Wikipedia talk:Articles for improvement

OverviewNominationsHolding areaScheduleParticipantsDiscuss

This Week's Article for Improvement (week 45, 2026)

Blueberry-flavored frozen yogurt

The following is WikiProject Articles for improvement's weekly selection:

Frozen yogurt

Please be bold and help to improve this article!

---

Get involved with the project. You can: Nominate an article • Review nominations

---

Upcoming articles

AFI automated checklist

Item	This week	Next week	Week after
1. Weekly page created
2. Article specified
3. Picture chosen
4. Caption written
It is currently week 45 of 2026 ()

Spinning top

Week of 9 November – 15 November

• Google
• News
• Books
• Scholar
• JSTOR

Add your sources

Garbage

Week of 16 November – 22 November

• Google
• News
• Books
• Scholar
• JSTOR

Add your sources

Mass media

Week of 23 November – 29 November

• Google
• News
• Books
• Scholar
• JSTOR

Add your sources

[[File:|alt=]]

The AFI article for Week 47 of 2026

Collected garbage at Attero, Wijster, the Netherlands

The selection for week 47 (starting 16 November 2026) is Garbage. — MusikBot ^talk 00:05, 26 October 2026 (UTC)

Discussion (week 47)

Please any comments or concerns about the selection here. If three net members elect to reselect the article, then the article will be reselected and the former one removed from the list of articles.

Originally I manually changed it to Solar storm following Online encyclopedia's removal. Then when Bob the Builder got pushed back to week 49, I changed it again to Garbage to test what would happen with semi protected articles, and also seeing if we need a new AFI nomination rule preventing articles that would most likely be vandalism magnets. The bot changed it to Hide-and-seek, and I manually changed it back to Garbage again. If anyone wants it changed back to Hide-and-seek (the bot's chosen article) or Solar storm, I am fine with changing it back. dualcerbtalk 15:37, 26 October 2026 (UTC)

Changed it back to Hide-and-Seek after spinning top got removed. dualcerbtalk 11:39, 27 October 2026 (UTC)

@Dualcerb Axle doesn't look like a very interesting article for newcomers. Shall I add the Chihuahua article to week 48 instead? They're both C-class, and the AFI project will likely be settled in by then. NewAccount7295 (talk) 22:03, 27 October 2026 (UTC)

Just wait for the bot to pick one. dualcerbtalk 12:32, 28 October 2026 (UTC)

But you picked axle for that week? The bot isn't going to pick one since it is already listed. NewAccount7295 (talk) 14:19, 28 October 2026 (UTC)

It's gonna pick one, just wait. dualcerbtalk 14:23, 28 October 2026 (UTC)

@Dualcerb Is there any reason you're opposed to running Chihuahua that week? You yourself said that articles about dogs would be a good choice for AFI. NewAccount7295 (talk) 15:13, 30 October 2026 (UTC)

Not opposed actually, you can change week 48. dualcerbtalk 15:15, 30 October 2026 (UTC)

Post-start RfCBEFORE (and relevant discussions)

There has been a ton of discussion about AfI and what needs to be changed, and the consensus appears to be that we need a RfC to re-assess the project. Hence, yet another topic. Here are the major questions I've seen at this point are, which should probably be part of the RfC, and all the links I could find (feel free to edit if anyone finds more):

1. Should AfI continue on the Main Page? (here, here, here)
2. Where should AfI be on the Main Page? (above, and here)
3. How will we balance to compensate for AfI? (here, here)

Here are some other questions to keep in mind:

4. What are the criteria for an article to be featured? (here, here, here)
5. How long should an article be featured? (dispersed among other links)
6. Should we protect/unprotect pages before to going to AfI? (here, here, here)

I assume we'll need the RfC up by the time the next AfI is on? 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 19:56, 21 October 2026 (UTC) edited 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 20:36, 21 October 2026 (UTC)

I think 4-6 don't really need an RfC and can be worked out here on this page. We should definitely start working on the other questions however. InfernoHues (talk) 20:00, 21 October 2026 (UTC)

Agreed. FaviFake (talk) 20:10, 21 October 2026 (UTC)

I also agree with you on that. Covering only 1-3 would make the RfC less complicated and more to the point. NewAccount7295 (talk) 20:32, 21 October 2026 (UTC)

I think five needs to be in the same discussion as 1-3. Staleness will be a decisive factor in many votes on location and whether it should be on there. 1brianm7 (talk) 23:03, 21 October 2026 (UTC)

From what I'm seeing with the current article, I believe a week is needed for it to fully develop, but to fully decide that, waiting a few weeks for the RfC would be good. However, the opening statement to this thread suggests starting it next week. (If there is enough demand for one, I am not opposed to starting it around then) NewAccount7295 (talk) 23:07, 21 October 2026 (UTC)

I disagree. ALL of these should be brought up at an RfC, as they are critical parts of AfI. Especially 4 - it is important to have a clear CENT discussion of how/why an article should be "featured" on the MP. Natg 19 (talk) 23:22, 21 October 2026 (UTC)

I was actually thinking of starting a similar RFC workshop, but it looks like you have beaten me to it!

Anyways, I see that there are two main categories of questions:

• Questions about the Main Page and the placement of each of the sections/modules

• Allow me to add @Chorchapu's list from the Idea Lab:

1. TWAFI to the left side; expanded RD on the right side (we'd have to figure out where exactly)
2. TFA expanded to full width and placed at the top; TWAFI to the left side
3. TWAFI full-width below the blue/green columns, above TFL/TFP
4. Remove TWAFI altogether (unlikely, since it has unambiguously worked)
5. Another module on the left side (the Signpost, WP:CENT, or something else?)
6. Status quo of 46/54 columns
7. AFI above all, with white background
8. All featured content together (TFL/TFP to the top); total reoganisation of the page

• Questions about the AFI project itself, including:

• How often should the AFI be changed (one day, half a week (3.5 days), or one week?)
• What should be the protection settings be used, before, during, and after its appearance on the Main Page?

• Going to add @Dualcerb's suggestion from above (which I believe could be part of the RFC):

1. When the page leaves AFI, the indefinite semi protection is reinstated.
2. When the page leaves AFI, indefinite pending changes is applied instead of indefinite semi protection. This allows new editors to continue editing the page even after it leaves AFI.
3. When the page leaves AFI, no protection is reinstated. Same reasoning as the 2nd option, just putting more faith in the new editors to not vandalize the article.
4. We don't allow indefinitely semi protected pages to be on AFI in the first place.

• Are there any other pages or systems (including automation) within the AFI WikiProject that we still need to sort out?

Hason-LEK-SIN ● 💬 ● 🧱 ● 🧪 00:54, 22 October 2026 (UTC), edited: Hason-LEK-SIN ● 💬 ● 🧱 ● 🧪 01:07, 22 October 2026 (UTC)

Oops, the ping didn't go through, trying again: @Dualcerb. Hason-LEK-SIN ● 💬 ● 🧱 ● 🧪 01:08, 22 October 2026 (UTC)

What I probably need to do now is to add a preview of these options (in the Main Page alternatives), but that's going to take some "coding" and time... Hason-LEK-SIN ● 💬 ● 🧱 ● 🧪 02:01, 22 October 2026 (UTC)

I was just thinking about that. But I certainly lack the skills to do so - at most I could make a half-assed attempt at generally where things are. Chorchapu (talk | edits) 02:03, 22 October 2026 (UTC)

I do not think we should spend valuable editor time asking whether we should allow a seven-day appearance on the main page to constrain admin's decision-making on whether to protect frequently vandalized pages. 1brianm7 (talk) 02:05, 22 October 2026 (UTC)

Okay, other possible questions since these should probably have proper, visible consensus instead of being thrown around here:

• How many votes is needed for an article nominated at Wikipedia:Articles for improvement/Nominations to be approved? (Currently three net supports)
• Should articles that talk about a brand be removed from AFI because it is "promotional"?
• What should we do with articles that become ineligible for being an AFI on the Main Page (for eg, because it was nominated before AFI's goals became something for newcomers)? Should we remove them, or should we keep them but separate them into a "harder" list?

Hason-LEK-SIN ● 💬 ● 🧱 ● 🧪 02:26, 25 October 2026 (UTC)

I've started an RFC workshop: Wikipedia:Articles for improvement/TWAFI RFC workshop

I don't think the first is that big of a deal. All you need is a consensus to approve it, like at AFD or any other venue. I suggest mostly sticking to basic questions. NewAccount7295 (talk) 02:37, 25 October 2026 (UTC)

This is the RFC workshop, but sure. Hason-LEK-SIN ● 💬 ● 🧱 ● 🧪 02:56, 25 October 2026 (UTC)

Do you think it would be better to make removal a part of the placement question (e.g. A. Remove B. Full row below WTW C. Single col. above ITN... etc), or keep it separate? 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 03:00, 25 October 2026 (UTC)

No. The first question is about whether TWAFI will remain on the Main Page or not. The second is where it will be placed on the Main Page, under the condition it is kept. NewAccount7295 (talk) 03:03, 25 October 2026 (UTC)

For option 8, I think we could split it between "move TFL/TFP alongside TFA" and the wildcard "redesign the Main Page from scratch". There have been quite a few proposals for an updated Main Page look (including one by @Rjjiii which I quite liked but don't remember the link of), and this could be a good opportunity, which could open the door to further discussions on how to proceed. Chaotic Enby (in solidarity · talk · contribs) 16:05, 1 November 2026 (UTC)

It was probably this mockup of a card-based Main Page from my sandbox. I don't intend for it to be an actual concrete proposal as it's quite messy, and because actual designers are editors and I imagine could do better, but to show how we could do something that allows for more photos and more flexibility. The mockup has a spoken article of the day to show how we could trial run new ideas. Also, at DYK at least, there are more people locating public domain photos and taking their own photos than we have spots to run photos but it does feel like something valuable that we should encourage rather than discard. Rjjⁱⁱⁱ (talk) 19:11, 2 November 2026 (UTC)

Whoa, that looks sick! FaviFake (talk) 22:58, 2 November 2026 (UTC)

Love the design language, currently thinking about how to incorporate it (and the more visual approach in general) while keeping a more compact Main Page. Chaotic Enby (in solidarity · talk · contribs) 23:15, 2 November 2026 (UTC)

I strongly agree on answering these questions or similar questions via an RfC. Same for the questions ^ shared by @Hason-LEK-SIN. - Wil540 art (talk) 16:39, 22 October 2026 (UTC)

I don't see why we need to be hasty. n = 1 articles for improvement on the main page is not a big sample size. Let's wait for passions to cool on all sides, at least, before we move forward with a new RfC. Bremps... 19:56, 23 October 2026 (UTC)

Agreed. Before we start a months-long discussion, we should at least wait for three or four TWAFIs to see how they went. FaviFake (talk) 20:26, 23 October 2026 (UTC)

RE my original comment, since a lot of complaints about TWAfI concern its displacement of ITN, and considering that ITN will have one or two RfCs which may change its layout or existence in the near future, perhaps it'd be a good idea to make our RfC coincide with theirs? 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 23:43, 24 October 2026 (UTC)

The RFC would (partly) focuses on Main Page layout, so that might be a good idea, but it's going to be difficult to coordinate this. And to clarify, the RFC where I would ask the above questions would be held at the Village Pump. Hason-LEK-SIN ● 💬 ● 🧱 ● 🧪 02:08, 25 October 2026 (UTC)

The only thing I feel really strongly about is this: WP:RFCBEFORE (if you read it) isn't about workshopping questions. It's about deciding whether your question could be answered through some non-RFC process, like asking at the Teahouse.

(One of these days, we should probably rearrange the Wikipedia:Requests for comment page so that shortcut points to whatever people expect to find there, instead of to what's always been there.) WhatamIdoing (talk) 03:09, 25 October 2026 (UTC)

 Done (and reverted :(   ) FaviFake (talk) 11:39, 25 October 2026 (UTC)

I think the TWAFI should be moved elsewhere, have it small when the mention of the article that needs attention and remove it since having it above the ITN is a bad place for it due to inconvenience issues for many users, where it should be below it or elsewhere. BattleshipMan (talk) 03:20, 26 October 2026 (UTC)

I believe TWAFI should be moved to the left — Ultimate Ultimatum (talk) 22:51, 2 November 2026 (UTC)

Section break

The ITN RfC is being held right now, but it is only for changes to the project, not about removing it. If any of the reforms pass, an RfC on abolishing ITN will not be held for at least three months.

The dust has started to settle pretty quickly, so I actually think this week or the next would be a decent time to hold the RfC for TWAFI. Chaotic Enby proposes that for higher visibility than last time, and the need for reader feedback, a small notice be added to the Main Page about it, see Template:Article for improvement/sandbox. There has been no opposition to doing so.

The ITN RfC isn't too active so I don't know if it would affect this one too much being held at the same time, if there are concerns about that. NewAccount7295 (talk) 02:55, 29 October 2026 (UTC)

@FaviFake @Axolitl @7amithorn @Chorchapu @Hason-LEK-SIN @Eman7blue42 NewAccount7295 (talk) 14:37, 29 October 2026 (UTC)

I'm neutral on when we want to run the RfC. I would like to match it to the ITN one, so changes can adjust for each other, but we really only have one full week of data to show for it.
Besides the notice on the template, WT:MAIN, and WP:VPT (due to the long discussions there), does anyone have any suggestions for notifications? 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 15:34, 29 October 2026 (UTC) edited 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 15:43, 29 October 2026 (UTC)

The ITN RfC is on CENT; I assume we'd want to do the same here too. Chorchapu (talk | edits) 15:37, 29 October 2026 (UTC)

why did I forget that 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 15:41, 29 October 2026 (UTC)

As shown in the linked AFI sandbox, @Chaotic Enby plans to put a small notice above the AFI section on the Main Page mentioning the discussion to solve any issues with that. NewAccount7295 (talk) 15:39, 29 October 2026 (UTC)

Yes, I had suggested putting a TfD-style notice for the AFI (and ITN) discussions on the Main Page, for the benefit of readers and non-guru editors. It's been removed from the sandbox for now, but it can be seen at Special:PermanentLink/1382425254. Chorchapu (talk | edits) 15:42, 29 October 2026 (UTC)

Courtesy link to the ongoing discussion at Wikipedia:Village pump (policy)/ITN reform/RFC phase that relates to a similar proposal. Chaotic Enby (in solidarity · talk · contribs) 15:46, 29 October 2026 (UTC)

I'm willing to start the RfC today if that works. We're halfway through the second week of AFI, so it will only be a few days until we get data for the mechanic article, and RfC will last much longer than that. Anything that needs to be changed in the workshop before then? NewAccount7295 (talk) 15:58, 29 October 2026 (UTC)

The size limit question needs to take into account that there was consensus in the last RFC for the current layout of the AFI box. If we are asking that we should probably ask what the layout of the AFI section should be in the future.

I also noticed there is no placement option for below DYK and OTD but above TFL and POTD. Eman7blue42 (talk page | recent edits) 16:12, 29 October 2026 (UTC)

Question 2, i.e. Where should AfI be on the Main Page? is really the major issue by far, given the very weak placement consensus of the last RfC. The gun was jumped in posting it in the current placement; now we must run after it. AfI has not been given enough of a trial period to adequately contemplate removing it completely, rendering Question 1 as too-soon. Question 2 must be answered before question 3 can be addressed. Question 3 will happen naturally with spacing hotfixes anyhow (mayhaps? I have it hidden, so I have not been keeping up). Adding more content of some sort to compensate for the space is an entirely different beast; honestly, consensus to add and where to add something is necessary before balancing work can really take full shape. 4-6 are extremely minute in comparison; suitable for a different time. Thus, I would suggest just focusing solely on question two. — Godsy ^(TALK_CONT) 12:26, 29 October 2026 (UTC)

There was heavy demand for that question, with many people calling for it to removed from the Main Page after it was added. At the very least, I think the RfC needs to ask both Question 1 and Question 2. NewAccount7295 (talk) 14:21, 29 October 2026 (UTC)

I mentioned this in a discussion about AFI on another page, but Question 5 ("How long should an article be featured?") is also going to be related to placement. I've seen editors bring up both issues together, in the sense that they are objecting to giving one article a full week above the fold. If there's consensus to go with a horizontal banner below OTD and DYK or even further down, below the Picture of the Day (POTD) and Today's Featured List, I think that will change a lot of people's opinions on it. I occasionally come across comments from experienced editors who don't even realize that the POTD is on the main page, so I don't imagine there will be the same level of concern about running anything that far down for multiple days. Rjjⁱⁱⁱ (talk) 15:04, 29 October 2026 (UTC)

I think there are too many questions being suggested there, but I'm okay with that one being asked in the RfC. NewAccount7295 (talk) 15:06, 29 October 2026 (UTC)

I think we've addressed this with a new question. FaviFake (talk) 01:17, 30 October 2026 (UTC)

If no options to improve TWAFI are included in the RfC, I will !vote to remove it so improvements can be made without taking up a fifth of the front page of the seventh most visited website in the world. 1brianm7 (talk) 15:47, 29 October 2026 (UTC)

• Maybe this is a distinction without a difference, but I wonder if Should articles whose primary subject is a for-profit corporation, publication, consumer product, software application or game, fictional character, or media franchise be eligible for nomination under the articles for improvement criteria? conflates fictional characters, consumer products (this is especially vague), and corporate entities in a way that isn't reflective of the project's goals, or may cause confusion/discussion among participants. For example, should all of these articles be forbidden under the same rule? Teresa (Barbie), Shinsegae, Jack Frost, Jenga, Titania (A Midsummer Night's Dream), Easter Bunny, Hacky sack. Wracking ^talk! 03:35, 4 November 2026 (UTC)

  I am open to feedback to reword that question. It was hard to write. FaviFake (talk) 04:01, 4 November 2026 (UTC)

  I don't want to overcomplicate things, but I think it's worth considering what the goal would be of such a rule, and how to write the question to reflect that. Is the question meant only to prohibit articles related to for-profit corporations and their products? (I read the adjective "for-profit" as only applying to "corporation", which could be the source of my confusion.)

  Could a noncommercial fictional character (e.g., Easter Bunny, Titania (A Midsummer Night's Dream)) be eligible, so long as editors agree it meets other AFI criteria including "highly important" and "interesting" ("regional topics" if passed)? (These are just examples, I didn't look hard for what might be the best AFI noms in this area.) Wracking ^talk! 04:27, 4 November 2026 (UTC)

  For me at least, the goal is to prevent the appearance of promotion. "Help us improve Bob the Builder!" could sound like a paid ad or sponsorship to a reader or new editor. From your list above, I would be opposed to including Teresa (Barbie), Shinsegae, and Jenga. The rest are fine. InfernoHues (talk) 04:31, 4 November 2026 (UTC)

  Exactly. I think we can resolve this by simply adding "commercial" or "for-profit" before every item on that list. FaviFake (talk) 09:14, 4 November 2026 (UTC)

  That makes sense to me. Wracking ^talk! 19:04, 4 November 2026 (UTC)

  There will also be edge cases that come up where you have properties that long-since lapsed into the public domain, but remain a trade-marked commercial entity in some areas. Cthulhu, Peter Pan, and Reversi both come to mind as articles on something that is clearly in the public domain, and also legally trademarked as an active product. Rjjⁱⁱⁱ (talk) 19:31, 4 November 2026 (UTC)

  I've asked this before, but, what is this push to avoid the appearance of promotion when TFA and DYK already features such articles all the time, and articles on media properties are some of the most popular ones on Wikipedia, including for new editors? NewAccount7295 (talk) 19:09, 4 November 2026 (UTC)

  The push to avoid promotion is because people are seeing TWAfI as promotion. No idea why TWAfI in particular, and not others, but it is happening (tried looking for relevant diffs, got lost in the pipes of the village pump), so we're trying to avoid the misconception. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 20:12, 4 November 2026 (UTC)

  What if we moved TWAFI into a page of its own and left a link to it on the Main Page so that no sections are obstructed? Either that or move it to the left where TFA was. — Ultimate Ultimatum (talk) 20:18, 4 November 2026 (UTC)

  Your first proposal was tried, and failed (c.f. this report). The second is an option, but imho pushing TFA would cause far far more backlash than pushing ITN. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 20:24, 4 November 2026 (UTC)

  For some outside perspective on this, a couple of news sites and social media discussions have touched on how often Wikipedia's Main Page showcases the work of fashion designer Alexander McQueen. And then their conclusions were just that Wikipedia's a volunteer project, and someone is willing to thoroughly document McQueen's work. Also, even for existing projects, this gets raised often at WP:ERRORS. In this case, it is bit different because the AFI spot is requesting work rather than showcasing it. Rjjⁱⁱⁱ (talk) 20:26, 4 November 2026 (UTC)

  I believe that simply has to do with it being a new section on the Main Page. If TWAFI were not new and people were used to it, I don't know that editors would see it as being so promotional. NewAccount7295 (talk) 20:40, 4 November 2026 (UTC)

Data duplication

We currently have AFI data duplicated and spread across two template formats, the current, more complete {{Article for improvement/data/2026/44}} (which is parameter-based), and the legacy {{Wikipedia:Articles for improvement/2026/44}} (which is subpage-based). Should we transition fully to using the former, or is there some benefit to using both? Chaotic Enby (in solidarity · talk · contribs) 19:47, 23 October 2026 (UTC)

I think we should try to stick to the decade-old format as much as possible, mostly for backwards compatibility. In the future, if we want to do, idk, a cool showcase of AFI over the years, we won't need to handle two completely different systems.

That is, {{Article for improvement/data/2026/46}} should be merged into Wikipedia:Articles for improvement/2026/46. I'm not sure whether MusikBot substitutes a template to create those subpages: if that's the case, we can edit that template so, when subst'ed, it transcludes the full blub that appears on the MP on the given week.

The name of the TWAFI should probably remain at Wikipedia:Articles for improvement/2026/46/1, so in case we want to switch to two a twice-weekly format we can just use /2 as was done a decade ago. (Actually, in that case we'd have two MP blurbs in one week, so how would those be archived if we only have one page in the format Wikipedia:Articles for improvement/2026/46? This is harder than it sounds.)

I think this should ideally be done sooner rather than later, or else we'll have to move and reorganise more than just a couple TWAFIs once we settle on one structure.

I've posted this elsewhere, but FYI the old system is documented at:

• Wikipedia:Bots/Requests for approval/Theo's Little Bot 14
• Wikipedia:Bots/Requests for approval/MusikBot 8
• Wikipedia:Bots/Requests for approval/MusikBot 7

FaviFake (talk) 20:23, 23 October 2026 (UTC)

Problem is, if we have two weekly articles, we'll need things like /1/picture and /2/picture, etc. which means again dealing with a different format. Plus, so many more subpages to keep track of. A big change is the opportunity to switch to a simpler data format, and I'll be more than happy if MusikBot could create subpages based on the data template for backwards-compatibility. It currently doesn't look like it is being done, and all the Wikipedia:Articles for improvement/2026/46-format subpages have been empty for now. Chaotic Enby (in solidarity · talk · contribs) 20:34, 23 October 2026 (UTC)

Oh, right. That would be impossible to manage.

But I think we should at least keep the blurbs of specific days as subpages of Wikipedia:Articles for improvement. I think this is more similar to what WP:FA and WP:OTD are doing. The underlying code of these subpages should be changed, of course, but at least we could try to stick to the existing page format, at least a little bit? FaviFake (talk) 20:42, 23 October 2026 (UTC)

That definitely works for me! Chaotic Enby (in solidarity · talk · contribs) 20:44, 23 October 2026 (UTC)

AFI automated checklist slightly outdated?

Moved from a new section below

 – FaviFake (talk) 11:07, 25 October 2026 (UTC)

In the project management section, there is a checklist that looks like it is intended to be fully completed before an article for the week should be posted to the main page. Currently the "caption written" row is all Xs and hovering over it tells the reader that a page needs to be created. However for this week, Template:Article for improvement/data/2026/43 already has a caption that says "Candy store in Maryland, United States" which also appears at other templates like Template:Article for improvement banner/Picture box.

Since the row seem to be inaccurate now to current practice, how should it be handled? --Super Goku V (talk) 08:46, 25 October 2026 (UTC)

Recent main page AFI edit stats

Found myself up early this morning so did a quick and dirty look at the recent edits to confectionery shop:

• ~750 edits in first 56 hours
• Mostly registered accounts, only ~25 from TAs
• Around 40 edits tagged as rollbacks and 10 tagged as manual reverts
• ...so presumably another 50 or so unconstructive edits in total
• Surprisingly few obvious LLM edits - about ten tagged as LLM check or paste check

Just on the general shape of the edits, I think that looks pretty positive so far! Well done everyone who worked to get this set up.

One thing that seems to have provoked a lot of back and forth is unsurprisingly ENGVAR - leaving aside the name there's also been a bit of a dispute about exactly how to pick the right variation. Might be a good thing to clearly settle before putting the next article on the mainpage? Andrew Gray (talk) 08:05, 21 October 2026 (UTC)

Might be a good thing to clearly settle before putting the next article on the mainpage?
Great idea! Where do you think these kinds of discussions should be held? FaviFake (talk) 16:25, 21 October 2026 (UTC)

I confess I'm not sure :) I think for ENGVAR and such there's a lot to be said for just having a clear decision and sticking to it, even if it's arbitrary, rather than having people getting bogged down going back and forth over it on a heavily edited page. And I suspect that AFI articles are the sort of topic that's more likely to have disputes because there's less chance of clear national ties.

It could perhaps be part of the checklist for selecting front-page AFI ("is there a stable and clearly stated ENGVAR style? If not, agree what it is") and then we can try to have a temporary moratorium on changing it whilst up at AFI, much as I think was done with the renaming discussion for this article? Andrew Gray (talk) 17:09, 21 October 2026 (UTC)

Thanks for doing this! Do you know how many unique users edited, and what their median edit count was? I noticed a lot of accounts with 10 or less edits participating. InfernoHues (talk) 16:59, 21 October 2026 (UTC)

I don't - I decided that was a bit hard to work out by hand so I'll try and do it by Quarry once that's back up (it's caught up in the PHP security mess at the moment). But anecdotally, yes, I think a lot of newer editors, which is really heartening to see. Andrew Gray (talk) 17:11, 21 October 2026 (UTC)

Oh I didn't know you did the above by hand! No worries then. InfernoHues (talk) 17:13, 21 October 2026 (UTC)

I can get you a count of unique users for the moment - up to 1700 UTC today (65 hours) there had been 832 edits by 298 distinct users. 27.5% of editors were TAs, and they made 27.6% of all edits.

53 edits were manual reverts or rollbacks, 76 undos (which I forgot to count earlier). So even in the worst case of assuming all those undos were removing unconstructive edits, that'd be something like 130 unconstructive edits, 130 reverts, 570 neutral or positive contributions. So that's 1-2 unconstructive edits and 9-10 neutral/positive per hour, depending on how you count the undos. Andrew Gray (talk) 18:05, 21 October 2026 (UTC)

Looks like Quarry should be up sometime next week, so I'll get those numbers ASAP once it is. Andrew Gray (talk) 09:10, 24 October 2026 (UTC)

Like an hour before you made this post I was on this page thinking of how to phrase the proposal that we settle all English variety and date stuff before it reaches the main page. At the top of Talk:Donald Trump, they have a long list of conventions on stuff like the Oxford Comma, English variety, em-dash, etc. Like two days before TWAFI goes live, I'd say an editor should just standardize everything one way or another. Then, at the top, we just say "these are the conventions. It is not needed to try and adhere to them, other editors will fix the article to comply once it is off the main page. Do not attempt to alter any conventions." 1brianm7 (talk) 19:13, 21 October 2026 (UTC)

Right, I think the guidelines for WP:CITEVAR, WP:DATEVAR, WP:ENGVAR and so on just apply as usual. One note: three templates will not just inform editors but also auto-format many citations: {{use DMY dates}}, {{use MDY dates}}, and {{CS1 config}}. CS1 config adjusts things like punctuation with full stops (HELP:CS1) versus commas (HELP:CS2), and author names. It works on the CS1 templates, {{Citation}}, and the many wrapper templates. The two date templates also work on these templates as well as other less common citation templates like {{cite patent}} and {{cite comic}} via Module:Auto date formatter. If you set these templates up at the top of the article it will make it easier for new users who use the machine-generated citations where you just input the URL, doi, ISBN, etc. Rjjⁱⁱⁱ (talk) 20:31, 21 October 2026 (UTC)

Just wanted to chime in and add my thanks and praise to everyone who made this possible. By that I mean everyone who approved this on the admin side, everyone who helped write and test the bot code, everyone who advocated for this from the AFI project, and everyone who did the countless other tasks that I probably don't even know exist.

		The Main Page Barnstar
		A sparkling Main Page Barnstar is due to each and every one of you. You know who you are.

Placing Articles for Improvement on the top of the main page is the sort of idea that is so genius it seems obvious in retrospect. It blew me away when I saw it. I was mindlessly opening Wikipedia to check the news and stopped dead in my tracks. For a moment I was annoyed, before it dawned on me just how good of an idea this was. Not to sound grandiose, but I think in the long term we will look back on this as a positive turning point not just for the AFI project, not just for the main page, but for Wikipedia's culture as a whole.

Yours sincerely,

SilverStar54 (talk) 09:41, 23 October 2026 (UTC)

How many new accounts were created after the confectionery shop was added to the main page, with the account’s first edit being to the confectionery shop article? Some1 (talk) 03:49, 26 October 2026 (UTC)

Okay, here's some more numbers (cc @Some1 @InfernoHues @FaviFake) - this is the source data.

• 1569 edits overall, of which 18.5% were later reverted
• 453 edits by TAs, of which 40% were later reverted
• 55 edits by brand new user accounts (created while AFI on front page), of which 43.6% were later reverted
• 261 edits by slightly older new users (account earlier in 2026), of which 9.6% were later reverted
• 800 edits by all other users (account created before 2026), of which 7.5% were later reverted.

This is checking for the mw-reverted tag, which doesn't necessarily mean that particular edit was explicitly rolled back - it can eg include "several edits then the page is manually rolled back to an old version" (see mw:Manual:Reverts#Revert types). Edits undone after some time has elapsed may also not get marked as reverted; I'm not sure how long that threshold is. But it's a decent rough-and-ready measure for what we're looking at.

So revert rates definitely picked up a bit through the week compared to what we were seeing initially.

Editcounts are tricky because we can only see "that user's edit count today", not sat the time of editing. But there were 487 distinct users (including TAs), who currently have a median of 185 edits. 119 of those users had all their edits to the page reverted; the 367 users with surviving edits have a median of 303 edits overall.

There were 150 distinct TAs, of whom 89 had all their edits reverted. TAs made up 62.4% of all reverted edits. Andrew Gray (talk) 22:07, 27 October 2026 (UTC)

I must’ve missed this, but what about the number of distinct new editors whose first edit were to that confectionary shop article while the article was on the MP? Some1 (talk) 00:34, 28 October 2026 (UTC)

@Some1 Ah, sorry - I didn't work that out because it got complicated fast; there's no "this was the first edit" flag so you'd have to assess the full edit history of each account and it's hard to do that in Quarry.

I checked manually -

• 11 user accounts created during the AFI period only ever edited that page (so therefore it must have been their first edit)
• Seven more made their first edits on the AFI
• Two more made their first edits to a sandbox then immediately to the AFI
• Two more made their first edits to an article but edited the AFI within a few minutes
• The remaining six made their first edits elsewhere

So 18 of the user accounts created during the AFI period made their first edit to the AFI, and another 2-4 might have been motivated to create it by AFI tho it wasn't their exact first edit.

Interestingly, there were another seven user accounts that only ever edited the AFI, but were created earlier - four earlier in 2026, and three pre-2026. Andrew Gray (talk) 19:07, 28 October 2026 (UTC)

Interesting stuff, thank you!! Some1 (talk) 20:24, 28 October 2026 (UTC)

@Andrew Gray, do you plan on gathering statistics for mechanic as well when it goes off the Main Page soon? Maybe you could compare them to the confectionery store statistics. NewAccount7295 (talk) 16:02, 1 November 2026 (UTC)

Yes, I'll do a comparable set, probably on Tuesday. Andrew Gray (talk) 18:21, 1 November 2026 (UTC)

Amazing, thank you so much! You should consider collecting all of your findings about the AFIs at the RfC workshop, so once it will be opened !voters will have clear data about how it's going before deciding where the section should be changed. FaviFake (talk) 23:02, 1 November 2026 (UTC)

Will do! I've rerun the numbers for mechanic: in total, 820 edits by 313 distinct accounts, 13.2% reverted

• TAs: 146 edits, 39.7% reverted
  • 88 distinct accounts, 43 entirely reverted (no surviving contributions on the page)
• This-week new editors: 39 edits, 46.2% reverted
  • 25 distinct accounts, 12 entirely reverted
• This-year new editors: 185 edits, 6.5% reverted
  • 56 distinct accounts, 12 entirely reverted
• All other editors: 450 edits, 4.4% reverted
  • 144 distinct accounts, 5 entirely reverted

So about half the edits overall compared to the previous week, proportionally fewer TAs and proportionally more "new this year" editors. Overall revert rate is a bit lower but that might just be because of less frothy editing rather than less vandalism.

I'll write something up a bit tidier in the next day or two for the RFC. Andrew Gray (talk) 20:35, 2 November 2026 (UTC)

And RFC notes updated. One thing I found particularly interesting: there were around 70 pageviews per edit.

In the nine months January-September, the same two articles had ten edits and 39,652 pageviews, or about 4,000 pageviews per edit (and a significant fraction of those edits were cleanup bots).

The baseline average across all of Wikipedia - in the two weeks these were on AFI, the whole site had 2.7bn pageviews (counting mainspace only) and 914k mainspace edits, or around one edit per 2,900 pageviews. (This does not count deleted edits, but it's still fairly indicative) Andrew Gray (talk) 23:09, 2 November 2026 (UTC)

PS @FaviFake I'm assuming you wanted these to go in the stats section at the top - please feel free to move them around as appropriate. Andrew Gray (talk) 23:14, 2 November 2026 (UTC)

Yes exactly! Feel free to edit the in any way, even for other AFIs or other sections of the workshop, don't worry! The data will probably be collapsed in the end, leaving uncollapsed a short summary (feel free to write that too if you want) FaviFake (talk) 23:17, 2 November 2026 (UTC)

I've tweaked it around a little. I think the first paragraph there would serve as an overview summary and then the actual data/notes can go below the fold. Andrew Gray (talk) 00:10, 3 November 2026 (UTC)

@Andrew Gray If just discovered this page in case it's useful for comparing Main Page AFI vs old AFI: Wikipedia:Articles for improvement/Accomplishments § Weekly statistics FaviFake (talk) 01:06, 3 November 2026 (UTC)

@Andrew Gray Can you also start calculating edits per day for the articles to list on the statistics page, kind of like of how it lists pageviews per day? NewAccount7295 (talk) 15:36, 5 November 2026 (UTC)

Yes, that'll be easy to do - I think I can even do surviving edits + reverted edits per day. I'll run up a graph showing edits and pageviews, since it might be easier to read than the table of numbers, and we can regenerate it next week when the third set of data comes in. Andrew Gray (talk) 17:37, 5 November 2026 (UTC)

I wonder how the stats relate to the featured articles or the DYKs. – The Grid (talk) 17:48, 5 November 2026 (UTC)

When I originally did this I ran the numbers for 30 October (selected on the basis that the DYKs were midnight to midnight that day so we have a clean comparison) - TWAFI had a total of 5,434 views and 54 edits, as it had already hit the end-of-week drop.

By comparison there were 4,611 pageviews of the bolded articles, and just two edits to them (!). TFA got 47.8k views and 18 edits (it had about 5k/day baseline views as it was a reasonably popular bio). Our overall rate of views to edits in mainspace is about 2,900 to one, so these are both roughly in line with what we'd expect.

Main page that day was 4.1m, so a pretty low click through rate even for TFA! Andrew Gray (talk) 18:06, 5 November 2026 (UTC)

Before and after: Confectionery store

Moved from a new section

 – FaviFake (talk) 17:18, 28 October 2026 (UTC)

https://en.wikipedia.org/w/index.php?title=Confectionery_store&oldid=1380944878 (this was the last edit before being on the Main Page)

https://en.wikipedia.org/w/index.php?title=Confectionery_store&oldid=1381998415 (this was the first edit after leaving the Main Page)

• 12,687 bytes to 86,761 bytes.
• Stub class to B class.
• 1,569 edits.
• 486 editors.
• 113,699 page views.

All in the span of a week.

Beautiful... dualcerbtalk 14:35, 28 October 2026 (UTC)

It's clear to me that this was a good idea. The main page is our tool, which we haven't been using effectively at all. 3df (talk) 06:13, 29 October 2026 (UTC)

Apologies if it's already been mentioned, but Confectionery store was mentioned in The Signpost as the fifth-most edited article on the site for the period of September 23 – October 23. (That timeline doesn't include the full TWAFI period, as Confectionery store was the AFI from October 19–25.)

For the last thirty days (as of writing), Confectionery store is #3 (1661 edits), Frozen yogurt #5 (998), and Mechanic #8 (931). Wracking ^talk! 23:22, 5 November 2026 (UTC)

Accomplishments

I'd propose getting the bot back to updating the Wikipedia:Articles for improvement/Accomplishments page and splitting it by month for sake of readability. –LaundryPizza03 (dc̄) 08:46, 26 October 2026 (UTC)

+1, it's this kind of data that the success of TWAfI is gauged by. (BTW, does anyone know what language MusikBot is coded in?) 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 15:13, 26 October 2026 (UTC)

You could ask @User:MusikAnimal, MusikBot's creator. - Wil540 art (talk) 15:55, 26 October 2026 (UTC)

From its userpage, it sounds like MusikBot is programmed in Ruby. Chorchapu (talk | edits) 20:56, 27 October 2026 (UTC)

Too many new nominations

Following the addition to the Main Page, there has been a surge of nominations. The project went from not being known by almost all editors, to being linked directly from the main page of the entire wiki.

At this rate, the pool of successful AFI candidates will be huge, and the 1 article a week will not be able to remove them quickly enough. This will inevitably lead to articles remaining in the backlog for years, possibly becoming already "improved" by the time the bot picks them. It's better to have much stricter criteria and a smaller backlog of ~100 articles which are really good for new users, than to have 10,000 articles which technically meet the criteria but aren't really the best picks for new users.

I would like to ask others, and in particular SVcode who has been helping the project for a long time, how they think this could be solved. FaviFake (talk) 17:24, 19 October 2026 (UTC)

If a week is too long, 3.5 days might work. NewAccount7295 (talk) 17:32, 19 October 2026 (UTC)

Even at that rate, the backlog would not be able to keep up. Just today, we got 8 new nominations, which means we'd need to feature 7 articles every week to keep up. FaviFake (talk) 17:39, 19 October 2026 (UTC)

For example, should be make the criteria stricter? If so, how? Should the nominations page be semi-protected, so that new editors coming directly from the main page need to get a minimum of experience with editing to be able to nominate an article? We use a similar system at XfD venues where, due to the fact that editors can't use Twinkle or create the discussion pages without an account, unregistered editors are not allowed to nominate articles until they're autoconfirmed. FaviFake (talk) 17:34, 19 October 2026 (UTC)

I think both of these ideas are good. SVcode(Talk) 17:37, 19 October 2026 (UTC)

Thank you! I'll make a request for protection in a few days if there's consensus for it.

Could you kindly elaborate on which specific changes you would like to see? For example, when you decide whether to support or oppose a nomination, is there another factor you consider besides the criteria that are already at the top of the page? When you're ambivalent, what do you rely on to decide whether to support or oppose? How can we translate that in a concrete rule?

In addition, I think the 3-new-supports rule needs to be ditched. Even if the criteria are made stricter, if a nomination receives two supports from inexperienced users, it should not be added to the list just based on two votes. But I really have no concrete proposals in mind, so I'd really appreciate if you could explain your thoughts or ideas a bit more! This change should be pretty significant if we want to reduce the flood of nominations.​ FaviFake (talk) 17:45, 19 October 2026 (UTC)

I think it might make more sense to require around 7-10 different support votes for an article to be included, and then we could retroactively apply that to articles that have already been approved? Some of the ones that were approved seemed to have barely gotten any actual votes... TheAwesomeHwyh 18:06, 19 October 2026 (UTC)

@TheAwesomeHwyh Retroactively applying that standard won’t work because the way the system was set up, once a nomination achieved three support votes for a certain period of time, the bot automatically moved it to the nominated list. Look around on this talk page, a couple independent audits of the remaining nomination list have been done. That said, I support updating the three support vote limit with this newfound MP traffic. - Wil540 art (talk) 18:15, 19 October 2026 (UTC)

I like the idea of revisiting the current picks. We could set up a system where 1 of these old nominations could be re-nominated again every day, and in half a year we would've revisited all of them.

However, this is not urgent. What is urgent is the flood of new nominations and their quick approval. Would an increased vote limit really work? FaviFake (talk) 18:44, 19 October 2026 (UTC)

We could revisit this idea. NewAccount7295 (talk) 18:49, 19 October 2026 (UTC)

Unfortunately not. We don't have a script to create these kinds of pages and there are few active script maintainers.

I think the archives are working fine and I'm not sure how exactly this idea would solve the problem. FaviFake (talk) 18:53, 19 October 2026 (UTC)

Maybe we could implement a triage list? Instead of picking articles at random, we could give priority to articles that fit our criteria the most (e.g. a Top-importance article for several WikiProjects, or a Stub/Start-class article, is given higher priority), alongside those with explicit support by editors (which could help when dealing with more subjective criteria, or edge cases).

In general, having better "lines in the sand" (e.g. needs to be at least High-importance for a project or a Vital article, needs to be at most C-class) might stem the tide of articles that just won't cut it, instead of giving false hopes to editors submitting them by having them wait practically forever. Chaotic Enby (in solidarity · talk · contribs) 19:22, 19 October 2026 (UTC)

Of course, these are just examples. A C-class vital article could already be very thorough and not have any good spots for newcomers to work on, so maybe "Stub/Start-class or has a clear cleanup banner" – although more complex metrics can be more confusing and conversely easier to game. Chaotic Enby (in solidarity · talk · contribs) 19:24, 19 October 2026 (UTC)

Gonna compile the suggestions for AFI nominations here:

• Nominations can only be made by autoconfirmed users
• Triage list of nominated articles based on class, Wikiproject importance, (maybe cleanup banners present and vital article status?)
• Requiring that articles must be either High-importance from a WikiProject or vital articles, and at most C-class
• Requiring that articles must be stub/start-class or has a clear cleanup banner
• Reviewing prior nominations under new criteria (1 a day) to see if they fit the new criteria
• Scrapping the 3 net support votes rule (and replace it with what?)
• Raising the votes rule to require more votes (7 to 10)

To add on to this, I really think we should be prioritizing articles with cleanup banners in the criteria. There was discussion of scrapping the 3 net support votes, which I agree with, but no replacement option has been suggested. Perhaps we allow an AFI coordinator (or some other role that can be created) to close nominations. I also think that there should be some sort of rule that allows us to quickly reject nominations that clearly fail to meet criteria (e.g. Nazi morality). Eman7blue42 (talk page | recent edits) 01:57, 20 October 2026 (UTC)

I think we should wait for a couple of AfIs to run before making changes. It's possible that the increased nominations are due to the novelty of it being on the main page. If so, once AfI becomes more typical the nominations should slow. InfernoHues (talk) 02:00, 20 October 2026 (UTC)

you're probably right, we should wait. Perhaps once the dust settles about the whole placement thing. However I at least think for now we should require nominators to be autoconfirmed or have an account. Eman7blue42 (talk page | recent edits) 02:09, 20 October 2026 (UTC)

Remember that for every approved AFI nomination, there is usually one that isn't approved. Not all of the new nominations are going to make it to the Main Page. That means there won't be as many new AFI articles as you might expect. There will probably still be quite a few though. NewAccount7295 (talk) 18:57, 20 October 2026 (UTC)

A potential new criterion: The article topic should have good, freely available, online sources that are not already in the article. New editors likely not have access to academic journals or paywalled newspapers. The current AfI suffers from this. Most of the free online sources on Confectionery store are not very good; the good sources are paywalled. Something like Solar storm, which I supported as a nomination, would not suffer from this problem because there are free reliable sources available. InfernoHues (talk) 22:19, 20 October 2026 (UTC)

Courtesy link to French Wikipedia prominently inviting editors to look for sources for upcoming AFIs, so they play a dual role of source-finding experience for slightly more experienced editors, which could be a good thing for us to implement! Chaotic Enby (in solidarity · talk · contribs) 08:59, 21 October 2026 (UTC)

I like this general idea! It could be added as a criterion, e.g., there are at least X freely available sources online about the topic. FaviFake (talk) 16:20, 21 October 2026 (UTC)

Change number of net votes required for list inclusion based on class

Moved from a new section below

 – FaviFake (talk) 14:01, 23 October 2026 (UTC)

We have barely any stub class articles in the list, and we have a lot of C class articles.
I think that the requirements for inclusion should be 4 net supports for stub class, 5 net supports for start class, and 6 net supports for C class. (these numbers can be changed later)
Just a little idea that I had. I don't know if the bot can even work like that but I am just bringing it up that it should be easier to get a stub on the list than a C class. dualcerbtalk 13:58, 23 October 2026 (UTC)

I think that this is a good idea. Also, maybe we should tweak and combine the other ideas. SVcode(Talk) 01:04, 24 October 2026 (UTC)

+1 we should probably also re-assess articles coming in. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 14:49, 24 October 2026 (UTC)

Full agree. Of course, article class (Start vs C at least) isn't an exact science, but that's something to take into consideration. Also support the idea of a combined score. Chaotic Enby (in solidarity · talk · contribs) 15:50, 29 October 2026 (UTC)

Any suggestion to raise the number of votes required I'd support, but this suggestion seems very good. I like the idea of prioritizing stubs over starts over C-class. Eman7blue42 (talk page | recent edits) 15:57, 29 October 2026 (UTC)

I like the spirit of this idea, but I think the requirements for inclusion should depend on the level of importance and not on its class at the time of nomination, because articles usually immediately become C class after being on the MP for a few hours, so for the following 6 days what really matters is the importance of the topic of the article. Maybe we could decide based on the level of importance or the WP:VITAL level.

I also like that the WP:VITAL criteria include a minimum time limit before closing as approved and use "two thirds" instead of 50/50. These requirements would ensure that articles are considered much more carefully, to prevent incidents like these, where a nomination was approved and archived before anyone had time to point out that it was a medical topic.

What do you think of this system?

The required quorum depends on the article's importance at the time of nomination:

• At least 8 votes for level-5 vital articles, and at least 6 votes for other vital articles;
• At least 10 !votes for all other articles.

Neutral votes and the nominator's vote are not tallied. Once quorum is met, a nomination is:

• Approved if it has been open for at least 7 days and supports make up at least three quarters of those responses.
• Not approved if the majority opposes it, if it has been open for at least 2 days.
• stale if it reaches neither outcome within two weeks.

This system would be best suited for the dozens of nominations that we receive each week, which are more than we can realistically feature. Since only one article can be selected each week, my goal was to be structure AFI so it's more selective about which nominations make it onto the list, rather than allowing the list to grow indefinitely.

The 20-day limit would also prevent nominations from remaining open indefinitely when there is no clear consensus either way, similar to WP:VITAL. (If it is so controversial, then it shouldn't be approved since we have so many better articles to choose from.)

I think this would make the nomination process sufficiently robust and selective. (the green part was improved during the discussion) FaviFake (talk) 14:24, 30 October 2026 (UTC)

I would also force a re nomination for all articles added before TWAFI was on the Main Page, with the regular 3 net support rule. This would probably weed out a few more articles that would be less viable for help. dualcerbtalk 14:28, 30 October 2026 (UTC)

I agree. I think we can re-nominate one old article every 2 days, so eventually we'll have reviewed all of them under this new system (and we won't have to complain when the bot "chooses" a terrible article). FaviFake (talk) 14:30, 30 October 2026 (UTC)

Or maybe instead we do nominations for removal instead, where anybody can nominate an article to be removed from the list. dualcerbtalk 14:32, 30 October 2026 (UTC)

That could prove too contentious. NewAccount7295 (talk) 14:34, 30 October 2026 (UTC)

Is Fast food B class? It looks B class to me. (this is relevant to the discussion) dualcerbtalk 14:46, 30 October 2026 (UTC)

An entire section is unsourced. NewAccount7295 (talk) 14:47, 30 October 2026 (UTC)

I wanna know, are there any C class articles on the list that should be B class? (This is still relevant to the discussion) dualcerbtalk 14:49, 30 October 2026 (UTC)

Wouldn't that be the same thing? I don't think non-members would want to nominate existing articles for removal. it's better if we just create neutral nominations ourselves. FaviFake (talk) 14:49, 30 October 2026 (UTC)

I want the list to be trimmed again, but I do not want to suggest articles for removal given that I already nuked 70 of them a week ago and 31 new articles have entered the list since AFI entered the main page. To be honest if you could suggest articles instead of me that would be better. dualcerbtalk 15:01, 30 October 2026 (UTC)

Instead of trimming articles from the list perhaps focus more on the ones from the second part of the list (the ones chosen after AFI was placed on the Main Page) NewAccount7295 (talk) 15:03, 30 October 2026 (UTC)

We don't "focus", we just let the bot pick an article. The bot should pick good articles to feature on the MP. FaviFake (talk) 15:37, 30 October 2026 (UTC)

I don't really understand, when choosing one per week takes up minimal editor time, and the article is going to appear on the Main Page as part of such a major project, why they need to be chosen randomly?

Confectionery store was chosen by an editor. I chose mechanic and frozen yogurt, and the next few articles after that have also been chosen manually. NewAccount7295 (talk) 15:42, 30 October 2026 (UTC)

They don't need to be, but there's no reason why we should let the bot suggest bad articles. FaviFake (talk) 15:55, 30 October 2026 (UTC)

The primary reason why I supported class as a factor is because it's a general heuristic for "how much needs improving". C-class articles should be harder to nominate because they're less likely to be improvable as a Start would. How about this:

The required quorum depends on the article's importance and class. The baseline is 3 votes for approval.
Votes are added based on importance like so:

• 1 !vote for vital articles above level 5;
• 3 !votes for level-5 vital articles;
• 4 !votes for articles rated Top-level importance for a Wikiproject;
• 5 !votes for all other articles.

Votes are added based on importance like so:

• 0 !votes for stubs;
• 2 !votes for Start-class articles;
• 4 votes for C-class articles.

Neutral !votes do not count. Once quorum is met, a nomination is:

• Approved if it has been open for at least 5 days and receives at least two-thirds support.
• Not approved if a majority oppose it, even if it has not remained open for a week.
• stale if it reaches neither outcome within 20 days.

The numbers, of course, can be tweaked. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 15:19, 30 October 2026 (UTC)

The question should not necessarily consider class alone, since some articles may be assessed inaccurately. NewAccount7295 (talk) 15:20, 30 October 2026 (UTC)

1. We should try to reassess articles before nominating them. (We have Rater, it's not that hard)
2. That's why my version of the criteria keeps FF's importance scaling.

7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 15:23, 30 October 2026 (UTC)

Your suggestion is fine, but can't we just do this through standard consensus like at AFD? NewAccount7295 (talk) 15:25, 30 October 2026 (UTC)

Not going to lie, this is very confusing, this should be simplified. There probably shouldn't be a penalty for start class either. dualcerbtalk 15:25, 30 October 2026 (UTC)

I'll be AFK for a few hours (work), but if you have a revision, feel free. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 15:29, 30 October 2026 (UTC)

I think this is unnecessarily complicated. We're talking about quorum requirements, which will need to be taken into account for every single discussion. For example, for the current Furby nomination, the "closer" would need to make this calculation: 3+2+3=8, just to get the quorum. (Ironically, my system gives the same result in this case.)

I think the quality of the article and its scope for expansion should be determined by the participants, not the tallying rules. FaviFake (talk) 16:14, 30 October 2026 (UTC)

the "closer" would need to make this calculation

well, unless we programmed MusikBot to do it

But TBH, I'd rather have a guideline heuristic like Start and stub-class articles, as well as vital and top-importance articles, are considered better for an AfI than a voting count quorum; WP:NOTBUR, and KISS principle, to use the LAs. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 03:00, 1 November 2026 (UTC)

well, unless we programmed MusikBot to do it
Impossible, the formatting for the votes is usually all over the place.

But TBH, I'd rather have a guideline heuristic
At this point, this does sound better, you're right! We can scrap the different quorum requirements and just adopt the rest. I also realised that making the quorum lower for vital articles would also make it more easy to oppose these articles compared to other nominations, which is the exact opposite of what we want. FaviFake (talk) 15:56, 1 November 2026 (UTC)

I meant it would be easy to program MusikBot to go to a nom page, check article class and vitality, and then calculate and declare the needed votes, but that's no longer here nor there.

just adopt the rest ironically, it is already: criteria #1 and #3 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 18:52, 1 November 2026 (UTC)

I like this, but to keep it simple we just require 8 votes for vital articles and 10 for all others. The less complexity we introduce into the system the better. I'll also note that under the proposed system a nom with 4 support and 2 oppose votes would pass, though it wouldn't under the current system. Eman7blue42 (talk page | recent edits) 16:45, 30 October 2026 (UTC)

In case it's not clear this is regarding FaviFake's proposal. 7amithorn's proposal adds way to much complexity. Eman7blue42 (talk page | recent edits) 16:46, 30 October 2026 (UTC)

Good idea! There isn't really the need for so much complexity. I've struck the 6-vote quorum. FaviFake (talk) 16:52, 30 October 2026 (UTC)

👍 Hason-LEK-SIN ● 💬 ● 🧱 ● 🧪 01:47, 3 November 2026 (UTC)

Front page link

This is the only front page item without a link to the nominations, should be added. Militant.Insurgency (talk) 15:56, 29 October 2026 (UTC)

What should be added where? FaviFake (talk) 20:21, 29 October 2026 (UTC)

Similar to ITN nominations and DYK nominations, there should be a link to WP:AFI nominations on the front page. Where it is placed would not be my choice, that is up to the admins (or whoever does the front page stuff) :) Militant.Insurgency (talk) 21:01, 29 October 2026 (UTC)

Why should such a link be added? FaviFake (talk) 21:04, 29 October 2026 (UTC)

So people can find the nominations page easier and add items they deem good for AFI. Militant.Insurgency (talk) 21:07, 29 October 2026 (UTC)

That could lead to an overflow of nominations though. NewAccount7295 (talk) 21:08, 29 October 2026 (UTC)

We don't want newcomers to find the nominations page easier, for obvious reasons. FaviFake (talk) 21:11, 29 October 2026 (UTC)

Brands

On discussions like Wikipedia:Articles for improvement/Nominations#Elmo, people have been opposing articles just on the basis of them being about a licensed property. I feel like we should at least try with articles about them, and see how it works out. Like I was saying at that discussion, if we're trying to appeal to new editors, take a look at what they edit. Many of them start by editing articles about brands they're interested in.

There has been a lot of back-and-forth discussion about this the last two weeks. What is your opinion? NewAccount7295 (talk) 15:08, 30 October 2026 (UTC)

I wouldn't allow brands at all. Brands would be fine on the main page for a day, not a week. dualcerbtalk 15:10, 30 October 2026 (UTC)

Would you be okay with running an article about a brand for a week just as a test to see how it works out, similar to your plan to use garbage as a test for vandalism? NewAccount7295 (talk) 15:14, 30 October 2026 (UTC)

It's not a question of "working out", it's a question of whether we want brands to appear on the main page. FaviFake (talk) 15:54, 30 October 2026 (UTC)

Having some data and experience to go on could help answer that question though. NewAccount7295 (talk) 16:17, 30 October 2026 (UTC)

Even if new editors LOVED editing articles on brands, I still wouldn't want Wikipedia to feature brands on the Main Page. FaviFake (talk) 16:24, 30 October 2026 (UTC)

Well, if we end up asking that question in the RFC, it would reasonable to try it beforehand to get more informed editor opinions about how it went. NewAccount7295 (talk) 16:39, 30 October 2026 (UTC)

I disagree that more statistics will sway RfC !voters. Not sure whether we should ask this at th RfC FaviFake (talk) 16:45, 30 October 2026 (UTC)

I'm now leaning toward asking that question in the RfC. FaviFake (talk) 19:45, 30 October 2026 (UTC)

I'm willing to ask the question in the RFC but would not want to do so unless there were some actual data backing it rather than just the assumptions of editors about what they think would happen, in other words running such an article on TWAFI first; @Eman7blue42 makes a good argument on the nomination page .

It may be worth a separate RFC however depending on how long that takes. NewAccount7295 (talk) 19:48, 30 October 2026 (UTC)

Again, I disagree that more statistics will sway RfC !voters because they will not !vote based on whether a brand succeeds at converting readers into editors, they will !vote based on whether they think brands should be prominently advertised on the front page of our wiki. You keep ignoring this point.

If someone is disgusted by the idea of having a brand on the front page, they won't care about the new editors. FaviFake (talk) 19:53, 30 October 2026 (UTC)

Thank you for highlighting that, as there were multiple points I was addressing.

To me, that sounds partly like the personal opinion of some editors, but not something that would necessarily be a major issue when looking at the bigger picture. While testing that, we'd also be able to see if running an article about a media property would cause a large reaction about that. I can see a few readers being concerned, but I don't think the entire encyclopedia would be. Still, the best way to find out would simply be to try it once or twice and see. NewAccount7295 (talk) 19:57, 30 October 2026 (UTC)

we'd also be able to see if running an article about a media property would cause a large reaction about that
Wait, so you want to feature a brand on TWAFI just to see how angry the other editors would be to see it on the main page? That's just backwards: first, we ask if it should be advisable, and then we feature it in case people want it. FaviFake (talk) 22:23, 30 October 2026 (UTC)

I'm not expecting there to be as strong of a reaction as you describe though, both from editors and the general readership, even if some people do express their concerns.

I actually believe featuring a brand on AFI would be more successful for the project than a lot of people opposing it think, but the only way to find out would be to actually do it. NewAccount7295 (talk) 22:28, 30 October 2026 (UTC)

I don't think it wouldn't be successful, I agree with you. Biographies of living persons would also be successful imo, especially if the person has been in the news recently, but we also don't want to feature those, and we also decided that with an RfC. Same for medical topics.

In response to your reply below, I do not think the consensus would be an issue. Just a few days ago, Bandage was successfully nominated and archived despite it clearly being a medical topic. It was later removed, but that's still an example where the AFI team can decide on a case by case basis regardless of what the criteria say.

And, again, I do not think the consensus being uninformed would be an issue, because most editors who would be against brands would not be convinced by higher statistics. FaviFake (talk) 22:47, 30 October 2026 (UTC)

But maybe they'd be convinced if the sponsorship concern did not turn out to be as much of an issue as they believed? NewAccount7295 (talk) 22:49, 30 October 2026 (UTC)

How can it possibly "turn out to be as much of an issue"? They believe that the main page would be defaced if we allowed a brand to appear in it for a week, how could seeing exactly that change their mind? FaviFake (talk) 23:12, 30 October 2026 (UTC)

This doesn't seem like a question to be asked in the Main Page RFC, I see it more as an internal issue with interpreting the criteria and the goal of the project. If editors want to have a say in what articles get featured, then they should participate in the nomination discussions. I don't think just because an article could be associated with a brand then suddenly we can't consider it for AFI. Honestly "brand" is such a vague term; as a note we should be focusing on how to make the criteria much more clear.

While I have the same worries as FaviFake about the reaction to an article like Furby or Elmo being featured on the Main Page for a week, we don't know what the reaction will be until it happens. Honestly the backlash against AFI has mostly been regarding its placement on the Main Page rather than the articles chosen. I say we run one article (one that passes the nominations page) and see how it plays out. Its a risk, but it'll give us much needed answers. Eman7blue42 (talk page | recent edits) 20:33, 30 October 2026 (UTC)

This doesn't seem like a question to be asked in the Main Page RFC, I see it more as an internal issue ...
Wikipedia uses the RfC process when local discussions are not sufficient to resolve a dispute.

just because an article could be associated with a brand then suddenly we can't consider it for AFI.
That's not really true; the RfC would just decide whether it should be added to the rest of the criteria. We can't outright ban a category of articles; AFI voters will decide how to vote based on the criteria. FaviFake (talk) 22:20, 30 October 2026 (UTC)

My concern is that running an RfC about this, with editors purely guessing about whether it would actually be a problem, could create a permanent, albeit less informed, consensus. On the other hand, if we were to run an RfC after having done a trial process for this, there'd actually be a proven basis to base arguments on.

Since doing so might take some time, that is why I've suggested running it as a separate RfC as not to hold up the other.

So far, what I'm seeing from a lot of editors seems to be a concern that if articles about any licensed properties or brands are ran on AFI, editors will mistake them for being paid promotions. Articles about them are already featured at TFA and DYK, and, while I understand AFI may be different as some have argued, they're still shown on the Main Page all the time already. If that concern is not as widespread as they are expecting, would it still be an issue? NewAccount7295 (talk) 22:37, 30 October 2026 (UTC)

Isn’t that contrary to the point of Articles for Improvement? (Oinkers42) (talk) 20:12, 30 October 2026 (UTC)

I don't see any reason to run things that you reasonably expect would be contentious, just to find out exactly how contentious they will be, when there are so many good options already available. The list currently at Wikipedia:Articles for improvement/Articles is great - out of a hundred or so entries, there's nothing commercial (with a slight question mark on 'jumbotron'), nothing particularly regionally-focused except maybe Zinnia, no people. Exactly the sort of articles everyone can take an interest in, and so the sort of things that are perfect for AFI.

More to the point, I think that when you go to the community and say "AFI has been a great success, should we keep it on the main page?", there is an implicit "we'll be doing the same sort of thing going forward", and if that changes in future to include different sorts of articles, there might well be pushback. Andrew Gray (talk) 00:18, 31 October 2026 (UTC)

My point was that I don't actually believe such articles would be contentious to run, and they have the chance of being among the most popular ones on AFI. NewAccount7295 (talk) 00:21, 31 October 2026 (UTC)

Sorry to be negative here, but I'd be very surprised if it wasn't contentious. People feel very strongly about the front page, and they are generally very conservative about what appears on it. (As we have seen with a lot of the AFI discussions in the past couple of weeks!) The fact that people are raising concerns now almost guarantees that you'll get the same comments, louder and from more people, should it turn up on the front page.

Pragmatically, any tensions that arise around most main page selections can be defused (or even just avoided) with "well, it'll be something else in a few hours". That avenue isn't available here - if you complain on Monday, it's not going away for six days - so arguments about "this choice doesn't feel appropriate" could potentially build up quite a bit more than they ever would with eg a single DYK hook. Andrew Gray (talk) 11:07, 31 October 2026 (UTC)

Well, that just means AFI itself is contentious anyway, regardless of what articles are featured. However, I feel like some of the contention is actually starting to settle down. This makes me think, if we're refusing to do things like this out of fear of them being contentious, then why did we add AFI to the Main Page in the first place, knowing it would also be contentious? NewAccount7295 (talk) 14:11, 31 October 2026 (UTC)

Because there was consensus to do it from an RfC. Once we get consensus for featuring brands from an RfC, we can do that too. FaviFake (talk) 15:12, 31 October 2026 (UTC)

Yes, but there is a big difference between adding a new section to the Main Page, something which requires a massive RfC, and featuring a specific article or two on that section for a week as a trail run to see how it works. One of the reasons we're waiting on the RfC is so we have statistics to show when it happens, which is the same argument with this question. NewAccount7295 (talk) 15:15, 31 October 2026 (UTC)

Honestly, the list seems like some of the most boring, inoffensive articles that I can’t imagine anyone getting too particularly excited to edit, particularly newcomers. (Oinkers42) (talk) 15:03, 2 November 2026 (UTC)

Inviting some more editors from both sides from the nomination page if they want to give their thoughts on this: @Wracking @PrimalMustelid @Thunderforge @AirshipJungleman29 @Tt2460 @Chaotic Enby @SVcode @TheAwesomeHwyh NewAccount7295 (talk) 01:08, 31 October 2026 (UTC)

Noting that I have prepared the RfC question at Wikipedia:Articles for improvement/TWAFI RFC workshop § Topics. FaviFake (talk) 01:11, 31 October 2026 (UTC)

Why? Are we lacking in other articles to such an extent that we need to draw upon potentially controversial topics? ~~ AirshipJungleman29 (talk) 15:52, 1 November 2026 (UTC)

There isn't a problem with brands being featured articles. What's the difference with brands as articles for improvement? Is it just the duration on the main page, or is it something else? If, for the sake of argument, articles for improvement were only on the main page for a single day, would we be having this discussion? Thunderforge (talk) 04:03, 6 November 2026 (UTC)

The difference is that we don't want to tell editors "help us improve this specific for-profit corporation's article!!" FaviFake (talk) 04:06, 6 November 2026 (UTC)

Are these articles B or not?

Using Rater for the first time, and these articles were given a B rating with at least 70% confidence by Rater. Because B rated articles shouldn't be on the list, I am wondering if these articles are actually B (and should be removed from the list) or if the bot is being inaccurate.
Debate (B, 74.2%)
Bull (B or higher, 72.9%)
Carbon footprint (B or higher, 95.8%)
Construction worker (B or higher, 73.2%)
Earth science (B or higher, 76.2%)
Commerce (B or higher, 96.8%)
Blue cheese (B or higher, 91.4%)
dualcerbtalk 15:57, 30 October 2026 (UTC)

Where is it written that B rated articles shouldn't be on the list? I think some of these should remain. FaviFake (talk) 16:26, 30 October 2026 (UTC)

Prior consensus during the first article cut on the 3rd of October which left us with ~140 articles remaining. dualcerbtalk 16:29, 30 October 2026 (UTC)

Well starting from now I think we should not remove articles based on their class, the nomination process will be sufficiently robust. We can allow nominating one or more articles for removal at the /Nominations subpage, where the standard nomination process will apply except that the nominator will be in opposition. FaviFake (talk) 16:34, 30 October 2026 (UTC)

For now I decided to change Carbon footprint and Commerce to B due to them having over 95% confidence, but not remove them yet. dualcerbtalk 16:52, 30 October 2026 (UTC)

We really should wait to remove articles from the list until we have consensus on how to do so. I frankly don't think a strong consensus was there for the original removal. Eman7blue42 (talk page | recent edits) 17:27, 30 October 2026 (UTC)

Agreed. Once we're settled with the new nomination rules, I think we should procedurally re-nominate the articles that were previously removed from the list, one by one or even in small batches, to give all of them a fair chance, since they were successful. FaviFake (talk) 17:31, 30 October 2026 (UTC)

Earth science isn't. It has a high ORES score but it doesn't meet 2 or 3 on the assessment scale, which are things ORES scores can't check for. Several of these have had cleanup banners of some type for over a year. Rjjⁱⁱⁱ (talk) 15:12, 31 October 2026 (UTC)

upload campaign?

This is perhaps something for down the road, once this main page experiment is more settled, but I was looking at all the pictures in the confectionary store article and wonder if it's been considered to have a short line in the TWAFI box along the lines of "Have you taken a useful photo of this subject?" It could point to a campaign on Commons that automatically populates categories, and typically includes a dedicated campaign category for maintenance purposes (e.g. in case, despite some custom communication to avoid this, it turns out to be a copyright nightmare). — Rhododendrites ^talk \\ 13:59, 31 October 2026 (UTC)

Chihuahua

It doesn't make sense for User:NewAccount7295 to add an article they nominated straight to the queue for week 4648, articles are selected randomly, thanks. An article I nominated almost two years still hasn't been featured for example. Sahaib (talk) 21:29, 1 November 2026 (UTC)

I was given permission to add it. Articles are chosen randomly, but this is because the pre-Main Page AFI was constructed this way. Now that the project is much more active and has a lot of editors overseeing it, I don't believe there is any reason to have it that way anymore, and I don't think there is a consensus for that. In fact, none of the currently scheduled articles were chosen randomly. Confectionery, mechanic, and the next AFI which goes up in a few hours, frozen yogurt, were all chosen manually. (Also, the article was added to week 48, not week 46). NewAccount7295 (talk) 21:33, 1 November 2026 (UTC)

@NewAccount7295: one user agreed, that's not consensus. I'm not opposed to Chihuahua featuring, I just think that articles should be chosen randomly or in the order of when they were first nominated. Sahaib (talk) 21:41, 1 November 2026 (UTC)

@NewAccount7295 No, you weren't given permission to do anything. The discussion you linked to literally contains a hatnote that says If three net members elect to reselect the article, then the article will be reselected. Two people are not three people, especially if you're the nominator.

If you believe the process should change and the articles shouldn't be selected randomly, then you need consensus for that. You are welcome to add that question to he RfC. Bremps' original proposal from the get-go has always been that the selection will be done by RNG, as it is currently. [...] No one gets to play favorites or horse-trade support votes. I believe that otherwise, this will be a major issue as everyone tries to get their pet article in the line for improvement. This is exactly what you are doing.

You have repeatedly added articles you've nominated to the queue after being told by multiple editors that it is neither necessary nor welcome. I've reverted your addition yet again. Please stop. If you disagree about how the process is run, you create a proper discussion.

Sahaib, thank you for pointing this out. FaviFake (talk) 21:47, 1 November 2026 (UTC)

I feel like this project is becoming increasingly contentious, when it shouldn't be at all. I'm not happy that it looks as if I've done that. I've only added two articles I nominated. I'd prefer not to do that. You are right that it says "If three net members elect to reselect the article, then the article will be reselected", but I think this was a rule established before AFI became TWAFI.

While I agree with Bremps' concern, I don't think articles should be chosen randomly. However, I think this is something that can be resolved in a separate RfC, not the upcoming one. If you look at Wikipedia:Articles for improvement/Schedule, you will find that all of the articles there were chosen manually. To avoid the issue Bremps' proposed, I'd suggest a rule for AFI that disallows people promoting articles they have nominated unless they have permission from several users to do so. I believe it would mostly solve that issue.

The performance of mechanic as an AFI showed that how interesting a topic is may affect activity. It will take more AFIs to determine that for sure, but that is a hypothesis.

Alluding to that, I feel like some of the articles being chosen (not all of them) aren't the most interesting (such as axle, which was scheduled before Chihuahua). Meanwhile, there are some very interesting articles on the AFI list.

Now, @FaviFake @Sahaib @Dualcerb, not regarding the fact that I nominated it, do you believe Chihuahua (dog breed) would be an okay choice for week 48? It looks like a very popular subject, with a very high number of pageviews, something new editors would take an interest it. We can still run it if there we get three net votes to do so. NewAccount7295 (talk) 22:02, 1 November 2026 (UTC)

Even if it's an older rule, it's still a perfectly reasonable rule. I actually think it's too weak, as three editors should not be able to change an article now that it's on the Main Page.

I think the criteria for changing the randomly selected article should be the same as the criteria for approving a nomination. These "nominations for changing" (or whatever we want to call them) should happen on the /Nominations page and should be subject to the same criteria as any other nomination, which will be updated soon per § Too many nominations given the new activity on AFI. I don't think the nominators should be disallowed from suggesting their own article as long as there is an actual discussion.

I agree that most articles chosen randomly by the bot should be changed manually.

I think it's a decent choice, not great but also not bad either. I do not think we need to discuss an article that's so far into the future at this time, and, as I said, I think the 3 supports rule is outdated just as much as the current nomination rules.

What do you think of my suggestion to match the criteria for approving? FaviFake (talk) 22:14, 1 November 2026 (UTC)

I think it is a good idea on paper, but I'm concerned that a formal discussion with a certain number of votes that would have to take place to change the article could (ironically) make things even more contentious.

Simply opening a discussion on this page and gaining consensus to do so might be a better route. Maybe we should have open talk page section for each week and discuss things there. NewAccount7295 (talk) 22:19, 1 November 2026 (UTC)

a formal discussion with a certain number of votes could (ironically) make things even more contentious.
How so? The more a process is deterministic, the less people will be concerned about the result. FaviFake (talk) 22:27, 1 November 2026 (UTC)

I'm willing to try it. I just thought of how such discussions could turn into debates, but other venues could cause that as well. NewAccount7295 (talk) 22:29, 1 November 2026 (UTC)

Yes, that is ... precisely the goal. To have a proper debate about which article should appear on the first row of the Main Page. How else would that be decided, boldly? FaviFake (talk) 22:37, 1 November 2026 (UTC)

Since there seems to be no opposition to the article itself being ran that week, shall I change it back to Chihuahua? NewAccount7295 (talk) 16:21, 2 November 2026 (UTC)

Are you serious? After everything that has been said? There are still only two supports for your article, and it is still your article. Please be patient. FaviFake (talk) 17:06, 2 November 2026 (UTC)

I really don't care about what week 48's article is. dualcerbtalk 17:11, 2 November 2026 (UTC)

One reason in the project getting more contentious may well be editors overruling fundamental aspects of the project without any discussion because they ”don't believe there is any reason to have it that way anymore, and don't think there is a consensus for that.” Again, without any discussion. Just a thought. ~~ AirshipJungleman29 (talk) 08:23, 2 November 2026 (UTC)

+1 There's no reason to not do random selection from this point on. Let's chalk up the first two or three times of non-random selection to the growing pains of getting a new project off the ground.

There was a case for non-random selection when the pool of articles had some non-ideal entries. Those have since been removed and this shouldn't be a concern going forward. If it is, we should remove the problematic article in the queue and randomly draw the replacement.

I say we do random selection for the very next one to build a good habit early. I believe Chihuahua should be placed back into the larger pool for future draws and we should redraw next week's entry. I don't want this project to generate big fights over whose article goes first, or who jumped the queue, etc. Bremps... 21:52, 2 November 2026 (UTC)

I don't think there is any contention over the article Chihuahua itself, just how it is selected, so running it that week would be fine; it is a popular subject anyway. However, there may be other reasons for changing an article, like having one run a specific week for an anniversary, or because of it being too similar to the prior one (like what is said in the "Variety" subsection).

For week 49, I went ahead and selected Van. Pretty random, something different for variety, not something I nominated, and in heavy need of improvement. If anyone is opposed to me doing that, you may revert that. NewAccount7295 (talk) 22:09, 2 November 2026 (UTC)

WHY DID YOU GO AHE—

Ahem.

NewAccount7295, would it be ok with you if we used {{Random number subst}} when there is consensus from enough people that the AFI bot picked a bad article? This template is truly random because a bot will tell us the result. If that's alright with you, I will replace "Van" with the template, and a bot will pick a number between 1 and 99 (the current number of AFI candidates).

If there is consensus among a lot of editors that even this new article sucks, then we can use the same method a second time. FaviFake (talk) 22:22, 2 November 2026 (UTC)

Great idea to start using that template, I didn't realize it even existed. Do you have any issue with van though? I am okay with replacing it if that is what the users here want, but I am wondering. NewAccount7295 (talk) 22:49, 2 November 2026 (UTC)

Yes, I have an issue with it because it was not picked randomly, thereby not giving a fair chance for those who nominated other articles.

Here we go: (check the history to see if the bot did its thing) 30 FaviFake (talk) 22:56, 2 November 2026 (UTC)

@NewAccount7295 The bot did the thing and the result is 30, Chain! FaviFake (talk) 23:59, 2 November 2026 (UTC)

It looks like it could have room for improvement but are we sure it will be interesting enough for that week? NewAccount7295 (talk) 00:01, 3 November 2026 (UTC)

Actually, we've already featured two similar ice cream articles, people would get mad if we made two articles about mechanics and a mechanical component. Since these are the most critical weeks for AFI, I'm willing to roll the dice again. 26 FaviFake (talk) 00:08, 3 November 2026 (UTC)

The answer to that would be rapids. What do you think? NewAccount7295 (talk) 01:49, 3 November 2026 (UTC)

great find, go for it! Kowal2701 (talk, contribs) 14:01, 3 November 2026 (UTC)

Wait, in all this mess I didn't realise the bot hadn't picked yet! Why in the world did you create the page BEFORE the bot? I assumed you didn't like the bot's pick but no, you decided the article before the bot. (Please don't do this ever again without consensus.)

The bot just chose mass media so we're going with that unless enough people want to change it. I see no problem with it. FaviFake (talk) 18:45, 3 November 2026 (UTC)

Let's chalk up the first two or three times of non-random selection to the growing pains of getting a new project off the ground.
Yes, exactly. Most of the nominations removed by Dualcerb were terrible and I think they did a great job with the list, and now we even have a formal process for nominating an article for removal from the list (for example, see the most recent nominations, those with the red icon), so we don't even need to worry about the process: if someone wants to remove an article, they nominate it like any other.

I think we need to elect Bremps as the founder and president of AFI; just like Jimbo seldomly comments on the process he created. (This is a joke, of course. (or is it?)) FaviFake (talk) 22:16, 2 November 2026 (UTC)

Well, I've found my prime minister. Bremps... 00:13, 5 November 2026 (UTC)

Y'all have my ballot, provided Chaotic Enby runs as speaker of the house. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 01:44, 5 November 2026 (UTC)

Yeah, we must prioritise explain xkcd editors! FaviFake (talk) 01:48, 5 November 2026 (UTC)

uh... 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 01:50, 5 November 2026 (UTC)

Bremps, this is your last chance to prepare for the coup. FaviFake (talk) 01:53, 5 November 2026 (UTC)

screw this, I'm voting for Mr. Starfleet Command 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 01:58, 5 November 2026 (UTC)

and I'm nominating him for renaming! FaviFake (talk) 02:00, 5 November 2026 (UTC)

7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 02:13, 5 November 2026 (UTC)

Dammit, I wish WP:RFC/N hadn't been shut down! FaviFake (talk) 02:35, 5 November 2026 (UTC)

Variety

The third selection of Frozen yogurt seems too similar to Confectionery store aka Candy store -- essentially about more sweet food popular in the USA. And I see that there's a stack of nominations for tools and furniture like axe and chair. These have merit but will seem monotonous if they come in a wave. There should be some curation to ensure that there's a mix of topic types. This is what is done for each month's FA selection, as I understand it. And DYK does something similar. Andrew🐉(talk) 09:52, 2 November 2026 (UTC)

Your suggestion is similar to NewAccount7295's replies above. However, note that, once approved, the articles are moved into the candidate list, which contains hundreds of articles. It is unlikely that articles nominated in the same period will be selected by the bot one after another. Still, I think we should ask a question about curation in the RfC to settle the matter. Note that a separate question on whether articles about regional topics should be successful is the first place is already planned. You can participate workshopping the RfC, which is being discussed in a section above. FaviFake (talk) 10:13, 2 November 2026 (UTC)

NewAccount7295 points out that all three main page selections have been manually curated and so they are not random. The old process has been destroyed and we are now shaping a new one. Andrew🐉(talk) 10:50, 2 November 2026 (UTC)

We are still following the old process, which allowed random selections to be overturned according to the opinions of a certain number of participants. Now that the project is ten times larger, that number is naturally too low to be effective: we now need to decide whether it should be updated as a result or scrapped entirely, and such decisions are made through an RfC if the selections appear on the main page. FaviFake (talk) 11:15, 2 November 2026 (UTC)

Where is this overturning process documented? The instructions at Wikipedia:Articles for improvement/Articles just say

• Successful article nominations are added to the end of the list
• Each week, one article is randomly chosen from this list and scheduled as the weekly collaboration.
• Articles are AFI entries from 00:00 UTC Monday to 23:59 UTC Sunday, unless otherwise indicated.

(Note: These are not firm rules, the process may be adjusted over time)

Andrew🐉(talk) 11:25, 2 November 2026 (UTC)

You would be looking for the policy on consensus-building. ~~ AirshipJungleman29 (talk) 11:38, 2 November 2026 (UTC) Struck: misunderstood the question.

It is documented in the week's article discussion:

Place any comments or concerns about the selection here. If three net members elect to reselect the article, then the article will be reselected and the former one removed from the list of articles.

I agree that it should also be noted in the article's list, I have now done that. FaviFake (talk) 12:30, 2 November 2026 (UTC)

People seem to have trouble counting to three. Andrew🐉(talk) 14:34, 2 November 2026 (UTC)

We probably need to cut more articles.

We are already at 100 articles again. I think that we should start with a separate subpage of the nominations page (similar to WP:RFPPD) to also nominate articles for removal from the list. This would allow people to get major consensus for removing articles instead of asking on this talk page. dualcerbtalk 15:44, 2 November 2026 (UTC)

To help, I divided the list into a section for articles approved before AFI was added to the Main Page, and after. NewAccount7295 (talk) 15:47, 2 November 2026 (UTC)

We probably need to cut more articles why? what detriment will we have by using a larger list? 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 16:07, 2 November 2026 (UTC)

I think that having the bot pick articles is better than having people pick manually as it causes less arguments. Shortening the list would make the bot pick an article that most people agree would be good for the main page, which would make people content instead of changing it manually. dualcerbtalk 16:16, 2 November 2026 (UTC)

I also don't see the issue of having a long list, they all gain consensus beforehand, bot can pick randomly. Ofc some might languish in the list for a while but oh well (could skew the bots decision towards older entries) Kowal2701 (talk, contribs) 14:00, 3 November 2026 (UTC)

The problem is that we can only feature a small number of articles on the Main Page. If every week we featured one article and approved 20 more, the list would spiral out of control and the bot would pick between worse candidates. We wouldn't be able to monitor these articles and make sure they still need improvement.

Instead, imagine a scenario where the list only contains 20 articles, and every week we add one or two, the absolute crème de la crème of our nominations: all of the bot's picks would be utterly perfect, and we wouldn't need to set up a process to overturn the bot's choice (which has happened for every single Main Page selection so far due to the previous criteria being too lax, and has proved very controversial). FaviFake (talk) 14:12, 3 November 2026 (UTC)

Should we just stop allowing nominations until we figure out new criteria and a new process? We seem to have plenty of articles in the backlog to last us. Maybe we shouldn't have a big list at all and instead just nominate articles for each specific week... Eman7blue42 (talk page | recent edits) 16:47, 2 November 2026 (UTC)

Nominating articles for specific weeks could potentially be a good idea. (Although it might cause arguments...) dualcerbtalk 16:50, 2 November 2026 (UTC)

Or another way to see it is it might cause consensus building. Eman7blue42 (talk page | recent edits) 16:54, 2 November 2026 (UTC)

We don't need to stop nominations, the new process recently went into effect so it will be more rigorous.

However, I have updated the {{AFI nom}} template so now it accepts nominations to ratify a previous bold removal after an article had been approved under the old process. It also accepts nominations to remove an article from the list, so now @Dualcerb can finally have a way to propose removing an article without so many people complaining.

Please see WP:AFINOM § Bed and breakfast to see how it looks. I will update the instructions later to accommodate for these new types of nominations. Let me know what you think! FaviFake (talk) 19:00, 2 November 2026 (UTC)

I think its great for now. Eman7blue42 (talk page | recent edits) 19:26, 2 November 2026 (UTC)

Replied to one, looks good, though it'd be nice if text flowed around the box properly. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 19:34, 2 November 2026 (UTC)

I think I fixed it, I had mistakenly left a {{clear}} template at the end of my signature. FaviFake (talk) 19:53, 2 November 2026 (UTC)

how do you do that? 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 20:03, 2 November 2026 (UTC)

Lol, my bad, I meant to say that I left it intentionally but mistakenly thought it would work well with DiscussionTools (the "Reply" button). It did not in fact work well. FaviFake (talk) 20:08, 2 November 2026 (UTC)

It did not in fact work well is basically the anthem of the reply tool. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 20:59, 2 November 2026 (UTC)

Better to wait until the upcoming RfC decides on the Main Page rotation frequency. No point shutting off the inflow and then finding the community wants a new article to appear every day. ~~ AirshipJungleman29 (talk) 14:59, 3 November 2026 (UTC)

The person you replied to never suggested shutting off the inflow. Besides, even if they want a new article every two days (which i think is highly improbable), the current process still allows more nominations to be approved than we can feature. FaviFake (talk) 17:14, 3 November 2026 (UTC)

Curious what you think “cut more articles” means. ~~ AirshipJungleman29 (talk) 01:57, 4 November 2026 (UTC)

It means remove articles from the Holding area, not prevent people from creating new nominations. ​ FaviFake (talk) 02:00, 4 November 2026 (UTC)

Yes, and thus shuts off the inflow from the holding area to the main page. ~~ AirshipJungleman29 (talk) 09:38, 4 November 2026 (UTC)

What? Articles from the holding area would not stop being featured on the main page if we removed a few articles among the one hundred articles listed. FaviFake (talk) 18:14, 4 November 2026 (UTC)

Arguments to avoid re nominations?

Maybe when the norms wrt !voting on nominations are cemented there can be a page like WP:ATA to point people to if necessary?

Just my 2c, but I don't think arguments like "I don't find this interesting" should hold much weight, it's whether an article appeals to a broad audience, either because it's commonly known or has a range of aspects (i.e. sections). Mundane =/= boring, such topics can be intriguing precisely because of their mundanity ("what on earth does an encyclopedia article on dustbins look like"), and the main page needs more serendipity like this. Also, I know pageviews are somewhat useful, but just servicing the most popular articles really isn't in keeping w the spirit of an encyclopedia, imo the focus should more be on how vital the article is (though not necessarily deferring to WP:VA because that's a bit of a basketcase). And these articles typically get neglected, so TWAFI can fill a big gap in our coverage while onboarding newcomers. Thoughts? Kowal2701 (talk, contribs) 17:26, 4 November 2026 (UTC)

I agree with you that we are starting to see some questionable arguments on nominations. If there is an argument to avoid it might be something like "Oppose I don't think this article should be featured because it is about a brand". There is currently a "companies" and "products" section; it doesn't feel fair to reject such articles solely on that basis; that question will be asked at the RfC. I don't think we need an "arguments to avoid" page though. The better option would be to give some guidance at the top of the page in a single paragraph. NewAccount7295 (talk) 18:08, 4 November 2026 (UTC)

There is currently a "companies" and "products" section AFAICanTell, it's a legacy from the older form of AfI that was never removed. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 18:43, 4 November 2026 (UTC)

I know that, but it is still there for users to add nominations to, and they are doing that, just to be told those articles shouldn't be there in the first place. NewAccount7295 (talk) 18:45, 4 November 2026 (UTC)

Then, Proposal: remove the Companies and Products sections until the RfC has passed a decision on them. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 20:03, 4 November 2026 (UTC)

People will still nominate companies, just in other sections. I have no opinion on whether the section should be removed, but until we add something to the criteria, people will continue to post them. And we can only edit the criteria after the RFC has ended. FaviFake (talk) 20:13, 4 November 2026 (UTC)

I renamed the section to Food and drink. Furbchess (talk) 21:15, 4 November 2026 (UTC)

a page like WP:ATA to point people to if necessary?
Great idea! Please feel free to create that. It could also be added below the current criteria, in a "non-criteria" section or similar (maybe collapsed; we'll figure it out along the way!)

norms wrt !voting
Note that AFI nominations, similarly to RfA, receive votes, not !votes, as they all count equally. (There's no time for formal closures based on consensus.)

arguments like "I don't find this interesting" should hold much weight
No, we need articles to be interesting so that newcomers will want to edit it. Dustbin is an example of an interesting article. Maybe "interesting" isn't a great word here, and I agree that mundane topics can qualify. We can reword the first criterion if you have some ideas on a better adjective or way of expressing this quality. FaviFake (talk) 18:39, 4 November 2026 (UTC)

maybe "intriguing"? Like w DYK we want the main page segment to hook people in, make them curious, but w TWAFI it's the subject/title itself. My point wrt that was that "I personally don't find this interesting" isn't a good one, since it's projecting one's own interests on everyone, and the sample size of people at /Noms is tiny and unlikely to be representative Kowal2701 (talk, contribs) 18:57, 4 November 2026 (UTC)

Automatically notify the article about its nomination

Since every other XfD venue already automatically notifies the related pages (e.g., RfD notifies its target, AfD notifies the page itself, and TfD notifies pages that use the template), I propose we make the bot add {{subst:AFI notice}} to the talk page of a nominated article. I assume this could be done by checking for the ====Article name==== inside a new ==== Level-4 heading ====. The {{afi notice}} template looks like this:

Example

Hello! The Articles for improvement article has been nominated to be featured on the Main Page for seven days as This Week's Article for Improvement. You are welcome to participate in the discussion at WP:AFI/Nominations § Articles for improvement and help us decide whether it meets our criteria. If the nomination is successful, Articles for improvement will be added to our holding area and you will be notified again once it is scheduled to appear on Wikipedia's Main Page. For the Articles for improvement project, ~~~~

If there's consensus I'll ask the bot operator. FaviFake (talk) 18:47, 4 November 2026 (UTC)

I was also planning to re-enable the shuffling of the sections, which used to be performed by the bot every day (simply, the last two sections like Geography and Arts would be moved above all other sections). FaviFake (talk) 20:15, 4 November 2026 (UTC)

Looks pretty good. It has my support. Bremps... 22:57, 4 November 2026 (UTC)

This looks great! Perhaps could be more concise though. Eman7blue42 (talk page | recent edits) 17:13, 5 November 2026 (UTC)

Thanks! I tried to make it shorter, let me know what you think or what else should change. FaviFake (talk) 04:04, 6 November 2026 (UTC)

Fixing the edit notice

Updating edit notice automatically

Does anyone with technical skills know how to update Template:AFI edit notice/article info automatically? For the last two weeks I've manually updated it when the AfI changed, which led to the old edit notice appearing on the new AfI until I got to it. InfernoHues (talk) 02:55, 4 November 2026 (UTC)

Would it be easier if we made you automated instead?

Maybe there is, but I'm sure it would take time to code and organise. We still have to update two pages just to change the TWAFI for a given week, so on the technical side we're definitely not there yet. FaviFake (talk) 03:08, 4 November 2026 (UTC)

I wish :P

It's a lot more important for this one to be automatic since it can't be done in advance. Changing the TWAFI can be done weeks ahead of when it runs, but the edit notice can only be changed when the article actually switches on the main page. InfernoHues (talk) 03:10, 4 November 2026 (UTC)

{{Interval}} can be used with a {{#switch}} parser function. For example:

{{#switch:{{interval|n={{#time:d F Y}}|format=time|1 November 2026|8 November 2026}}|1-2={{AFI edit notice/article info/1 November 2026}}}}

More than two valuies can be passed to the {{Interval}} template and the switch statement can test for 2-3, 3-4, and so forth, or you could just have multiple switch statements, each with the start and end dates for a given article. The text could be included directly in the switch statement rather than in a template, but it might be easier to manage in a template (plus I think it would avoid wikitext restrictions with template parameters). isaacl (talk) 04:42, 4 November 2026 (UTC)

@Isaacl Thanks! Would you be able to implement that, or even a rough version of it? FaviFake (talk) 17:38, 4 November 2026 (UTC)

I think it would be better for someone involved with the day-to-day operations to do the implementation. The rough version is to copy and paste the example I provided, and repeat for the next week with the dates and template name changed, so the changeover will happen automatically. (I'm not sure if extra white space will get added if there is a newline between the {{#switch}} statements. Some testing on a test page would be good, before putting it on the edit notice page.) isaacl (talk) 18:25, 4 November 2026 (UTC)

Larger warning

I tested what TWAFI would be like as an unregistered temp editor. There is an edit notice, but it's a polite message telling me that there are "suggestions" on the talk page. It's a pleasant but easy-to-overlook yellow.

Could someone please make a mock up of a more aggressive popup? Preferably, it's bright red, and titled PLEASE READ in bold font. It should inform users of our WP:LLM policy (in short, don't) and unequivocally, enthusiastically tell them to "visit the talk page".

Additionally, it would be great if it didn't vanish as soon as the editor began changing the page. Is there any way for it to briefly intercept the user as they hit publish?

Best, Bremps... 22:54, 4 November 2026 (UTC)

Is Template:AFI edit notice/new not showing? I don't know how edit notices work, do they have to be added manually to each article? If so, the notice only showed on the first one I think. I think you need to be a template editor to add it, so if anyone has that, it'd be great if you could do so. InfernoHues (talk) 23:05, 4 November 2026 (UTC)

I don't see that on my end. I do think that we should have the more aggressive popup, though; this design is sleek but we really need to grab attention. Bremps... 23:12, 4 November 2026 (UTC)

InfernoHues Crap, Bremps is right, the notice is not shown. Try editing the article; there's just the notice produced by the {{refideas}} template on the talk page. FaviFake (talk) 23:15, 4 November 2026 (UTC)

I made a request at ERRORS to add it. InfernoHues (talk) 23:17, 4 November 2026 (UTC)

I have requested for the creation of the page and it was granted. Thank you for spotting this! |)are2|_eap (Talk) 02:13, 5 November 2026 (UTC)

Template:Editnotices/Page/Frozen yogurt does not exist currently - probably someone forgot to do it. 7amiþ = reform(solidarity).__init__(t, c) # Leave a talkback, TWAfI's wild rn 23:21, 4 November 2026 (UTC)

Moved to WP:VPT § The editnotice for the current AFI is not updated automatically

 – Since the rest of the discussion belongs there. FaviFake (talk) 00:22, 5 November 2026 (UTC)

The bot isn't botting

The bot seems to have stopped doing its job (see WP:AFIN's history). It's supposed to archive nominations within 4 hours of their last comment but there are days-old nominations still listed. Maybe we should start doing things manually if it doesn't wake up soon (i've already contacted the botop, hasn't responded yet). FaviFake (talk) 03:28, 6 November 2026 (UTC)