Wikipedia:Bureaucrat discussion

This is an information page. It is not an encyclopedic article, nor one of Wikipedia's policies or guidelines; rather, its purpose is to explain certain aspects of Wikipedia's norms, customs, technicalities, or practices. It may reflect differing levels of consensus and vetting.

Shortcut WP:CRATCHAT

This page in a nutshell: Bureaucrats have the option of holding a discussion with other bureaucrats when they are unsure as to how to interpret consensus at an RfX discussion.

A bureaucrat discussion or 'crat chat is a procedure by which bureaucrats on Wikipedia can discuss whether consensus is present at a particular request for adminship (RfA) or request for bureaucratship (RfB). They can also be held for resysop requests if there is a doubt concerning the suitability for restoration of admin permissions.

Bureaucrat discussions should not be confused with the regular discussions, involving bureaucrats, at the bureaucrats' noticeboard and on other pages.

Bureaucrat discussions are only held when a bureaucrat is unsure whether consensus has been reached in an RFA or RFB. Typically, these are borderline cases, or ones in which unusual circumstances apply.

Before restoring the administrator flag, a bureaucrat should be reasonably convinced that the user has returned to activity or intends to return to activity as an editor. Should there be doubt concerning the suitability for restoration of admin permissions, the restoration shall be delayed until sufficient discussion has occurred and a consensus established through a bureaucrat chat.

Holding a bureaucrat discussion ensures that the decision making process in such cases is transparent. However, as bureaucrat discussions delay a decision and can take considerable time on the part of bureaucrats, they are used only as a last resort, after measures such as extending the RFA or RFB have been considered. They are not used for all controversial cases; if the closing bureaucrat can determine consensus alone, there is no need for a bureaucrat discussion.

A bureaucrat discussion is started by one bureaucrat placing an RFA or RFB on hold. This ends the discussion on the RFA or RFB and prevents any more comments in support or opposition being added.

The bureaucrat then creates a subpage of the RFA or RFB's page, on which to hold the bureaucrat discussion. They start a discussion by giving an explanation of why they feel unable to determine consensus, and highlight some areas which the discussion may need to address. The discussion is linked from the RFA or RFB. Finally, the bureaucrat informs other active bureaucrats of the discussion, both on their user talk pages and on the bureaucrats' noticeboard.

To keep the discussion focused, it is open only to bureaucrats (other editors may comment on the associated talk page). Bureaucrats who have taken a position on the RFA or RFB itself are expected to state this and recuse themselves from the discussion. The talk page of the bureaucrat discussion is available for all users to comment on the ongoing discussion, and bureaucrats will often respond directly to comments made there.

Past experience has shown that bureaucrat discussions usually work best when they operate over a short time frame. Once several bureaucrats have participated, if agreement arises, the RFA or RFB is closed as usual. If no agreement among the bureaucrats emerges after one to two days, a common proposal will be to close the RFA or RFB as "no consensus", given that as a group, the bureaucrats cannot determine consensus.

RfA's discretionary range has been set from 65% to 75% since 2016, and was de facto 70% to 75% before that.

• 3 bureaucrat discussions have been held for RfAs below this range.
  • 2 ended in promotion.
  • 1 found no consensus.
• 21 have been held for RfAs within this range.
  • 7 ended in promotion.
  • 14 found no consensus.
• 3 have been held for RfAs that were above the range. All ended in promotion.

The lowest passing percentages are 64.1% post-2016 and 68.4% pre-2016. The highest failing percentage is 73.9%

RfB had a discretionary range of 85% to 90% until 2010, when it was modified to a passing mark of "somewhere around 85%". All four RfB bureaucrat discussions happened before that change:

• 3 were for RfBs below the range.
  • 1 resulted in promotion.
  • 2 found no consensus.
• 1 was within the range. It had no consensus.

The lowest passing percentage is 82.7%. The highest failing percentage is 85.9%.

12 administrators and 1 bureaucrat have been promoted as the result of a bureaucrat discussion.

• 7 admins remain admins.
• 1 admin resigned uncontroversially.
• 2 admins were desysopped for inactivity.
• 1 admin and 1 bureaucrat had their tools removed for cause.
• 1 admin resigned while facing desysop.

Out of 14 RfAs and 3 RfBs where a bureaucrat discussion found no consensus, 5 would-be administrators and 1 would-be bureaucrat were promoted as a result of a subsequent RfX.

Only one user has been the subject of two bureaucrat discussions.