Steve Bennett wrote:
On 3/31/06, Ilmari Karonen nospam@vyznev.net wrote:
Well, no, the part about going on a vandalism spree was hyperbole. It's more about protecting ourselves against users who might use their admin povers for more subtle undesirable things, such as POV pushing, or for ends incompatible with the project (like the folks who think the best thing about Wikipedia are the userboxes), or who might simply use them carelessly or thoughtlessly, say, by rangeblocking all of Europe.
And we are going to detect such users by ensuring we only select admins who use edit summaries 95% of the time and have made at least 1500 distinct edits in the minimum 6 months they have had an account at en?
Not really, though there probably is _some_ correlation. For example, edit summaries exist for explaining one's actions, something that is desirable for admins in general. People who assume that others will know why they did what they did without an explanation may often not make very good admins.
Similarly, someone whose only edits are to user pages and templates is obviously not likely to be a viable admin candidate. Whereas someone who's only been around for a month is simply a wild card: we know they're sensible now, but will they continue to act sensibly when they get caught in an edit war, or someone insults their mother, or they go off their meds? Waiting three, or six, or twelve months improves the odds.
Of course, voting solely on the basis of numeric metrics like edit counts and account age is not only silly, but also easily gamed. Fortunately few people do so, except as minimum criteria to filter out the most unlikely candiates, and in any case the fact that everyone uses different criteria tends to balance things out.
I think I'd rather that each RfA required a neutral person to review the person's entire edit history, noting the number of edit wars, whether edit summaries were accurate, their apparent stance on controversial topics like userboxes etc, then publishing those facts for everyone to decide on. Rather than (incorrectly) assuming that each person voting does such a review for themselves.
Now there's a suggestion I would certainly support, if we could only figure out who this poor fellow would be. For non-selfnoms we might consider putting the obligation to provide detailed information about the nominee on the nominator; while not impartial, they are at least likely to be somewhat less biased than the nominee.
Of course, the way it currently works is that people provide reasons for their votes, and others may comment on them. While this system is far from perfect, and tends to weight memorable isolated incidents over long-term trends, it still generally results in at least some amount of public background review being done.