Wikipedia:Wikipedia Signpost/2007-04-23/RFA reform

RFA reform

Efforts to reform Requests for Adminship spark animated discussion

In the last two weeks, several discussions and experiments to reform the requests for adminship process have occurred. In the standard RFA process, users discuss a candidacy by commenting in three separate sections: "Support", "Oppose", and "Neutral." Bureaucrats usually promote candidates with more than 80% support and reject candidates with less than 75% support. Many users have objected to this format because it resembles a vote, and Wikipedia is not a democracy.

Temporary removal of vote tallies

To mitigate the appearance of voting, Thebainer removed the vote tally from the RFA subpages. (For example, the vote tally of 46/0/0 at Thebainer's own RFA indicates 46 unanimous votes to support.) He explained: "The key reason why the tallies at the top are a bad idea is that, as the first thing anyone sees when they come to an RfA, they will undoubtedly have a psychological impact on that person when they come to voice their opinion, conscious or no."^[1] He added that users could still find the vote tally manually or by looking up the bureaucrats' noticeboard. The tallies have since been restored.

Experimental RFA formats

Moralis

In a bolder move, Durin reformatted two RFAs with permission of the candidates involved. First, he removed the "Support", "Oppose" and "Neutral" sections from Wikipedia:Requests for adminship/Moralis shortly after the discussion opened. Instead, users wrote their comments roughly in chronological order, similar to the standard procedure in deletion debates. This resulted in a lengthy hodgepodge of various opinions.

Moralis's RFA failed with no consensus; supporting and opposing comments were roughly even in number. Much of the opposition focused on lack of experience because Moralis had made fewer than 1,000 edits.

In closing the discussion, Warofdreams explained:

In general, I found that the new format wasn't particularly helpful in determining consensus. On the plus side, it did encourage users to explain their reasons for support or objection, but on the negative side, it led to a large number of comments mixed throughout the discussion repeating issues which had appeared earlier. This made it far more time-consuming than an ordinary RfA to determine which issues had been raised and how many users felt that these were serious concerns.^[2]

Matt Britt

Second, and more drastically, Durin created a new style of RFA for Wikipedia:Requests for adminship/Matt Britt, modeled after the requests for comment process. Users could not vote on Matt Britt's candidacy as a whole. Instead, separate discussions were created to focus on individual aspects of the candidacy. Some propositions were uncontroversial ("Nominee appears to be trustworthy", "Demonstrated need for tools"). However, the candidate's level of experience with deletions inspired no fewer than four sections of discussion ("Nominee has no experience with XFDs", "Nominee has little experience with XFDs", "Nominee has little experience with XFDs but this doesn't seem to be much of a problem", "Nominee's experience with XFD/DRV is not sufficient to promote"). Two other sections were devoted to Matt Britt's usage of edit summaries.

The unorthodox format generated much consternation within the community. Many users supported the proposition that "This method of RFA is so confusing that I am unable to participate." For example, Black Falcon complained, "It's impossible for me to participate within this format in a manner that will actually reflect my view of the candidate. If I were to try, it'd come out to endorsing 5 views and opposing 5 others. Yet, my comments would be a product of the way the views were worded rather than my actual opinion on the candidate." GRBerry added, "While not prevented from participating, the format makes it much more difficult to participate. It also makes it much harder to determine whether the community believes that the candidate should be promoted. As an experiment, it has proven that this format is a really really bad idea, and should be discarded from further consideration." Feydey quipped, "When a simple Rfa gets to the size of a small novel I know which I prefer to read. And good luck to the closing bureaucrat."^[3]

Despite the criticism, Durin defended the need to experiment with new methods. He wrote,

...come on, it's just an experiment. There's no damage to the project being caused by this. For far too long, RfA has wallowed in a complete inability to come to any agreement on whether RfA is broken or not, whether we should reform it or not, what that reform would entail if anything...The amount of discourse on these subjects could fill a small library. It's really rather absurd.

So, I got tired of hashing these endless debates out. Instead, I decided to DO something about it and actually try doing something different for a change. Unless someone can show me how these experiments constitute some threat to the project (especially a threat that's worse than the normal, already damaging RfA formats), then I intend on trying others if I can find willing guinea pigs.

In the very least, you have to acknowledge these experiments are fostering a considerable amount of discussion on actual attempts at reform rather than theoretical notions of how something might work.^[4]

Kim Bruning humorously supported Durin's controversial experiment. He wrote, "RFA can use some spring cleaning. And besides, isn't this fun? :-)"^[5]

Dan declined to promote Matt Britt. He explained:

I have closed Matt Britt's request as unsuccessful because the community appears far from convinced of the validity of this format. I am sorry to Matt, who I suggest should run again under the regular format until such a time as consensus favors the new format -- which, while intriguing, certainly has its issues. My mandate as a bureaucrat permits me to promote administrators under very specific circumstances; with so many users objecting to the very premise of this request, I cannot in good faith promote Matt. I have, however, studied the request in an attempt to reason out how I would evaluate the request, were the format accepted as valid; I believe I would have promoted him. If this format, or some revised version thereof, ever gains the community's acceptance (and I think it has much promise), I will start dialogue among the bureaucrats to decide on a standard way to evaluate these requests.^[6]

Centralized discussion

In a related development, a centralized discussion on RFA reform has garnered 23 proposals and numerous comments. This follows after a survey on adminship from February 2007. The discussions are currently active, and none of the proposals has reached consensus yet.