and [11], and they both support the text and raise no obvious copyvio issues (to the point where it took me awhile to match the first one to your text!). I don't think the article provides any context as to what the "the two halves of the Confederacy" are, but that's just a since-we're-here comment and it is (similarly unexplained) in the source. CMD (talk) 03:02, 6 June 2023 (UTC)

Talk:Jack Critchley/GA1

G'day all, the subject review (of a bio of a South Australian and Australian politician between 1930 and 1959 who briefly served in WWI) has effectively been abandoned, as the reviewer hasn't edited WP since just after the review began in December last year. It seems to me to be relatively complete and comprehensive, just needs a look-over. I'd appreciate it very much if someone would take over the review and finish it up. Thanks in anticipation, Peacemaker67 (click to talk to me) 01:13, 11 June 2023 (UTC)

I'll do this. Steelkamp (talk) 07:46, 11 June 2023 (UTC)

Possible abandoned GA review

They've done it again. User 420Traveler looks to be abandoning another one of my GA reviews, although they did manage to post comments this time. It's been six days since they last edited, and although I'd give them at least a few days, I'd like to get a second opinion on whether this is too long before someone else takes over. Bneu2013 (talk) 16:46, 5 June 2023 (UTC)

I'd say, if it were my nomination, I would ping the reviewer after a week of no action and if it's two weeks since a response from them I would consider asking for someone else to step in. (t · c) buidhe 20:13, 5 June 2023 (UTC)
Looking at their contrib history, it seems they frequently have breaks in editing for periods of a few weeks, so given this hasn't even quite been a week yet, and they at least posted meaningful review feedback, I wouldn't be getting impatient just yet. I think there would need to be several weeks (perhaps 3+, or even a month) before someone would consider stepping in to do another review. Hopefully it won't get to that though. Bungle (talk • contribs) 20:36, 5 June 2023 (UTC)
For the last one they did, which was picked up a little over a week ago, they disappeared for over a month before they made another edit. That being said, while I appreciate their willingness to review nominations, if you are unable to complete a review in a timely manner, then you shouldn't be reviewing GA nominations. Bneu2013 (talk) 21:22, 5 June 2023 (UTC)
I know it can be frustrating when the reviewer doesn't respond promptly but personally I prefer it when someone reviews my nom and takes a few months to do it than when it sits waiting for review for 6+ months (which is also a common occurrence). My own GAN Sorley Maclean—GAN 18 August 2018, passed 23 September 2019. Anything that is discouraging competent reviewers including arbitrary time limits should be avoided. (t · c) buidhe 21:27, 5 June 2023 (UTC)
We say 7 days in the GA documentation. I really can't agree that it's fine to wait months for a reviewer after they've started a review. That's time someone else could have come in and actually done the review in a timely manner. Trainsandotherthings (talk) 02:25, 6 June 2023 (UTC)
Indeed, and I left them a message about this recently. Continuing this behavior would bring me towards supporting some sort of sanction - this really is antithetical to a collaborative process to start a review and then just disappear for a month plus. Trainsandotherthings (talk) 02:23, 6 June 2023 (UTC)
In the case of that review, it didn't seem like it was even meaningfully started and so a 4-week period with no feedback is fair enough to hand to someone else, in that instance. The review in question was at least started and, while it has been only a week thus far, I do think it seems reasonable to suggest that this editor is mindful about taking extended periods of leave if making a review committent. There are no explicit instructions for review time-frames, only an encouragement of responsiveness. The reviewer may well question which element of the reviewing policy or guideline they're falling foul of if this isn't stated explicitly. Bungle (talk • contribs) 06:01, 6 June 2023 (UTC)
Exactly. Further, if a nominator waits several months and then happens to be away for a while when their GAN is finally addressed, that can readily be understood; but if a reviewer voluntarily takes on a commitment to a GAN and then promptly disappears, one may reasonably ask whether they are behaving at all appropriately for the process. Chiswick Chap (talk) 06:27, 6 June 2023 (UTC)
If you start a review, it's pretty reasonable to expect that you will finish it, yes? Trainsandotherthings (talk) 02:13, 7 June 2023 (UTC)

Just for the sake of full disclosure, I want to make it clear that I do not believe that all GAs should be reviewed in seven days (or a month or two months). It is indeed a process that should not be rushed. I won't deny that I am guilty of being slow to finish a few reviews myself. The only reason I raised a concern was based on past behaviors of this reviewer. Bneu2013 (talk) 15:39, 6 June 2023 (UTC)

That make sense. A one-off event could easily be an unfortunate real-world situation (someone got sick, unexpected overtime at work, etc.). But a pattern would suggest that the reviewer might be more useful contributing in other ways (e.g., at GAR or offering second opinions). WhatamIdoing (talk) 04:24, 14 June 2023 (UTC)

Adding the Notability Criteria for GAs

TL;DR - Should notability be added as a criteria for Good Articles? As in, the topic of an article must unambiguously meet any of the notability criteria, whether WP:GNG or WP:SNG in order to be evaluated and pass a Good Article review?

I've wanted to ask this question for a while on why the GA (and FA) criteria doesn't include any mention that an article must meet the notability guidelines. I assumed it wasn't necessary because it was implied that an article should obviously be notable in order to meet GA status. However, I saw an AfD of a FA where participants concluded the article somehow doesn't meet GNG in spite of the FA review. Considering that a FA review is nothing to take lightly, the notability of the topic should've been addressed much earlier than this. And right now, there's even a back-and-forth debate over a topic's notability despite the article meeting GA status already.

I think this issue is definitely something of concern for two reasons: one, it implies that an article of a dubiously notable topic can technically reach GA status. And two, it also implies that any GA can still potentially be deleted in an AfD. I don't believe both of these implications benefit Wikipedia editors nor readers in the slightest, because if GAs/FAs are meant to represent the best work Wikipedia has to offer, then having such articles still be fallible to deletion only indicates that this site itself isn't secure enough about its own encyclopedic content. As an inclusionist I admittedly have more "tolerant," for lack of a better word, standards for notability. And I don't think that a topic must meet GA status in order to be notable. But isn't it reasonable to state that a topic must be notable in order to meet GA status?

Finally, if I had a specific solution, I'd imagine that an implementation of notability in the criteria would go something like this:

  • Criteria #2 (and perhaps #3): at least multiple sources must be present in the article that significantly cover the article's topic. Of course, this goes without saying that these multiple sources must be reliable, independent and secondary.
  • Quick fail policy #6: an article may immediately fail GA if the article's topic is of questionable notability
  • Reassessment: A GA should not be sent to AfD or merged/redirected unless a reassessment is made first that shows that the article not only fails the GA criteria, but the article's topic may potentially be non-notable.

PantheonRadiance (talk) 08:05, 8 June 2023 (UTC)

I think the current policy, is that if you are going to review an article and find that you believe it does not meet notability, put it up for AfD instead of doing the Good Article Review. I don't think I see the issue with the current plan.Rjjiii (talk) 08:09, 8 June 2023 (UTC)
As Rjjiii says, if a reviewer (or potential reviewer) has a notability concern, then they should start an AfD, as that is the most suitable place to discuss notability. It would be inappropriate for either of these processes to weaken AfD (by suggesting that GAs and FAs can't be directly subject to the AfD process). I would oppose any effort to add notability as a criterion for GA or FA. Harrias (he/him) • talk 08:19, 8 June 2023 (UTC)
But if a reviewer has such a notability concern, wouldn't it be appropriate to Quick Fail the article so that the nominator may know in advance that the article's topic may not meet the guidelines? It would give the nominator a chance to at least address the notability issue first before using an AfD to do so. And what about when an article does pass the criteria but as the notability policies change, it may not meet the guidelines anymore? Why wouldn't a GAR suffice before an AfD? PantheonRadiance (talk) 08:35, 8 June 2023 (UTC)
Contrary to how GAN often acts in practice, there is no requirement that every article be either quickfailed or painfully dragged past the finish line. A subject where the article is not in a quickfail-eligible state, but the reviewer has concerns about its appropriateness for an independent article, can be hashed out perfectly well in the review, and the review can be closed non-quickly if there are irreconcilable differences on the subject. It is not universally agreed that articles close enough to GA they can't be quickfailed are likely to be non-notable outside unusual circumstances, and such unusual circumstances are generally too complex and subjective to codify at this level of resolution. 'Notability policies [sic] changing' is also a fairly complex situation that isn't universally agreed to imply either of 1. notability guidelines necessarily tightening, rather than having complex up-and-down courses on many subject matters or 2. something generally applicable to articles capable of passing quality-assessment processes, rather than e.g. stubs. Vaticidalprophet 08:45, 8 June 2023 (UTC)
@PantheonRadiance: No. AfD is the only suitable venue to discuss notability. Doing so at GA or FA is superfluous. What would you propose happen if at GAN or GAR an article is failed or delisted because of notability concerns, but then it survives an AfD? Would the decision at GAN/GAR be reversed, or would it stand? It would be too messy. Let AfD deal with notability, and GA/FA deal with quality. Harrias (he/him) • talk 09:10, 8 June 2023 (UTC)
If, as a potential reviewer of an article you come across something at GAN you have notability concerns about, then I would say the best action is to list it at AfD (after carrying out BEFORE checks, obviously) and then add a |note= to the {{GA nominee}} template on the article talk page, saying something like "Article currently undergoing an AfD due to notability concerns." Harrias (he/him) • talk 09:13, 8 June 2023 (UTC)
Harrias My idea was meant for the case where a topic's notability is dubious insofar it would not survive an AfD in the first place - as in, it decidedly doesn't meet a single notability guideline. In that edge case however, I don't see what the issue is with a future renomination after all the issues are properly addressed including notability itself. Perhaps between the fail/demotion and the AfD's close, clear evidence was found which proves the topic's notability, such as previously existing sources not found during the GA review but found in the debate. In that scenario I would d find it fair to renominate after rewriting the article with such sources that slipped past editors.
Vaticidalprophet I think that a non-QF GA review that involves concerns about a topic's worthiness for an independent article is somewhat along the lines of what I meant, in that if a reviewer's concerned about its notability, it should be addressed in the review just like the rest of the criteria. My quick-fail idea was meant to be along the lines of "it wouldn't come close to surviving an AfD because it obviously doesn't meet the guidelines."
I will concur however that it this is a fairly complex situation because of how notability in and of itself varies in interpretation, and how consensus changes. But in a more general sense, if a topic doesn't seem to meet any notability guideline, it should be addressed in the review itself to avoid instances of nominating current GAs for deletion. PantheonRadiance (talk) 10:00, 8 June 2023 (UTC)
I struggle to imagine an article that hits "blatantly obvious AfD fail" and doesn't hit any existing QF criterion. Keep in mind what the GNG is and what its implications are; "significant coverage in independent reliable sources" is the de facto minimum to write a decent article on a subject, such that 'uncontroversial GNG failure' and 'GAN that obviously can't be quickfailed' are an unlikely combination. This is not to say that no GA or FA has ever had a non-keep AfD outcome, but the specific circumstance being posited is something I haven't ever really seen. Vaticidalprophet 10:04, 8 June 2023 (UTC)
A significant issue for handling notability in GAN in particular is that notability discussions involve a range of perspectives balancing different aspects of en.wiki's notability-related policies and guidelines. A process involving only two editors (a reviewer and a nominator) can not be expected to emulate that. CMD (talk) 09:33, 8 June 2023 (UTC)
I agree that editors have varying perspectives that make this issue complex. Admittedly, the only ideas I had to rectify it were to either have more editors chime in during a GA review (per step 4: "other editors are also welcome to comment and work on the article"), and/or have the reviewer also be from the WikiProject the topic originates from (ex. Book GAN -> WikiProject Books). PantheonRadiance (talk) 10:20, 8 June 2023 (UTC)
I don't disagree that non-notable articles shouldn't be promoted to GA or FA, but I don't know that formalising "must be notable" as a GA criterion is a workable solution. How many articles are sufficiently clearly non-notable that a reviewer would fail a nomination on that basis, and yet meet the existing criteria 2 and 3 of WP:GACR? My suspicion is as close to zero as makes no difference. Most GA/FAs deleted or merged at AfD, IME, are either (a) promoted a long time ago, when the community was more willing to accept articles which would not be created today, (b) promoted based on inadequate reviews or (c) of borderline/questionable notability.
Changing the rules to explicitly allow failing articles for non-notability wouldn't affect classes (a) or (b), but I can think of at least two detrimental effects on class (c), which I suspect may outweigh the benefits. Firstly, it would lead to a potentially messy situation where a reviewer fails an article in good faith for lack of notability, but either the article is never actually AfDed, or is AfDed and kept (or worse, there is no consensus). What happens then? If an article has previously survived an AfD (even through a no-consensus close) does that mean that it automatically meets the notability criterion? Secondly, we already see "keep: it is a GA" votes being made at AfD, and any encoding of notability in the criteria would I suspect only encourage them; I do not want GA and FA to be seen as some special extra layer of protection against deletion whereby you have to convince reviewers to delist an article as non-notable before it can be deleted. Caeciliusinhorto-public (talk) 11:18, 8 June 2023 (UTC)
I would say the AfD keep votes are in fact a reason to add the criteria explicitly, since it's currently (erroneously) being used as a shield anyhow to keep articles that aren't notable. We might as well require a basic sanity check at the GAN level. Der Wohltemperierte Fuchs talk 17:12, 8 June 2023 (UTC)
This has been raised in the past, such as in this proposal last December. Article quality or status is an invalid deletion rationale, and a half decent closer at AfD would throw out that !vote without further consideration. This is an issue that needs to be solved at AfD: there's a major problem with editors giving deletion rationales that don't cite notability guidelines. Such editors need to be reminded that they are participating disruptively, and if they ignore the warning, they need to be topic banned from AfD until they demonstrate an understanding of WP:N. Thebiguglyalien (talk) 17:47, 8 June 2023 (UTC)
Article quality or status is an invalid deletion rationale, and a half decent closer at AfD would throw out that !vote without further consideration is a rather blasé description of genuinely differing positions on "is it possible to write an article that does not pass GNG, but is in compliance with what a 'good Wikipedia article' is". (Lower-case and added word intentional to avoid confusion between 'good articles' and 'Good Articles'; I have seen GAs I think are Start-class, and many high-quality articles outside assessment processes, but so it goes.) My question for this issue is what articles you've seen that are both "something that couldn't possibly be quickfailed at GAN" and "something that obviously couldn't have a chance in hell of surviving a deletion discussion". The biggest situation that regularly comes up regarding objectively-quality-assessed or subjectively-high-quality articles going to AfD is about a specific SNG that prescribes a stricter standard than GNG, and none of the (several) situations I've seen this happen have been anything like the easy cases people imply exist when these discussions happen. Vaticidalprophet 18:12, 8 June 2023 (UTC)
It's also worth noting that a lack of notability is not the only valid reason for deletion, contrary to the implicit suggestion in there's a major problem with editors giving deletion rationales that don't cite notability guidelines. It is one of a canonical list of 14 reasons that is explicitly non-exhaustive. TompaDompa (talk) 21:33, 8 June 2023 (UTC)
Never seen one deleted for lack of notability, but GAs have been deleted as the creation of a sock (criterion #1), due to there being a copyvio (criterion #2) and under WP:IDONTLIKEIT (criterion #14) Hawkeye7 (discuss) 09:19, 9 June 2023 (UTC)
I think notability problems more commonly end up with the article getting merged/redirected than with straight up deletion (the question is "should we have a standalone article about this Pokemon/episode/book", not "should this topic not be mentioned anywhere in the encyclopaedia"). The only straight deletion of a (former) FA that I am aware of is Wikipedia:Articles for deletion/ANAK Society. —Kusma (talk) 09:34, 9 June 2023 (UTC)
And as a hoax (criterion #6): Wikipedia:Articles for deletion/Bicholim conflict. Caeciliusinhorto-public (talk) 10:12, 9 June 2023 (UTC)

This proposal is not needed, as the current rules already accommodate for this scenario. We have rule 3: "It has, or needs, cleanup banners that are unquestionably still valid". An article of questionable notability should be tagged with {{Notability}}, and that article has to be either fixed or deleted before even starting GA nomination discussions. Cambalachero (talk) 18:42, 8 June 2023 (UTC)

That tag should not be used. There is no such thing as "fixing" notability; either a subject is notable or it is not. The proper place for discussing notability is WP:AfD. If the result at AfD is "kept", then tagging the article for notability is vexatious editing and grounds for a block. Hawkeye7 (discuss) 09:19, 9 June 2023 (UTC)
The tag is for articles where it is unclear whether the subject is notable or not. Poorly sourced articles that do not clearly demonstrate that the subject passes WP:GNG often attract it. At AFD, WP:HEY is a nice outcome if the subject is notable but the article did not show it. —Kusma (talk) 09:29, 9 June 2023 (UTC)
I think it'd be fair for reviewers to consider an AFD template as a special case of failing GACR #5, "Stable". A page that might stop existing in a few days is not the epitome of stability. WhatamIdoing (talk) 04:28, 14 June 2023 (UTC)

"Problematic reviewer" appears to have abandoned my nomination

Early last month, after four months of waiting, ErnestKrause took on the task of reviewing my nomination of The Exorcist, which I am hoping to get to FA by the end of the year so as to mark the film's 50th anniversary. I will say a lot for him: he was a careful reviewer who was never anything less than completely civil and collegial. Many of suggestions were helpful, particularly splitting off separate articles for the film's production and themes (both currently still in draft space), just as we have done with other film articles. I also have done a great deal of trimming that has gotten the article down in size from nearly 280K to almost 230.

He let me take my time, and with several other things going on both online and off-, I needed it. I did not finish all the things he asked me to do (and that, in process, occurred to me to do) until two weeks ago. As per his request, I pinged him that I was ready for the next stage of the review ...

... and waited, and waited. I saw that he had abruptly stopped editing around May 18. I could not figure out why, as there were no long gaps in his edit history since he began contributing a few years ago.

But then I decided to look over his talk page, and that's when I understood why. There was a link to this discussion here about his issues reviewing the noms for the articles on individual Federalist Papers, which then also led to this AN/I thread where a topic ban from GAN and reviewing was seriously discussed, although ultimately not implemented. I sat down to read them; it took a couple of hours last week that I had not planned to spend that way.

Hoo-boy. No wonder EK is gone ... he appears to have slunk off after that one, and I can't say I wouldn't understand. I don't think anyone can honestly say when he'll be back, or even if he will. I wish I had known at that time ... I have lost a month on my timeline.

So obviously I am in the same situation as Bneu2013 is, above, and as they have I would like to request that someone else consider finishing the review. Unlike his situation with 420Traveler (who I should put in a good word for from previous collaborative experience), I think there is no slack left to give ... it's been longer and the reasons for the editor's absence are probably more understandable and even if Ernest resumes editing I'm very doubtful he'll want to return to reviewing GA noms any time soon.

That said ... after reviewing the two discussions linked above I should like to point out a couple of aspects of this review that struck me as a little odd, including one that I think is highly problematic and should be kept in mind should he return, start reviewing noms again and raise similar questions.

  • I was a little surprised at his initial insistence that I should put the account of the script's development in the book article. I can understand someone with no experience (as far as I can tell) reviewing film articles not being familiar with MOS:FILMPRODUCTION. But most heavy readers of Wikipedia, including all of us regular editors (I'm sure), will have read quite a few of our film articles, and where those have been adequately developed, maybe not even to GA level yet, there is almost always some detailed discussion of how the script was developed, especially from a previous medium (novel, play, videogame, whatever): who worked on it even though they might not have been credited, whether the script went in a different direction at one point, why they had to change something from the source etc.

    And even imagining a reviewer who has never read a single film article in their wikilife, it still doesn't make sense, on the face of it, that any history of a film's script development would be located in the article about the source work. How could you reasonably expect to find it in that article rather than the article about the film made from that script, a document specifically created for that purpose. And what of screenplays written directly for the screen? Certainly accounts of their development would not be anywhere but in the article about the film, as they could not rationally go anywhere else? What would be the point of such an arbitrary distinction that would make readers' experience needlessly difficult and confusing?

    To be fair, I'm sure Ernest would have understood this, and yielded on the matter, had he returned, but it just strikes me as the same sort of "huh?" thought process that finds no issue with quickfailing a GA nom or three because the parent article isn't good enough.

  • Some of his suggestions seem motivated more by fandom than editing, and were a little outside the scope of what I usually get from reviewers. I yielded on not treating Father Merrin as a supporting character (even though, to me and I think most viewers of the film, he is, since Karras has the character development he does not) after I reviewed the credits and onesheet and saw how prominently played Max von Sydow's name was ... OK, maybe he's got a point, and I'll humor him ... what's it to me, after all? But I still think it was not necessary to insist that I get a photo of van Sydow in that section (Commons has very few of him from that time period), especially when there's that famous photo of Pierre Teilhard de Chardin that's in there which easily illustrates what Merrin looks like and why von Sydow was cast.

    Frankly, his argument here seems to be reducible to "Father Merrin is not a supporting character because he is played by an internationally respected actor who is now dead" which to me would be akin to saying Obi-Wan Kenobi is not a supporting character in the first Star Wars film because he's played by Alec Guinness.

  • In the same vein, I saw no particularly compelling editorial reason for moving the graf about "Tubular Bells" to the top of the music section just because he always associated it with the film. Of course, I did it anyway because he was reviewing it and it really didn't have too much effect on the article. But frankly the story of the music in The Exorcist (one of many significant aspects of it, historically, as the article explains in the legacy section) is to my mind better told chronologically, in the context of its legendarily troubled production, in which account "Bells" didn't come into the film until near the end of post, and then only by chance (and frankly, there's really not a lot of it in the film, and that should be further qualified with the note that (by design) there isn't a lot of music in the film to begin with).
  • Now, at the end of this list, here's what raises both eyebrows, for me (and I can't imagine I'd be the only one): Shortly after opening the review, Ernest went over to my talk page and asked me if I'd go over to the FAC he'd started for James Madison and look the article over and make some comments. I'm not averse to taking part in those processes, but the article and FAC were pretty well-developed, and with everything else going on in my life at the time, it was ridiculous that I would have the time to do what I always feel such a process deserves—a printout and run-through with a red pen, followed by comments at length (surprise!) in the nom. I do all my own GARs this way, and I know the nominators appreciate it.

    I was able to make a couple of small wording edits, which Ernest told me he appreciated. And the article has subsequently been promoted.

    However, I felt uncomfortable (and still do) making those edits, because they had been solicited by an editor actively reviewing a GAN of mine at the time. Do you think it didn't enter my mind that doing so might well be beneficial to the chances of my own nomination getting reviewed quickly and/or favorably? It's really the equivalent of having the IRS agent auditing your tax return hit you up for a contribution to the United Way while they're doing it (something they are not allowed to do in real life, for so many good reasons). Are you really going to say no?

    I have never had this kind of request before during a GAR. I will say that I don't think Ernest was really looking for a quid pro quo, but his intentions are irrelevant. It's a bad look all around, with the inherent appearance of impropriety.

    Whatever might happen, or not happen, with Ernest ultimately, we need to make it clear in some way, if we haven't already, that reviewers are absolutely not to solicit nominators to review their own nominations for any sort of content recognition (not even DYK, I would say). I seem to recall that some years ago we had an affair where a group of editors palling around on IRC, at one point, agreed to review each other's GANs ... unsurprisingly they were approved within ten minutes of each other, with what seemed to be minimal review. Did it cost someone their adminship, or a chance at one? (Certainly a lot of community trust, though; I don't think they edit much, anymore, if at all).

Good night (NA EDT) and happy editing! Daniel Case (talk) 06:35, 20 June 2023 (UTC)

Replying specifically to your last paragraph: the way FAC is currently set up, you're expected to solicit reviews from others. If you can't get a handful of reviews within a few weeks, the FAC coordinators will automatically fail your article. Naturally, there's quite a lot of editors reviewing each other's nominations at FAC, particularly within the more insular topic areas. Thebiguglyalien (talk) 07:39, 20 June 2023 (UTC)
Yes, but soliciting a review from someone whose article you're reviewing should be avoided, no? Daniel Case (talk) 15:39, 20 June 2023 (UTC)
Addendum: I should point out also that at the time Ernest solicited my eyes on that article, the FAC was well underway, with many reviewers adding their thoughts. This was not something he needed to do to keep the nomination active. Daniel Case (talk) 19:39, 20 June 2023 (UTC)
Whatever Ernest's issues - and I can't say I have never had my own issues with him - it feels deeply unfair to excoriate him and his opinions at such length on a high-traffic talk page when he is not here to defend himself. It is one thing to say that a reviewer has abandoned a review and to ask about next steps. It is another entirely to accuse someone of impropriety when, as Alien notes, trading or soliciting reviews is fairly common in GA/FA spaces. ♠PMC♠ (talk) 08:26, 20 June 2023 (UTC)
He could, as far as we know, return today to defend himself. What I said about him is less harsh than some of the things people said about him earlier. Daniel Case (talk) 15:41, 20 June 2023 (UTC)
Those things were said when he was here, had recently been actively editing, and might still have been expected to participate. Posting this kind of thing when you know he isn't here may be allowed technically, but it certainly does not encourage him to return, and it creates an atmosphere of ABF that may be intimidating to other editors. ♠PMC♠ (talk) 18:09, 20 June 2023 (UTC)
Well, to return to your original comment, can you not allow that I am not without a reason for feeling that this was deeply unfair to me, as well? Currently the GAN table automatically notes when a nominator has apparently fallen off the edge of the encyclopedia ... but not, it seems, when a reviewer does.
It should not have taken a month for me to find out just exactly why this nomination of an article I have put a great deal of work into and have greater plans for had apparently stalled. To find out, and then to have the nomination failed through no fault of my own (ought there not to be some way we can indicate in the ArticleHistory procedural fails as a result of cockups like this?) is really still a bitter taste in the mouth. I mean, what would happen if a review was abandoned because the reviewer got blocked indefinitely? Topic-banned from GAN, as was proposed in this case? Or died in real life? What does GAN owe a nominator, especially one who has been cooperating with the reviewer, in those situations? Surely something so they don't have to find this out for themselves some twilight hour later.
My comments about Ernest notwithstanding, I was actually looking forward to what he had to say. I had done a lot of work, above and beyond what he asked, and I believe we were getting to the point of a favorable outcome. Had the whole issue regarding his quickfailing of those noms not come up here and at AN/I, I would never have been in the position to have made those comments. So please excuse me if I resent being made to feel partly responsible for his departure by many of those who actually were (which, to be fair, was not the intended result, I'm sure) well after the fact. Especially by a community of people intimately involved in a process that initiates no assistance to editors in my situation. Daniel Case (talk) 19:57, 20 June 2023 (UTC)
I am sympathetic to your frustration with Ernest not completing or formally closing the review when he left. As I said, I have had my own issues with Ernest and am not defending his actions in general or at this specific review. That being said, given the scrutiny he was under, I can imagine why he would have simply logged off and never returned without making formal goodbyes.
If you had issues with Ernest's review before he departed, you had plenty of time to request a second opinion. WP:GAN/I#2O gives instructions on how to do so, but I see no indication that you did. There is no central coordination at GAN; no one is responsible for monitoring your nomination but you and the reviewer. The GAN list does note inactive reviewers, you'll see if you Ctrl+F "reviewer inactive". I note that you say you were busy IRL, so perhaps you could not have done any checking yourself, which is unfortunate. Once you did notice that Ernest was unlikely to return, you did the correct thing in posting here seeking assistance. It's what we generally suggest to anyone in any of the unusual (but not unprecedented) cases you mentioned, in fact. My objection is to the manner in which you went about it.
I have never attempted to make you feel "partly responsible for his departure". He was already well gone by the time of this thread. I am trying to get you to you consider that posting a lengthy public criticism of someone who has chosen to exit the project and is not here to defend themselves is not fair. ♠PMC♠ (talk) 23:21, 20 June 2023 (UTC)
If he returns, and based on this thread there is another effort to topic ban him, he will have the opportunity to defend himself against what I said.
OK, I clarify ... you did not attempt to make me feel partly responsible for his departure, which did occur before this thread. You are, instead, attempting to make me feel partly responsible for a decision we have no way of knowing if he has yet taken: whether to return. We cannot say he even knows of this thread. We cannot say what effect it will have on a hypothetical, unknown decision.
And I think that prioritizing the unknown emotional state of editors who have voluntarily left over someone else trying to bring attention to shortcomings, shortcomings that could be improved easily, in this aspect of the project reflects a skewed sense of priorities (Tell me ... if Ernest's inability to respond was the result of being banned or blocked indefinitely, would you still have started this conversation?)
All this notwthstanding, I do thank you for bringing my attention to the "reviewer inactive" comments. But, are those added by a bot or manually? I see one currently says "22 days"; if a bot takes care of that it wasn't doing its job on mine. I gather from the instructions (which, minor correction you linked to the wrong one), which, to be fair, I had not reread in a very long while since I have been nominating articles for GA for a very long while (and never had this happen before) this is done manually. Maybe it might be a good idea to have a bot do this after a certain period has lapsed? And, better yet, it would be even easier for a bot to tell if a reviewer has been blocked for a long period and avoid that problem? Daniel Case (talk) 19:48, 21 June 2023 (UTC)
(Tell me ... if Ernest's inability to respond was the result of being banned or blocked indefinitely, would you still have started this conversation?) Yes! That's called WP:GRAVEDANCING and it is almost universally frowned upon! An editor of your tenure ought to well be aware of that. If you want to bring attention to shortcomings of a process, then do that, don't write an essay about how strongly you disagree with one particular absent person's reviewing style and opinions.
The entire GAN page is bot-maintained; I don't recall what the precise activity criteria are. I linked to the correct section, thanks, I was linking you to the instructions for getting a second opinion to remind you that you could have done that. By all means make a suggestion somewhere for a bot to make notifications about inactive GANs, that sounds like it might be useful. ♠PMC♠ (talk) 20:03, 21 June 2023 (UTC)
FYI, the bot marks a reviewer as inactive when they have not edited for more than 21 days. In this case that was on 9 June, and the bot added the notification in this diff. That doesn't serve as a notification to the nominator, of course; they would have to notice it, and there's no reason they should look at the GAN page again. It was intended mostly to alert GAN regulars so if anyone wanted to investigate they could, but we have no GA coordinator so it is nobody's responsibility to do so. I could change the bot to tell the nominator that their reviewer is inactive if there's a consensus that's a useful thing to do. Mike Christie (talk - contribs - library) 22:46, 21 June 2023 (UTC)
I understand (and somewhat share) the conflict of interest concern regarding 'trading reviews', but this is a tolerated practice at these (GAN/FAC) venues provided it is done with a neutrally worded message. The review has been closed as abandoned and returned to the nomination pool by Buidhe. I think that is all that is necessary here. Mr rnddude (talk) 15:31, 20 June 2023 (UTC)
That's fine. I can live with that, with the caveat that if it has not been reviewed by September I will withdraw it and take it straight to FAC.
And I am rather dismayed that this should be a tolerated practice. I have no problem with soliicting reviews from others generally ... recently I was solicited by the nominator of British logistics in the Western Allied invasion of Germany as a result of having protected British logistics in the Normandy campaign while it was on the Main Page. That's fine, ethically ... apparently he pinged everyone who edited the FA while it was on the Main Page; I have no problem with that. I might not necessarily have a problem with two people deciding to review each other's active FACs; perhaps some disclosure would be required.
But GAN is a different boat because it's usually one-on-one. The reviewer holds all the power ... yes, the nominator can always request a different reviewer, but a lot of people would just not want to be seen as difficult. Or they have timeframes they are working within (as I do).
If the issue is a shortage of reviewers, I understand, but I have always addressed this myself by reviewing one GAN for every nomination of my own. As I will do this time. Daniel Case (talk) 15:56, 20 June 2023 (UTC)
The quid-pro-quo review-and-nomination rule seems to work over at DYK. I wonder why it doesn't work at GAN? SilverTiger12 (talk) 18:28, 20 June 2023 (UTC)
DYK requires you to review one statement (the proposed hook) within an article and its corresponding source(s), while GAN requires a review of a whole article (typically thousands of words) with an accompanying spotcheck of a selection of sources. The workload is a magnitude (minimum) higher for GAN than DYK. There is substantially more work involved in writing a GA than a DYK new article as well. In brief, it is a much greater demand than many editors will support. Mr rnddude (talk) 18:55, 20 June 2023 (UTC)
Well, that's not everything a DYK reviewer has to review ... there's also the article's history (to make sure it was created, moved to mainspace or promoted to GA in the 7 days before nomination) its length, and (very strongly advised) a copyvio check.
But generally the DYK criteria are more objective, and do not require that you pass a subjective judgement on article quality (though reviewers can, at their discretion, invoke that as a reason for failure if it's bad enough; I have done that on a couple of occasions at DYK when the article's English was so atrocious it was not worth the potential embarrassment of having it linked from the Main Page. Yes, evaluating the hook is a bit more subjective ... I have lost count of the amount of times I've pointed out to a nominator that their hook was dull, sometimes requiring some explanation on my part. I had this idea once for April Fool's Day that instead of doing what we usually do (making up clickbait, deliberately misleadingly worded hooks) we should change the name of the section to "Do You Really Care ..." for a day and filling it with hooks built on obvious, unsurprising information ("... that the XXXX River will make your hands wet if you touch it?")—thankfully, perhaps, we have never done that (Yet).
All the same, in years of nominating and reviewing DYKs, I have never had a reviewer ask if I would review their nom while they were still reviewing mine. Nor do I recall such a situation with other reviewers. I have not yet completely gotten over my shock of this morning (my time) which appears to me to be based on a misunderstanding. So I will reiterate yet again that what disturbed me about what Ernest did was not the QPQ aspect of it, it was that he did it while his review of my nomination was pending.
It's important to note also that reviewers who do QPQs at DYK are directed to the queue, where they can easily find an unreviewed nom; this has done wonders for clearing the backlog and, indirectly, discouraging backscratching. Also, reviewers are generally encouraged to do their QPQ before nominating their own article (once they reach the point where they are required to submit one) Yet here there is no such process to encourage QPQ reviewing, much less indirectly regulate it. A strange thing, giving the increased workload noted above. Daniel Case (talk) 19:35, 20 June 2023 (UTC)
An explicit QPQ has been discussed here many times. One reason that it gets rejected is that it seems likely to lead to substandard reviews. On the other hand, the reason that the GAs and reviews for each nominator are listed on the GAN page is to let reviewers choose to review nominators who are active reviewers. It encourages reviewing, without making it an explicit requirement. It's not clear how much beneficial effect that change has had (the review counts have been listed since the end of last year) but there are certainly reviewers who are using it to select what to review, so it is a little like QPQ in that way. Mike Christie (talk - contribs - library) 20:17, 20 June 2023 (UTC)
I have never had a reviewer ask if I would review their nom while they were still reviewing mine - That you had not previously had this experience does not mean that anything untoward has occurred, and I really wish you would stop implying that it has. Several people have now chimed in to say that these kinds of solicitations and exchanges are not particularly unusual in the GA/FA areas. Unless there is rubberstamping or extortion going on (neither of which appears to have occurred here), asking someone else to review your content is not unethical. If one doesn't wish to, one can simply say no or ignore the request. ♠PMC♠ (talk) 22:45, 20 June 2023 (UTC)
You are still missing my point. I said above, several times, that I do not have a problem with QPQ reviews per se—I have always reviewed one GAN for every nomination I've made (in fact, due to having done more years ago when those reviews were less stringent, and some noms that failed, I have currently twice as many reviews as GAs), and my comments about the QPQ system at DYK (which as I've noted is structured to largely avoid this being an issue) should make clear that I have no problem with it there.
"That you had not previously had this experience does not mean that anything untoward has occurred" comes across as almost willfully obtuse. The point is that I should not have been asked to participate in an FAC started by the reviewer then considering my GAN because it doesn't look good. It facilitates reasonable doubt about the integrity of the GAR process even if it ultimately were to be found that nothing untoward took place—in fact, that doubt would understandably extend to . This sort of "if there wasn't anything obviously wrong, don't worry about it" thinking fuels low social trust, which is what has gotten the Western world into the current political state we're in. I am sort of resigned to it for the time being from the current Chief Justice of the United States, among many others, but not here.
I mean, imagine if Ernest's FAC hadn't gotten any !votes when he made that ask. Suppose I had gone over and, taking what I believed to be his hint, cast a "support" based on a very superficial, vague analysis, the FA equivalent of "it's got a good beat and you can dance to it" ... after which, within a few hours, Ernest had then promoted my nominated article regardless of what suggestions of his I'd implemented. How would that look? And would anyone have looked? But "looked" is all it could have been because there would have been no evidence in the record of any explicit exchange.
If indifference to this continues, I can assure you that one day there will be another scandal over something like this that blithe nonchalance will be nowhere near sufficient to deal with. And there will be questions asked as to why no one had a problem. And if I am around and I know about, I will not hesitate to link this discussion from wherever it is discussed.
We can and should do better. Daniel Case (talk) 19:22, 21 June 2023 (UTC)
Addendum: "If one doesn't wish to, one can simply say no or ignore the request." Bless your heart. I can only conclude from this you've never been in a situation where some sort of facially innocuous request has been made of you by someone with the power to give or deny you something you want. The understanding is always clear; nothing need be spoken or made implicit, and it never is. If I am correct, you should count yourself fortunate, and accept this as a learning experience on how condescending this comes across to anyone whose lived experience on this is a little different. Daniel Case (talk) 19:55, 21 June 2023 (UTC)
It is so cute that you are calling me condescending in a post where you go to the trouble of linking "bless your heart". It's disappointing to see an editor with your tenure gripping firmly to the assumption that others must be acting in bad faith, as opposed to acting in a spirit of collaboration in onwiki processes. I'm sorry that your life experiences have left you so cynical. ♠PMC♠ (talk) 20:07, 21 June 2023 (UTC)
I've never been one to mince words, so I'll say exactly what I'm feeling - what the hell was the point of any of this? It is truly ironic that you whine about "condescending" behavior while engaging in it yourself. Were you here just to complain that your review was abandoned? If so, we could have done with about 2,000 fewer words from you. As the one who started the ANI thread, I actually think you're being unfair to ErnestKrause and you shouldn't have written much of what you did here. All you had to say was "the reviewer appears to have abandoned my nomination, would another editor be willing to take over?" If you have an issue with de facto QPQ, well either get used to it or expect nominations to have to wait far longer than they do now (easily 6+ months). Trainsandotherthings (talk) 23:11, 21 June 2023 (UTC)
  • @Daniel Case: - Thanks for bringing this up. It has been nearly a month since this reviewer has made any edits; their second to last edit was starting my review. While I appreciate their willingness to review my nominations, I think it is important to not rush the process, but also to review in a timely manner. If someone else is willing to pick up this review, I would be okay with that. On a related note, The Exorcist is one of my all time favorite films, and I do agree that this would be an extremely valuable article to get to FA status by its 50th anniversary. So whenever you nominate, I'd be more than willing to take a look at the review. Bneu2013 (talk) 18:32, 21 June 2023 (UTC)

GA review exceptions June 4–10

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

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)

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)

Women in Green Editathon conclusion

WP:Women in Green just completed its fourth editathon to write and review good article nominations about women and women's works. In one month, members of the project nominated 24 articles about women and took on 42 GA reviews. Among these reviews was the entire backlog of GA nominations about women and women's works from September through January, which has now been cleared. We should consider facilitating more organized efforts like this, as they're clearly helpful in improving article quality and promoting them to good article status. Thebiguglyalien (talk) 01:48, 1 July 2023 (UTC)

Also clearly helpful in reducing backlog despite the new nominations. —David Eppstein (talk) 02:47, 1 July 2023 (UTC)

Template-protected edit request on 2 July 2023

Change criterion 2(b) 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 that could reasonably be challenged, 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).

Per RfC consensus. BilledMammal (talk) 07:58, 2 July 2023 (UTC)

Done, with the addition of a link to WP:LIKELY for "could reasonably be challenged". See next section. Mike Christie (talk - contribs - library) 13:28, 2 July 2023 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Good_article_nominations/Archive_28&oldid=1164617746"