This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

I have concerns about the practice of tagging new categories (less than seven days old) for speedy deletion under C1. While I understand this helps maintain Wikipedia, and the category Empty categories awaiting deletion has been created, it still doesn’t seem right. The text of C1 states, This criterion applies to categories that have been unpopulated for at least seven days. So, according to this criterion, a category is ineligible for deletion under C1 if it is less than 7 days old, since it can’t be empty for 7 days if it hasn’t existed for 7 days, and therefore it should not be tagged for speedy deletion. Yet, this still happens. Why don’t we change it to clarify that categories must have been empty for at least 7 days to even tag it in the first place? We could instead use a template to be placed on the creator’s talk page that informs them the category will be nominated for speedy deletion if it remains empty after 7 days. MrSwagger21 (talk) 04:03, 23 June 2020 (UTC)

Because then categories can't be speedied for being empty unless they've been empty for 14 days. Tagging them C1 doesn't immediately make them speedyable, it starts the timer. —Cryptic 05:24, 23 June 2020 (UTC)

@Cryptic: That may be correct, but where is that stated? I don’t see anything about having to wait 7 days on the C1 criterion, only that it must have been empty for that long. Or if it’s not stated anywhere, why? MrSwagger21 (talk) 06:14, 23 June 2020 (UTC)

@MrSwagger21: At Empty categories awaiting deletion it says "eligible for deletion after that tag has remained in place for seven days". That said, I don't think it is a great idea to tag categories that have only just been created. I would favor giving new categories a bit more time to get populated; we are all busy people. Zero^talk 12:00, 23 June 2020 (UTC)

The concerns here sound like they could be addressed by adding something like "This criterion does not apply to categories less than seven days old." That effectively gives new categories 14 days to be populated but doesn't increase the wait time for existing categories. Thryduulf (talk) 12:50, 23 June 2020 (UTC)

Agreed. Zero^talk 13:00, 23 June 2020 (UTC)

Can someone give an example of why you'd create an empty category in the first place? Surely the #1 reason for creating a category is if you already have an article in mind that should be in it. Reyk _YO! 13:02, 23 June 2020 (UTC)

Tracking category. Created one, and it was speedied 3 days later. I don't think I've ever had a tracking category (with or without ((empty category))) nominated that quickly. Primefac (talk) 13:14, 23 June 2020 (UTC)

@Reyk: Maybe the article is still under development and not in main space yet. Zero^talk 13:18, 23 June 2020 (UTC)

Those are both interesting scenarios, yes. Reyk _YO! 14:26, 23 June 2020 (UTC)

When you tag a category page ((db-c1)) or ((db-catempty)) it gets put in Category:Empty categories awaiting deletion. Humans should normally ignore that page; it's patrolled by bots periodically which will perform a WP:NULLEDIT on the pages listed there. This null edit may transfer the page from that cat to both Category:Candidates for speedy deletion and Category:Candidates for speedy deletion as empty categories and a page in either of those two categories may be acted upon. The test for whether or not a cat page gets transferred in this manner is based upon ((REVISIONTIMESTAMP)) + 7 days being earlier than the current system time. So if a cat page that already bears a ((db-c1)) is subsequently edited (for example, by the removal of a parent category) the seven-day timer is reset to the beginning. --Redrose64 🌹 (talk) 16:10, 23 June 2020 (UTC)

@Reyk: Categories with lists of users are affected by this, since they may take time to populate (manually or automated), or if they’re a category that users need to voluntarily add themselves to.

@Redrose64: That may be the way we have it set up, but it still doesn’t align with the text of C1. In this situation, the tag is being used to tell a bot that it needs perform a certain action on a page. However, this isn’t what tags are for. They are for nominating a page for speedy deletion that is eligible for speedy deletion, to be manually deleted by an administrator. If you really think about it, it goes against the principle of “speedy”. Nominating a page and then deleting it after 7 days sounds more like PROD to me. We need to change the text of C1 and/or find another way to get a bot to carry out those automated functions. MrSwagger21 (talk) 18:09, 23 June 2020 (UTC)

Several speedy deletion criteria have a grace period, this isn't the only one ... five of the eleven F criteria (specifically, F4/5/6/7 & F11), for example. --Redrose64 🌹 (talk) 18:21, 23 June 2020 (UTC)

Yes, but just like C1, they still don’t tell users to tag them before they’re eligible. They only specify when they actually are eligible. Is this practice generally accepted knowledge, or something we ought to add in to the criteria themselves? MrSwagger21 (talk) 18:31, 23 June 2020 (UTC)

• While I don't at all agree that this should be removed just because it's different from most of the other criteria I'm not sure if it's actually a good criteria. I've been following CAT:T3 for many months and my impression is that it's mostly navigation templates where all the links are in a different template and templates created by newcomers such as hardcoded infoboxes and a few test templates. While these sound uncontroversial to delete I'm not completely comfortable with deleting them with at least some overview. The navboxes for example may be inappropriate to delete as many navboxes are too large with a more specific navbox actually being an improvement. Most of the rest should be possible to handle with our existing criteria (G2 for testpages, G6 for drafts in the wrong namespace) and the small fraction left is small enough to just deal with at TfD even though they may be completely non-controversial. If these instead were handled at TfD there will be some review from TfD regulars, a notice would be transcluded with the template, the seven day hold would actually be enforced and it would show up on article alerts attracting editors from relevant WikiProjects. There would be no problem for TfD to handle the influx with there rarely being a significant backlog especially for less contentious discussions such as the T3s.
  While it's definitely plausible to handle these using TfD or other criteria it is also worth noting how bad the criteria is from an uncontestability standpoint. While the hardcoded templates clause mostly apply to infoboxes I've also seen it used for less obvious cases such as hardcoded shortcuts similar to ((Unprotected)) which was kept at TfD last year and has occasionally be used to justify deleting article text templates which I would prefer doing through TfD to make sure attribution is handled properly and deal with possible Labeled section transclusion needed after the deletion. With regards to the substantial duplication clause that could mean anything from having two identical navboxes (where the best course of action would be a redirect) to complex templates with the same purpose such as Aerospecs. Thankfully people have been quite restrained with what they nominate here but it just shows how broad the current criteria is. --Trialpears (talk) 13:30, 1 July 2020 (UTC)

  Also pinging TheImaCow who uses this criteria the most currently. --Trialpears (talk) 13:31, 1 July 2020 (UTC)

• @Boing! said Zebedee: I hadn't realised there were more like this - I mentioned several of them just a few days ago, in the section immediately above; see my post of 18:21, 23 June 2020 (UTC). --Redrose64 🌹 (talk) 20:11, 1 July 2020 (UTC)

Is there a consensus as to how many YouTube subscribers is sufficient to pass A7? I'm seeing articles on YouTubers who have millions speedied under A7. Adam9007 (talk) 19:21, 15 June 2020 (UTC)

WP doesn't consider viewership count alone as a factor of notability, so even if someone had a billion subs, with no other sources to talk about it, it should be deleted or taking out of mainspace. --Masem (t) 19:27, 15 June 2020 (UTC)

Subscriber numbers are easily gamed(it's not hard for someone to create more than one YouTube account) so they are not useful in determining notability. What matters is coverage in reliable sources, not subscribers. Someone can have 5 subscribers and merit an article, and someone can have 5 billion and not merit one. 331dot (talk) 19:30, 15 June 2020 (UTC)

Both answers mention "notability" which A7 is explicitly not about... Regards SoWhy 19:37, 15 June 2020 (UTC)

According to this statistic (blog, I know), there were 16,000 channels with more than a million subscribers in January 2020, with each day adding four channels on average. So just having 1m+ subscribers might not in itself be sufficient to pass A7 if there are no other potential claims of significance. Regards SoWhy 19:36, 15 June 2020 (UTC)

Well, I saw an article on a YouTuber that claimed to have 11 million subscribers. The actual channel shows that the person has nearly 12 million. A quick Google search shows that this person is covered by seemingly reliable sources (albeit ones I'm totally unfamiliar with and are in a language I don't understand). IMHO that should have been AfD's or ATD-I'd. Adam9007 (talk) 21:35, 15 June 2020 (UTC)

I think subscriber numbers and views are inherently unreliable and should not be included in our articles in most cases. However that view doesn't have consensus. And as such I would consider a million subscribers to be a CCS for A7 purposes. Best, Barkeep49 (talk) 23:51, 15 June 2020 (UTC)

At the very least, such a claim should prompt taggers and reviewing admins to do a quick search because very high numbers of subscribers usually means there is RS coverage out there. Regards SoWhy 08:13, 16 June 2020 (UTC)

This page might be worth considering - only 40% (rounded) of YouTubers with 1-2mil subscribers are kept at AFD. Primefac (talk) 00:16, 16 June 2020 (UTC)

8% kept at < 100,000 Subscriber count means that subscriber count is not good enough for a CSD decision. --SmokeyJoe (talk) 00:37, 16 June 2020 (UTC)

(edit conflict)8% kept at < 100,000 Subscriber count means that subscriber count is not good enough for a CSD decisionWhat do you mean? The one I'm talking about has almost 12 million subs. Adam9007 (talk) 00:51, 16 June 2020 (UTC)

I mean, number of subscribers does not appear to be a defining correlation with whether the youtuber is kept at AfD. Therefore, number of subscribers should not be considered, specifically, a low number of subscribers should not be considered for meeting the criterion A7 to delete it speedily. I'm note sure about the opposite, but I'd suggest a very low number, 1000, 10000, for the subscriber number to make it A7-proof. More AfD data would need to be provided. --SmokeyJoe (talk) 02:31, 16 June 2020 (UTC)

Kept at AfD and eligible for A7 are two very very different standards. But in this case, the fact that 40% are kept is an indication that it's a CCS that needs further thought and input and not speedy deletion. The guy who can't believe he's citing inherently unreliable and easily faked statistics approvingly: Barkeep49 (talk) 01:32, 16 June 2020 (UTC)

But in this case, the fact that 40% are kept is an indication that it's a CCS that needs further thought and input and not speedy deletion I was once told the A7 bar is much higher than 40 per cent kept at AfD. More like 60-70 per cent. Just saying. Adam9007 (talk) 01:49, 16 June 2020 (UTC)

Well I would disagree that we should, without community input, be speedily deleting something with a 40% error rate but that's just me. Best, Barkeep49 (talk) 03:19, 16 June 2020 (UTC)

Yeah, I too am against throwing the baby out with the bathwater, but I suppose that's just me too. Adam9007 (talk) 03:26, 16 June 2020 (UTC)

I would consider 100,000 to be a credible claim of significance, and would consider declining an A7 even on the basis of merely tens of thousands of subscribers. _signed, Rosguill ^talk 00:50, 16 June 2020 (UTC)

I would consider 100,000 to be a credible claim of significance, and would consider declining an A7 even on the basis of merely tens of thousands of subscribers. In that case, the one I'm talking about definitely wasn't an A7. I actually considered DRVing it, but decided to take a detour here first. Adam9007 (talk) 01:02, 16 June 2020 (UTC)

1000, in combination with half of another claim, should make it A7-proof. CSD-deletable should be deleted 99-100% of the time at AfD, if they were to be sent to AfD. --SmokeyJoe (talk) 02:34, 16 June 2020 (UTC)

FWIW, my reasoning above is that A7's criterion of a credible claim to significance is essentially asking whether the chance of a subject being notable is less than SNOW. A Youtube channel with 10,000 subscribers isn't going to stand out from other Youtube channels by virtue of its subscriber count, but it's still somewhat plausible that 3 reliable sources would find out about it and choose to write about it. _signed, Rosguill ^talk 02:55, 16 June 2020 (UTC)

I think we are agreeing. I think that given thousands to tens of thousands of subscribers, it demands a little more investigation, as you get with AfD. If trivial, use PROD. --SmokeyJoe (talk) 04:38, 16 June 2020 (UTC)

yeah, I was mostly responding to Adam, but figured that it would be easier to respect the natural threading and put my response below yours. _signed, Rosguill ^talk 04:45, 16 June 2020 (UTC)

Enough of this. If you need to discuss whether a specific article satisfies A7 or not, the answer must be "no"; and the way to discuss it is by starting an AfD. --Redrose64 🌹 (talk) 07:56, 16 June 2020 (UTC)

I think partly an issue may be that there's so many youtubers now, and so many articles for generally non-notable youtubers being created, that if you let every youtuber with a couple thousand subs pass speedy you'd be flooding AfD. ProcrastinatingReader (talk) 17:13, 25 June 2020 (UTC)

Adam9007, how much is it for a thousand subscribers now? Used to be under a hundred dollars. Guy (help!) 21:27, 16 June 2020 (UTC)

This discussion honestly demonstrates how poorly understood A7 still is. The distinction between significance and notability (in the context of Wikipedia jargon) continues to be lost constantly. Glades12 (talk) 18:50, 16 June 2020 (UTC)

The distinction between significance and notability (in the context of Wikipedia jargon) continues to be lost constantly. I agree wholeheartedly. It's been my opinion for quite some time that we should do away with the significance/importance terminology, but I don't think that's likely to happen. Adam9007 (talk) 20:34, 16 June 2020 (UTC)

Even if we're going to talk about significance or importance, anything based solely on factor that can be gamed like directly reading the youtube subscribe counts for significance or importance is not good. Now, if it is the NYTimes that comes along and says a channel had a million subs, okay, thats different and I would say with the third-party mention that that implies something. But if you have a third-party source talking about a channel that way, you probably have more to say about the channel than just sub counts. --Masem (t) 21:26, 16 June 2020 (UTC)

TwitFaceTubeGram subscriber numbers are not an indication of either significance or actual notability, since they can be purchased.[1] [2] "I spent a few bucks" is not an assertion of significance, and we have no way to know if the follower numbers are organic, purchased, or both. Reliable source coverage remains the way to determine significance. Seraphimblade ^{Talk to me} 11:06, 18 June 2020 (UTC)

@Seraphimblade: No it doesn't. Reliable source coverage remains the way to determine notability. Significance is explicitly a lower standard than notability, and whether claims of significance can be verified is irrelevant, all that matters is whether they are plausible. Passing A7 just requires that two questions can be answered "yes" for at least one claim in the article individually or for any combination of claims taken together:

1. Is it plausible that this claim is true?

   For example it is plausible that a YouTube channel has 1 million subscribers, it is not plausible that it has 1 quintillion subscribers. It is plausible that a 16 year old from Ohio is a state chess champion, it is not plausible that they are King of France.

2. Assuming the claim is true, is it plausible that it might make them notable?

   If the only claim of significance is subscriber count, then it is plausible that a YouTube channel with 1 million subscribers will have significant coverage and so be notable. It is not plausible that a channel with 10 subscribers would.

However if there are multiple claims to significance, e.g. the youtube channel is the first one in a language that only has 50 speakers and has 10 subscribers then it is plausible that there will be reliable source coverage of that so it doesn't meet A7. Once again remember it does not matter whether there are reliable sources, only that there could be. Thryduulf (talk) 20:58, 18 June 2020 (UTC)

Thryduulf, A7 is the most confusing and vague of the CSDs I work with (and for that reason, I tend to shy away from using it). Your explanation above is the best one I've ever read, thanks for posting that. Can that be enshrined on the CSD page? -- RoySmith (talk) 22:18, 18 June 2020 (UTC)

RoySmith, WP:CCS, linked to from the CSD page, says basically the same thing. Adam9007 (talk) 23:28, 18 June 2020 (UTC)

Thryduulf, I agree with you in general for significance vs. notability. However, a subscriber/follower claim determines neither. Before knowing they have a million subscribers, they might or might not be significant. If we assume the claim to be true—they still may or may not be; that doesn't tell us. It's not like, say, a claim to be the department chair at a major university or the primary screenwriter of a well-known film, where the assertion, if true, asserts at least a fair degree of significance. Seraphimblade ^{Talk to me} 21:56, 19 June 2020 (UTC)

@Seraphimblade: If they have a million subscribers they might be notable, therefore it is a claim of significance and so the article is not eligible for A7 speedy deletion. Any plausible claim that means someone might be notable is a claim of significance by definition. Thryduulf (talk) 22:11, 19 June 2020 (UTC)

@Thryduulf:, the claim that something or someone exists is a claim that it might be significant, in which case A7 would be a dead letter since presumably the simple existence of an article asserts that the subject exists in some way or another. To be a claim of significance, the claim must be, if presumed true, something that would very likely indicate significance. (It doesn't need to be proven true, but be at least somewhat plausible and indicate clear significance if it were true, not just possible significance.) Seraphimblade ^{Talk to me} 22:14, 19 June 2020 (UTC)

No it doesn't ned to be something that would "very likely" indicate significance. It just needs to be something that means it might be notable - speedy deletion is explicitly only for things that will always be deleted at AfD. A mere claim of existence is sufficient in some cases (e.g. geographical places) it is not sufficient in the case of everything - and YouTube channels are a good example: "FooVision is a YouTube Channel started by Joe Blogs" is a claim of existence but not of significance (there is no chance that this will result in anything except "delete" were it to go to AfD). "FooVision is a YouTube Channel started by Joe Blogs that has 1,000,000 subscribers" is a claim both of existence and significance. Thryduulf (talk) 22:25, 19 June 2020 (UTC)

We purposely do not consider sub counts as any measure of notability alone in any GNG or SNG. It is a metric that can be played with. I would believe that if someone has a million subs it should be fairly easy to produce a third-party source that at least passes a bar that prevents A7 deletion that is related to why they have 1 million+ subs, even if that is not a good GNG/SNG passing article. As soon as you go down the route that "sub count > N" is sufficient to avoid deletion, we are going to get plagued by people demanding to have articles because they have lots of subs or followers. We've spent far too much energy on making sure NCORP and NBIO are tuned to prevent social media from abusing WP, and the route I'm seeing here is working against that. --Masem (t) 23:17, 19 June 2020 (UTC)

Agreed exactly with Masem. This is very open to gaming, and does not provide any kind of substantial indication of significance. If all the article claimed was "1 million+ subscribers" (even were that confirmed), but made no other indication of significance or source coverage, I cannot see such an article ever surviving AfD. Seraphimblade ^{Talk to me} 23:33, 19 June 2020 (UTC)

@Masem and Seraphimblade: You are still confusing and conflating "significance" and "notability". Nobody here is claiming that the number of subscribers alone is a good indicator of notability, I (and I think everyone else in the discussion) agree it is not. Whether an article is or is not notable is irrelevant for speedy deletion purposes and just because something is not eligible for speedy deletion implies absolutely nothing about whether the subject should or should not have an article. Sources are explicitly not required to demonstrate a claim of significance, and indeed the comment that it should be easy to find coverage in reliable sources for any notable channel with a million subscribers is exactly the reason why: sources are probably available (Wikipedia content is required to be verifiable not verified) and it is plausible that those sources will contain other claims to significance and/or notability and so they need to be examined by AfD not by a single patrolling admin. Please actually (re)read and try to understand the criterion, the comments in this discussion and the explanations at WP:CCS before commenting further. Significance is intentionally a very low bar that is very easy to pass because CSD deals only with the most obvious cases that will always be deleted - so by definition if there is any disagreement about whether a page does or does not meet a criterion then it does not. As multiple editors have in good faith expressed the opinion that subscriber count alone can be a claim of significance, any article making such a claim is not eligible for speedy deletion. Thryduulf (talk) 02:00, 20 June 2020 (UTC)

I'm fully aware of the distinction here between significance and notability, and if we were talking anything else like, say, books, newspapers, magazines, or a similar route of publication that was difficult to game, I would see no issue with using some simple sales/subscriber metric as a measure of significance. But social media is a whole other beast. A good marketing company can force numbers to look good for the right amount, and that's what we want to absolutely avoid. Users can beg their existing subs to help boost their numbers which easily can break numbers, but that doesn't change their WP-significance or notability. I have no problem in having some "second chance" language that says "if this person or organization really does have 1M+ subs, it should be easy to find at least one third-party source to back up why they are significant" which I do believe is a fair test. That's not showing notability, but its gets away from just a number that can be toyed with. It is this specific situation around social media that is of concern, not any other situation. It's a unique problem that we know what issues it has created, and despite the fact that CSD is meant to help be a low bar to clear, we need to be more enforcing against "easy" self-promotion that can come off self-claims of popularity from social media. --Masem (t) 02:31, 20 June 2020 (UTC)

You keep saying you understand the difference between significance and notability, but then follow that up with comments that conflate the two so I'm not convinced you truly do. Why do these need to be speedily deleted? Any of these "self-promotional" articles about channels with fraudulent subscriber numbers will get deleted at AfD because they aren't notable, without having to make any changes to CSD criteria or interpretation and without imposing new burdens on articles about social media topics that actually are notable. CSD is the exception in the deletion policy not the rule, every page that is deleted must have consensus for that action, CSD is a list of very specific criteria that, when narrowly interpreted, allow that consensus to be presumed not tested. Thryduulf (talk) 02:43, 20 June 2020 (UTC)

No, it is not conflating the two. An assertion of significance must be plausible, and must demonstrate likely significance if it is considered to be true. Anything and everything might be significant, so just "might be" cannot be the standard; otherwise we have no A7 criterion at all. In the case of YouTube subscribers, those can be (and often are) purchased, and purchasing something does not make something or someone significant except perhaps in the rarest edge case. That is different from notability, which requires that there be available a substantial quantity of reliable and independent reference material about a subject. But "significance" doesn't just mean "Kinda sorta might be", as the answer to that is "yes" in 100% of cases. Given that we have an A7 criterion to begin with, it is clearly obvious that is not the intent. In addition, where's the cutoff? If a channel has 900k subscribers, is that significant? 500k? (I know of one I watch sometimes that has over 700k now, and it is neither significant nor notable. It may go over a million by year's end, and chances are very good it will still be neither significant nor notable.) Seraphimblade ^{Talk to me} 02:56, 20 June 2020 (UTC)

While the logic that is being presented here is internally consistent with CSD and all that, what I'm saying is that from external to WP, if I were a person that wanted to up my SEO and had no coverage my social media but were, say, 100k short of this 1M number that I know that WP will not rapidly delete an article about me due to these CSD guidelines, I would game my audience to get them to get that last 100k and then convince one of them to create the article on me or my channel, and sit back knowing it won't be deleted. That may seem farfetched but we've seen how the call to social media has influenced WP. We don't want to create something easily gamed, that's the issue, and creating a CSD that allows for a presumption on significance only on subscriber counts will be a serious problem towards that. All that is fixed by just asking for a third-party source to confirm # of subs, for example, to show that there's some recognition that the channel is seemingly legit. --Masem (t) 03:10, 20 June 2020 (UTC)

• I* am conflicted here. I certainly don't think subscriber counts should ever be accepted as evidence of notability. But should thy be accepted as a claim of significance? Such a claim must be plausible, we don't seem to have disagreement there, and a channel, having 1`00,000 or even 10,000,000 subscribers ism not implausible in general.Secondly, such a claim must have at least some degree of likelihood that, if supported by sources, it would persuade some to judge the topic to be notable, or it should be b=plausibly linked to circumstances that would establish notability. Say if five percent of all topics with property X would be kept at AfD, then having X is a claim of significance. By that standard a million subscribers would be a CCS I think, but I would prefer not to regard it as such, and insist on some o0ther, any other, claim of significance. Any article that would eventually be found notable would normally have other claims that could be made. However, if reliable sources cover the number of subscribers, I would be inclined to treat that is a claim of significance. DES ^(talk)_{DESiegel Contribs} 03:02, 21 June 2020 (UTC)

  This is some top notch thinking until you decide to chuck it away by saying you don't want to use it (hence the conflicted thinking no doubt). I continue to think that subscriber counts do not belong in our articles except when referenced by reliable independent secondary sources - I hope a new RfC passes where my last attempt failed. However, what we're talking about here is whether or not an article should be speedily deleted. Not even given the chance, over a week, of people finding better sources and/or linking those subscriber counts to reliable independent secondary sources. In the context of DES' analysis I think the anser is of course it should above a certain subscriber count - 1 million give or take. It might still turn out to be completely not notable. But I am, as a matter of principle more inclined to say that we should be erring on the side of notable topics at the expense of editor time (though editor time is precious too and we need to grapple with the fact that we don't have enough of it to support our processes so here I am maybe chucking away my analysis at the end, but at least I'm doing it in a parenthetical comment :) ). Best, Barkeep49 (talk) 04:11, 21 June 2020 (UTC)

  In terms of numbers, I'd say that where the number of subscribers is the only claim in an article, I'd ays 10,000 or above definitely is and below a few hundred definitely isn't. What about the middle range? Well, that would depend on the rest of the article - 1,000 subscribers for a Spanish-language channel about mediaeval Dutch poetry is more significant than 9,000 subscribers for an American channel about video gaming. Note that if there are other claims in the article then the raw subscriber numbers are going to be less important to determining whether a credible claim of significance has been made. In my opinion there is no question that not speedy deleting notable topics is more important than editor time. Thryduulf (talk) 13:13, 22 June 2020 (UTC)

This discussion seems to have stalled, but I'd like to say that Thryduulf has hit the nail right on the head. If we really insist on speedying things that are statistically more likely to be kept at a deletion discussion than not, then we might as well change Speedy deletion is intended to reduce the time spent on deletion discussions for pages or media with no practical chance of surviving discussion. in the lead of WP:CSD to Speedy deletion is intended to reduce the time spent on deletion discussions for pages or media that might not be kept. Sounds ridiculous I know, but I sometimes do wonder.... Adam9007 (talk) 00:28, 4 July 2020 (UTC)

• Yes, the purpose of speedy deletion seems to be often misunderstood, especially by people who do a lot of speedy deletions or new page patrol. Every deletion of every page must have consensus. Speedy deletion is simply a list of situations where that consensus can be presumed because there is no chance that a discussion would ever reach any other conclusion. For example, it is completely unthinkable than any discussion would ever reach consensus to keep or merge a copyright violation, an attack page or unintelligible nonsense, regardless of how many times it is discussed - and they have been discussed enough times that we know this. The same is very much not true of articles about YouTube channels - it is completely plausible that the article about a channel with 1500 subscribers might be kept so it is not eligible for speedy deletion. Thryduulf (talk) 10:28, 4 July 2020 (UTC)

  Thryduulf, Only 1500? But you're absolutely right about speedy deletion. I think this AfD demonstrates spectacularly that some editors (including admins) neither know nor care about what speedy deletion is meant for. Adam9007 (talk) 16:32, 4 July 2020 (UTC)

  Thryduulf, it seemed like we were on the same page but I guess we're not. 1500 subscribers is a CCS? Can you point me to an AfD where someone in the range of 1500 subscribers was kept at AfD? At the risk of getting back-up on my soapbox but at 1500 a significant percentage of those subscribers could be bots and other manipulated subscriptions. The fewest subscribers of any page kept according Wikipedia:WikiProject_YouTube/Notability#Keep_/_delete_tally is 175,000 and that was on NMUSIC grounds. 1500 seems WAY below any possible CCS. Best, Barkeep49 (talk) 16:42, 4 July 2020 (UTC)

  @Barkeep49: I don't know how to find AfDs where subscriber numbers were the only claim of significance, but as I stated above it really depends on the details - 1500 subscribers for a channel on an obscure topic (e.g. medieval Dutch poetry) is a clear claim of significance, 1500 for a channel about the current hottest video game isn't. It is not possible for a single administrator to make any reliable judgement about whether subscribed claimed numbers are genuine or bought so it is not relevant for the purposes of speedy deletion - it is something that can be discussed at AfD if it is relevant. Speedy deletion is the exception not the norm and a credible claim of significance is explicitly intended to be a very low hurdle to clear. Thryduulf (talk) 16:54, 4 July 2020 (UTC)

  I agree we cannot be making individual determination about the reliability of subscription counts. It's why we shouldn't be including them in the first place. But that's not the discussion here which is why I have not been pressing that point. However, the idea that 1500 is a CCS for medieval Dutch poetry streaming strikes me as extraordinary. It is 100x less than anything kept on that chart. This seems like a stable link and is a search of channels with between 1250 and 1750 (e.g. +/- 250, or a 16% range, from 1500). Can you show that any of these have a reasonable claim of notability? I'm not sure what the right cut-off for a CCS around subscribers is, but 1500 seems FAR below that threshold. Your idea that we need to think of CCS relative to the type of channel they have is an interesting one but even there I'd love to see something that suggests it plays out that way at AfD as my experience has been that "YouTuber" is the bucket they're put in and it's not more finely broken out than that unless other SNGs (e.g. Music) might be in play. Best, Barkeep49 (talk) 17:34, 4 July 2020 (UTC)

  Notability is completely and utterly irrelevant. SNGs are completely irrelevant, because they deal with notability not significance, and notability is irrelevant. Whether the channels in that list are notable is irrelevant, because notability is irrelevant. If a channel about an obscure topic has 1500 subscribers then it's far more likely that sources exist for it than they will for a similar sized channel in a vastly more crowded field. It is much, much better that dozens of non-notable channels get deleted at AfD than for one notable one to be speedily deleted, and if we are serious about countering systematic bias (and we all should be) then we need to be serious about things like this. Thryduulf (talk) 17:48, 4 July 2020 (UTC)

  Again, the issue around subscriber count is not a notability issue directly, it is that using the subscribing counts - directly sourcing that from YouTube - is so open to gaming that it should be considered suspect and not a reliable source to meet WP:V, and thus we can't use that to meet significance so a CSD is appropriate. Now, if a reliable source mentions the channel - even in passing - and says it has, say, 100k subs, then we can presume that the source doesn't consider that number gamed and the significance is there, and the CSD shouldn't be used. That's the issue to be focused on. --Masem (t) 18:16, 4 July 2020 (UTC)

  Since when has A7 required a source? I was under the impression that a believable indication that there could be sources is enough to pass A7? Adam9007 (talk) 18:23, 4 July 2020 (UTC)

  If CSD is allowing articles through without any sources , which a core policy WP:V requires... that's a problem. But let's assume that a different aspect of the article about the topic is sourced but the "significance" is still about the subscriber count and that's left only based on looking directly to the YouTube channel, that's still an issue of the gaming factor, especially if CSD sets a figure like "anyone with more than X subs can be considered significant". This is the area where Internet technologies can work against simple allowance that CSD tries to allow. --Masem (t) 18:36, 4 July 2020 (UTC)

  WP:A7 explicitly states: The criterion does not apply to any article that makes any credible claim of significance or importance even if the claim is not supported by a reliable source or does not qualify on Wikipedia's notability guidelines. Adam9007 (talk) 18:40, 4 July 2020 (UTC)

  And what I'm trying to say is that simply claiming a number of YT subscriber counts that exceed some prechosen number, without any sourcing, can't be considered a significant claim because sub count is something that can be gamed easily, particularly if we publish that prechosen number; if gives a reason for a person who wants to get an article on WP to seek a means to game sub counts to get there. Now, I do agree that when that number starts getting high enough, gaming that becomes harder - a 10M sub channel is near impossible to get there from gaming alone, but the impression here is that the number is going to be much smaller and that's an issue. --Masem (t) 19:05, 4 July 2020 (UTC)

  Well for starters there is no finite number and never will be, because details (e.g. matter type of channel) and this discussion is anyway about where number of subscribers is literally the only claim of significance. As for the main part of your comment, you've fallen into the same trap as was repeatedly done above - it's completely irrelevant whether the claim is correct, only that it is plausible that it might be (and given that at least some subscriber counts are genuine, almost all realistic figures could be correct - and if the channel claims to have got say 1m subscribers in less than a week, well that's a clear claim of significance). It also doesn't matter if an article about a truly non-notable channel that has purchased all its subscriber numbers is not speedily deleted, because it will go straight to AfD where it will be deleted when nobody, after searching, can find any sources. Passing speedy deletion does not guarantee an article, it doesn't even guarantee it wont be nominated for deletion, it literally means only that it does not meet the letter AND spirit of a speedy deletion criterion. Thryduulf (talk) 19:46, 4 July 2020 (UTC)

  Thryduulf, I philosophically agree with you that a notable channel being speedily deleted under A7 is a huge problem. I think my patrolling of that category shows that I put that into practice. Further, you are correct that notability and CCS are different. I believe I made that same point up thread and am sorry I slipped here. But what evidence do you have that If a channel about an obscure topic has 1500 subscribers then it's far more likely that sources exist for it than they will for a similar sized channel in a vastly more crowded field? I have pointed to evidence that suggests 1500 is not a CCS, what evidence can you point to, other than your belief, that 1500 subscribers on an obscure topic is a credible claim of significance? Best, Barkeep49 (talk) 18:31, 4 July 2020 (UTC)

  It's very, very simple: For any given subject the channels most likely to be written about in reliable sources are the biggest ones. The chance of a 1500-subscriber channel about a very popular topic being one of the biggest is infinitesimal. The chance of a 1500-scubscriber channel about a very niche subject being one of the biggest is exponentially highest - especially as the sources that are going to write about that channel are going to themselves be smaller and less well known. With probably 2 minutes research I could tell you the most prominent magazines about contemporary video gaming, despite knowing very little about the topic, but it would take me very much longer than that to tell you the most prominent magazines about medieval Dutch poetry even though I know not much less about it (not least because the sources are less likely to be in English). None of that requires specific evidence because it is blindingly obvious to anyone who chooses to look at it with an open mind. Thryduulf (talk) 18:58, 4 July 2020 (UTC)

• Would you folks agree that business's article that said it had 1500 regular customers was a CCS? How about a web forum with 1500 members? —Cryptic 18:27, 4 July 2020 (UTC)
  • @Cryptic: What type of business? What type of web forum? How is that in any way relevant to this discussion? Thryduulf (talk) 18:58, 4 July 2020 (UTC)
    • Because it's the same question. Where is the line drawn? Now you want to start saying that a 1500 subscriber claim can be a claim of significance in some cases? Then that's the problem. A subscriber count isn't, with no further information than that, in and of itself a claim to significance, since every YouTube channel has a certain number of subscribers. That's just inviting spam, and YouTube channels are a substantial source of "profile"/"get the word out" style spam. (And again, if we have such a criterion, where do we draw the line? If a claim of 1 million subscribers is a claim to significance, how about 999,000? If one isn't and one is, why are we using 1 million as our arbitrary line? If both are, how low before it stops being one, and why are we drawing the line at that number and not one less or one more?) Seraphimblade ^{Talk to me} 19:58, 4 July 2020 (UTC)
      • That is exactly why I've spent the past goodness knows how many comments saying that you can't put an exact number on it because what is significant depends on circumstances - what is massively significant for a channel about medieval Dutch poetry is trivial for a channel on current mainstream video games. If the article is literally just "FooVideos is a youtube channel with 1500 subscribers" then that is not a credible claim of significance because it tells you nothing (indeed it's borderline A3), but "FooVideos is a small-African-language youtube channel with 1500 subscribers." is because that could be a significant proportion of the speakers of that language who have access to youtube. If an article is just spamming or advertising then it should be speedily deleted under criterion G11, regardless of what claims of significance it does or does not make. Thryduulf (talk) 21:55, 4 July 2020 (UTC)
• Thryduulf, while this discussion continues, another one continues elsewhere about how to increase participation at AfD. Many articles already get constantly relisted, and we'd waste editors' time having to spam "delete a non-notable YouTuber" a dozen times a day. Currently some admins use discretion over YouTubers and CSD many of the articles before they hit AfD, which keeps numbers manageable. A consensus that would result in this practice stopping and all those being thrown into AfD instead would be disastrous. We'd have tons of YouTube articles on everyone's favourite YouTuber being thrown around, sourced by celebrity gossip blogs or, arguably more likely but not any worse, not sourced at all. If you're willing to pick up the extra work when these YouTubers hit AfD that'd be a big help. Though I suspect AfD participation is only going to go in one direction. AfD is already a messy pit. Yes, in some cases 1500 can be a CCS, but in the vast majority of cases it isn't even close. And even many YouTubers with hundreds of thousands / millions of subscribers are regularly deleted, even with sourcing (eg see my recent [Guzman AfD], 1M subs). They may be producing good quality educational content, but that doesn't mean they give rise to notability to be on Wikipedia. The vast majority of 1500 sub YouTubers obviously won't even have the RS coverage needed to sustain their articles from OR anyway (the SNG assumptions don't really logically fit the same for YouTubers). I'm struggling to think of YouTubers who somehow only have 1500 subs, yet still meet notability, with their prime claim to notability being by virtue of their YouTuber-ing. They may be notable, but it sure won't be for YouTube, in which case this wouldn't apply anyway. ProcrastinatingReader (talk) 22:10, 4 July 2020 (UTC)
  • Once again notability is explicitly not a concern of speedy deletion. Also, regardless of how many articles there are at AfD, any admin using IAR or similar to speedy delete pages that do not meet one or more speedy deletion criteria are abusing their admin powers and must cease doing so immediately. According to policy, every page that is deleted must have consensus for deletion. The criteria for speedy deletion is a list of the few cases where there is consensus that the consensus to delete does not have to be explicit. The criteria explicitly state they are to be narrowly interpreted. This means that, by definition, any speedy deletion that does not meet the letter and spirit of the criterion of one or more of the criteria are being done contrary to consensus. If there is any doubt that page meets the criterion it does not. In this thread you have multiple people saying that subscriber counts can be a claim of significance, therefore by definition they can be. It is very significantly more important that notable articles are not speedily deleted than it is to reduce the number of AfD discussions.
    If you think that there is a need to speedily delete more articles about youtubers and/or youtube channels than can be deleted currently then what you need to do is to propose a new criterion that meets all four of the requirements listed at WP:NEWCSD. Thryduulf (talk) 22:41, 4 July 2020 (UTC)
    • Thryduulf, as far as I understand, they use A7, sometimes G11, not IAR. And the whole point of this discussion seems to be surrounding the application of A7 to the cases, so I'm not sure it's as clear as the blue sky to say they're abusing their tools. As for notability, you're right, it isn't, but a credible claim of notability is. Feels a bit like arguing over semantics, since I thought I was being reasonably clear in addressing that, although I'm not the best writer so I can appreciate any confusion. To reiterate my main point, I'm not sure any cases exist where a person's prime claim to notability is YouTube, yet they only have 1500 subscribers. The hypothetical you gave above is of "a small-African-language youtube channel with 1500 subscribers". But, they won't be notable by virtue of their YouTube channel, regardless of the reasons why they have low subscribers. This doesn't mean this person/organisation isn't notable at all. They can have a YouTube channel, which isn't so popular, and their claim to notability can be by virtue of something else, and that may be credible. But, again, if a person's main claim to notability is YouTube, and they only have 1500 subscribers, clearly they aren't going to be a notable YouTuber. Even if sources picked up on it, which would be the assumption, they'd almost certainly have more than 1500 subscribers by virtue of that coverage alone, if said coverage is deemed to be "independent, reliable and significant". Do we have any actual examples of articles of this summary which have went through AfD and survived? The page Primefac linked shows 8% under 100k being kept. So that's less than 1 in 10 surviving. I would bet the statistic is far lower for 1500 subs. edit: according to table, 0% under 10k subs are kept at AfD ProcrastinatingReader (talk) 23:12, 4 July 2020 (UTC)
      • @ProcrastinatingReader: Yet again, whether they are notable or not is irrelevant. Whether they survive AfD is irrelevant. I agree that it is exceedingly unlikely that someone whose only claim to fame is a youtube channel with 1500 subscribers is notable. However, some people or organisations who are notable for other reasons will have a youtube channel with 1500 subscribers. If the channel is by a 15 year old from the USA who vlogs about videogames then no, simply having 1500 subscribers is not a credible claim of significance. However if 1500 subscribers is a significant fraction of the potential audience for the topic then it is plausible that this person/organisation is notable for other reasons - perhaps this person is the world expert on medieval Dutch poetry - and the article will be kept for that reason when someone investigates and finds the sources - that cannot happen if the article is speedily deleted. Please stop fixating about the number and read everything I and others are saying. I shouldn't need to be explaining the purpose of CSD for the 10th time. Thryduulf (talk) 00:29, 5 July 2020 (UTC)

        With all due respect, your fixation on 'notable' and ignoring my clear mentions of credible claim of notability in both my responses, as well as your fixation on "they might be notable for something else" which is something I brought up myself and addressed, then I repeated it again in the same reply... this is just perplexing to me. Either I'm so awful at wording things and I can't see it, or you've misinterpreted me twice. I could be wrong, but I don't see how your comments even tries to address the points I've made. And keep in mind, if you're right, clearly it's not just me who is confused, but a large number of admins and respondents to this section, so I'd appreciate slightly less attitude (esp in the edit summaries). ProcrastinatingReader (talk) 00:36, 5 July 2020 (UTC)

        You keep mentioning claim of notability, but A7's standard is lower than that, which is why it says significance/importance instead. Adam9007 (talk) 00:54, 5 July 2020 (UTC)

        I'm aware what A7 is. This is arguing over semantics, whether you wish to call it a credible claim of significance or a credible claim of notability, as I did, what we're discussing is exactly the same thing. I'm referring to the same discussed in WP:CCS and paraphrased by Thryduulf above, in that the idea is: is there a plausible claim made, and if the claim were true would it (or anything said claim implies) suggest that said claim would lead to notability. These are the exact words used in CCS, but it would not pass part B, since the claim is highly unlikely to lead to notability. I hope with this definition, which I didn't believe was required since WP:CCS exists and we established this foundation earlier in the discussion, my argument is clearer. 0% of YouTubers with under 10k subs have survived AfD, per the table Primefac linked. So calling these 'controversial deletions' or 'likely to result in notable articles being deleted' are statements unsupported by the evidence, and all the other points I made which I really don't want to repeat all over again. ProcrastinatingReader (talk) 01:07, 5 July 2020 (UTC)

        And what Thryduulf is saying is that 1500 subscribers won't lead to notability for a video gaming channel, but it might do so for a mediaeval Dutch poetry channel (not saying I agree or disagree). How many articles on channels about mediaeval Dutch poetry have we AfD'd? I'm willing to bet that the answer is 'not many', so they won't be in the statistics. Also, if the topic really is that obscure, the channel isn't likely to have more than a few thousand subscribers at most, so we can't just put a firm number on what is not significant. Adam9007 (talk) 01:22, 5 July 2020 (UTC)

    • (edit conflict) No, just because a few people say "We don't agree" doesn't represent a consensus against doing so, especially when many others here are telling you no. Now, I'm not advocating IAR speedies (except in maybe a few absolutely egregious cases of BLP abuse or the like that doesn't technically fall under one), but these are still A7s and will still be treated accordingly. Now, if it makes an actual claim ("FooTube is the second-largest YouTube channel on the subject of..."), then that would constitute a claim of significance. A simple number does not, and those will and should still be A7d. Seraphimblade ^{Talk to me} 23:15, 4 July 2020 (UTC)
      • @Seraphimblade: No, the entire point of CSD is that it is uncontestable if there are people contesting it (and there are) then it is not uncontestable, by very definition. We agree about claims that are not simply numbers - that's not what this discussion is about. This discussion is about cases where the number of subscribers is only claim, and the argument is that they can be (note not are) a claim of significance. This is obvious because nobody is disputing that a claim of say 4 million subscribers is a credible claim of significance (regardless of whether they are notable or not, or whether it's true or not). As noted above, multiple times, an arbitrary number is not a good idea for various reasons. All this means that some claims of subscriber numbers are credible claims of significance, and some are not, but it is not possible to make a hard and fast rule about what is. Therefore the only conclusion that makes any logical sense is that you have to read the article and, in the context of that, determine what is and is not a claim of significance and that this will be a lot lower numbers in some circumstances than others - i.e. what I've been explaining this entire time (mixed in with repeated reminders that notability and sources are explicitly irrelevant). Thryduulf (talk) 00:29, 5 July 2020 (UTC)
• Based on Wikipedia:WikiProject YouTube/Notability#Keep / delete tally, I would say that 100K subscribers is a prima facie claim to significance. Speedy deleting something with a 30% chance of getting kept is too risky. -- King of ♥ ♦ ♣ ♠ 00:51, 5 July 2020 (UTC)

  King of Hearts, not saying I disagree with your point, but I believe the value for 100k is 8% kept, not 30%? 30% is for <1M subs. ProcrastinatingReader (talk) 01:13, 5 July 2020 (UTC)

  No, that's for less than 100K. All the brackets with 100K+ are 30% or more. -- King of ♥ ♦ ♣ ♠ 01:14, 5 July 2020 (UTC)

  That is, however, only the ones which didn't get speedied, so the best of the best so to speak. I suspect if you looked at the total including speedies, that percentage would be half that or less. Seraphimblade ^{Talk to me} 06:22, 5 July 2020 (UTC)

  Pages that are not nominated at AfD are irrelevant for determining anything regarding AfD statistics. Pages should only be speedily deleted if they will always get deleted at AfD - as in >99% of the time at minimum. Even if we take your guess that only 15% of pages are not deleted, that's still 15% more than would indicate speedy deletion would be appropriate. Thryduulf (talk) 12:51, 5 July 2020 (UTC)

  The essay Wikipedia:Why_I_Hate_Speedy_Deleters#Why_do_editors_have_to_jump_through_the_hoops? says: The rationale that people will often give for ignoring the rules is they feel that the article would be deleted anyway. Thus we might as well delete the article today instead of wasting people's time at Articles for deletion or waiting for a 7-day proposed deletion (PROD) to expire. Given my experience, I really do have to wonder if that's the prevailing wisdom? Adam9007 (talk) 17:13, 5 July 2020 (UTC)

  Pages that are not nominated for AfD are relevant for determining "What proportion of articles written making this claim actually stick around as articles?". Even then, though, that's a very flawed measurement—the articles in question may also make other claims that actually are dispositive, or may even already cite a bunch of reliable sources but someone careless or cutting corners nominates it for deletion anyway. Saying "An article making this claim has ever survived AfD" does not make that a claim of significance, and that's a frankly ludicrous standard. Seraphimblade ^{Talk to me} 19:41, 5 July 2020 (UTC)

  You might think "An article making this claim has ever survived AfD" is "ludicrous" but I don't understand how anything else is possible: If Page X making claim Y was kept at AfD then clearly at least some pages that make claim Y are notable. Any page that might be notable is ineligible for speedy deletion under criterion A7 by definition. Thryduulf (talk) 00:17, 7 July 2020 (UTC)

  It depends though. I mean lots of articles that says someone is an American have been kept at AfD. But I would hope we both agree that an article which merely says "Barkeep49 is an American Wikipedia editor" would merit an A7 despite the fact that there are American Editors who have survived AfD. The full context matters. Best, Barkeep49 (talk) 02:04, 7 July 2020 (UTC)

  Indeed, I've been saying all along that context is required - what is a claim of significance - after all you cannot determine whether a claim is credible or not without it (a 20 year old being a professor emeritus is not credible, a 60 year old being one is). Thryduulf (talk) 10:43, 7 July 2020 (UTC)

• Based on the discussion above, I'd say something > 100,000 subscribers is a sufficient claim of significance to overcome an A7. We keep a fair number of those (I'm hearing 8%?) and that's probably enough we should have a discussion. Clearly it doesn't count toward notability itself. Hobit (talk) 17:03, 6 July 2020 (UTC)

The essay WP:CCS says: the inclusion of reliable secondary sources may itself be an indication of significance, but it doesn't say under what circumstances this is the case in. Am I right in thinking that this means sources that would count towards WP:GNG? What about cases where the SNG has additional requirements (e.g. WP:CRIME)? Adam9007 (talk) 16:59, 6 July 2020 (UTC)

• Significance is explicitly a much lower standard than notability, so if an article has sources that plausibly might meet either the GNG or an SNG then that's very clearly a claim of significance. For example if there is an article about a crime and the article includes a reference to coverage of that crime in a source that is (or plausibly might be) reliable then that article cannot be speedily deleted under criterion A7. Thryduulf (talk) 00:11, 7 July 2020 (UTC)
• The test of A7 says: he criterion does not apply to any article that makes any credible claim of significance or importance even if the claim is not supported by a reliable source or does not qualify on Wikipedia's notability guidelines. Thus the absence of a source is not highly relevant. The presence of any reliable source may strengthen the claim, or make it clear that the claim is credible (if it can be supported by a source, it is surely credible), but mostly the presence or absence of sources should not be relevant to whether an A7 (or an A9) is valid or not. DES ^(talk)_{DESiegel Contribs} 20:28, 7 July 2020 (UTC)

  So reliable independent sources can establish notability but not significance? That makes no sense. Adam9007 (talk) 21:42, 7 July 2020 (UTC)

On rare occasions (I don't have an example) an article may have sources that establish notability but despite this the article has been written so poorly that no claim of significance is made. No, I don't like this either. Thincat (talk) 21:54, 7 July 2020 (UTC)

I cannot conceive any such example either. Mentioning sources that cover the subject is imho in itself a credible claim of significance (i.e. "look, this subject has attracted coverage in reliable sources!"). Regards SoWhy 07:30, 8 July 2020 (UTC)

I think User:Ritchie333/Plain and simple guide to A7 says it better (but then I would, wouldn't I?) where my criteria is the (IMHO) more flexible "could any independent editor reasonably improve this article so it would not be deleted at AfD?" Ritchie333 ^{(talk) (cont)} 20:43, 7 July 2020 (UTC)

• Adam9007 I would put it that no source is needed to establish significance. A plausible claim that if sourced would likely persuade some editors at an AfD that the topic is notable, or that suggests a reasonable,probability that source that would establish notability exist is a claim of significance. I would quantify this as "If of all topics that have property X, 10% or more would be kept at AfD, then X is a claim of significance". That is my opinion, it has no particular consensus.

However, as to Thincat's point, I would ay that if sources already present in the article pretty clearly establish notability, they also serve as a claim of significance. Or to put it another way, if a topic is clearly notable, it shouold not be deleted via an A7.

I would also say that A7 is normally about what is already in the article, not about what could be found with a search. An article that said only "John Knox was a popular preacher." giving no details and citing no sources would be a reasonable A7 in my view, although John Knox is clearly notable . Now if an editor knows that a topic is notable, the editor should not tag the article about it for A7 nor delete it as A7, but rather should add a claim that s/he is confident is accurate. But a good-faith editor is nmo9t required to do a search before tagging or deleting under A7. DES ^(talk)_{DESiegel Contribs} 23:16, 7 July 2020 (UTC)

While WP:BEFORE does not explicitly apply to A7, WP:PRESERVE at least indirectly asks editors to try and fix problems before tagging or deleting. So expecting taggers and admins to do a quick search does not seem unreasonable. Regards SoWhy 07:37, 8 July 2020 (UTC)

• If there's a source that covers the topic to such an extent that it might be part of argument towards WP:N, then no, A7 shouldn't be applied (if it's a totally trivial mention, that's probably not true). Note that CRIME isn't an inclusion/deletion criteria, but just a content organisation guideline, so no, stuff like that is irrelevant. WilyD 06:48, 8 July 2020 (UTC)
• Absolutely. "Randy Marsh is an American geologist" is speedyable, but "Randy Marsh is an American geologist^[1][2][3]" is not if any of those sources could plausibly count towards GNG and/or an SNG. -- King of ♥ ♦ ♣ ♠ 13:25, 8 July 2020 (UTC)

• Amorymeltzer I found it when trying to compile a list and I have to say it has several things that probably would have been missed otherwise. Also thank you for taking care of the twinkle part. I'm guessing it gets implemented in the July update July 17? --Trialpears (talk) 10:49, 14 July 2020 (UTC)

  Yeah, but I'll just do it this afternoon. ~ Amory (u • t • c) 11:30, 14 July 2020 (UTC)

S Marshall boldly deprecated X2 earlier today. I have no issues with this deprecation, especially since S Marshall is by far the most active editor in dealing with CXT cleanup and I fully trust their judgement, but there will need to be some significant cleanup to update tools, templates, categories and documentation. How should we handle this? I am willing to do the cleanup since I'm probably the person who knows best what needs to be done seeing how I did the same for T2 a few days ago, but I would really prefer not having to do a mass revert if the removal doesn't have consensus. --Trialpears (talk) 14:17, 15 July 2020 (UTC)

I'm a little uneasy about this deprecation. SMarshall noted in his summary that there was consensus to draftify the remaining articles in the X2 eligible category. As far as I'm aware, this never happened, and there are a bunch of articles that are problematic as a class still sitting unchecked in mainspace. After this deprecation, there is no special tool available to deal with them. If the consensus to draftify is valid, let's implement it, then depricate the criterion. If that consensus is too old, or otherwise inadequate, let's establish a consensus for what should be done with those articles in a quick discussion, then implement that. I'm not going to revert the deprecation, because I haven't been active on-wiki recently, so I may be missing context, but I am going to ask for a tapping of the brakes. Tazerdadog (talk) 14:55, 15 July 2020 (UTC)

I'd certainly welcome your help with the implementation, Tazerdadog. But AFAICT X2 hasn't been used for ages.—S Marshall T/C 17:08, 15 July 2020 (UTC)

Yep, I'm not arguing to let X2 linger, I just want to make sure we've dealt with the base issue. Do we need to re-establish the consensus to draftify, or can we move straight to asking for a bulk move (probably bot-assisted) to draftspace? Tazerdadog (talk) 19:01, 15 July 2020 (UTC)

I think the consensus to draftify is still very safe. Of the initial 3,600 there are about 1,400 left; but no-one's using X2 to deal with them any more. I prefer to do it manually because I'm finding myself deciding about half of them can stay in the mainspace.—S Marshall T/C 19:16, 15 July 2020 (UTC)

According to quarry:query/46313 it looks like the last deletion using "X2" in the deletion summary was Oval Dreams back in January. I'm a bit suspicious about this data since it indicates that only 142 pages have been deleted with X2 which seems way too low. Perhaps many deletions didn't actually include X2 in the deletion summary before templates and tools were set up to handle it properly? Nonetheless I think it isn't used much and draftification is a suitable option in the remaining making it safe to deprecate. --Trialpears (talk) 19:47, 15 July 2020 (UTC)

I laughed to see that there were no X2s at all in 2018. I didn't edit that year; I had a long old Wikibreak.—S Marshall T/C 20:33, 15 July 2020 (UTC)

The standard of dual fluency in the source language and English was quite high, and probably severely narrowed the number of people who could do it. Any objection to moving them all manually, then depricating X2 after that is done?Tazerdadog (talk) 07:10, 16 July 2020 (UTC)

No actual objection, but I would not understand the purpose of the delay. There's consensus to stop using X2.—S Marshall T/C 11:21, 16 July 2020 (UTC)

I'm just becoming more convinced that this was a good decision from the above. Given that the X2 log is basically a subset of S Marshall's CSD log and they think it is obsolete there is no need to retain it. If it hasn't been reinstated or this discussion becomes significantly more controversial by this weekend I plan on updating templates tools and documentation to reflect this change. --Trialpears (talk) 22:29, 16 July 2020 (UTC)

I'll go ahead and do it now. Didn't think we would deprecate two criteria in the same week ever considering that we've only deprecated one in the past decade. --Trialpears (talk) 11:34, 18 July 2020 (UTC)

So, here's one example Draft:Mary Jane Holland (song), but I see this a lot, where (with some automated tool) articles get moved from the mainspace to the Draft name, ostensibly to incubate it, but probably in fact to delete it by G13. This doesn't sit right with me, but I struggle with figuring out what (if anything) I should do. WilyD 07:01, 13 July 2020 (UTC)

• Why? It's an unsourced microstub- exactly the kind of thing that should be sent to draftspace. If you're suggesting people are using draft space moves as an end-run around the normal deletion discussions, I'd like to see some proof of that. The whole idea seems farfetched to me at first glance; you'd need to think the article is hopeless, trust that nobody's going to touch the draft in six months but also expect a bunch of people to come to AfD to go keepkeepkeep. Pinging @Jmertel23: to get their thoughts on this particular draft space move they performed; I suspect it won't be "I was tryna sneak around AfD". Reyk _YO! 07:25, 13 July 2020 (UTC)
  • I don't think they're typically intentionally trying do an end-run around AfD (given the articles sit for six months, you probably could successfully PROD them). But that is what's actually happening. So if the purpose isn't to get the article deleted, but the practice is to get the article deleted, then we should probably look at the practice. WilyD 07:50, 13 July 2020 (UTC)

Other current examples Draft:Quiniela (Argentina) Draft:Claude Cehes so this is happening a lot. I'm not sure whether they should be deleted, but I definitely feel funny about deleting them per G13. WilyD 08:10, 13 July 2020 (UTC)

Well, I don't think Draft:Claude Cehes should have been moved to draft, but certainly it shouldn't have been tagged for G13 3 days after it has been moved to draft space. Fram (talk) 08:49, 13 July 2020 (UTC)

The wording of G13 is a bit unclear there. Aside from being moved into draft space it is true that it hadn't had any human edits in over six months, so I can see where the misunderstanding came from. Reyk _YO! 09:47, 13 July 2020 (UTC)

I'd certainly support making it explicit that a move into draft space counts as a human edit for G13 purposes. Thryduulf (talk) 09:59, 13 July 2020 (UTC)

Prematurely tagging with G13 is one problem. The initial problem is draftifying without tagging the pages for at least two projects. How were the articles to get eyeballs-on from interested editors without project tags? Evad37, would it be possible to get MoveToDraft.js to prompt/remind the mover to provide 2 projects? Cabayi (talk) 10:30, 13 July 2020 (UTC)

That's a requirement I've never heard of, and which seems rather impractical considering that most projects are moribund and getting two active, working projects for any article seems rather hard. Examples

• Draft:Abdullah bin Ahmad al-Wazir was just moved to draft space without any project tags: which two projects would that be? Biography? Yemen? Wikipedia talk:WikiProject Yemen has had three edits this year...
• Draft:Kalinga Literary Award: same, no projects
• Draft:OdishaDiary: same
• Draft:Sacred Heart High School Mothkur: same
• Draft:Walusimbi(personality): same
• Draft:Bangladesh Jatiya League: same
• Draft:Honey Trehan: same

It looks as if no one tags pages for projects when they move the articles to draft space, and I can't blame them. It's exactly the same when e.g. adding a Prod to an article, it also doesn't get or need project tags at that time. It seems like yet another hurdle in dealing with problematic content, something we can do without. Fram (talk) 11:24, 13 July 2020 (UTC)

Either we take steps to get editors' eyeballs on problematic drafts and act on the promise that draftifying is done with the intent of improving the draft or we say straight out that draftifying is G13 with a 6 month delay. Projects aren't the perfect solution, but they're what we've got. Cabayi (talk) 11:35, 13 July 2020 (UTC)

• Draftifying has always been an end-run around the deletion policy. There's no effort to gain consensus (as with AfD). No restrictions on its use or admin oversight (as with PROD and CSD). You just have single editor with the NPR right with full discretion to send any article to limbo and automatic deletion in six months, most likely without a single other human being properly looking at it in the meantime. But these points have been made by many people for many years, and it still seems to have consensus, so... – Joe (talk) 11:54, 13 July 2020 (UTC)

  Just to clarify: NPR right is not necessary for moving an article to draftspace, nor for using the User:Evad37/MoveToDraft draftifying script. Schazjmd (talk) 14:18, 13 July 2020 (UTC)

  If there was any evidence at all of people sending articles to draftspace with the intent of dodging AfD, as opposed to genuinely thinking draftspace is the right place, it wouldn't be long before there was consensus to do something about it. Looking at some of the examples raised in this discussion, it's clear that the problem is actually the reverse: there's a lot of shite getting sent to draftspace that should be speedied. In particular G11 needs to flex its muscles more. Reyk _YO! 14:24, 13 July 2020 (UTC)

  This has come up several times before, most recently at Wikipedia talk:Criteria for speedy deletion/Archive 77#G13 and articles moved to draftspace. Just go to an archive page and look for G13. --Redrose64 🌹 (talk) 15:26, 13 July 2020 (UTC)

  I don't think anybody is intentionally using draftspace to dodge the deletion policy. The problem is that draftifying is the seductive "easy option". NPPers tend to work fast. When you're presented with an article that is in bad shape but isn't blatantly unencyclopaedic, you have limited options:

  • Fix it yourself and mark it as reviewed, but this is time-consuming and unappealing work when the subject isn't interesting;
  • Send it to AfD, but you'll have to do a thorough WP:BEFORE and write a convincing nomination that shows it will never be a good article, or you'll get roasted but AfD regulars;
  • Use PROD, but you have to be sure that the nomination is "uncontroversial";
  • Use CSD, but you have to make sure it exactly fits one of the criteria, or you'll screw up that all-important-for-RfA CSD log;
  • Send it to draft, simply declaring it "not ready for mainspace" and washing your hands of it.

  I think in that context it's obvious why draftifying has become a crutch. The problem is, there is a reason we make all the other deletion processes difficult: we don't want to delete people's work unless it really is unsalvageable. But unfortunately NPP is structured in such a way that it de-emphasises concepts like WP:PRESERVE, WP:FIXIT, and WP:NOTFINISHED, in favour of a quality control process that separates "reviewers" that check work from "editors" that do it. – Joe (talk) 08:18, 14 July 2020 (UTC)

  This is a good analysis and it's not the first time that I've seen one where encouraging NPPers to slow down will do a lot of good - how do we do that? Thryduulf (talk) 09:09, 14 July 2020 (UTC)

  I'd love suggestions. Quality over quantity is important is stressed in the NPP instructions. Concerns over this topic are why Wikipedia:New_pages_patrol#Drafts basically duplicates substantial parts of WP:DRAFTIFY. I have added bold language about not being a substitute for AfD. I also emphasize this point when discussing draftification with everyone who comes through my New Page Patrol school. On a side note but very few NPP these days are high volume reviewers and to the extent that there are high volume reviewers it tends to be of redirects. This is a reason that we've been struggling to maintain with the pace of new article creation. Best, Barkeep49 (talk) 16:47, 14 July 2020 (UTC)

• Moving to Draft is effectively a 3-month PROD. Exactly as with PROD, it can be trivially fixed by remedying the problem that caused it to be draftified. Draftify --> G13 is an excellent way of getting dross out of the main encyclopaedia with ample opportunity to fix if anyone is even remotely interested. Guy (help!) 08:24, 14 July 2020 (UTC)
  • That requires though that people who can fix it are aware that it needs fixing, and we are doing an appalling job of doing that. Thryduulf (talk) 09:09, 14 July 2020 (UTC)

    Thryduulf, that's a generic issue with any deletion process of unwatched articles. Guy (help!) 16:22, 16 July 2020 (UTC)

  • No, this isn't right because anyone can contest a PROD for any reason, but once it's moved to draft the barrier is much higher. WilyD 09:39, 14 July 2020 (UTC)
    • Anyone can move an article from draft to mainspace, just like anyone can do the reverse. If people routinely (or too often) do this incorrectly (moving decent aricles to draft, or moving unacceptable articles (about acceptable subjects) to mainspace, then we can educate and if necessary restrict these editors. But the process shouldn't be rejected (or mischaractrized) because some people misapply it. The same thing is true for Prod, CsD, AfD, or even article creation; if people too often get it wrong, they will be asked to change, and if that doesn't help they will be forced to change. While "move to draft" may be used too heavily or fast sometimes, it often is also used to be more gentle, to give articles which could be deleted a "grace" period, a second chance (e.g. for promotional articles about notable subjects). Instead of limiting its use, it should be expanded to e.g. articles in other languages (which now get a week or so in the mainspace before action is taken). Fram (talk) 09:53, 14 July 2020 (UTC)
  • Yep, if you ignore the fact that, for most of that six-month period, there is no category to look for (and possibly save) these inappropriately "prodded" articles that are not indexed by search engines. Also, what you wrote is almost directly contradicted by WP:Drafts#Moving articles to draft space. (You did mean six, right?) Glades12 (talk) 20:15, 16 July 2020 (UTC)
    • Sorry, the part I struck now is mistaken. The categories that do exist aren't as well-known nor easy to browse as the PROD-related ones, however. Glades12 (talk) 20:21, 16 July 2020 (UTC)
• Draft:Claude Cehes, discussed above, is on Wikipedia:Administrators' noticeboard/CXT/Draftification list July 2017 which is an old task list that I'm slowly working through -- apparently on my own. Consensus to draftify the articles on that list was reached in this discussion, against the background of the colossal Wikimedia Foundation software balls-up documented here. Please note that virtually all the articles on that list are (a) subject to CSD X2, so technically speedyable and putting them into draft space is my less-deletionist way of decrapifying the encyclopaedia; and (b) in violation of the TOU, because they're translations that don't credit the original contributors. (They could be brought into compliance with the TOU by putting the correct variant of ((translated)) on their talk pages.)
  More generally, I'm appalled by the tone of this discussion. I mean, how dare the little people move things to draft space without jumping through a series of hoops? Clearly, those who're doing NPP or working through old backlogs have infinite time to spend on this stuff. Please: get real. As a result of the project's various failures of recruitment and retention, nowadays there are only about 500 active mainspace editors. The options are trust them, or drown in spam.—S Marshall T/C 15:58, 14 July 2020 (UTC)

•  Note: I crunched some numbers: turns out 464 draftications happened during the last 1 week period. A detailed report is available at User:SDZeroBot/DraftifyWatch containing a brief excerpt from the article and the comment used by the moving editor. I can also get the bot to update the page periodically so that we can at least some oversight on the process. SD0001 (talk) 17:29, 14 July 2020 (UTC)
  • A quick glance through that list shows some drafts that really shouldn't be there - Draft:Dry toast and Draft:Calso Water should have been prodded, Draft:106.2 FM should be at AfD. Thryduulf (talk) 20:50, 14 July 2020 (UTC)
• I agree that mainspace articles should not be moved to draft unless there is some realistic chance of them being worked on in the near future. If the page has potential, it should be left in place until someone comes across it and is inclined to work on it. If not, prod or AFD. People who move pages to draft for "incubation" will invariably deny that this is a means of deletion, but in practice that is commonly what happens. Unless the page has been created by a regular editor who is still around, the chances of it getting noticed in draft are slim to non-existent. It has a much better chance of someone giving it love in mainspace. That could still take years, but we shouldn't worry about that because there is no deadline. SpinningSpark 11:27, 1 August 2020 (UTC)
• Unilateral draftification is an end-run around deletion policy, yes. It’s a hole for abuse. I was very worried about that. I wrote some sort of practice-based guideline at WP:Drafts#Moving articles to draft space, not very well publicised, but basically what NewPagePatrollers we’re doing already. I’ve watched and not seen any abuse. Basically, everything draftified would be deleted, whether via AfD PROD or CSD, if it weren’t for DraftSpace. —SmokeyJoe (talk) 11:53, 1 August 2020 (UTC)

Has anyone else noticed (mis)use of G2 as a catch-all (especially in draftspace) recently? Adam9007 (talk) 18:20, 1 August 2020 (UTC)

I've run into a couple, yeah. I think a common problem is that people come across a blatantly unencyclopedic article in draft space, often something like a one-sentence autobiography. In many cases, these would be slam-dunk A7s in mainspace, but there's no criteria obviously suited to them in draftspace. The patroller is sure it doesn't belong on Wikipedia, and G2 can often be stretched to meanings of "This person doesn't know what they're doing here", so they tag it and hope. The reality is that a blatantly terrible draft sitting there just doesn't matter, it is NOINDEXed so it won't show up in search, and it will either be worked on or it'll get G13ed down the line, and MFD is there for any corner cases. By all means politely call people out if they're making inaccurate speedy tags, as you would in any namespace. ~ mazca ^talk 20:42, 1 August 2020 (UTC)

At Wikipedia:Deletion review/Log/2020 August 1#Weavers' cottage (Kleinschwarzenbach, Zum Weberhaus 10), User:DESiegel told me to come here with my opinion that contested speedy deletions should be auto-referred to XfD.

I think this is simple logic, IF the editor making the contest could have removed the CSD tag before the deletion. I think it is further a good idea, if the contest is good enough to be worthy of a discussion.

In practice, this happens sometimes. More often, a discussion occurs on the deleting admin's user_talk page, which is non-ideal if the substance of the discussion are the details of the topic. Important topic source discussions are not easily found when located in user_talk. Very often, the discussion progresses to WP:DRV, where it bogs down to an AfD-style debate, but on unclear lines of debate with respect to whether the test is normal deletion (should it be deleted) or speedy deletion (was it so clear cut that speedy was right). This makes the discussion quite confusing. This is the case at the [[Weavers' cottage DRV. There are reasonable statements such as "this isn't technically speediable, there's also no chance it could survive a proper deletion discussion at AfD", but that is a technical side point to the purpose of DRV.

A speedy deletion should not be sent to XfD if:

(1) The complaint concerns failure of the deletion process itself, or the behavior of the deleting admin, etc

(2) It was a G10 or a G12.

I propose that the following statement, or similar, be agreed to: "For most speedy deletions, if the deletion is contested by an editor in good standing, it is usually better to immediately list it at XfD, and to have the discussion in the XfD format."

If that is agreed, I would proceed to encourage "Speedy close and list at XfD" of DRV discussions like this one. --SmokeyJoe (talk) 04:04, 6 August 2020 (UTC)

Actually, we should go further and say: Any user other than the creator may appeal for an article to be restored at WP:REFUND. The attending admin should evaluate the situation and either deny the request if they believe the deletion was clearly correct or restore and list the article at AfD otherwise. -- King of ♥ ♦ ♣ ♠ 04:10, 6 August 2020 (UTC)

I'd agree with that, but this is often coming from contests by a creator, and where the benefit of discussion is explaining to the newcomer what acceptable sources are. At AfD, the sources are the focus of discussion. At DRV, the discussion turns to CSD nuances, and this does not help the newcomer. --SmokeyJoe (talk) 04:18, 6 August 2020 (UTC)

I am not opposed to this idea but I think it plays out much more clearly when we're thinking A7, A9, A10, and G11 than many of the other criteria. I do have some questions. So if I delete someone's test edits as G2, I'm obligated to restore and send to AfD? Am I also obligated to do a full BEFORE as I normally would do before nominating something for deletion? And just what is an "editor in good standing"? At what point does an editor asking why an article was deleted turn into a contested deletion? Best, Barkeep49 (talk) 04:28, 6 August 2020 (UTC)

Hi Barkeep. Definitely talking mostly about borderline A7 and G11. Also files I expect, but few file speedies go to DRV. Not A10, lets stay clear of A10. A10 disputes are worth the level of attention found at DRV.

If you delete someone's test edits per G2, and they explain to you that they they were not mere test edits, and you disagree but it could be debatable, then yes, undelete and send to AfD. If you think something is speediable, surely that is a trivial BEFORE task? Check the history for vandal content removal.

What is an "editor in good standing"? It is a Wikipedia term of art. Minimally, a non-blocked, non-banned editor, but it allows discretion to ABF with an IP or an account with no edit history. It couples with "reasonable contest", it is subject to interpretation and discretion. If you get a clueless protest on your talk page, don't send it to XfD. However, if a reasonable protest gathers any sympathy at a formally launched DRV, it belongs at XfD first. — Preceding unsigned comment added by SmokeyJoe (talk • contribs) 05:41, 6 August 2020 (UTC)

No BEFORE is a trivial task. There can be test edits on a notable topic and if I'm sending something to AfD, G2 or otherwise, I'm spending the time to do a BEFORE the righ tway. As for editor in good standing, I don't think an editor of 1 month with say 30 editors who attempts to write an article that fails A10 is necessarily here in bad faith. I also don't think they have the standing to make an accurate challenge. The proposed wording doesn't talk about the protest itself. It talks about the editor making it. I would much prefer something akin to what you wrote in this reply than what you originally wrote. I still think I could get behind the core of this proposal if limited to the four CSD I listed before and with some other improvements to the wording (though Wily's point below about the value of DRV is a really good one). Best, Barkeep49 (talk) 15:14, 6 August 2020 (UTC)

• I'm opposed to this. Especially where G11 is concerned, this could keep articles that need to be deleted on the site for up to an additional three weeks just because someone disagrees with the speedy. Contested is different than controversial. SportingFlyer T·C 06:07, 6 August 2020 (UTC)
• Oppose from the principles at the top of this page: Most reasonable people should be able to agree whether a page meets the criterion and It must be the case that almost all pages that could be deleted using the rule, should be deleted, according to consensus. "Most people" and "consensus" don't translate to "unanimous", and it is very frequent for someone to object, such as the creator. If some page genuinely meets the criteria for speedy deletion then it's unacceptable in its present form and shouldn't be kept around. Allowing PROD-like contests for speedy deletion would lead to perverse consequences of the type that speedy deletion is supposed to prevent, e.g. an article written by a kid about his pet hamster, a promotional article written by a spammer or even a piece of blatant vandalism would have to be kept around for weeks just because somebody objected to speedy deletion. Hut 8.5 06:45, 6 August 2020 (UTC)
• While I understand where the sentiment comes from - a deletion being contested is often a sign that it is controversial and CSD is only for uncontroversial cases - I don't think that any disagreement automatically makes a decision "controversial". I'd expect that many spammers would try to string the removal of their spam along if such a rule were to become reality. Also, many times creators of a CSD-ed page contest the deletion because they don't understand our expectations. That means that they need to be educated on our standards but it doesn't imply that the deletion itself is controversial. Jo-Jo Eumerus (talk) 07:51, 6 August 2020 (UTC)

• Oppose- I would not want spammers and unrepentant copyright violators to be able to force an AfD even when the CSD criterion clearly still applies. This proposal also seems to mandate an XfD discussion even when a speedy is contested and the reason for deletion no longer applies. Reyk _YO! 09:17, 6 August 2020 (UTC)

• I kinda dislike this for another reason - DRV reviews give some feeling for what is and isn't an appropriate deletion for those of us closing XfDs and/or processing speedies, so I like being able to see what "controversial" deletions are seen as appropriate and which are inappropriate, you know WilyD 09:46, 6 August 2020 (UTC)

• I am opposed to this idea for several reasons, although I would accept it if a wide consensus for it develops, and possibly i could be persuaded to change my view.

  First let us consider the current process, which is fairly clear IMO, but perhaps should be better documented. If an article or other page is speedy deleted, and an editor objects or questions this, the editor should normally go first to the user talk page of the deleting admin. In some cases the admin, considering the objection, may restore the page. I have gotten a number of pages restored in this way, and I have seen less experienced editors get such a result also. More often the admin will explain the reasons for the deletion mo9re fully, and the objecting editor will accept them, or at least decide not to take the matter further. In the majority of cases that ends the matter one way or the other without a full-dress discussion at any venue. I think this is highly desirable.

  If an editor is not satisfied, or if the deleting admin i8s not available, the matter may be taken to WP:DRV. That forum serves, in significant part, as a review of admin actions in doing speedy deletions. When an admin becomes over-hasty with the delete button, DRV can say so. DRV discussions were a significant part of the evidence in a recent ARBCOM case where an admin was de-sysoped for a patter of over-hasty deletions. I hope this will be rare, but DRV discussions can also serve as precedents for what is and what is not a valid speedy deletion. IMO it does a reasonably good job at this, better than an AfD, where the prime emphasis would be on the merits of an article, would do. DRV might find that an A7 speedy, say, was not justified, even though AfD later found that the topic was not notable. These are two different determinations, with different standards, and it makes sense to have them made in different discussions. A wide admin will take note if a speedy deletion is overturned at DRV, and consider that in making future speedy deletions. (WilyD said much the same above.)

  I should mention that DRV recently clarified its procedures so that consulting the closing admin, while strongly recommended, is not required. If the requesting editor feels, from past interactions, that no useful,purpose would be served by consulting the deleting admin, that step may be skipped. Or it may be skipped for other reasons, or though ignorance of the procedure, without leading to a speedy close of the review for not touching second base. Thus the deleting admin is not a gstekeeper who can block a review, or delay it indefinitely.

  Now let us consider the overall purpose of speed deletion. It is largely to avoid a full-dress XfD in cases where there is a wide consensus that certain kinds of pages should be deleted, with no need for individual discussion. If any editor can force a full, XfD on any page after a speedy deletion, it doesn't save so much. Hut 8.5 is right about this in comments above. Also, speedy deletion allows for quick removal of pages that we really are not willing to tolerate during a full discussion. A10 attack page, G12 copyvios, G11 spam, and G3 hoaxes come to mind, particularly in article space. An Afd would require that these be undeleted, and moreover be open to editing, for a full week or more, as a significant part of an AfD is the ability to improve an article while it is under discussion. That would mean it would not be hidden behind ((tempundelete)) as pages restored for DRV discussion are, and would be picked up by search engines. This is a significant cost -- is any gain worth it? Or do we need to have a complex of rules to specify when a contesting editor has the right to force an XfD, and some forum to determine how these rules apply in a particular case? I think DRV, which can and does asses whether a speedy deletion was valid, and if it is found not valid, may restore the page or send it directly to an XfD, does this better than any such complex of rules would. We are not being so overwhelmed with DRV discussions that we need to send them to XfD instead, in my view.

  In short I see several problems with this, and not enough gain to make it worth while. I also think that it would be a sufficiently major change in procedure tom require a centrally advertised RfC, but that is a detail. DES ^(talk)_{DESiegel Contribs} 15:39, 6 August 2020 (UTC)

• This is reasonable for some pages - categories, files, redirects; perhaps user pages, drafts, and A7s - but ranges from nonsensical to utterly disastrous for most speedied articles. In particular, applying this to G11s would be a major victory for those abusing our encyclopedia for self-promotion. Speedy deletion explicitly considers only the current and past versions of a page, while AFD evaluates a topic. Meaningful improvement while at AFD and deletion on solely WP:TNT grounds are both rare, especially compared to a typical 75-100 G11 deletions per day; and taking something to AFD based on utter lack of NPOV almost always results in the "keep, is googleable, AFD is not cleanup" fallacy. —Cryptic 16:04, 6 August 2020 (UTC)

  Note, when Cryptic says perhaps user pages, drafts, and A7s might be appropriate for XfD on demand, one should recall that A7s are frequently restored, either to article space or to draft space, on request at WP:REFUND. Drafts are often deleted for G11 and g12, and sometimes user pages are also, as well as U5. I actually think U5 is overused, but the creating user is likely to be the one complaining, and their views would probably not be trusted to restore an alleged U5. So I'm not sure that even this list of exceptions isn't too wide. DES ^(talk)_{DESiegel Contribs} 16:55, 6 August 2020 (UTC)

• My experience over 12 years is that most admins are more conservation about deleting speedy, now that so many other routes are available. Most A7s are not restored, and the ones that are, generally do get fixed; very few other speedies are restored. Anyone can in good faith appeal, and that is all that's required. Most admits are receptive to appeals that make any sense at all, though I recognize that some admits never restore an appealed speedy regardless ofthe merits; Having every appeal automatically go to AfD is an attempt to solve a minor problem by creating what could easily be a very major one--afd is now, and always has ben, too busy to give most items sent there adequate attention. Sending hopeless ones there will not make things better. DGG ( talk ) 05:12, 8 August 2020 (UTC) .
• Ok, thanks. —SmokeyJoe (talk) 08:11, 8 August 2020 (UTC)

Hi,

I've noticed that we have R2 for inappropriate redirects from mainspace to other namespaces, but other inappropriate cross-namespace redirects are often handled by G6. Is there a reason we don't just have one criterion that covers all inappropriate cross-namespace redirects? Adam9007 (talk) 18:27, 4 August 2020 (UTC)

The two criteria can service different purposes. If I move Example title to Draft:Example title and leave the mainspace redirect up temporarily so the user who created the page has an easier time finding the draft, but I forget to remove the redirect after a short time, then it makes sense to delete that G6. On the other hand, if somebody creates a redirect in mainspace that points to Wikipedia, User, or even Draft space, then R2 makes more sense. —C.Fred (talk) 18:36, 4 August 2020 (UTC)

@Adam9007 and C.Fred: Also, there is a general consensus that there should not be redirects, or (for the most part) links, from the main article space. MOS:DRAFTNOLINK says: Do not create links to user, WikiProject, essay or draft pages in articles, except in articles about Wikipedia itself (see Wikipedia:Manual of Style/Self-references to avoid). There is not any such consensus for redirects or links between other namespaces. When a page is moved from draft space (or from a userspace draft) to the main article space, a redirect is normally created and may well stay in place indefinitely, possibly for years. Mostly I use G6 to delete redirects only when they obstruct a page move, not just because they are cross-namespace. There will be some cases where a cross-namespace redirect is inappropriate, but that cannot be automatically assumed, and probably should be done by discussion, not by speedy deletion, in most cases. DES ^(talk)_{DESiegel Contribs} 20:55, 4 August 2020 (UTC)

@DESiegel: That consensus is why CSD R2 exists. There are still cases where a redirect might be appropriate. In many such cases, there will be a comment on the page explaining why the redirect is there. I make sure to check page history before deleting a redirect, since sometimes the page is only a redirect because the previous content of the page was blanked and replaced with a redirect. Speedy deletion is appropriate particularly in cases where it's new redirects. There are many cases, though, where other criteria also apply, including G6, G11, and G3. —C.Fred (talk) 21:08, 4 August 2020 (UTC)

C.Fred I think we are in general agreement. I would put it that a redirect out of article space is almost never appropriate, and that is why we have R2. A cross-namespace redir that is into article space, or that does not involve article space at all, may sometimes be inappropriate, but often is perfectly appropriate, and any deletion, whether speedy or not, must be individually justified, and R2 cannot be used as the reason. Would you agree? DES ^(talk)_{DESiegel Contribs} 22:35, 4 August 2020 (UTC) @C.Fred: DES ^(talk)_{DESiegel Contribs} 22:36, 4 August 2020 (UTC)

CNRs to other namespaces than draft, user, etc. shouldn't always be handled by G6, IMO. Deletions of them are often controversial. Glades12 (talk) 07:43, 5 August 2020 (UTC)

Indeed being a CNR isn't, in and of itself, harmful. The harm comes from potential confusion if someone expecting an article arrives at something that isn't reader-facing. RfD recently had a few article → category redirects nominated, a couple were kept, one was retargetted to a template and one was retargetted to an article. Wikipedia ←→ Help redirects are almost always appropriate, Draft → Article redirects are usually kept when nominated, redirects from user subpages to almost anything else are almost never deleted without a request from that user (and many of those that aren't kept are converted to soft redirects, which is what happens to most main user pages that redirect elsewhere). So it's clear that simply being a CNR does not mean it meets the G6 criteria and so should not be routinely deleted unless they really are obvious errors or in the way of page moves. Thryduulf (talk) 21:47, 9 August 2020 (UTC)

MW1: Blank JSON, CSS, JS and Lua module pages with minimal or no history worth investigation. MW2: Content same as other pages or modules. Reason of proposing: Some user MW pages can be useless and will clutter up the server. Although ENWP does not have the Delete page permanently extension, those stuff should be removed from users. ThesenatorO5-2^{argue with me} at 02:31, 10 August 2020 (UTC).

I'm not sure if there is an actual problem here, or if we are just setting up more stuff for admins to delete to an already long list. G2/6/7 seem more than capable to handle these two proposed criteria. Either way, if server space is a problem, the devs would tell us. -- Amanda - mobile (aka DQ) 03:48, 10 August 2020 (UTC)

IMO, this fails both "frequent" and "nonredundant". If you disagree, can you give some examples of pages that would be deletable under your new proposed criteria but not any existing ones? Jackmcbarn (talk) 03:51, 10 August 2020 (UTC)

@ThesenatorO5-2: Did you read WP:NEWCSD, particularly item 3? Can you point to a reasonable number of WP:MFDs (or WP:TFDs, for modules) where such pages were frequently deleted with little or no opposition? --Redrose64 🌹 (talk) 13:20, 10 August 2020 (UTC)

Not only do we not need speedy deletion criteria for these, as my colleagues have explained, but there are blank MediaWiki pages with interesting history, and I don't see why we should hide these from non-sysops and waste server space by marking them as deleted. Here, by the way, are all MediaWiki pages whose deletion has ever been debated: Special:Prefixindex/Wikipedia:Miscellany for deletion/MediaWiki: —Kusma (t·c) 19:35, 10 August 2020 (UTC)

@Jackmcbarn: There are currently not existing ones though. I have changed the wording of the criteria, making it better. — Preceding unsigned comment added by ThesenatorO5-2 (talk • contribs) 01:31, 11 August 2020 (UTC)

They don't have to still exist. Do you have any examples of pages that had to be deleted via a deletion discussion that these criteria would have allowed deleting? Jackmcbarn (talk) 04:08, 11 August 2020 (UTC)

• I don't think any speedy deletion criterion for the Mediawiki: namespace can pass the "frequency" requirement. Every deletion in that namespace ever is at quarry:query/47347; I was initially surprised at how many there were, until I remembered that the earliest implementation of templates were in the Mediawiki: namespace, which accounts for fully half of those deletions (up to page 29). Half the remainder (to page 43) were from a single developer bot run on 7 January 2007. That only leaves 1505 "interesting" deletions afterward, which is barely a hundred a year. Usually we'd want something more like at least five to ten a day for a new speedy criterion. —Cryptic 04:35, 11 August 2020 (UTC)

There have often been articles created about unremarkable mobile apps. Instead of having a new criteria, or using a generic criteria, including them in db-web would make the most sense.

Adjusted wording could be "This template may meet Wikipedia's criteria for speedy deletion as an article about a website, blog, web forum, webcomic, podcast, browser game, mobile app, or similar web content that does not credibly indicate the importance or significance of the subject. See CSD A7."

Elliot321 (talk | contribs) 20:30, 8 August 2020 (UTC)

• The last substantial discussion about mobile apps I can find was in 2014, with a brief and ill-informed followup in 2016. (We've discussed and rejected both web apps and software in general more recently.) This seems ripe for reexamining.
  I think most admins are able to identify and agree on importance/significance indicators for mobile apps these days, especially compared to six years ago; that's always been the major hurdle for enacting a speedy criterion for software in general. However, I can't remember the last time I saw a mobile app article IAR-tagged A7 that wasn't also written so promotionally that I felt no qualms about speedying it as a G11. I also don't watch AFD or prod very closely these days, and it's been many many years since I looked directly at the new pages feed, so I don't have any handle on how frequently such articles are deleted by other means, or should be deleted but survive because the red tape at AFD is too much of a hassle.
  (The proposed wording is unsuitable, of course, since mobile apps aren't web content. But we can worry about that later.) —Cryptic 21:13, 8 August 2020 (UTC)

  Cryptic, We've discussed and rejected both web apps and software in general more recently If that's the case, then we ought to remove browser games from db-web, seeing as they're web apps. Adam9007 (talk) 00:51, 9 August 2020 (UTC)

• Mobile apps are not web content; they are software. If mobile apps were to be included in A7, we'd need a new template: e.g. db-app or something like that. Adam9007 (talk) 21:24, 8 August 2020 (UTC)
  • Hence, we can worry about the wording and template later. —Cryptic 21:30, 8 August 2020 (UTC)
  • The line is definitely not clear. For example, PWAs straddle it. Jackmcbarn (talk) 00:54, 9 August 2020 (UTC)
• I think we should just include any content or product which is distributed exclusively over the Internet. So A7 would continue to exclude software sold in CD form, for example. -- King of ♥ ♦ ♣ ♠ 21:39, 8 August 2020 (UTC)

  That would be most software entirely in this day and age, and that seems like a pretty large expansion. (That's even if it's obvious from the article as-written that there isn't an alternative distribution.) --Izno (talk) 23:02, 8 August 2020 (UTC)

  I think we should just include any content or product which is distributed exclusively over the Internet. That was indeed the old definition of web content, as it was stated at WP:WEB. It was changed after an RfC: Wikipedia_talk:Notability_(web)#RfC:_How_can_non-web_content_be_classed_as_web_content?. Adam9007 (talk) 23:44, 8 August 2020 (UTC)

  I think we should keep the narrower focus. As Izno points out, distributed exclusively over the Internet is almost any software these days with physical copies being increasingly rare. Software is imho a too complex field for A7 which often requires some discussion. Regards SoWhy 08:19, 12 August 2020 (UTC)

• That would cover Netflix etc, so just keep to the present definition as G11 captures quite a lot, imv Atlantic306 (talk) 23:15, 8 August 2020 (UTC)

  I long believe that people are way overusing G11, tagging articles that were merely written by people with a COI or having a little bit too positive of a tone. A lot of them would also be eligible under an expanded A7, and the ones that wouldn't belong at AfD instead of being speedily deleted. -- King of ♥ ♦ ♣ ♠ 23:20, 8 August 2020 (UTC)

• I oppose this. In fact I strongly oppose it. In my view judging the significance of creative works such as software needs a discussion, not a single admin. The suggestion above to extend A7 to cover things distributed exclusively by the internet would mean covering not only almost all software, including large complex commercial software, but all ebooks, although books have long been specifically excluded from A7. Remember that A7 is not about notability, it is about the presence of a claim of significance in the article as written. A badly written initial article about a highly notable topic may be a valid A7. With creative works such as apps and other software, this is even more likely IMO than with articles about people and companies. I think the costs are more than the gains here. DES ^(talk)_{DESiegel Contribs} 16:51, 13 August 2020 (UTC)

Per Wikipedia talk:WikiProject Articles for creation#Copy-Paste from Draft into Mainspace, User:Robert McClenon and others have observed instances in which an article is created in draftspace and review is sought, the draft is declined, and then another editor posts a copy of the draft in mainspace. I gather that the subjects posted are usually commercial, raising the specter of paid editing as the motivation for this odd sequence of events. Either way, one editor copying another editor's draft to mainspace without attribution clearly violates the GFDL, and creates a mess of issues raised in the linked discussion. Wherever this happens, it should be a speedy case, even if none of the existing speedy deletion criteria apply. BD2412 T 02:50, 11 August 2020 (UTC)

• I don't oppose a WP:NEWCSD if it doesn't overlap, but to check: Does the following cover the problem?

(1) If it is a clean copy-paste, meaning no new edits to the draft after the copy paste, then solve everything with a WP:History Merge; or

(2) If it is messy, messy to fix with a history merge, or messy as in it doesn't belong in mainspace, then WP:CSD#G12 the mainspace page.

The another editor must be confirmed, because they created a new mainspace page? Threaten to block them (gently or escalate) for the copy-pasting in future, refer them to WP:MOVE.

This is a reminder to not delete duplicates, if there is any chance of losing the required attribution history. If in doubt, redirect, and fix later. --SmokeyJoe (talk) 03:02, 11 August 2020 (UTC)

I would guess that the article-space version would have the usual modifications of an article moved from draft to article space (draft templates removed, dormant categories made active). The issue raised with a history merge in the prior discussion is that it creates a false report that the article has been reviewed. If it is merged from draft to mainspace, the article is in mainspace despite not being accepted as a draft. It could be moved back to draftspace at that point, but that's more work to accommodate a mainspace version that never should have existed. BD2412 T 03:29, 11 August 2020 (UTC)

An alternative to 2b - doesn't belong in mainspace - that's sometimes appropriate is to history merge into the draft. I happened to do one such earlier today, as an alternative to either deleting as R2 (it had been redirected to the draft version by another editor and left that way, which is how I found it) or G4 (a very recent AFD had been closed as draftify). As I understand it, that should also fix the inadvertent autoreview. —Cryptic 03:34, 11 August 2020 (UTC)

It would be useful to have a policy statement somewhere that explicitly lays this out as the correct solution. By the way, I have no disagreement with SmokeyJoe's comments on dealing with editors who carry out such copy-paste moves. BD2412 T 03:41, 11 August 2020 (UTC)

I think we must already, somewhere - WP:CUTPASTE is procedural instructions, not policy, and even WP:Moving a page is the same. The only wrinkle here is that sometimes you don't want to leave the end result in the main namespace. —Cryptic 03:53, 11 August 2020 (UTC)

I did look at WP:CWW (guideline) which indicates using templates... Jo-Jo Eumerus (talk) 08:05, 11 August 2020 (UTC)

Thank you, User:BD2412. New CSD criteria should be: objective; uncontestable; frequent; nonredundant.

• Objective - Copied from draft to mainspace without attribution seems clear.
• Uncontestable - Something needs to be done when this happens, because the copying without attribution violates the GFDL.
• Frequent - I do not have statistics, but it has happened several times in the past two weeks, and I will report each time that it happens.
• Nonredundant - Yes. At present it requires a histmerge, a PROD, or an AFD.

Robert McClenon (talk) 03:56, 11 August 2020 (UTC)

By the way, there are at least two possible reasons for a copy without attribution from draft space into article space. The first is plagiarism, one editor ripping off another. The second is improper collaboration, meatpuppetry or sockpuppetry. Robert McClenon (talk) 03:57, 11 August 2020 (UTC)

I can imagine a new editor unaware of the page move protocol thinking that this was the correct way to move the content from a draft to an article, without ill intent, though I would be hard-pressed to believe that with an editor with any amount of experience. BD2412 T 04:07, 11 August 2020 (UTC)

• This isn't typically a situation where we'd want to delete the page. The attribution issue can be fixed with a single edit (WP:CWW), so that's not a great reason for deletion. If the draft was declined at AfC then there's likely to be something wrong with it, but there isn't anything stopping any editor from moving a declined AfC draft to mainspace. Hut 8.5 18:25, 11 August 2020 (UTC)
• This is excessively complicated. The much easier solution is to history-merge the pages and move the resulting page back to the original location. If abuse is suspected, create-protect the article. I can't say I do it all the time but I've definitely done it. Ivanvector (^Talk/_Edits) 18:33, 11 August 2020 (UTC)
  • If you history-merge the pages, and then move them back to draftspace, that is an action that many would perceive as a de facto speedy deletion. Apparently some editors question whether they can do this without it being laid out in policy. BD2412 T 20:05, 11 August 2020 (UTC)

• I'll ask my standard question for new criteria - can you supply five recent examples that were a pain in the neck to deal with, and would benefit from a new speedy tag? Ritchie333 ^{(talk) (cont)} 20:29, 11 August 2020 (UTC)

• I'm not seeing the non-redundancy with WP:CSD#G12. A recently created blatant non-attributing fork? CSD#G12 does not explicitly exclude "fixable", and if the fix is more work for a negative result than deletion, that seems completely reasonable. --SmokeyJoe (talk) 03:31, 12 August 2020 (UTC)
  • Could G12 be slightly expanded, then, with a statement like "including unattributed copying within Wikipedia"? My sense is that a lot of people don't think of this as a "copyvio" in the way they think of copying a chunk from some commercial website. BD2412 T 04:55, 12 August 2020 (UTC)
    • Unattributed copying within Wikipedia is easily fixed by adding an attribution. In probably most cases, you shouldn't be deleting copying within Wikipedia, you should be teaching the person how to do attributions correctly. It's a baby bathwater casse. WilyD 05:04, 12 August 2020 (UTC)
• A WP:ATD possibility is to draftify, redirect to the source draft version, and warn the editor. —SmokeyJoe (talk) 07:34, 12 August 2020 (UTC)

• In addition to those reasons brought above by Hut 8.5 and SmokeyJoe as well as others, this would be a fundamental change in policy. As Hut 8.5 correctly points out, AFC is not a requirement for new articles and there is no rule that a declined or non-reviewed draft cannot be placed in mainspace anyway. Creating a speedy deletion criterion like that would essentially create such a rule for copy and paste moves. The correct way to handle such articles imho is to merge the history for attribution purposes and then handle it as one would handle it if it had been moved to mainspace or created there in the first place. There appears to be no policy based reason to handle copy and paste moves differently from actual moves (with regards to speedy deletion). After all, most new editors are most likely not aware of such things as attribution and the "move" button. Regards SoWhy 08:15, 12 August 2020 (UTC)
• Is there a technical preventative solution. When a large amount of text is added to a page, software checks for that the text is not a very close match to a preexisting page. If it does, warn the editor about copy-pasting. —SmokeyJoe (talk) 08:20, 12 August 2020 (UTC)
  • An edit filter might be able to do that, but comparing the text with every existing page sounds like it would be very expensive (computationally) to do, especially if the requirement is for a "close" match rather than identical, remembering that the comparison would have to be done for every edit. There is also the possibility of false positives, especially when splitting an article or creating one that is very similar to an existing one (e.g. someone creating articles about a set of items - railway stations on a new line, individual year editions of award shows, warships of a certain class, etc) - although not all of these should not be separate articles it is always something that requires human judgement. Thryduulf (talk) 09:21, 12 August 2020 (UTC)
    • What if the edit filter is restricted to draftspace and userspace pages with a matching title? --SmokeyJoe (talk) 10:46, 12 August 2020 (UTC)
      • You'll have to ask someone more knowledgeable about the technical aspects of edit filters than me, e.g xaosflux or DES. Thryduulf (talk) 11:53, 13 August 2020 (UTC)
        • @DESiegel: because I failed to do so above. Thryduulf (talk) 11:54, 13 August 2020 (UTC)
        • @Thryduulf: for the most part, the edit filter can only look at the contents of the diff being made - more expensive can look at the entire new page source - and more expensive can do some limited things with the recent history of the page being edited; the edit filter can't compare your diff to one page with the source of a different page even if you knew the page name. So this isn't really the right tool for the job - you might want to check with some of the copyright violation bot operators to see if they would want to deal with this though. — xaosflux ^Talk 14:50, 13 August 2020 (UTC)
• To the original point - if the only "bad edit" happening is a copy-paste move, I'd think the best resolution would be to move the history to fix it - not speedy deletion. If the content is otherwise not appropriate, other deletion criteria should already be able to be used, basically the copying part itself isn't really the problem requiring speedy deletion. — xaosflux ^Talk 14:53, 13 August 2020 (UTC)
• I object strongly to this, or any part of it, being made into a speedy deletion criterion. The only actual problem, the loss of attribution, is almost trivially easy to fix using m((Copied)), or one of the other methods listed in WP:CWW, or with a history merge.

  BD2412 says above that a history merge creates a false report that the draft was approved by an AfC Reviewer. It does nothing of the sort. An approving AfC reviewer leaves several specific entries in the page history, which a history merge would not, as well as adding an AfC banner to the new article talk page, which this would not do. One of those normally uses the words "approving AfC draft" in the edit summery. If a non-reviewer took the significant trouble to fake those entries, and knew how to, that is a very different problem than any discussed above, and I have not heard of such a thing. (Note that only approved reviewers and admins can use the AfC review script normally used for such approvals.)

  Robert McClenon wrote above of improper collaboration, meatpuppetry or sockpuppetry. What excludes perfectly proper but ignorant collaboration? That is, an editor thinking in good faith that a draft should be moved to mainspace and that a copy&paste move was an acceptable way to do that. I have seen semi-experienced editors who thought a copy&paste move was an acceptable way to rename a page in good faith. I have even seen experienced editors advise others that where there was no attribution issue -- where the editor involved had made all significant edits to the page -- a copy&paste move was perfectly acceptable, and editors getting such advice may not understand the limitations on it. Please to remember to WP:AGF.

  Several of the above comments seem to assume that is is improper to move a page from draft space to mainspace without AfC approval, or that it is improper to move a page submitted for AfC review and declined by an AfC reviewer to mainspace without approval. That is not correct. Any autoconfirmed user who believes in good faith that a draft is ready for mainspace may make such a move at any time. AfC is not draft jail. We discourage doing so, because unless the reviewer is incorrect, such a page is likely either to be speedy deleted or deleted by AfD fairly promptly. If it survives an AfD, the move is justified in hindsight. Even if not, until we require non-autopatrolled users to start all new articles in draft space, and use AfC (which I have heard proposed), a user is entitled to take the risk of deletion is s/he so chooses. The creating user does not WP:OWN a draft.

  I would have no objections to some project-space page, perhaps a guideline, perhaps a mere supplemental page, spelling out how to hanle this case. In my view, such a page should say something like:

  1. If the nbew mainspace page clearly fits any of the speedy deletion criteria, such as A7 or G11, it may be tagged and deleted just as if it were not a copy.
  2. If not speedy deleted, fix the attributions, by using ((copied)) or a dummy edit, or a history merge. Consider redirecting the draft to the mainspace article.
  3. If the mainspace page does not seem a proper article, use PROD or AfD to suggest deleting it; the draft copy can in that case remain for futher work.
  4. If the copying editor was clearly trying to game the system and knew better, having been previously warned about such actions, that would be disruptive editing and could be dealt with as such.

  That is my view of the way that this sort of thing should be handled. DES ^(talk)_{DESiegel Contribs} 16:26, 13 August 2020 (UTC)

• Oh, an edit filter was mentioned above as a possibility. That could be done. This is not then plae for detailed discussion of a new proposed filter, and I am not an expert in creating filters. But do remember that a filter must be run on every edit in the project, and that if a filter runs too long, the result is that later filters (or maybe all filters) are simply not run and edits that would have been caught by a filter go through. Anything that must check other pages that the one where the edit is being made adds to the time-cost of a filter, I believe, and needs a storng justification. many filters do a quick check for auto-confirmed and exit early if an editor has that right. A filter for this could not do that -- I suppose it could check for extended-confirmed and exit early in that case. The edit filter people generally want diffs of say 10-20 instances that the filter would block, to design the most efficient tests possible. DES ^(talk)_{DESiegel Contribs} 16:34, 13 August 2020 (UTC)

  @DESiegel: not to dive too down a technical hole, but I'm not seeing a variable for an edit filter to say "IF (this added text) IS PRESENT IN (any other page)" - did you have an idea around that? — xaosflux ^Talk 16:55, 13 August 2020 (UTC)

  xaosflux I think you know edit filters much better than I do. Perhaps I was wrong when I wrote above That could be done. But I would think that that when an edit creates a new page X in the article name space, a filter could check if a page Draft:X exists, and if so if it is identical to the content of X. The more we want that comparison to be "fuzzy" -- that is to detect content that is similar but not identical -- the more expensive such a filter would be, I would think. But if it is limited to checking a draft page with the identical name, as was suggested above, I would think it was possible. Whether this happens enough and is a big enough problem to justify an edit filter is another question. I tend to doubt it, myself, but that would be discussed on the edit filter noticeboard, I would think. DES ^(talk)_{DESiegel Contribs} 17:13, 13 August 2020 (UTC)

  @DESiegel: That's above the capability of the edit filters - but like I noted above, it is something that perhaps one of the copyright violation bots could take up (they are a bit 'slower' since they work off of recent changes feed and won't actually prevent the save), but catching and tagging this type of action may be enough to address the problem? — xaosflux ^Talk 18:55, 13 August 2020 (UTC)

  Very well, it seems I was mistaken about the possibility of an edit filter. A bot might be a goodf way to deal with such cases, either a new bot or an add-on to an existing bot, but this isn't really the place to discuss that. In any case, I still think that a new CSD is not the way to handle such cases, nor is deletion under G12, when attribution can easily be provided. DES ^(talk)_{DESiegel Contribs} 22:25, 13 August 2020 (UTC)

• I strongly oppose this, this is a violation of the principle that any autoconfirmed user can move an article to mainspace. Devonian Wombat (talk) 10:49, 15 August 2020 (UTC)
• I also oppose this. These moves should be treated like any other moves. If it's a copy paste move, fix the attribution issue. If it's not, there's no problem with it. If another CSD applies, use that. If not, prod/AFD. Calliopejen1 (talk) 16:34, 19 August 2020 (UTC)
• Oppose. Don't see why this is necessary when a history merge fixes the problem of attribution. SD0001 (talk) 02:34, 20 August 2020 (UTC)

I propose to change:

This applies to pages that are blatant and obvious misinformation, blatant hoaxes (including files intended to misinform), and redirects created by cleanup from page-move vandalism.

to:

This applies to pages that are blatant and obvious misinformation, recently created blatant hoaxes (including files intended to misinform), and redirects created by cleanup from page-move vandalism.

See Wikipedia:Administrators' noticeboard#8 years-undetected hoax article for context on the article that inspired me to bring this up, Battle of Ceber. Even if it turns out to be a hoax in the end, the amount of research required to ascertain that means that it is clearly not a blatant hoax. Often, whether a hoax is "blatant" is subjective, and if a page has survived for several years then it's good evidence that it's not a blatant hoax, if a hoax at all. Given that the purpose of CSD is to reduce the workload at AfD/PROD and we don't have a large backlog of several-year-old hoaxes waiting to be deleted, I think we should just make it a rule that old pages deserve more scrutiny at AfD instead of being speedied as a hoax. -- King of ♥ ♦ ♣ ♠ 17:40, 8 July 2020 (UTC)

Why on earth should we make it more difficult to remove false information? Absurd. Praxidicae (talk) 17:43, 8 July 2020 (UTC)

@Praxidicae: The phrasing used is such that pages that are blatant and obvious misinformation would still come under the criteria even if they were not recently created, it's fair to say. Naypta ☺ | ✉ talk page | 17:55, 8 July 2020 (UTC)

@King of Hearts: The idea behind this is a good one, but would it not be more sensible to add to the end of the G3 criteria, after Articles about notable hoaxes are acceptable if it is clear that they are describing a hoax, something like Non-recently created pages are not generally blatant hoaxes, and should normally go through other deletion processes? I can see a conceivable scenario in which a genuinely completely blatant hoax could just be missed in NPP, so a speedy was still valid; doing it through a clarifying statement rather than making it part of the criteria would suit that scenario better, and equally give deleting sysops a policy line to point to in declining a CSD of a non-recently-created article. Naypta ☺ | ✉ talk page | 17:46, 8 July 2020 (UTC)

• (edit conflict) Support per nominator. In the context of R3 the precise meaning of "recent" is not defined and there is no clear consensus beyond "a few days old" definitely is recent and "more than about a month old" definitely isn't. I don't think that the timeframes for this need to be that short, but I don't have any firm opinions in what it should be. "Recent" meaning "less than about 6 months old" is what first comes to mind but as I say I'm more than willing to consider alternatives. However I do firmly believe there should not be an exact cut-off. Thryduulf (talk) 17:50, 8 July 2020 (UTC)
• I tend to agree with Naypta here: An article that requires extensive research to determine that it is a hoax is not a blatant and obvious hax, and speedy deletion is not the way to deal with it. An article that has been around for more than a year in mainspace, say, is probably not a blatant and obvious hoax, or howm was it missed for such a long time. But an article describing, say, how King Kong is currently being exhibited by Barmum's Circus in Albany New York is a blatant hoax, even if it has somehow been missed for years. Whether gross misinformation on a more obscure subject is a blatant hoax I am not sure, say an article about an 11th century reigning Duke in one of the Germanies that places him in the 13th century instead, or better an article about a person saying that he is an 11th C ruler of a small state, when he in fact never existed and no source so much and claims that he did, a vandal made him up out of nothing, or perhaps imported him form a historical novel. Surely that is a hoax, but it takes some research to confirm that it is -- is this blatant enough to speedy? Fortunately, such cases seem to be rare. DES ^(talk)_{DESiegel Contribs} 19:15, 8 July 2020 (UTC)
• Oppose- I echo Praxidicae's befuddlement: why in the name of $DEITY would we want to make it harder to remove misinformation??? It's bad enough that obstinate individual editors make removing hoaxes difficult without enshrining such obstructionism in our policies. Reyk _YO! 19:23, 8 July 2020 (UTC)
  • It's probably also worth pointing out that the longest-enduring hoax article we ever uncovered was deleted by G3 just last year- conclusive proof that the system works as it currently stands and is in no need of changing. Reyk _YO! 19:56, 8 July 2020 (UTC)
• Oppose If it's a hoax, it should be speedyable. SportingFlyer T·C 19:53, 8 July 2020 (UTC)

• as the criterion stands now, SportingFlyer only blatant hoaxes are subject to G3 speedy deletion. Would you change that? Would you care to engage with my hypotheticals above of obscure hoaxes, or the Battle of Ceber example above? DES ^(talk)_{DESiegel Contribs} 21:10, 8 July 2020 (UTC)

Per Naypta's comment, I think the following would be sensible:

This applies to pages that are blatant and obvious misinformation, blatant hoaxes (including files intended to misinform), and redirects created by cleanup from page-move vandalism. Articles about notable hoaxes are acceptable if it is clear that they are describing a hoax.

This applies to pages that are blatant and obvious misinformation, blatant hoaxes (including files intended to misinform), and redirects created by cleanup from page-move vandalism. Articles about notable hoaxes are acceptable if it is clear that they are describing a hoax. For suspected hoaxes which are not recently created, consider using other deletion processes instead, as the page may not be a blatant hoax.

In strictly WP:LAWYERly terms, this doesn't change anything about the policy. However, because people have a tendency toward too liberal an interpretation of CSD, especially when subjective words are concerned, this provides guidance that old pages generally should not be G3'd, but if it's obvious enough then G3 is still OK. It's saying in essence, if your gut tells you it's a hoax, check again, there might be a reason why it has survived for so long; if you're still sure after double-checking, then you can nominate it for speedy deletion. -- King of ♥ ♦ ♣ ♠ 22:59, 8 July 2020 (UTC)

• Support, with thanks to King of Hearts for taking on the feedback! Perhaps this should be an RfC? Naypta ☺ | ✉ talk page | 23:01, 8 July 2020 (UTC)

  This talk page is well enough attended that we'll probably get enough traffic even without an RfC. But that's certainly an option if it looks like we don't have enough voices. -- King of ♥ ♦ ♣ ♠ 23:06, 8 July 2020 (UTC)

• Support per my comments on the first proposal. Thryduulf (talk) 23:55, 8 July 2020 (UTC)
• I support the revised proposal by King of Hearts, based on the comments of Naypta above. DES ^(talk)_{DESiegel Contribs} 00:58, 9 July 2020 (UTC)
• Oppose - it was a bad deletion, but this fixes nothing. The problem was that something that wasn't a blatant hoax was deleted as a blatant hoax, it would've been just as bad if the article was eight minutes old as eight years old. This is just bureaucracy for the sake of bureaucracy. If people fail to follow rules, adding more rules won't improve anything and usually makes it worse, since the more rules there are, the harder they are to know, and the less likely they are to get read. WilyD 09:03, 9 July 2020 (UTC)
• Oppose As WilyD says, this was first and foremost a bad deletion; what is needed is for admins to follow the policies and guidelines we have (most do most of the time), rather than follow their own interpretations (plenty do too often). It's preventing bad deletions that's important, not changing the rules to make it easier for them to continue to happen. ——Serial ^# 09:17, 9 July 2020 (UTC)

  @Serial Number 54129: How would this chang[e] the rules to make it easier for [bad deletions] to continue to happen? Naypta ☺ | ✉ talk page | 09:28, 9 July 2020 (UTC)

  I don't understand why adding guidance saying basically "If the article is old, chances are a lot of people have seen it and not identified it as a hoax so take a second look and make sure it really is a blatant hoax before applying this criterion." will make anything worse? Thryduulf (talk) 10:37, 9 July 2020 (UTC)

  The longer a page is, the less well it gets read. And adding special notes for every rare occurance just piles on clutter. If there was some reason to think this was a persistent problem it'd be different, but this is closing the barn door after the horse escaped ... via a secret tunnel. It wasn't the problem in this one instance, and it's not an ongoing problem. WilyD 12:22, 9 July 2020 (UTC)

• I'm definitely opposed to actually tightening the guideline; my read on the specific Battle of Ceber incident is that it was a bad deletion based on several parties under-rating the meaning of the word "blatant". It was therefore already not really an appropriate G3 candidate as the rule is currently written. The proposed reminder about thinking extra hard about non-recently-created pages is perfectly reasonable advice, but I do see WilyD's point about it just being another piece of clutter in an already long and complicated page. This is the first genuinely wrong G3 hoax deletion I've seen in a while, I'm not sure it's even warranted. I'm fairly neutral either way on that one. ~ mazca ^talk 15:22, 13 July 2020 (UTC)
• Oppose. I don't find the justification compelling enough for this criterion specifically. Any speedy deletion criterion is less likely to apply to a page with greater longevity than a page more recently created, because the premise is longevity indicates acquiescence. If we want to include that advice for speedy deletion generally, I'd find that acceptable. --Bsherr (talk) 00:21, 14 July 2020 (UTC)
• Oppose new policies to prevent the stupid thing Bob did once. Guy (help!) 09:05, 14 July 2020 (UTC)
• Oppose, no real need to change policies and tie admin hands further. Stifle (talk) 09:58, 23 July 2020 (UTC)
• I agree in principle with the revised wording, but also with the oppose comments above that this may not be required. I'm left somewhere in the middle. We should expect some sensibility from admins to exercise reasonable and proper discretion, given the process they go through for the right. Unless there's a pattern of improper G3 deletions (there might be?) I don't know if this is required? ProcrastinatingReader (talk) 11:35, 23 July 2020 (UTC)
• One cannot count on the nominator to WP:BEFORE. Admins must also perform due diligence-- look at past revisions and consider the possibility of vandalism. It does not hurt to Google search as well. One should view with skepticism the idea that an article that has been around for years is a hoax.04:28, 20 August 2020 (UTC)

I am a big opponent of WP:CREEP, and one way to oppose instruction creep is to remove processes that are no longer needed or used. These two criteria certainly qualify: after last year's portalspace paredown, I don't think we have had any portals deleted under either of these criteria, and certainly not a volume that exceeds MfD's ability to handle, easily, any that come up in the future. I look at it this way: if proposed today, would either of these criteria be added as CSD criteria? Certsinly not: both would fail the frequency requirement easily. We can always revisit and consider re-adding in the unlikely event this is no longer the case at some future date. Thoughts? UnitedStatesian (talk) 02:17, 21 August 2020 (UTC)

• Comment With just 500 portals around, its indeed questionable whether any P* criteria can meet the frequent criterion. SD0001 (talk) 07:12, 21 August 2020 (UTC)
• All P1/P2 speedies. Since there aren't any other portal-specific criteria, we don't have much to gain from removing these unless they're being abused. I don't see evidence of that. Certainly, we don't want to divert A7s into the just-as-indexed portal namespace once self-promoters discover how much more work it's become to remove their autobiographies from there than it is to write them. —Cryptic 10:35, 21 August 2020 (UTC)

I assume that applying an MfD tag to a Portal immediately NOINDEXes it, correct? UnitedStatesian (talk) 01:59, 23 August 2020 (UTC)

• Cryptic's query shows there's only been five P1/P2 deletions in the last year, and two of those qualified for other criteria (Portal:Stony_Creek_Consulting,_LLC could have been G11 or G7 and Portal:Fatimid_Caliphate/Related_portals is really a G6). I suggest we get rid of them. For P1 I suppose we could modify the introduction to the article criteria to make it clear they apply to portals as well, which would remove the need for a separate criterion. Hut 8.5 12:11, 21 August 2020 (UTC)

I am asking where is CSD U4? I only saw U1 U2 U3 and U5. -- PythonSwarm ^{Talk | Contribs | Global} 00:51, 24 August 2020 (UTC)

PythonSwarm, U4 was rescinded yonks ago. Wikipedia:Criteria_for_speedy_deletion#Obsolete_criteria. Adam9007 (talk) 00:59, 24 August 2020 (UTC)

Yonks? You mean years, right? Glades12 (talk) 05:58, 24 August 2020 (UTC)

Sorry, I didn't know that yonks was a word. Glades12 (talk) 06:00, 24 August 2020 (UTC)

Alternatively, make a link to WP:CSD#U4 and see where it takes you. --Redrose64 🌹 (talk) 17:09, 24 August 2020 (UTC)

G8 currently includes: "Unused editnotices of non-existent or unsalted deleted pages"

This would seem to imply that we do allow editnotices to be retained for titles that are protected against recreation. Does that describe current practice? This seems a bit strange to me, given that it is somewhat redundant to protection-reason summaries.

For sake of due diligence, I searched the archives and found all mentions of the exact word "unsalted" to concern other criteria. I glanced briefly at some revision history. --SoledadKabocha (talk) 04:04, 26 August 2020 (UTC)

I can see having an edit notice for a salted page explaining why the page has been salted and what the user interested in writing about that subject should do. It might not be done often, but it does make sense. Primefac (talk) 09:42, 26 August 2020 (UTC)

It was added 28 June 2016 here by User:Andy M. Wang. —SmokeyJoe (talk) 10:03, 26 August 2020 (UTC)

So it looks like the BOLD addition was basically intended to allow non-controversial deletion of edit-notices, previously unmentioned by G8, by pre-excluding any even slightly contentious cases. The whole thing seems sensible to me, while I'm not sure if there are many examples of salted pages with editnotices, it seems like a reasonable possibility, and hence worth it to avoid making it an unambiguous speedy. ~ mazca ^talk 22:32, 26 August 2020 (UTC)

As I was reviewing some of my old edits, a thought came to me: If one of the reasons why WP:FILEREDIRECT exists is to prevent external linkrot, then I suppose by default, in most cases, that means any redirects that are ((R from move))s from the "File:" namespace should not be eligible for speedy deletion if they have been around for a long period of time. So ... with that being said, it seems there should probably be a few criterion that should not apply to redirects in the "File:" namespace.

From what I'm seeing per my aforementioned comment, the criterion which should probably not apply to redirects in the "File:" namespace where either the redirect has existed for a "long period of time" or the target "File:" page had been at the redirect's title "for a long period of time" are G6 and G7. (Note: Criterion R3 already covers some cases of ((R from move)) redirects in the "File:" namespace since it relies on the age of the redirect or target page, and thus will not be mentioned any further here.) Here are some example scenarios on why these three criteria (not in conjunction with any unmentioned criteria) seem to cause issues if "File:" namespace redirects are deleted due to them:

• G6: An editor notices that an image of a dog chasing a ball had the name "File:Usnjestybxsthb.jpg", and thus renamed it "File:Dog chasing ball.jpg" per reason "2" for renaming a file. However, the file had been sitting at "File:Usnjestybxsthb.jpg" for about 5 years, so even though the title makes no sense, that does not necessarily mean there are no external links to the title.
• G7: An editor uploads a file in 2010 at "File:Abc.jpg", and no one else edits it. About 10 years later, the file uploader moves the file to "File:Xyz.jpg". Since the file had been sitting at "File:Abc.jpg" for 10 years, the file could have external links to that title.

...So, I'm proposing that the G6 and G7 criteria have exceptions added for redirects in the "File:" namespace where the redirect is a ((R from move)) and either:

1. The redirect has existed for a "long period of time", or
2. The target "File:" page had been at the redirect's title for a "long period of time"

Steel1943 (talk) 20:21, 25 August 2020 (UTC)

• That makes sense, support. Thryduulf (talk) 13:41, 27 August 2020 (UTC)
• Much of the reason why keep any long-standing redirect is to prevent external linkrot; it's not like we have a current need for AfricA and so on. Why limit this to the File namespace? —Cryptic 13:48, 27 August 2020 (UTC)
  • @Cryptic: Files are quite often poorly named but remain useful nonetheless. We can assume that if a mainspace redirect gets very few views, it never gets used. Deleting such redirects is less hassle than say a file called File:Africaa. I support this proposal as-is, but would not be opposed to nother discussion about non-file redirects. Anarchyte (talk • work) 07:56, 28 August 2020 (UTC)