Wikipedia talk:Requests for adminship

Proposed addition to RFA boilerplate

I've proposed adding Please keep criticism constructive and polite to the RFA template. Please comment here. Thank you.--Kchase T 12:31, 22 January 2007 (UTC)

Ryulong's RFA passing at 69.4%!?

Why has Ryulong's RFA been closed as successful at 69.4% support, given the magnitude of concerns of the opposition? – Chacor 06:20, 24 January 2007 (UTC)

I have promoted Ryulong to sysop. While some might object that he was below 70% support (with 69.4%), I believe, based on my interactions with him, that he would make a good sysop, and it would be a mistake not to promote him. Raul654 06:22, 24 January 2007 (UTC)

Your interactions with him DO NOT overrule community opinion, especially when MAJOR concerns have been raised. – Chacor 06:23, 24 January 2007 (UTC)

Besides, whatever happened to bureaucrats having to remain neutral? Your promoting him based on your past interactions with him throws that out the window. – Chacor 06:29, 24 January 2007 (UTC)

• I agree - What is the point of an RfA if the bureaucrats are going to ignore it and promote users anyway? 69.4% is close to 70%, yes, but the opposition was huge.. --Veesicle (Talk) (Contribs) 06:32, 24 January 2007 (UTC)

Er, Raul, quite bluntly, my interactions with him lead me to exactly the opposite conclusion. There are some major civility issues here, and I see that this is going to become a time bomb waiting to explode. Titoxd^(?!?) 06:25, 24 January 2007 (UTC)

Here we go again... Grandmasterka 06:28, 24 January 2007 (UTC)

Before the debate begins, let us remember that we're all friends outside of the ring :) — Deckiller 06:30, 24 January 2007 (UTC)

Oh, we are, but I'm just... surprised. Titoxd^(?!?) 06:31, 24 January 2007 (UTC)

I've asked Ryulong to hold off using the tools until we sort this out. The last time this happened we had a whole bunch of people asking you-know-who to respect consensus by stepping down. There isn't a need for that, but I think Ryulong should be courageous enough to accept consensus and wait for us to sort out this mess. – Chacor 06:35, 24 January 2007 (UTC)

I think that's very fair, especially in this situation. I didn't participate in the RfA (I rarely hang around the RfA room anymore), but I've been skimming it for a couple minutes. I'm not going to take a position yet, however. — Deckiller 06:39, 24 January 2007 (UTC)

Respectfully, Gimme a break. This is why we elect 'crats - to determine consensus. If it were just the numbers we could have a bot do it. The decision has been made.  Glen  06:41, 24 January 2007 (UTC)

But there was no consensus. Raul himself admitted that he only passed him because "based on my interactions with him, that he would make a good sysop". – Chacor 06:42, 24 January 2007 (UTC)

Yeesh. Picky, picky. Ok, if you want to play the Wikilawyer game, fine - consider myself as supporting his request for adminship, and discounting the opposition from the newbie with a dozen edits. That gives a final tally of 126-54, which is 70%, which is in the bureacrat's discretionary range.

As to a point raised above - Chacor's understanding of policy is wrong. If a bureacrat was supposed to be a bean counter, we could hand the job over to bots. By definition (and, for that matter, the 2005 poll which defined the job) the role of a bureaucrat involves using his discretion. Raul654 06:43, 24 January 2007 (UTC)

And, of course, then, you couldn't close the RFA. – Chacor 06:44, 24 January 2007 (UTC)

That is not correct, check the archives and historical practice if you don't just accept me saying it. You chose to start the wikilawyer game with your heading stating 69.4%, and on that front he's got you checkmated. While I appreciate people are being a little more civil this time, they are still failing to put this in proper perspective to the overall goals of the project. The proper perspective would be focusing on what's the most important thing we're here for and going to work on that. I'm continually surprised at how much overimportance we place on internal processes versus content. - Taxman ^Talk 14:33, 24 January 2007 (UTC)

(after ~250 edit conflicts) With all due respect, guys, I hate to say this, but he's the bureaucrat, and we're not. Quite often regular users like myself have to put up with an admin action that we think is lousy and can do nothing about it; but that's the reason they're admins, and they make the decision - there have been some afd decisions that made my blood boil (*cough*), but that's just how life goes. He was elected to this position because people trusted him in the roll of deciding who would make a good admin. It may not be right, from what we see, but he is the bureaucrat. Just like admins have to make controversial decisions at XfD's regarding WP:ILIKEIT material, bureaucrats must do the same thing with RfA's (and for that matter, I'm glad they do, even when I disagree with them, like in the case of Kazfiel). That's my 2 cents. Patstuart^talk|edits 06:43, 24 January 2007 (UTC)

The bureaucrat discretion range is 75-80%. In the RfA archives, they settled on 80%, but some bureaucratic actions created a gray area of 75-80%. I read the archives too. :-) Grandmasterka 06:46, 24 January 2007 (UTC)

Apparently, it has been expanded to 70%-80%. And what happened to adminship not being a big deal? And yes, I'll be expecting an onslaught of "but he might abuse the tools" and "community opinion" responses. --²¹⁰physicq (c) 06:49, 24 January 2007 (UTC)

We don't need more uncivil or newbie-biting admins, two points raised in opposition. Personal interactions doesn't change anything. – Chacor 06:51, 24 January 2007 (UTC)

Based on the Carnildo precedent, this should be seen as a forgone conclusion that such candidates are going to be promoted. Get Taxman here to back it up and the uproar, I think, should quiet down... Hbdragon88 06:45, 24 January 2007 (UTC)

So basically, have the bureaucrats and ArbCom like you, then get at least 50% support in an RFA. Should we change the RFA page to say so, instead of lying about a discretion range? *rolleyes* – Chacor 06:47, 24 January 2007 (UTC)

And Carnildo has turned out to be a fine admin. Which shows exactly why the 'crats have the discretion, and we don't. Patstuart^talk|edits 06:49, 24 January 2007 (UTC)

Well of course, this time Carnildo has not yet been mired in anything hugely explosive. People complained here too, because unlike Ryulong's RFA (I say that 69.4% is actually pretty good, and we're splitting hairs with the supermajority of 70%) Carnildo was at a low 60%. Also, Danny voted for Carnildo and then was involved in the discussion of whether to promote Carnildo again, also raising outcries of personal involvement (similar thing being thrown here, with Raul's reasoning based in part off of "personal interactions"). Hbdragon88 06:55, 24 January 2007 (UTC)

Personally, I feel that were Ryulong is lucky that he is a high profile candidate. I can think of many nominations from 70% up to 76 or even 77% that failed, where the candidate was inundated with negative campaigning or political factional reasons etc, over things much smaller than incivility or overaggression with respect to blocks, and no bureaucrat even took a second look at it. Either people use a bean counter for all, or examine all people carefully, not simply those who are high profile. Blnguyen (bananabucket) 06:52, 24 January 2007 (UTC)

Supposing that all things being equal, 80% is a 4:1 ratio, 75% is a 3:1 ratio and 70% is a 2.333:1 ratio. So that's quite a big difference between 70% and 75%. It irritates me that many users have failed close to 75%, in the face of some thinly disguised opposition, or smaller magnitudes of concern arise, and not get the consideration that Ryulong did. There are only 1-2 grey zone RfAs each week, but all of them are routinely slaughtered on numerical basis, apart from two or three special considerations each year, typically of very famous users. Since it is only 1-2 grey RfAs each week, I think everybody should be treated the same - either thoroughly examine all of them, and promote in a less numerically correlated way - or promote everybody in a numerically correlated way. Blnguyen (bananabucket) 07:00, 24 January 2007 (UTC)

...which might invite the debate on where such a numerical correlation line stops. --²¹⁰physicq (c) 07:02, 24 January 2007 (UTC)

I agree. I support a larger discretion range for bureaucrats, but bureaucrats should consider using their discretion on all 70%+ candidates (many of which are closed as failure "by the numbers" although no significant oppose reasons are presented), not just those with 40k+ edits and their third RfA. Kusma (討論) 06:57, 24 January 2007 (UTC)

I also agree. We show our support/opposition for bureaucrat candidates because we trust/distrust their decision making skills. In the grey area, Bureaucrats should be given a little leeway if they can support their moves and if they do it for everyone (for instance, if any candidate has a rate in between the grey area, the b-crat should explain why the user has been promoted/denied). — Deckiller 07:03, 24 January 2007 (UTC)

I have the utmost respect for Mark. Had he indicated that he determined consensus (despite the numerical vote tally), I'd have taken him at his word. Instead, he acknowledged that he promoted Ryulong because of his personal opinion "that he would make a good sysop." He might be right, but the closing bureaucrat's job is to interpret and implement the community's will, not to substitute his/her own. —David Levy 07:14, 24 January 2007 (UTC)

I agree that Mark may have not given sufficient (or appropriate) reasons in this case, but I can't say if it was a good or a bad promotion on my end (mainly because I did not participate in the RfA). — Deckiller 07:21, 24 January 2007 (UTC)

Likewise, I have no opinion as to whether or not this was the correct outcome. It's how it was reached that troubles me. —David Levy 07:29, 24 January 2007 (UTC)

Indeed. Wouldn't it be somewhat awkward, however, to ask him to come up with reasons after this? We'll probably just have to take this incident as a "lesson learned", that b-crats must provide a list of neutral (as in no personal preferences) reasons on this talkpage for grey area promotions. — Deckiller 07:36, 24 January 2007 (UTC)

(quadruple edit conflict) RfA consensus is skewed anyway. There is no way that we can actually determine community consensus using RfA. The reason is simply that (I will not name names, or aim this at anyone in particular) a great number of those who participate regularly in requests for adminship do not understand what is required of an admin (or make choices based on political, rather than pragmatic motives), and a great number of those who do are turned away by much of the drama (like this, for example) that surrounds it. This is nobody's fault. I think that one of the reasons why many people avoid RfA in general is simply because it is inherently unproductive, and there are far better ways to contribute to Wikipedia. This issue is by no means limited to RfA. I'm sure there are similar issues in AfD as well. For these reasons, I think that a 5-10% difference in RfA consensus has a rather low correlation to actual community trust in the user. — Werdna talk 06:58, 24 January 2007 (UTC)

You can't be serious. 185 people participated and the candidate couldn't even hit 70% when the page talks about 75% as a minimum. Seems to me that consensus for promotion was lacking. More participation would not have changed that. --JJay 07:07, 24 January 2007 (UTC)

True, with one comment. If editors are turned away by drama, then they aren't really suited to be admins, because being an admin means getting yourself mired in drama (as is the case here). --²¹⁰physicq (c) 07:01, 24 January 2007 (UTC)

I think you missed me there. I was saying that people who would be good voters are turned away (either because it's unproductive and a waste of time, or because it's full of drama). — Werdna talk 07:03, 24 January 2007 (UTC)

Oh I see. Sorry for the mixup. Never mind my confusion. --²¹⁰physicq (c) 07:04, 24 January 2007 (UTC)

I feel at the very least the role or lack thereof of numbers in RfA needs to be better defined. If we bother to include a counter at the top tallying support/oppose/neutral and have users add their comments with #s, we imply that vote-counting is a part of the process, if not the be-all and the end-all. Comments at Deskana's RfB also demonstrate that people feel that vote-counting does play a role, yea, even the most important role. If indeed it's pure consensus that counts, with vote-counting not even playing a substantial role, RfAs should be formatted completely differently. The process here bears very little resemblance to what's discussed at WP:CONSENSUS.

As a second point, I note there doesn't really seem to be any accountability for bureaucrats' decisions here. I don't think that's a good. There needs to exist the possibility of such decisions being challenged, and not just through discussion on a talk page that ultimately leads nowhere. (Please note that this is a general concern, and that I am not taking sides on this promotion. I did not contribute at all to Ryulong's RfA and, while I'm skeptical, I still remain neutral on his promotion.) Heimstern Läufer 07:05, 24 January 2007 (UTC)

I would assume that Bureaucrats are as accountable to arbcom as the next user is. --tjstrf talk 07:09, 24 January 2007 (UTC)

Is the ArbCom willing to consider cases like this, including being prepared to undo promotions and censure bureaucrats if necessary (again, I am not asserting that this is the case here)? I was under the impression they were reluctant to do that; please correct if I'm wrong. Heimstern Läufer 07:16, 24 January 2007 (UTC)

Likewise. I don't think the ArbCom dare to set such a precedent. – Chacor 07:17, 24 January 2007 (UTC)

That was rather my point actually: Bureaucrats are supposed to be the epitome of trusted users, we don't get to just challenge their decisions on a whim. If you wish to challenge his closing, then gather evidence and convince the arbcom that it is worth the time and precedent. If it makes you feel better, you can consider that extra .6% subtracted from Raul's 96.15% 'cratship. --tjstrf talk 07:25, 24 January 2007 (UTC)

I wonder, then, if we've made a position that requires too much trust in humans. At any rate, though, I'm certainly not going to seek a challenge to this decision. As I've said, I'm neutral. These are just ideas I've had for a while, and this seemed like a time to speak them. I think it's probably time for me to just quiet down now, so I will. Thanks for listening, and I do mean that sincerely. Heimstern Läufer 07:36, 24 January 2007 (UTC)

The incredible thing here is that this closing was done absolutely without regard to Wikipedia process. This is exactly the kind of thing many of us fear when forming an opinion on any kind of promotion request. The reasoning given (above) was, quote "I believe, based on my interactions with him, that he would make a good sysop, and it would be a mistake not to promote him." Straight from the horses mouth, as they say, this reasoning doesn't even close to touch what right to discretion bureaucrats actually are given by the community: discretion to determine consensus. Apparently, Raul654 took that to mean discretion to determine, completely on his own, whether or not Ryulong is fit for sysophood. This is not to mention that Raul took the unusual step of acting in Bureaucrat duties: He had only performed 6 such promotions in the last year prior to Ryulong. -- Renesis (talk) 07:42, 24 January 2007 (UTC)

Interesting. While you think he disregarded every opinion and went his way, I think he considered to be in the "bureaucrat margin" (you know, the range for which we trust humans to decide and not a bot) and weighted his personal opinion on the user. You know, the same personal opinion an administrator uses to close an AFD when he knows about the topic of the article. I like when people "take risks". Wikipedia would not exist if Jimbo did not take one, would it? Or Linux. Or Firefox. Just like in those cases, you can only do one thing: wait and see. -- ReyBrujo 08:08, 24 January 2007 (UTC)

That's a strawman - I never said and never would say that I don't support "risks". And bureaucrats are intended to use the discretion range to weigh the opinions given "you know, like in AFD", not to base it on his own. -- Renesis (talk) 08:03, 24 January 2007 (UTC)

There is where we won't agree. As I stated below, they are not bots, they can take their own decisions. And they can weight their own opinion if they are in such "range". Or you think the closing 'crats at Carnildo's RFA did not? -- ReyBrujo 08:08, 24 January 2007 (UTC)

Of course they can include their own opinion; I am fine with that, but nothing about "based on my interactions with him" seems to indicate that it weighed any other of 55 editors' interactions with him, and certainly not all. -- Renesis (talk) 08:16, 24 January 2007 (UTC)

I think there are cases when discretion should be used in the 'grey area'. However, the difference between 70% and 75% is actually bigger than say, 50% and 55% (using ratios). I do not think that the discretion was used in the right way, though, and this should not have been an exception to the '75-80%' rule. A number of people (including me) have switched their vote from support, an indication that the candidate can't be trusted as much as before. I have nothing against Ryulong, although the indef block on an IP with lack of evidence is highly questionable. Insanephantom (my Editor Review) 08:46, 24 January 2007 (UTC)

The IP block is not consequential in any case right now. Please strike that bit out? – Chacor 08:48, 24 January 2007 (UTC)

• Just to add my belated two penn'orth: I agree with those who find this business depressing. Some of the (verifiable and verified) accounts of Ryulong's behaviour should ensure that he not be given admin tools; community opinion might overturn that (though it's worrying that it would), but that the RfA be closed in this fashion, on the basis of one person's personal experience, is unacceptable. (Kim Bruning's extensive and rambling defences of Raul654 serve, if anything, to push me further towards suspicion and outrage.) --Mel Etitis (Μελ Ετητης) 09:33, 24 January 2007 (UTC)

Usage of tools by Ryulong

As this discussion is not related to the way RFA works, I have closed it. Please use the proper dispute resolution channels instead. Thank you, Kusma (討論) 07:23, 24 January 2007 (UTC)

"Margin of discretion"

The above is in quotes because it's obviously a loose term... but does anybody else realize that (per JJay's analysis above) that the candidate would have needed to have 42 more supports to be above the 75% "discretion range"? -- Renesis (talk) 07:26, 24 January 2007 (UTC)

Right, that shows that 75% is a very high threshold (every oppose vote cancels three support votes). Perhaps it is too high. Kusma (討論) 07:28, 24 January 2007 (UTC)

Good point. Logic dictates that one oppose should cancel two supports for consensus. At least, that's the logic in my mind :) But then again, our debate here isn't about policy changing (we should carry that over to the respective page). — Deckiller 07:30, 24 January 2007 (UTC)

Or perhaps that shows why the candidate shouldn't have been promoted. – Chacor 07:30, 24 January 2007 (UTC)

Hmm, how about we discuss the comments and not the votes - if there are serious objections in the content of the comments, bring them up. Simply saying that because some number wasn't reached that there is no reason to do anything either way is absurd. I think it's a little premature to call for anyone's blood - let Raul post his rationale based on the comments, not some vote number. Honestly, give him some time, he hasn't done anything controversial yet, don't assume that he's going to be a horrible admin, we have a policy for assuming that sort of thing... -- Tawker 07:42, 24 January 2007 (UTC)

There is merit to discussion of both numbers and comments, but one thing we know from Raul654's own comment is that he disregarded all 178 opinions and made the promotion based on his own. -- Renesis (talk) 07:45, 24 January 2007 (UTC)

I doubt that very seriously.--MONGO 07:49, 24 January 2007 (UTC)

Then read his very first comment, above. His promotion was based on his feeling, not the community's. -- Renesis (talk) 07:59, 24 January 2007 (UTC)

Frankly, I find this claim insulting. For one thing, a supermajority of people did support him adminship. Futhermore, before I promoted, I made it a point to read through all the people opposing his adminship. If it had been 65%, would I have promoted him? No. So I'll thank you to stop putting words in my mouth. Raul654 08:43, 24 January 2007 (UTC)

Undent I'm sorry, but the claim that I'm putting words in your mouth is completely ridiculous. You said nothing besides "based on 'my interactions...", and I haven't claimed you've said anything else. And I'd like to know why 65%, but not 68%, 70% or 75%? Why any mention of numbers, if this entire thing is based on "discretion"? -- Renesis (talk) 08:52, 24 January 2007 (UTC)

It's not based *solely* on my interactions with him, and yes, that's the word you are putting in my mouth. I said I promoted him based on my interaction with him. I did not say I did so to the exclusion of the community's commentary. My interaction with him was the deciding factor, but not the only one. Raul654 08:55, 24 January 2007 (UTC)

I don't think interaction with one person is a valid way even for just a deciding factor when it is 40 support votes away from the 75% threshold. Insanephantom (my Editor Review) 09:01, 24 January 2007 (UTC)

The problem, Mark, is that your personal opinion of the candidate should not have been the deciding factor. Bureaucrats' discretion doesn't extend that far. We trust you to interpret the community's will, not to insert your own. —David Levy 09:18, 24 January 2007 (UTC)

Raul seems to think the lower end is 70%, and personally I have seen mentions of both 70 and 75%. – Chacor 07:29, 24 January 2007 (UTC)

People who support a large discretion range regularly get shot down at RfB (which doesn't mean that consensus supports a small discretion range, it just says that more than 10% of people support a small discretion range) so it has been getting smaller recently. Kusma (討論) 07:34, 24 January 2007 (UTC)

• 75% may seem high. Yet, the fact that almost every current admin passed that 75% cap, most without any problem, says something. 55 reasoned oppose votes is considerable and should not have been discounted. We should set a difficult standard for promotion. --JJay 07:38, 24 January 2007 (UTC)
  • A quick glance at the deletion backlogs shows that we need more active administrators. I think both promoting and demoting should be easier than they are right now. Kusma (討論) 07:43, 24 January 2007 (UTC)
    • We need more deletionist admins, not admins in general. About 50 admins do 90% of the deletions, and this is the worst in the pictures area. I did a little calculation in July after I became and admin to see which new admins had the highest deletion rates (Delete/day)- they were Kimchi.sg, Jaranda, myself, AmiDaniel, Crzrussian. I don't know about Daniel, but the others are strongly deletionist, and might explain why they can't sit still when they saw a backlog. Kimchi in particular slaughtered 20k items (mostly pics) in three months. Blnguyen (bananabucket) 08:17, 24 January 2007 (UTC)
• I believe we have over 1,000 admins. An argument that standards must be lowered because of a perceived lack of admins, must have as a corollary an easy mechanism for removing admins. --JJay 07:46, 24 January 2007 (UTC)

  (super ultra edit conflict, reply to JJay comment about setting a difficult standard) Wait, wasn't Wikipedia the place where there was that long "Something is not right..." or so topic saying that only 3% of our editors were admins? Sometimes I wonder how we can tell new users "Adminship is not a big deal" and then erupt like this happens. -- ReyBrujo 07:49, 24 January 2007 (UTC)

• We don't tell new users that adminship is not a big deal. That's a canard. --JJay 07:53, 24 January 2007 (UTC)

• Well, it is slapped here, one of the pages we tell users who want to be admins (being them established members or relatively new users) to read. -- ReyBrujo 08:03, 24 January 2007 (UTC)

  It is my observation that people are considerably more difficult on admin candidates than they were, oh, say 2.5-3 years ago, when most regulars were supposedly admins. Granted, we had more rouge admins by percentage, but stuff still got done (and we came up with some great policies outside of bureacracy, like [[WP:IAR]). I'm not opining either way on the change; but I do believe it's far less of a trivial matter than it used to be. Patstuart^talk|edits 08:08, 24 January 2007 (UTC)

I guess nobody remembers Wikipedia:Requests for adminship/Kafziel 2 either, and how discretion was used to determine "no consensus" because it had been near (but above) the 80% range during most of its time, and only fell below after it was supposed to be closed (down to 77%). -- Renesis (talk) 17:06, 24 January 2007 (UTC)

I remember. And I don't have a problem with, in theory, establishing a wider range of b'crat discretion in closing RfAs. What I have a big problem with is unequal application of that discretionary interval. People have been denied with far higher numerical percentages and (in my opinion) weaker opposes than Sean Black, Carnildo, and now Ryulong. Ryu said he would join administrators open to recall so I'm not too tweaked about the decision here, but many many other candidates have been denied the same privilege Raul exercised here. And that frustrates me more than I can express. -- nae'blis 18:50, 24 January 2007 (UTC)

Reading the guide...

I would like to quote two sentences of the first paragraph:

It is important to understand that consensus within this context is determined by the bureaucrat who is closing the RfA. RfA is not a voting process.

In other words, the closing bureaucrat is the one that "reads" and "understands" the request, considers whether the objections have been sustained enough, and determines the outcome. The people who participates in a RFA knows this. If not, we can begin adding {{Not a ballot}} to the requests. Also...

As a group, they are careful in their decisions. In general, candidates with over 80% support are likely to succeed, and it is unusual for those below 75% to succeed, but all cases are weighed on their merits.

Again, the bureaucrat is given a flexibility similar to the admin closing a deletion discussion.

Now, some comments:

1. Bureaucrat are different, just like admins. While an admin may speedy delete an article, another may send it to AFD. While one may block a user for 3 hours, another would block it forever.
2. We welcome diversity. If we were looking for bureaucrats or admins acting always in the same way, we would either use a bot, or have a single administrator and a single bureaucrat.
3. Because of this diversity, yes, it is possible that another bureaucrat could have closed this request as No consensus. But, at least down here, Argentina, in real life, trials may have different outcomes according to attorneys and magistrates.
4. Exceptions exist, and we all know that. Death is the only certain thing in life, all others can be modified. As long as the exceptions are not abused, I do not think Wikipedia will explode. I have been here for 18 months (14 if you are picky about edit rates) I have seen this happen two or three times. And I must have seen over 200 new admins created. As stated somewhere above, if we were searching for 80.0% or above always, without exception, User:AdminMakerBot would have congratulated me.

Just some thoughts before going to sleep. -- ReyBrujo 07:36, 24 January 2007 (UTC)

Thanks, Rey. As I go to sleep, I must say, I wish I'd seen admins do this more often with other candidates. Ryulong aside, I've seen some awfully good candidates who would have been exemplary admins get rejected because old edit disputes. I'm glad 'crats are not taking it as a vote. If we want to change policy to say absolutely only between 75-80 or 70-80%, then let's change it. Otherwise, no policy has been violated, and I for one agree with the said non-policy. Patstuart^talk|edits 07:40, 24 January 2007 (UTC)

wikiparadox

Wow! Wikiparadox. On the one hand we have an editor with over 10000 edits. Those Never, ever ever pass RFA cleanly (drop by on irc or my talk page if you never heard of this rule of thumb before and are curious as to why). On the other hand we have someone with >100 supports, worthy of mention on WP:100. People with over 100 supports generally pass RFA with flying colors.

Sooo... yeah, this was one very strange request for adminship. I figure Raul654 gets a little more discretion here, and he probably tilted in the correct direction at that.

Of course, Ryulong is going to have to listen most carefully to what some of the people who were opposing have to say, and perhaps some of them can keep a bit of an eye on him to make sure he's really as nice as the supporters said he is?

--Kim Bruning 08:14, 24 January 2007 (UTC)

Not always; I had more than 10k edits during my RfA (1 oppose) :) — Deckiller 08:17, 24 January 2007 (UTC)

Darn. Do I have to up my rule of thumb to 20000 by now, do you think? --Kim Bruning 08:22, 24 January 2007 (UTC)

Nah, the RfA was more than 10 months ago, when WP:100 was a three user club :) — Deckiller 08:24, 24 January 2007 (UTC)

I don't think that there's any question that Ryulong will be closely watched for slip-ups. The issue here (and also raised at WP:BN) has to do more with Raul's closing of the RFA as successful without even 70% (and even then most people ask for 75% before a crat can use discretion) support, and apparently completely based on his own opinion. – Chacor 08:18, 24 January 2007 (UTC)

Well, human beings have intuition and opinions. Sometimes those are useful, since else we could just let a bot do the closes, right? --Kim Bruning 08:22, 24 January 2007 (UTC) Either that, or you could wear this t-shirt at the next wikimania, and show it to all the bureaucrats! [1] ;-)

Really, "I like him" is a pretty damn lame closing reason. Isn't this just the kind of argumentation we discourage at AFD? I expect a better explanation for such a controversial close. I'm terribly disappointed in Raul. He's let the community down on this one. Regretfully, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 08:24, 24 January 2007 (UTC)

You don't think there was a need to correct for process bias, at the very least? --Kim Bruning 08:26, 24 January 2007 (UTC)

If Raul had said "I carefull considered this request for adminship and decided that the consensus of argumentation seemed to be support" then I wouldn't have had a problem. However, he did not. He said "I have interacted with Ryulong and like him" which is favourtism, at best, and an abuse of power, at worst. I fully expect demand a more thorough explanation of this closing, addressing the points raised by the support and oppose !votes.

In fact, while we are discussing accounting for process bias, if we strike out all the "ILIKEIT" votes from the support side and all the "IDONTLIKEIT" votes from the oppose side, then I think we find a number that fails even a 50% guideline dismally. That, to me, is the disconcerting part. The opposition votes were generally well thought out, supported, and brought by respected members of the community in a respectful fashion. They should have been given a lot more respect than the scarity Raul showed by glossing over them so quickly and nonchalantly.

Regretfully, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 08:34, 24 January 2007 (UTC)

Hmph. Well, we could always just replace Raul with a really small shell script, I guess! --Kim Bruning 09:21, 24 January 2007 (UTC)

he point is simple - we can bandy about the process and how broke it is and blah blha blah - the numbers don't amount to a pile of dung and I think everyone knows that. The problem is that bureaucrats are chosen for their fair judgement, integrity, social clue, and good sense. Raul wasn't showing much of a clue by not sitting down and taking a moment to really think this through. It seems to have been a knee-jerk reaction on a "hunch", and hoever good that "hunch" is, it's not how Wikipedia works. That's with good reason, too. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 09:42, 24 January 2007 (UTC)

to Raul

I know Ryulong uses IRC, I know your use IRC, I put 1 and 1 together I get 2, pertinent to Ocam's razor you personally get along with Ryulong, and for a 'crat who has never? (recently?) closed an RfA you picked your moment well, personally there was absolutely no consensus to promote him and what Chacor says is extremely "scary" - Raul I can say this without even knowing you: You've let me down. thanks/Fenton, Matthew Lexic Dark 52278 Alpha 771 08:29, 24 January 2007 (UTC)

• Fishy, at best. Right loss issues at worst. I'd put it somewheres in between. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 08:38, 24 January 2007 (UTC)

Raul is a big guy. (boy is he ever, in more ways than one :-P ), and generally does what he thinks is the right thing, and watch out people who try to influence him. That's why he got the bureaucrat bit, I'm sure!

So while I'm sure Raul will also have been aware of this situation on IRC, I'm positive he thought this one out well by himself as well.

(That, and when in doubt, I sometimes use irc to probe people and see what they're made of too. Which is actually what I did in this case, come to think of it! Hmm! :-P ) --Kim Bruning 08:45, 24 January 2007 (UTC)

To reply to Matthew Fenton's comment above - yes, I have interacted with Ryulong on IRC, as I have interacted on IRC with many hundreds of other people. Most of the converations I have had with him have involved him asking me to checkuser and/or block Boabaobo (sp?) sockpuppets. (Boabaobo apparently tends to edit the same articles as Ryulong)

As to this RFA, the first time Ryulong knew of my interest in his RFA was when I told him I intended to promote him, about 5 minutes before I did. Raul654 08:57, 24 January 2007 (UTC)

...thereby dragging him into a mess of your creation. —David Levy 09:18, 24 January 2007 (UTC)

Only if you're the one choosing to do the dragging! ;-) --Kim Bruning 09:20, 24 January 2007 (UTC)

Clearly, he didn't - and it's a shame too. Even if Ryulong is a fine admin, goes on to get every special right there is and gets eleted to ArbCom, the Board, whatever, this controversy is going to haunt him for a long, long time. It's not very fair to Ryulong, and it's not very fair to the community either. Regretfully, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 08:49, 24 January 2007 (UTC)

The best we can hope for is that he proves the opposers wrong and becomes one of the best leaders on the 'pedia :) — Deckiller 08:51, 24 January 2007 (UTC)

Indeed. His admin actions don't have "I passed with less than 70% support" written on them, and most people will just judge them on their own. I don't see people going around challenging every single of Carnildo's decisions. Kusma (討論) 09:00, 24 January 2007 (UTC)

Carnildo had many opposes from trolls, et al, I'd say this is different. There were real valid strong concerns raised over WP:BITE, etc., and they won't go away any time soon. – Chacor 09:03, 24 January 2007 (UTC)

He has already proven to me he is willing to abuse the powers, he's firmly placed his user talk page into the admins only club, when I see stuff like that it screams out to me "I don't trust you not to move my page and I am better then you as I can protect my own pages!" thanks/Fenton, Matthew Lexic Dark 52278 Alpha 771 09:01, 24 January 2007 (UTC)

No offense, but that is hardly abusing any sort of power. Do you have any kind of clue the amount of move vandalism that goes on w/ talk pages of "well known" sysops? AntiVandalBot's talk has been move protected for eons now, my talk page has been move protected for eons. Before said protections I had move vandalism waay too fequently. Really, move protection of a talk page which has no business being moved (and come on, I can't think of a legit case to move a user talk page like that.) is no big deal and should hardly be seen as admin abuse. -- Tawker 09:10, 24 January 2007 (UTC)

Agree, move-protection is not uncommon. No reason for user talks to be moved, anyway. – Chacor 09:15, 24 January 2007 (UTC)

Adding on to that, it's an easy page for him to test the protection system on, providing another valid reason for him to have move protected his page. --tjstrf talk 09:20, 24 January 2007 (UTC)

Actually correct move. You're supposed to be discussing here, for now at least. --Kim Bruning 09:06, 24 January 2007 (UTC)

If that's all you have to offer as evidence for "abuse", then Ryulong is doing a great job so far. Kusma (討論) 09:07, 24 January 2007 (UTC)

Actually I've given two people barnstars for redirecting conversation from elsewhere to here. Ryulong will have to wait for his. :-P --Kim Bruning 09:09, 24 January 2007 (UTC)

Oh, only move protected even. Nevermind. --Kim Bruning 09:20, 24 January 2007 (UTC)

Am I the only one who starts mentally screaming and throwing things whenever I hear someone mention secret IRC conspiracies? --tjstrf talk 09:11, 24 January 2007 (UTC)

Oh I just go someplace quiet and do that physically. That or taking a jog around the block, or what have you. Works wonders! --Kim Bruning 09:18, 24 January 2007 (UTC)

I find this deeply disturbing. It's simply not on that someone sees fit to draw "secret IRC conspiracy" from "User A and User B both use IRC. User A helps User B". It's not on that people have to put up with half accusations like this. With all due respect, as far as I'm concerned, it's "hard evidence, or keep quiet until you've got some". — Werdna talk 10:18, 24 January 2007 (UTC)

Result of this uproar

[2] --²¹⁰physicq (c) 08:30, 24 January 2007 (UTC)

Adminship is no big deal, however the tools in the wrong hands are a big deal to me, you can do a lot of damage with those "tools" (i.e. blocking a country, blocking a whole ISP) thanks/Fenton, Matthew Lexic Dark 52278 Alpha 771 08:33, 24 January 2007 (UTC)

I have no comment regarding this contemplation. --²¹⁰physicq (c) 08:35, 24 January 2007 (UTC)

Same. — Deckiller 08:38, 24 January 2007 (UTC)

Put simply, judgement and discretion is required in administrative actions. Incidentally, judgement and discretion is required of Bureaucrat actions, something which seems dismally lacking here. Regretfully, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 08:39, 24 January 2007 (UTC)

My first thought upon learning of this matter was that regardless of whether Ryulong would make the best sysop or the worst, this thing was going to haunt him. It really isn't fair (as the situation certainly isn't his fault), and I wish that Mark had considered this before performing a closure that he must have known would be controversial. Assuming that there was valid reason to deem this case borderline, a consultation with other bureaucrats clearly was called for. By announcing that he had promoted Ryulong based on his personal opinion, Mark basically painted a bull's-eye on Ryulong's back. —David Levy 08:44, 24 January 2007 (UTC)

Like before, I completely agree; a bull's-eye that Ryulong should not have. We need our members to be accepted within the community - especially our new admins, regardless of who opposes. If this promotion stands, we need to help Ryu and work with him, not criticize his every move and haunt him about minor blunders. Let's not drive away yet another user as a community. By the way, this comment about the bull's-eye (at least my aspect of it) isn't aimed at Raul; this is aimed at the uproar in response. — Deckiller 08:47, 24 January 2007 (UTC)

Um, it's not Raul654s fault if people were to take aim at Ryulong, it would be *peoples* fault. Blame where blame is due! --Kim Bruning 08:48, 24 January 2007 (UTC)

This was an entirely foreseeable consequence of the closure. I don't blame Mark for other people's inappropriate behavior, but I do hold him accountable for creating a situation that he must have known would trigger such a response. Whether Ryulong deserved to be promoted or not, he certainly didn't deserve to be dragged into the center of this controversy. —David Levy 09:18, 24 January 2007 (UTC)

Well, it's definately a way to quickly discover where all the troublemakers are hiding. <pulls a slightly worried face -- did we really want to discover that?> --Kim Bruning 09:45, 24 January 2007 (UTC)

Are you suggesting that the people who have problems with this closure are "troublemakers"? Ad homimem arguments aren't cool at all. Really, really. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 09:51, 24 January 2007 (UTC)

Duh no. But some people may well try to use the mess as an excuse to act up, of course! --Kim Bruning 11:45, 24 January 2007 (UTC)

Perhaps, in lieu of Ryulong resigning, he could use the tools for a month or two and then open a Request for Comment on his administrative actions to that point. The community seems happy with Carnildo and Sean Black being admins, despite the way they were re-sysopped. After some real world experience with the tools, another opportunity for community review could provide some necessary closure and clear whatever cloud Raul has placed Ryulong under. If he turns out to be a good sysop despite my and other opposers' concerns, then WP will be better off in the long run.--Kchase T 11:44, 24 January 2007 (UTC)

Exactly. And hopefully this will prove once and for all that RfA is too strict. — Deckiller 12:29, 24 January 2007 (UTC)

Disappointment

I am starting to get disappointed in how people handle things here. A consensus is not a vote, a consensus is simply a determination of a debate. Ryulong has been granted the tools, and has promised to use them well. IF he breaks policy and goes nutso; then take it up with arbcom. Raul654 didn't cabalize it, and his name should not be tarnished over determining consensus. Frankly Chacor, I have seen issues with you since your own ArbCom ruling came down and your mop was taken away for policy you broke. I respect your opinions, but you have seem to have come into some attitude issues.

Do I support and continue to support Ryulong? Yes.

Am I worried he may cause issues with blocks? No, or I would not have nominated him.

Do I hear and recognize the community issues? Yes, of course I do; and they are quite valid. But on the other hand, a consensus has been reached and Ryulong HAS been made an admin. The only way this is going to change is with an ArbCom ruling.

Should Ryulong step down? Not at this time, if he shows bad faith and becomes a problem then yes. But stepping down would be a disappointment to those who did support him, and would be a show of bad faith from the community; as he had not done anything to justify a resignation.

I hope we all can make the best of this, and besides cross-posting this to the WT:RFA page, the only time I will post on such issues. Somitho 09:36, 24 January 2007 (UTC)

How the hell is this turning into an attack on me when I'm not the only one who has expressed disappointment at this? This is an unacceptable ad hominem attack. – Chacor 09:37, 24 January 2007 (UTC)

A few things, if I may. First of all "I like him" is an unacceptable argumentation for a promotion. It practically oozes a conflict of interest. Secondly, Somitho, please retract the personal attack. I'ts inflammatory, and it's unneccesary. ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 09:49, 24 January 2007 (UTC)

The problem is that Mark didn't determine consensus. By his own admission, he determined that there was no consensus and used his own personal opinion of the candidate as "the deciding factor." A bureaucrat's discretion allows him/her to analyze and interpret the preponderance of the available comments, not to cast a "super-vote" that weighs more than anyone else's. —David Levy 17:20, 24 January 2007 (UTC)

Based on people informing me that Frankly Chacor, I have seen issues with you since your own ArbCom ruling came down and your mop was taken away for policy you broke. I respect your opinions, but you have seem to have come into some attitude issues. could have turned into a personal attack, I formally retract such statement with apologies. It was never meant to be a personal attack, just an issue I pointed out. Somitho 09:57, 24 January 2007 (UTC)

Copied from WP:BN

(Didn't know about this discussion) Having not participated in the discussion and not known about it had I not had this page on my watchlist I would also like to voice my misgivings that there was not a clear consensus in this case and it therefore in my view should not have been closed as a sucessful RfA. It certainly did not meet the arbitary standard for RfAs these days (80% support) and many if not all of the oppose views were from respected editors expressing their views, not socks/meats or disrupters. Viridae^Talk 08:09, 24 January 2007 (UTC)

Good arithmetic?

"consider myself as supporting his request for adminship, and discounting the opposition from the newbie with a dozen edits. That gives a final tally of 126-54, which is 70%, which is in the bureacrat's discretionary range." That leads to two questions: are bureaucrats allowed to vote in RfAs that they close? Did Raoul check all the support votes to see if any of those were from newbies?--Runcorn 19:41, 24 January 2007 (UTC)

Only if they register with the review board sub-committee three weeks in advance and receive their written confirmation a minimum of five business days before the closing. A minimal processing fee is required. Jkelly 20:01, 24 January 2007 (UTC)

Change to AfD style RfA

I am disappointed with this, even though I only voted neutral. I think it's time for a change. There should be no numbers, if the 'crats make their own decisions, as indicated here, and Carnildo 3, Sean Black 2 etc. Make it more like AfD, and the result can be reviewed if people disagree, much like DRV. It'll at least make the process easier to reverse if enough people disagree. It's one thing complaining, but another actually changing. Comments please. --Majorly (talk) 12:42, 24 January 2007 (UTC)

The problem with RfA is that support votes don't state reasons (and in many cases, it's hard to see how they could), and so they all end up identical. It's clear to me that an AfD-style RfA would result in the promotion of almost nobody. You might also want to look at why Wikipedia:RfA as RfC failed; I think that might be relevant to your point in an indirect way. (Although I supported and helped draft that proposal, it became clear what its problems were when it was tried in practice.) There was also the addition of the 'Discussion' section to the RfA for much the same reason as you suggested, but people seem reluctant to use it (it's probably a matter of tradition, and of adding in the same place as other users). --ais523 12:48, 24 January 2007 (UTC)

I think we need to do one of two things. Either make it a straight out vote and throw away the discussion part (this is what we do for ArbCom already), or go the other way and have a pure discussion with none of this support/oppose/neutral dickery. Thoughts? ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 12:51, 24 January 2007 (UTC)

Our best option to ensure a more realistic (as to the future of this project; it's going to continue growing and accumulating waste) and fair way of promoting is probably to create a more realistic grey area; 65-75 percent. 75 percent and above can be almost always automatic, below 65 percent never (barring significant trolling and whatnot), and 65-75 can be that grey area. — Deckiller 12:53, 24 January 2007 (UTC)

Anything with numbers ends up just being incongruent dickery, in my opinion. As the quote goes: There are three types of lies: lies, damned lies, and statistics.' ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 12:55, 24 January 2007 (UTC)

Perhaps, but because we already have numbers as a factor most of the time, we can take this time to at least tailor the current system so that it works with the needs of the project. — Deckiller 12:58, 24 January 2007 (UTC)

(edit conflict) What with the problems with giving bots admin rights, it would be even harder to promote a bot to 'crat (only half joking here); if, as I think most people here would agree, we use human 'crats, there's always going to be some discretion used. Besides, there's suffrage/sockpuppetry problems in straight votes (although I suspect they wouldn't be insoluble). I would be more inclined to prefer a straight discussion than a straight vote, although it might be instructive to look at why Wikipedia:Requests for bureaucratship/Grandmasterka failed (as far as I can tell, mostly because the platform involved taking the quality of votes into account, which will always be controversial due to disagreements on what makes a high-quality vote, or I suppose !vote under the Grandmasterka plan). --ais523 12:59, 24 January 2007 (UTC)

I really would not like changing to an AFD style because it wouldn't solve anything. All it would do is make the thing harder to read. --BigDT 13:35, 24 January 2007 (UTC)

Responses to various points above

I'd like to address a few points raised by some people in the above discussion.

• Re: Consensus correlation and trust: This is an interesting point raised by User:Werdna who said "I think that a 5-10% difference in RfA consensus has a rather low correlation to actual community trust in the user". Earlier in the conversation, User:Blnguyen noted that 80% equated to a 4:1 ratio, 75% to 3:1, and 70% to 2.33:1. I think this shows that indeed a 10% range from 70% to 80% shows a dramatic difference in consensus. There's a reason the normal accepted range is where it is; consensus tends to form around that level.
• Re: No accountability for bureaucrats: This was raised by User:Heimstern. Wikipedia is not a democracy, and we do not have any sort of balance of powers really. There is no hierarchical structure here of editor-->admin-->bureaucrat-->arbcom. All of these people in these various categories are equals. Some simply have additional responsibilities that have been entrusted to them through filtering processes of one sort or another. Admins aren't any 'better' than editors in good standing. Neither is ArbCom any better than editors. There are no oversight processes for any duty within Wikipedia. What there is are processes to handle breakdowns in the system. If an editor goes bad and refuses warnings, we block. If an admin abuses their powers, they're taken to ArbCom to make the decision to desysop. If (hasn't happened but if) an ArbCom member went off the deep end in some way, Jimbo would probably step in and ask them to step down. If a bureaucrat gets out of line, they can have their bureaucrat duties removed (and yes this has happened). So, while there is no accountability per se, there is always accountability in general for all things to keep our focus on what it is we are trying to achieve here.
• Not very fair to Ryulong: I raised this point on the Carnildo promotion. However, in this case I don't think it applies. Raul654 did talk to Ryulong before the promotion. Ryulong's well experienced enough to know that this promotion would be controversial and would understand the consequences. Apparently now he is reconsidering his decision.

Personally, I would not have promoted Ryulong. An RfA can be an opportunity to improve various aspects of a Wikipedian and there's no policy in place preventing a second, third, fourth or even fifth RfA. It's been done multiple times before with no apparent harm to the project. The vast majority of the time, RfA percolates along just fine and bureaucrats are not questioned in their decisions. This is important because bureaucrats have a difficult task of weighing different aspects of a discussion with impartiality. Bureaucrats must pass a high threshold of consensus before becoming a bureaucrat precisely because they must be trusted and continue to maintain that trust. A controversial decision such as this RfA tends to cash in some of the trust given to bureaucrats and potentially sets the stage for more argumentation against bureaucrats being entrusted with divining consensus. This is of course a balancing act; sometimes tough decisions need to be made which is what bureaucrats are entrusted to do, but the tough decisions erode community confidence in their ability to make...tough decisions. Bit of a catch-22. --Durin 14:51, 24 January 2007 (UTC)

I think that's a very interesting point you raised which we haven't really considered. Ryulong was told of this before he was promoted. He knew he was below the threshold for consensus. I'm concerned that he had the choice of saying "I haven't achieved consensus to become an admin", but didn't take it. It's certainly not crossed my mind until you brought it up, and I think this makes this promotion all the more controversial - the candidate could've said "no, I haven't made it". – Chacor 14:57, 24 January 2007 (UTC)

Doesn't the fact that he was asked if he wished to be undermine the whole RfA process? thanks/Fenton, Matthew Lexic Dark 52278 Alpha 771 16:11, 24 January 2007 (UTC)

Not sure if this is the place, but as this is the RfA trainwreck du jour I just must chime in. What bothered me was the initial comment that the b'crat just liked Ryulong, only as literally an afterthought was community consensus mentioned. I like to think b'crats don't just cast supervotes, but admittedly it is a fine line between interpreting consensus and using the claim of doing so to do what you want anyway. A lot of the issue to me has to do with the history of RFB, and this hasn't really been mentioned here... if you look at the pre-2005 RFB, which is when a lot of the current b'crats were promoted, it's hard to imagine a lot of them passing today's RfA, let alone today's RFB. I have a lot of confidence in people like Taxman who came in under the modern system, not so much so in b'crats who got elected in the initial days of RFB and haven't really done much as b'crats over the years. To give such automatic confidence to people who really didn't pass anything resembling modern RFB just seems a bit arbitrary. I admittedly speak more of the b'crats who never do anything than Raul who is somewhat active at RfA. --W.marsh 16:13, 24 January 2007 (UTC)

Ditto, not sure where to add my twopennyworth, but this will do. I respect Raul654, but I think this was a poor decision. 69% is not much more that 2:1, so this promotion effectively rides over the opinions of nearly a third of the participants.

On probationary adminship, Carnildo was controversially re-adminned with about 61% "on a probationary basis, for a period of two months, after which his activities will be reviewed by the arbcom" (see Taxman's comment in this archive). Presumably this "review" happened on or around 5 November 2006 - was there a report of its findings anywhere? (A dog that did not bark, no doubt?) -- ALoan (Talk) 20:07, 24 January 2007 (UTC)

Poor judgment

I believe it was extremely bad taste to promote somebody with such low support percentage. It does nothing to improve people's faith in the objectivity of the process (and people already think RfA sucks). The moral of this story is that it does not matter how many people in the community think you are not fit to be an admin, all that matters is that a friendly bureaucrat thinks you're a nice guy. Apologies if I repeat something from the above, that discussion was too long to read completely. Either way, I think this was a very poor judgment on Raul654's part, a bureaucrat who has not promoted for ages, and this should not be repeated. Oleg Alexandrov (talk) 16:05, 24 January 2007 (UTC)

Exactly why I find this situation so fishy, it just doesn't make sense to me. thanks/Fenton, Matthew Lexic Dark 52278 Alpha 771 16:11, 24 January 2007 (UTC)

So you'd rather have a bot do the promotions based on a number instead of judgement? I guess I should have put my earlier comments down near the bottom here so they don't get lost, but either way they'll just get passed on by the volumes of material that are going to come out. Too bad really, and I'm adding to it, so I'll stop now and instead go work on some articles. - Taxman ^Talk 16:31, 24 January 2007 (UTC)

Nobody said a bot should do the promotions. The point is that we have a system in place, according to which people are promoted. As far as I know, bureaucrat's discretion is in the 75%-80% range. Granted, the system is arbitrary, and the numbers are arbitrary. However, it is profoundly unfair that some people did not get promoted with almost a 3/4 support ratio, while somebody got promoted with much less because a certain bureaucrat thought that no matter what people said that person was a nice guy. It is surely not the fault of the person who got promoted, but it is hard not to agree that the bureaucrat who did the promotion made a mistake. Oleg Alexandrov (talk) 17:51, 24 January 2007 (UTC)

On Bcrat accountability

People are mentioning a lack of accountability, but do we really know whether this is the case? Surely the bcrats, as a group, want to have the confidence of the community. If there was significant opinion that a bcrat should step down and/or go through RFB again, why are people assuming the crat would refuse? Surely community trust in the crats as a group is more important than any one editor's privilege level. Friday (talk) 16:22, 24 January 2007 (UTC)

Yes, I think Raul (and other 'crats who are not particularly active) should go through RfB again. --Majorly (talk) 16:32, 24 January 2007 (UTC)

Personally, I trust the bcrats and rarely disagree with their judgment calls. However, I would prefer that they had an attitude closer to "When in doubt, don't promote". The reason is that an editor can always submit another RfA if they don't succeed, but an improper promotion can shake community confidence. There's probably only a handful of these true judgment calls a year, but I think it would be better for such a candidate to wait a few months and resubmit an RfA. Anyway, just my 2 cents. SuperMachine 16:31, 24 January 2007 (UTC)

The problem has to do with the rules of RfA system and their implementation; not with one person

• "Futhermore, before I promoted, I made it a point to read through all the people opposing his adminship. If it had been 65%, would I have promoted him?" (by Raul).
• "While some might object that he was below 70% support (with 69.4%), I believe, based on my interactions with him, that he would make a good sysop, and it would be a mistake not to promote him." (again by Raul)

I stress on this two assertions of Raul, because I think that they are at the heart of the problem here. IMO we don't have to do with a problem concerning one person, Ryulong in this case, but with a problem having to do with the way the whole RfA system works. First of all, I want to make clear that I honestly believe Ryulong will do a great job as an adm, and I regard his promotion as a closed case. Whether we agree or not with the way his RfA was closed (I must say I disagree), Ryulong is now a confirmed sysop, and should be allowed to do his job as a sysop properly. I do not agree with the limitations on the use of the sysop tools proposed here; an adm either is an adm either he isn;t. And Ryulong is one whether we like it or not. Limitations on the use of adm tools would ridicule the role of adm, and the credibility of the whole RfA system. Enough harm has been done already. We can't have adms of two velocities. The only limitations on Ryulong are those arising from WP guidelines; namely, the same that exist for all adms and users. For me, we have to deal with another problem here ...

Raul says he exercized discretion. But is discretion as an argument enough to nihilate the rationale of a whole system; because IMO discretion the way was exercized here did exactly that. Raul argues that "while some might object that he was below 70% support (with 69.4%), I believe, based on my interactions with him, that he would make a good sysop". But RfA doesn't speak about a 70% potential threshold; it speaks about a 75% potential threshold: "The numbers of people supporting, opposing, or expressing another opinion on a candidacy are a significant factor in determining consensus (few RfAs succeed with less than 75% support), but a request for adminship is not a ballot". Yes, RfA is not a ballot, and I would accept discretion exercized by Raul or any bureaucrat if Ryulong was about 74%, 73%; maybe 72% and with low number of objections. But here support is below 70%?! And 70% is already faaaaaaaar below the potential threshold of 75%. And Ryulong had more than 50 objections and many neutral votes (Don't we take them into consideration? The also matter IMO!).

And when does this discretion stop? Who tells me that another bureaucrat will not promote a nominee with 68%, 67%, 66%, 65%? Raul asks "If it had been 65%, would I have promoted him?" My answer is: I do not know Raul, but your discretion in the way you exercized it establishes IMO a negative precedent. It gives the right to bureaucrats to exercize their discretion in a much looser way. And, yes, maybe you will not do it, but, in the future, another bureaucrat based on your precedent may promote nominees with 65% of support and even lower. Maybe this is what we want: a lower "threshold" (I know the term is not accurate). But, if this is the case, let's discuss the whole matter and decide it accordingly.

The problems are numerous here; I recognize that. And I must say I watched Ryulong's RfA too closely (I was also a nominee at the same time!), because I expected something like that to happen; it was from the first moment a close call. The issues this RfA opened are many, and this is good, because it will make us think: What are the limits of bureaucrats' discretion? Are there any? Do we have the right to question their decisions or are they "sacred figures" beyond criticism? Can we have adms "into probation" as it was proposed here? What is the "threshold" for adm: 75% as RfA says or 70% as Raul argues? And if a nominee is around 80%, but with many objections (more than 30) raising serious issues for the credibility of the nominee, can the bureaucrat exercize his discretion in the opposite direction by not promoting? And in any case, we must never forget Jimbo Wales' wise words.--Yannismarou 16:26, 24 January 2007 (UTC)

The wise words were certainly true in the good old days, but looking at the discussions above, it is evident that things have changed a lot. — Ambuj Saxena (☎) 17:23, 24 January 2007 (UTC)

RfA isn't broken.

Whenever something like this occurs, we see radical proposals of how we can "fix" RfA. Well, there's no need to turn the process upside down; it usually works. The problem is that the correct process wasn't followed here. Mark, being human, made a mistake. It happens, and it will continue to happen occasionally under any RfA system.
Furthermore, while Mark's decision was made for the wrong reason, we don't even know that it was the wrong decision. Maybe Ryulong will be a terrific sysop. Can we please give him a fair chance and not blame him for Mark's misstep? —David Levy 17:20, 24 January 2007 (UTC)

Mark made a mistake? I don't think it's quite that simple. When someone makes a mistake, the best response is to admit the mistake and attempt to correct it. This doesn't seem to be the case here. Mark believes that he made the right choice and therefore there is no mistake to fix. Perhaps Ryulong will be a terrific sysop, but that doesn't change the fact that this decision isn't just an isolated "mistake". I believe that it illustrates a disconnect between the way that at least one bureaucrat views consensus and the way it's historically been defined. SuperMachine 17:45, 24 January 2007 (UTC)

My point is that this was an instance in which the correct process wasn't followed, not an instance in which it failed. When the correct process is followed, we have very few problems. An occasional bicycle accident is not a valid reason to reinvent the wheel.

I agree that Mark's actions are indicative of an underlying problem, and I believe that he needs to acknowledge this and promise not to act similarly in the future. —David Levy 18:01, 24 January 2007 (UTC)

Servants of the community, not masters over it?

It seems to me that this whole palaver stems from one problem: a misperception that crats are the masters of the community. They aren't, or, at least, if they are, they shouldn't be. We elected these people to do jobs for us, on our behalf - namely promote the people whom we had chosen to be admins to sysop status, and to close the RFAs, among other things.

In this case we had clearly not chosen an admin, within the boundaries of RFA's definition of "chosen". Given the rationale above, the crats had no right to promote against the consensus of the community as that consensus is defined on the RFA page, especially when the rationale for promotion was nothing more than ILIKEIT on a spectacular scale. If I had taken a similar rationale to RFA, or AFD, or anywhere else on Wikipedia, I'd have been laughed out of the house and possibly blocked for disruption.

More generally, every system in the world is broken. You will never, ever be able to come up with a perfect system, because systems are by definition the creations of men, who are not possessors of maximal excellence. On a positive note, there is no reason to think that RFA is massively broken and IMO we are getting better and better sysops, and will continue to do so for a while, though self-evidently that can't last forever. In this case I'd be surprised if Ryulong abuses the tool after this, when doubtless a fair few people will be watching everything he does. But the principle on which he was promoted is not on. Moreschi ^Deletion! 20:04, 24 January 2007 (UTC)

Scientific analysis of RfA

Well, I brought this issue up at Wikipedia_talk:Requests_for_adminship/Archive_72#Scientific_and_Emotional_judging_and_bureaucrat_intervention and again at Wikipedia:Requests for adminship/Agathoclea when some niggles turned up. Basically, I feel that if a user is low-profile, it naturally makes people less inclined to instinctively support them, and if they are more high profile, then they are more likely to instinctively support, without probing around for weaknesses. For a low-profile user, people tend to try and dig up all sorts of obscure reasons to oppose. Without reprinting all the stuff I said again in the old archives, I can remember people saying "not enough WP, not enough talk edits" on one user, who had dozens of start class articles, and an FA, (Sam Vimes) and he only scraped through at around 78.1%. However, there was another RfA at the same time where the relatively famous user only had two stubs, 8 times less article edits, and only half the amount of (allegedly too few) WP and talk edits which Sam had. Despite this, the other RfA passed at 98% close to the century of supports. It is a human instinct which is causing these things, a human difficulty of separating heart from head, which gives "dashing", "spectacular", "superstar", "flashy" users a much easier run than "dour", "self-effacing", "low-key" editors. If Wikipedia is to fulfil its potential to the maximum, then certainly it must be more meritocratic. Now that a certain controversy has arisen, I feel that it is time to give all candidates a fair go, in terms of ALL grey-zone candidates receiving a close examination of their case, instead of only high profile ones, or ones who are lucky to have had good experiences while working with bureaucrats. I would like to see in grey cases, bureaucrats to analytically compare the issues raised in rock hard, objective terms, so that merit rules the day. It is irritating that because of emotional !?"voting", some candidate X are opposed by one opiner A for some criteria, while another candidate Y failing that same criteria is supported by opiner A, because the opiner simply goes on perception/image rather than rock-solid contributions. If some people suddenly go all soft on some "cute" candidate and then nitpick on some "ugly duckling" candidate despite the "ugly duckling" meeting a given opiner's "criteria" more thoroughly; I feel the crats should take these abnormal phenomena into account in non-landslide RfAs, so that the most skillful selection results for the good of wikipedia. There is no easy way to make people have the same standards, but I think the candidates deserve consistency from a given opiner, or in this case "umpire". Umpires have different interpretations of certain rules/laws in sport, but in a single game, they should apply their interpretation to both sides equally. I would like to see this in RfAs also, so that a superior candidate who is opposed by the same opiner that supported a weaker candidate using the same criteria is not negatively affected by their profile in a close call. This also gives the advantage of filtering out attack/vendetta votes implicitly: If a person decides to disguise their attack oppose, or is unconsciously biased against a candidate they clashed with, they will suddenly use a higher bar on one candidate ( conscious or not) than they did before, and this can be taken into account and dampened. Certainly I would favour this system, and would operate as such if I was a bureaucrat, although, some people don't like arbi-crats, so I doubt I would be able to. Blnguyen (bananabucket) 07:38, 24 January 2007 (UTC)

You know that's interesting you'd say that, because I thought I'd noticed the exact opposite of the first part of your statement. People who keep low profiles and never participate in any admin matters (i.e., the stuff that can get people mad at you) will do great: (e.g., User:Kuru, who, BTW, I whole-heartedly support), while more well known people have the opposite problem (e.g., User:BostonMA). I'm completely surprised to hear you say this, as I must disagree with your assessment. As for the second part: I agree too much. I hate to see vendetta votes or inconsistency votes (usually it's a matter of WP:ILIKEHIM, which I've seen certain admins actually admit to voting to).

Patstuart^talk|edits 07:43, 24 January 2007 (UTC)

BostonMA edited religion related pages and disputes. Kuru was mostly a gnome. I'd say that high-profile failed RfAs such as Tariqabjotu, BostonMA, HRE, Khoikhoi failed because the community is very averse to any edit-disputes whatsoever, and if the people were in any non-trivial religion or ethnic dispute, everybody will run away and assume the user is troublemaker or stirs the pot. If you compare people who do undisputed edits, I would say that I have seen many cases when a popular but weak user sailed through, whereas an obscure user, like Wikipedia:Requests for adminship/Agathoclea and [[Wikipedia:Requests for adminship/Sam_Vimes2 and Wikipedia:Requests for adminship/Christopher Sundita picked up more damage on "lack of WP/talk edits" whereas other famous users who also edited far away from disputes, with less of the same thing, went almost unanimous. Blnguyen (bananabucket) 07:51, 24 January 2007 (UTC)

It appears that we completely agree on this, at least on the religion/edit dispute matter, which is a shame in its truth. However, I've noticed quite a few obscure users sail through precisely because no one knew any of their bad edits (maybe they were just obscure to me, as they only edit articles, and I don't see them on the message boards). Patstuart^talk|edits 07:59, 24 January 2007 (UTC)

Generally I feel if you make bad errors, people will always notice, but the problem seems to arise over what they did not or have yet to accomplish, rather than explicit failure. Having said that, if nobody ever knew that someone made a mistake (if it's serious enough, their enemies or simply neutral observers will make a note of it), the system and the bureaucrats would be irrelevant. Blnguyen (bananabucket) 08:11, 24 January 2007 (UTC)

There is a very subtle aspect to the first perspective that renders it tremendously difficult to implement. Note the academic nuance I am deliberately invoking here: what is the difference between equity and equality? Historically, no Wikipedian has been able to precisely characterize this with respect to RfAs. There is reason for this: it hints at a larger challenge to be overcome. --HappyCamper 07:50, 24 January 2007 (UTC)

Voter biases will always exist unfortunately no matter how much this matter is discussed. Whether "low-profile" or "high-profile" as you call it, I believe the outcomes vary in many circumstances as Patstuart mentioned above. I'm not trying to insinuate that discussion is pointless but a large majority of Wikipedians have their opinions, including myself or yourself. If as RfA supporters/opposers we can begin to show a little less biases, that is progress enough in my opinion. Unifying what we expect or examine in prospective admin candidates seems a tad more impossible to achieve.¤~Persian Poet Gal ^(talk) 08:00, 24 January 2007 (UTC)

Not sure if there's a misunderstanding, as is often the case each time I raise this issue - every member of the community is free to have their own rationale as to what makes a good administrator - That is fine with me. I only expect that a given person is self-consistent. Many sport umpires have different interpretations of the grey zone, but if they interpret this in one way and give many penalties/fouls to one team, and interpret the grey area a different way for another team, then people will complain. Blnguyen (bananabucket) 08:11, 24 January 2007 (UTC)

If you mean consistency with their opinions, then that is indeed a good thing to stress. I'm just saying that unfortunately biases will not die down easily if ever.¤~Persian Poet Gal ^(talk) 08:21, 24 January 2007 (UTC)

If it is a first time RfA, it may be easier for a bureaucrat to make the close call as unsuccessful. This only amounts to a postponement of adminship, and many candidates will come back in a few months and have an RfA not in the close call zone next time. There is no real harm to postponing adminship if a consensus is not clear on a first RfA. However, if a candidate has had a series of close RfAs, nothing is likely to change in a few months, and it is all the more important that a close study of the discussion be made. So perhaps it's only the cases of continued stalemate over the course of at least 2 RfAs that we should be discussing, and it is not necessary to force a permanent decision on cases that are not stalemated over multiple RfA attempts. NoSeptember 09:10, 24 January 2007 (UTC)

We have too many inactive bureaucrats

At the moment we have 23 bureaucrats (not including Danny's office account). According to tbe bureaucrat's page (WP:BCRAT) 10 are active. Of these, I would say the most active (that is, performing bureaucrat actions) are Essjay, Redux and Taxman, who I personally think adequately get the job done fine. The rest, particularly the inactives do not need to be bureaucrats. I'm saying this, obviously due to today's events, that Raul654 promoted Ryulong, and his last promotion was of Messedrocker, way back in September. The tools are there to be used regularly, not occasionally; my point is, users like Raul, Danny, Angela etc. do not need the bureaucrat tools. In particular, all three were promoted years ago, when times were different. Bureaucrats, like stewards, should only have the tools if they actually use them. My proposal is all relatively inactive bureaucrats must go through an RfB again if they want to retain their powers. --Majorly (talk) 17:02, 24 January 2007 (UTC)

Just wondering, we have around 300 inactive administrators too, as far as I know. Should they be kicked as well? -- ReyBrujo 17:06, 24 January 2007 (UTC)

And WP:DUU90 (which gained consensus, but was rejected for technical reasons) can be seen as a method of removing 'autoconfirmed' from inactive autoconfirmeds... --ais523 17:09, 24 January 2007 (UTC)

But with crats, we're not talking about hundreds or thousands of people - we're talking about twenty tops. Re-confirmation for admins is completely impractical as it would flood RFA and it would result in admins going out of their way not to tick anyone off. But with bureaucrats, you're talking about a small number and a position of extreme trust. Asking those who are not regularly active and thus perhaps not up-to-date on the current evolution of RFA to either stand for reconfirmation or retire is not out of bounds. --BigDT 17:14, 24 January 2007 (UTC)

• Angela, Raul, et al. fine, I'll heartily agree with that - but Danny needs to be able to change names or otherwise do stuff on a dime in conjunction with WP:OFFICE. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 17:07, 24 January 2007 (UTC)

  I disagree with kicking out the semi-active ones. I mean, if we did that with admins there would certainly be problems. Ones that haven' done anything or have been gone forever can be demote and have to go through the process again, I have no problem with that.--Wizardman 17:09, 24 January 2007 (UTC)

ReyBrujo, yes admins as well, but I think that's been proposed before. Inactive admins (particularly ones who have not been around for over two years would be too out of touch to start again. --Majorly (talk) 17:10, 24 January 2007 (UTC)

Angela and Danny are both stewards are they not? If they did need powers here aren't they enable to set the flag +crat at meta? thanks/Fenton, Matthew Lexic Dark 52278 Alpha 771 17:11, 24 January 2007 (UTC)

(edit conflict) Or for that matter, +sysop, +bot, or -bot; I think stewards can also rename users, so there's nothing a 'crat can do that a steward can't. --ais523 17:15, 24 January 2007 (UTC)

Actually, thats untrue. Stewards have changerights which can set any user to any rights, whereas bcrats have makesysop and renameuser - the second isn't included in the steward package. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 17:31, 24 January 2007 (UTC)

You're right; I was confused by the 'rename' section on m:RFP. --ais523 17:51, 24 January 2007 (UTC)

Promotions on enwiki should be done on enwiki. We want the information in our bureaucrat log, not on the meta log. And remember that Redux and Rdsmith4 and also both enwiki bureaucrats and stewards. NoSeptember 17:43, 24 January 2007 (UTC)

• (edit conflict x5) Last I knew, the community had repeatedly turned down proposals for desysopping inactive admins, it should be no different for inactive bureaucrats. Besides, aren't Danny and Angela bureaucrats due to other portions of their status here? --tjstrf talk 17:12, 24 January 2007 (UTC)

Bureaucrats are different to admins. It's a different idea totally. --Majorly (talk) 17:13, 24 January 2007 (UTC)

Different? Howso? A quick buzz over to meta:bureaucrat gives no such indication. In fact, the entire status is rather nebulous seeming aside from that they take a bit of load off of the stewards in promotions and name changes. If you wanted to determine the original intent here, I'm thinking you'd need to go dig through foundation mailing list posts someplace. --tjstrf talk 17:28, 24 January 2007 (UTC)

Bureaucrats and admins are not different. They can do one or two important things that adminstrators cannot. In fact, the Spanish Wikipedia does not have admins, everyone is either editor or bureaucrat. -- ReyBrujo 17:46, 24 January 2007 (UTC)

Right well this is the English wikipedia. Bureaucrats here are different: if they were the same, all admins would surely be able to promote as well? Or, no one would be able to and we'd have to rely on stewards. --Majorly (talk) 17:54, 24 January 2007 (UTC)

I am just demostrating that, if Administrators do not require "confirmation", neither Bureaucrats. And since many proposals to remove admins have been turned down, I don't see why 'crats should be different. If you want to be touchy, the English Wikipedia is different from Meta, where you need to be administrator in another Wiki in order to be administrator there. Somehow the "many inactive, remove them" mantra sounds like a way to have more possibilities for a future RFB. -- ReyBrujo 18:06, 24 January 2007 (UTC)

Bureaucrats are admins with three extra buttons. If these aren't used, they do not need them. Admins, however, are in constant shortage. When was the last time there was an RfA backlog? Or any 'crat backlog at all? You shouldn't be comparing bureaucrats to admins because they just aren't the same. --Majorly (talk) 18:10, 24 January 2007 (UTC)

The community trust users when they request 'crat powers. They say why they want them and the community determines whether they are worth or not. We don't accept or reject based on "active/existing" rate. Otherwise, we would just remove the RFB section until someone steps down. Heck, in my own RFA I stated I would not use the tools extensively, only to deal with some backlogs and nothing more, and made it crystal clear that I would stay as WikiGnome, yet the community thought I would not misuse them and accepted me. One thing I learned from conversations with others: Contrary to Meta or Commons, where your tools are removed if you are innactive, here you get the tools because you are trustworthy, you have an immediate need of them, and because the community approves [of] you. I want to see you falling with all your rage in the first admin that comes back after a wikibreak for deleting an article when he was at the "bordeline" consensus and he was in the gray area to decide. -- ReyBrujo 18:28, 24 January 2007 (UTC)

Yeah, they were trusted when they were promoted. --Majorly (talk) 18:33, 24 January 2007 (UTC)

And inactive bureaucrats are not in any immediate need. --Majorly (talk) 18:35, 24 January 2007 (UTC)

Regardless of where the "gray area" lies, Mark didn't utilize his bureaucratic discretion to determine that consensus had been reached. He determined that consensus had not been reached and promoted Ryulong because he liked him (citing his personal opinion of the candidate as "the deciding factor"). —David Levy 18:38, 24 January 2007 (UTC)

I will refer you to my comment almost at the end of a section above: he was in the gray area (according to his own perception) and used his personal knowledge about the topic to close the RFA. Since bots don't promote admins, I guess it is fine. If you want to change the way 'crats are chosen, create a RFC or a proposal at the VP. Discussion is good, but remember, to reach consensus, everyone must give up something. Postures stating "Demote everyone, reconfirm every year, assume bot mind" are not open for discussion apparently, which makes this discussion a good way to release stress and dispair, but achieve little. -- ReyBrujo 18:51, 24 January 2007 (UTC)

1. It is not a closing bureaucrat's job to "[use] his personal knowledge" to decide whether a sysop candidate is qualified. It's his (or her) job to gauge community consensus and respond accordingly. Discretion plays a role only in interpreting the discussion's outcome, not in creating it. Mark's personal opinion of Ryulong is worth no more than that of any other Wikipedian in good standing.

2. What gave you the idea that I wish to change the bureaucrat selection process or demote anyone? I've said nothing of the sort. —David Levy 19:48, 24 January 2007 (UTC)

And no, Danny uses his Dannyisme account for office use. I don't know about Angela. --Majorly (talk) 17:14, 24 January 2007 (UTC)

We should propose more bureaucrats :) HIZKIAH (User &#149; Talk) 17:25, 24 January 2007 (UTC)

No, we don't... that's the whole point of this discussion. --Majorly (talk) 17:31, 24 January 2007 (UTC)

This seems to be a solution looking for a problem. If there is a particular user you don't trust with bureaucrat tools, you should take it up with that person. Jkelly 18:15, 24 January 2007 (UTC)

Changes in the importance of voting and discussion on RfA and what bureaucrats do

I think the above discussion is another example that shows that the community doesn't really know what RfA should be like or even what it is like. The two possible extremes I see are:

1. Pure (open) voting, bureaucrats just remove SPA votes, test whether the support percentage exceeds a certain level, and promote or not based on that number (i.e. the function of the bureaucrat is mostly technical, with very little judgmental skills required)
2. Pure discussion where advantages and disadvantages of promoting a certain candidate (including acceptance in the community, which we usually measure by number of support pseudovotes) are discussed by the community. The bureaucrat then decides carefully whether promoting the candidate will be a net benefit (this includes considering the social impact of annoying the people who opposed the candidate) and whether there is potential for abuse of the tools, and then promotes (i.e. the bureaucrat actually makes the decision, so needs to have great knowledge of what is required of an admin, be able to judge community reaction etc.)

Actual practice is somewhere inbetween. We ask that bureaucrat candidates have the skills necessary for the second kind of system, yet expect them to follow the first system in practice, with a very small range for discretion. Much of the support section of a typical RfA consists of votes, much of the oppose section of actual discussion (either of what the candidate actually did, or of why that oppose reason is silly). However, people's RfA contributions have changed over time, and perhaps we should acknowledge that process change.

If we look at old RfAs from 2004, RfAs were a lot more vote-like than now. Old RfAs were not so much discussions, and many people passed with a three-line nomination statement and not much further information from the support voters. As RfA has changed and discussion has become more important (and typical support and oppose counts have changed), our criteria for promotion should change as well. As it is insanely hard to become a bureaucrat these days, only the most trusted users can become one, so we should trust them to judge how to close RfAs and not ask for pure vote-counting. If we don't want to allow them this discretion, we should return to a more explicitly vote-like RfA system without a discretion range. In any case, we shouldn't do both - it is silly to pretend that bureaucrats need an insane amount of trust while at the same time giving them only the tiniest of discretion ranges so their judgment skills are never actually needed. (note: I know that bureaucrats have used their discretion to extend RfAs, but that usually brings in as many oppose as support voters, so extending an RfA often means to decide it should fail but to pass the responsibility elsewhere). The thing our bureaucrats should do at the end of the day is to make sure we get enough decent admins that help everybody build the encyclopedia, and the RfA process should do that in a way that causes as little friction as necessary. If we can't decide whether we want a vote-counting or a more actively deciding bureaucrat, we will continue to have problems finding people who can actually pass RfB. Kusma (討論) 17:25, 24 January 2007 (UTC)

I agree with Kusma. Well said. However we should have more bureaucrats. --HIZKIAH (User &#149; Talk) 17:27, 24 January 2007 (UTC)

Sorry Kusma, I don't understand what you're getting at. What's the point of your argument? And, HIZKIAH, we don't. See the thread above. --Majorly (talk) 17:30, 24 January 2007 (UTC)

My point is that we should allow more bureaucrat discretion, as that is what is needed while RfA is evolving from a vote to a discussovote. Kusma (討論) 17:34, 24 January 2007 (UTC)

In that case I agree, only if the bureaucrats get reconfirmed every year, in the style of stewards. We have too many bureaucrats doing nothing. --Majorly (talk) 17:37, 24 January 2007 (UTC)

Sure, they can all reconfirm each other every year if that makes you happy. I doubt it would change anything though. --tjstrf talk 17:43, 24 January 2007 (UTC)

Hmm I don't know what you mean there. I mean they are confirmed by everyone else in an RfB. --Majorly (talk) 17:50, 24 January 2007 (UTC)

That would be fun. I don't think many would pass (hint:RfB is broken). Kusma (討論) 17:54, 24 January 2007 (UTC)

An election style would be better, like they do on Meta. --Majorly (talk) 17:55, 24 January 2007 (UTC)

Stewards reconfirm each other. They don't go up for public vote every year, they go up for confirmation by their peers. If Bureaucrats were to be grouped with them and treated the same way, then they would similarly be expected to reconfirm each other. --tjstrf talk 17:59, 24 January 2007 (UTC)

• It almost sounds like we either have to change it bureaucrat voting bots, or no support/oppose votes at all... We're right in the middle now though.--Wizardman 17:57, 24 January 2007 (UTC)

tjstrf, what I'm saying is they should be confirmed, in a vote, by any users if they should remain a bureaucrat. We only need 3 in reality, so whoever has the highest three % stays a bureaucrat. --Majorly (talk) 18:03, 24 January 2007 (UTC)

My only problem with that is if we cut it off at 3, that's gonna be seen as a status symbol. If it's the right 3 that shouldn't be a problem, but I'm just sayign that for reference.--Wizardman 18:07, 24 January 2007 (UTC)

3 would just be the number we cut down down to. After, any more can be added. --Majorly (talk) 18:13, 24 January 2007 (UTC)

Ah, okay then. There's very few active anyway so that should work.Wizardman 18:14, 24 January 2007 (UTC)

• Having only three is in my opinion far too low. During the calendar year of 2006 we had 16 bureaucrats performing at least one RfA promotion. Seven of those had more than 10. Certainly limiting the number of bureaucrats to 3 produces Single-point failure problems. Limiting the pool of bureaucrats to such a limited set reduces consultation possibilities when faced with a potentially controversial decision. Further, I find arguments that we only need X number of bureaucrats unpersuasive. If a candidate is able to successfully pass RfB, they are trusted to perform the duties of a bureaucrat. Having bureaucrats handy is a good thing. The duties were spun off from stewards to help the project. Limiting the pool of bureaucrats significantly reduces the potential positives and goes against the original intention of the project in creating bureaucrats. Having a reconfirmation every year might not be a bad thing, but limiting them from even getting to that point is I think counterproductive. --Durin 18:16, 24 January 2007 (UTC)

• You read me wrong. I meant we demote the inactive bureaucrats and leave the most active (i.e. about 3, but the number doesn't matter). When/if there's a need for more, someone can apply. --Majorly (talk) 18:20, 24 January 2007 (UTC)

• Demoting the current bureaucrats implies there's a problem with them...because of this particular controversial RfA. That doesn't follow. The bureaucrats as a group have not demonstrated incompetence nor any compelling reasons to consider older bureaucrats as less capable. I'd prefer to see more evidence that older bureaucrats are less competent than more recently promoted bureaucrats before stripping ~20 people of the bureaucrat flag. And again, "need" is a bad basis on which to constrain the bureaucrat pool. If they are capable and trusted and can pass RfB, there's no reason to deny them because of a perceived lack of need. Single point failures can and do happen, and we saw a microcasm of that recently on Wikipedia:Changing usernames. Whenever User:Essjay takes a break from it, we get a backlog. There isn't anything wrong with having 15 bureaucrats if 10 are capable of covering all the bases, so long as all the bureaucrats are acceptably skilled. --Durin 18:34, 24 January 2007 (UTC)
  • I do agree with you very strongly about one thing - this is solving the wrong problem. The issue at hand is that a bureaucrat has the unchecked power to do whatever they want and ignore consensus. The solution to that problem isn't to de-crat 20 people who had nothing to do with this decision. The solution certainly isn't to introduce an AFD-style debate into the process. (Could you imagine having to sift through the unordered, verbose comments of 100+ people?) The solution is to lay down stricter standards for promotion. If 75-85 is the margin of discretion, then we need a rule for anything outside of that range. For example, prior to promoting someone with 74% support (or not promoting someone with 86%), a crat must discuss it with and gain the approval of three other crats. That way, there is a check in the event of puppetry, ballot stuffing, etc, but a single bureaucrat cannot unilaterally make the decision. Personally, I support the promotion in this case, but I think the much bigger issue is a lack of confidence in the promotion itself and a solution like this would help restore that confidence. Retiring crats is unrelated to the problem. --BigDT 19:14, 24 January 2007 (UTC)

• This is the solution that makes most sense to me. The vast majority of cases are uncontroversial in their closing and there's no problem, but in the few cases where there is doubt about the consensus, more than one 'crat should be involved (like for Carnildo's RfA). I disagree with the promotion and particularly with the reasons given, but to de'crat a bunch of unrelated people makes no sense. Trebor 19:21, 24 January 2007 (UTC)