Huh, only 91 articles? That's actually pretty impressive, and axing it definitely possible, as long as people continue to work on it. ARandomName123 (talk)Ping me!18:41, 17 March 2024 (UTC)[reply]
Oh, looks like the dream is dead (for now). There's been a big uptick in new additions over the last few days, and now we're sitting at nearly 500. ARandomName123 (talk)Ping me!15:32, 27 March 2024 (UTC)[reply]
On March 1 and 2, I was trying to patrol the category and made great headway. I've probably knocked at least 30-40 out. A lot of the 'discovered' unsourced articles are sort of in batches of related articles, probably mass created... if we can find a good source for one of them, that should make serious progress on the category. ForksForks (talk) 01:57, 31 March 2024 (UTC)[reply]
Happy spring, party people! I hope everyone is doing swimmingly.
Headline: As a result of the first annual drive, which was a smash success, we have cleared 18,096 articles between today and January 4th, the equivalent of what we might do in ~9 months (based on previous averages). For yourself and your fellow editors, please clap.
Highlights: All categories from October 2007 to May 2008 have been emptied! We are finally down to backlog that is less than 16 years and below 100,000 unreferenced articles for the first time...well, since we went over 100,000 articles.
Low-hanging fruit: June 2008 is hovering at an enticing 88 articles left, a very lucky number.
High-hanging fruit: Everyone's favourite BFC (Big Friendly Category), December 2009, is a willowy 12,925 articles as of this writing - half the calories, but just as much body as the original. The other high-hanging fruit are, still, the Frustrating Five (name open for revision): February 2016 (921), April 2019 (1095), May 2019 (2250), June 2019 (4699), and September 2020 (1432); September 2020 had the lowest percentage of change from January to April. Godspeed to anyone working on these.
Announcements: The village pump discussion started by comrade in arms @CactiStaccingCrane is ongoing, see above.
Results: July 2012 edged out August 2012, 534 to 539.
New challenge: not one, not two, but THREE ties: September/October 2010 (314); April/May 2012 (364); and September/October 2018 (395). Some kind of triple Jellybean (my cat) in a hat prize awaits the person who tips all three of these.
@Kazamzam Nice. The consensus around that discussion seems to be draftifying articles, so I will propose that articles should be formally draftify if they don't have an inline source. CactiStaccingCrane (talk) 13:45, 5 April 2024 (UTC)[reply]
@Boleyn - it's part of the project banner text {{WikiProject Ecuador|importance=Low|imageneeded=no|unref=yes}} We would, I believe, need to go in and address each one manually. Doable if time-consuming. Kazamzam (talk) 19:55, 4 April 2024 (UTC)[reply]
Greetings, Last week at Category:Articles lacking sources from June 2008, for the first dozen or so articles, I added a "Notice" tag (being Bold). Today, I fine-tuned it a bit into a transcluded "Ambox", just the first 3 articles. With the four criteria more visible, hoping it helps both beginners and a call-to-action for more advanced editors. Asking for feedback and discussion here. Regards, JoeNMLC (talk) 14:40, 8 April 2024 (UTC)[reply]
Please improve this Wikipedia article by using multiple sources that meet four criteria. The sources should be: (1) reliable, (2) secondary, (3) independent of the subject, (4) talk about the subject in some depth. Encyclopedic content must be verifiable through citations to reliable sources.
The four criteria seem to be better suited for articles tagged for notability issues. Citations just have to be reliable. Primary, nonindependent and passing mentions are all usable (within wp:primary).
Obviously, sources that satisfy all criteria are preferred, but we don’t want people avoiding adding an otherwise fine reference just because it doesn’t have sigcov or something. ARandomName123 (talk)Ping me!14:57, 8 April 2024 (UTC)[reply]
You're conflating notability and verifiability here. Citations in an article are there to verify the content of the article: they have to be to reliable sources, but need not meet your three other criteria. "Significant coverage in independent, secondary reliable sources" is how we judge notability, but that's not something readers need to worry about. If sources showing notability need to be presented somewhere, the talk page is probably best. But in the vast majority of cases that doesn't come up until notability is challenged at WP:AFD.
And just to add, if what you're concerned about is notability, there are already a range of tags for that. {{GNG}} is probably the closest to your wording. – Joe (talk) 10:20, 9 April 2024 (UTC)[reply]
I don't think this addition to the existing maintenance tag is necessary or helpful. The second sentence of the unreferenced tag is 'Please help improve this article by adding citations to reliable sources.' with the link to the 'Referencing for Beginners' essay so I think this is a total non-issue that is just creating more work. Why do you feel this is necessary? Also having multiple maintenance tags removes the automatic search links to Google, Books, Scholar, JSTOR, etc. which I find very convenient so as an experienced editor, it is making my work 1% more difficult. Also+also agree with the above notes about notability and verification from Joe and ARN123. Kazamzam (talk) 20:18, 8 April 2024 (UTC)[reply]
I agree with Kazamzam. We already have tags that cover these various issues. I understand the desire to have more refined tags that cover more subjects but I think this ultimately is overly confusing for readers and editors. Also consider that we track various types of issues by categories and a tag like this muddies that approach. Stefen Towers among the rest!Gab • Gruntwerk16:16, 9 April 2024 (UTC)[reply]
Thank you, everyone! - For clarity & explaining why "not" to do what I attempted. Today, I did "undo self" for those June 2008 unref. articles. Going forward, I'm adding parts of this discussion to my "Article cleanup" notes. So as not to forget. Cheers, JoeNMLC (talk) 12:20, 10 April 2024 (UTC)[reply]
Empty references sections?
Is there any guidance on whether or not unreferenced articles should have: "==References==, {{reflist}}"? It's pretty convenient not having to type it out, especially if I'm using the visual editor. ARandomName123 (talk) 22:54, 12 April 2024 (UTC)[reply]
I see people add these in relatively often. Sometimes the random article patrollers that add unreferenced tags to stuff throw them on. I don't really ever see people removing them, so I assume it's okay to do if you want. ForksForks (talk) 01:22, 14 April 2024 (UTC)[reply]
I add them, as when looking through to add references, I am looking for easy references to add, and to see if there is a references section. This way at least one is completed, even if I struggle with the references themselves. I have had them removed by a couple of editors just seeing it as an empty section, though for me this is an essential section awaiting completion, as opposed to an unneeded section. Boleyn (talk) 13:05, 14 April 2024 (UTC)[reply]
I think somebody already suggested this, but I can't remember where: it'd be great if {{Reflist}} on a page with no references displayed something useful. "This article does not contain any inline citations or footnotes. Please help improve this article by adding some" – something like that. – Joe (talk) 10:03, 20 April 2024 (UTC)[reply]
I've been removing these because I don't think they help readers and could cause confusion. The idea to make {{Reflist}} smarter may help. In the meantime, there is {{Empty section}} which has a |find= option that could help with sourcing. ~Kvng (talk) 22:05, 29 April 2024 (UTC)[reply]
Backlog
Today it has reached 94,999! I realise it may have bounced back by the time people see this, but it is nice to see it fall to this. Unref BLPs now also under 1000. Boleyn (talk) 07:48, 20 April 2024 (UTC)[reply]
Boleyn It's going to reach 94500 soon... At this rate, we cite 500 articles every week. This means that it will take 3 years and 33 weeks (3.5 years) to completely clear the backlog.
How do we feel about two backlog drives a year? One in February and one in June/July (TBD)? Assuming each drive reduces the backlog by around 10k, each drive is worth around 20 weeks, meaning we can bring the time to completely clear the backlog down to just 2 or so years. ARandomName123 (talk)Ping me!16:54, 26 April 2024 (UTC)[reply]
Also we don't have many sub-categories based on topic, or a search box for keywords - I don't know how easy it is to do those or if they exist and I am just missing them! Boleyn (talk) 09:37, 26 April 2024 (UTC)[reply]
They do exist! If you take a look at Category:Articles lacking sources there is a yellow box showing articles based on certain topics. You can also type your search term in the normal search box, then paste incategory:"All articles lacking sources" next to it. For example, if you want to find banana articles in the backlog, you would search banana incategory:"All articles lacking sources". CactiStaccingCrane (talk) 14:13, 26 April 2024 (UTC)[reply]
Ok this is interesting... Currently there is 93,481 uncited articles, so a decrease of 1000 articles in 5 days. So right now we are working on these articles at the rate of 200 articles per day. If that's the case then we only need to take 467 days or around 15.5 months to finish the backlog. CactiStaccingCrane (talk) 15:48, 30 April 2024 (UTC)[reply]
So I (roughly) plotted the monthly count starting in January 2022, and assuming a backlog drive every 6 months, we are headed for a completion date of around early to mid 2026. Without any backlog drives (so assuming FEB24 didn't happen), we would only finish up by January 2030. ARandomName123 (talk)Ping me!18:59, 30 April 2024 (UTC)[reply]
It's always nice to work on something when there is some momentum. It looks like it could dip under 93,000 today. I'm aware I have been using the 'incategory' search to pick up some low-hanging fruit, and that I'm not working at the moment, so I am removing more from the category than I would be able to keep up. Maybe an extra editor or two working on it could help, if people know there's a big push at the moment. Boleyn (talk) 10:30, 3 May 2024 (UTC)[reply]
Should we get wikiprojects to join the effort?, would someone be able to do some analysis on which projects have the most unreferenced articles in their scopes? As well as directly helping with sourcing - they could provide the best guidance on where to find good sources. If it turns out that there is any one project that covers more than, I dunno, 15% of the unfererenced articles then that could be a theme for the next drive. -- D'n'B-t -- 11:55, 4 May 2024 (UTC)[reply]
It would be great if there's like a government source that has a list of villages. That way I don't need to search for eternity for sources for each villages. CactiStaccingCrane (talk) 15:40, 3 May 2024 (UTC)[reply]
Idea discussion: Teaming up with specific Wikiprojects
Let's continue the discussion about teaming up with other WikiProjects. This came up in the context of the WP:FEB24 backlog drive, but also outside of it. Maybe a more automagical editor can do some more efficient querying but I thought I'd get the ball rolling by throwing a few queries into PetScan based on some familiar faces I've seen along the journey. Note that it's even more haphazard as I've sometimes had to experiment to pick a subcategory tree depth that makes sense.
Added percentages based on @D'n'B's suggestion above. Category total at the time of writing is 91,777 articles.
Sturmvogel 66 maybe, but both Millitary History and Football seem to be pretty live and kicking. Thanks for the number crunching Turtlecrown. Let's see what sort of collaboration (alliance? team?) they'd be most interested in. -- D'n'B-t -- 06:27, 14 May 2024 (UTC)[reply]
I had skipped over albums as I was initially confused by Category:Unreferenced album articles which states "Articles are added to this category by setting the "unref=yes" switch of Template:WikiProject Albums". This is a similar situation to the Ecuador discussion above, except the category is empty. I was trying to find out if the WikiProject already had some infrastructure to deal with articles that land in the category and if we could work with that. So I'll summarise my findings here (TLDR: probably not).
The parameter and category were added by formerly very active (till 2015) user @Keraunoscopia in 2011 as part of the B-Class article checklist.
A least one editor at WP:Albums, @Champion, was actively adding standard unreferenced templates instead in 2018, providing this search which currently shows 6,759 articles with Infobox Album and an unref template.
Deletion was NC in 2019 as the parameter is still part of the template though not actively in use. Another 2021 suggestion to remove the param went unanswered.
It sounds like the category is only still there as nobody has removed the parameter from the template. However, I see potential here in collaborating with such WikiProject tracking efforts. This could also tie in with desires from this WikiProject (eg 1, 2, 3) to have more specific topic categories of articles needing references, whether or not a project ultimately has capacity to assist. Turtlecrown (talk) 09:23, 14 May 2024 (UTC)[reply]
The problem is partly with the web-like (rather than strictly tree-like) structure of categories. The depth on your above search was set to 9 which is very high for such a large category that "contains" other large categories such as Category:Cultural heritage. The above query returns 22019 articles. Setting the depth as 4 gives PSID 28295171 with 633 articles. It works better with more specific categories, eg Category:Archaeology by location which gives 237 articles at PSID 28295201. If we're strictly talking about articles already in WikiProjects it's definitely better to use Talk page templates, which gives 80 unreferenced WP:ARCHEO articles per PSID 28295129. Turtlecrown (talk) 09:56, 14 May 2024 (UTC)[reply]
I added some more queries based on the inclusion of a WikiProject template on the article's Talk page (where a project exists specifically for that topic), but it doesn't always work and I don't know why. I also can't get categories from Category:Articles by WikiProject to work despite trying a few approaches from the PetScan docs. This approach of course assumes that unreferenced articles have always been assigned to a WikiProject. Apart from ones that might have been missed, WikiProject Albums for example doesn't claim about 100 Category:Albums unreferenced articles that are EPs and music DVDs (the remaining ~200 are discographies). Turtlecrown (talk) 12:35, 14 May 2024 (UTC)[reply]
I think that if you allow for both categories sometimes being too broad and project templates sometimes missing, then the "true" number of articles that could be said to belong to a topic is somewhere between the two, and anyway, both sets of numbers are in agreement about which are the largest topics. But I think the choice of topic to theme a backlog drive around lies in where there's the most enthusiasm. Albums/musicians/music might well be the one. -- D'n'B-t -- 17:50, 14 May 2024 (UTC)[reply]
Filter: For any of the above PetScan, I like to run with additional filter of: Page Properties: Smaller or equal, usually set to 1000, or 500 to look at small articles. These may be candidates for Merge or Prod. if not able to find references. Cheers, JoeNMLC (talk) 14:34, 6 June 2024 (UTC)[reply]
Subject matter expertise is greatly appreciated, and one-off contributions are often huge and Barnstar-worthy as well. Hoping for someone who is mathematics-savvy like @XOR'easter to weigh in on finding a reference for everyone's favorite Degenerate bilinear form. Cielquiparle (talk) 12:11, 26 May 2024 (UTC)[reply]
A shout out as well to all the unsung heroes who are quietly contributing to the total count of ~200 referenced articles per day (regardless of when they were first tagged). Cielquiparle (talk) 10:25, 27 May 2024 (UTC)[reply]
June 2024
Happy summer! I hope everyone is doing swimmingly.
Headline: Less than the drive but a notable increase from our usual month-to-month, we cleared 5664 articles since April! For yourself and your fellow editors, please clap.
Highlights: June, July, and August 2008 are in the bin! For the categories that remain, October and December 2012 had the biggest decreases (47.51% and 51.57% respectively). December 2016 also had a big drop at 38.31%.
High-hanging fruit: Everyone's favourite BFC (Big Friendly Category), December 2009, is a debonair 12,092 articles as of this writing - built for speed and for comfort. The other high-hanging fruit are, still, the Frustrating Five (name open for revision): February 2016 (897), April 2019 (1071), May 2019 (2181), June 2019 (4579), and September 2020 (1418). Once again, September 2020 had the lowest percentage of change between updates. Godspeed to anyone working on these.
Announcements: The September 2020 change, or lack thereof, is particularly concerning given that it is a large category that gets very little attention and the articles are difficult to reference. Personally I think a lot of them don't meet notability and should be merged or deleted. However, from a project-wide perspective, it is important that we are working on categories across the board, not just the largest/smallest/newest/oldest. It would be like cleaning your home and a few rooms are spotless but the kitchen is a mess. I would like to propose that there is a benchmark decrease for all categories of at least a 10 article decrease for every two month update. The momentum from the drive has diminished but I think maintaining this kind of standard is important.
Results: September 2010 beat October 2010, 286 to 303; April 2012 defeated May 2012, 348 to 352; and September 2018 narrowly edged out October 381 to 383.
Benchmark: I agree in theory about working on categories across the board. For me personally, I don't think I would do as much if I just tried to be more diffuse. It's motivating to feel like I'm working together with people on something, and just bookmarking a prominent (usually the oldest) category is the easiest way of doing that. There are currently 198 months open, so it's hard to get an overview of progress as such a diffuse group. Ten from each would be 1980 articles, which is a lot, but doable. I'll try to give it a go. (In thinking what would help me direct my focus, I looked into ways of having an automatic and compact tracker to see in overview whether each month category has hit the benchmark, but had no success yet. It's possibly an over-complication but I would be interested to know if anyone has ideas!)
@Turtlecrown if I recall correctly, September 2020 is a bunch of stub articles about streams and brooks in the German-speaking world, many of which have the same name. Absolutely maddening. Kazamzam (talk) 14:53, 6 June 2024 (UTC)[reply]
I also find it more motivating when I can tell other people are working on a category. I would support working across multiple categories, but am wondering if more people would be willing to announce "streaks" – so that other people can join in, if they are so inclined? Cielquiparle (talk) 15:36, 6 June 2024 (UTC)[reply]
Or, is there an automated way (analogous to the Page curation log) where we could see all the "Unreferenced" Category tags as they are getting removed...? Cielquiparle (talk) 15:53, 6 June 2024 (UTC)[reply]
I love this idea. It would also, potentially, enable us to see other clean-up tags that could be improved if that's something people are interested in. Would the NPP editors be able to advise on this idea? It would be a great thing to put on the home page where visitors could see the tick-tick-tick of articles getting referenced and the count going down like sand in an hourglass. Kazamzam (talk) 14:15, 7 June 2024 (UTC)[reply]
This tool can show all removals of {{Unreferenced}}, but it doesn't give any information on the date categories. If you wanted to have a list logging such changes on-wiki then you would need a bot, I'm not sure but it sounds feasible. the wub"?!"23:51, 10 June 2024 (UTC)[reply]
We are so close to getting under 90,000!!! I have been working on articles tagged for both notability and no sources which is about 3,600 articles. More than half are notable, it's just not always apparent at first glance because of the lack of sources. I have started some AfDs and merge proposals, but a lot I have been able to add sources. Hopefully this will get two chronic backlogs down at the same time, but it's a large amount!
Now down to the final two! Both are Indian village articles, which are fairly annoying to cite. I'm pretty sure they exist, but I can't find any RS for them. The villages are also sometimes spelt differently, like in Tadawas, which didn't show up in the census search under its original title of Tadavas. ARandomName123 (talk)Ping me!02:31, 27 June 2024 (UTC)[reply]
Dinosaur hunting
Alright I think I've done it. I set my record. Just found references for an article that was created on 13-June-2002. Almost 22 years of being unreferenced! This article is old enough to buy me a beer of congratulations in all U.S. states!
Any other contenders for longest unreferenced article? I'm almost mad I didn't wait two days to reference it so it could have been a perfect two years exactly. Kazamzam (talk) 22:14, 11 June 2024 (UTC)[reply]
My oldest currently seems to be The Memory of Earth, unreferenced (though an EL was added later) since its creation in 20-March-2002. The first revision even included a spoiler warning. It wasn't tagged until April of this year, and I added a ref about a week later.
Yes, there's loads! I tried searching for keywords like encyclopedia, but using 'Retrieved' incategory:"All articles lacking sources" got up hundreds with a reference, where the unreferenced tag just hadn't been removed. I have nearly finished those. Boleyn (talk) 15:19, 24 June 2024 (UTC)[reply]
I tried about 20 of the mistagged articles, but those did seem to be tagged correctly, not sure if there is a way I could use the tool better? Thanks, Boleyn (talk) 14:44, 25 June 2024 (UTC)[reply]
Perhaps this is a common idea. But why don't we just... try to keep July 2024 empty for the whole month? If only 200-400 articles will be in it by the end of the month, that means only about 10 articles are added per day that won't get removed later. That's not too bad! If the project focused our efforts, couldn't we make it so all new tags are taken care of, thereby preventing new unsourced articles from entering the backlog?
Kudos to the team that played non-stop whack-a-mole throughout the month of June 2024 so we finished sub-200 (177 unreferenced articles at last count). I reckon that means the actual number of articles tagged as "unreferenced" last month was closer to 500. Possibly even more. Well done to all. Cielquiparle (talk) 14:09, 6 July 2024 (UTC)[reply]
Next backlog drive?
It's been a couple of months since our first backlog drive, which was a great success. Since then, the project's been doing pretty well, and it's been reduced by another 10k. Given the success of the first backlog drive, should we hold another one soon (Sep/Aug?; so presumably biannual drives), hold one next year (so annual, maybe in Feb again) or something else? Open to ideas. Pinging previous editors: @CactiStaccingCrane, Broc, Kazamzam, DreamRimmer. Cheers, ARandomName123 (talk)Ping me!13:31, 7 July 2024 (UTC)[reply]
I'd be happy to help organizing another drive, the previous one had overall great participation and helped reduce the backlog significantly. After FEB24, some had suggested a yearly frequency, other 4-months or milestone-based. I think a 6-month interval is a good compromise. A few points that were raised during the previous drive and that we should address ahead:
Set up automatic update of the leaderboard at regular intervals
Reviews should be better organized, so one can see if a page has been reviewed already or not. This was raised by multiple editors during the drive.
Criteria on types of acceptable sources should be clear.
Then an extra one from me:
Strongly encourage replacing {{unreferenced}} with {{more sources needed}} or {{one source}} if large parts of the page are left unsourced. Tags are important to highlight remaining issues in an article.
I'd be game for one more sooner – as the August, September request suggests! I'll just add that we should continue the clarifications added through the drive; one was to give a point if the unreferenced tag existed and so did a source, but no in-line cites and the editor added in-line cites. --Engineerchange (talk) 14:06, 7 July 2024 (UTC)[reply]
A few thoughts:
I think a semiannual schedule would be great so we don't overextend our asks for the larger editing community. It also gives us plenty of time to review and tweak features/aspects from the previous drive, i.e. the ones raised by @Broc, the initial bugs that @CactiStaccingCrane and others identified while planning the first drive, feedback from participants, etc.
I strongly agree that the types of acceptable sources should be more explicit - it might even be worth having a two-source minimum to weed out people who are tempted to do drive-by sourcing to excel on the leaderboard without really contributing to the project overall. Having a tutorial (perhaps even a video) would probably be really helpful.
Advertising the drive on high-traffic places (the Signpost being a big one) is also important to consider and I believe was largely overlooked last time.
I think August is too soon if we're just beginning the pre-drive discussion (on the project page at least) but September is doable if everyone gets behind it as an idea. I think if we can agree in the next...72 hours that September 1 at 00:00 EST is our start time, we can really get in gear for planning, and if not, October will have to be in (permanent schedule to follow from there?). Let's also set a goal for the number of cleared articles - 15,000 sounds good to me! Cheers, Kazamzam (talk) 17:47, 7 July 2024 (UTC)[reply]
I like the two source minimum requirement, so we don't just move it from this queue to the one source queue. I think a video tutorial is a bit too much of an ask. An explicit statement that you need to add a least two sources with in-line citations if there is no existing source (or the citations are not in-line) is sufficient enough instruction. --Engineerchange (talk) 18:04, 7 July 2024 (UTC)[reply]
Well FEB24 worked out with a similar timeframe for August, but September sounds fine. Any objections? If not, we can start creating the drive pages. We did have a mention in the Signpost last time in the News and Notes section, and I feel we might be asking for too much if we want a larger presence - I don’t think many drives are advertised on the SP in the first place.
I don’t think having a two source minimum requirement is the way to go, mostly because some stubs only need one source to support everything. Yes, this will increase the backlog for articles with only one source, but that backlog is probably easier to deal with than this one. ARandomName123 (talk)Ping me!23:27, 7 July 2024 (UTC)[reply]
The way to encourage (not require) a two-source minimum requirement is to award 1 point per source added. We just need a different way to track multiple sources added per article. Cielquiparle (talk) 05:20, 8 July 2024 (UTC)[reply]
Not sure if this the right time to throw in another suggestion, but I think there could be extra points for extra old tags. (eg. article has been tagged for 10+ years). -- D'n'B-t -- 06:00, 8 July 2024 (UTC)[reply]
Maybe we could create different "tracks" or different paths for different types of barn stars. I had this idea for a Treasure Hunt Citation Barn Star, where you had to demonstrate citations in each of the following categories:
I like this idea! It would be a nice way to bring something to other wikiprojects, instead of collaboration just being about asking for something. Turtlecrown (talk) 21:53, 8 July 2024 (UTC)[reply]
I’m fine with this idea. We could just have editors leave a different edit summary. Though now that I think about it, we’ll just end up moving articles to the additional citations backlog instead of the single source backlog. ARandomName123 (talk)Ping me!09:11, 8 July 2024 (UTC)[reply]
Maybe something like a two-source minimum and/or one source + one external link? I think saying one source is sufficient not only moves things from the single source backlog but it allows a large portion of the article to go unreferenced if it's at all longer than a stub. I'm less concerned about the backlog than the reference quality; articles that state a basic location and then give a bunch of utterly unconfirmed facts being marked as 'cited' because the basic location is confirmed really grind my gears. While that's a personal view, I don't want to encourage drive-by editing from this drive and I think that's a valid concern if we're going to have a large number of participants in a somewhat competitive mindset with minimal oversight or instruction. Not saying this as a doomsday foreteller but because I have unfortunately seen this before.
Perhaps instead of trying to deter with guidelines that might not fit every article, we could award more points for good referencing. I did a lot of work on Die Kuranten back in September 2023 and I think if someone puts in that kind of overhaul effort because of our drive, it would be great to reward or at least acknowledge it, which so rarely happens. Kazamzam (talk) 14:02, 8 July 2024 (UTC)[reply]
OK. I could live with the two-citation minimum. (Not so keen on the external link exception, as it seems open to abuse.) If nothing else, finding two citations tends to improve accuracy because it allows you to cross-reference.
The Reviews part of the last Backlog drive was very instructive and revealing; it definitely gave you a sense of the range in quality of citations, also by editor. It wasn't clear to me that all of the points deductions were actually applied as a final step, as it seemed more important to fix the bad citations (and to give editors a chance to either explain or to fix the bad citations themselves). But we should make it clear that if they haven't addressed any outstanding N marks by a specific date, points will be deducted from their final total.
We need to discourage people from slapping a footnote onto an entire paragraph, when it's only one sentence within the paragraph that is verified by the citation. To me it's potential grounds for N deductions as well.
I like the idea of giving extra points for good referencing but not sure how to implement. Maybe if you get a YY during the reviews process for a specific article, you get awarded the extra point. (At the same time, it seems subjective and open to abuse and collusion.)
August works well for me because there's an NPP drive scheduled for September, and I might not have time for both. Doing this in August would be ideal. – DreamRimmer (talk) 11:47, 8 July 2024 (UTC)[reply]
@DreamRimmer - that's really good to know. Maybe this makes October a more appealing option instead of August so we're not overlapping with other big drives? Is there a schedule that we can reference to find our niche? I know November is Asian Month which is a big one for me. Kazamzam (talk) 14:03, 8 July 2024 (UTC)[reply]
There is a GAN backlog drive scheduled for October. You can check out all the scheduled NPP drives here and find the GAN backlog drives here. August would be ideal because it allows us to plan another drive six months later, in February 2025. I apologise if you already have other plans for August. – DreamRimmer (talk) 14:36, 8 July 2024 (UTC)[reply]
August seems fine imo. We managed to get FEB24 up and running within a similar timeframe. With most stuff reusable from FEB24, we could be fine. ARandomName123 (talk)Ping me!14:47, 8 July 2024 (UTC)[reply]
@ARandomName123 - exactly, we "managed" and we "could" be fine. It was a pretty last minute scramble because we were committed to something that we had not prepared for; the edit history will attest to that. If we're still discussing basic aspects like 'how do we award points' and 'how do make sure people know if a page has been worked on', I do not think we're ready for this. We're a week into July and still discussing when the drive should start. I'd rather have a really good drive in October/November than a slapdash one in August for no reason other than that puts us on a semiannual schedule based on the February drive which was called at the 11th hour. Looking at the GAN backlog drives, they're not on a fixed schedule and I see no reason why we should nail ourselves to one if it's not actually conducive to a good product/outcome. August is also an extremely popular vacation month so we might get lower turn out by virtue of a lot of people not wanting to edit from a beach. Kazamzam (talk) 15:15, 8 July 2024 (UTC)[reply]
Ok, I see your point. October then? We'll have more than enough time to set everything up, and there shouldn't be much conflict with the GAN drive or other drives. ARandomName123 (talk)Ping me!00:18, 9 July 2024 (UTC)[reply]
The URA Backlog Drive 2 (Backlog vs. Jason) is now in planning stages. Comrade in arms @CactiStaccingCrane brilliantly made a backlog drives tab for the project so I think we can shift most of the discussion over there; it might be good to add a message on the project landing page as well for anyone who would like to be involved. There are several points for discussion that I think all previous planners are aware of and hopefully, since we have a lot of time, we can really iron them out and have an excellent drive (15k reduction is my personal goal) and then be on a good course for a semiannual schedule. Cheers, Kazamzam (talk) 14:37, 18 July 2024 (UTC)[reply]
For this drive, could we include a secondary goal of clearing all unreferenced BLPs? There’s only about 800 left, and with a push, we can probably bring it to zero. Regarding the drive pages, I’ll start creating them using February’s as guidance in a few hours, unless someone wants to get to them first? Looking forward to working with you all again! ARandomName123 (talk)Ping me!10:16, 20 July 2024 (UTC)[reply]
Getting through unreferenced BLPs seems like a great idea. I reckon once its down to zero it'll be possible to keep it there which would be good for the whole project. -- D'n'B-t -- 11:38, 20 July 2024 (UTC)[reply]
Do we want to create a November 2024 drive-specific talk page? The February 2024 one redirects to the main project but it would be useful to have a dedicated space for planning purposes. For example, I think the phrasing of 'To encourage additional reviews, the drive will be extended for one week (December 1-8) as an additional review period.' is wonky but I don't want to clog up the talk page here. However the drive talk page should be the place for drive participants to discuss issues they might be having, not just to how to see how the soyrizo is made. Maybe we start the discussion there and then archive everything on October 31st to start fresh (fresh-ish)? Kazamzam (talk) 11:46, 27 July 2024 (UTC)[reply]
I put in a RM to do so 2 weeks ago, but admittedly forgot about it until I noticed its recent close (as no consensus), and realized I should have posted here after initiating it, so am posting here now for visibility & historical purposes, and in case anyone else feels the need to re-RM. ~Tom.Reding (talk ⋅dgaf)06:58, 17 July 2024 (UTC)[reply]
Yes, {{WikiProject Unreferenced articles}} works for me too, and it's probably the most intuitive option, but I'm agnostic about whether or not "improved" is kept, since it's been used since 2010, so I'll defer to project participants on that. Redirects for either case can be made if desired. ~Tom.Reding (talk ⋅dgaf)17:09, 18 July 2024 (UTC)[reply]
@ARandomName123 @Tom.Reding I think alerting people to the project's existence by changing the template name but also the discussion would be very useful, especially as we gear up for the drive. We could also revise the edit summary that we use for main space for a talk page summary as well. Kazamzam (talk) 11:40, 27 July 2024 (UTC)[reply]
Shorter update than usual but happy 17th anniversary to the launch of the Phoenix spacecraft! I hope everyone is doing swimmingly.
Headline: We cleared 5148 articles since June! For yourself and your fellow editors, please clap.
Highlights: September 2008 is in the bin! There was also tremendous work done by editors @MIDI on geography stubs, reducing September 2013 from 1,489 to 627 (57%!), and to @Gnisacc for whittling December 2023 down to 73 articles, a 71% decrease.
Low-hanging fruit: October 2008 is only a measly 8 articles left, dangling precariously like a sinner in the hands of an angry God. Give it a nudge.
High-hanging fruit: Everyone's favourite BFC (Big Friendly Category), December 2009, is a shapely 11,593 articles as of this writing - so much freshness and only 2 calories. The other high-hanging fruit are, still, the Frustrating Five (name open for revision): January 2013 (1,118), April 2019 (1,037), May 2019 (2,103), June 2019 (4,433), and September 2020 (1,393). Once again, September 2020 had the lowest percentage of change between updates. Godspeed to anyone working on these.
Announcements: The second URA drive is in the planning stages! We will begin 1 November 2024 and end 30 November 2024, with a proposed extra week for review where participants can earn extra points. The planning is active and ongoing so please feel free to get involved.
New challenge: no ties this update! Carry on.
Next update will be 4 October 2024 and then, after a pause for the drive, 4 January 2025. My personal goal is to have 2008 entirely finished by the new year and it seems more doable than ever. Happy editing! Kazamzam (talk) 14:10, 4 August 2024 (UTC)[reply]
Thenks for the update, @Kazamzam! (And for dealing the final blows to October 2008.) Curious whether we collectively succeeded in decreasing every category by at least 10, as you suggested back in June? Cielquiparle (talk) 04:23, 14 August 2024 (UTC)[reply]
@Cielquiparle - thanks so much for the follow-up! There were eleven categories that didn't make the cut. Both February and June 2009 only decreased by 7 articles, which is...not great but that means there's just more room for improvement. Kazamzam (talk) 22:56, 14 August 2024 (UTC)[reply]
Can we send reminders to the Talk pages for everyone who already registered, letting them know the backlog drive is now open? I know @CactiStaccingCrane did it last time and I found it very helpful. (I would offer to do it but don't know how.) Cielquiparle (talk) 02:07, 1 November 2024 (UTC)[reply]
@CanonNi The {{Unreferenced}} tag should remain...unless you or another editor adds at least one citation, in which case the tag could be upgraded to {{more citations needed}} or equivalent. Cielquiparle (talk) 13:31, 1 November 2024 (UTC)[reply]
I have added sources to some unreferenced articles as part of this drive. Will these contributions be counted from now onward, or will only the changes made starting in November be reflected? TheBirdsShedTears (talk) 10:16, 25 October 2024 (UTC)[reply]
@TheBirdsShedTears - thank you for your question and your work. As it says at the top of the drive page, the drive has not started yet. Nothing will begin to count for points until November 1, but you are getting excellent practice! If you have any other questions, please reach out here or on the drive talk page. Thanks, Kazamzam (talk) 11:06, 25 October 2024 (UTC)[reply]
@Cielquiparle - you’re absolutely right. I think the biggest issues from February were how to judge an adequate reference (and how to make sure we have enough reviewers), if one reference is sufficient (I don’t love it for a big article but I think it has to be sufficient given the premise of the project and the campaign), how to make sure edits are counted (we used the edits summary function but personally I like the idea of people submitting through a checker in the same way that Asian Month does) and if there’s a way to get multiple points for one article (I’m on the fence). If anyone else has big ticket items to address, please bring them up! Kazamzam (talk) 10:52, 11 September 2024 (UTC)[reply]
@Kazamzam: I'm assuming the checker you're referring to is https://fountain.toolforge.org/? I haven't used it before, but if by checker you mean there's a function for us to review edits in the tool, that'd be really nice. I also noticed there's https://hashtags.wmcloud.org/, which imo is better than the tool we used last time, which searches for an edit summary (with a hashtag) over all users (ex), instead of one-by-one as we did previously. I'd personally prefer edit summaries for the convenience though.
The extra review week last time went pretty well imo, though I'm not sure if we actually applied the point deductions for failed reviews.
In total, we had 14.3k articles reffed, with just under 2k reviews (14% or so). This could probably be a bit higher, as it was kind of hectic last time.
@Cielquiparle @ARandomName123 - I was just looking through this before I saw your ping; great minds think alike! Looking at the checker (Fountain), it seems like it relies a lot on individual checking and the Asian Month competition I've referenced (rather egregiously, now that I look at the metrics) had fewer than 200 articles submitted. Given that we had over 14,000 articles submitted for the last backlog drive, I have doubts about the volume because I haven't seen any submissions on that scale (the other recent projects that I clicked through were also rather small). I'm going through the hashtag search now to see how that works - that might be our best bet but we need to have a 100% unique and distinct hashtag that no one gets wrong. Seems doable as long as we're clear from the jump and really hammer it home. Kazamzam (talk) 21:14, 18 September 2024 (UTC)[reply]
Agree on hashtags, and we could just do the same as last time, something like #NOV24. There's currently no edits with that hashtag, though I did do a small test with #NOV24 test. Volume shouldn't be an issue, as the #WPWPARK has a fairly similar revision number (16k) and is still fairly fast. ARandomName123 (talk)Ping me!21:25, 18 September 2024 (UTC)[reply]
@ARandomName123 - we love to see it! Something more distinctive than a hashtag might be helpful + an edit summary that uses the template like WP:FEB24 would be good for directing potential participants to the drive. I think the hashtag is the best route for counting but we should encourage people to use a template either way. Kazamzam (talk) 19:16, 19 September 2024 (UTC)[reply]
@Kazamzam: I was thinking we could just combine the two (#[[WP:NOV24]] or #WP:NOV24), but after testing it a bit, it unfortunately doesn't seem to work. "[[WP:NOV24]] #NOV24" does work, but it's a bit long imo. ARandomName123 (talk)Ping me!19:31, 19 September 2024 (UTC)[reply]
@ARandomName123 @Kazamzam Are we really doing this backlog drive in November? If hashtag search isn't giving us what we want, should we just revert to the way we did it back in WP:FEB24? Or...should we postpone this backlog drive after all given how busy everyone is and aim for February 2025...? Cielquiparle (talk) 02:16, 2 October 2024 (UTC)[reply]
@ARandomName123 @Cielquiparle I'm still game for November. We can go with the #NOV24 since Random (if I may) tested it and found that the tool can handle the volume. Other than that...can we hammer out the scoring and otherwise start building the scaffolding of the drive? Kazamzam (talk) 16:56, 2 October 2024 (UTC)[reply]
@Cielquiparle, Kazamzam (sure!): November's still good for me as well. How do we feel about completing the drive planning by October 23/24, about a week before the drive starts? As for scoring, do we want to give an extra point if an extra reference is added? Scoring for reviews is probably fine as is, and we should probably keep the extra review week. The criteria for reviewing should just be that the reference is reliable and supports the content. In FEB24 I think there were a few rejections for being non-independent, which I don't think was necessary. I'm open to suggestions though. ARandomName123 (talk)Ping me!17:49, 2 October 2024 (UTC)[reply]
Drive planning to continue through October 23/24...but could we open signups before that, like October 15? Or even sooner? To drive awareness and participation. Cielquiparle (talk) 20:56, 2 October 2024 (UTC)[reply]
@ARandomName123 @Cielquiparle - I like those dates. I think something that might keep people from going bonkers with adding a dozen references to score extra points is if the initial reference is one point and then every subsequent reference is half a point, or something to that effect? What do you think?
I will say that if it’s just the three of us as organizers since it seems like @CactiStaccingCrane is not active and I haven’t seen @Boleyn in a minute (hello if you see this!), I will limit my activities to reviewing and troubleshooting the drive rather than clearing articles myself. Unfortunate part of party hosting. Kazamzam (talk) 02:36, 3 October 2024 (UTC)[reply]
@Kazamzam Your point system proposal sounds OK...but how are we tallying points then? (It was so much easier when it was 1 point per article.) Or can the new tool automatically track points somehow...? Cielquiparle (talk) 19:24, 3 October 2024 (UTC)[reply]
@Cielquiparle, Kazamzam: As with FEB24, DreamRimmer has kindly agreed to provide the tech and bot needed to track the points. According to them, tallying half points for subsequent references would require a separate hashtag. They believe that we should stick to the method in the previous drive, or 1 point per NOV24 edit summary. ARandomName123 (talk)Ping me!03:04, 4 October 2024 (UTC)[reply]
@ARandomName123 @DreamRimmer How about a separate hashtag called NOV24PLUS or NOV24BONUS...and a rule that it's only 1 hashtag per edit summary line? So for any additional half-point, you would have to add them in sequence in multiple edits, which I think makes it easier to check later too. Cielquiparle (talk) 04:05, 4 October 2024 (UTC)[reply]
@Cielquiparle: Well this would necessitate adding each reference in a separate edit. Hopefully for future drives, we could analyze the diff itself to check for the number of refs added, but I don’t believe we have enough time to develop that for this time. ARandomName123 (talk)Ping me!19:54, 4 October 2024 (UTC)[reply]
@ARandomName123 Yes I'm over it now. The beauty of the February 2024 drive really was its simplicity. People will naturally be motivated to add "extra" references here and there as they go; it's not all about racking up points. Cielquiparle (talk) 01:04, 5 October 2024 (UTC)[reply]
I agree with @DreamRimmer that sticking to the previous method would be for the best. Individually counting each reference and/or making everyone do separate edits is probably too onerous given the reviewers are few and we're likely to get 10,000 articles cleared. Maybe in the future we can do a smaller drive and work on articles with the single source or refimprove tags. Kazamzam (talk) 14:49, 4 October 2024 (UTC)[reply]
@Kazamzam, Just let me know when everything is ready, and I will handle the bot part. If you want to do some testing, please use a hashtag, and I will update the leaderboard with the bot to check if anything needs fixing. – DreamRimmer (talk) 15:01, 4 October 2024 (UTC)[reply]
@DreamRimmer: I’ve been using #FEB24 for a bit of testing if you’d like to try it out. On a side note, could you let us know when the bot is ready? Hopefully the changes aren’t too complicated. ARandomName123 (talk)Ping me!19:57, 4 October 2024 (UTC)[reply]
@Cielquiparle, Kazamzam, DreamRimmer and any others: During FEB24, there was some dispute over what qualified as an acceptable for reference for reviews. Any suggestions? My pov is just that the source has to be reliable and verifies the immediately preceding sentence. Also, does anyone have any suggestions for the drive page, WP:NOV24 (including rules section)? ARandomName123 (talk)Ping me!04:48, 9 October 2024 (UTC)[reply]
Agree with that definition of acceptable reference. Might include as a sub-bullet to that, the example that if you add a citation to the end of a paragraph but the reference only applies to that one sentence, you need to go back and add {{citation needed}} to the sentence prior. Cielquiparle (talk) 04:55, 9 October 2024 (UTC)[reply]
Final touches
Buongiorno! I think we are ready to start promoting the drive on editors' watchpages, remove the TBD from the rules section, update the progress chart date (it's still February 2024), and...anything else we might need to do? Am I missing anything? I suppose monitoring the talk page for the drive closely would be a good idea - perhaps a message on the drive mainpage saying 'if you have questions or concerns, please address them to the URA organizers on talk page' or something to that effect?
Are we still giving out barnstars? We had some very nice ones made if anyone wants to get back on that horse.
@Kazamzam: Do you have any other comments/suggestions for the rules? I didn't want to remove the TBD until it's mostly finalized and agreed upon. I've added a message as you suggested, but feel free to edit it if you'd like.
Yes, I'm assuming we're going to be giving out barnstars. The ones from last drive should work fine. Once the drive is about to start (Oct.30/31) we should probably send out a mass message like last time to make sure the participants are aware that we changed the edit summary. ARandomName123 (talk)Ping me!15:15, 17 October 2024 (UTC)[reply]
1) I think the rules are good to go. I will clarify that the talk page referred to is the drive-specific talk page rather than this one - we saw a ton of activity here and it got a bit bloated imho.
2) Reading over the drive page more thoroughly, these are the issues I've identified that are still outstanding. We can cross them off as they are addressed:
Removing the [insert] from the lead section and updating that with the number of articles - can we automate this? If not, we should maybe make this sentence invisible until October 31st 23:59.
Making a drive talk page
Related: putting the info about where to ask questions as its own separate section so it doesn't get lost in the shuffle of the somewhat overly long lead section.
Updating the progress chart date to November 1st
Updating the milestones. Proposal: first 1,000 articles cleared; the 2008 categories finished; 2,000 articles; 5,000 articles; 10,000 articles; 15,000 articles (MAYBE even 20,000?). I would love to see 2009 in the dust but with BFC December 2009, that seems unlikely.
Suggestion: unreferenced BLPs are a serious issue. We don't have many of these but they are a big concern. Perhaps a section that focuses on these? This could be a 1.5 point category, with a caption like NOV2024BLP(?)
Sorry I'm a little late to the party. I've made a few edits to the drive page intro (mainly adding context and trimming) and sectioned off 'Where to get help'. Looking forward to November. Thanks for all your hard work. Turtlecrown (talk) 20:09, 22 October 2024 (UTC)[reply]
Also, if a message is sent, it would be best to add another reminder of the correct edit summary to include (#NOV24) in case people assume the old summary format is still being used. ARandomName123 (talk)Ping me!19:54, 31 October 2024 (UTC)[reply]
Sounds good, I’ve received the mass message. I removed the under construction banner, and apart from some minor changes mentioned above, there shouldn’t be too much else. ARandomName123 (talk)Ping me!01:10, 20 October 2024 (UTC)[reply]
@DreamRimmer - if there’s still time to edit, I would say something like ‘sigh up for [whole spiel], help add sources, improve Wikipedia, and earn barnstars!’. Maybe I’m overly pedantic but I think it’s important to say what will be happening and give a hint of what to expect. Kazamzam (talk) 17:33, 20 October 2024 (UTC)[reply]
@Kazamzam: We're now only a couple of hours out till the start of the drive, and we only have around half the number of participants as we did last time. Should we lower the goal back down to 10k? ARandomName123 (talk)Ping me!18:19, 31 October 2024 (UTC)[reply]
@ARandomName123 - looking at the signups from the February 2024 drive, many people signed up on February 1st and continued to sign up well into February (the last one was on February 28th). We also have an extra day and fewer articles to cite: in January 2024, we had 114,214 and now we are just north of 80,000. I think 10,000 is confident but realistically feasible, 15,000 is striving, and 20,000 can be our moon-goal. What do you think of that? Kazamzam (talk) 18:35, 31 October 2024 (UTC)[reply]
@Kazamzam: I think it's a bit late too make new additions to how the points work, but I've added a button specifically for BLPs to the top of the drive, next to the "Fix a random page lacking sources" button. ARandomName123 (talk)Ping me!20:27, 22 October 2024 (UTC)[reply]
Unreferenced articles by number of views?
Hello. I'm new to the project, and wondered whether it was possible to get a list of unreferenced articles sorted in descending order of page views. Maybe one already exists and I just haven't found it yet!
The idea being that then one could choose to work on articles that are looked at a lot, and ensure that they are decently referenced. Thanks, and pleased to be part of the project. SunloungerFrog (talk) 09:02, 4 October 2024 (UTC)[reply]
Quick update as we get ready for the November 2024 Backlog Drive, and happy 67th anniversary to Sputnik 1, the first artificial satellite to orbit the Earth![1]
Headline: Smaller decrease in articles than usual. Nevertheless, we cleared 2821 articles! For yourself and your fellow editors, please clap.
Minutiae: For anyone interested in a more detailed breakdown of the numbers - average was 21.0 articles or a 5.40% decrease; median 15; mode 14.
Highlights: October 2008 is in the dustbin of history! November 2023 dropped by exactly 100 articles, from 407 to 307, and everyone's littlest friend, December 2023, further decreased from 73 to 48, a 34.25% decline.
Low-hanging fruit: Other than December '23, November 2008 is still loitering about with 45 articles left, dangling precariously like a sinner in the hands of an angry God. Give it a nudge.
High-hanging fruit: Everyone's favourite BFC (Big Friendly Category), December 2009, is a svelte 11,115 articles as of this writing - officially below the 90% completion point! The other high-hanging fruit are, still, the Frustrating Five (name open for revision): January 2013 (1,103), April 2019 (998; below 1,000 for the first time!), May 2019 (2,038), June 2019 (4,311), and September 2020 (1,365). This time, January 2012 had the lowest percentage of change between updates. Godspeed to anyone working on these.
Announcements: URA Drive #2 is in the works and we want YOU to volunteer as an co-organizer and/or reviewer of submissions! Comrade-in-arms @DreamRimmer is doing the noble work of running the bot to keep track of submissions using the hashtag #NOV24. We will begin 1 November 2024 and end 30 November 2024, with a proposed extra week for review where participants can earn extra points. The planning is active and ongoing so please feel free to get involved.
New challenge: two ties this update! January 2024 and February 2024 are 301 apiece and December 2015 and January 2016 are 323. This could be something to highlight for the drive if anyone finds this at all compelling (or wants a picture of my cat, Jellybean, in her special hat). As a personal challenge, I would like to see 2008 done by the end of the year, so we can have a backlog that's only, uh....16 years.
There will be a post-mortem for the drive - given the success of the last one, I'm personally hoping to hit 20,000 articles - and then regular updates will resume in January 2025. Thanks, as always, for the amazing work! Kazamzam (talk) 16:52, 4 October 2024 (UTC) Kazamzam (talk) 16:52, 4 October 2024 (UTC)[reply]
References
^Burgess, Colin; Hall, Rex (2009). The First Soviet Cosmonaut Team: Their Lives, Legacy, and Historical Impact. Berlin: Springer. p. 1. ISBN9780387848235. LCCN2008935694.
Thanks, Kazamzam! That's us at 81,934 - perhaps we will get under 80,000 during the November backlog drive. It's heartening to see the numbers still going down. Boleyn (talk) 08:16, 11 October 2024 (UTC)[reply]
This is likely somewhat controversial, but I would be inclined to say: If it was already tagged as "Unreferenced" despite having the list of general references, and you actually add an inline citation, it should count. Cielquiparle (talk) 08:11, 1 November 2024 (UTC)[reply]
I would disagree with that; an article having {{no footnotes}} is a separate issue from one being {{unreferenced}}, even if the former is erroneously tagged as the latter. When I came across such articles in the FEB24 drive I just corrected the tags.
The point of the drive is to cite information that is uncited, and if there is a non-inline reference, there's a good chance that it is backed up by a citation (though ofc ideally it should be made inline so it's possible to verify efficiently). novovtalkedits09:38, 1 November 2024 (UTC)[reply]
I agree with the first sentence as a general rule.
I just think in the case of a Backlog Drive, we should reward good-faith addition of inline citations, rather than splitting hairs over whether an external URL or book title which already appeared on the page qualified as viable references or not, etc., etc. It gets slippery and is a buzzkill.
Most important thing is that in-line citations are added, referencing reliable sources...as we bring down the overall backlog by at least 10,000. Let's do it. Cielquiparle (talk) 13:39, 1 November 2024 (UTC)[reply]
This is what we did with the last drive. If an article had general references and the unreferenced tag, but no reflist or in-line citations, and the user added at least one in-line cite, it would count towards the drive. The sentiment was it's better this happens than just moving it to the {{no footnotes}} queue. As one of the users who did a lot of this type of thing last drive, it didn't boost my numbers by any means since there was still a lot of editing being done to add the in-line cites, as appropriate. --Engineerchange (talk) 21:20, 1 November 2024 (UTC)[reply]
Hello everyone, I have created a user script called User:DreamRimmer/DriveEditSummary.js. This script adds a +NOV24 button right above the Save, Preview, and Changes buttons in the source editor. Clicking this button automatically inserts a preloaded edit summary with the #NOV24 hashtag. Please note that it only works in the source editor, not in the visual editor. I hope you find it useful! – DreamRimmer (talk) 13:41, 1 November 2024 (UTC)[reply]
Great, thanks for this tool. Just did an extra page to try it. (Success!) Just a thought: "Adding reference(s) ... using" does make it sound like the script was used to add the references rather than the edit summary. Turtlecrown (talk) 03:45, 2 November 2024 (UTC)[reply]
I'm not sure if these ones should also be AfD'd - sourcing seems to be pretty sparse on the ground, but I'm wondering if it's a transliteration/spelling issue? They're notable figures and I can find oblique references to them here and there, and there are more recent figures in this royal family. So I'm not sure what the best way to proceed is here. Thoughts? Smallangryplanet (talk) 11:47, 4 November 2024 (UTC)[reply]
We love to see it! @Smallangryplanet, just a note, can you be sure to post question like this on the talk page for the drive itself? That will hopefully help more participants see any issues as they pop up. Kazamzam (talk) 20:28, 4 November 2024 (UTC)[reply]
@Kazamzam Oops, I thought this was the talk page for the drive itself! I think I got turned around because the "Discussion" tab on the drive links to this page, not the drive's talk. Is that correct or should I change it? Smallangryplanet (talk) 20:31, 4 November 2024 (UTC)[reply]
Why is the drive only rewarding editors for adding inline citations? General references are a legitimate citation style according to policy, preferred by many for short articles (and preferred for everything on some sister projects, e.g. dewiki), and have the same result of removing the article from the "unreferenced" backlog. – Joe (talk) 09:38, 21 November 2024 (UTC)[reply]
Well, with in-text or parenthetical citations. But more to the point, it is not a policy requirement that we attribute specific statements to their source except in limited circumstances. This is "WikiProject Unreferenced articles" not "WikiProject Articles with no footnotes". – Joe (talk) 10:23, 21 November 2024 (UTC)[reply]
@Joe Roe - that's a great question and a very interesting point that I have wondered about myself. I have definitely added footnotes in addition to (and sometimes instead of) inline citations, but I do then add the 'no inline' clean up tag. The following are my personal thoughts and not necessary those of the project. Happy to discuss further and take this as feedback for the URA and future drives going forward!
Short answer, per WP:INTEGRITY and WP:GENREF: The point of an inline citation is to allow readers and other editors to see which part of the material is supported by the citation. The disadvantage of general references is that text–source integrity is lost, although this is sometimes mitigated if the article is very short and the general references is very direct. General references are frequently reworked by later editors into inline citations; it saves time to just do it now.
Longer answer: Looking over the WP:MINREF guidelines, I think the URA is holding up articles to a higher standard than might be strictly necessary by focusing on inline citations. And mulling it over in the pre-caffeine morning, I am largely fine with this. The guidelines note that "it is typical for editors to voluntarily exceed these minimum standards", and I'm pleased that we are going above and beyond with this drive. Editors both in and out of the project can help add citations to articles as they see fit but we are choosing to exceed that minimum and reward to that effort for this drive; there's no prescription against general references, in-text, or parenthetical citations. But more importantly (in my opinion) is the following statement: "any material lacking an inline citation to a reliable source that directly supports the material may be removed and should not be restored without an inline citation to a reliable source". Yes, general references and non-inline citations might be sufficient in some cases and maybe would solve a lot of headaches around tiny stubs of German streams and rivers, but I think we are a) holding ourselves as a project to a higher standard that is ultimately beneficial to the encyclopedia and b) mitigating the need for future work by providing inline citations to statements that will hold up when editors are inclined to delete statements or articles that are unreferenced or only ("only") have footnotes.
Beyond the perspective of the drive, because the URA will still be here with tens of thousands of articles when this is over, adding general references is quite common but from a clean-up perspective, it punts the article from "unreferenced" to "lacking inline citations" and this, given the existence and widespread use of that specific tag, seems to be considered insufficient by the community. I think many if not most of us are here to improve Wikipedia, not shuffle articles from our preferred category to someone else's. If we're going to improve the encyclopedia by cleaning up these articles, we should actually do that rather than shuffling things around and hoping no one notices the level of untidiness is actually unchanged.
There's more to say on this and doubtless other people will have different opinions but those are my two cents (or with inflation, 12 cents). Again, this is a great question and thank you for raising it. I hope we have some really excellent discussion as a result. Cheers, Kazamzam (talk) 13:10, 21 November 2024 (UTC)[reply]
I don't think "shuffling things around" is an accurate description: you're moving something from one of our most pressing cleanup categories ({{unreferenced}}) to a less severe one ({{no footnotes}}). That is an improvement to the encyclopaedia. It's also not a given that the article with general references will or should be tagged with {{no footnotes}}: a stub with no quotes or potentially controversial claims, for example, will not benefit from inline citations unless and until it is expanded.
URA participants are of course welcome to pursue a higher standard than that required by policy or what a literal reading of the name of the project would imply. It just seems inefficient to me. One of the things I really like about this project is that it is tightly focused on one problem, rather than trying to 'tidy' everything about an article at once. – Joe (talk) 13:26, 21 November 2024 (UTC)[reply]
November 2024 Backlog Drive Planning
Popular article instruction/issue
T-8 days! Much obliged to @Turtlecrown for cleaning up the rules. One thing I noticed from this is that the 'Popular articles' section is a little wonky. It also needs to be updated daily - is there someone willing to commit to this? I'm hesitant to do so only because I couldn't get the WMCS link to work at first pass, which is somewhat concerning. Kazamzam (talk) 16:51, 23 October 2024 (UTC)[reply]
Assuming it is fixed before the drive, I can help with that section. It's fairly simple, just copy-paste the 50 (I think?) articles in. The wikitext would need to be added, but that's easily done with some code or just ChatGPT. ARandomName123 (talk)Ping me!04:07, 24 October 2024 (UTC)[reply]
@ARandomName123 The list of "Popular articles" is very popular among our Backlog Drive participants. Let me know if you need help updating it daily – if so, please let me know how best to go about grabbing the list. It seems we tend to get stuck at the top part of the alphabet, which is unfortunate. It would be great if we could mix it up a bit – either just wipe the list and start over completely fresh once a day and/or grab a more random selection from throughout the alphabet on a daily basis. Cielquiparle (talk) 02:33, 2 November 2024 (UTC)[reply]
@Cielquiparle: Yea, sorry about the delay, just updated it a few minutes ago. I'm not sure how we'd mix it up in terms of the alphabet, but you could change the timeframe the view count is counted. The default looks like the past 3 weeks.
If you see it hasn't been updated by around 00:10 UTC, feel free to do it. My usual workflow is just to let it run, then copy and paste the first 50 entires, including the misc. stuff so there's no need to copy-paste one by one, into ChatGPT (I'm too lazy to write a script lol) with a prompt along the lines of Please format the follow list in a code block in the format *[[article title]]. ChatGPT will then add the required wikitext and remove the unnecessary pageview numbers/rankings. Then just click the Copy Code button that pops up, and paste it into the section. Should be fairly quick, but if you find a method that works better for you, feel free to use it. ARandomName123 (talk)Ping me!02:41, 2 November 2024 (UTC)[reply]
@ARandomName123 Thanks for the instructions. Re: Alphabet bias, I think what's happening is that we're trying to run the Massviews analysis on too large of a category (exceeding 20k items), so it's always taking only the first 20k items in alphabetical order. Without writing a script, the best bet may to narrow by sub-categories...like everyone's favorite "Articles without references from December 2009", for example. So I just went ahead and added a handful to the top of the list (also because I had to delete several from the "new" list you refreshed which put back a bunch of articles that are no longer Unreferenced because of the WMCS lag and because #TeamBacklogDrive is so incredibly fast). Cielquiparle (talk) 03:31, 2 November 2024 (UTC)[reply]
@Cielquiparle @ARandomName123 In case it is a bit useful, to work around the 20K alphabetical limit in Massviews, you can do a Petscan against the category, sorting the output by number of incoming links in descending order (as incoming links is a sort of proxy for popularity) and then taking the top 1000 into a Pagepile, and running Massviews against that. For example:
@Cielquiparle no problem, very glad to help. The combination of Petscan and Massviews is really useful. In a similar vein, I added a section Popular biographies below popular articles that uses this Petscan 29822085 Pagepile and Massviews in a similar way. If you think the section is overkill, please do remove it - I won't be at all offended. SunloungerFrog (talk) 11:37, 13 November 2024 (UTC) P.S. @ARandomName123 your ChatGPT hack is awesome. Best use of AI yet![reply]
Watchlist notice and invitations
Would like to post a watchlist notice (text ad at the top of every editors' Watchlist), with the words the "November 2024 Unreferenced Articles Backlog Drive has begun. How is it done? Cielquiparle (talk) 08:27, 2 November 2024 (UTC)[reply]
Cielquiparle: If you'd like to request a watchlist notice, you can do so at MediaWiki talk:Watchlist-messages. Just a quick note: we can only display a watchlist notice once per event, and it stays up for seven days. Since we already displayed one for this drive from 20 October to 28 October, we won't be able to request it again. – DreamRimmer (talk) 09:36, 2 November 2024 (UTC)[reply]
For info: I put a short invitation note on the Talk pages of four of the large WikiProjects from here (Germany, Military history, Football, and Albums) with a link to bambots' /bycat/[WikiProject]#Cites no sources. Turtlecrown (talk) 13:05, 2 November 2024 (UTC)[reply]
Review page instructions
Again, very late here. But I'd like to change the review page instructions to the following, which:
presents the process as ordered steps,
tells reviewers what to check,
distinguishes between 'submission' and 'reference',
gives more prominence to the Tool (which many missed at the start),
and (optionally) covers a couple of FAQs (well, once-asked, but still good to know).
Part of the reason it's so late is that I was slow to understand the process at the start of the drive.
Welcome to the reviews page. Your role on the drive is extremely important and we are glad that you dedicated your time to do this grunt work. Each reviewer should create their own dedicated section for reviewing.
It is mandatory that you don't review your own submissions (obviously) and it is recommended that you spread out your reviews to many participants. Here is the procedure for reviewing a submission.
Find a submission from the drive. A tool is available to locate unreviewed submissions (credit: ARandomName123). Submissions can also be found via the "Tally" link for a participant in the Leaderboard.
Review the submission. Check that the article was unreferenced and the submission contains an inline citation from a reliable source, and verifies the immediately preceding sentence.
Use of primary sources is permitted, so long as it follows WP:PRIMARY.
If the citation only verifies part of a paragraph, a {{citation needed}} should be added to the prior part.
Duplicate #NOV24 tags will be caught by the bot and don't need to be flagged in review.[3]
In a submission with multiple references, a Y may be given if one citation passes these criteria, as long as any other errors are addressed.[4]
(recommended) Cleanup and refinement. If you need to make any adjustment to the article, you can prefix your edit summary with #NOV24REVIEW. Warn editors about their mistakes and sloppy work; if needed, you can raise concerns at Wikipedia talk:WikiProject Unreferenced articles. Consider re-reviewing user corrections to submissions made within the drive period.[5]
Document your findings in this page. Use the Wikitext format:
That's a wrap on the November 2024 drive! Thanks to all participants for their help; details on progress and benchmarks reached can be found on the drive page. Help is still needed on reviewing, so please sign up if you are interested! This update covers the changes since October so the totals are a little different.
Headline: We cleared 10,489 articles and are now officially below 72,000! For yourself and your fellow editors, please clap.
Minutiae: For anyone interested in a more detailed breakdown of the numbers - average was 114.8 articles; median 40; mode 39.
Highlights: the entire year 2008 is in the dustbin of history, along with January and February 2009! With the heroic work of editor in arms, @Turtlecrown, the stubborn, German river-centric category of September 2020 saw a 25.3% drop. And everyone's littlest friend, December 2023, further decreased from 48 to 37, a 22.9% decline.
Low-hanging fruit: Other than December '23, June 2024 is a skinny 115 articles, dangling precariously like a sinner in the hands of an angry God. Give it a nudge.
High-hanging fruit: Everyone's favourite BFC (Big Friendly Category), December 2009, is a lean, mean 9,568 articles as of this writing - bringing all categories to under 5 digits. The other high-hanging fruit are, still, the Frustrating Five (name open for revision): January 2013 (1,039), April 2019 (875), May 2019 (1,822), June 2019 (3,915), and September 2020 (1,020). This time, January 2013 had the lowest percentage of change between updates. Godspeed to anyone working on these.
Challenge results: February 2024 beat January 2024 264 to 270, and December 2015 sneaked a win from January 2016, 286 to 289. Live sports are just riveting, aren't they.
New challenge: No ties this time. Have a wonderful end of 2024 and I hope to see you all in 2025!
Announcements: Something discussed prior to the drive was building a regular, semiannual schedule for drives, which would put us on target for May 2025. I see this as a reasonable amount of time to prepare and make adjustments based on editor feedback about the strengths, weaknesses, and areas of opportunity for improvement. However, I see no reason to rush this as we are still tallying this drive's results. If this date is amenable to people, we can get started on a draft page in January. I would especially like to address the points raised by @Joe Roe about the prioritization of inline citations and how we can address this in the future.
I had a great time working on citations during this drive. Thanks to everyone working to organize these events. See you all in the field between now and the next drive! I will defeat December 2023!! Gnisacc (talk) 19:03, 5 December 2024 (UTC)[reply]
I take it back. Let's have as many backlog drives as we can in 2025 and clear the backlog. Why not schedule 3–4 next year? February, May, August, November. Cielquiparle (talk) 13:09, 7 December 2024 (UTC)[reply]
@Cielquiparle - I LOVE this attitude. Personally I'm open to it but I wonder if, by having (too) many drives, we run the risk of diminishing returns via lower editor engagement/participation. We cleared fewer articles in the November compared to the February drive but we also had a much larger backlog. February started with 111,643 and cleared 14,300 (12.8%); November started with 80,645 and cleared 8,511 (10.5%) - not much of a difference proportion-wise, plus we are probably clearing out the easier articles sooner and so what's left are the tougher nuts to crack. My concern is that if we did a drive every few months, would the proportion drop to more like 5-6%? How to do we make sure to encourage and sustain engagement throughout the drive? We definitely lost momentum by the halfway point of both drives, so I think making sure that we keep that up (somehow) for future drives is a concern to prioritize and I'm worried that might be difficult if we do 4+ drives a year. I think it's doable, but we have to be smart about it. Setting smaller goals that are likely to be met through the drive to keep momentum up is probably a good idea; at the same time, we can't have a party every time we go down 100 articles.
Maybe this is the point when we start organizing off-wiki. I also agree with your point that May might not be a good idea for exam reasons (excellent foresight!). Personally April is a wash for me but I could be game for June. Cheers, Kazamzam (talk) 16:52, 8 December 2024 (UTC)[reply]
Updates like these are the type of stats I love to see and make these things much more fun to participate in. Any chance that you have stats on the biggest decrease for categories by topic? Wozal (talk) 16:18, 8 December 2024 (UTC)[reply]
@Wozal - I wish :/ the way that the search function assigns categories by "topic" is pretty hit or miss. For example, if you try to filter the All articles lacking sources category for Biography, the first five results (by "relevance"): Fourth Dynasty of Egypt family tree, List of monarchs of Pontus, Single suiter, Quantitative notrump bids, and Entry-shifting squeeze. Not until you get past the first 20 hits do you get an actual biographical article of an actual person (Avgust Tsivolko, for anyone curious). So providing the stats on decrease by topic, let alone 200+ categories by topic, would be an incredibly time-consuming task. But this raises the larger issue of how the articletopic search function is basically useless; I think this is detrimental to potential WP members who might be really passionate about, say, Ancient Roman history or wetlands or Judaism or whatever but get frustrated and turned off by this filter that is supposed to be helpful. It's something I'm hoping we can address before the next drive, either to improve or replace it with something better.
I enjoyed the update very much, and always do. Thanks for all the hard work. And well done on the backlog drive, everyone!
My two cents on inline citations is that they make sense as a backlog drive task because of a need for what I'll call "meta-verifiability" - the ability to check each other's work and see if it does what it says it does. I've also seen a fair few articles in the wild that use only general references in which the article topic is not mentioned, and hunting down these resources, where possible, is time-consuming, and I don't want to assume intent but it can make content seem more reliable than it is. Maybe it pays in general to emphasise the difference between {{no footnotes}} and {{unreferenced}} though. Turtlecrown (talk) 23:10, 8 December 2024 (UTC)[reply]
Kazamzam thank you for the update. As a new editor, I really enjoyed working on the drive, and I learned a lot of things by reading through and adding references to a fairly broad selection of articles about different topics. I was always pleasantly surprised when it turned out that there was an interesting twist on an otherwise unprepossessing-looking article, and it was very satisfactory to be able to keep them in the encyclopedia. On the flip side, I was rather surprised to come across a few dreadful articles of almost no merit that had hung around for a long time; it was equally as satisfying to get those into the various deletion processes and bid farewell to them.
I was particularly pleased that we reduced the known backlog of unreferenced BLPs to zero, and I firmly think that, in time, (and maybe it is not as far off as we might fear), we can do it with the general unreferenced article base too.
On the issue of how the articletopic search function is basically useless, I note that our sibling project Unreferenced BLP Rescue have set up various Petscan searches to do this for Unreferenced BLPs. Maybe that is an approach that we could consider adopting, and I would be happy to do what I can to help set it up.
@Cielquiparle - I’m showing my age/status as the elephant who never forgets but it’s definitely not the first non-chronological clearance. Back in 2022, I got rid of July 2007 and then went back for the rest of 2007. I believe there’s a mention of it in Archive 4 but I can check the spreadsheet when I’m not on my phone. Either way, very exciting! I wonder if it’s appealing to people to pick a category from the middle of the pack rather than just old - maybe? I know that December 2023 was listed as a clean up project which is why it got so much attention. Now we can just add December 2009… Kazamzam (talk) 10:39, 11 January 2025 (UTC)[reply]
