Wikipedia:Proposal for mentor–learner administrator selection process

I would like to propose a new administrator selection process. The current system has failed miserably, due, in my opinion, to voting, voting openness to every single person regardless of whether or not they are familiar with candidates, and failure to represent the best interests of the community and the real meaning of administratorship.

At this time, administrator candidates must undergo an extremely rigorous process of checking, questioning, interrogation, and voting by community members who, in many cases, do not even know the candidate. Can this system really be called fair? Why should people enforce their own edit count standards for candidates when they have no knowledge of the candidate? Why should people even be permitted to vote on a candidate unless they are familiar with the person in question?

In addition, administratorship is seen by many as an end (administratorship) to a means (getting administratorship), not a means (administratorship) to an end (building a reliable encyclopedia); this is an issue that is solvable only by education, not by high edit counts.

All in all, requests for administratorship (RfA) is, among other things:


 * a bureaucratic mess,
 * a system that fails to carry out constructive discussion and interpret consensus,
 * a system that expects administrators to comply with process, policy, and polls more than with the Wikimedia Foundation's vision and community consensus,
 * a system that treats administratorship as a final position as opposed to a learning process,
 * a standards-based system that stops more good or badly-educated candidates than it does ill-motivated candidates, and
 * a system that can easily be gamed through allowing people to get revenge on candidates they don't like or don't agree with.

I believe that a far better and more intimate method of selecting administrators would be to have a mentor–learner process, where an experienced administrator could act as a mentor to an administrator candidate, and, eventually, nominate this candidate for consideration on RfA. I suggest that voting on RfAs be abolished, and replaced with pure discussion and natural consensus. Finally, I suggest community members only be permitted to discuss requests made by candidates they are familiar with.

Thus,


 * experienced administrator mentors teach non-administrators,
 * these mentors eventually nominate their respective learners for administratorship (self-nominations are not permitted),
 * voting is abolished and replaced with pure discussion and natural consensus,
 * community members are permitted to discuss requests made by candidates only if they are familiar with the candidates in question, and
 * bureaucrats interpret discussion and consensus, not votes, and act accordingly.

How can such a system be worse than what we already have – in fact, how can it not be far better? Relationships are built. There is no reason why mentoring, official or unofficial, cannot continue after a candidate is accepted or rejected for administratorship, so administrative learning is a continuous process rather than a once-off, get-adminstratorship-and-go experience. I invite your discussion.

(The general idea of having experienced administrator mentors was inspired by User:Durova's comments in Episode 15 of NotTheWikipediaWeekly (May 9, 2008) about increasing mentorship in RfA. – Thomas H. Larsen 23:54, 23 May 2008 (UTC)