Tuesday, 13 January 2009

Template RFA opposes

There's a lot of people who like voting in RFAs. Contrary to popular opinion, I have never been one to vote in every single RFA up there. I, like many, only tend to comment on people who I am familiar with, or who I have seen around. I also started extending support to candidates whose RFAs were getting tanked by opposers with poor rationales.

There's a particular poor rationale I can't stand and that's the old "lack of project space edits". I'm not entirely sure how Wikipedia: space matters in the grand scheme of things. We could run our project well enough without the entire project space for certain. This is fairly old news, but to those who are unfamiliar, Stifle, a Wikipedia admin and OTRS volunteer, is a regular RFA voter. What's wrong with that, you may ask. It was brought to my attention some time ago, that Stifle uses an algorithm (as can be seen on his userpage) for voting on RFAs. He has just opposed Kanonkas with a templated rationale, that he often uses (Lack of Wikipedia: space edits indicates likely lack of knowledge policy). There are several things wrong with that. First, it's an oppose that is uncertain. Stifle has no idea of the knowledge of the user, as indicated by the word "likely". He is simply assuming that because a user has fewer than 500 edits in the Wikipedia: space, they must automatically be stupid and completely clueless about how Wikipedia works. Second, knowledge policy is gained not by making tons of "per nom" edits in Wikipedia: space. It's gained by writing the encyclopedia, something that Stifle does very little of. It's also gained by reading the policy pages, something that cannot be proved the candidate did, even through the use of questions. Even as an admin myself in later days, I still had to check the policy page just in case when I wasn't sure. I'm certain many admins do the same. It's unreasonable to expect admins to know all 40+ policies by heart.

I'm sure Stifle is a lovely person, but in my opinion, his ideas and approach to requests for adminship is all wrong. Candidates need reviewing properly, not just judged by their edit count in one namespace.


Whiteknight said...

Let me ask you a question about WP, since I really don't know how things work over there: What part do the bureaucrats play in this whole process? Are they treated like mindless automatons who must obey the final bean count, or are they empowered to interpret the results?

At Wikibooks the process is a little difference because the final promotion decision is made by the bureaucrats. They could look at templated votes and say "this vote is meaningless and shouldn't be counted". We've had entire RFAs shut down and restarted because too many people were voting based on criteria that wasn't wasn't considered to be "on-topic".

On Wikibooks, the only real question in an RFA is whether or not the person in question is trusted to use the tools they're requesting for the overall good of the project. A vote like "Oppose. Has fewer then 500 project namespace votes" is absolutely meaningless to the question at hand and will be ignored.

Are the bureaucrats at WP intelligent and professional enough to discount the votes that aren't well-supported and pay more attention to the votes that are thoughtful and relevant?

The success of the RFA system is dependent on the quality of the bureaucrats who implement them. If the system is broken, they are the people who are responsible to fix it.

Majorly said...

The latter in theory, but the former in practise.

Wikibooks sounds like a wonderful place for RFAs. I don't think restarting an RFA on enwiki would be a wonderful idea - people will oppose simply for the fact it restarted, I assure you.

I'm active in a few places other than enwiki, but the RFA processes are all fairly similar - meet a certain percentage, and you pass, fail to meet it and you fail. On Meta, I became a bureaucrat, and put forward the idea that every admin should be granted the right without another process. It has worked well for the most part. This is because every admin on Meta is trusted by everyone else. We don't have a lot of RFAs, but the majority pass unopposed. This high standard of passing, plus the fact Meta admins are all trusted elsewhere anyway, made it easy to persuade the community to agree. We did add an ability to the bureaucrat role that other wikis oddly lack, and that's ability to desysop.

I think the RFA system on Meta, while it is still a vote, is a fairly good one because it generally passes people with complete community confidence. On enwiki, this is just impossible, unless you get thousands of voters. Nor is every admin on enwiki trusted, which is pretty stupid considering they're admins. I know Wikibooks has a desysop policy of some sort. Meta does too - inactives are removed. Again this would never work on enwiki. Not because it would be impossible to do, but because the community has said no so many times.

On enwiki, certain opposes are ignored. On the particular RFA I refer to, Stifle is the only opposer (which perhaps is an indication of how at-odds he is with the community). It would be pointless ignoring it because it would make no effect on the result. However, when an RFA is close, I do think the bureaucrats on enwiki do read through. But it's exceedingly rare for them to ignore the magic 75% boundary, even if the opposes are all fairly petty.

I disagree that it's up to the bureaucrats to fix the system. The system is fine, I think. As I was writing this, I was thinking to myself how lovely it would be if only unopposed RFAs passed, which is a clear indicator of community confidence, like on Meta. But that would discount so many fine admins, it would simply not work. Anyhow, bureaucrats are merely the ones to decide and press the button at the end. It's the community that decides how it's run. Currently, there are too many people trying to ruin RFAs with protest votes that do nothing (as Stifle did), or piling on over one minor issue (as on Enigmaman's RFA).

Giggy said...

It sounds like 'crats on wikibooks have a lot more balls than on wikipedia, and that's part of the problem.

ChrisiPK said...

Are templated votes based on hard criteria really that seldom on the English Wikipedia? We have them all the time on the German Wikipedia, many users create a special subpage just to show their criteria around (see http://www.google.com/search?q=wahlkriterien+site%3Ade.wikipedia.org) and also only put templated reasons on their votes, not even saying why they oppose or support. Most of them look like "per my criteria" and include a link to the subpage.

These are valid votes on the German Wikipedia. Bureaucrats usually don't discuss the outcome of an RfA, but just check the suffrage of the voters. If a 2/3 majority is reached, the nominee is promoted.

I think it would be much better to leave bureaucrats more freedom for their decision. As it is now, counting votes can be done by anyone, that is not what we need bureaucrats for. Maybe this will change in the near future, as new bureaucrat elections will start shortly, which will hopefully increase the number of active bureaucrats (currently 2) up to 5 or above and thus make it possible to actually assign them some tasks other than counting.

Majorly said...

No, they are fairly common. People have their criteria pages. I'm sure I used to have such a page once, but deleted when I came to realise we should be evaluated people on their own merits rather than some arbitrary criteria.

It's very rare for a bureaucrat to promote someone with less than 75% support, and when they do there is nearly always uproar. It would be good if bureaucrats had more freedom. There are far too many frivolous opposes these days. Supports need not be so details, as all you're saying is you agree with the nomination. There's not much you can add. Opposes however are trying to prevent something, so should provide solid reasoning.