Jump to content

Wikipedia:Deletion review/Log/2024 August 4: Difference between revisions

From Wikipedia, the free encyclopedia
Content deleted Content added
Line 58: Line 58:
****My original point is that controversies about transition isn't some kind of special case in our rules. Do you disagree? But further, why should issues of transition and family relationships have less coverage here than other topics? I think it's a very important topic and not one our coverage should discriminate against. Obviously you disagree, but I don't understand why. [[User:Hobit|Hobit]] ([[User talk:Hobit|talk]]) 02:56, 8 August 2024 (UTC)
****My original point is that controversies about transition isn't some kind of special case in our rules. Do you disagree? But further, why should issues of transition and family relationships have less coverage here than other topics? I think it's a very important topic and not one our coverage should discriminate against. Obviously you disagree, but I don't understand why. [[User:Hobit|Hobit]] ([[User talk:Hobit|talk]]) 02:56, 8 August 2024 (UTC)
*****Other P&Gs that play into my reasoning include [[WP:BLPGOSSIP]] and [[WP:10YT]]. I'd question by what standard you are asserting topical importance. Publicly codifying a dispute between family members reduces odds for reconciliation, which I'm sure everyone would hope for this situation. At a point we both remember, Wikipedia wouldn't publish Star Wars Kid's name out of a sense--Well, Jimbo's sense--that even if something was RS'ed, the human cost to the article subject outweighed the fact that it met notability guidelines and sourcing policies. [[User:Jclemens|Jclemens]] ([[User talk:Jclemens|talk]]) 04:12, 8 August 2024 (UTC)
*****Other P&Gs that play into my reasoning include [[WP:BLPGOSSIP]] and [[WP:10YT]]. I'd question by what standard you are asserting topical importance. Publicly codifying a dispute between family members reduces odds for reconciliation, which I'm sure everyone would hope for this situation. At a point we both remember, Wikipedia wouldn't publish Star Wars Kid's name out of a sense--Well, Jimbo's sense--that even if something was RS'ed, the human cost to the article subject outweighed the fact that it met notability guidelines and sourcing policies. [[User:Jclemens|Jclemens]] ([[User talk:Jclemens|talk]]) 04:12, 8 August 2024 (UTC)
****:The entire article could be 2 sentences: {{tq|Person changed her surname to avoid association with her dad. Person addressed something her dad said about her on twitter}}. If this was anybody else whose dad wasn't famous (and famously transphobic at that), no newspaper would give the slightest of shits and their would be no coverage of her. If by some miracle they did, and there was coverage, I'm guessing it would be a speedy delete per [[WP:NOT]]/[[WP:NOTDB]]/[[WP:NOTPROMO]]/[[WP:NOTWHOSWHO]]/[[WP:BLP1E]]/[[WP:LOWPROFILE]].
****:If the media cared enough to report on every trans woman who 1) changed their name and 2) had a poor relationship with their father, we'd have GNG sources for 90+% of trans women. The only reason she has any claim to notability whatsoever is she's Musk's daughter. Since she has never been notable outside of that, we shouldn't have an article.
****:This is not about special cases in rules, this is about the fairly obvious BLP issues with an article that is just {{tq|person changed their surname + person said her famous dad lied about her}}. If she was not trans, that would still be an awful deletion-worthy article because the entire premise would be "two trivia facts about this famous person's daughter and her relationship with him". [[User:Your Friendly Neighborhood Sociologist|Your Friendly Neighborhood Sociologist ⚧ Ⓐ]] ([[User talk:Your Friendly Neighborhood Sociologist|talk]]) 04:28, 8 August 2024 (UTC)
* '''Overturn''': I'm not impressed by the LLM argument either, but a bad argument doesn't erase the very good argument that there was not a consensus even by the most optimistic reading of the discussion for "redirect". 55-45 isn't a consensus and even that is cherry-picking when you consider that 55% is a bunch of different types of votes lumped into one category, and that the plurality option was "keep". AFDs aren't votes, but ignoring the vote should require some significant argumentation for doing so, which there wasn't. [[User:LokiTheLiar|Loki]] ([[User talk:LokiTheLiar|talk]]) 22:37, 4 August 2024 (UTC)
* '''Overturn''': I'm not impressed by the LLM argument either, but a bad argument doesn't erase the very good argument that there was not a consensus even by the most optimistic reading of the discussion for "redirect". 55-45 isn't a consensus and even that is cherry-picking when you consider that 55% is a bunch of different types of votes lumped into one category, and that the plurality option was "keep". AFDs aren't votes, but ignoring the vote should require some significant argumentation for doing so, which there wasn't. [[User:LokiTheLiar|Loki]] ([[User talk:LokiTheLiar|talk]]) 22:37, 4 August 2024 (UTC)



Revision as of 04:28, 8 August 2024

Vivian Jenna Wilson (talk|edit|history|logs|links|watch) (XfD|restore)

I must object that consensus was met on Wikipedia:Articles_for_deletion/Vivian_Jenna_Wilson#Discussion.

First the raw tallies. They were:

Keep: 16
Do not redirect: 1
Redirect: 12
Redirect and merge: 2
Delete and redirect: 1
Ideally delete, if not, redirect: 2
Delete: 4

Grouping categories, we get to:

Keep: 16
Do not redirect: 1
Redirect: 15
Delete: 6

If we further merge categories down to just two:

Keep: 17
Redirect: 21

A 45%:55% split is not consensus. Consensus is not simple majority.

Of course, Wikipedia's consensus process is not simply about numbers, but argument. Individuals can vary on their judgement of an argument's strength, of course. But overall, factoring this in this does not change the picture. Removing "trivial" support / oppose votes without much argument, I get that "Keep" loses one trivial vote and "Redirect" loses 2 trivial. Wikipedia policies (in base voting arguments, I didn't check in subthreads) were mentioned 14 times by the keep side and 18 times by the oppose side. Contrarily, the keep side cited 22 external sources while the oppose side cited only one.

To get further opinions on whether consensus was established, I fed the discussion to the most powerful free LLMs (after verifying that they know Wikipedia's policies well, which they do - unsurprising, given how Wikipedia is prominently used in their training datasets). Under the discussion, I asked each, "Under Wikipedia's policies, does it appear that consensus is established?".

Claude: "Based on my analysis of the debate, there does not appear to be a clear consensus established yet. The opinions are quite divided between keeping the article, deleting it, or redirecting it. Here's a summary of the key points..."
Gemini: "... Analysis of Consensus: While there is a strong case made for keeping the article based on WP:GNG, the counterarguments about the subject's primary notability being derived from their relationship with Elon Musk and the potential application of WP:BLP1E are also significant. It appears that consensus is not clearly established. The debate is closely divided, with strong arguments presented on both sides. ..."
ChatGPT: Seems to be down at the moment
LLaMA 3.1 405B: "No, it does not appear that consensus is established. There are a significant number of editors arguing both for keeping the article and for deleting or redirecting it. The discussion is ongoing, and there is no clear majority or overwhelming agreement on a specific course of action. In Wikipedia, consensus is typically determined by the weight of argument and the quality of reasoning, rather than a simple headcount. However, in this case, the discussion is complex, and there are valid arguments on both sides...."

I believe User:RL0919 has erred, and that the closure should be undone. Consensus has not been met.

Starting a deletion review because the user archived the deletion discussion and it explicitly says not to comment on the deletion there, but to instead comment here. Rei (talk) 20:01, 4 August 2024 (UTC)[reply]

  • Closing admin comment: Large language models do not have opinions, and I will not defer to their output. I also don't put much stock in vote-counting that is carefully massaged to avoid noticing that over 60% of the participants did not want the article to be kept. --RL0919 (talk) 20:33, 4 August 2024 (UTC)[reply]
    • You're free to dismiss any arguments you choose to. The fact is that there is nothing even close to a consensus in that page. And even "glorified Markov chains" (**they're not) can see that. -- Rei (talk) 21:22, 4 August 2024 (UTC)[reply]
    • Yeah. Admin work is time-consuming and disproportionately thankless. You don't have to agree with the decision, but saying it's worse than the output of these glorified Markov chains is an unnecessary insult. —Cryptic 21:04, 4 August 2024 (UTC)[reply]
  • Endorse Any time a closing admin thinks consensus favors no article for a BLP-related reason--in this case controversy about transition and family relationships--then that's a really high bar for us to overcome, doubly so when the numerical total favored, no matter how narrowly, not maintaining a separate article. Jclemens (talk) 20:37, 4 August 2024 (UTC)[reply]
    • The article was not controversial due to "transition and family relationships", it was purely controversial over noteworthiness. Not a single person objected to it over its subject matter, and the subject in question refers to herself as a "Professional H-list celebrity" and is taking interviews, so clearly has no objection to being a public figure. Lastly, if a single admin's word is effectively final, what is even the purpose of deletion review? -- Rei (talk) 21:22, 4 August 2024 (UTC)[reply]
      • If it were objectively unreasonable, we would tell that admin they were wrong and why. I'd like to think that we tend to have less admins getting overturned here because the collective participants here give diverse feedback that represents multiple points of view, so admins who get overruled here tend to take our input under advisement and not make similar closes in the future. Jclemens (talk) 21:29, 4 August 2024 (UTC)[reply]
        • Given that nobody was objecting to the article for anything having to do with a BLP reason, only over noteworthiness, would you care you revise your original post? We should hold this discussion only on the facts of the (almost perfectly evenly split) debate about noteworthiness in the RFD. Thanks  :) -- Rei (talk) 21:45, 4 August 2024 (UTC)[reply]
          • Nope. Just because the discussion didn't explicitly take something into account doesn't mean I can't, in DRV, come up with a separate reason why I believe the close was correct. You don't have to be convinced by my opinion, nor do I have to strike it if you're not. Jclemens (talk) 16:29, 5 August 2024 (UTC)[reply]
        • ED: Well, BLP1e, but that's about noteworthiness. One person mentioned BLPNAME once, but only in regards to her siblings, not her. -- Rei (talk) 21:47, 4 August 2024 (UTC)[reply]
      "H-list celebrity" is pretty clearly a joking self-description of someone who does not consider herself a public figure. Funcrunch (talk) 00:01, 5 August 2024 (UTC)[reply]
      It's also not something who someone is seeking to stay out of the limelight is putting in their profile. Nobody can plausibly read her posts or any of the interviews she's done and possibly think that she is seeking to remain private.
      And I will say that we're straying quite far from the purpose of a deletion review here.. -- Rei (talk) 00:40, 5 August 2024 (UTC)[reply]
    • @Jclemens: As I've mentioned below, I'm confused about what you are claiming is the policy or guideline-based justification for the deletion. I know BLP, but I'm struggling with how this doesn't meet BLP. Is this more of an IAR-endorse on your side, or is that mischaracterizing your !vote? Hobit (talk) 14:04, 7 August 2024 (UTC)[reply]
      • What part of controversy about transition and family relationships was unclear? I'll note that Where is Kate? had far more coverage, worldwide, and yet was deleted and redirected on BLP grounds. Just because the media covers something extensively enough doesn't mean we should have an article on a topic. Based on the totality of my experience, I'd say this is a less suitable topic for an article than that. Jclemens (talk) 19:14, 7 August 2024 (UTC)[reply]
        • My original point is that controversies about transition isn't some kind of special case in our rules. Do you disagree? But further, why should issues of transition and family relationships have less coverage here than other topics? I think it's a very important topic and not one our coverage should discriminate against. Obviously you disagree, but I don't understand why. Hobit (talk) 02:56, 8 August 2024 (UTC)[reply]
          • Other P&Gs that play into my reasoning include WP:BLPGOSSIP and WP:10YT. I'd question by what standard you are asserting topical importance. Publicly codifying a dispute between family members reduces odds for reconciliation, which I'm sure everyone would hope for this situation. At a point we both remember, Wikipedia wouldn't publish Star Wars Kid's name out of a sense--Well, Jimbo's sense--that even if something was RS'ed, the human cost to the article subject outweighed the fact that it met notability guidelines and sourcing policies. Jclemens (talk) 04:12, 8 August 2024 (UTC)[reply]
          The entire article could be 2 sentences: Person changed her surname to avoid association with her dad. Person addressed something her dad said about her on twitter. If this was anybody else whose dad wasn't famous (and famously transphobic at that), no newspaper would give the slightest of shits and their would be no coverage of her. If by some miracle they did, and there was coverage, I'm guessing it would be a speedy delete per WP:NOT/WP:NOTDB/WP:NOTPROMO/WP:NOTWHOSWHO/WP:BLP1E/WP:LOWPROFILE.
          If the media cared enough to report on every trans woman who 1) changed their name and 2) had a poor relationship with their father, we'd have GNG sources for 90+% of trans women. The only reason she has any claim to notability whatsoever is she's Musk's daughter. Since she has never been notable outside of that, we shouldn't have an article.
          This is not about special cases in rules, this is about the fairly obvious BLP issues with an article that is just person changed their surname + person said her famous dad lied about her. If she was not trans, that would still be an awful deletion-worthy article because the entire premise would be "two trivia facts about this famous person's daughter and her relationship with him". Your Friendly Neighborhood Sociologist ⚧ Ⓐ (talk) 04:28, 8 August 2024 (UTC)[reply]
  • Overturn: I'm not impressed by the LLM argument either, but a bad argument doesn't erase the very good argument that there was not a consensus even by the most optimistic reading of the discussion for "redirect". 55-45 isn't a consensus and even that is cherry-picking when you consider that 55% is a bunch of different types of votes lumped into one category, and that the plurality option was "keep". AFDs aren't votes, but ignoring the vote should require some significant argumentation for doing so, which there wasn't. Loki (talk) 22:37, 4 August 2024 (UTC)[reply]
  • Endorse the close of Redirect:
      • After considerable tweaking and diddling with the !votes, the appellant still shows that a majority favored some action other than keeping the article. The closer was justified in concluding that there was a rough consensus for redirection.
      • The large language models are correct that there was no obvious consensus from counting the votes. Therefore the closer was expected to consider strength of arguments and determine what was the rough consensus.
      • An alternative approach to the large language models would be to send their output to a bit bucket.
      • DRV is not AFD round 2. However, having read the redirected article, it is my opinion that the paragraph in the redirected article is enough coverage, and is better than keeping the article.
      • I have not asked my cat for an opinion. Robert McClenon (talk) 00:47, 5 August 2024 (UTC)[reply]
  • Endorse. While it's an interesting exercise to see how our robot overlordsnon-human colleagues might close a deletion discussion, the human who did close the discussion explained his reasoning perfectly well: those wanting a standalone article were not able to articulate why the contents merited a standalone BLP rather than coverage in an existing article, and WP:N is clear that there is no obligation to have an article about someone who is in news coverage for a narrow reason (quoting from the close.) And as the closer points out above, even if you ignore "strength-of-arguments" and go with "strength-of-numbers," most participants in the discussion (myself included) felt a standalone article was not justified. Bottom line, was the closer's judgment reasonable? Yeah, I think it was. Certainly it was not so unreasonable that overturning it would be justified. (Jclemens also makes a good point above.) 28bytes (talk) 00:49, 5 August 2024 (UTC)[reply]
  • Endorse while I deliberately opted not to close that discussion because everything Musk related is contentious, I'd have probably closed it the same way. The majority of participants did not believe that Wilson is independently notable and therefore did not require a distinct page. Star Mississippi 01:10, 5 August 2024 (UTC)[reply]
  • Comment My only remark was with the closure statement that stated that opinion was divided between Delete and Redirect when clearly it was between Keep and Redirect. There were plenty of participants arguing to Keep this article and I don't recall seeing any justification for dismissing them. Still it was a close call and I don't like to second guess other closers over close calls. But I can also see why this closure was contested. But I don't think a Redirect closure is unreasonable. Like Star Mississippi, I skipped closing this discussion because I guessed it would end up here no matter what the closure was. And I don't the introduction of how LLMs would have closed this discussion helps the the argument for why this closure should be overturned or relisted. Liz Read! Talk! 03:14, 5 August 2024 (UTC)[reply]
  • Weak endorse: There was rough consensus for not keeping the article and a "no consensus" close would not have been very helpful in this case. The LLM is right in that "there was no clear consensus", but controversial topics like this will never have "clear" consensus or be easy closes. Some of the delete/redirect votes were nothing more than WP:NOTNOTABLE in my opinion, but that doesn't allow the closer to Supervote the other way instead when it has the majority. Maybe it should've been left alone for a bit or relisted to attract further opinion, but that's not a requirement and there was no error in the close. C F A 💬 05:50, 5 August 2024 (UTC)[reply]
  • Endorse - reasonable close, even if I would have preferred something else. --SarekOfVulcan (talk) 13:52, 5 August 2024 (UTC)[reply]
  • Endorse - not enough sources to back undeletion of the article. MSM should wait for Vivian's side of her background. Ahri Boy (talk) 22:11, 5 August 2024 (UTC)[reply]
  • Endorse. Closure was a reasonable conclusion. I don't give a stuff what LLMs say. Stifle (talk) 07:53, 6 August 2024 (UTC)[reply]
  • Overturn I don't see how this is close. The arguments for redirecting and/or deleting are basically "Doesn't meet GNG" (which I find to be false and no attempt was made to explain why the sources don't count). "Only known because of who her dad is" (which at best is a misunderstanding of WP:INHERIT). And "The topic is better covered by a merged article" (which is a fair argument, though I disagree with it). The keep arguments are largely "Meets GNG" and "Meets BLP inclusion guidelines". GNG is clearly met and subject likely doesn't qualify as a low-profile individual (as we define it) and it's not clear what the "event" is, so BLP1E restrictions are conquered. Given the strength of arguments keep is a much more accurate outcome than redirect. NC is probably the right outcome however. Hobit (talk) 15:29, 6 August 2024 (UTC)[reply]
  • Endorse per Jclemens. WP:BLP absolutely was a consideration in where and how in depth to cover a dispute between two living people, so the closing admin was correct to evaluate whether there was consensus to keep the article. I haven't seen LLMs used to evaluate consensus like this before, but I'd like to recommend against it in the future. hinnk (talk) 19:10, 6 August 2024 (UTC)[reply]
    • Maybe I'm a bit too rules-based, but I've no idea where we document that coverage of a dispute between two living people is handled differently than normal BLP issues. Is the GNG met? Is there a BLP issue? We have sources that go over the GNG bar by a fair bit and I'm not seeing anything in WP:BLP that isn't met. Do you disagree? WP:IAR is a fine thing, and I'm a big fan. But it's a weak argument that needs serious numbers. I'm seriously not seeing a guideline or policy-based argument for a redirect. Could you cite what you think are the guidelines or policies that support the redirection? Hobit (talk) 23:49, 6 August 2024 (UTC)[reply]
      I'm going to avoid derailing the deletion review, but you're welcome to look at the original discussion for my thoughts on whether WP:GNG is met and how it relates to the decision to keep or redirect. The number of WP:BLP arguments presented (on either side), as well as the closing admin's comments, indicates to me that the standard Jclemens describes is appropriate. hinnk (talk) 00:16, 7 August 2024 (UTC)[reply]
      I'm stuck with why?. I've read the discussion. I'm seeing claims that because they are related she shouldn't have an article. That's not what WP:INHERT says. I'm seeing claims the GNG isn't met, but that's laughable given the provided sourcing. So what guideline or policy gets us to a redirect? I'm sorry to be picking on you, but you're the one that !voted after me and I thought I laid out a pretty clear argument that there is no specific policy or guideline that supports the redirect/delete side. So those arguments are weak and there is no way a closer should get to "redirect" from here. If no one can explain what text of the GNG or BLP (or BLP1e) that the article violates, we really shouldn't get to a redirect outcome without numbers that support invoking WP:IAR. And we aren't close there. Help a guy out, what am I missing? Hobit (talk) 01:05, 7 August 2024 (UTC)[reply]
      I don't think you missed anything. I think you disagreed with it and are trying to relitigate it in the replies of a comment addressing a different topic. I'd like to suggest more firmly that you consider why that's not constructive. hinnk (talk) 02:00, 7 August 2024 (UTC)[reply]
      I'm not sure I'm relitigating. I'm asking for exactly what policy or guideline you feel the discussion concluded wasn't met. You're endorsing the outcome but, AFAICT, not explaining how the outcome aligns with our rules. I don't think you've done so and frankly I'm fine if you are arguing IAR or something (which is how I would characterize Jclemens's argument (which you cited)). Hobit (talk) 14:01, 7 August 2024 (UTC)[reply]
  • Endorse The closure seems like a reasonable reading of consensus based on the strength of arguments, and also roughly aligns with the numbers. The BLP/NOPAGE/NPF arguments were not adequately refuted by the Keeps, who were primarily focused on WP:GNG which is not in dispute. Since this is a BLP-related discussion, we need to be extra careful to get it right and err on the side of not including material. While the burden of proof isn't quite reversed at AFD like it is with other BLP discussions (though there was a time when no consensus, default to delete was a valid outcome for BLPs), I think the level of consensus reached was more than adequate for this result. The WordsmithTalk to me 17:47, 7 August 2024 (UTC)[reply]