RfC Proposal discussion

@JPxG, Dlthewave, FOARP, Blue Square Thing, Ovinus, Thryduulf, Levivich, DavidLeeLambert, Aquillion, S Marshall, Paradise Chronicle, and Avilich: Do any of you have thoughts on this proposal? I'm pinging you because you joined discussions at Wikipedia talk:Arbitration Committee/Requests for comment/AfD at scale about deleting or draftifying articles as a group. I know that some of you will oppose this proposal but I hope you will be willing to provide constructive input to the proposal so that if it does pass the result will be functional. BilledMammal (talk) 04:58, 24 February 2023 (UTC)[reply]

I would of course support the proposal. But I believe this wouldn't come through. So my thought is that there could be some sort of recognition for anyone who expands a stub to some other expanded level. Currently article creators are listed here and they receive the notifications of the article. If those "incentives" are also given to the main article expanders like the top 3 or 5 (or anyone who contributed over 5000, 10'0000 bytes...) the incentives for article creation will not be as exclusive as before.Paradise Chronicle (talk) 08:53, 24 February 2023 (UTC)[reply]
Support. At worst the proposal will not pass, though hopefully the closer will take into account the newest policy changes concerning source citation and mass creation. Avilich (talk) 14:59, 24 February 2023 (UTC)[reply]
Some thoughts:
  • Wikipedia:Miscellany for deletion/Mass-created portals based on a single navbox is the example that came to mind of a 1,000+ mass-deletion, in case it helps.
  • I agree with S Marshall below (1k at a time will take too long), but it's not bad to start with a relatively small batch to test the concept
  • I would try and cut the overall proposal length down by like half; probably don't need a lot of the "background" section (some of that can be saved for a below-the-fold comment); you might describe the selection criteria in the form of a bullet or numbered list, so editors can tell at-a-glance how this set was chosen
  • If the info is available, maybe say something about how these articles were created ... e.g, all in one day? one hour? etc.
Needless to say, I'd support. Thanks for putting this together! Levivich (talk) 20:32, 24 February 2023 (UTC)[reply]
I've made some tweaks to the details. I'll modify it further, and consider how to cut out most of the background section. The articles were created in many batches, a few hours at a time; this list is split across many batches, because when Lugnuts was creating Olympian articles he normally worked by sport, not by year. I'll think about how I can include that information.
Thank you for your input! BilledMammal (talk) 02:30, 25 February 2023 (UTC)[reply]
Yes, I think the tweaks make it better. IIRC, one of the things that persuaded editors to mass-delete portals is the realization that very little thought/effort went into the portals (they were mass-created very quickly), and also the realization that after creation, no other editors really edited them (the "unmaintained portal" argument). I think that's what convinced people they were a bunch of pages quickly made that nobody else uses. A few more thoughts:
  • Maybe break it into subgroups by creation (these XXX articles were created in YYY time)
  • Maybe look at how many other editors have made non-bot, non-minor, non-automated edits
  • Maybe compare size-at-creation with size-currently to show if they have/haven't been expanded
  • Maybe note how long ago the articles were created
  • Maybe note page views for these articles
Put it all together you might be able to show (what I suspect is the case) that these were quickly made and have largely sat unread and unedited ever since. You might also somewhere explain whether you're proposing these because (a) you believe they're not notable, or (b) you believe they should be draftified regardless of notability. Also, I'm not sure what the solution is, but if it's an RFC instead of a bundled AFD, I expect some editors will object based on lack of delsort; there's probably some way to delsort a draftification RFC (e.g., manually notifying wikiprojects that would otherwise be notified via delsort). Or maybe just a WP:CENT listing will be enough notification to satisfy everyone. Levivich (talk) 15:46, 25 February 2023 (UTC)[reply]
Good idea; I haven't worked out how I'll do the subgroups yet, and I need to get the page views from elsewhere, but the articles haven't been expanded, and they haven't been edited by other editors; the average and median article has grown by 325 bytes, the average article has had 0.138 significant edits by editors other than Lugnuts (defined as greater than 200 bytes and neither reverted nor an undo), and the median has had 0.
I've added a table containing the article by article information, although I'm not sure if that data should be with the list or separate. BilledMammal (talk) 01:50, 27 February 2023 (UTC)[reply]
Great job. I think if you make the table sortable, people will sort by date and see the groupings for themselves. For pageviews I've seen this and this. I couldn't figure it out in quarry, but I'm not really good with that stuff. Levivich (talk) 02:16, 27 February 2023 (UTC)[reply]
I've made it sortable, and added a column providing the number of articles created on the same day, as I think without that the context is lost. I'll have another look into those API's, thank you, and if that fails I can get the dumps. (Quarry doesn't contain data on pageviews, unfortunately) BilledMammal (talk) 10:41, 27 February 2023 (UTC)[reply]

I don't have much to add aside from general support for the idea. We should clearly explain why these shouldn't stay in article space (what harm are they doing?), why we should accept that some notable topics may be in the mix and why we can't do this through our normal deletion process. I think the added statistics in the table explain this well. One thing to note is that WP:NSPORTS requires at least one SIGCOV source and draftification is standard practice for newly-created articles on possibly notable topics that aren't fleshed out yet; why should we treat these incomplete articles differently by allowing them in mainspace? –dlthewave ☎ 16:22, 28 February 2023 (UTC)[reply]

That rule was not in effect when these were created. BeanieFan11 (talk) 16:38, 28 February 2023 (UTC)[reply]
And would these have been allowed if it was? Levivich (talk) 16:40, 28 February 2023 (UTC)[reply]
That's debatable. The policy says all mass-created articles (except those not required to meet WP:GNG)... – back then, passing NOLYMPICS was usually considered enough for notability (I can't recall a deleted olympian prior to the removal of the presumption of notability for non-medalists/the NSPORTS RFC). BeanieFan11 (talk) 16:46, 28 February 2023 (UTC)[reply]

First reaction

My first reaction is that dealing with over 1000 articles in one go is never going to fly, no matter your intentions or the quality of your arguments, many people will oppose as just far too much. I suggest that you need to split this into multiple discussions, perhaps starting with an AfD of a randomly selected 10 so that people can check them and verify that you've done your preparation correctly, there isn't something you've not thought of and that the principal is something they agree with.

If that results in a clear consensus to delete then you can do a second, slightly larger group (say 25), at AfD with (if necessary) a revised selection criteria, pinging participants of the first AfD. If the first AfD doesn't result in a consensus to delete, stop there as it's clear the community doesn't agree with you. If the first AfD consensus is delete but weakly, make your second the same size and listen to the feedback. If after a few (at least 3) AfDs there are strong consensuses to delete and no more feedback about your selection criteria then you could move an RFC for the rest.

Once you're over about say 25 don't put them in a single list, split them by event and Olympics as this will make it much easier for people to parse and to spot false positives, and this will reduce the overwhelm factor that will produce knee-jerk opposition. In all the discussions be open that deleting many more is your intent, but also make it clear why you are doing it in multiple steps and that you will listen to feedback. Do listen to feedback.

Another thought is that maybe you should exclude any (or deal separately with any) that have articles in other languages. Yes, this will take much longer and be much more work but there is no viable alternative to that.

I am going to be very busy with real life from now until about Monday evening so I might not be able to give any more comments until then. Thryduulf (talk) 08:22, 24 February 2023 (UTC)[reply]

"stop there as it's clear the community doesn't agree with you" -- not sure "the community" ever agreed to their creation in the first place. The double standard that it is possible to create but not delete an unmanageable number of non-notable trivial stubs no longer applies, since it is now established that mass created articles need at least one GNG source directly cited in them. Avilich (talk) 14:48, 24 February 2023 (UTC)[reply]
@Thryduulf AfDs on Olympic stubs have been done before by myself and I was encouraged to just redirect them. I'd argue that if the community leaves such articles in the current state, other editors might believe this it is for what wikipedia is, and will also create similar basic stubs.
I'd support to draftify them for that the editors who are interested in saving them, can expand and submit them.Paradise Chronicle (talk) 18:56, 24 February 2023 (UTC)[reply]
The purpose of this is to determine an alternative to AfD, because AfD cannot handle mass created articles without creating large amounts of drama. BilledMammal (talk) 02:31, 25 February 2023 (UTC)[reply]

First reaction (S Marshall)

Lugnuts started more than 94,000 articles. If we deal with them at the rate of 1,000 a month, the cleanup project will take just under eight years. I feel that this is completely excessive, and in order to make the task proportionate, we have to address more like 10,000 in a one-month RfC. I must say, I'm utterly confounded and bewildered by Thryduulf's take above.—S Marshall T/C 15:00, 24 February 2023 (UTC)[reply]

You're right about 1,000 a month being too small, but I think starting with a smaller test case and then increasing the scale if there is consensus for this one is the right path to go down. BilledMammal (talk) 02:32, 25 February 2023 (UTC)[reply]

First reaction (FOARP)

My first reaction is I am deeply tired of this topic. We've been working on this for years now without resolution. Here's some simple and basic principles which are well established at WP:

  • WP:PROVEIT - it's the job of people who want to include content to show that it should be included. There is no requirement in the opposite direction - no existing basis for inclusion means it can be deleted. This goes for articles as well.
  • WP:FAIT - editors should not present the community with a fait accompli and insist that it cannot be addressed.

Both of these point to a very basic solution to the Lugnuts problem - delete the currently-failing articles. Lugnuts even told us that he had purposefully included errors in them. Frankly I don't care that some of them can potentially be saved, when as it stands they all fail. People can re-add them as real, actual articles that weren't created simply as a way of chasing stats - deleting them does not prevent this. FOARP (talk) 15:44, 24 February 2023 (UTC)[reply]

My thoughts (BeanieFan11)

I of course would oppose this. I have a feeling that a large amount of these are notable - as I've found a large percentage of Olympians nominated at AFD (at least for ones I have access to sources for) pass GNG. For example, I randomly selected one of the 1000-something on your list, Albert Bechestobill, and quickly found full-page coverage in major newspapers. I feel this is just a route to backdoor get rid of massive amounts of sports articles because you don't like them. BeanieFan11 (talk) 22:47, 24 February 2023 (UTC)[reply]

Both are atrocious solutions, although the one you're suggesting more so. BeanieFan11 (talk) 00:03, 25 February 2023 (UTC)[reply]
You could prod a bunch of them, you could redirect some, you could afd a bunch (for these, I could perhaps see more than 10 considering the circumstances), you could tag a bunch of them with the nosigcov template (and some of these are clearly notable like Bechestobill). There's several logical different ways they could be dealt with. Getting rid of them by the thousands? No. BeanieFan11 (talk) 00:24, 25 February 2023 (UTC)[reply]
I'm curious about your view on what reasonable thresholds are. How many could I prod, redirect, or afd at one time, and how long should an article have a notability template before someone should move on to one of the other steps? Levivich (talk) 01:31, 25 February 2023 (UTC)[reply]
Just noting that I'm aware of the issue with medalists still being in the list, and will resolve that before opening the RfC - thank you for that edit. BilledMammal (talk) 10:42, 27 February 2023 (UTC)[reply]

2nd reaction Paradise Chronicle

We could see if Lugnuts applied as mandated by WP:MASSCREATION for the permission of mass creating articles here or in a Good Faith assumption somewhere else. I believe Lugnuts was aware where such a permission was to be granted, Lugnuts contributed to three discussions of the venue here, here and here, but those do not appear involving a permission of mass creation. I don't know if the last consensus on GNG also works retroactively but... why not. Else, of course, in case such opposes like the ones of BeanieFan11 pile up, we could add incentives such as "notifications that usually only the creator receives" to the article expanders, which received a majority of support votes in the RfC of 2022, but the discussion closers at the time didn't know what to do "divine" of it. The solution of the WP:CSD#X3 is also good. Just get it over with, I am observing discussions on at scale since months and I believe they go on since years.Paradise Chronicle (talk) 01:53, 25 February 2023 (UTC)[reply]

I searched RfA; a number of editors applied for mass creation approval, but Lugnuts was not one of them. I don't know if he applied elsewhere, but I believe it would have been brought up at the ArbCom case if he did. BilledMammal (talk) 02:39, 25 February 2023 (UTC)[reply]
Am I correct that in this case the articles that create months of discussions were created violating policy? Its ok to apply WP:IAR, but months of discussions could well be seen as disruptive to the project.Paradise Chronicle (talk) 03:32, 25 February 2023 (UTC)[reply]
It's debatable. I think they were, but other editors disagree - I would note that most mass-creators have either stopped mass creating articles or been banned, which supports the notion that it is disruptive. BilledMammal (talk) 03:35, 25 February 2023 (UTC)[reply]
No, they were not violating policy. Mass creation was then often considered acceptable and additionally Lugnuts' creations always met the notability guideline at the time. We used to hold that all Olympians were notable and mass creation of them was commendable. BeanieFan11 (talk) 14:33, 25 February 2023 (UTC)[reply]
@BeanieFan11 Could you link to the relevant discussions? For any large scale article creation a permission was mandatory since 2009.Paradise Chronicle (talk) 15:24, 25 February 2023 (UTC)[reply]
I can't link exact discussions but as far as I know his articles were acceptable at the time (I didn't think they were automated creations, either) – otherwise he would have been blocked before creating 93,000. BeanieFan11 (talk) 15:27, 25 February 2023 (UTC)[reply]
Without links your "knowledge" is just blah, blah rather a weak argument against the linked consensus from the Village pump (policy).Paradise Chronicle (talk) 15:33, 25 February 2023 (UTC)[reply]
But do you have evidence that his creations were automated? BeanieFan11 (talk) 15:35, 25 February 2023 (UTC)[reply]
Also semi-automated needs permission. Paradise Chronicle (talk) 17:13, 25 February 2023 (UTC)[reply]
BeanieFan11/Paradise Chronicle - ALL of these mass-creations were in violation of WP:MASSCREATE. WP:MEATBOT is explicit that rapid-editing by hand is also mass-creation. For some reason Lugnuts thought that if he just said that he wasn't using any tools (and he probably was using a template) then it would be OK but this was not the case. It's the same for the stuff made by Carlossuarez46, in fact if you look at the top article-creators of all time (Dr. Blofeld, Ruigeroeland, Starzynka etc.) they were all violating this except for the ones who were approved bots (although those got shut down eventually because the approval was removed) and the ones who racked up their article counts before it was implemented. The fact that most of the top-ten article-creators are now banned/retired under a cloud shows how the community feels about this kind of behaviour (in fact, only one of the top ten article creators is still an active editor in good standing, and they stopped mass-creating ages ago). FOARP (talk) 10:12, 27 February 2023 (UTC)[reply]
Seeking permission for semi-automated mass-creation of articles just isn't a thing that anybody has been doing. I'm not disputing that seeking permission has been called for since 2009/2010. But in practice, apparently nobody had actually sought permission until July 2022 and permission has only be sought one other time in August 2022 (I had asked for examples of permission in the August discussion where I was pointed to the July discussion; other examples of permission for fully-automated mass-creation were supplied as well). In the 2015 analysis of Wikipedia's 5 millionth created article there wasn't a peep about blatant MASSCREATION (although one editor did express some concern that one of 25 articles about different camera lenses created in a single minute might not survive a "deletionist attack"); in that single minute, 56 articles were published among the 7 editors who submitted 2+ articles. The Wikipedia communities attitude about semi-automated mass-creation of stubs isn't the same in 2023 as it was during the run-up to 5 million articles in 2015, but .
MASSCREATE is vague; is 25-50 articles in a day? Or in a subject area in a single editors Wikipedia career? Lugnuts is one of a very small number of editors to get into a consistent 25-50 article/day creation rate. Lugnuts claims to have introduced errors in their articles. There's no guarantee that a Lugnuts article with substantial edits by other people has been thoroughly fact-checked. I'm fine with draftifying Lugnut's Olympian stubs, but I'd favor that on concerns about falsehoods/hoaxes, not a MASSCREATION protocol for semi-automated creations that has basically never been followed. Plantdrew (talk) 02:43, 28 February 2023 (UTC)[reply]
He did claim to have introduced errors, but I've checked many Lugnuts articles in the past and have rarely found errors. I think he was just saying that to piss off the editors against him, knowing he was about to be blocked. BeanieFan11 (talk) 16:42, 28 February 2023 (UTC)[reply]
More