Jump to content

Wikipedia talk:Arbitration Committee/Noticeboard

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Wikipedia Arbitration
Open proceedings
Active sanctions
Arbitration Committee
Audit
Track related changes

Behaviour on this page: This page is for discussing announcements relating to the Arbitration Committee. Editors commenting here are required to act with appropriate decorum. While grievances, complaints, or criticism of arbitration decisions are frequently posted here, you are expected to present them without being rude or hostile. Comments that are uncivil may be removed without warning. Personal attacks against other users, including arbitrators or the clerks, will be met with sanctions.

Off-wiki misconduct in Palestine–Israel topic area

[edit]
Original announcement
  • I get the sense Ivana either refused to respond or they were unrepentant and doubled-down. Either tends to be a very good way to get a site-and-topic-ban combo. —Jéské Couriano v^_^v threads critiques 18:34, 9 December 2024 (UTC)[reply]
  • Huldra was very publicly accused of misconduct, is there any public statement on whether or not that evidence indicated any malfeasance on her part? And if it did not, is there any reason why such a public accusation is not met with just as public a refutation? nableezy - 18:47, 9 December 2024 (UTC)[reply]
    The Committee does not generally publish exonerations or dismissals of claims via motion. We did however let Huldra know that we closed this without action in regards to her, and that we appreciated her patience and responsiveness throughout the process. CaptainEek Edits Ho Cap'n!⚓ 18:51, 9 December 2024 (UTC)[reply]
    The committee also generally doesnt allow public accusations to be made based on private evidence. nableezy - 18:52, 9 December 2024 (UTC)[reply]
    Which is why we removed the accusations and made the issue private. CaptainEek Edits Ho Cap'n!⚓ 18:53, 9 December 2024 (UTC)[reply]
    It has been on my mind that perhaps the committee should make such statements. Maybe not in private cases but in public ones. For example, findings of fact that explicitly reject items submitted in the evidence phase that were off-topic or not at all compelling. Just Step Sideways from this world ..... today 19:03, 9 December 2024 (UTC)[reply]
    I recall suggesting something similar during my time on the committee (2015). However one of my fellow arbitrators (I don't remember who) was firmly of the opinion that every finding of fact should directly support at least one remedy. The same argument was made (I think by someone else) more recently when I suggested that things which were directly necessary for the situation being adjudicated to occur but which were themselves neutral (I don't remember the context for this, but I do remember using the analogy of a rail accident investigation into why a train ran away - that it was on a gradient was necessary for the runaway to occur, and so a finding of fact was made to this effect, even though it was not relevant to any remedy). Every remedy should be supported by at least one finding of fact, but I still disagree that the only purpose of findings of fact is to support remedies. Thryduulf (talk) 21:19, 9 December 2024 (UTC)[reply]
    I do kind of agree that the parts of a PD should lead into one another, but maybe it doesn't have to be part of the proposed decision. . A simple statement on the evidence talk page, for example, something along the lines of "we examined evidence related to <user> and did not find it compelling." Striking out evidence that doesn't show what is claimed could be considered as well. This is probably a bad test case as we don't actually know what the evidence was. Just Step Sideways from this world ..... today 22:36, 10 December 2024 (UTC)[reply]
    @CaptainEek, not before a lot of us read the accusations. Should there be consequences for parties making public accusations on the basis of private evidence when those accusations are not upheld? TarnishedPathtalk 23:03, 10 December 2024 (UTC)[reply]
    Speaking generally, as obviously I have not seen the private evidence, there are multiple reasons why evidence presented may not lead to action being taken, including:
    • The evidence shows the opposite to what is claimed
    • The evidence clearly doesn't show anything (of consequence)
      • This could be due to fabricated or misinterpreted evidence, or genuine evidence of things that are not problematic (e.g. claims of vandalism that turn out to be just removal of unsourced promotional material)
    • The evidence is completely unclear
    • The evidence shows evidence of something, but not conclusively enough to take action
      • e.g. there is clear evidence of bad actions by someone but it is not clear (enough) who that someone is
    • There are no actions that it is possible for Arbcom to take
      • e.g. the evidence points to the bad actions exclusively being done by someone not on en.wp
    • All the actions that Arbcom could take are moot
      • e.g. a person who would be sanctioned already has been (whether for these actions or some other).
    In several of the above, it's possible for evidence to have been presented in good or bad faith. What (if any) consequences there should be for the person submitting the evidence will depend on the circumstance. Fabricated evidence presented in bad faith is very different to genuine evidence presented in good faith that is simply insufficiently conclusive to take action. Thryduulf (talk) 23:29, 10 December 2024 (UTC)[reply]
    I think that presenting public accusations on private evidence should be prohibited and offenders should be sanctioned, whether or not the private evidence checks out. There is no excuse for it, and there wasn't an excuse this time, so I don't know why the offender was (afaik) allowed to get away with it. Zerotalk 11:16, 11 December 2024 (UTC)[reply]
    Technically it is prohibited, as it is casting aspersions without on-wiki evidence. My concern was that the whole laundry list of evidence-free accusations sat there for over three days, with at least two arbs commenting on it, before Primefac finally removed it. It's still there in the page history, as well. Black Kite (talk) 11:54, 11 December 2024 (UTC)[reply]
  • As I said in my response to the inquiry, I accept the judgement of the committee regarding me. I am however not pleased the committee seems to acquiesce to such clearly politically motivated acts of intimidation. Nothing new here I'm afraid. Tashmetu (talk) 18:54, 9 December 2024 (UTC)[reply]
    I will note in my opinion, the evidence for Tashmetu was weaker and their response made me unwilling to support the revocation, unlike with the other revocation. Sdrqaz (talk) 03:34, 10 December 2024 (UTC)[reply]
  • While I am limited in what I can say, this coming out now, in December, has been for me a prime example of the limitations (which I mainly attribute to capacity) of this year's committee. Despite how long it took this seems to be a thoughtful and considered response. I have high hopes for next year's committee and hope they live up to (or exceed them). Best, Barkeep49 (talk) 19:08, 9 December 2024 (UTC)[reply]
  • I think we can all thank the committee for taking a reasoned and considered action, announcing it, and remember that people can and do learn and change and get 2nd chances and even 3rd chances. Andre🚐 03:00, 10 December 2024 (UTC)[reply]

I have some questions about the 2 "For gaming the extended confirmed restriction"

  • Do they tell us something useful about the statute of limitations for gaming? In those cases, the gaming occurred many months ago. There seems to be a lack of clarity and diversity of views on this issue. The staleness question is relevant to a current AE case for example. Gaming is often spotted long after the EC grant is issued (or not at all probably). If there is something like a statute of limitations, it may make actively searching for accounts that look like they may have employed gaming worthwhile, at least for accounts that went on to edit in a contentious topic area.
  • Is gaming enough by itself or does there need to be another element of the "crime" to trigger removal of the privilege e.g. leapt into a contentious topic area post-EC grant, or was involved in edit warring, or off-wiki coordination etc.? Wikipedia provides several tools that people can use to pretty rapidly make 500 perfectly legitimate edits, so there seems to be a fuzzy boundary between ok and not-ok for the first 500 edits.

Sean.hoyland (talk) 04:02, 10 December 2024 (UTC)[reply]

There is no one sized fits all solution to ECP gaming I'm afraid. In this case of course, there was an off-wiki element which makes it perhaps a poor comparison to other cases. CaptainEek Edits Ho Cap'n!⚓ 04:19, 10 December 2024 (UTC)[reply]
I think if a) it can be shown that an editor employed gaming and b) the editor went on to make a significant proportion of the edits in a contentious topic area, then running a checkuser might reduce the ban evasion rate. I think this might be the case for a couple of reasons
  • The likelihood that a revision to an article was done by a ban evading actor is far higher in PIA than across Wikipedia in general. This is clear when you compare the PIA topic area to 500,000 randomly selected articles. For a PIA article revision, the chance that it's a sock edit is 5.9%, whereas for the random sample it's 2.9%, at least for the 2020 until now period. For the post-Oct 7, 2023 year, 2023-10-07 to 2024-10-06, it's 6.82% vs 1.96%.
  • EC requirements appear to concentrate ban evading actors in the subpopulation with EC rights. This subpopulation is relatively small compared to the general population, which could be a useful feature. Annual grants for EC are generally only in the 3500 to 4500 accounts range as far as I can tell, very substantially less than the total number of new accounts each year. And the chance that an EC account is blocked for ban evasion is high, in the 5 to 10% range, depending on the year. Furthermore, the speed of EC acquisition, the number of days from registration to extendedconfirmed, tells you something about the likelihood that an account will be blocked for ban evasion i.e. the quicker someone acquires EC, the more likely they will be blocked for ban evasion. You can see that relationship here for all newly acquired EC grants across Wikipedia from 2018 onwards.
So, maybe gaming followed by contentious topic area editing could be an indicator of an increased likelihood of ban evasion. Sean.hoyland (talk) 16:56, 10 December 2024 (UTC)[reply]
Sean.hoyland, where are you getting these numbers from? Ironic, I know, but could you cite your source? And. not to a gif but some source that explains how this data was gathered and evaluated. Thank you. Liz Read! Talk! 02:58, 11 December 2024 (UTC)[reply]
Liz, voices in my head mostly. Also, from SQL executed on the analytics server from my laptop using my toolforge account, sometimes directly, sometimes with some extra processing using Python. hmmm...I'll think about how provide some background. I have been looking at the topic area for a while now. Sean.hoyland (talk) 03:11, 11 December 2024 (UTC)[reply]
  • The first background thing is the answer to the question "what is the topic area?". We don't know precisely so we (BilledMammal, Zero0000 and I) have used approximations and gathered stats based on those. The approximations use article titles and are normally limited to namespaces (0,1), but not always. The article titles are selected by looking for ARBPIA templated talk pages and pages that are members of both the Israel and Palestine wikiprojects. This approximation has been slightly expanded recently to include pages that are part of the Israel-Palestine Collaboration project and pages that are in both the Israel and Lebanon wikiprojects and both the Israel and Syria wikiprojects. This could be expanded further of course to improve sampling, but the existing approximation seems to catch a large number of titles.
  • The second thing is the answer to the question "what is a ban evading actor, a sock?" In a perfect world (which is what external ML research projects appear to assume), you should be able to find ban evading actors by looking at the category graph because all actors blocked for sockpuppetry should be categorized as such. In reality, the category graph is incomplete. There are many accounts that have been blocked as socks that are not categorized as socks. This seems to be partly culture-related. Some editors seem to think that adding a sock template (which automatically categorizes the account) somehow gives credit to the actor. This is not helpful in my view, especially for ML project, partly because the category graph is the most efficient way to link accounts to sockmasters, which is very useful information. So, to handle this, actors are labeled as ban evading actors based on a combination of information from the category graph and from the block logs (looking for terms like checkuser, sock, multiple accounts etc. in the comments). Once you can label actors as ban evading actors you can easily count their revisions, both inside the topic area and outside, and distinguish them from accounts not blocked for sockpuppetry. You can also make timelines of their activities.
  • The third thing is how to count extendedconfirmed grants. It is easy to see who has been granted the EC privilege and when. So, it is easy to e.g. track this over time, measure the difference between registration and the grant issue, and the difference between grant and a block for ban evasion.
  • Another thing is identifying gaming, which is quite an interesting unresolved issue for me. This was discussed a bit here.
We have quite a lot of information about the topic area. Some examples follow. There are some plots in here for interest for Top and High important articles for the Israel and Palestine projects. Some information about account ages for various articles here, or over the whole topic area here. Unique editor counts over time. Ways to look at protection and talk page ARBPIA templating. Sean.hoyland (talk) 04:15, 11 December 2024 (UTC)[reply]
Forgot to mention the random sampling. It is easy to select n pages randomly and examine them. It only takes a few minutes to select, for example, 2 million pages and count the number of revisions by ban evading actors and normal editors, with an optional time range limit. I have code to do this in chunks to avoid getting hate mail from the cloud services people. For 2 million randomly selected articles, for the post-Oct 7, 2023 year, 2023-10-07 to 2024-10-06, 1.97% of the revisions are by ban evading actors blocked for sockpuppetry (211,546 revisions out of 10,732,361). Sean.hoyland (talk) 05:09, 11 December 2024 (UTC)[reply]
Also, in the meantime, lets enjoy the open canvassing on X (link) for "Israel Forever Foundation on Dec. 19 about "Wikipedia’s anti-Israel bias and how to fight back against it." "Learn the tricks of the trade with Aaron Bandler, Jewish Journal", (lets sign up) (My bolding), Huldra (talk) 22:05, 12 December 2024 (UTC)[reply]
"Learn the tools necessary to engage in online activism through much-needed Wikipedia editing to fight back against the hijacking of truth in the open learning forums".
No surprise that some editors are wary new editors coming into the topic space. TarnishedPathtalk 23:42, 12 December 2024 (UTC)[reply]
I am sure whoever the valiant Wikipedia neutrality protector who stalked pro-Palestinian activists and doxed them is going to be on this case too! Afterall they were clearly only motivated by a desire to ensure adherence to Wikipedia code of conduct and had no other motives! Tashmetu (talk) 07:31, 13 December 2024 (UTC)[reply]
It says right on The Israel Forever Foundation mission page "...enlightening, experiential, and apolitical learning and activism", so although I'm not quite sure I really know what the word experiential means, it all sounds like harmless fun. Just a bunch of enthusiastic people wanting to improve Wikipedia. Perhaps a source of fresh ideas like ronald, "The solution: Wikipedia should be blacklisted as a sponsor of terror and its staff should be imprisoned." It'll be fine. Sean.hoyland (talk) 10:41, 13 December 2024 (UTC)[reply]
I'm sure. TarnishedPathtalk 11:25, 13 December 2024 (UTC)[reply]
Pease WP:AGF; I'm sure User:BilledMammal will report it if anything even close to canvassing occur, Huldra (talk) 21:39, 13 December 2024 (UTC)[reply]
Huldra, I don't know if you are serious but I don't think you can rely on BilledMammal for regular reports. They have been on a WikiBreak (or retirement?) since November 14th. They haven't been involved in this case since it opened. Liz Read! Talk! 04:56, 14 December 2024 (UTC)[reply]
@Liz, there's a tad bit of sarcasm from Sean, Huldra and myself. TarnishedPathtalk 05:04, 14 December 2024 (UTC)[reply]
@Liz, I apologise for using this page for a bit of a banter (I think it is called: I am not a native English-speaker). In addition to this arb.com case, BilledMammal has made 14 AE reports only in 2024 (according to this). AFAIK, every single one of them against editors deemed not pro-Israeli enough. The chance of BM making a case against pro-Israeli canvassing is, IMO, approximately zero. That is something editors not following the IP-area knows nothing about, therefor our (Sean, TarnishedPath, Tashmetu and myself) comments here were at the wrong place: my apologies, Huldra (talk) 21:32, 14 December 2024 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Arbitration_Committee/Noticeboard&oldid=1264819537"