Wikipedia:Wikipedia Signpost/2018-01-16/Arbitration report

Conduct of Mister Wiki editors case
Conduct of Mister Wiki editors was opened on 1 December, and closed on 7 January. As reported in the previous two issues of The Signpost, this case centered on the provisions of WP:PAY (paid editing guidelines), conflict of interest, and administrator conduct.

A majority of the arbs voted to de-sysop Salvidrim!. Salvidrim! then by bureaucrats less than hour after the majority vote was cast. Salvidrim! and Soetermans were found to be in breach of WP:PAY. Both editors were prohibited from reviewing articles for creation drafts, or moving Articles for Creation (AfC) drafts created by other editors into mainspace. A proposed prohibition on paid editing by these editors was only supported by and failed.

Arb Opabinia regalis, who abstained from the vote to de-sysop, had this to say about the case and the arbs' reasoning in it.

[I]t's hard to avoid the conclusion that we have to do something serious about a volunteer administrator using both the admin tools themselves and his standing within the community for personal gain... [An Arbcom] case doesn't evaluate "community trust", there isn't even really a reliable way to do so - we all know RfA is broken ... I also really dislike making arbcom decisions with a view toward "how things look from the outside" or out of "pour encourager les autres" considerations, but in thinking this through I find it hard not to consider the larger implications of the case.

Notes on a proposed decision began to appear on 31 December. Included in the notes was based on private correspondence indicating that Salvidrim! had initiated a checkuser block review at the request of an unnamed banned editor. based on private correspondence indicated that he had "repeatedly coached [a Mister Wiki executive] on how to avoid drawing community scrutiny" (Finding of Fact 2A). Although several arbs had already voted to desysop, these actions appeared to galvanize the final few votes and 2A was specifically cited by at least one.

Community discussion during the case was robust with over a dozen editors (not including the committee or named parties) providing input. Discussion covered not just conduct specific to the case, but broader questions of community response to paid editing in general, the onus on admins and other trusted roles to provide transparent and well-separated actions, and the roles of various reporting and enforcement mechanisms such as noticeboards. Process vulnerabilities such as and sockpuppet investigations were concerns for some.

The case was closed without new major policy. Two items concerning paid editing were fine-tuned in the committee's decision (emphasis that of the Signpost, not the committee):
 * 1) Paid editing is defined as "an edit made, or an on-wiki action taken, by an editor in return for payment to or for the benefit of that editor".
 * 2) Conflict of interest extends to "any article or subject that [a] firm has been retained to edit, even if they were not directly paid to take action in relation to that specific article or subject".

It was left to the community at two open RfCs to decide whether administrators or others with advanced rights may use their privileged toolset when engaged in disclosed paid editing ( VPP RfC and WT:ADMIN RfC ). – B

Invited op-ed

 * By Jytdog

A lot of people thought this was a case about the principle of whether it is OK for admins to edit for pay, commercially, but it wasn't. It was really, really local – just about how Salvidrim! conducted himself and what the consequences would be. It involved Soetermans a bit but there was no real controversy there.

The most surprising thing in the case to me, is the 2nd stated principle, which I will quote here: "2) Because Wikipedia is intended to be written from a neutral point of view, it is necessary that conflicts of interest are properly disclosed, and articles or edits by conflicted editors are reasonably available for review by others. Editors are expected to comply with both the purpose and intent of the applicable policies, as well as their literal wording."  You won't find that in writing anywhere else in Wikipedia, but I believe it expresses the living consensus in Wikipedia, and it is present in every other publishing institution that takes its responsibilities to readers seriously.  We as an editing community are living our way into what that means concretely.

The key findings of fact were about Salvidrim!'s poor judgment in avoiding community review. In the surface of the decision it is clear that this was about avoidance of review of conflicted edits. In discussion of the desysop decision, you will see that this was also about avoidance of review, in the sense of Salvidrim! not going to RfA himself to assess the level of community trust in him; the fact that we were having an Arbcom case at all, became part of the issue. Salvidrim! was transparent about his paid editing, and he was very forthcoming in providing private evidence to Arbcom, and everybody appreciated both.

The desysop was not a happy outcome but in my view was appropriate. I think even those opposing the desysop, wished that Salv had put his admin status up for community review after all this broke, and before the Arbcom case was filed.
 * Views expressed in this op-ed are not necessarily shared by the Signpost; responses and critical commentary are invited in the comments. (Want to write an op-ed of your own? See our submissions page.)

New cases
No new cases were accepted by the committee. A request titled "Michael Moates White House press corps" was December 22. – B

Discretionary sanctions procedure updated
ArbCom enacted a change to the discretionary sanctions procedure, adding new requirements for administrators when imposing page-level restrictions. – B