Wikipedia talk:Speedy deletion

See also Wikipedia talk:Speedy deletions for discussions which took place in 2006-07 before it was redirected here.

Discussion potentially impacting criterion F5 (orphaned non-free use files)

At Wikipedia:Village pump (policy)#Non-free images should be permissible in draft space changes are proposed to the policy about non-free images that would have (almost certainly minor) implications for criterion F5 (e.g. changing "not used in any article." to "not used in any article or draft." Please leave any comments at the linked village pump section rather than here to keep discussion in one place. Thryduulf (talk) 22:35, 16 June 2025 (UTC)

G8 on modifications of redirects

G8, as applied to redirects, is basically limited to a redirect with a nonexistent target, either because the target never existed or because it's been deleted. Wondering about expanding it to cover "related" redirects.

Birmingham, North Warwickshire, and Stratford-upon-Avon Railway Act 1895 redirects to North Warwickshire Line, as does Birmingham, North Warwickshire and Stratford-upon-Avon Railway Act 1895 (without serial comma). Imagine that someone took the with-comma variant to RFD, arguing that we shouldn't redirect laws to railway articles, and the RFD was successful. We shouldn't delete the without-comma variant based on the RFD, since it wasn't nominated, and it's not G8-eligible because its target is alive and well. However, it only exists because I created it a few minutes ago as a variant of the with-comma title, and if the with-comma title were a freestanding article, without-comma would redirect to it. (If double redirects weren't a problem, I would have created without-comma as a redirect to with-comma.) So, if one of them is deleted for reasons unrelated to punctuation, it seems reasonable to delete the other, even if there's no discussion, but no existing criterion covers it. It's likely frequent, since people are quite likely to encounter and nominate a redirect without being aware of the existence of a parallel redirect. But is there an objective and uncontestable way to do this?

A clear but complicated process would be a template — redirect A is marked with a template saying "if redirect B is deleted, this can be deleted under G8". However, that would require an additional edit to every existing page, and the creation of new redirects would take more work, because you'd have to mark it with the G8 template as well as including the normal redirect code and any normal redirect tagging. But if we declared that variations of existing redirects were automatically deleteable, I can see plenty of wiggle room: for example, we wouldn't want to delete A if B were deleted strictly for punctuation reasons or to encourage the creation of an article, and either we'd need long and careful criteria to ascertain what was a variation, or we'd end up with arguments over the same question. In the perfect situation, we wouldn't limit it to tiny variations; for example, if we deleted Broken Hill Proprietary (just an example; I can't imagine anyone wanting to delete it), we'd probably want to delete Broken Hill Proprietary Company too.

Any ideas? Can this be done in a practical way? Or is it just too complicated to make it objective and uncontestable? Nyttend (talk) 06:24, 28 June 2025 (UTC)

PS, this originates from a specific situation. Wood v. Georgia (1981) used to redirect to List of United States Supreme Court cases, volume 450, but it was deleted some time ago. Wood v Georgia (1981), a redirect to the same list, is at RFD; it can't be G8-deleted because it isn't dependent on the deleted "v." title, but because it's a variation, it probably should be deleted. Nyttend (talk) 06:32, 28 June 2025 (UTC)

So, we do kind of have a template, {{avoided double redirect}}, but it includes ADRs from distinct topics, which wouldn't be suitable here (e.g. character → book → series might be appropriate to change to character → series if book is deleted). But I would argue that when the variation is purely typographic, the plain wording of G8 (dependent on a non-existent or deleted page) does apply. The undotted-v version of the Wood redirect absolutely was dependent on the dotted-v version. There's no world where it would make sense for the former to exist while the latter doesn't, just like there's no world where it makes sense for Talk:Foo to exist while Foo doesn't. Noting that the list of G8 examples is non-exhaustive, and in my view already covers this use case, I would support adding a bullet of

• Avoided double redirects that are alternate names of a parent redirect that has been deleted at redirects for discussion

I don't think this would need to apply only to redirects tagged with {{avoided double redirect}}; Wikipedia is not a bureaucracy and in a case like the Wood redirects it's common sense that one is an ADR of the other. But the template would certainly help make G8's applicability clear.

There might be some cases where the solution to this situation would be to designate a new redirect as the primary one in the ADR family, rather than delete the ADRs; this could be noted in a footnote if desires. There might be some other very rare exceptions, but G8 already notes that {{G8-exempt}} exists, so it could be used in those cases. -- Tamzin^{[cetacean needed]} (they|xe|🤷) 06:51, 28 June 2025 (UTC)

Oppose. Far simpler than working out which ones should and should not be under G8 (because unless it's all of them then it's not suitable for speedy deletion) there is currently a bot being programmed that will check for avoided double redirects and typographical variants of redirects nominated at RfD and alert the discussion to their existence. Humans can then add those that should be discussed together to the nomination and consensus applied to all of them at once, while those that don't have that consensus (or have a different consensus) won't be incorrectly deleted. Another bonus is that this will also work for things other than deletion.

See Wikipedia talk:Redirects for discussion#Avoided double redirects of nominated redirects and Wikipedia:Bot requests#Redirects related to those nominated at RfD. Thryduulf (talk) 09:21, 28 June 2025 (UTC)

It's all of them that are "dependent" on the deleted page, i.e. only exist as variants of it. This is consistent with our practice of G8-deleting redirects to articles that get deleted. -- Tamzin^{[cetacean needed]} (they|xe|🤷) 09:32, 28 June 2025 (UTC)

See the WP:NEWCSD criteria. The issue is that there needs to be an objective definition of what is "dependent" that includes only pages that should always be deleted. {{R from avoided double redirect}} is not that as you noted in your first comment, and it absolutely is not any redirect that someone thinks is dependent on a deleted redirect, again the book/character/series redirects are examples of this. Thryduulf (talk) 09:41, 28 June 2025 (UTC)

I'n not proposing a new CSD (not that I think WP:NEWCSD is or ever has been a good summary of the community's expectations for a new CSD). G8 is already a broadly-worded criterion, and I'm saying this scenario falls under it. If you want to make the list of examples under G8 exhaustive, rather than explicitly non-exhaustive as it currently is, you should propose that. -- Tamzin^{[cetacean needed]} (they|xe|🤷) 09:52, 28 June 2025 (UTC)

NEWCSD explicitly applies to modifications of existing criteria as well as brand new ones, and your comment is the first time I've heard someone suggest it doesn't represent community consensus. The question is whether avoided double redirects are both "dependent on a deleted page" and "should always be deleted" and because we both agree that the answer to the second question is "no, some of them should not be deleted" then we need to either agree that none of them can be speedily deleted (which is the current consensus) or come up with some criteria that distinguishes those which should be deleted from those that cannot.

My point is that determining this criteria is pointless as the bot currently in active development will avoid the need to speedy delete any of them because avoided double redirects will be discussed at RfD the same time as the redirect they are avoiding. It isn't going to catch all untagged avoided double redirects, but enough of them that the frequency requirement of NEWCSD will not be met. Thryduulf (talk) 10:02, 28 June 2025 (UTC)

NEWCSD applies to expansions of existing criteria, which this would not be, because we are talking about whether to include an example in a non-exhaustive list to make explicit something that's already allowed by the plain wording of the criterion. But we've both made our opinions clear. Let's hear from others. -- Tamzin^{[cetacean needed]} (they|xe|🤷) 10:15, 28 June 2025 (UTC)

This would be an expansion because the consensus currently is that G8 only applies to redirects that target deleted or non-existent pages. Thryduulf (talk) 10:18, 28 June 2025 (UTC)

What consensus are you referring to? -- Tamzin^{[cetacean needed]} (they|xe|🤷) 10:31, 28 June 2025 (UTC)

This has been discussed a few times and every time there has been no consensus to expand G8 to cover redirects to pages that exist. See for example Wikipedia talk:Speedy deletion/Archive 89#G8 definition of dependent, Wikipedia talk:Speedy deletion/Archive 89##Do redirects avoiding double redirects to deleted redirects fall under G8? and other people have also told you that this does not fall under G8. See also Wikipedia talk:Speedy deletion/Archive 74#Tightening G8 with respect to redirects where G8 was narrowed to the current wording. Thryduulf (talk) 10:47, 28 June 2025 (UTC)

Thryduulf, I have no problem with you disagreeing with me, but please don't assert "consensus" and then link me to one discussion (archive 74) that isn't about this and two (89 and what I assume was meant to be 79) where there is no consensus on this point, and where most of the opposition is coming from you (as is often the case on this page). I will quote Anomie from the Archive 89 discussion: You're applying your own idiosyncratic definition of 'dependent on' and asserting it's the only possible 'literal meaning'. I don't see anything at your lightly-attended RFC [Archive 74] that's relevant here either. If you want consensus, hold an RfC. But please don't say I'm acting against a consensus that is 90% you. You do not have a veto over changes to WP:CSD. -- Tamzin^{[cetacean needed]} (they|xe|🤷) 11:05, 28 June 2025 (UTC)

My point was "there has been no consensus to expand G8 to cover redirects to pages that exist." and linked you to examples of discussions where there was no consensus to expand G8 to cover redirects to pages that exist (and the discussions are about expanding the consensus, not confirming what it already is). I might be the most vocal person opposing a change to the consensus but that does not indicate that the consensus doesn't exist - if you think there is a current consensus for the words in the policy (as gained consensus in archive 74) to mean something other than the literal meaning of the words in the policy then please show me that consensus. Thryduulf (talk) 11:22, 28 June 2025 (UTC)

I think we should just have a bot that adds all ADRs to the RFD nomination page so it can be decided at the RFD whether the ADR is one that should obviously be deleted (like your example of serial comma; that one would fall under WP:NOTBURO) or one where something else should be done, like Tamzin's "book" example. The default could be that all ADRs that nobody speaks up for during the RFD are deleted when the main redirect is deleted. Basically if the ADRs should be deleted after a different redirect is deleted after discussion, just include the ADRs in the discussion. —Kusma (talk) 09:21, 28 June 2025 (UTC)

Meh... I think the other redirect is covered under {{db-xfd}}. If Dewey, Cheatham, and Howe is deleted via RFD and Dewey, Cheatham and Howe exists, the latter should also be deleted unless the rationale was specifically "this isn't a valid use of a comma", since the substance of the RFD holds for both pages. This is true whether the second redirect is discovered the same day or a year later. Primefac (talk) 11:22, 28 June 2025 (UTC)

I don't think these make good speedies; they're too broad and ill-defined a class of a redirects, and there's too much judgment involved to let it become precedent. For the really obvious ones, like Nyttend's missing period and comma examples, you could probably get away with IARing them. Putting the previous RFD (or whatever reason the first one was deleted for) in your deletion log with an explanation that it applies exactly as much to this redirect too is way more honest than picking the not-really-applicable G8 - or, worse, G6 - from the drop-down. —Cryptic 11:42, 28 June 2025 (UTC)

redirect A is marked with a template saying "if redirect B is deleted, this can be deleted under G8" When AnomieBOT creates redirects for en-dashed titles, it applies a custom bot-specific template that does exactly that. 😀 Although the template also mentions G7 and G6 since, as we see above, some people don't like to use G8 for anything other than "redirect target is currently a redlink" and don't accept "redirect would be a redlink if the double redirect wasn't bypassed". But if we want something not specific to bot creations, we should probably add a |dependent=yes parameter to {{R avoided double redirect}} instead. Anomie⚔ 13:39, 28 June 2025 (UTC)

Largely echoing what Tamzin has already said, my position is that {{R avoided double redirect}}s that rely on a redirect that has been deleted already qualify for G8. If not for a software limitation regarding double redirects, it would literally match the Redirects to target pages that never existed or were deleted wording. I would be in favor of making that explicit, and think that a parenthetical or footnote after 'target pages' stating (including avoided double redirects) would be the most efficient way to do it. I also want to emphasize that G8 already has the wording this criterion excludes any page that is useful to Wikipedia which would cover any exceptions that shouldn't be deleted. -- Tavix ^(talk) 19:40, 28 June 2025 (UTC)

Adding myself to the list of people who agree that these qualify for G8. The intro text says examples include, but are not limited to, implying there are other reasons—including, but not limited ot, {{R avoided double redirect}}s. If someone has a reason why a particular redirect should persist, they can remove the rcat. HouseBlaster (talk • he/they) 19:47, 28 June 2025 (UTC)

Rethinking G8 entirely

This discussion, in which multiple admins are interpreting the same criterion and coming to diametrically opposite conclusions, is proving that the current wording of G8 is unworkable and we should consider breaking up the criterion entirely.

Restore the original WP:R1 criterion for broken redirects.

Move Timed Text pages without a corresponding file (or when the file has been moved to Commons) to F2.

Editnotices of non-existent or unsalted deleted pages is within the scope of T5 as well.

Leave G8 just for "subpages or talk pages of a nonexistent page".

If there's want for a new criterion for "avoided double redirects" as proposed here then just make that R5. That would allow us to explain what is and isn't covered in more than a single sentence and not have to rely on vague penumbras.

* Pppery * _{it has begun...} 17:07, 29 June 2025 (UTC)

I agree regarding the R1 and T5 proposals and am completely neutral regarding F2. If people genuinely think that all or some avoided double redirects should be speediable they can get consensus for a new criterion that meets all the NEWCSD criteria. Thryduulf (talk) 19:44, 29 June 2025 (UTC)

"Diametrically opposite" is a bit of an exaggeration, and the conclusion that the current wording of G8 is unworkable is even more so. If anything, all that's needed in G8 is a parenthetical one way or the other for clarity. -- Tavix ^(talk) 20:25, 29 June 2025 (UTC)

Suggestion for additional clause to G13 criterion

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

The G13 criterion, at present, applies to articles in the draftspace that have not been edited for 6 months by a human.

I propose an additional clause that allows draft AfC's that have been rejected and do not fit the criteria for inclusion or creation, to be removed from the draftspace, irrespective of how long they have been left unedited.

Why? Well, because a draft for the chess move Nxe3 can be created, be rejected for not being noteworthy, and hog server space and resources for six months until someone can contest for speedy deletion — which usually goes under the radar.

This proposition would reduce usage of server storage, and would be critical in cleaning up the AfC categories. from Piperium (chit-chat, i did that) at 07:07, 29 June 2025 (UTC—edited)

Someone brings this up about every three weeks. The last time hasn't even been archived yet. See that, and the big honkin' "Search archives" box at the top of the page.

Also, deleting things increases, not reduces, data usage. (And even if it just dropped the data and all the associated logs and revisions and so on it wouldn't be enough to matter - we'd be talking maybe billionths of the total data usage per page deleted.) —Cryptic 07:21, 29 June 2025 (UTC)

I am actually so blind — literally the second discussion on the page 😭

And I actually was unaware of how the "Search archives" button functioned, so thank you! :) from Piperium (chit-chat, i did that) at 08:27, 29 June 2025 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Order of (obsolete) criteria

I reorganised the list of obsolete criteria to be in alphabetical order, for ease of reference - twice in the past ~week I've failed to easily find what I was looking for (I had to search) because it wasn't in alphabetical order despite appearing like it ought to be so. This was reverted by Tavix with an edit summary saying it should be "listed in the same order as the criteria above".

I can see the logic in the main criteria starting with general and ending with exceptional circumstances, but I don't really see much benefit to not having the rest in alphabetical order (articles would be in second place either way) although I'm not opposed to leaving them how they are if there is logic I'm not seeing. What I really don't understand though is what benefit comes from having a single list of obsolete criteria in what is little more than semi-random order? Thryduulf (talk) 20:37, 29 June 2025 (UTC)

The benefit is consistency. -- Tavix ^(talk) 20:42, 29 June 2025 (UTC)

Why does consistency bring benefits here? Thryduulf (talk) 20:45, 29 June 2025 (UTC)

If you don't want to answer that question, perhaps "What benefits does consistency bring here?" would be easier? Thryduulf (talk) 03:11, 1 July 2025 (UTC)

Support alphabetical. Alphabetical is much easier to browse and search. “Consistency” here is unclear as to with what, with the order of the level two headings, which I guess are conceptual, in order of importance? Consistency in general is a good reason, but is a weak good reason that should defer to any other good reason. SmokeyJoe (talk) 13:46, 3 July 2025 (UTC)

Minor alteration of G4 heading

Earlier, someone submitted an article I made for speedy under G4 after an RFD deleted a redirect with the same title. They agreed with me that it does not fall under G4 because it was not sufficiently identical. However, I think they were mislead in part by the header of G4. The header does read Recreation of a page that was deleted per a deletion discussion. Now, of course, they should have read the section closer; I don't think they would dispute that. But I think the header can and should be slightly more specific. Accordingly, I BOLDly changed it to Identical recreation of a page that was deleted per a deletion discussion. This was reverted because it does not need to be word-for-word identical to qualify for deletion under G4. That is a fair point, but I think the absence of the word also suggests something that is not 100% accurate: deletion simply because the page is a recreation. Of the two options, I think adding "Identical" makes more sense. "Sufficiently identical" is also an option, but that gets wordy, IMHO. lethargilistic (talk) 21:02, 8 July 2025 (UTC)

I think the current wording is fine. That speedy was downright wrong but it's wrong even without "identical"; a redirect is not in any sense the same page as an article, or vice versa. * Pppery * _{it has begun...} 21:05, 8 July 2025 (UTC)

I would support clarifying the G4 header in some way to get the word "identical" in there. "Sufficiently identical" doesn't seem too wordy to me.

More generally, I'm starting to be concerned that different parts of the community have startlingly different interpretations of how G4 should be used, and I think we need some more discussion about it. Does anyone else share that concern? Firefangledfeathers (talk / contribs) 22:29, 8 July 2025 (UTC)

I do not like the current G4 wording "sufficiently identical". To me, "identical" is like "unique": it is or it is not identical, but it is not something that can be partly true. Adding a comma would make it not identical. I guess "sufficiently similar" is too vague, but I would prefer "substantially similar" or something like that. —David Eppstein (talk) 22:59, 8 July 2025 (UTC)

If we go with "similar", I'd prefer a qualifier at least as strong as "substantially", but probably more like "overwhelmingly". If we stick with "identical", maybe the move is to get an "almost" in there. Would that address the "partly true" concern, or am I way off? Firefangledfeathers (talk / contribs) 23:03, 8 July 2025 (UTC)

It would at least be better than "sufficiently" to me. —David Eppstein (talk) 04:51, 9 July 2025 (UTC)

"Substantial recreation of a page deleted per a deletion discussion." ? Best Alexandermcnabb (talk) 05:06, 9 July 2025 (UTC)

I don't understand how someone can compare a current article to the deleted version in order to assess whether they're substantially the same. The deleted one isn't available for comparison. Is there an obvious answer that I'm missing? FactOrOpinion (talk) 00:58, 10 July 2025 (UTC)

Don't nominate for a G4 deletion unless you have a good idea, eg posted by the same person, or you remember the content of the article from before. You may also find an old copy of the article on archive.org. The deleting admin should check before deleting the page. Graeme Bartlett (talk) 01:57, 10 July 2025 (UTC)

@FactOrOpinion You can ask an admin to check if G4 applies, or you can check if the Internet Archive or similar sites have the deleted content. Toadspike [Talk] 20:09, 13 July 2025 (UTC)

When can G6 be used on an RFA?

Can an RFA be tagged as G6 if it's not obviously an error? I think it doesn't fall under any of the G6 criteria, but @Fortuna imperatrix mundi is sure that untranscluded RFAs are a valid use. The tagging that brought me here today was on Starfall2015's RFA, which I'm pretty sure does not fall under G6. SarekOfVulcan (talk) 16:37, 11 July 2025 (UTC)

Regardless of anything else, it clearly isn't an uncontroversial deletion since the candidate is edit warring to remove the speedy tag. So give them their wish. * Pppery * _{it has begun...} 16:41, 11 July 2025 (UTC)

If something does not clearly fall under any speedy deletion criterion it is not speedily deletable. Speedy deletion is only for "the most obvious cases" which means that if there is doubt about whether it meets a criterion is does not. Additionally, if someone (excluding the creator in some cases) expresses good-faith opposition to deletion then it would be controversial and so again it is not eligible for speedy deletion. Thryduulf (talk) 16:54, 11 July 2025 (UTC)

Note that per WP:CSDCONTEST, G6 is one of the few cases where the creator is allowed to remove the template themselves. --SarekOfVulcan (talk) 17:03, 11 July 2025 (UTC)

• Thank you for not notifying me of this discussion, SarekOfVulcan. Still smarting from WP:Miscellany for deletion/Wikipedia:Requests for adminship/Abhiraj010, eh? —Fortuna, imperatrix 17:12, 11 July 2025 (UTC)
• Incidentally, at the time I tagged, it was uncontroversial and untranscluded—the edit war etc was subsequent to it. —Fortuna, imperatrix 17:14, 11 July 2025 (UTC)
• G6 in particular is defined by the deletion being uncontroversial. If anyone is able to contest it, even after the fact, then it's not only not a good speedy, it was never a valid speedy to begin with.
  This doesn't mean you can't go ahead and delete it anyway if you're certain that'll improve the encyclopedia and that an MFD would be a net detriment. WP:Ignore all rules is policy too. But don't label it with a speedy code, and be ready to back up your reasoning. —Cryptic 17:15, 11 July 2025 (UTC)

Off-topic

(ec)Sorry, Fortuna, I assumed the ping above would be sufficient. And I voted delete on that discussion, and it was SNOW deleted, not speedied, so I'm not sure what your point is. --SarekOfVulcan (talk) 17:16, 11 July 2025 (UTC) Pings are clearly insufficient; Usder:Materialscientist has his completely turned off, so don't ever ping him to a discussion. My point is that G6 has been customary and practised for some time. If that changes, fine; but it is done so through consensus, not you sounding off on me on my talk. —Fortuna, imperatrix 17:19, 11 July 2025 (UTC) The thing is, it's not customary to use speedy for anything that could be legitimately objected to. If someone creates an RFA framework, doesn't fill it in, and disappears for two weeks, that might be G6-worthy. A legitimate-but-misguided attempt isn't. --SarekOfVulcan (talk) 17:28, 11 July 2025 (UTC) Separate from the topic at hand: Fortuna, this is a rather astonishing show of bad faith from a long-tenured editor. A ping is absolutely sufficient for just about anything short of Arbcom or ANI. Sarek made a good-faith attempt to notify you. If you're one of the few people who have chosen to turn off a feature that's been part of Wikipedia for over 12 years, that's your choice. Ed [talk] [OMT] 17:32, 11 July 2025 (UTC) Sorry. But. Then file a neutral post requesting clarification; not naming someone in the OP is a guaranteed path to good faith. I mean, hey―it might result in an unprejudiced discussion as to the pros and cons, rather than... this. —Fortuna, imperatrix 17:50, 11 July 2025 (UTC) FWIW admins are explicitly not required to have pings turned on (although it is regarded as best practice), but if they do disable them they are strongly encouraged to note this on their user page. I know you (Fortuna) are not an admin, but it still wouldn't harm to follow that advice. Thryduulf (talk) 19:41, 11 July 2025 (UTC) "not naming someone in the OP"—Sarek named you in the OP. Ed [talk] [OMT] 02:50, 12 July 2025 (UTC)

Ignore all rules is never a reason to speedy deletion something. Speedy deletion is explicitly only for things that uncontroversially improve the encyclopaedia. Speedy deletion of anything that does not meet one or more of the criteria is always controversial. Thryduulf (talk) 17:45, 11 July 2025 (UTC)

• IAR deletions of clearly misguided RfAs have had precedent for years now—see e.g. Wikipedia:Requests for adminship/Antny08 (deleted per G6), Wikipedia:Requests for adminship/Davelo15 (deleted per G7 despite 1 oppose vote), Wikipedia:Requests for adminship/J.Turner99 (deleted per IAR). Because of this, I do not fault Fortuna imperatrix mundi for attempting to apply a speedy deletion tag on this RfA. However, it is also true that these deletions are technically out-of-policy. Maybe a compromise solution could be to move these RfAs to userspace without leaving a redirect behind. Deleting/moving these RfAs has the benefit of keeping the RfA page a redlink for the user—in the event that these editors want to run RfA for real at some point in the future, they don't need to have their run colored by a clearly ill-advised attempt when the editors were still brand-new here. Mz7 (talk) 22:23, 11 July 2025 (UTC)

  And, indeed, Wikipedia:Requests for adminship/Starfall2015 (the RfA which catalyzed this thread) has now been IAR-deleted. In my view, this is a valid deletion under the WP:IAR policy, following the yearslong precedent I mentioned above. Mz7 (talk) 09:09, 15 July 2025 (UTC)