I was thinking about my recent proposal to require RFA candidates to answer yes or no to questions about their prior or alternate accounts. It occurs to me that the community's opposition is based not solely on the merits of the case, which I believe strongly weigh in favor, but also on the community's perverse incentives in maintaining lax safety standards.

Everyone knows that "anyone can edit" is a very insecure, unsafe system. But there are some safeguards. "Anyone" can't edit the Main Page - only admins can. "Anyone" can't edit the article about Barack Obama - only autoconfirmed users can. Anyone can edit a few hundred thousand Biographies of Living Persons, but I believe that should change.

"Anyone can edit" places a heavy burden on a small number of individuals. The most enthusiastic supporters of "anyone can edit" do not bear the heaviest burden of making sure Wikipedia continues to function while torrents of vandalism flow in constantly. I cannot tell you how many tens of hours I spent fighting vandalism, both directly at Recent Changes Patrol (old-school style, without automated tools) and by performing analysis of user contributions at what is now Sock Puppet Investigations (SPI). I felt strongly while doing this work, free of charge, that I could more productively contribute as a writer and editor, but the administrative side of things desperately needed someone with my skills and motivation to deal with sockpuppet cases that literally sat on a noticeboard for weeks without comment. I discovered only too late that nobody understood the effort I invested in this work, and few people thanked me for it.

What happened this past week on WT:RFA reflects the pattern that supporters of radical openness ignore the potential adverse effects of their foolhardy carelessness. The kibitzers at WT:RFA want to engage in pleasant discussions with candidates. They don't want four lines of Q&A regarding potential sockpuppetry to clog up a discussion about the candidate's strengths and talents. Nobody was quite willing to express it this way, but they don't want to install safety equipment because it doesn't look nice. The red fire extinguisher disrupts the color of the room. If a fire happens, they'll deal with it - but hopefully it won't happen, and meanwhile they don't like the color (or cost) of the fire extinguisher.

Further to the topic of perverse incentives, I was very amused by the endless drama reqarding Betacommand, who recently requested permission to run an archival bot at SPI. Betacommand is forbidden to run bots, but he requested an exception to assist with archiving at one particular high-volume page which requires regular maintenance. I read the discussion, and in line with the majority, I judged the request reasonable under the circumstances, and I would have approved it. However, a vocal minority will never let Betacommand (a.k.a. Greek symbol "delta") anywhere near automated tools. The vehement opposition of "Verbal" is especially amusing.

Someone should tell Verbal that what Betacommand's bot does on SPI (and nowhere else, only that one domain) is none of Verbal's business. He can just go on doing whatever he does elsewhere in the project. We have here the opposite incentive game-theoretical construct as I described previously. Verbal does not profit or lose anything by what Betacommand does with his bot - and yet, Verbal expresses his vehement opposition repeatedly. One can only concluded that Verbal is motivated by factors other than the merits of Betacommand's request, which (under the full circumstances) argue strongly in favor of acceptance. Thus, Verbal and others like him are motivated by power trips - preventing Betacommand from doing what he wants to do, regardless of whether any risk is involved - or by irrational fear that Betacommand will somehow abuse his permission to program the bot to do unauthorized activities. Or maybe they just want to err on the side of strict safety standards. If only they felt the same about sockpuppet candidates at RFA.

Crickets chirping...

Hello?

Petty Wikipedian politics. -> Annex.

Betacommand has a proven record of bad judgement, misuse of bots and poor social skills. Therefore opposition is reasonable.

Asking a silly question that a) 99% of people will be able to answer "no" to and b) 1% will be able to say "yes" to, but probably won't is nothing like a safety guard, at least compared with banning Betacommand from running a bot.