This case is now closed and pages relating to it may no longer be watched To request an amendment or clarification of an arbitration decision, see Wikipedia:Arbitration/Requests/Clarification and Amendment. To report a violation of an arbitration decision, see Wikipedia:Arbitration/Requests/Enforcement. To request the assistance of an arbitration clerk, see Wikipedia talk:Arbitration Committee/Clerks.

Wikipedia Arbitration
About arbitration Talk Arbitration policy (CU/OS) Guide to arbitration Noticeboard +  Talk Search archived proceedings Ban appeals Report
Open proceedings All open requests Case requests Clarification and Amendment requests Arbitrator motion requests Enforcement requests
Active sanctions Contentious topics ( Log) General sanctions Editor sanctions
Arbitration Committee Discuss + Contact Members Procedures Elections Talk History Clerks +  Talk Clerk procedures
Audit Talk Statistics Talk Reports (pre-2016)
Track related changes
v t e

Behaviour on this page: Arbitration case pages exist to assist the Arbitration Committee in arriving at a fair, well-informed decision. You are required to act with appropriate decorum during this case. While grievances must often be aired during a case, you are expected to air them without being rude or hostile, and to respond calmly to allegations against you. Accusations of misbehaviour posted in this case must be proven with clear evidence (and otherwise not made at all). Editors who conduct themselves inappropriately during a case may be sanctioned by an arbitrator, clerk, or functionary, without further warning, by being banned from further participation in the case, or being blocked altogether. Personal attacks against other users, including arbitrators or the clerks, will be met with sanctions. Behavior during a case may also be considered by the committee in arriving at a final decision.

To update this listing, edit this template and scroll down until you find the right list of arbitrators. If updates to this listing do not immediately show, try purging the cache.

Active:

1. Aoidh ( talk · contribs)
2. Barkeep49 ( talk · contribs)
3. Cabayi ( talk · contribs)
4. CaptainEek ( talk · contribs)
5. Firefly ( talk · contribs)
6. Guerillero ( talk · contribs)
7. HJ Mitchell ( talk · contribs)
8. Maxim ( talk · contribs)
9. Moneytrees ( talk · contribs)
10. Sdrqaz ( talk · contribs)
11. Z1720 ( talk · contribs)

Inactive:

1. L235 ( talk · contribs)

Recused:

1. Primefac ( talk · contribs)
2. ToBeFree ( talk · contribs)

Given the nature of this case, I think some guidance on the evidence page as to what may be posted publicly, and what should instead be submitted privately (with instructions on how to do that) on the evidence page would be very helpful, and hopefully avoid the evidence page itself becoming the source of yet more contention and messes. Seraphimblade ^{Talk to me} 20:08, 6 March 2024 (UTC) reply

Thanks – while the case navigation header already mentions an e-mail address, I think this could be made more clear and detailed. ~ ToBeFree ( talk) 20:15, 6 March 2024 (UTC) reply

Probably also worth noting that if you are unsure whether something is public or private evidence, send it by email and ask. The Committee should advise whether it can/should be posted publicly. Thryduulf ( talk) 20:27, 6 March 2024 (UTC) reply

I agree with Thryduulf - and indeed someone had sent us such a piece of evidence during the case request which has now been replied to with "this needs to be done publicly". Beyond that Seraphim, I welcome more thoughts about what guidance could be given, beyond noting the email address, which I've also added as a bullet point to the evidence page. Barkeep49 ( talk) 20:46, 6 March 2024 (UTC) reply

I think Thryduulf's "If in doubt, ask privately first" suggestion is excellent. Maybe also a pointer to and a brief synopsis of the outing policy? Seraphimblade ^{Talk to me} 21:01, 6 March 2024 (UTC) reply

I think they deserve to be heard, and they raise issues that arbs should consider before making a final decision, even though I see a lot I disagree with... but it's not evidence. It should belong somewhere else, workshop or one of the talk pages. —  Usedtobecool  ☎️ 04:27, 9 March 2024 (UTC) reply

@ Dennis Brown Regarding the question about whether owning AT&T stock is a COI; like so many things, there's no black and white answer. I'm assuming that the amount of AT&T you own is an inconsequential fraction of the shares outstanding. And that it's a small portion of your total portfolio. And nothing you wrote on AT&T had any measurable effect on its stock price. And you don't know anything about the company which isn't public knowledge. The less true these assumptions are in aggregate, the more COI becomes an issue. RoySmith (talk) 13:46, 9 March 2024 (UTC) reply

You don't know how much AT&t stock I own. AT&T stock pays a dividend quarterly. It is not the only stock I own that pays dividends. You do not know how much I earn from these stocks but I will say that it exceeds what an average paid editor would earn for going in and changing a few articles. And unless you've investigated my edits, you don't know if anything I wrote could have affected the stock price or dividends of any of these stocks. I have not disclosed what stocks I own, other than this one example, nor how many shares I own. This is what I mean when I say most everyone has a conflict of interest of some kind, but does that make it paid editing? Because if it is, then every administrator needs to make sure not to edit any article remotely related to a stock, or a competitor to a stock in their 401k, IRA, brokerage accounts, among other things, as this would be inconsistent with adminship, and grounds to remove the bit. And should admin disclose what stocks they own? Or how many shares? After all, we're supposed to disclose any conflict of interest, right? If we overreach and define paid editing as any conflict of interest where there is potential financial benefit, you will render policy unenforceable. And before you dismiss this as an absurd example, you need to look at it more closely because it is the natural outcome of that decision, and in time it will come back to bite us. Please note, I have not said anything about whether Nihonjoe's actions are wrong or right, as I don't have access to the private information. My concern is how we frame it. Farmer Brown - 2¢ (alt: Dennis Brown) 22:43, 9 March 2024 (UTC) reply

The issue really starts with Joe Roe's evidence, which appears to be an extended exercise in trying to construe WP:PAID to say something that is contradicted by both the plain meaning of its text and common sense. -- JBL ( talk) 20:02, 9 March 2024 (UTC) reply

“Users who are compensated for any publicity efforts related to the subject of their Wikipedia contributions are deemed to be paid editors, regardless of whether they were compensated specifically to edit Wikipedia” is a direct quote from WP:COI Jessintime ( talk) 20:21, 9 March 2024 (UTC) reply

Yes and? -- JBL ( talk) 20:31, 9 March 2024 (UTC) reply

Sure, but this does not yet directly mean "Users who edit articles about their employer are paid editors". Ymblanter ( talk) 20:32, 9 March 2024 (UTC) reply

Just noting that the committee/drafters are discussing this evidence, Joe Roe's evidence, and the first section of Fram's evidence. Barkeep49 ( talk) 22:56, 11 March 2024 (UTC) reply

Fram - you are over your standard 1,000 words for your evidence section. You will need to request an extension if you wish to post any additional evidence. firefly ( t · c ) 15:29, 12 March 2024 (UTC) reply

Thanks, if I need to post more I'll ask for an extension first. Fram ( talk) 15:32, 12 March 2024 (UTC) reply

Noted - thanks! firefly ( t · c ) 15:56, 12 March 2024 (UTC) reply

Fram is extended an additional 200 words to post part of a private submission. Barkeep49 ( talk) 21:57, 13 March 2024 (UTC) reply

Thanks, done. Fram ( talk) 08:17, 14 March 2024 (UTC) reply

I think this question has directly or indirectly come up during this case. Here seems to be an OK place to bring it up again, given the submissions from Tryptofish and JoelleJay. I've written an essay, User:Maxim/The dark abyss, to describe what happens. It's more a brain dump than an essay, but I hope it makes that part of ArbCom a bit less mysterious. Maxim ( talk) 13:34, 14 March 2024 (UTC) reply

Thanks for linking that, Maxim, it's very informative and allays some of my concerns over how private evidence is handled. It's good to know that (if I'm understanding correctly) there's some sort of actively-triaged action list visible to all arbs, and that it's normal for complex COI allegations to take a while. If it ends up being in the scope of this case, I think it would be helpful to float ideas on how progress in an investigation might be transmitted to involved parties without much additional hassle for arbs. Something like a deidentified phabricator ticket (or even more simply like a DoorDash delivery timeline) that just automatically shows where in the pipeline an investigation is. JoelleJay ( talk) 14:12, 14 March 2024 (UTC) reply

I've thought for several years now that the capacity of any given arbitrator is not fixed. They will have more or less capacity given not only the things you'd expect but their interest in given topics. However, I would argue the capacity among arbcom as a whole for administrative tasks is pretty fixed and an increase in administrative tasks (as thoughtfully proposed here) leaves less capacity for other work which in this instance would be handling the concerns themselves. Barkeep49 ( talk) 15:41, 14 March 2024 (UTC) reply

I understand it would be more work for arbs to keep track of notifying interested parties without some easily-portable automated system doing this. My thought was mostly along the lines of some template email, like the acknowledgement of receipt one, being sent out if and when an investigation reached some particular stage or endpoint. Nothing individualized. Perhaps even limited to cases where arbs can be reasonably sure a nontrivial number of editors have seen some particular private evidence (like certain off-wiki content, or when the committee receives multiple independent reports on the same evidence) and thus the likelihood of editors agitating for a response and potentially posting outing material on-wiki is high.

I just know I was left wondering whether my very brief email to the committee was even sufficient to trigger anything at all, or if I should have provided a curated list of the major allegations and diffs in the email body rather than linking to a webpage for y'all to dig through yourselves. JoelleJay ( talk) 16:47, 14 March 2024 (UTC) reply

That I can answer: providing a brief and well curated explanation is always going to get faster action than a more generic link (for more on that see User:Barkeep49/ArbCom_Guide#How_to_submit_effective_evidence). Speaking for myself there are certain things I stopped doing because I didn't want to do the administrative work that was to accompany them. For instance I stopped acknowledging appeals because updating the table of appeals felt like too much of a pain. Now using a templated message is easier than updating a table with multiple embedded templates so in that sense you're right. But as Thryduulf points out what the landmarks where an update is sent out isn't always clear and things can go backwards not just forwards (one arbs completes their look into it, discussion on those findings start, another arb does other investigation and presents different findings, thus potentially stopping discussion into yet more arbs can look into it) and this is on top of the administrative burden concerns I have. Barkeep49 ( talk) 17:01, 14 March 2024 (UTC) reply

Unless things have changed significantly since my time on the committee, there isn't really a usefully granular pipeline between acknowledging the email and taking final action/determining no action is required as there are many variables that determine what action(s) by the committee are needed and even more variables that determine how long each will take. I recall one allegation we received that required a lot of work to investigate, and then the results of that investigation were not clear cut so the committee was roughly evenly divided on whether the allegation was sufficiently proven or not so dealing with it took several weeks. During that time we received a similar allegation related to a different editor which was resolved in (iirc) a few days start to finish, despite requiring roughly the same number of steps, because everything was much clearer. Thryduulf ( talk) 16:08, 14 March 2024 (UTC) reply

Thanks for the essay. I've had the impression for a very long time that it has been difficult for ArbCom to get to where there is a Committee consensus on how to respond to an email request, as the essay describes. And that stands to reason, because that's the nature of a committee, especially a volunteer one. If ArbCom wants to kick some ideas around, internally, about ways to improve the process, I think it would be beneficial to both ArbCom members and to the community if that particular kind of logjam could be loosened up. One way to do that might be to have a process where there would not always be a need to have an official quorum and vote to designate a response as more than "in an individual capacity". Perhaps one member could draft a response, circulate it among the Committee before sending it, and if there are no objections after a defined amount of time, the reply could be sent, as something more than an individual response but also something less than the official response from the Committee as a whole. (Defining, in your procedures, the amount of time before the reply will be sent will also have the beneficial side effect of motivating quiet members to stop procrastinating.) I think that the more efficiently emails get processed, and the smaller the backlog, the easier the whole thing would become. You'd probably need to work out some details that I've overlooked, to flesh out such a procedure. -- Tryptofish ( talk) 18:46, 14 March 2024 (UTC) reply

Perhaps one member could draft a response, circulate it among the Committee before sending it, and if there are no objections after a defined amount of time, the reply could be sent, as something more than an individual response but also something less than the official response from the Committee as a whole. this is essentially how things work for correspondences already. Barkeep49 ( talk) 18:49, 14 March 2024 (UTC) reply

I allude to this in the essay, but finding that one member to draft a response is sometimes tricky. Maxim ( talk) 19:00, 14 March 2024 (UTC) reply

Hmm, I didn't anticipate Barkeep49's reply. My impression from Maxim's essay and from things I've heard in the past goes more along the lines of where Maxim wrote: Unless an issue can be reasonably handled by an individual arbitrator acting in an "individual capacity",... it requires multiple arbitrators to agree on an action... But, depending on what the matter is about, it may be difficult to get to the "multiple arbitrators to agree" step, not as much because there are too many competing ideas, but because none are acceptable (or suggested). What I suggested is to (at least in some circumstances) replace "multiple arbitrators to agree" with "no arbitrators objected after a reasonable amount of time". That seems different to me.

Yes, I can see how it can be hard to get the one member to draft the response. I suppose that sometimes that's because it's just not clear what the right course of action would be, and everyone feels a need to discuss it before making up one's mind. I suppose there's no way around that. But I can also picture situations where one member might think "well, it's pretty clear to me that we ought to do xyz", but there is concern that "xyz" would not have "multiple arbitrators to agree" on it. That's where I think my suggestion might be a starting point, especially if there aren't really "too many competing ideas". -- Tryptofish ( talk) 19:19, 14 March 2024 (UTC) reply

Trypto: as someone who is often the one drafting it, if there's not affirmative response to the message (but there's not opposition) I will often do a "My plan is to send this at X time unless someone objects". But you're 100% correct that this could be systematized. Barkeep49 ( talk) 20:07, 14 March 2024 (UTC) reply

I agree. -- Tryptofish ( talk) 20:17, 14 March 2024 (UTC) reply

I've hacked away at my draft, but I can't get it below 650 words without just throwing down bare diffs. May I please request an extension in advance? Yngvadottir ( talk) 11:49, 16 March 2024 (UTC) reply

Thanks for asking. No problem going to 700 words. Barkeep49 ( talk) 02:15, 17 March 2024 (UTC) reply

@ Yngvadottir re:your last bullet point, I don't think bad faith actors should get to dictate things we do. However, if good faith actors want to assume responsibility for the content that can be a good thing. We saw this play out on the evidence page with content getting reverted by a banned editor and reinstated by SN with SN ultimately taking "ownership" of the content. For me that's the right model in these situations. Barkeep49 ( talk) 16:10, 17 March 2024 (UTC) reply

@ Barkeep49: The removal is a standard admin action following determination that the editor is "a bad faith actor" and blocking, yes. However, taking responsibility for any edit removed by an admin in such circumstances carries the risk that the edit itself may have been malicious; in this case, there may have been OUTING concerns. Doubly so when the removal is by a member of ArbCom during ArbCom's consideration of a case concerning OUTING. Hence I thought it better to ask Firefly to consider self-reverting the particular edit, even though it was not revision-deleted. Note that I subsequently asked for an extension at the case request page, but there was no response. The removed question was about a publisher that so far as I can determine using Wayback and "What links here", has never had an article or a draft, and if another editor has evidence of COI edits by Nihonjoe with respect to it, they will presumably submit the evidence privately, so I won't reinstate the question. I will however make a clarification and addition in my evidence section once I've drafted and counted the words. Yngvadottir ( talk) 21:23, 17 March 2024 (UTC) reply

@ El C: I'll reply here to avoid clogging up the evidence page since this reply isn't evidence. Your request is illogical. ArbCom and the 15 humans that it comprises do not exist in a vacuum; the nature of our community is such that highly active members will cross paths in multiple places wearing different hats. Nihonjoe made me an admin almost 14 years ago; Primefac frequently beats me to the punch on oversight tickets; Fram and I have had differing opinions on policy and philosophy issues; and I've crossed paths with Kashmiri in various places in project space. You and I have worked together at places like AE before. It's inevitable that arbs will have prior knowledge of many of the parties to the case and probably desirable that they are aware of trends within the community.

Specifically to my comment at ANI, I don't think it's reasonable to read that as a defence of Nihonjoe; in fact I expressed no opinions at all. I merely pointed out that this is an area where policy and community expectations have changed over the years and it would be unfair to judge an action from even a few years ago by today's standards. I also enquired about more-recent actions because much of what was initially presented was a few years old, which can be indicative either of a problem that is dead and buried or of a current problem that has been going on for a long time. I was one of several people who encouraged Joe to post a list of articles where he had a potential conflict so that those articles could be evaluated by the community. HJ Mitchell | Penny for your thoughts? 16:19, 19 March 2024 (UTC) reply

I will just note that we have policy about how recusals are to be requested and I posted that info on El C's talk page after I first saw his evidence. Barkeep49 ( talk) 16:44, 19 March 2024 (UTC) reply

If I thought there were grounds for recusal, I wouldn't make my colleagues drag me kicking and screaming. But since it's come up, I feel it's worth explaining my thoughts, even if this is the wrong place. HJ Mitchell | Penny for your thoughts? 16:53, 19 March 2024 (UTC) reply

For me it's the fact that in this case we've had suggestions of requests for recusal at both the request for arbitration stage and now the evidence stage. Neither of those is the right venue ot make an initial request; I think the policy was intentionally written in a way to make such requests potentially "less drama" by having them be on a user talk page rather than a case page. If that's not working for the community any more it can certainly be changed, but I don't think we should IAR the existence of the policy away. Barkeep49 ( talk) 16:58, 19 March 2024 (UTC) reply

I merely pointed out that this is an area where policy and community expectations have changed over the years and it would be unfair to judge an action from even a few years ago by today's standards.

Not exactly. Here's what you wrote:

Diff #1 is eight years old. The rest (which are mostly formatting) are five years old. The community's attitude to COI/paid editing has changed significantly in that time (largely as a result of professional marketing firms using significant resources to covertly influence our content). Is there anything more recent or that was explicitly against the rules as they stood at the time the edits were made?

First, you were incorrect about the community's attitude to COI/paid editing having changed significantly from 5-8 years ago. This was a disruptive bit of misinformation because for a while thereafter, people just believed this was true, I think merely because you (and some other editors) said it. I remember this because I had to correct this misinformation in my preliminary statement, where I linked to WP:COI as of 12/30/2015 to show that disclosure was already mandatory. I'm a little concerned that as of this date, you don't seem to understand that you were incorrect about the COI guideline changing over this time period?

Thus, and secondly, you were also incorrect about the edits not being explicitly against the rules as they stood at the time the edits were made. They explicitly were. My preliminary statement was a direct rebuttal to your comment from a week earlier, which was necessary because two of the initial arb responses when I made my preliminary statement expressly repeated the "wasn't explicitly against the rules at the time" line of argument.

Third, you absolutely expressed opinions: that policy and community expectations have changed over the years and that the edits were not explicitly against the rules as they stood at the time the edits were made.

Fourth, you absolutely defended Joe. Your comment was a defense of Joe: specifically, you said the evidence was stale, and that the edits didn't violate the rules at the time they were made. Not only did you defend him, but you defended him with incorrect assertions.

This was pre-judgment, and it was incorrect pre-judgment. Whether Joe's edits were against the rules at the time they were made is a core issue in this case. You already made up your mind about that issue prior to seeing evidence, and even now, still seem to hold the same opinion, which to me suggests unconscious bias, of the kind that may call for recusal. Levivich ( talk) 16:51, 19 March 2024 (UTC) reply

This is less about the specifics of the last paragraph (which since I'm not being asked per policy to consider I've not done so and thus have not prejudged the merits or not of this recusal request) and more a general musing. I've long wondered when I'm "allowed" to have an opinion about stuff as an arb. What opinion am I allowed to have at the case request for instance? Am I allowed to be of the opinion that the evidence already present is sufficient to merit a sanction even if I'm not proposing a motion to do that sanction? Or do I need to wait to have an opinion? What role does the fact that I'm willing to change my mind play? For instance in this case I have changed my mind, thanks to closer examination of evidence and the analysis/explanation given to me about a couple of core elements of the case. Am I allowed to say what those change of opinions are now or should I be waiting to share?

This all makes me sound less sympathetic than I am to the underlying concerns. There really is a danger in someone prejudging evidence or forming an opinion and being unwilling to change their minds. Those would both be bad to do as an arb. But I also think there is danger in having an arb form a private opinion that they feel coerced into not sharing and thus there's no real chance for them to be persuaded otherwise until it really is too late and I think the line is a squiggly evershifting one rather than something clear. Now all that said, I tend to avoid participating at AN/ANI precisely because any number of things could end up before ArbCom and I want the community to believe (because it's true) that the case is getting a fair hearing. Barkeep49 ( talk) 17:12, 19 March 2024 (UTC) reply

My twofold 2c: First, I've long seen that arbs will generally avoid commenting on anything having to do with admin misconduct of any kind (involved, adminacct, etc.) specifically to avoid having to recuse if the admin issue escalates to arbcom (which is the only place that they escalate to). I think this is good practice, I've seen it from multiple arbcoms over the past few years, and I'm rather surprised that in this instance, though not the only instance, several arbs have not followed that practice. Here we see the results. ($0.02)

Second, arbs, like all other persons in any kind of judge-like position--hearing officers, arbitrators (in the real world), actual court judges, administrative judges, basically everybody in any kind of position like this--should not judge the evidence or the case, nor should they express any opinions about the evidence or the case (except as is necessary to make a preliminary decision), until all the evidence is in. This is a totally essential requirement and one that is followed, AFAIK, in literally every "real" dispute resolution system in the world, public or private, government or otherwise. As an evaluator of evidence, one must keep an open mind until one can evaluate all the evidence.

The only exception to this is a preliminary decision, e.g. whether a case can go forward, which is analogous to the real world concept of probable cause. Like real-world adjudicators, Wikipedia arbs should only decide the narrow issue of "probable cause" ("should a case be opened") at the preliminary step, without deciding (in their mind) whether anybody should be sanctioned.

The only exception to that is if the evidence is so overwhelming that no case is required at all, e.g. "handle by motion."

So, basically, at the outset, an arb should choose one of three paths:

1. "Below probable cause": Deciding that the preliminary request doesn't show enough evidence to merit action by arbcom. This is saying "we will not look into this further."
2. "Meets probable cause": Deciding that the preliminary request has enough evidence to open a case and conduct a full investigation. The arb should not opine at all about what the outcome of that investigation will be. This is saying "we will look into this further."
3. "Overwhelming": Deciding that the preliminary request is, for whatever reason, so "overwhelming" that no case is needed at all, and it can be done by motion. This is saying "no more evidence will come in."

For #2 situations, it's paramount that arbs keep an open mind, don't pre-judge the outcome, don't express opinions beyond "will look into it," "won't look into it," "don't need to look into and can act now." And if the arb find, "in their heart of hearts," that they can't not pre-judge the outcome, they should recuse. Even if they don't have an actual conflict of interest or prior involvement, they should recuse just because they know they aren't capable of keeping an open mind about this particular case. ($0.04) Levivich ( talk) 17:27, 19 March 2024 (UTC) reply

There are situations where matters are more complicated than that. For example:

1. There are multiple parties, evidence is overwhelming against some but only showing cause to investigate for others
2. There are multiple accusations against a single party, the strength of evidence for them varies significantly such that it's clear sanction will be needed regarding at least one but not necessarily all.
3. It is clear that some sort of sanction is needed but not clear what/how significant it needs to be.
4. The evidence presented appears to overwhelmingly show one thing but evidence presented subsequently changes that.

Any or all of the above can of course apply in combination too. Thryduulf ( talk) 17:38, 19 March 2024 (UTC) reply

Not complicated.

1. Arbs should make the "probable cause" determination as to each party separately.
2. That's exactly the thing that arbs should not prejudge. If they aren't handling any of it by motion, they need to not prejudge the strength of any of it, except to determine that it's strong enough to open a case (e.g., the "probably cause" determination)
3. That's what arbs should not prejudge
4. That's why arbs should not prejudge

There is really no "but.." argument for why an adjudicator should not judge the evidence before all the evidence is submitted. You won't find any support for that notion anywhere in the real world. Look at the ethics rules for, and I mean this literally, any dispute resolution system, and "don't prejudge the outcome" is like Fairness 101. Levivich ( talk) 17:44, 19 March 2024 (UTC) reply

hearing officers, arbitrators (in the real world), actual court judges, administrative judges tend to be adversarial systems, at least in places based on English legal traditions which is what I'm guessing both of us are most familiar with, and ArbCom is not designed as adversarial in the same way. I appreciate you putting forward the case in a thoughtful manner for "Arbs should never signal their opinions until it's time to rule". I will give it more thought. Barkeep49 ( talk) 19:04, 19 March 2024 (UTC) reply

No, not just in adversarial systems such as common law (England, US), it's the same in inquisitorial systems such as civil law (France, Italy), and in other dispute resolution systems, like private arbitration, because they are universal concepts of fairness. Levivich ( talk) 15:15, 20 March 2024 (UTC) reply

France is the non-English derived legal system I'm most familiar with, and prosecuting judges are not neutral in the way you describe. Among other things this is why they are not permitted to serve as a judge in the future with the same defendant. I am sure we could collectively create a long list of the ways ArbCom differs from legal systems - in a longer reply I initially wrote and then trimmed down to the response above, I noted the fundamental role that lawyers play in legal systems of all sorts and which are not present on Wikipedia. The fact that we are both investigating judges and a jury is unusual as well and probably not for the better.

Returning to the larger question at hand, I think what your proposing would in our current system with its many imperfections, at least for me, lead me to make worse decisions. By sharing my thinking, I give everyone, including those who might be sanctioned, a chance to respond and change my mind. And I do change my mind. But that doesn't happen in a vacuum and often if I said nothing, the evidence/analysis that changed my mind wouldn't have been produced. In fact I can think of more than one occasion where someone felt blindsided when the proposed decision was produced and a remedy they weren't expecting was being seriously considered; this is especially acute for me in cases where the sanction relies in major part on conduct during the case itself. I dislike when that happens and this is why I've moved away from pretty much the model you describe to the more ambivalent place I am at today. The moral certitude that you write with is causing me to think carefully and re-examine those practices but I wish you would consider the ways that what you propose would violate other elements of "universal concepts of fairness" as you've defined them. Barkeep49 ( talk) 15:47, 20 March 2024 (UTC) reply

France's ethics rules for judges ( English PDF, see p. 10), Chapter II: Impartiality, No. 5:

5. Judges may not publicly express a belief through their comments or behaviour until the decision has been given.

When I say every dispute resolution system in the world has this rule, this is an example of what I'm referring to.

In terms of the alternative approach, I think you give too much credit to an arb's ability to change their mind, and too little credit to the power of various biases like unconscious bias and confirmation bias.

Also, once a person publicly expresses an opinion (especially in writing), it's going to be harder, not easier, to change that opinion.

Further, once an arb expresses an opinion, it can have a chilling effect on further participation in the arbitration process. What's the point of putting in evidence if you know the judges have already decided?

Finally, a premature expression of opinion by an arb can seriously skew the evidence, as editors end up responding to arbs initial "hot takes" rather than just presenting what they view as relevant evidence. (As we saw here, my preliminary statement is an example of that.)

Arbs shouldn't be expressing an opinion until they've had a chance to review all the evidence. (For the same reasons, nobody else should be proposing remedies on the Workshop phase until the evidence phase closes.) Editors shouldn't worry about trying to change arbs' opinions until after that point, as well. Otherwise, arbs end up thumb-scaling, and arb cases end up being about changing arbs' initial expressed opinions, rather than just "following the evidence wherever it may lead." Levivich ( talk) 17:07, 20 March 2024 (UTC) reply

agreed Buffs ( talk) 22:18, 20 March 2024 (UTC) reply

Thanks for the diff. I'm glad I made as much sense as I thought I did. I don't see anything there as prejudging anything whatsoever. I guess you could say I judged the age of the evidence, but that's why I asked if there was more—it's rare that what someone did five years ago is actionable now, but the old stuff can be used to demonstrate an ongoing problem. And the community's attitude to COI has absolutely changed, even if that's not reflected in changes to what's written in policy. Orangemoody and before that Bell Pottinger were landmark moments in COI management, and there have been other smaller cases before and since. It's much more of a hot-button issues than it was even a few years ago and of course in the very early days of Wikipedia it wasn't considered an issue at all. Of course, one aspect of this case is that we expect admins to keep up with evolving policies and expectations. HJ Mitchell | Penny for your thoughts? 17:42, 19 March 2024 (UTC) reply

Is it relevant whether or not the community's attitude has changed? If not, why bring it up? If so, are you submitting evidence that the community's attitude has changed? Or are you going to conclude that the community's attitude has changed based solely on your own personal experience (and even when that is contradicted by the fact that the guideline has not changed in material respects since the time at issue)? Your own personal experiences are what creates your biases, and you're just repeating your biases here and asserting them as fact, and this is bad because the truth of this "fact" weighs directly on whether or not Nihonjoe did something wrong 5-8 years ago (and since). Levivich ( talk) 17:49, 19 March 2024 (UTC) reply

Is it relevant whether or not the community's attitude has changed? yes. Behaviour can only be fairly judged by the standards of the time, so if (I've not looked at the evidence) Nihonjoe did something wrong 5-8 years ago then it matters how severe the community regarded infractions of that nature at that time - the wording of the policy has not materially changed but the attitude towards breaches of it most certainly has.

Separately, someone doing something wrong 5-8 years ago is irrelevant in isolation. It is only relevant when placed into context with recent behaviour, and ideally with behaviour in the intervening period. Thryduulf ( talk) 20:01, 19 March 2024 (UTC) reply

@ HJ Mitchell: Both Orangemoody and Bell Pottinger concern paid editing: a ring of paid promoters and a publicity firm. Aren't you conflating that with COI in general? The intersection point has been an issue of disagreement in this case, both at AN and in evidence. There are other forms of COI than monetary, and I can think of other forms of monetary COI than being paid to edit as defined in the TOU. Yngvadottir ( talk) 21:02, 19 March 2024 (UTC) reply

I take your point. I mention them just as seminal moments that arguably contributed to a hardening of attitudes towards COI in general. There is, of course, a spectrum of COIs. HJ Mitchell | Penny for your thoughts? 21:07, 19 March 2024 (UTC) reply

This is just my impression, not empirical fact, but I think that circa the Jytdog case, there was a pretty strong community feeling that preventing outing was much more important than uncovering COI – whereas in the present, the community has become less tolerant of COI than it was then, with a less clear view about how COI reporting and the avoidance of outing should be balanced. -- Tryptofish ( talk) 21:17, 19 March 2024 (UTC) reply

HJ Mitchell, sorry but I can't spare the time for an in-depth follow up right now. If it was up to me, it'd be in a week or two, so that I could do that, but time was running out (literally the last day). Still briefly: I am of the opinion that you, all three of you, conflicted yourselves—yes, for you and ToBeFree, even at that early stage (solely). Because those statements and what they convey (which for your part, Levivich addressed above in some detail) are from the same person: admin/commentator -and- arb are the same person. And I think the bar there should be purposefully high there, for a number of reasons and a number of potential problems it preempts. Sorry again that I'm unable to engage on this further presently, truly, but it was now-or-never for me. Thanks. El_C 02:22, 20 March 2024 (UTC) reply

( edit conflict) Putting this as a response to Trypto, but it's also further to HJ Mitchell. I would say community attitudes on paid editing and business promotion have hardened over time, and that's bled over into attitudes toward COI in general, but that one reason is something that rears its head often in considerations of both bias in general and civility norms: those who work in offices consider office work normal and find it very hard to take into account that other editors are numerically unlikely to also be office workers, and quite likely not to be employed by a for-profit business, or at all. (The same assumption of normalcy of course also goes for generational experiences including still being in full-time study, but we're more aware of that kind of myopia). I see David Fuchs has alluded in recently submitted evidence to the WMF's own imperfect definition of paid editing and linked to the then-ArbCom's dissenting statement from 2017. I would say myself that there are many intersecting forms of COI (the unnameable site in question has recently been publishing some uninformed statements about academic COI on its forum) but a continuum of financial COIs, not all of which are paid editing. (This relates to Dennis Brown's examples.)

David Fuchs has also mentioned that the Fram case involved OUTING concerns; I had been intending to note that here. If we're doing ArbCom case history, OUTING vs. ADMINCOND also arose in the Gamaliel case. (I'm on record as disapproving of ArbCom actions and inaction in both those cases.) But ArbCom decisions do not establish binding case law precedents. In fact ArbCom isn't a law court at all. Nor should it be setting policy, although one of its functions is to solve conflicts between our policies and guidelines. The community has discussed admins and paid editing outside ArbCom, in at least one binding RfC. The changing attitudes toward editing on behalf of businesses have largely been expressed and encoded in the PAG outside of ArbCom cases, which is as it should be.

Two things fundamental to how Wikipedia works are that (a) editors are expected to do the right thing (hence we extend good faith that they will; and hence we discuss at RfA and weight ArbCom candidates' statements and answers to establish whether we trust the candidates); and (b) editors check other editors' contributions for errors (whether typos, misunderstandings, forgetfulness, or reflections of either conscious or unconscious bias). We have to trust ArbCom to take notice of e-mails rather than black hole them, but this is an extension of admin accountability itself (which in turn is an extension of Communication is Required). I personally don't think it makes much sense to require requests for recusal to be on the Arb's talk page; Arbs are flooded with pings of various sorts even more than are normal active admins. I'm sure Firefly didn't notice my inquiry on their talk page, and that the clerks didn't notice my late request for an extension at the case request made after Firefly's response didn't answer my question. Better have the requests for recusal be on the evidence page, which the Arbs are reading (I assume!) Also, Levivich's point about ethics is well taken—I'm much less informed on such matters and a woolly holistic thinker, which is high among the many reasons I decline to run for ArbCom. We should be able to trust ArbCom members to know not only when they can't be dispassionate but when there will be an appearance that they are not disinterested, exactly as we should be able to trust admins and bureaucrats to know when they have a COI (part a). But we also need to be able to address it when they appear to be blind to problems or barking up the wrong tree (part b): such as the previous statements highlighted by El_C (wisdom of the crowd, again; I had forgotten ToBeFree is also an arb), whether expectations are higher for those with advanced permissions (ADMINACCT), and also the relative importance of "case law". Yngvadottir ( talk) 02:56, 20 March 2024 (UTC) reply

• What was said at the original AN thread matters because it was the first (and only) attempt at dispute resolution. The fact that, of the very few but very influential editors who downplayed the issues even as much of the evidence wasn't in and/or tried to shut that discussion down, three of them were arbs out of only fifteen that we have, seems pertinent irrespective of whether it was, especially since the concerns being raised were about a similarly high-ranking member of the community. Primefac is recused already; that's good. ToBeFree, asked at the original AN thread, (paraphrasing)"COI isn't even a policy, what do you want us to do?" It was disappointing, and not at all what I was expecting to hear from a member of the arbcom, but it was materially not incorrect. COI isn't policy. I am of the opinion that whether COI is a policy, or what it said eight years ago is an irrelevant distraction. The fundamental question, because it concerns an admin and a bureaucrat, not a run-of-the-mill spammer, is what the hell are we here for and what are our values? Are we here to build an enyclopedia for the benefit of humanity and should reflect in our actions the best interests of that goal, or are we here to promote our own personal and professional interests while remaining within the confines of written policy so no one can stop us? It wasn't less shocking to hear ToBeFree's position echoed by more arbs. But ok, that was not directly to do with the actual case and its subject. Arbs, I am happy to assume, are open to changing their minds; perhaps the fact that they're hearing this case at all is evidence to that. Therefore, ToBeFree need not recuse. Firefly's actions were run-of-the-mill and would not have been a big deal in a less messy case. So, Firefly need not recuse either. I am giving arbcom some leeway here under the assumption that they're still settling down. I wasn't happy either, that I was told not to post further in the case request without asking for an extension, and when I did, no one responded, even though it was clear, arbs and clerks were reading all the posts that were coming in. Maybe all cases are like this. But, it was weird being ignored, being someone who's been critical of arbcom and its members' handling of the very case. HJ Mitchell ought to recuse. They were one of the few editors who participated in the original AN thread early on, trying to downplay the issues and shut off discussion. Their arguments even now reek of involvement. Involvement cares about how it appears, and it gives zero weight to the intentions. And yes, it does look like HJ is submitting evidence. Usedtobecool  ☎️ 04:00, 20 March 2024 (UTC) reply

• What seems to get lost here is that COI isn't a singular thing. Some will argue otherwise, but I can only chock that up to a lack of real life experience. You are oversimplifying it. As I've said dozens of times (and many have argued against), everyone has some kind of COI. The world is not a binary place. We all have a vested interest in something. The phrase "COI" has been conflated with "financial gain" at Wikipedia, and so we dilute, or mutate, the actual meaning of the phrase into something entirely new. Everyone has temptations to push their agenda (financial or other), and admins must rise above the temptation, or be wise enough to avoid those topics. "Paid editing" is incompatible with adminship. We all agree on these two points, which are not the same thing.

COI is part of a larger issue of neutrality. We have a tendency to redefine words at Wikipedia. We modify words and narrow their meaning here at enwp, words like "bludgeon", "wikilawyer", "disruption", "neutral" and "COI" for internal use, but those words still have meaning outside of these pages. I Googled "Conflict of interest" and the first definition was "A conflict of interest occurs when an individual's personal interests – family, friendships, financial, or social factors – could compromise his or her judgment, decisions, or actions in the workplace.". Financial is only one aspect. As our definitions changes over time, these sweeping generalization will come back to bite us. "Drawing a line in the sand" with COI is not workable. Even saying "okay then, if you have financial COI, disclose it" won't work, because no one is going to declare their stock portfolio and others will ignore the rule because it is unenforceable. You end up with less disclosure, not more. If you edit where you work/own, declare it, of course, but making nebulous rules outside of this won't work. Abusive COI is kind of like pornography, it is hard to define, but I know it when I see it, and unfortunately, that is the best we have to offer. Arb's job is to decide if they can see it, in each case. Their job isn't to define it or regulate it. Dennis Brown - 2¢ 06:59, 20 March 2024 (UTC) reply

This confuses conflict of interest with bias or worldview. Or looked at another way, it misrenders the meaning of interest in conflict of interest, with a notion of interest attributed to individual choice (a base sort of neoclassical self-maximising individual). I might be a devout Rastafarian with Jamaican parents and an Ethiopian partner working as a doctor in a hospital in Leeds (just so there's no mistake, I'm not) - it doesn't mean I have a conflict of interest when I edit articles about Haile Selassie. The point is I am a volunteer and I am working freely; the source of my contribution is my free labour. That changes dramatically when the source of labour is not free, but purchased labour - whether direct or indirect - or connected to subsequent financial gain (the proverbial stock portfolo... I don't have one BTW... but yes, if one owns shares in a company, then one does have a potential conflict of interest in editing about that company). I'm less concerned, for example, with the consequences of a conflict of interest in the case of editors being related to a subject (or subjects editing about themselves) in that it will most often simply reflect a particular worldview, not qualitatively different than that of an editor obessessed with a particularly famous actress or singer of the moment. The production of knowledge changes fundamentally once we shift from a free basis to a paid basis (or one related to indirect financial gain) because the social relations that exist for production under paid labour or free labour are fundamentally different. Think of the extent to which the limitation of rights automatically ensue upon entry into employment (limitations to free speech, limitations to free association). I don't expect that we can perfectly police everything, but there's very good reason why there is such strong concern around the effects of editing outside of a voluntary basis. Part of the reason the problem has been exacerabated in this particular case *appears* to be a generalised discounting of the effects of conflict of interest - treating it as something with too much grey; akin to jay-walking, everyone does it in some way, so let it slide. Maintaining an expectation of a high degree transparency with regard to COIs has to be our starting point - if an editor does not feel they can be transparent they should probably not edit. Regards, Goldsztajn ( talk) 12:27, 20 March 2024 (UTC) reply

I have yet to opine on the merits of this case at all, which isn't really public anyway, and I doubt anything we say here will change the fate of Nihonjoe. My fear is that we DON'T treat COI as the grey area it is, by trying to claim it is black and white, when it isn't. And most financial COI is more subtle than "my boss said to edit this article", and I would argue, they are typically handled properly. This case appears to be one in which the COI was not handled properly, but I don't want to see Arb start defining things and creating policy, which is not their role. Dennis Brown - 2¢ 12:42, 20 March 2024 (UTC) reply

The committee interprets policy and can set precedent ... I'm happy for it to do its job. To the extent that I'm a member of the peanut gallery, I do not wish to see a loosening of standards with regard to COI. Regards, Goldsztajn ( talk) 13:16, 20 March 2024 (UTC) reply

No, Arb doesn't set precedent and no committee is bound by the rulings of a prior committee. They aren't a court. This is firmly established in the policies that govern Arb. Dennis Brown - 2¢ 03:06, 21 March 2024 (UTC) reply

Binding precedent isn't the only kind of precedent. WP:ASPERSIONS and WP:BRIE, for example. Levivich ( talk) 03:29, 21 March 2024 (UTC) reply

What are you talking about? You absolutely do have a COI when you edit articles about Haile Selassie... Its just not a COI which (by common sense) rises to the level of being significant (unless say you're whitewashing Haile Selassie, then common sense says that your religious COI is now significant). You are confusing a conflict of interest and a conflict of interest significant enough to matter on wiki. Horse Eye's Back ( talk) 16:14, 20 March 2024 (UTC) reply

COI on enwiki is defined as Conflict of interest (COI) editing involves contributing to Wikipedia about yourself, family, friends, clients, employers, or your financial and other relationships. Haile Selassie is not any of those things to Goldsztajn as I understand it. This is different from challenges editors might have, due to their investment in a topic, in obtaining neutral point of view. Barkeep49 ( talk) 16:20, 20 March 2024 (UTC) reply

How can he be a devout Rastafarian without having relationship with Rastafarianism? If you kept reading COI you would have found "Any external relationship—personal, religious, political, academic, legal, or financial (including holding a cryptocurrency)—can trigger a COI." Horse Eye's Back ( talk) 16:26, 20 March 2024 (UTC) reply

You're right that I didn't give enough weight to your sentence contrasting a COI and a COI enough to matter on wiki. Barkeep49 ( talk) 16:51, 20 March 2024 (UTC) reply

For what its worth I wasn't trying to contribute something high minded or philosophical about COI (I have such thoughts, but this is not the venue) I was basically just summarizing the sentence after that one in COI "How close the relationship needs to be before it becomes a concern on Wikipedia is governed by common sense." Horse Eye's Back ( talk) 17:00, 20 March 2024 (UTC) reply

Yes I think that sentence often gets lost. Barkeep49 ( talk) 17:02, 20 March 2024 (UTC) reply

For me its the most important but also hardest to quantify part... If not only because there is a certain type of new/difficult editor who consistently mistakes their own interests for common sense. Even being charitable I must say its the most nebulous part of the guideline. Horse Eye's Back ( talk) 17:11, 20 March 2024 (UTC) reply

It's the most nebulous because it's treating bias as synonymous with conflict of interest, WP:COINOTBIAS: Beliefs and desires may lead to biased editing, but they do not constitute a COI. Austrians editing the Austria article don't have a conflict of interest which in itself potentially prohibits them from acting in accordance with our policy and guidelines; a bureaucrat working in the Austrian Ministry of Foreign Affairs clearly would. The problem with this discussion is that by defining every relationship as a conflict of interest, everything *is* nebulous. We need to focus on the trigger - the threshold where a particular relationship imperils a capacity to uphold policy and guidelines. For me, any financial relationship is an automatic trigger with a low threshold for risk of COI. Once we leave the area of financial, I would have a higher threshold, that is, relatively speaking, treat the COI risk level as lower (but not non-existant). The point of conflict of interest is not relationships per se, but how particular relationships are far more likely to bring one into conflict with policy and guidelines. If I have a critique of the current COI page it is that it is very heavy on instructions on what to do (has an engineer's manual feel to it), but with scattered, abstract (few?) elaborations of what constitutes COI editing. Most bar and solicitors' associations have actually quite detailed examples of what constitues COI (eg [1]) - we should be elaborating something of that order. I'd be happy to be pinged to work on that. Regards, Goldsztajn ( talk) 21:52, 20 March 2024 (UTC) reply

They aren't synonymous but they do significantly overlap, biases can result in conflicts of interest and conflicts of interest can result in biases. I don't think they're as cleanly distinct as you're claiming they are. Horse Eye's Back ( talk) 21:57, 20 March 2024 (UTC) reply

Regarding the requests for recusals, there is definitely some probing warranted given the involvement of various arbitrators outside of this venue. I will give my thoughts based on what I have seen here on the evidence page from diffs and provided comments. I believe that Firefly's removal of the comment was definitely inappropriate given it was on the talk page of a party involved in an active case and was even relevant to the case at hand. Surely there could have been another admin to handle the situation since the block and comment removal were not as time-sensitive as the oversight block Primefac had performed on Fram. Given the question's relevance to the case, its removal by an active arbitrator on the case could definitely be construed as an attempt to protect one of the parties. I don't believe this to be the case, however. I believe the response given by Firefly is unsatisfactory as it left it entirely up to Nihonjoe as to whether or not the question would even be restored, let alone answered, which would effectively result in it being hidden from anyone who either hadn't seen it or checked the history. I believe this could be indicative that there is some bias, but at the very least, Firefly should have made better choices. As others have said HJ Mitchell's involvement and comments at the AN thread seem to indicate there is a bias here since they attempted to downplay the issues rather than look into them. I believe both of these arbitrators should recuse themselves in the interest of this case's arbitration being neutral in nature. Noah, AA ^Talk 18:57, 20 March 2024 (UTC) reply

I would like to further state that I am also of opinion that members of the committee should not be involving themselves in issues of either admin or bureaucrat misconduct outside of arb com since they are the only venue of appeal if the community can't deal with the problem itself. If an arbitrator has already expressed their opinions in a thread at AN for example, it gives off a sense of impropriety since it may establish that the arbitrator has a potential bias towards either a party or a particular outcome. Noah, AA ^Talk 19:31, 20 March 2024 (UTC) reply

I'll reply to your endnote about me briefly before it takes one a life of its own: I made a single comment in that ANI thread, not the comments (plural) you refer to and I reject any suggestion that I was downplaying any issue or even offering an opinion—the evidence presented before my comment was four diffs which were all multiple years old; old evidence is not of much value of itself but could be suggestive of an ongoing problem, which is why I asked if there was anything more recent. HJ Mitchell | Penny for your thoughts? 19:50, 20 March 2024 (UTC) reply

I hadn't meant to pluralize comment. Your attitude in the comment seemed dismissive which is what I believe others have been pointing at as an opinion. Whether or not the evidence is old is besides the point. The fact that COIs weren't even disclosed until after the allegations came to light should have been concern enough. Unlike other situations, this was an ongoing breach of COI guidelines until Nihonjoe admitted to having the COIs. In most cases, one may breach a policy or guideline and then would be in compliance if they stopped doing the behavior in question. Noah, AA ^Talk 20:14, 20 March 2024 (UTC) reply

• I don't know about any particular request for recusal. But any arbitrator who participated in the AN thread about Nihonjoe, either to comment in it or to attempt to close it, should recuse. Recusal is a "double-sided shield"—it keeps ArbCom from making any decisions that may be questioned as partisan, and it also keeps the individual arbitrator from being seen as participating while WP:INVOLVED. Whether or not there is actual bias there, it certainly raises the appearance of impropriety. Seraphimblade ^{Talk to me} 21:36, 20 March 2024 (UTC) reply
  • Agreed to a point. Buffs ( talk) 22:14, 20 March 2024 (UTC) reply

  Yes, that would seem appropriate in most situations. Noah, AA ^Talk 13:12, 21 March 2024 (UTC) reply

• Can the Arbs clarify what is happening here? As I understand it, the following has happened/is happening:
  1. HJ Mitchell has been requested to recuse and has declined
  2. The Arbs are currently considering whether they should be made to recuse.

BilledMammal ( talk) 02:08, 24 March 2024 (UTC) reply

As far as I know 1 hasn't happened and 2 is definitely not happening (probably because I'm not the only Arb who thinks 1 has occurred). Now maybe HJ Mitchell sees it differently - though despite the negative words someone had for me when I said this yesterday I will repeat what I said up thread when I noted I think there's a high barrier to IAR'ing policy which hasn't been met here especially in a situation that seems to be why the policy was written in the first place. In that case I am willing to say that this thread more than satisfies that part of policy. Barkeep49 ( talk) 02:19, 24 March 2024 (UTC) reply

To avoid any misunderstanding, your position is that no request to recuse has been made because it was made here rather than on HJ Mitchell's talk page?

WP:NOTBURO would appear to apply, specifically A procedural error made in a proposal or request is not grounds for rejecting that proposal or request; editors have clearly requested that an Arbitrator recuse, and have provided evidence for the request. With that said, if the paperwork needs to be submitted I'm more than willing to copy and paste the rationale to HJ Mitchell's talk page - while I haven't given much consideration yet to whether I believe he should recuse, I am very concerned about any failure to consider an otherwise valid request on the grounds of misfiling. BilledMammal ( talk) 02:26, 24 March 2024 (UTC) reply

I have been trying to lay out the principles I see behind the policy - foremost the community has traditionally taken a dim view of when people try to stretch ARBPOL to meet their needs (often done by arbitrators themselves) and the policy seems to me to have been created to avoid exactly the situation which happened here - which is what NOTBURO says we should do. I even wrote to El C, before this thread had started, the policy in hopes he would fix it. Barkeep49 ( talk) 02:38, 24 March 2024 (UTC) reply

Looking into the history of that wording I think you may be reading more into it than exists - it was a change as part of an omnibus modification with that specific aspect receiving, as far as I can tell, no discussion.

However, if you are saying the only way this will proceed is if someone copies and pastes the request to HJ Mitchell's talk page, then I will do so tonight, although I will say that using the wrong venue appears to be the sort of procedural error that NOTBURO warns us against denying a request for. BilledMammal ( talk) 03:18, 24 March 2024 (UTC) reply

They would still ignore it anyways since anything stemming from this discussion would be tainted. This is a case of just drop it and move on. Noah, AA ^Talk 03:25, 24 March 2024 (UTC) reply

I'm not that cynical yet; I still trust the committee and believe that if all the procedural forms are followed they will give the request the consideration it deserves - although I do believe they should have considered the request when it was made rather than refusing it on the grounds of a procedural error. BilledMammal ( talk) 03:28, 24 March 2024 (UTC) reply

To meet procedural requirements, I have now posted requests to recuse on the talk pages of Firefly, ToBeFree, and HJ Mitchell per El C's evidence. While I do not yet know whether I believe these editors need to recuse, I am certain that these requests need to be given proper consideration. BilledMammal ( talk) 05:04, 25 March 2024 (UTC) reply

@ Barkeep49: Will the committee now consider proceeding to step 2 since proper procedure has been followed and both ToBeFree and HJ Mitchell have refused to recuse on their talk pages (noting that Firefly has not responded yet)? There are multiple people here who feel these arbitrators should be recused from the case. Noah, AA ^Talk 15:57, 25 March 2024 (UTC) reply

Just noting for those who have not seen it yet, my response on my talk page. firefly ( t · c ) 12:40, 26 March 2024 (UTC) reply

Five days ago when Harry wrote @El C: I'll reply here to avoid clogging up the evidence page since this reply isn't evidence. Your request is illogical..., that was #1 happening. Harry's OP was an exceptionally clear decline of the recusal request, with reasons given. At this point, there is no point in repeating this request and response on another page on this website. I don't see any impediment to arbcom beginning #2 if that hasn't already begun. Levivich ( talk) 03:36, 24 March 2024 (UTC) reply

Case for recusal

The criteria I am supposed to use for evaluating the need to recuse is Typically, a conflict of interest includes significant personal involvement in the substance of the dispute or significant personal involvement with one of the parties. Previous routine editor, administrator or arbitrator interactions are not usually grounds for recusal. I want to make sure I understand the requests. They are based off of this diff for HJM, this diff for TBF, and this revert and this reply (or arguably lack of reply) From Firefly. And based on this people are suggesting that these are "significant involvement in the substance of the dispute" or that something about this situation isn't typical, though I admit to not quite understanding the "not typical" claim if that is what's being advanced. I want to make sure I understand the strongest possible case for recusal so I can make a fair assessment. Barkeep49 ( talk) 16:02, 25 March 2024 (UTC) reply

"I want to make sure I understand the strongest possible case for recusal so I can make a fair assessment" if thats the case I would be remiss in not reminding you of [2]. Horse Eye's Back ( talk) 16:11, 25 March 2024 (UTC) reply

@ Horse Eye's Back Wikipedia_talk:Arbitration/Requests/Case/Conflict_of_interest_management#Drafter's note. Barkeep49 ( talk) 16:13, 25 March 2024 (UTC) reply

Yeah thats more or less how I feel about it (note that I am not requesting any recusals) Horse Eye's Back ( talk) 16:22, 25 March 2024 (UTC) reply

I appreciate that you abstained from the vote to expand the scope. I think that was the right call (not recusing from the case overall, but recusing from certain aspects of it). Levivich ( talk) 16:36, 25 March 2024 (UTC) reply

Two points. First, there is a significant grey area in that policy; everything between routine editor, administrator or arbitrator interactions and significant personal involvement. I suggest you err on the side of exclusion, rather than inclusion; caution is never a bad thing.

Second, I encourage you to consider the purpose of our involved policies, which include preserving the communities trust in the process. If a non-trivial number of editors hold a belief, that is not unreasonable, that an arbitrator is involved, then I believe that in line with both the purpose and the spirit of our policies it is better for the arbitrator to recuse themselves. Failing to do so will lead to a lack of community faith that the result of the process was the right result, and will damage the legitimacy of the committee impeding it from effectively resolving disputes in the future. BilledMammal ( talk) 02:20, 26 March 2024 (UTC) reply

ARBPOL and INVOLVED differ in key ways that this reply doesn't seem to acknowledge. I also read the policy as having a clear standard; any standard is going to have edge cases but that isn't the same, for me, as having a grey area. In fact one of the differences between the two policies for me is I find there to be more of a grey area in INVOLVED than in ARBPOL. As to the second paragraph, asking an 8 member committee to rule on this case also carries risk - for instance that Arbcom gets the result wrong and that damages the legitimatacy of the committee in the future. Barkeep49 ( talk) 04:43, 26 March 2024 (UTC) reply

My reading is that this policy clearly defines when an arbitrator is involved (significant personal involvement), and clearly defines when an arbitrator isn't involved (routine editor, administrator or arbitrator interactions) - but leaves significant space between the two where an arbitrator may or may not be involved.

As to the second paragraph, asking an 8 member committee to rule on this case also carries risk - for instance that Arbcom gets the result wrong and that damages the legitimatacy of the committee in the future.

You face a difficult choice. On one hand, if the editors concerned about involvement are right, then these arbitrators not recusing will significantly increase the chance the result is incorrect. On the other hand, if they are wrong, then these arbitrators recusing may increase the chance the result is incorrect.

However, regardless of whether they are right or wrong, if these arbitrators do not recuse and their participation changes the result of the case than a sizable portion of the community will be convinced that the result was incorrect and the legitimacy of the committee will be guaranteed to take a substantial hit. In a process like this it is not enough to do the right thing; you have to be seen to do the right the thing.

As a final note, I wouldn't be too concerned that eight is too small; at the risk of comparing ArbCom to a judicial system, arbitrators are much closer to judges than to juries, and typically no more than three judges sit on a bench trial. BilledMammal ( talk) 05:17, 26 March 2024 (UTC) reply

( edit conflict)That may be so, but I know that when I was an arbitrator, I kept myself out of contentious situations that I thought might wind up at ArbCom. I failed at that once, specifically the Media Viewer case, where I commented on it extensively since I figured a WMF issue wouldn't wind up at ArbCom, but I was wrong and immediately recused myself there due to my clear involvement. I think it was an error in judgment for so many arbitrators to get involved in this situation, but that ship has sailed, and those arbitrators do need to recuse, even if the result of that is suboptimal. Seraphimblade ^{Talk to me} 05:25, 26 March 2024 (UTC) reply

I haven't weighed in on recusals, but to think 8 Arbs isn't enough seems silly to me. 5 would be more than enough if the community had full trust in them to handle the case. The whole purpose of having so many Arbs it to make sure we have enough after half of them recuse or are too busy in real life. We don't need all or most on every case, even if it works out that way most of the time. The community having full faith in whoever is overseeing the case is infinitely more important the number participating. Dennis Brown - 2¢ 06:36, 26 March 2024 (UTC) reply

I would have to agree with what has been echoed by others in that 8 would definitely be enough arbitrators to handle this case. I believe that all three of the arbitrators whose recusals have been requested should be recused for the sake of maintaining the community's faith in any outcome that is reached here. These arbitrators have refused to recuse and many here have found the explanations to be unsatisfactory since it does not negate the reason why the recusals were requested in the first place. The risk posed by keeping these arbs on the case would far outweigh the cost of excluding them. The risk is not only that the committee makes a bad call here but also that the community loses faith in future decisions made. Noah, AA ^Talk 14:29, 26 March 2024 (UTC) reply

I will name two decisions that I think suffered from having a small number of arbs weigh in and which reached a decision that arbs involved don't look back on with pride: Wikipedia:Arbitration/Requests/Case/Antisemitism in Poland and Wikipedia:Arbitration Committee/Noticeboard/Archive 14#Arbitration motion regarding Mark Ironie and CorbieVreccan (technically there were more than 10 active arbs, but this is more a technicality than a reality). My experience is that when the committee is at 10 or less the quality of decision making goes down rapidly. Just getting this discussion going with 8 arbs shows some difficulties when you have a limited subset of arbs. Barkeep49 ( talk) 16:26, 26 March 2024 (UTC) reply

I dunno, BK...

You, last week: An arb speaking on their own is much easier than ~7 arbs agreeing on wording which is what it takes to post at ACN.

You, today: Just getting this discussion going with 8 arbs shows some difficulties when you have a limited subset of arbs.

7 arbs is too many, but 8 arbs is not enough? Poor decisions may have nothing to do with the number of people making the decision. Levivich ( talk) 17:02, 26 March 2024 (UTC) reply

I admit I'm surprised you didn't understand the distinction between these two things but since you didn't I'm happy to explain. 1 arb speaking on their own is much easier than 7 arbs agreeing on wording. That speaks to the agility of the committee and how much time is worth to spend on an announcement once a majority of the committee has already agreed on the course. Because, as I pointed out today, getting any discussion/decision is hard and the point of having a committee is to ensure a diversity of perspectives, hopefully reflecting the community as a whole (or at least the ACE electorate) and not just the handful of people who are motivated enough to comment are considered rather than relying on the judgement of any individual. Barkeep49 ( talk) 17:04, 26 March 2024 (UTC) reply

Perhaps I misunderstood but you seem to have said that it's hard to start a discussion with as few as 8, but hard to come to final agreement on language when there are as many as 7. I'd suggest that the difficulties are the same whether there are 7 people or 11 involved: that's not a delta that makes a meaningful difference. There is no magic that happens when you move from 7 or 8 people to 11 or 12. It'd be easier if there were only 2 or 3, and harder if it were 50 or 100, but between 8 and 11, the gap is too small to matter. Levivich ( talk) 17:17, 26 March 2024 (UTC) reply

I agree that coming to an agreement with 8 arbs is easier than 15. My experience as an arbitrator suggests that the difficulties are the same whether there are 7 people or 13 involved is wrong. The reason is wrong is because ArbCom is a global asynchronous committee with an intentionally decentralized decision making process (e.g. no chair) made up of people unusual enough to be Wikipedians and who are unusual even among Wikipedians for wanting the job of arbitrator. I have also provided evidence of how the quality of a decision (rather than merely the alacrity with which it can be reached) declines with a smaller committee. "Ease of decision making" needs to be balanced against "quality of decision reached" and when we're in a month long process to begin with I think we've already decided to prioritize "quality of decision" over "ease of decision making". Barkeep49 ( talk) 17:23, 26 March 2024 (UTC) reply

OK but your evidence is poor: two bad decisions in 5 years in which one had 9 arbs voting and the other had 6. That might be "evidence" but it's not conclusive and one might even say it's "cherry picked." Betcha I can find two bad decisions in the same time period with 10+ arbs. One would have to go through all the decisions in the last 5 years, rate them as good or bad, and then see if there is any correlation between the number of arbs voting and the quality of the outcome, to really come out with any conclusive or even convincing evidence--and even then, it would be evidence of correlation, not causation. Still, I highly doubt that such an analysis is going to find that there is something that happens when we go above or below 10 (or any other numerical threshold between 8 and 11).

Look at this series of recent events:

1. Arbcom didn't do or say anything (or didn't do or say enough, e.g. open a case, post a statement that it was investigating, etc.) when it received private evidence of undisclosed COI editing (and admin tool abuse? idk what exactly they rec'd via email), which AFAIK was no later than December 2023 -- this silence/inaction led directly, in my estimation, to off-wiki outing, which led to the community forcing arbcom's hand
2. Arbcom still didn't say anything official for two months despite repeated requests from multiple people (including from me in late Feb)
3. During that period, four of 12 active arbs (a full third of the active panel!) got directly involved in the underlying dispute, three of whom expressed "nbd" opinions, and two of whom used their admin tools (in one case, OS)
4. Only one of those four recused (the OS, good on him)
5. We are told that the recusal requests for the other three on the evidence page, and one recusal decline on this page, did not officially happen because they happened on the wrong page on this website
6. Once the recusal requests were put on the right pages, the remaining three all declined to recuse
7. We are told that one reason those three should not recuse is because arbcom will make a poorer decision if it has 8 arbs instead of 11

The above sequence is the kind of thing that would make people lose faith in arbcom. So this is one example of there being 11 arbs, and arbcom still coming to (multiple) bad decisions.

So is arbcom going to be voting on these recusals? I'm almost afraid to ask this next question: if arbcom does vote on these recusals, all three arbs who are being asked to recuse are going to abstain from all three votes, right? Levivich ( talk) 17:53, 26 March 2024 (UTC) reply

ARBPOL says only Should the arbitrator not respond, or not recuse, the user may refer the request to the Committee for a ruling. so there is no binding requirement for a formal vote (e.g. an informal discussion is allowed), nor any mention of who should or should not participate in such a discussion or vote. It seems obvious that anyone recused from the case as a whole would not participate, and equally obvious that those whose recusal is being discussed will participate in discussion to provide their views, answer questions, etc. but would not be voting regarding themselves.

In general, whether they should abstain from voting regarding their colleague's recusal would depend how similar the basis for the requests is. e.g. if the basis for asking Arb A to recuse is because of their prior involvement with Party X over a different matter, and the reason for requesting Arbs B and C to recuse is because they took part in the same discussion of relevance to the case. In this case B and C should abstain regarding each other but not regarding A, while A would not need to abstain regarding anyone other than themselves. Thryduulf ( talk) 18:16, 26 March 2024 (UTC) reply

I'd believe there is an expectation at this point that the arbitration committee would have to take up this issue and move to a vote on whether any or all three of these arbitrators should be recused from the case. This isn't a case of one person asking an arbitrator to recuse and then referring it to the committee. There are a substantial number of editors here who feel these arbitrators should be recused from the case. It would honestly look very bad if the committee decided to not vote and simply proceed given the amount of tension that has already occurred over this. I believe refusing to vote would only serve to erode the trust that the community has in this arbitration committee. Noah, AA ^Talk 18:30, 26 March 2024 (UTC) reply

For my part, I consider myself recused on any discussion or any vote (if it were to happen) about the matter. firefly ( t · c ) 18:22, 26 March 2024 (UTC) reply

@ Levivich, @ Hurricane Noah and others, we're currently discussing and voting on the recusals. The relevant Arbs are abstaining from the vote. Moneytrees🏝️ (Talk) 18:32, 26 March 2024 (UTC) reply

Is there any reason this discussion and vote isn’t being held publicly? I think transparency would be beneficial here. BilledMammal ( talk) 18:34, 26 March 2024 (UTC) reply

Thanks to FF and MT for the updates. @BM: I was thinking about this when I'd asked a similar question at WT:ACN, and I was re-reading WP:ARBPOL, and, to my surprise, it literally says at Wikipedia:Arbitration/Policy#Transparency and confidentiality that Committee deliberations are often held privately, though the Committee will make public detailed rationales for decisions related to cases, unless the matter is unsuitable for public discussion for privacy, legal or similar reasons. So it seems private deliberations are the norm, and only public decision rationales are expected, and even then, only for those cases that are suitable for public discussion. I think the community should review this language, but that is in fact the policy right now: private deliberations are the norm. Levivich ( talk) 18:37, 26 March 2024 (UTC) reply

Thank you, I was not aware of that; I think when this case is over the community will need to review various aspects of ARBPOL, including when arbitrators should recuse and how transparent the committee should be. BilledMammal ( talk) 18:40, 26 March 2024 (UTC) reply

...and since I've already looked this up, might as well share: the language "deliberations are often held privately" has been there since 2005 (!). Kind of some interesting ancient wikihistory: Up until March 2005, it said (emphasis in the original), "Initially, we will keep our deliberations private, based on a semi-formal vote amongst Arbitrators. However, both Fred Bauder and The Cunctator have expressed strong distaste for this option, so the Arbitrators are far from unanimous on this point." Then on March 29, 2005 that was changed to "Deliberations are often held privately, but Arbitrators will make detailed rationale for all their decisions public," and it's remained in more or less that form since. Levivich ( talk) 18:51, 26 March 2024 (UTC) reply

Well said. This whole thing has been a bit ridiculous. Lulfas ( talk) 22:09, 27 March 2024 (UTC) reply

many here have found the explanations to be unsatisfactory it is worth pointing out that multiple other editors have found some or all of the explanations to be adequate. Harry's situation is the one I have paid most attention to and I do not see his involvement as even approaching the level at which he would need to recuse. Thryduulf ( talk) 17:18, 26 March 2024 (UTC) reply

Having read through this case and its preceding threads, it is disappointing that in a case centred on COI that ARBCOM members who intervened in the original discussion are not recusing in the spirit of INVOLVED. There's a thematic reflection at play; the spirit of COI and INVOLVED are both about understanding situations of personal entanglements, and further about the potential appearance of personal entanglements. It is also perplexing to read in a section titled "El C's request for recusal" the claim that there has been no request for recusal. Perhaps these stem from a similar cause, some disconnect between the private discussions that are necessarily taking place and the public ones here. Perhaps they are unrelated. In any case, it is worth considering the interaction of COI and INVOLVED here, the discussions surrounding the case show clearly the import these policies around personal entanglements have for the community. CMD ( talk) 16:23, 26 March 2024 (UTC) reply

Seeing a couple of editors making the argument that, because some editors have expressed the view that there should be recusals, it is necessary to recuse in order to maintain community trust, I don't think anyone here can say with certainty that the community shares the views of the editors who have been vocal here, or that the community will lose trust if some Arbs end up disagreeing with the editors who have been vocal here. Just one fish's opinion, but I don't feel strongly about anyone further needing to recuse. -- Tryptofish ( talk) 20:58, 26 March 2024 (UTC) reply

I haven't seen that argument, nor is it made in my comment above so I'm not sure what this is replying to. CMD ( talk) 11:50, 28 March 2024 (UTC) reply

Vote on recusal

See the results of the vote for recusal at Wikipedia_talk:Arbitration/Requests/Case/Conflict_of_interest_management#Votes_on_recusals. Barkeep49 ( talk) 20:59, 1 April 2024 (UTC) reply

I've been thinking about this on and off for a decade (and I am completely, genuinely serious): why are we concerned about someone's conflict of interest?* Why aren't we simplifying this down to assessing and addressing their behavior? Which is more inappropriate? A person who edits an article about the company in which they work but doesn't disclose that or someone who edits an article in such a manner that endorses their own political/religious/etc opinion. To be blunt, it all depends on what was actually done. If the first edit is noncontroversial (such as a change in CEOs), I don't see the problem. If the second edit is appropriate (such as "<Trump/Biden> said '<insert quote here' on <insert date>.<insert source>"), I don't see a problem there either.

In the instance listed above, someone allegedly put in their company info, someone else removed it, the original author put it in again and berated the editor on their talk page while threatening them. The issue of the company info in the article is nearly inconsequential. We can discuss it as a group and come to a consensus. The issue is that the original editor abused their position to push an agenda. Whether it was COI or pushing a POV and whether or not they were an admin, IMHO, the behavior was the significant problem, not the edit.

* I get that at higher levels of responsibility, we're potentially dealing with applying "force" against those who have a vested interest "against" you/your interests. These can represent very real conflicts of interest with the potential for legal implications. I get that those need different criteria. For example, if you are a Democrat/Republican and you block Republicans/Democrats, you are potentially engaging in the suppression of information. If done at the behest/benefit of a political party, it could be involving the Wikimedia foundation in an international scandal. It's no different for companies in competition.

Anyway, there's my thoughts. Let's focus on the problems rather than divining someone's motivation or potentially outing personal info about them. Buffs ( talk) 22:14, 20 March 2024 (UTC) reply

I've long felt that "COI edits" are a problem only to the extent that they are NPOV violations. Sometimes, however, COI talk page edits can be a problem in ways that are not obvious, with a greater risk of deception. As you allude to, COI admin or functionary actions can be a problem more generally. (My third sentence is not about anybody in particular in this ArbCom case, just a general view.) -- Tryptofish ( talk) 22:23, 20 March 2024 (UTC) reply

I think there are at least a couple reasons.
1. Sometimes it's not immediately clear that there is an NPOV issue. It can be difficult for the average editor to assess niche and complex topics, especially when it comes to determining whether the sources in an article are accurately represented and whether the choice in sources/content is a BALASP reflection of the larger corpus of mainstream RS on the topic. When someone has declared a personal, cultural, religious, etc. COI on a topic, we know to give their edits more scrutiny. If they're editing in good faith, they might not even realize their COI is affecting their choices. Declarations thus help us monitor and prevent potential walled gardens.
2. Having guidelines on this allows us to sanction users who are found to have a serious undeclared COI. They can also serve as an enhancer when sentencing an editor whose edits violated NPOV or other content policies. JoelleJay ( talk) 23:19, 20 March 2024 (UTC) reply

Having guidelines on this allows us to sanction users who are found to have a serious undeclared COI. Having a conflict of interest, declared or otherwise, is not something that is or should be sanctionable. What matters is whether someone has or has not made edits that harm the encyclopaedia or its users. If the CEO of MegaCorp edits the article about MegaCorp, without declaring they have a COI, but all they've done is correct typos, updated figures, and added independent reliable sources to support material that is neutrally written and clearly DUE then no harm has come to the encyclopaedia at all. In contrast an editor who adds a slew of unreliably sourced undue material positive/negative towards MegaCorp to the article and removes material of the opposite POV because MegaCorp sponsors the sports team they do/don't support has harmed the encyclopaedia without having any COI. Thryduulf ( talk) 23:32, 20 March 2024 (UTC) reply

By "later found to have a serious undeclared COI" I mean "has a significant actual COI and has been editing significantly in that topic". A low number of minor edits like you describe would be easy to adjudicate re: violation of content policies and would not be "significant"; many major edits, especially spread over years, would be much more difficult to comb through and suggest much greater influence on how the topic is presented on WP.
I also thought of another reason for guidelines: they serve as a deterrent for more subtly problematic COI editing by any editors who actually want to stay in this community. JoelleJay ( talk) 23:51, 20 March 2024 (UTC) reply

( edit conflict)} I agree very strongly with Buffs' comments about focusing on behaviour. If behaviour is inappropriate it is inappropriate whatever its motivation. If edits improve or harm the encyclopaedia then they are good or bad regardless of why they were made. Thryduulf ( talk) 23:21, 20 March 2024 (UTC) reply

That is nice in an ideal world, where every edit gets thoroughly scrutinized to determine whether it accurately reflects its sources. We do not live in that ideal world where we have an unlimited resource of volunteer time. In the real world, COI editors make edits which slant articles (perhaps even without intending to), and those edits on more obscure topics can stand for years, sometimes even when they are absolutely blatant spam. So, editors are required to declare COI and restrict themselves on their COI topics to edit requests and AfC, so that other editors know that the edits in question need close scrutiny. Failure to declare COI, and editing mainspace when having a COI rather than going through review processes, is in itself misconduct. If you're doing things right, the articles and edits you propose should sail right through the review process without issue. If they wouldn't have, well, that's all the more reason they should have been checked. Seraphimblade ^{Talk to me} 00:05, 21 March 2024 (UTC) reply

Yes, as you say, the problem is bad editing not the existence of a COI and there is not a strong correlation between existence of a COI and edits being good or bad. Thryduulf ( talk) 00:48, 21 March 2024 (UTC) reply

@ Thryduulf what are you basing there is not a strong correlation between existence of a COI and edits being good or bad on? Barkeep49 ( talk) 01:37, 21 March 2024 (UTC) reply

The vast majority of edits to Wikipedia are good edits. The only a fraction of bad edits to Wikipedia are not done by people with a COI with the subject of the article. It's impossible to know how many edits are made by people who have a COI with a given subject, but most edits by people with declared COIs are not harmful (if it were otherwise then COI editing, declared and undeclared, would be completely banned not simply discouraged). When edits to an article are good (sourced, neutral, etc), nobody looks to see whether the editor has an undeclared COI or not (nor whether they are paid or not) only when the edits are consistently bad do we look into the editor's background (one-off bad edits just get reverted).

There are of course bad edits made by people with significant conflicts of interest, I'm not denying that. The point is simply that the mere existence of a COI between an editor and a subject is a poor predictor of whether the edits that editor makes to the subject will be good or bad; and the existence of a bad edit to an article is a poor predictor of whether the editor has a COI with regards that article. Also, even when the edits are bad and a COI is the cause, the problem is the bad edits not the COI. Thryduulf ( talk) 02:33, 21 March 2024 (UTC) reply

There's an unknowable to this problem - what an editor with a conflict of interest does *not* do. The CEO of Megacorp is unlikely to add material about cases of harassment or wage theft. This is why I want to know if Megacorp's CEO is editing - in part because we can, if needed, arrive at a view as to what they are not doing. Regards, Goldsztajn ( talk) 02:50, 21 March 2024 (UTC) reply

Someone doing or not doing something is not, in itself, evidence of a problem. If the CEO of Megacorp is adding positive material to the article that's a problem if it results in the article being unduly positive, it's not a problem if it results in the article being balanced (i.e. it was unduly negative previously). The CEO of Megacorp not adding negative material is not a problem if they are also not adding positive material, or if the negative material is already in the article, or it hasn't been covered by independent reliable sources. THE CEO of Megacorp removing soured, due, information from the article is usually going to be an issue - but not always, e.g. if it is a copyvio, or they're moving it to a spinout article. Context matters. Thryduulf ( talk) 03:07, 21 March 2024 (UTC) reply

Let me give a specific example: a minor, but notable company satisfying WP:NCORP from a part of the world where English is not used widely (say Indonesia). Megacorp Indonesia's CEO edits the company's Wikipedia article website to indicate that the total workforce has declined from 1,214 to 875. It's sourced to a short English language news report in the business press; the source is listed as reliable in our perennial sources and the source only mentions savings due to cost efficiencies. NPOV, no harm? However, the reasons for the reductions are layoffs due to a unionisation drive which are covered in local Indonesian language sources. If we are aware of the CEO editing their company's article we are far more likely to be vigilant over those edits and put in place oversight and spend the time and energy necessary - if we unaware of the CEO editing no one is going to have even the potential to put in the oversight necessary. Regards, Goldsztajn ( talk) 03:14, 21 March 2024 (UTC) reply

Thryduulf is correct, IMNSHO. If they are just adding that the "workforce has declined from 1,214 to 875", that is an objective and neutral fact. It IS NPOV. If we find out later that there is more to it, we can add more to it to provide context, but the statement added is not incorrect. Likewise just because they are trying to unionize doesn't mean the company is wrong to lay off people. NPOV runs both ways. Just because someone has a financial COI doesn't mean we should extra scrutinize their edits and wikistalk them.

There are two problems this mentality perpetuates:

1. The most obvious is that we push people away who are closest to the subject and know the most. In fact, in contentious areas, editors routinely use COI as a club to push away the "opposition". Let's hypothetically say someone is involved as a volunteer for a non-profit organization. They aren't getting paid, but they are highly vested in it. They don't have a financial COI, but they are more likely to push the positives of the NPO. How can we distinguish that from someone who is just a proponent? We can't. Instead, whether codified as a rule or not, the pervasive focus should be: "focus on the edits, not the editor". If someone has a COI
2. Such a mentality on COI can quickly lead to "well you're a fan and you only want the positive about subject X". If such behavior drives off proponents, you're left with only the people who despise the subject and you get a negative bias in your articles (something we are seeing in WP). Instead, we should promote collegial discussions instead of advocacy (whether pro or con for subject X). COI is used as a club by opposition.

Buffs ( talk) 16:58, 21 March 2024 (UTC) reply

People who despise the subject also have a COI. Just disclose, on the article talk. That tells other editors you'd welcome them taking a look. Valereee ( talk) 19:47, 21 March 2024 (UTC) reply

That's sort of the subject of the hearing. Forcing them to say who they work for can be a form of outing. As someone previously threatened by a well-known international terrorist organization by name with my address and a picture, I'm pretty wary of keeping my anonymity as much as possible. Buffs ( talk) 20:05, 21 March 2024 (UTC) reply

If you can't say you work for the organization, don't edit the article about it. Problem solved. No one is forcing you to edit that article. Valereee ( talk) 21:39, 21 March 2024 (UTC) reply

Aren't those who chose to edit topics they have a COI with self-outing to some degree from the general (employer, place of residence, or religion) to the almost unavoidably explicit (e.g. making a page about yourself)? Horse Eye's Back ( talk) 16:44, 25 March 2024 (UTC) reply

Its all about trust, the community by the large believes that massive unrestricted COI editing would destroy the credibility of wikipedia and their enjoyment of editing... That means that COI editing outside of the restrictions the community has set is disruptive regardless of whether or not the individual edits have any issues at all. Also note an interesting outcome of that framing: undisclosed COI editors are de-facto engaged in harassment of the community (from that angle there's not that much of a difference between an undisclosed COI editor and a sock master, and we don't let sock masters off the hook even when their edits are on their own a net positive). Horse Eye's Back ( talk) 17:08, 21 March 2024 (UTC) reply

"massive unrestricted COI editing would destroy the credibility of wikipedia" Why? Most people who edit WP have an interest in the subject they are editing and have some form of COI. To be blunt, I don't see a problem with a paid editor. They don't control the content on Wikipedia. Anyone can edit it, even if they are paid employees. Buffs ( talk) 20:09, 21 March 2024 (UTC) reply

@ Buffs, I'm sorry, but Most people who edit WP have an interest in the subject they are editing and have some form of COI is just not true in my experience. You may only edit things you have a COI for, but being willing to research and write about a subject is not the same as having a COI about that subject. Valereee ( talk) 21:36, 21 March 2024 (UTC) reply

WP:COI states "Conflict of interest (COI) editing involves contributing to Wikipedia about yourself, family, friends, clients, employers, or your financial and other relationships...Any external relationship—personal, religious, political, academic, legal, or financial (including holding a cryptocurrency)—can trigger a COI." So, by this logic merely holding a cryptocurrency somehow means you can't write about it; those who write about it are not owners of cryptocurrency. If you vote for/against someone, that's a COI. If you're a member of a religion, you can't write about it. etc

Now, that's the exact "letter of the law".

In practice, that's not how it's applied. Buffs ( talk) 23:45, 21 March 2024 (UTC) reply

If you own crypto, you shouldn't write about it on Wikipedia. "Member" of a religion isn't a thing for most religions (they don't keep membership lists, one doesn't sign up), but if you hold very strong religious beliefs, you shouldn't write about that, either. Voting in and of itself creates a weak COI, but if you hold very strong political beliefs, or very strong feelings for/against a candidate in an upcoming election, you shouldn't write about that, either. You should generally edit Wikipedia topics about things where you don't have strong partisan feelings, close personal connection, or any kind of financial interest. Write about things that don't involve you, because it helps maintain neutrality. Levivich ( talk) 00:10, 22 March 2024 (UTC) reply

So if you own something, you shouldn't contribute about it? If you're highly motivated, you shouldn't write about it? If you're a history professor, you shouldn't write about history? By that logic, the only people writing about crypto are people who don't own crypto; no one motivated by a subject is writing about it, only the disinterested; and experts in the field/those most knowledgeable about a subject are disqualified in favor of those least qualified to be accurate. I think you think much too highly about our editors as most contribute to pages they are highly interested in and editors are not even as close to as altruistic as you seem to think they are.

My point is that we shouldn't care what someone's motivation is, just what they contribute. Buffs ( talk) 05:23, 22 March 2024 (UTC) reply

That's not what Levivich wrote and this confuses an interest in something with a conflict of interest. The degree of one's personal interest in a topic is not necessarily related to how one views an issue. I can be highly interested in stamps, but I may hold no particular view on the quality of Hungarian stamps of the 1930s. Strong interest in something is not the same as strongly held views about something. Regards, Goldsztajn ( talk) 05:53, 22 March 2024 (UTC) reply

Buffs, there are editors who write entire series of articles about roads, railway stations, lakes, insects...you name it. None of these folks have a COI. One woman has written literally dozens of articles about TV and radio stations, including four FAs. Some people specialize in certain areas because they have an interest in or knowledge of that area, some because they have access to research materials. I write a lot about food, chefs, restaurants because those are things I'm interested in and have enough familiarity with to assess notability. That doesn't mean I've got a COI for subjects around cuisine. Your hypothetical history professor doesn't have a COI with history. They have an interest in it. You are confusing 'motivated' with 'opinionated'. You're confusing 'interested' with 'involved'. Valereee ( talk) 12:15, 22 March 2024 (UTC) reply

let's agree to disagree. Buffs ( talk) 16:04, 22 March 2024 (UTC) reply

"My point is that we shouldn't care what someone's motivation is, just what they contribute." by this logic should we be changing the way we interact with sock puppets? This doesn't seem compatible with wikipedia as I understand it. Horse Eye's Back ( talk) 16:46, 25 March 2024 (UTC) reply

And suddenly starting a Wikipedia editing agency looks like an attractive business idea! Bon courage ( talk) 16:54, 25 March 2024 (UTC) reply

Only if its app based... "Find my Admin" has a nice ring to it, nothing wrong with paying for admin tool use of course... As long as the edits themselves are good. Horse Eye's Back ( talk) 17:27, 25 March 2024 (UTC) reply

I think the correlation between conflicts of interest and poor decisionmaking is nearly 1:1, that's why there are COI rules in almost every aspect of life: if a person has a COI, their conscious and unconscious biases will impact their impartiality. On Wikipedia, non-neutral editing is one example of the poor decisionmaking that comes from the biases that come from having a COI. The notion that most people with a COI are still nevertheless able to be unbiased is pure fantasy; if it were true, we wouldn't have COI rules all over the real world. Levivich ( talk) 18:36, 21 March 2024 (UTC) reply

For me it's about making sure those articles which need it get additional scrutiny. Disclosing a COI tells other editors, "Hey, I've got a COI here, done my best to be as objective as possible, but I'd welcome someone checking my work." We need to know which articles need that additional scrutiny. Valereee ( talk) 19:11, 21 March 2024 (UTC) reply

This ^^^ is the correct answer. Often, it someone with a COI that has the greatest knowledge about a subject, and thus is able to provide more quality content, but disclosure allows others to verify the neutrality of that information. A good editor with a COI will take criticism to heart and allow others to tweak their work without raising too much fuss. Dennis Brown - 2¢ 05:17, 26 March 2024 (UTC) reply

• Readers. Because readers deserve at least a chance to know. As the COI guideline says: "Readers expect to find neutral articles written independently of their subject" Whereas, if someone with a coi is writing it, editing it, participating in decisions about it, the reader should be able to find that information out, so for example, they can know they are reading autobiography, or the spouse's views, or the statements of the owner, or financial beneficiary, etc. -- Alanscottwalker ( talk) 21:34, 27 March 2024 (UTC) reply

The drafters have shared with Nihonjoe an anonymized summary of the private evidence we've received about him and offered him until the close of the workshop phase a chance to respond to any of it. Barkeep49 ( talk) 03:16, 21 March 2024 (UTC) reply

Too bad Fram never got this, eh? Lulfas ( talk) 16:48, 21 March 2024 (UTC) reply

^^^ this Buffs ( talk) 17:00, 21 March 2024 (UTC) reply

I'm glad this ArbCom has learned from the failings of the past ones in this respect. Der Wohltemperierte Fuchs ^talk Der Wohltemperierte Fuchs ^talk 18:26, 21 March 2024 (UTC) reply

absolutely Buffs ( talk) 19:59, 21 March 2024 (UTC) reply

Well, or they decided to (continue) helping Nihonjoe out. Arbs, especially Primefac, have really not covered themselves in glory here. Lulfas ( talk) 01:31, 22 March 2024 (UTC) reply

@ Lulfas and others, To be clear, we've been sharing these sort of summarized private evidences with parties since 2022's Skepticism and coordinated editing I think? Looking at the stretch of cases between Fram and SCE though, I don't think we really got a chance to otherwise... Moneytrees🏝️ (Talk) 02:06, 22 March 2024 (UTC) reply

I think ArbCom has, with the increase in cases that its ruling on because of private evidence, been working to find ways to provide increased transparency to the community and especially the parties. So that started in Skepticism and coordinated editing where links where shared publicly and where an FoF which relied on private evidence was shared with a party ahead of the proposed decision for feedback. To my knowledge this more comprehensive summary is new. Now one crucial element in allowing ArbCom to have done the summary in this case: this wasn't about harassment. So ArbCom didn't need to do any sort of balancing between those two editors which made it far more straightforward to just share. Despite that caveat, I hope it becomes an ongoing practice, in the way the links is now a practice (and happened as much as was possible in this case). Barkeep49 ( talk) 02:16, 22 March 2024 (UTC) reply

Yeah, what Barkeep said. Moneytrees🏝️ (Talk) 02:22, 22 March 2024 (UTC) reply

Being assessed based on secret evidence is unseemly. Buffs ( talk) 16:33, 22 March 2024 (UTC) reply

What do you mean this case wasn't about harassment? I submitted evidence of harassment as well as evidence of false claims of harassment, other editors did as well. I was harassed relentlessly for months by Nihonjoe et al. They made every effort to have me removed from the community or have my editing in this topic area restricted, they followed me to dozens of conversations on dozens of pages. And all of the harassment was because I was getting close to their secret which they desperately did not want to reveal: massive undisclosed COI editing by BYU/AML editors. Horse Eye's Back ( talk) 16:09, 22 March 2024 (UTC) reply

Seriously @ Barkeep49: WTF? The scope of the case according to the header on top literally is the interaction of COI and the harassment policy. Horse Eye's Back ( talk) 16:14, 22 March 2024 (UTC) reply

@ Horse Eye's Back you've submitted no private evidence of any kind let alone of harassment. You've said you submitted evidence to Primefac but he's recused from this case so that didn't submit it to the committee. I can't speak for him, but if I had to guess what you emailed him didn't make clear it was intended for the entire committee (people will email me "as an arb" sometimes wanting it to be shared with the committee and sometimes not). And my read of that was that what you submitted to PF was about the broader BYU topic which the committee decided not to include as the scope of this case. So what I said is correct: the private evidence did not include evidence of Nihonjoe harassing others and this meant the committee/drafters didn't need to weigh the protection of one side against informed knowledge on the other side. Barkeep49 ( talk) 17:47, 22 March 2024 (UTC) reply

Primefac was instructed to pass anything which hadn't already been submitted (I assume most or all of it has been submitted by others) on but yes the evidence of harassment is in general public. I now see what you mean, by case you mean situation vis-a-vis private evidence you didn't mean the case in general. I get what you're saying, it wasn't like someone had submitted evidence which could open themselves up to off-wiki harassment or whatnot from disclosing to an involved party. Horse Eye's Back ( talk) 17:58, 22 March 2024 (UTC) reply

What information I received was either not directly related to this case or was already presented to the Committee (some I had seen before the case was filed, which is how I knew the private evidence was already known). Primefac ( talk) 18:09, 22 March 2024 (UTC) reply

If you are recused, why are you making decisions about whether something was related to this case? Combined with your actions being a fairly important part of this case, that appears like yet another involved action? Why wouldn't you send it on to the committee and stay butted out? Lulfas ( talk) 11:51, 23 March 2024 (UTC) reply

Note that this information was sent to Primefac after they assured me that they were not involved. Horse Eye's Back ( talk) 16:14, 23 March 2024 (UTC) reply

Can the committee confirm whether they have received any external evidence regarding a conference held in Provo in 2014... in the Birch suite?! ——Serial Number 54129 19:29, 21 March 2024 (UTC) reply

I appreciate the discretion with which this post was written. Speaking only for myself, I personally don't feel comfortable revealing anything about private evidence at this time beyond what we've already shared. Perhaps my fellow drafters/arbs will feel different. Barkeep49 ( talk) 02:18, 22 March 2024 (UTC) reply

I was recently reminded of these diffs [3] [4] from my talk page. Requesting permission to enter them into evidence vis-a-vis Nihonjoe's campaign of harassment against those who questioned the BYU/ALM editing block. This is different from the other diffs I presented because Nihonjoe delivers a formal admin warning ("Consider this a formal warning about your behavior."), again without disclosing their COI. They were involved on multiple levels and yet they still took administrative action. And of course don't forget, they knew the whole time that the complaints against Helps were legitimate... They actually knew that such complaints were just scratching the surface, but they lied, obfuscated, harassed, and abused their position to cover up that bad editing. Horse Eye's Back ( talk) 16:01, 23 March 2024 (UTC) reply

@ Horse Eye's Back: This sounds like it may be related to the evidence that you emailed Primefac during the evidence phase but which I have not seen. Would you mind forwarding a copy of that email to arbcom-en-bwikimedia.org? - Aoidh ( talk) 16:39, 23 March 2024 (UTC) reply

IMO its more related to the public evidence provided during the evidence phase. I don't have a copy, Primefac can forward. Further context can be found at Wikipedia:Administrators' noticeboard/Incidents#Follow up from VPM. Horse Eye's Back ( talk) 17:11, 23 March 2024 (UTC) reply

@ Horse Eye's Back you've made increasingly strong accusations of harassment not all of which are supported by any evidence. Can you in the next day or two please collect all evidence of harassment not currently in your public evidence and the drafters/committee will decide based on that whether to allow you to submit out of process. Thanks, Barkeep49 ( talk) 19:20, 23 March 2024 (UTC) reply

That is my public evidence... That Nihonjoe harassed those who asked hard COI questions about Rachel Helps (BYU) and related accounts across a number of pages both talk and noticeboard. This pattern repeated over at least a four year period without Nihonjoe disclosing any of the relevant conflicts of interest. More diffs regarding SlimVirgin would be [5], [6], and [7] (note that while slimvirgin was a Wikipedia:Conflict of interest/Noticeboard regular Nihonjoe was not, they hadn't posted there since 2012 and have not posted there since these edits in 2020). The harassment of MoKo365 (apparently unrelated to the BYU/AML nexus but related to a different undisclosed interest) all the way back in 2008 was also documented in the evidence presented by Jessintime. I don't know how else to characterize this pattern of behavior other than harassment (primarily hounding as we call it) of editors who threatened Nihonjoe's undisclosed interests. The way Nihonjoe responded to the newest batch of concerns about COI wasn't uniquely aggressive, deceptive, or uncollegial... It was more of the same. Horse Eye's Back ( talk) 20:10, 23 March 2024 (UTC) reply

Do I have to enter Nihonjoe falsely accusing slimvirgin of hounding Rachel Helps (BYU) in their evidence submission [8] here or is that self evident? Even when dead Nihonjoe is hounding them, ironically with evidence free accusations of hounding (that appears to be the MO... To cover the harassment/hounding by preemptively accusing the other party of harassment/hounding, an almost untenable situation for the other party back when Nihonjoe was an admin). Horse Eye's Back ( talk) 20:26, 23 March 2024 (UTC) reply

@ Horse Eye's Back ArbCom has discussed your request, and while these diffs reinforce evidence submitted by yourself and others, it does not substantially introduce new elements to the case. Given this, the Arbitration Committee declines your request for an extension. - Aoidh ( talk) 16:32, 27 March 2024 (UTC) reply

Thank you, I had understood the standard to be reinforce or expand not "substantially introduce new elements." My apologies. Horse Eye's Back ( talk) 16:39, 27 March 2024 (UTC) reply