Misplaced Pages

This is an old revision of this page, as edited by Serial Number 54129 (talk | contribs) at 17:08, 2 April 2020 (→Comment on Bbb23 and symptoms of an overactive ArbCom: re). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Archives

Index 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 31, 32, 33, 34, 35, 36, 37, 38, 39, 40 41, 42, 43, 44, 45, 46, 47, 48, 49, 50 51, 52

This page has archives. Sections older than 10 days may be automatically archived by Lowercase sigmabot III.

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.

Misplaced Pages:Arbitration/Requests/Case/Motorsports closed

Original announcement

• Good solution IMHO, although it will be interesting to see how it turns out given that these to editors edit solely in the exact same areas... Cheers, --Puddleglum_{(How's my driving?)} 23:25, 30 March 2020 (UTC)

Workshop formatting suggestion

Shouldn't the "Analysis of evidence" come before the "Proposed final decision" sections? It seems to me that a natural progression is to first analyze the evidence, and only then propose a decision. WhatamIdoing (talk) 03:20, 31 March 2020 (UTC)

Psl631 unblocked

Original announcement

Comment on Bbb23 and symptoms of an overactive ArbCom

I posted this to CheckUser-l and ArbCom-en, but I think it worth posting publicly:

Just a general comment without being particularly aware of any specific incidents:

I’m sure there were likely checks that can be questioned, but that can be said of literally anyone who has the tool because the use of the tool is discretionary and discretion means that people are going to make errors in judgement calls. An error rate of 2-3% in any line of work requiring discretion is more than reasonable, but Misplaced Pages/Wikimedia’s system of accountability only looks at individual events.

Given the scale of his work, I’m sure whatever letter ArbCom sent could likely have enough individual events justifying it that would not reflect the scale of his good work. Given that he ran around 2,000 checks a month, even an error rate of 0.5%, which would be more than reasonable, would be 10 checks to point to a month when drafting something. There are of course individual events that might warrant this, but those are egregious ones involving using the tool to manipulate and control or intentional release of data, none of which I have seen here.

Having looked through his log, while he was more liberal than I may have been, but overwhelmingly he was right, and from a statistical standpoint likely was around the norm or better.

This is a symptom of an overactive ArbCom, probably in reaction to last one, but I am quickly losing confidence in the ability of this committee to be fair to anyone who is active in difficult areas, and have lost substantial respect for many of its members. TonyBallioni (talk) 13:54, 2 April 2020 (UTC)

One of the primary roles of the committee is to audit the use of checkuser and to provide appropriate corrective guidance where necessary. –xeno 14:10, 2 April 2020 (UTC)

That response assumes infallibility. Given the last 3 months, this ArbCom has proven itself to be the most unfair and overactive that we have seen in recent memory. We have a desysop based on blatantly false FoFs (BHG), an ArbCom case over an IBAN that should have been kicked to AN and resolved in 72 hours, and now this. Sure, it’s within your authority to do all these things. My point is that you all have rushed to get involved in matters faster than any ArbCom in recent memory, and have a remarkably high public error rate because of this haste. This was an unforced error. TonyBallioni (talk) 14:22, 2 April 2020 (UTC)

TonyBallioni, are you suggesting that ArbCom must be infallible in order to audit the use of checkuser? LEPRICAVARK (talk) 16:23, 2 April 2020 (UTC)

No. I’m saying that the response of “this is within our remit” when the criticism is that they’ve been rushing into things all year and making mistakes along the way is an argument from authority that assumes because they can do it that they get it right. TonyBallioni (talk) 16:27, 2 April 2020 (UTC)

Okay, I can see the reasoning behind that even if I don't agree. Personally, I've been very pleased with this committee's work and in light of the Framgate debacle from last year I don't think anyone should be surprised if ArbCom is more inclined to take initiative than they have been in the past. LEPRICAVARK (talk) 16:34, 2 April 2020 (UTC)

(edit conflict) Unfortunately the argument that the more work one does inevitably leads to a higher error-rate in doing so was found inadequate in RHaworth's arbcom, and I assume the same princile applies. Also—and I don't think I'm known as a cheerleader of this arbcom, probably far from it—but although we're only painting pictures here, if there were policy violations involved (IF—I'm not sayng their were) then they would specifically not be mitigated by volume of activity.I can well understand Bbb23's refusal to be micromanaged, if that was the case—it's an absolutely untenable position to put anyone, and absolutely impossible for anyone to accept. It would be odd, indeed, if anyone thought it would have another outcome than it did. ——SN54129 14:10, 2 April 2020 (UTC)

To summarize my arguments on CU-l:

Volume does not and should not negate intentional policy violations.

Enforcing established standards should be done even if it makes other admins/functionaries who also violate those standards uncomfortable and scared to continue violating those standards. Relevant for this case, and all the recent desysops.

Don't worry too much, he'll be back. The average enwiki retirement lasts about two weeks. But not before tens of users come out to lend emotional support, which becomes even more comical for the users on their 5th+ "retirement" who still get the same "oh the wiki will crumble without you" comments every time.

As to the specific case, I have enjoyed working with Bbb23 in the past, and hope he returns and continues with the 97% of his work that was excellent. -- Ajraddatz (talk) 14:50, 2 April 2020 (UTC)

The problem with that argument is there isn’t any evidence objective standards have been violated. Subjective ones possibly, sure, but very few things here are objective. The standard allow for discretion and are fairly vague. I could probably find checks in any individual CU log I wouldn’t have run, including arbs, stewards, and ombuds. That’s normal because there’s judgement involved and different people will have different judgement. The argument here is essentially that CUs are expected to make no errors, and that those overseeing the issue also make no errors. That’s not possible when dealing with matters of judgement. TonyBallioni (talk) 15:15, 2 April 2020 (UTC)

CheckUser does require judgment calls, but those calls still need to be made within the framework of the global and local policy. Nobody expects CUs to be perfect and error-free all the time, but problems arise when there is a trend of improper use of the tools, the user has previously been cautioned or made aware of what the problem is, and the behaviour continues regardless. Bonus points for sanctimonious dismissal of concerns and no self-reflection or attempt to understand what the problem is (Not necessarily saying that is the case here, but certainly has been with a lot of the recent desysops. And the same "judgement and vague policies" argument can be made with respect to those cases, in a similar irresponsible manner IMO). I think it is entirely acceptable for ArbCom to issue a warning to a user in that situation, if for no other reason than to reinforce what the standard is. As the local group responsible for oversight of functionary tools, that should be not only their right but their responsibility. -- Ajraddatz (talk) 15:30, 2 April 2020 (UTC)

Ajraddatz, there's no way to know for sure if the reason anyone returns at all is not precisely the kind of overwhelming emotional support they receive, is there? Usedtobecool ☎️ 15:54, 2 April 2020 (UTC)

No way to know for sure, but I expect it has far more to do with addiction to the website than the predictable outpouring of support every time an admin/long-term editor "retires". -- Ajraddatz (talk) 16:00, 2 April 2020 (UTC)

@TonyBallioni: just to be clear have checkusers been given more insight into what has happened here or are you speculating nearly as much (obviously you can see Bbb's checks in a way the rest of us can't so you have some more information) as the rest of us? Because from what Bbb has posted, I can speculate about two possible ideas behind ArbCom's actions. Option 1 is the error rate idea that seems to have been the focus here. If that's the case it appears like there are differing views among checkusers (sample size 2 in this thread but obviously also the checkusers of ArbCom deciding to warn but not revoke) about Bbb's performance. Alternatively, Option 2 is that there were certain categories (triggering events?) of checks that Bbb was performing that ArbCom wanted him to stop performing. I don't know which of these is more likely, but perhaps you do.Regardless, I don't blame Bbb for being really upset. Not in the least. But my hope is that in a few days or few weeks or even few months when he's in a better place that he can restart communication with ArbCom so he can find a way to not feel micromanaged and return to the valuable work he has done onwiki. Best, Barkeep49 (talk) 15:23, 2 April 2020 (UTC)

TonyBallioni, For those of us who don't follow every detail of ArbCom's daily actions, could you explain the context here? Was there a ArbCom case I missed? -- RoySmith (talk) 15:25, 2 April 2020 (UTC)

RoySmith, as I understand it, User_talk:Bbb23#ArbCom_and_me is the impetus here. creffett (talk) 15:28, 2 April 2020 (UTC)

I recognize that CU is a very privacy-sensitive area, but in the interest of transparency, would ArbCom be willing to comment generally on what CheckUser policies Bbb23 violated? creffett (talk) 15:27, 2 April 2020 (UTC)

Certainly...not  ;) ——SN54129 15:33, 2 April 2020 (UTC)

(edit conflict) @Barkeep49 and RoySmith: Roy, see User talk:Bbb23: Bbb23 has retired after a warning from the committee.Barkeep, we haven’t had specifics but have had discussions, but from what some arbs have said and what Bbb23 has publicly said, they have disagreements over interpreting the CU policy (particularly wikipedia:CheckUser#Grounds_for_checking.) That’s fine, but I also think if ArbCom were to propose a stricter interpretation of that in an RfC the community would overwhelmingly reject it, and so what we have amounts to ArbCom attempting to change policy by enforcement rather than its members trying to change the actual policy. I do not think the community as a whole supports reading the policy to effectively eliminate the fourth ground for checking, which places substantial discretion to the individual CU. It needs to be justifiable, but I don’t think Bbb23 was going around checking his enemies and I think he likely had a good idea why he was looking when he looked. TonyBallioni (talk) 15:37, 2 April 2020 (UTC)

@TonyBallioni: regardless of Bbb23 (right now), perhaps such a general RfC is warranted, vis a vis finding out/confirming exactly what the community wants/expects from CU discretion. A positive enforcement of the current wording—which I agree would be the most lkely outcome by a landslide—would presumably prevent this happening in the future. (Assuming, of course, that that's what's happened this time.) ——SN54129 15:57, 2 April 2020 (UTC)

What is this? Like I think the community would support giving CU discretion. I don't think the community would support checking on every first time RfA voter (as Misplaced Pages Online speculates happens but I've always understood doesn't happen). So using an RfC to rebuke this ArbCom based on facts that we don't know (and which it seems even CUs don't fully know seems unwise). Using the RfC, in a thoughtful manner to evaluate to what extent the community is and isn't comfortable with the use of checkusers using their powers, all within the confines of global policy, does seem like this could be helpful. So as with so much on wiki, the devil's in the details. Best, Barkeep49 (talk) 16:11, 2 April 2020 (UTC)

What's RfAs and rebukements to do with it?! ——SN54129 16:20, 2 April 2020 (UTC)

Serial Number 54129, RfA because I wanted to give an example of something that critics have speculated about and which I think if true (and I don't believe it is) the community wouldn't support. Rebuke because Tony has been critical about this decision and you used the phrase "prevent this happening in the future". Now maybe that's a misreading. That's why I started with the question of what the "this" that you used in your last two sentences would be. Best, Barkeep49 (talk) 16:26, 2 April 2020 (UTC)

Ah. Indeed. ——SN54129 16:43, 2 April 2020 (UTC)

This is a great loss & I hope Bbb23 returns. But many members of the "community", as I have read it, have been sensitive about some checkuser blocks which, by their nature, are often unexplained or are not supported by SPI reports which have been filed. The community has to trust (or at least accept) checkusers and their judgment to act within the existing guidelines. I don't think an RfC proposing a general extension of permission to checkusers to be more liberal in their checks of editor accounts would be supported by all. Liz 16:32, 2 April 2020 (UTC)

Speaking as one of the community members who share the sentiments you described, I certainly wouldn't be supportive of such an extension. When a CU says they have evidence of socking, the common folk are expected to believe them. ArbCom says they have evidence that Bbb23 violated policy. Well, I believe them. LEPRICAVARK (talk) 16:37, 2 April 2020 (UTC)

@Liz and Lepricavark: Unless I've missed something, no-one has mentioned an RfC to tighten the criteria, merely an RfC to re-confirm it. ——SN54129 17:07, 2 April 2020 (UTC)

I wouldn’t support expansion either. I also am pretty sure that an RfC to tighten the policy would be defeated in a landslide. This committee’s views on the policy I do not think are in line with the community view on it and sock puppetry. The issue isn’t only Bbb23. The issue is a committee doing via enforcement what individual members couldn’t get passed via RfC: namely tightening use of the tool beyond what is currently authorized by policy. There’s a fine line between interpreting policy and changing it, and I think they’re trying to cross that line. There’s a happy middle ground, and I think the existing policy is actually works quite well. TonyBallioni (talk) 16:48, 2 April 2020 (UTC)

• Who the checkusers are checking is always a black box to the rest of us, and has to be so. But my interpretation of Bbb23's statement is that ArbCom suddenly told him they didn't trust his judgement: "They have ordered me to limit my checking to 'non-discretionary' cases", and the "warning" was that they could take away his CU bit at any time. That's a serious attack on his integrity without possibility of rebuttal; I sympathise with his reaction and unless the committee—or the other CUs, who I know properly consult with each other frequently behind the scenes—had previously expressed concerns, I do think this was overly harsh. ArbCom is charged with disciplining and judging all of us, and has the power to destroy people's on-wiki reputations, but those reputations are the product of years of work here, and those with higher permissions have them because they are trusted and have the responsibility of exercising judgement in difficult situations. ArbCom telling a trusted editor they are unworthy of that level of trust is not a small thing, and their career can and should be mourned. Yngvadottir (talk) 16:42, 2 April 2020 (UTC)
• I also think there was a conversation to be had between the Committee and Bbb23 first, before escalating to an outright directive. El_C 16:49, 2 April 2020 (UTC)

  El C, this was part of an ongoing discussion - it didn't just come out of the blue. – bradv🍁 17:05, 2 April 2020 (UTC)

• According WP:AUDIT/STATS from January to December, Bbb23 performed 20,905 checks last year. Either he's CU'ing everything that moves, or then socking is a such a widespread issue that you can't really have proper grounds for checking and there's no way to audit that kind of amount. Of course this wouldn't be a problem if IP information was masked (T227733)... --Pudeo (talk) 17:00, 2 April 2020 (UTC)

• This current ArbCom seems determined to rid us of our best admins. So far they are succeeding admirably, and very few people seem to care or to realize what a ridiculous and insidious form of authoritarianism this is, that a mere handful of people can peremptorily drive the best, hardest working, and most trusted admins off the project with complete impunity. And apparently there's nothing we can do about it, and too few people that care or realize the danger to put a stop to it. Softlavender (talk) 17:08, 2 April 2020 (UTC)