A minefield at the best of times, Requests for Adminship are where Wikipedia gets the most personal, the most vindictive and the most stubborn. To be sure, Adminship should be "no big deal", as Jimbo says. To be sure, what a user did many months if not years before should rarely be an issue. Indeed, barely anything on its own should cause anyone to automatically vote "oppose".

But we do not live in a perfect Wiki-world, and we are not all entirely rational automata. And so RfA is in the state it is in. People bickering about long-forgotten arguments, candidates sometimes being judged more on the amount of edit summaries they fill out than the quality of their recent dispute-resolving behaviour. RfAs can quickly become battlegrounds, with two opposing camps of editors with diametrically-opposed views of a candidate's suitability. Often, the opposers' reasons border on offensive. To me, this represents what Wikipedia is not about. It seems to violate three key policies that sit at the heart of our beliefs: assume good faith, be civil and no personal attacks.

So what are my criteria for RfAs? And how can we move forward so they become less about personalities and more about suitability for Adminship?

My criteria are pretty simple, revolving around the sole premise that users with some experience, and with no pattern of evidence to suggest they will abuse it, should be given Adminship if they are nominated for it.

  • A reasonable number of edits, spread across the four key namespaces: Article, Article Talk, Project and User Talk. I would say 500 or so is sufficient
  • A reasonable amount of time on the project, which I would say is three months or so
  • A decent period in which no or very few disruptive activities have taken place, which I would also classify as three months; as such, a good editor should be able to achieve Adminship after three months on the project

For self-nominations, I would up the ante a little:

  • 1000 edits
  • 6 months on the project
  • 6 months of good behaviour

Moving forward is going to be hard. As I said before, people are human, and we all bear grudges. In my opinion, though, three things would help the RfA process enormously:

  1. No automatic enfranchisement, meaning that you need a certain number of edits and a set amount of time on the project in order to get a vote in RfAs. I would advocate 100 edits and 1 month with a registered account. Of course, anyone would be allowed to leave a comment about the RfA in question
  2. Disregarding oppose votes with no reasoning, although efforts should be made to contact the user via their Talk page and getting them to add a reason
  3. Changing the ethos of RfAs away from it being a vote on a user's popularity and towards it being a place where, barring serious concerns, a user should be granted Adminship rights. This is in line with the "no big deal" idea put forward by Jimbo, and accepted as policy