User:Maxim/ArbCom and desysops

This page presents two tables on all ADMINCOND cases, as well as other cases where specific administrators were sanctioned, from 2010 to present. The purpose of compiling this data was to assess the likelihood of an admin "getting off" when the subject of an accepted ADMINCOND case. During the review, it became useful to additionally note other cases, typically broader in scope, that have resulted in a specific administrator being sanctioned. A full survey of Arbitration/Index/Cases was performed from 2010 to present. The cut-off year of 2010, while slightly arbitrary, was used because it is when the committee could be said to have "matured". At this point, it started to take on significantly less cases per year (30 in 2009, 14 in 2010), and to not take "easier" cases that could be more easily handled by the community. A cut-off also helps compare like to like with respect to expected standards of administrator conduct, which have changed with time.

Some preliminary findings as follows:
 * The Arbitration Committee, on average, hears 3 ADMINCOND cases per year.
 * Since 2012–2013, there has been a precipitous decline in administrators being sanctioned during cases with a broader scope.
 * The chances of an administrator avoiding a loss of permissions as a central party to a case are very poor. While 5 of the 41 listed cases (~12%) did not end up in a desysoping or resignation, consideration must be made for some of these cases that weren't quite purely ADMINCOND cases (e.g. Magioladitis, Gamaliel and others) or where the final decision suggests that perhaps the case ought not have been accepted (Michael Hardy). As far as cases that most fit the mold of having a scope to "examine an administrator's conduct with regards to ADMINCOND", GiantSnowman is the only strong example of administrator not being desysoped as a result of such a case.

And some further thoughts (19 November 2021):
 * The idea to tabulate this data probably came out of some combination of an arbcom-en thread which was sort of a post-mortem on the RexxS case and an Iridescent-talk megathread. While there was a range of topics in those discussions—from case naming to RfA reform—the overall context was the aftermath of the contentious RexxS case.
 * The numbers are just that: numbers. This essay just appears to be the first time someone went and tabulated the outcome of what can be construed as "modern" ADMINCOND cases. I'm not trying to pass any value judgment on the numbers, but to give a population analysis for a given definition of a "modern" ArbCom.
 * I didn't consider declined cases, nor did I consider similar ANI threads, because that wasn't really the point of the table. To be specific, the point was to answer what happens if an administrator is involved in an accepted arbitration case, particularly an ADMINCOND one. Dear reader, you may make of these numbers what you will.

Further thoughts, 5 June 2023:
 * In the context of a new requests for arbitration along the "themes" discussed on this page, I note that there have been a couple of recent cases that were resolved by a desysop motion, namely Dbachmann (2023) and Athaenara (2022).
 * Within the 2010–present time frame, ADMINCOND-style or similar cases that were resolved by motions more likely ended up with warnings or admonishments, namely Timwi (2022), Future Perfect at Sunrise (2018), Kww (2014), Kevin Gorman (2014), Kevin (2013), Hex (2013), Nabla (2011), Trusilver (2010), and this 2010 mess. There is also a case dismissal by motion: YellowMonkey (2010).
 * In contrast to the 10 examples of warnings or admonishments above, there are three additional desysops-by-motion: Dreadstar (2015), EncycloPetey (2012) and Craigy144 (2010).