Jump to content

Wikipedia talk:Good article nominations

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
MainCriteriaInstructionsNominationsFAQDiscussionReassessmentReport
Reviewing initiatives:Backlog driveMentorshipReview circlesPledges
Good article nominations
Good article nominations
Shortcut

This is the discussion page for good article nominations (GAN) and the good articles process in general. To ask a question or start a discussion about the good article nomination process, click the Add topic link above. Please check and see if your question may already be answered; click the link to the FAQ above or search the archives below. If you are here to discuss concerns with a specific review, please consider discussing things with the reviewer first before posting here.

GAR one-month minimum: stats analysis and proposal to reduce

[edit]
Introduction

On January 15, 2025, GAR rules were ammended so that GARs were open for a minimum of one month. Advocates for the change wanted more time for editors to volunteer to fix the article. Detractors felt that reviews didn’t need to be kept open for that long, and reviews without a response would clutter the GAR page. Since this change has been implemented for 10 months, I want to find out what happened when this change was made, and to find out if keeping GARs open longer resulted in more articles being kept. My bias before starting this study was that the one-month minimum was too long and should be reduced. I also believe that during the time of this study, I nominated the greatest number of articles at GAR, although I have not quantitatively proven this.

Methodology

Using the GAR archives, I recorded the GARs that resulted in "kept" that were created after January 15 (when the one-month change took place). I then recorded when the first editor posted a comment that indicated they wanted to work on the article. This metric was selected because GARs will not be closed if someone indicates they want to work on the article. GARs will stay open while work is ongoing, even past the one-month minimum. I then recorded the number of days difference between the review being opened and the first indication comment.

Results

184 GARs opened between January 15 and November 15 were closed as "kept". One article was closed as "kept" without additional comments (Evansville tornado outbreak of November 2005) and was not included in these results. This leaves 183 kept GARs.

This chart displays the time between the creation of the review and the first comment that indicated that someone wants to work on this:

Time distance Number of articles (cumulative) % of articles that received an indication (cumulative)
Less than one day 95 51.9
One day 108 59
Two days 124 67.8
One week 150 82
Two weeks 165 90.1
Three weeks 171 93.4
One month (30 days) 177 96.7
Older than one month 183 100
Analysis

Over 50% of articles that are eventually declared "keep" receive the first indication within 24 hours. By two days, over two-thirds of the articles had an indication posted. By two weeks, over nine-tenths had an indication. Seven articles had the first indication after the one-month minimum and would have been delisted if closed at the minimum. The GAR with the furthest distance was Star Wars Jedi Knight: Jedi Academy at 40 days.

My proposal

Change the minimum amount of time a GAR can be open before delisting from one month to two weeks.

Rationale
  • When many GARs are open at the same time, it becomes difficult for editors to navigate and find articles they are interested in.
  • GAR nominators avoid nominating similar articles from the same topic at the same time: by closing GARs sooner, other articles from a similar topic can be nominated sooner.
  • The minimum amount of time an article remains open at FAR without comment is four weeks (28 days). At one month (28-31 days), GAR’s process to delist an article is longer than FAR, even though it is supposed to be a lower-stakes assessment.
  • After two weeks, the likelihood that an editor will indicate that they want to work on the article drops substantially.

I hope this was interesting information, and I look forward to feedback. Z1720 (talk) 20:38, 15 November 2025 (UTC)[reply]

Comments
I think I would support lowering the no-comment close threshold to no lower than two weeks so longer as the closer of the nomination is a different editor than the one who opened the GAR, to at least guaranty two sets of eyes on the article. (I think an editor who is closing a GAR, especially a low-participation one, needs to check the article history to make sure improvements are not occurring without mention of such at the GAR page and scan the article to verify that the problems alleged in the GAR are actually present). GA status is much more easy-come-easy-go than FAC. Since GAN operates on effectively a two-editor consensus (nominator + reviewer), then I see no reason why GAR should not operate on a two-editor consensus model as well. As to timing, I would certainly fail a GAN where significant work was needed and the nominator had made no reply for multiple weeks unless there were extentuating circumstances. We do need to be careful with how many GARs of a topic are listed at once, or in short succession after GARs closed as keep. Most of the GARs I have been involved with making the corrections to the articles in required substantial work, which can wear out an editor group. Hog Farm Talk 00:07, 16 November 2025 (UTC)[reply]
support per HF above. ~~ AirshipJungleman29 (talk) 00:18, 16 November 2025 (UTC)[reply]
Support very well reasoned proposal. I also think HF's thoughts are very reasonable. IAWW (talk) 00:39, 16 November 2025 (UTC)[reply]
Support The logic of the proposal checks out. I would also not object, based on the data provided above, to lowering it back to one week since an additional week beyond the first only leads to a 10% increase in the number of articles receiving input. If it is indeed the case that open GARs are rate-limiting when it comes to new GARs being opened, processing GARs that do not lead to significant article improvement more quickly means that we get to the GARs that do lead to such improvement in shorter order; in other words, the process becomes more productive from a perspective of article improvement. TompaDompa (talk) 01:01, 16 November 2025 (UTC)[reply]
Comments It would depend on what the value of a kept GA article is compared to the value of keeping the clutter down, and the relative amount of work in resubmittong for GA against fixing an article under review. At a rough guess, there should be little difference, but do we have any statistics about how many failed articles are resubmitted and how long it takes for that to happen, and what the success rate it? I would guess that we have no useful data on any of these considerations, but if we do, we should consider them. Do we even have any idea of the value of a GA?· · · Peter Southwood (talk): 06:11, 16 November 2025 (UTC)[reply]
Support a reduction to a 2 week minimum time before closure. I was opposed to the 1 month minimum waiting time from the beginning - my comment at the time was We're assessing articles, not selling guns. Forcing a 30 day hold no matter what is foolish extra bureaucracy. Searching through Archive 33 shows multiple threads lamenting the massive GAR backlog that was caused by this change. I've appealed to the community to get more involved in GAR but there hasn't been a significant difference in engagement. Most nominations end up being simple delists with little or no activity besides the initial nomination. I do agree with Hog Farm about the pace of nominations being potentially too fast. It doesn't take long to determine an article merits reassessment, but it's almost always far more work to address the issues which merited the assessment. As someone who closes GARs, I am perfectly happy to wait far longer than 2 weeks if there is work being done on the article or someone needs time. That's not what happens in most cases, as the data provided by Z1720 proves. I think two weeks is more than enough time for an interested editor to indicate their intent to work on improving an article listed at GAR. Trainsandotherthings (talk) 22:25, 17 November 2025 (UTC)[reply]
  • Oppose GARs are a "break glass in case of emergency" nuclear option, yet I've seen them being handed out like candy for trivial issues such as single unsourced sentences. The last thing we want to do is encourage more GARs by making them delist sooner rather than encouraging fixing the article oneself per WP:BEBOLD and WP:SOFIXIT. A month is a suitably major length of time for a major action. ᴢxᴄᴠʙɴᴍ (ᴛ) 02:34, 19 November 2025 (UTC)[reply]
  • Support, I think it's clear from the thorough statistical analysis presented above that the extra time does not provide a commensurate benefit in terms of retaining GAs/closing GARs as keep. ♠PMC♠ (talk) 02:36, 19 November 2025 (UTC)[reply]
  • Support, with HogFarm's so longer as the closer of the nomination is a different editor than the one who opened the GAR caveat. -- asilvering (talk) 01:46, 22 November 2025 (UTC)[reply]
  • Support with HogFarm's caveat. Thanks for this wonderfully detailed write-up, Z1720! ViridianPenguin🐧 (💬) 02:28, 22 November 2025 (UTC)[reply]
  • Comment: About the "closer is different from nominator" caveat: right now I don't close GARs that I opened or commented on, as I want a second opinion. Adding HF's stipulation wouldn't change my practices, and I think it would be a positive aspect to codify into the rules. If put in place, I would appreciate it if editors interested in the GAR process would consider patrolling the GARs so that they aren't stalled waiting for a closer or that they aren't being closed/checked by a small group of editors. Z1720 (talk) 04:09, 22 November 2025 (UTC)[reply]
    This is a good point, and appears to be a good practice, but it also produces situations where an editor responds, fixes all the issues to the extent that they understand them, and leaves the article in what they consider satisfactory condition, after which nothing happens. The nominator should at least provide feedback on whether the issues they identified have been rectified, so the responding editor has some confidence that the work is done (or not, as the case may be), and an independent closer knows what to look for.
    To facilitate this process, the responding editor/s should preferably leave a message indicating that they think they have fixed the listed issues.
    A checklist template for this could be useful. Also maybe a maintenance category for unclosed GARs that have been reported as fixed.
    Also, if both nominator and responding editor/s agree that the issues have all been fixed, I see no reason why the nominator should not also close. An additional opinion is only really necessary if there is uncertainty or disagreement. Cheers, · · · Peter Southwood (talk): 04:59, 22 November 2025 (UTC)[reply]
  • @Pbsouthwood: I can only speak for myself, but I try to follow-up on articles after concerns are addressed. Sometimes there is a long wait because I am not pinged, so I do not know that editors are waiting for me to comment. A checklist template might be difficult, as GARs aren't a full GAN review (for example, I rarely check image licences in a GAR). Also, if there is only one concern in the article (usually uncited statements) then a checklist is probably not necessary. There are also some GARs I open where other editors post additional concerns, so I want this process to give space for those editors to also comment, rather than sending articles to GAR repeatedly. Z1720 (talk) 14:54, 22 November 2025 (UTC)[reply]
    Z1720: FYI, I just closed a bunch more GARs last night. Bgsu98 (Talk) 14:57, 22 November 2025 (UTC)[reply]
    All fair comment, but a template providing a checklist can have non-relevant items either deleted or checked as complete (I prefer the former) and it is a good check that the criteria of the review actually match the GA criteria and are not arbitrary or scope creep.
    Not being pinged can happen, and sometimes seems to happen even when a reply should automatically generate a ping. One doesn't want to appear rude or impatient - other things exist, but sometimes when nothing happens for long enough, everyone forgets and nothing continues to happen. Cheers, · · · Peter Southwood (talk): 15:57, 22 November 2025 (UTC)[reply]
  • Support: the extra time beyond two weeks does not seem to be particularly useful. Thanks to Z1720 for the helpful statistics! Phlsph7 (talk) 09:28, 22 November 2025 (UTC)[reply]
  • Comment This is moreso aimed at Z1720 but would this include cases where an editor has shown interest in fixing the article but has not been able to have time to do so in the two weeks? for example with Cancer pain I do intend to do what I can to fix your concerns but I'm tied up with improving Coeliac disease which I started improving a couple weeks ago. Basically if I was to say "hey I don't have the time right now for this but if you could keep this open for a month that would allow me to improve thigs, if I haven't made any edits in that month then feel free to close it" would that mean I could still have a full month to find the time to improve it? IntentionallyDense (Contribs) 07:24, 23 November 2025 (UTC)[reply]
    Yes, even when the time period was one week, we would often leave articles open for months if there was hope they could be improved. —Femke 🐦 (talk) 11:14, 23 November 2025 (UTC)[reply]

Amount of articles per section: a discussion

[edit]

A positive aspect of increasing the number of GAs is that some section headings have lots of articles. Last year, I opened a discussion about how many articles should be in each section before editors consider splitting them. One consensus of the discussion (in which only a few editors expressed opinions) was that a section could be split if there are over 200 articles, but only if there is a logical place to split it.

In the video games GA section, I did a split of video games released in 1995-1999 into 1995-1997 and 1998-1999. I split that heading first because it was the first one chronologically to have over 200 articles. It was good-faith reverted by PresN: in the edit summary, the stated reason was that other video game sections were larger, that the categories should have consisten ranges of time (though noted that Music's categories are not) and that the two year range for a category is too small. Before conducting more splits, I wanted to get more opinions from the GA community.

Questions: When should editors consider splitting GA headings? Should the video games section headings be consistent (every five years) or should different time ranges be used to keep the number of articles below an arbitrary number?

Looking forward to reading your thoughts. Z1720 (talk) 14:50, 17 November 2025 (UTC)[reply]

If we do go with a cap on section size, I would prefer aesthetically not to go with the alternating 2-year/3-year setup, but maybe 3-year groupings starting at 1990 through 2019. I don't love tiny year ranges, but I do understand that there's not a lot of great options if we don't want sections to have more than 200 articles. --PresN 15:53, 17 November 2025 (UTC)[reply]
  • I would be fine with 3-year sections: it's a bit of a pain to set up (splitting a section into 2-and-3-year sections is easier than splitting three sections into five) so if implemented it would have to be done in one series of edits. Z1720 (talk) 16:15, 17 November 2025 (UTC)[reply]
When the title disappears off the top of the screen (on standard Vector2022 width) that is not ideal for easily remembering where you are in the sea of blue. CMD (talk) 02:49, 18 November 2025 (UTC)[reply]
I've been privately maintaining a variation of the section dividing the decades into early, middle and late segments (visible here), my reasoning being that thirds of a decade have clearer specific cultures than whole ones; nobody's going to mistake the late 90s for the early 90s, for example. If we were to do further splitting, I would suggest a format similar to what I'm using, though my method might come off as somewhat persnickety to some, since the divisions occur within a part of a year rather than between years (the divide between early and mid occurs at April/May XXX3, and the divide between mid and late occurs at August/September XXX6). A simpler and slightly more agreeable alternative might be to go by a "3-4-3" format for each decade that's a bit more clear cut, an example being 1990-92, 93-96, and 97-99. I have no strong feelings whether or not we go through with any split at all, but I figured suggestions for that route would be helpful. Cat's Tuxedo (talk) 01:42, 18 November 2025 (UTC)[reply]
I have no qualms with irregular time ranges, but of course if someone is willing to do the work of reformatting to implement a shift to 3-year groupings, I would prefer that for aesthetics. If we go with the latter approach, I agree with PresN's pitch to do ten 3-year groupings covering 1990 through 2019. ViridianPenguin🐧 (💬) 02:48, 22 November 2025 (UTC)[reply]
Ideally I'd prefer if the video game sections were split by console generations, although that may not be a very well defined term. If we end up splitting them by smaller time frames, one question I'd ask is how early access games (more specifically, fully released games that were formerly in EA) would be handled? Would we want to use the date the game enters EA, or the date the game fully releases? What would happen if the two dates were in different time frames? Gramix13 (talk) 05:54, 27 November 2025 (UTC)[reply]

Talk:D4vd/GA1

[edit]

The singer is currently involved in an investigation, and due to the nature of said investigation it's impossible for the article to remain stable. Changes in the investigation, and the attention it brings, will mean the article may change wildly. Thus I believe the GA review should be closed and delayed until this matter is sorted out. Harizotoh9 (talk) 01:51, 23 November 2025 (UTC)[reply]

The GA criteria specifies that "Stability is based on the article's current state, not any potential for instability in the future." The D4vd article as it currently stands is not unstable, given that there isn't an edit war or content dispute going on regarding his involvement in the Celeste incident. Rather, there appears to be a lot of normal, constructive editing towards keeping the article up-to-date, and the details aren't changing so dramatically that it would be a moot point to perform a review (see WP:GANOT). Thus, I wouldn't support closing the review on the sole basis of stability. Leafy46 (talk) 06:33, 23 November 2025 (UTC)[reply]
The article would have benefitted from a more rigorous shaking‑loose of minor issues before being brought to GAN. Wikilinking insignificant words like teenager, montages, Singapore, California, anonymous sources, album, maturity, and impounded are always best avoided, as is the practice of Wikilinking a word in one paragraph of an article to another paragraph in the exact same article — all issues found in the D4vd article. Errors like this are demonstrative of a lapse of focus, and runs the risk of draining a good reviewer's energy. I don't condone what appears to be the other reviewer's abandonment of this review, but then again, there's at least a residue of argument for not blaming them for it. (Still, clicking "begin review" is a contract that ought to be honored rain or shine.) The nominator should feel free to renominate this article at their earliest convenience (ideally, after the above mentioned issues are addressed).  Spintendo  12:10, 23 November 2025 (UTC)[reply]
This isn't the first time @Veko:'s disappeared mid-review. He did this at Talk:Niggas in Paris/GA1, which I ultimately reviewed.--Launchballer 13:09, 23 November 2025 (UTC)[reply]
Well, that's for another reviewer to deal with. It very well could be unsuccessful even with a new reviewer, but that's the point, it should have a reviewer to go through such issues and decide if the article has met the criteria. You don't quick-fail reviews for wikilinking too much - the whole point of the review is to bring issues like this up so the nominator can resolve them. It's unfair to have that permanent stain on the article history for something that is not the nominator's fault. jolielover♥talk 13:12, 23 November 2025 (UTC)[reply]
I meant to reply to @Spintendo: but it somehow ended up like this jolielover♥talk 13:13, 23 November 2025 (UTC)[reply]
Stability is one of the least important aspects of a GA in my experience, and it's more so concerning edit wars than new information popping up. Most articles on BLPs are going to have new information come up after their review, that's why you can check the permanent revision the review was conducted on. I recently had a GA pass in the midst of his album announcement (which was not intended), and obviously, new information to the page, but you can make it work. D4vd's article has been, for the most part, not too shabby in terms of fast changing information and edit warring. jolielover♥talk 13:16, 23 November 2025 (UTC)[reply]
I agree. What's happening with the article now is normal editing. This should be reopened.--Launchballer 13:40, 23 November 2025 (UTC)[reply]
Agreed. I do think we should readdress whether it should be a part of the GA Criteria. In my eyes, it basically suggests that there isn't an edit war going on, not that it is stable as in "complete". Lee Vilenski (talk • contribs) 14:00, 23 November 2025 (UTC)[reply]
@Jolielover: is absolutely right, over-Wikilinking should never form the basis of a quikfail. I have to admit that overlinking-issues in general form one of the first things I look for in gauging an article's robustness, so when I see a nominator who has invested an admittedly generous 192 edits into an article and yet still overlooks the overlinks, it naturally places me in the position of being unfairly biased against them. The GAN process is, as Jolielover rightly mentions, yet another opportunity for nominators to improve their issue-identifying acumen. At the end of the day, I think everyone agrees the more untenable position is that of the reviewer who promised a review that ultimately wasn't delivered (and who appears to have done this before). Thumbs down icon for increasing everyone's workload. If it were me, my contrition would be to offer 2 additional reviews as a penance.  Spintendo  19:56, 23 November 2025 (UTC)[reply]
Yup, I pretty much said the same above and I stand by that perspective. Leafy46 (talk) 23:53, 23 November 2025 (UTC)[reply]
Stability is more than just edit wars and content disputes; an upcoming, unreleased film is generally held to fail stability because in not very long it will be substantially different from what is reviewed. That isn't too say that this situation is so analogous that it should be failed on stability, but I personally would put off taking up a review until the dust settles in a few months. Rollinginhisgrave (talk | edits) 20:31, 23 November 2025 (UTC)[reply]

Oh and perfect example just came up. People just released a news article citing sources which contradicts information from news reports only a day or two ago which means the need for re-writes information added in the last few days. The recent reports are ultimately cited to TMZ, and some or all of it may be incorrect. Sources can't even agree upon Rivas' age (whether she was 14 or 15), and basic other details of the investigation. Stability is not possible under these circumstances since investigators have said very little about the case, and sources are sometimes in contradiction with each other. Harizotoh9 (talk) 08:47, 25 November 2025 (UTC)[reply]

Neither of People or TMZ are reliable for contentious claims about living or recently deceased persons. See also WP:RSBREAKING.--Launchballer 11:14, 25 November 2025 (UTC)[reply]
TMZ is not sourced once (I in fact phased out the source entirely). The citations were to KTLA, who have since posted a new article correcting their past article. The highest quality and most cautious sources for this article are People, ABC, NBC, AP, CBS and LATimes. . News sources are relying upon sources because officials have been very quiet and released very little official information. That will eventually change and investigator and medical examiner findings will be released, which will lead to heavy expansion of the investigation section. None of this is stable and it's expansion, re-writes piled upon re-writes. Harizotoh9 (talk) 19:04, 25 November 2025 (UTC)[reply]
I agree that the article cannot be inherently stable right now. Bgsu98 (Talk) 19:18, 25 November 2025 (UTC)[reply]

Need AfC review before GAN can be started

[edit]

Would anyone here be willing to do an AFC for Draft:Eric Gilbertson (climber) so I can nominate it for GAN? TIA (t · c) buidhe 23:09, 23 November 2025 (UTC)[reply]

Instead of waiting you appear to have promoted your own draft yourself. Perhaps you could explain what is new since Wikipedia:Articles for deletion/Eric Gilbertson (climber)? Especially re "The majority of sources constitute self-published material on the subject's blog or secondary sources based significantly on said blog"? —David Eppstein (talk) 02:14, 26 November 2025 (UTC)[reply]

Talk:Pinnacle Airlines Flight 3701/GA1

[edit]

This article was recently quickly failed at GAN with the claim that I "fundamental lack of understanding of the subject matter", which was reaffirmed at a talk page discussion where the reviewer flat out said that they "can only conclude that [I] do not understand" the topic. Now I contest this quick fail for obvious reasons (I was the one who wrote the article and have good understanding of what the sources say), but I am not here for that since I have an open peer review. What I am more concerned about is that if the same standard of review keeps up (where the review is quick failed because a perceived lack of knowledge/the problem with the article is really "I don't understand what I am writing about"), this article may never become a good article without a significant rewrite. I would like some other opinions on this; how valid was the quick fail justification and what major problems does the article really have that should be addressed by me before a future GAN? RandomInfinity17 (talk - contributions) 02:08, 26 November 2025 (UTC)[reply]

I don't claim to understand the technical terminology in play, but it sounds like Roy does, and I trust that he wouldn't QF on those grounds otherwise. It is not unreasonable for a reviewer to QF an article that contains numerous errors. Nobody needs to be an expert to write a Wikipedia article, but you do need to have sufficient understanding, especially in technical fields with lots of jargon and doubly so if you're going to take something to GAN. ♠PMC♠ (talk) 06:00, 26 November 2025 (UTC)[reply]

Suggestion for easing the backlog of reassessments

[edit]

Looking at the GAR page, it seems to me like a large number of articles listed there are being reassessed because of issues that were present at the time of the review. (I was specifically motivated to write this by Bury F.C., but it's far from the only one). It seems silly to me to list articles here with the expectation that someone will work on them to bring them back to their former glory, if they were never good in the first place. If I want a mediocre article worked on, should I get a friend to pass it as a GA and then list it here for reassessment in the hope someone will work on it?

Instead I propose some sort of mechanism to annul problematic reviews that didn't address an obvious issue that existed at the time of the review, to make it as if the article was never good in the first place (which it wasn't). It's much easier to take a glance at the reviewed versions of recent GAs to check whether there are any issues the reviewer hasn't addressed than find time to improve the article. Additionally, the nominator is not going to complain about an incorrectly passed GA as they would for an incorrectly failed one, so a mechanism for outside observers to object might be good.

Thoughts? Is there some reason I've overlooked why this is a terrible idea?

(Courtesy ping to Z1720 as the most active editor in this area.) JustARandomSquid (talk) 11:26, 27 November 2025 (UTC)[reply]

Request to remove stuck/malformed GA entry: Hushmand Dehqan

[edit]

Hello admins. Could you please help remove a stuck and malformed entry from the current Good Article nominations list?

I have fixed the nomination on the article's talk page (`Talk:Hushmand Dehqan`), but the bot is unable to process the correct nomination because of a faulty, stuck entry in the main list.

The faulty entry to be removed is:

`Hushmand Dehqan (talk | history | start review) (0 reviews, 999 GAs) Example (talk) 07:17, 29 November 2025 (UTC)`

The error persists because of the inclusion of the placeholder `Example` and the bot's inability to create the review page. Removing this line should allow the bot to process the nomination correctly from the talk page.Mojgoon (talk) 08:29, 29 November 2025 (UTC)[reply]

I have removed the malformed template from the article talk page. In the future when nominating Mojgoon, please follow the instructions at WP:GANI. Thanks, ~~ AirshipJungleman29 (talk) 11:39, 29 November 2025 (UTC)[reply]
Thank you for your help in removing the malformed template from the Talk page of Hushmand Dehqan. Mojgoon (talk) 12:06, 29 November 2025 (UTC)[reply]
Why have you reinstated the malformed template Mojgoon?? As I said above, please follow the instructions at WP:GANI. ~~ AirshipJungleman29 (talk) 12:18, 29 November 2025 (UTC)[reply]
Hello,
I apologize for bothering you. I tried several times to renominate my article for GA status, but it did not work. The WP:GANI page did not help me either. Could you please assist me? Thank you.
Best regards, Mojgoon (talk) 12:40, 29 November 2025 (UTC)[reply]
Could you please write the correct code for the GA2 renomination right here for me? Thank you again for your patience. Mojgoon (talk) 12:46, 29 November 2025 (UTC)[reply]
As outlined clearly at WP:GAN/I#N2 Mojgoon, please copy {{subst:GAN|subtopic=Philosophy and religion}} and paste it at the top of the relevant talkpage. ~~ AirshipJungleman29 (talk) 12:47, 29 November 2025 (UTC)[reply]
Thank you. It's done. Mojgoon (talk) 13:45, 29 November 2025 (UTC)[reply]
In the future you could use this script to automatically do everything properly IAWW (talk) 15:22, 29 November 2025 (UTC)[reply]

SWEEPS progress

[edit]

In 2023, SWEEPS2023 was established as a collaborative effort to evaluate good articles promoted in 2009 or earlier that might have unreferenced text. It started with 380 articles and is now below 25. Thank you to everyone who participated in identifying or fixing up these articles, and we appreciate any help with evaluating the last articles.

As a reward, we have given SWEEPS more work! The next iteration of the effort, SWEEPS2025, has been set up to evaluate good articles promoted between 2010 and 2016 that might have uncited text. The project is starting with 214 articles from a variety of topics. The effort is collaborative, so anyone can get involved with evaluating and fixing up articles on this list. Feel free to post below or on the project's talk page if there are any questions. Z1720 (talk) 15:58, 1 December 2025 (UTC)[reply]

Discussions on AI fact-checking quality articles

[edit]

The Signpost has published an article that they have titled "AI finds errors in 90% of Wikipedia's best articles", which has brought about two conversations:

Though the piece is framed around featured articles, the use of AI in supplementing spot-checking pertains to GA as well. Rollinginhisgrave (talk | edits) 03:46, 2 December 2025 (UTC)[reply]

We should be careful not to conflate this with spot-checking. Throwing the article into an AI is presumably a look at the entire article at whatever level of detail the AI uses, whole article rather than sampling. Further it is specifically looking for (supposed) factual errors rather than a more holistic check on source use. A useful tool but not a 1:1 replacement. CMD (talk) 06:29, 2 December 2025 (UTC)[reply]
I agree. AI can be really good for quick flagging TSI issues too through. Obviously only for flagging to be followed by human review. IAWW (talk) 12:10, 2 December 2025 (UTC)[reply]
That our articles contain some errors means that there is always something left to improve, always something for someone to come and fix, and always something else to discover. Fallible beings can't create perfect articles. What we can do is create living documents that are always growing, always improving, and always changing. Yours, &c. RGloucester — ☎ 12:18, 2 December 2025 (UTC)[reply]
I would love to work on a mini-project that goes through all GAs and FAs and uses AI to find small errors. Seems very productive. IAWW (talk) 12:12, 2 December 2025 (UTC)[reply]
If you want to look through existing GAs perhaps check them as part of the GA sweep mentioned above. CMD (talk) 15:56, 2 December 2025 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Good_article_nominations&oldid=1325345588"