Wikipedia talk:Requests for adminship/Archive 192

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 185

←

Archive 190

Administrator conduct in an Rfa

Am I missing something? I just took a look at the description of what an admin should be: Administrators are expected to lead by example and to behave in a respectful, civil manner in their interactions with others.

However, I am not seeing that in Coffee's continuing edits to the Rfa of Ronhjones. Where is the concern about requiring an existing administrator to follow their established standards of conduct? Or am I wrong to expect this? Coffee really got my attention when he said I "should take my head out of my ass"... (a conversation now found on the talk page of Ron's Rfa), and then Coffee's astonishing statement that he "hated" Ron's answer to a question and hoped he failed his Rfa "with all his heart", which led to my reacting, and his subsequent additionally questionable statements. Now I see still more vitriolic material posted, despite requests to give it a rest. I don't think it merits a noticeboard, but still... Won't someone take Coffee aside and get out the trout?

The larger issue, of course, is administrator conduct in an Rfa. I'd like to see some guidelines discussed, and perhaps established, to spare good faith Wikipedians who run in a Rfa the odd sensation of demonstrably uncivil behavior by the very people whose ranks he or she may join. Jus da fax 20:50, 24 November 2009 (UTC)

I like the way you took my comments out of context. But that aside, why should we make the standards for administrators any different than for regular editors? I've said this time and time again, the term "administrator" just mean someone with tenure who has extra tools, nothing more or less. --Coffee // have a cup // ark // 20:55, 24 November 2009 (UTC)

Not commenting either way on the conduct at hand, but don't we expect all editors "to behave in a respectful, civil manner in their interactions with others"?--Fabrictramp | talk to me 21:13, 24 November 2009 (UTC)

Please remain calm. Jusdafax, while several of the comments made are bordering on defensive here/there, I believe that there were simply some "blunt" edits which expressed strong feelings about the issues at hand. The issues of our BLP situation is a highly emotional one for many editors here, and I'd hate to see the passion for "getting it right" melt-down to any vitriolic rhetoric that may be regretted later. If we can all just tone it down a notch, perhaps we can avoid any further distasteful posts. Thank you. — Ched : ? 21:11, 24 November 2009 (UTC)

When dealing with people in article-space, where their buttons give them an air of authority, admins should maintain decorum at all times. However, the tools have no use at an RFA, nor do their words carry any more weight than normal editors. For all intents and purposes, their extra buttons are useless here, and so they should be held to the same standards as everyone else. I don't believe that the :higher standard" applies to RFA, where everyone is more or less equal. The Wordsmith^Communicate 21:15, 24 November 2009 (UTC)

(@Coffee) On the latter point you make I do agree with you, of course. Jimbo's comments to that effect, made in the early days of Wikipedia, were instructive; as I had never read the Wikipedia:Administrators page before. But I think there exists, rightly or wrongly, a feeling that admins should be held to a higher standard, and indeed, the present-day gauntlet of an Rfa seems to confirm that. By the way, if I have quoted you out of context, my apologies. Since Ron's Rfa currently stands in the crucial range of 75%, I feel it vital to give him as fair a shake as possible, and I again - sincerely and respectfully - ask you to join me in standing down regarding further comments on that currently sensitive page. Jus da fax 21:20, 24 November 2009 (UTC)

Don't worry I already put my stick away

. Look I respect you, I've seen you around a lot (you joined after I did), but what I didn't like was the way you responded to me in that RFA. You decided that my comments were humor that shouldn't be there, which caused me to make that original statement in the first place. I was deliberately being humorous, while at the same time being serious. Trying to lighten up the atmosphere, etc. As to the second comment, you did take that out of context, what I said was if Ron couldn't take more than one quick look over something like an AFD, then I knew, "with all my heart", that he shouldn't be an admin. I said nothing about the particular RFA, and wasn't trying to attack Ron. Now do you see where I'm coming from? --Coffee // have a cup // ark // 21:29, 24 November 2009 (UTC)

I echo Coffee here, that is indeed his sense of humor, blunt but humorous. ceranthor 21:37, 24 November 2009 (UTC)

*Bluntly beats ceranthor with a blunt baseball bat* --Coffee // have a cup // ark // 21:44, 24 November 2009 (UTC)

OK, fairly said. Operative phrase: "lead by example." (Salutes.) Jus da fax 21:48, 24 November 2009 (UTC)

Wait wut? You don't think I'm a good example? Darn, and to think I always wanted there to be "little Coffee cups"

. --Coffee // have a cup // ark // 21:51, 24 November 2009 (UTC)

Statistics page

Is there a page anywhere of statistics of Successful RfAs to Unsuccessful RfAs? --The High Fin Sperm Whale (Talk · Contribs) 05:12, 26 November 2009 (UTC)

User:NoSeptember/The NoSeptember Admin Project might help. –Juliancolton | ^Talk 05:13, 26 November 2009 (UTC)

One more thing, is it possible to be de-administratored without getting blocked? --The High Fin Sperm Whale (Talk · Contribs) 19:15, 26 November 2009 (UTC)

Yes; see WP:FORMER. –Juliancolton | ^Talk 19:18, 26 November 2009 (UTC)

Nom

I know there has been a lot of controversy at various times over people voting before the nominee has accepted the nomination, and I was thinking that a simple change in the nomination procedure would stop all that. South Bay (talk) 07:09, 26 November 2009 (UTC)

And that change is? (X! · talk) · @441 · 09:35, 26 November 2009 (UTC)

I am not aware of such controversy. If the nominee hasn't yet accepted the nomination, he/she will most certainly not have answered the standard questions. The RFA shouldn't be transcluded before the nominee is done. That's the current procedure and if it's done any different, then it's because people don't follow it. So I really wonder what you wish to change about this order of business. --Atlan (talk) 09:39, 26 November 2009 (UTC)

I could be wrong but I don't think the OP is questing that they shouldn't be transcluded before the nom is done. The question, I believe, is whether it's okay for people who have it watchlisted to vote before it's transcluded to the main page. I could be wrong, though, that's just how I read it. Regards, --—Cyclonenim | Chat 16:25, 26 November 2009 (UTC)

I'd be okay with people supporting before it's transcluded, but opposing beforehand, even if someone is 100 percent sure that they will oppose the candidate no matter what, strikes me as cruel. Though I think both of those should be avoided in general. (The only case of purposeful pre-transclusion support !votes that I'm aware of is Wikipedia:Requests for adminship/Maunus, and the impression I get is that the RfA was planned out well ahead of time. -- Soap ^Talk/_{Contributions} 16:29, 26 November 2009 (UTC)

I think that jumping the gun with supports, like having an excess of conoms is liable to be counterproductive. I'm less concerned about Opposing before transclusion. I can only think of one very controversial candidate whose RFA sat untranscluded for a while and attracted a lot of opposes, there may have been more - I suppose the advantage of opposing before tranclusion is that the candidate still has the option of not transcluding. Ϣere SpielChequers 16:45, 26 November 2009 (UTC)

I personally feel that the only supports that should be made prior to inclusion are ones from nominators. However, if the candidate has accepted and signed, then I feel it should be almost officially open, even if not yet transcluded (as it will most likely be transcluded very soon after the candidate has accepted/answered compulsory questions). Regards, --—Cyclonenim | Chat 18:24, 26 November 2009 (UTC)

As an aside, see Wikipedia:Requests for adminship/Can't sleep, clown will eat me too, for the concept of pre-transclusion support taken to extremes. Graham87 03:01, 27 November 2009 (UTC)

The Drought at RFA

Just in case there's anyone left who doesn't understand why many of us who watch the RFA process believe it to be broken, I've updated User:NoSeptember/Admin stats#Year to year comparison of promotions by months, copied it here and colour coded it.

Month	2003	2004	2005	2006	2007	2008	2009
December	15	24	68	19	34	9
November	9	27	41	33	56	11	13
October	10	16	67	27	27	16	7
September	17	29	32	22	34	6	8
August	9	12	39	26	18	12	11
July	11	17	31	26	31	16	10
June	24	13	29	28	35	18	12
May	10	23	17	30	54	16	12
April	6	20	25	36	30	12	14
March	8	31	16	34	31	22	13
February		14	9	28	35	27	9
January		13	15	44	23	36	6

Months with an average of 1 or more successful RFAs per day are in green, and those with less than 0.5 a day in red. I read this as showing a flow of circa 1 new admin a day from Mid 2005 to march 2008, then something happened and it has subsequently oscillated between half and a fifth of that rate. Ϣere SpielChequers 13:02, 11 November 2009 (UTC)

Let us note, however, that we are not substantially producing less new admins now than we were in 2003. Between March and October '03, there were 95 new administrators. Between March and October '09 there have been 89 new admins. So in total, we're still producing about 93.68% of the administrators we were then. If the project didn't fail under those circumstances, it seems unlikely that a 6.4% drop in admin rates is going to kill the system entirely. We need to figure out what occurred between '04 and '05 which made the process more favorable to candidates, and what happened since March '08 which seemingly threw us back to the dark ages of admin promotion. I'm probably not being logical here, just the way I see things. Regards, --—Cyclonenim | Chat 14:27, 11 November 2009 (UTC)

Yes but we have ten times as many articles and many times as many edits per day now than we did in 2003. For the moment we also have far more admins than we had in 2003, but the trend is clearly downwards. Ϣere SpielChequers 15:11, 11 November 2009 (UTC)

Note: I added November to the table, since it's now November 26 and there can be no more successful RfA's closing in November. -- Soap ^Talk/_{Contributions} 23:41, 26 November 2009 (UTC)

If anything the above would support a conclusion that, rather than experiencing a famine in terms of new administrators now, we simply experienced a feast of them in the 2005-2007 timeframe and that we've simply returned to previous levels. The overall activity on Wikipedia (gauged by overall number of edits) also seems to have peaked in 2007 and has leveled of (and even slowed down somewhat) in the meantime. One possible conclusion is that the steady increase in the admin corps in 2007 came in response to the rapidly increasing activity on the project as a whole, and when that leveled off, the rate of new adminships went down in response. However, part of the problem I am seeing is that we are coming up with a lot of charts, a lot of theories, but no real information. We are seeing what is happening but making very little progress as to the why and how to fix it. I personally suspect that the drop in successful adminship candidates is a reflection on a recent trend toward being increasingly demanding of our candidates; we are now looking for more in terms of edits, more in terms of activity, more in terms of "content building", more rigorous in our questioning, and less forgiving of minor errors. I would venture a guess that if the same generic standards applied to "modern" RfAs were in place back in 2007 the above chart would see a lot less green. Just my opinion of course. Shereth 16:16, 11 November 2009 (UTC)

I find no problem with having demands for content building and for more rigorous questioning of candidates; I do have a problem with being less forgiving of a candidate. Many of the editors I would like to have seen become admins would fail over that part due to past mistakes, one's that I'm willing to forgive and forget but that others aren't. The problem is, however, that there's no way to prevent this. We can't introduce a policy saying that the past is the past—people are still entitled to not forgive. I feel it would be detrimental to the project, too, if we prevent further questioning of candidates and prevent opposes over content building. All that said, what's the solution? We've been floating around with graphs for the past couple of months, but so far I see no actual constructive proposals to fix the problem; all I hear is that there is one. I sure as hell can't think of a solution either, so don't look in this direction. Regards, --—Cyclonenim | Chat 16:34, 11 November 2009 (UTC)

I'm not trying to say that increased scrutiny of new candidates is bad; nor am I saying that tougher requirements is a bad thing per se. All I am suggesting is that these are likely factors contributing to the decline in successful administrator candidates we've seen over the past couple of years. The harder we make it to pass RfA, the fewer candidates we are going to see pass, and that is an inescapable conclusion. We cannot have one without the other, and there is no solution to this "problem". Either we must accept the fact that we will have fewer and fewer active administrators (until an equilibrium is reached and our admin pool levels out) or we have to, as a collective, alter the way we put these candidates through the gauntlet to achieve a higher rate of passage (as well as making the process less onerous to prospective candidates). I agree with you that this sort of thing cannot be legislated in any fashion; it must be a conscious effort on the part of the community to ease up on candidates a bit if we want to see more pass. Shereth 16:44, 11 November 2009 (UTC)

It was more of a general comment than I reply, so I probably shouldn't have indented. I think it's a very natural thing for any process to become stricter, since whenever an admin goes rogue, it makes the community want to enforce some stringent guidelines for the next candidate. It only takes a certain number of these situations to create an intricate, complicated system like we have now in which it becomes difficult for anyone other than the perfect all-round candidate to pass. The community needs to accept that nothing is perfect in life, and that there will always be the odd person to go to the dark side. Ultimately, that can be anyone from someone who got 100% at RfA, to someone who passed on the borderline of 70%. I guess the only way for the community to accept this, and for it to become less of an issue, is for a better desysopping system, but that's not coming any time soon. Regards, --—Cyclonenim | Chat 16:56, 11 November 2009 (UTC)

<-Thanks for the data on article creation and edit counts. Both indicate remarkably stable activity:

Edit counts between 200K and 250K per day since mid-2006
New articles at 1600 a day since mid-2005.

When I look at admin stats, I see two issues

Admin creation rate down
Admin active rates dropping significantly.

In round numbers, we added 100 new admins in 2009, but the active count is down 100, which means 200 admins active at the end of 2008 are no longer active. That's well over a 20% annual attrition rate. While the desysoppings are painfully notable, they are a drop in the bucket.

It isn't clear to me that these are the same issue. As I and others have noted before, an increased hurdle to adminship may exist, but it doesn't fully account for the reduction in the creation rate. More importantly, I see no connection between the entrance requirements and the decisions of admins to walk away.

I think we have to address both issues, but I am presently more concerned about the attrition rate (unless someone can show me I mangled the data, or missed some plausible reasons that aren't alarming.)--SPhilbrick T 17:20, 11 November 2009 (UTC)

I agree that attrition is very high compared to offline volunteer projects; But I had heard that this was a widespread feature of online communities, and I'm not sure if its better or worse for admins as opposed to non-admins. I've tended to focus here on the lack of new admins because this is the talkpage for RFA, but there is a lot of talk on strategy about community health, maybe strategy:Participation/Attracting new participants and retaining existing participants would be a good place to raise this? Ϣere SpielChequers 02:13, 12 November 2009 (UTC)

We're getting somewhere with some of the stats, but I'm still questioning if there really are two "issues". Admin creation rate and admin active rates don't tell the whole story. What if the existing admins are doing work at a greater rate than necessary to keep up with the loss of admins? What if the bots and edit filters are taking up some of the slack? The objective evidence doesn't necessarily point to an issue; just this week I saw CSD empty and the PROD queue up to date (its only contents were < 7 days old). I rarely see AIV more than 5 vandals deep, and furthermore, I would estimate that at least 10% of reports to AIV don't wind up in an immediate block anyway. And, let's face it, DYK doesn't need to have 4 rotations per day. It isn't outside the realm of possibility that - especially with constant edit rates and new article rates mentioned above - the current crop of maintainers is actually sufficient. Frank | talk 02:27, 12 November 2009 (UTC)

Category:Temporary Wikipedian userpages is down from its peak of 31,000 pages to a stunning 26,000 pages and Wikipedia:Possibly_unfree_files is only backlogged to July 2009, so I might disagree with your assessment as to currency. MBisanz ^talk 02:30, 12 November 2009 (UTC)

Well, that looks like a 15% decrease to me :-) But seriously, folks...is there any indication that adding 500 admins would change either of those stats? Frank | talk 02:35, 12 November 2009 (UTC)

That, I think, might have more to do with the unrewarding nature of the tasks than the number of volunteers at work. It's hard for me to understand from the page Category:Temporary Wikipedian userpages how activity there helps Wikipedia achieve its mission. A homeless shelter can recruit a thousand volunteers but that doesn't mean the toilets will get clean - tasks very distant from the mission of bringing everyone free access to the sum of all human knowledge are always going to be slow going. (Unless of course they can be done by non-admins to build up their edit count in preparation for RFA.) Christopher Parham (talk) 18:05, 17 November 2009 (UTC)

As the RfA climate has become more and more exacting and hostile, fewer and fewer people want to endure the ordeal (personally, I'm always up for a run, but most people don't have such thick skin). I remember my first RfA in May 2004: I passed almost unanimously with virtually no fuss, and that was pretty typical for the time. I'd suggest two tendencies have contributed to the change: first, we've had lots of problems with abusive admins, yet we've never developed a way to effectively deal with abusive admins, and adminship is still perceived as a lifetime appointment; secondly, the community has become larger and the political dynamics have grown more complex, leading people to attribute more importance and prestige to adminship as part of the stratification of the community. The two basic remedies to this problem would be to institute a community desysopping procedure and to reduce the level of approval required to pass RfA; an ultimately more meaningful remedy would be to reduce the importance of individual admins by expecting them to make controversial decisions collectively rather than unilaterally. Of course, even if reforms like that were somehow accomplished, you'd still have residual high RfA standards that have been ingrained into the wikiculture and might take a long time to fade away. Everyking (talk) 04:33, 12 November 2009 (UTC)

I'm not a fan of the idea that desysopping is or should be connected to RFA, other than this process learning of any candidate behaviour to oppose over because it predicts that the candidate would be a poor admin. But if it were linked then I see the "can't be as bad as current admins" support as a more sensible approach than "oppose, has slight imperfection". Not least because appointing many more good admins would dilute the importance of any bad ones we have, whilst rejecting well qualified candidates and having a declining number of admins increases the effective influence of any bad ones we may have.

However I think that the backlogs of admin work may not be the best way to measure the problem. This is a volunteer community, and the idea of adminship is that experienced cluefull editors get the mop so that they can do some stuff themselves, and hopefully spend some of their time here doing chores like speedy deletion. If we have fewer admins to share the admin load we may see longer backlogs at some relatively low priority admin tasks. But I suspect we are more likely to see our declining number of active admins devoting an increasing proportion of their time here to doing admin chores. In my view this has several downsides:

The less time admins spend doing non admin stuff the more they risk becoming detached from the general editing community.
We risk creating psychological pressure to cut corners in order to keep up with the flow of stuff needing admin attention. When I go to CSD I focus on certain subcategories, and rarely delete or decline to delete more than half a dozen articles in one visit. One of the things I suspect we are seeing at Wikipedia:Newbie treatment at CSD is that some admins see a backlog of 100 or more at CSD as something to clear in an hour or two. As our number of active admins falls I fear we will see an increasing proportion of admin errors as a dwindling number of admins try to keep up with the demand for admin actions.
Burnout or admin burnout may increase, one of the things I most enjoy here is reviewing at FAC and I don't want my being an admin to totally takeover and stop me being there. If being an admin meant only doing admin stuff I would probably hand in my mop.
Firefighting at AIV, CSD and the Drama boards is likely to get priority over longer term solutions such as finding new candidates for RFA.

So I would suggest that an increasing admin error rate, increased criticism of the admin cadre as being detached from the community and less non admin contributions by admins would be better indications of an admin shortage than backlogs at AIV or Cat speedy. Ϣere SpielChequers 14:35, 12 November 2009 (UTC)

Concerning Everyking's first point, we don't really have more than anecdotal evidence that we get less RfA candidates because less people want to run the gauntlet. I believe it plays a role but really, this is just speculation and my hunch is that it's not the main problem. I don't really see any way of instituting a productive desysoping mechanism. In fact, I'm worried that it would reinforce the perception that admins are above the law. After all, some ArbCom decisions include stuff like "Administrator X is admonished for his egregious use of the tools and profound dickishness" which, all kidding aside, may often be the right way to fix the problem but certainly promotes the idea that you have to screw up really bad to lose the bit. But Everyking's second remedy ("ultimately more meaningful remedy would be to reduce the importance of individual admins by expecting them to make controversial decisions collectively rather than unilaterally") is what we should shoot for. This is not even so hard to do. We all tend to excuse admins we know and like when they cross the line but do so for what we see as legitimate reasons. We shouldn't. Next time your buddy blocks a troll who's harassing him, let him know that the block should have come from somebody else. Some of you may remember that the ANI page used to say "this is not the Wikipedia complaints department" but I sometimes think a department of complaints about admins would be a good thing. You go there, you make a short statement of the issue and if your complaint is legit than someone just goes and whacks the admin with the trout. That would be much healthier than an ANI thread.

But the real priority is the concern raised by WereSpielChequers. On one hand, the dwindling number of active admins is forcing the remaining admins into less and less article work, disconnecting them from the ultimate purpose of the project. On the other hand, RfA increasingly rejects candidates that aren't spending 10 hours a week editing the wiki, rejects candidates that avoid ANI, rejects candidates that aren't really interested in the finer workings of the wiki-bureaucracy. Now, we're still promoting people who turn out to be great admins but they're all in the same mold and are all people who spend way too much time in the non-article space (myself included). Just look at Wikipedia:Requests for adminship/Hawkeye7 where people are actually arguing that you can't be an admin if all you really want to do is stop bugging admins.The result is that the admin corps is slowly disconnecting itself from content editors and rapidly increasing the perception of that disconnect. Pascal.Tesson (talk) 23:27, 12 November 2009 (UTC)

Agreed, isn't that attitude another incarnation of the generally discredited "doesn't need the tools" argument? delldot ∇. 15:34, 13 November 2009 (UTC)

Does anyone happen to know if the number of candidates attempting RfA's has decreased in proportion with as the number of those who succeed? In other words, are we promoting fewer admins because fewer candidates run, or because we've gotten tougher, or both? (I suspect it's a combination of both but it could be that one factor is small enough to ignore. I would say that the pool of editors who could probably pass RfA based on current standards but don't want to is fairly small, as well.) -- Soap ^Talk/_{Contributions} 19:09, 13 November 2009 (UTC)

The stats themselves would be easy to compute and they might already exist somewhere but in any case, it would be hard to draw any of the conclusions you're looking for. For one thing, the stats are pretty fluctuating because of small samples and the rate of failure is largely a function of how many completely inadequate candidates run RfA without realizing their chances are 0%. Maybe we're just going downhill because there are just so many people on the planet that are even interested in being a Wikipedia administrator... Pascal.Tesson (talk) 21:22, 13 November 2009 (UTC)

Month	2003	2004	2005	2006	2007	2008	2009
December	15	24	68	19	34	9
November	9	27	41	33	56	11
October	10	16	67	27	27	16	7
September	17	29	32	22	34	6	8
August	9	12	39	26	18	12	11
July	11	17	31	26	31	16	10
June	24	13	29	28	35	18	12
May	10	23	17	30	54	16	12
April	6	20	25	36	30	12	14
March	8	31	16	34	31	22	13
February		14	9	28	35	27	9
January		13	15	44	23	36	6
Total Promoted	119	239	389	352	408	201	102
Total Unsuccessful		63	213	543	512	392	203
Total RfAs		302	602	895	920	593	305

Based on the information I added to the table, I believe soap is right. It's not just that not many people are getting promoted (although, in the past four years more people have been unsuccessful than successful), it's that not very many people are running. It started in 2008. The total RfAs in 2008 were 327 less than in 2007. That's not good. This year, we're already in November and, up to the end of October, there have only been 305 Rfas in total. And there have only been 5 RfAs so far this month. Pascal.Tension, I agree that the stats do fluctuate but I also believe that a drop of 327 RfAs in a year is not a small fluctuation. — Oli ^OR Pyfan! 21:57, 13 November 2009 (UTC)

I also believe that some people would like to be Administrators but don't want to go through the highly daunting (and stressful) process of RfA. You just have to look here to know that quite a few people still want to be administrators. All this boils down to one point I want to make. I think that, to get more administrators, we should make the RfA process seem more relaxed (I'm not saying we should let more people become administrators). We should have friendly constructive criticism, for example something as simple as, "I'm sorry, but I think you should do some more work in areas such as XfDs" instead of "not enough edits" or "per Majorly". — Oli ^OR Pyfan! 22:15, 13 November 2009 (UTC)

Of course, but I think the only way to make the process more relaxed is to make actual changes to the system in order to promote a shift in how people view RfA and adminship in general. When we have so many people involved and a tendency so deeply ingrained into the project culture, a mere appeal for everyone to "play nice" isn't going to have any effect at all. Everyking (talk) 05:54, 14 November 2009 (UTC)

Unfortunately, the best way to change the way people view adminship is to have a lot more admins so that the power of any individual admin is decreased or at least appears to decrease. I think we'll eventually go for a trial adminship or limited adminship system. The time and effort put in to the RfA system would be better spent monitoring new admins and advising them on ways to improve. Save a few stubborn morons, most editors and admins listen to good advice. Pascal.Tesson (talk) 17:26, 14 November 2009 (UTC)

I'm new here and was just looking at the RfA process and I have to say why would anyone want to bother? I was thinking that if I wanted to be an admin all the hoops to jump through would turn me off. Knowing all the rules of Wikipedia is like studying law. But that for me would be the easy and fun part. What would annoy me is the mindnumbing repetitive stuff that they seem to require of admin candidates. CSD? High % of edit summaries? What does that have to do with writing quality articles? I notice I have a less than 50% edit summary utilization, yet before I sought to check, I thought I was using it appropriately. It seems minor edits I didn't provide an edit summary for brought down my edit summary % dramatically. Do I need to provide an edit summary for this comment I'm making right now for example? To be an admin you have to prepare months in advance paying attention to such details which detract from stuff like actually providing useful edits to an article. Lambanog (talk) 19:56, 14 November 2009 (UTC)

Well, you answered your own questions. You do not need to jump through hoops to be an editor of articles. If that's all you want to do, then celebrate; no RfA in your future. But, if you want to close contentious AfDs, determine the notability of borderline BLP articles, work with the legality of image use, or block users for violating a myriad of policies, then you need to demonstrate that you know those policies, and use them appropriately. Should you provide an edit summary for the comment you just made? Yes. You should. Tan | 39 20:00, 14 November 2009 (UTC)

Then again, you also do need to have some article writing experience in order to pass an RfA... I didn't have enough when I last ran, and if I were to run again, I'm frankly still guaranteed not to have enough article writing experience, because I'm not an article writer. But with the current RfA crowd, there needs to be... a general balance of content, discussion, and maintenance experience. Audited content work is something that I've seen pop up lately as well. The thing that should not be 20:53, 14 November 2009 (UTC)

Well different people have different criteria. I believe that you should have tried out all the buttons a few times and at least had a look at the policies and forums so that you can talk to users about their issues. Graeme Bartlett (talk) 22:00, 14 November 2009 (UTC)

Actually, Thing, the soon-to-be successful Rfa of Amorymeltzer clearly shows that article writing experience is currently no big deal with the !voters. As one of the very few voting against, I will henceforth no longer make an issue of such things in an Rfa, as the trend of consensus appears to be overwhelmingly against a GA, FA or new article creation being of much concern as a qualification for new admins. And while I personally think this is a big mistake, it may have the effect of 'lowering the bar' enough to bring in a new surge of admins, which is both ironic (re: what happened to you in your August Rfa - have things changed that much in 3 months?) and moots the point of this discussion. I have absolutely nothing against Amorymeltzer personally and indeed feel he will be a fine admin in important yet limited areas, but rightly or wrongly I see his promotion as a big change in current standards, the long-term effect of which may well be regretted in years to come. Jusda fax 22:46, 14 November 2009 (UTC)

We've been promoting candidates with little article creation experience for quite a while now, and the number of promotions remains low. I'd link to examples, but I would rather not single out particular RfA's when there are many to choose from. -- Soap ^Talk/_{Contributions} 23:05, 14 November 2009 (UTC)

If so, Thing/Sleeps was dealt with rather harshly, in my view. Lack of content work seems to have been the major objection. Jusda fax 23:12, 14 November 2009 (UTC)

In response to Tan: don't shoot the messenger!

I make the observation as someone who hasn't turned on automatic prompt for edit summaries but who in good faith tries to include meaningful edit summaries when I feel they are called for. Correct me if I'm wrong but I could turn on the prompt and add a next to meaning less "x" character on my edit summaries every time and over time it would dramatically change my score on the Edit Summary calculator although I would argue my current practice gives far more pertinent summaries. You might of course say people approving RfAs should drill down and not rely exclusively on a tool like the Edit Summary calculator. Fine but what really are you going to see? A sea of curt edit summaries which are really little better than the "x" character in most cases. I notice in your edit summary for your comment you placed "reply" and use it probably almost as often as I leave a blank space. I could conceivably recreate a similar looking edit summary history as yours if I was allowed to automate replacing a blank edit summary with the word reply. A more logical and efficient system to help assess the quality of edit summaries would be a complaint system where instances where edit summaries were not sufficient are tagged or flagged by other users maybe via a checkbox or button. Reviewers can then go directly to the identified case to verify. As it is lots of keystrokes are wasted on generic "reply" edit summaries which while fostering the habit also tends to make the comments more perfunctory. Lambanog (talk) 23:28, 14 November 2009 (UTC)

I am willing to cede some of the edit summary argument to you. If a <100% edit summary usage is why someone is opposing an RfA, I heartily agree this is a shoddy reason at best. However, your original point wasn't solely about the edit summary issue - maybe I should have noticed that you weighted that portion of your original comment heavily. That said, when parsing a candidate's contribs, a lack of edit summaries is extremely annoying. Yes, stating "reply" might seem useless - but when there is nothing there, what was going on? An epic, industrial-grade flamewar? An edit-war? The addition of uncited material? Or even a positive thing - helping a newcomer? I don't know unless I actually check the diffs. Of course someone could hide behind generic edit summaries, but this is very rare. The bottom line for me is that edit summaries are extremely helpful for anyone reviewing your edits. If you don't plan to submit yourself to RfA, then don't worry about having A+ edit-summary usage. However, if you do plan on this, your contributions will be heavily scrutinized.... and people will be frustrated by a lack of transparency. It's just the way it is, and as someone who does do the scrutinizing, I side with the generic "reply". :-) Tan | 39 23:41, 14 November 2009 (UTC)

One of the misconceptions that most plagues RfA is the idea that adminship requires superhuman intelligence and years of study in Wikipedia policies. But really, it's not rocket science and if you're open to the idea that tough decisions should be made carefully and in consultation with others, the chance of screwing up is pretty slim. Let's face it: admins that run into serious trouble are not people who misunderstand policy, they're people who don't feel bound by it. That's why I hate the boilerplate RfA questions: knowledge of policy is not the right thing to test or to look for. Pascal.Tesson (talk) 23:44, 14 November 2009 (UTC)

Well, probably predictably by this point, I disagree with that. Policy knowledge is essential. Because a lack of policy knowledge doesn't typically result in desysopping is a poor reason not to require it. Tan | 39 23:52, 14 November 2009 (UTC)

But he didn't say that you shouldn't know policy, only that it isn't the thing which should be focused on. And I have to agree with his statement, and (without pulling names out of a hat) could back that up with examples. --Izno (talk) 00:29, 15 November 2009 (UTC)

I'm with you 100% there Tan, but Pascal.Tesson makes two good points. One, that it's not that hard to get a good grounding in policy, and two, that sysop issues usually aren't a misunderstanding of policy but rather ignoring it/acting anyway. Knowing policy is one thing but applying it is another, and that's something that RfA doesn't always test very well. I'm not a huge fan of hypotheticals, especially because after once or twice they're useless, but they can provide a better litmus for what the two of you seem to be agreeing on. ~ Amory (u • t • c) 03:29, 15 November 2009 (UTC)

I'm going to generalize from my own perspective: <full disclosure> I've been biting my tongue long enough about this. I've in past times past !voted frequently but not so much recently. Three or four years ago if a user wanted to be an admin and wasn't on "probation" for doing stupid things and had some name recognition then he became an admin - no big deal (Boothy aside). Not so much anymore - it's become a big deal indeed. I'll not name names but IMHO there were many many admin promotions made back then that would not stand today. Those promotions, without today's gauntlet process, produced some of our best admins and with only a few screw-ups. They would not stand today (if the candidate even agreed to be a candidate) in today's RfA environment because they weren't "perfect" enough. I've pretty much quit this process just because of that environment - a gotcha process that many potentially fine admins are saying "screw that". If the vetting process becomes as it has become then many potentially stellar admins won't submit to the inevitable "spanking" from some aggrieved user and their followers. Fewer candidates, you bet. </full disclosure> hydnjo (talk) 03:02, 15 November 2009 (UTC)

Arbitrary section break

To put it bluntly, with no disrespect towards the OP (a general observation), I think these self-pitying threads are doing more harm than good. How to Lie with Statistics is quite relevant. –Juliancolton | ^Talk 04:21, 15 November 2009 (UTC)

Yup. Stifle (talk) 11:09, 15 November 2009 (UTC)

As a statistician, I don't see anything particularly misleading about the statistics portrayed above. That said, I agree wholeheartedly with both of your sentiments about these self-pity threads being unhelpful.

To the community: If we were able to draw some clear inferences from these statistics, and somehow determine cause-effect relationships, that would be one thing. As it stands, however, these threads still leave us without concrete actions to take. Fewer people seem to be running, sure: Well, then nominate some people! Stop spending time moaning at WT:RFA and fix the problem! I am extremely picky about who I take the time to nominate, and yet I still managed to nominate (or co-nominate) two admins—Natalie Erin and Hmwith (both of whom were promoted without opposition and have served as distinguished administrators). If people are not running because they view RfA is a hostile environment, then make it a little less hostile by providing some social validation ahead of time by nominating them. If each participant in this thread spent the equivalent amount of time recruiting new admin candidates, I bet we'd have a bunch more admin candidates right now. IronGargoyle (talk) 18:17, 15 November 2009 (UTC)

Question: Why stop at two? What stops you from inviting and nominating ALL the people you deem worthy and who are willing? Lambanog (talk) 19:59, 15 November 2009 (UTC)

I didn't say I was going to stop at two. Your attempt at pointing out my own hypocrisy is skillfully deflected. ;) I do have tough standards in terms of who I nominate though (and yet I managed to nominate two people). I bet that a lot of people at WT:RFA haven't nominated anybody (and yet may have less stringent standards). IronGargoyle (talk) 21:53, 15 November 2009 (UTC)

Ultimately, the problem with RfA can't be solved by anything less than a community vote on how the process should operate. I've been saying that for years, and I keep saying it because the problem never gets solved any other way. RfA is critical to the health of the community, and we need a definitive community decision on how it should function. Everyking (talk) 06:11, 16 November 2009 (UTC)

Wikipedia:RfA Review was supposed to generate that, but stalled. Stifle (talk) 09:23, 16 November 2009 (UTC)

You're right Everyking. We were gearing up to kick off such a vote at the village pump back in mid Sept. But then the alternative recruitment campaign idea emerged, which took the steam out off our plan. A classic case of the good being the enemy of the best; the recruitment idea was too good a compromise not to be allowed to play out. By early next year, assuming our active admin corps is still declining, it will hopefully be mostly accepted that its not wise to rely on recruitment drives alone and we can have another go at achieving community consensus for making the process easier and / or less hostile. In the meantime these excellent stats from Mat are useful to help track the problem. FeydHuxtable (talk) 19:46, 17 November 2009 (UTC)

I'm curious to know what "harm" these threads are doing, that they could be doing "more harm than good". Chris Cunningham (not at work) - talk 14:08, 16 November 2009 (UTC)

Indeed, RfA is critical to the health of Wikipedia. I'll ask a question: is the process actually broken, or has it failed to evolve in the right direction? Many people say "RfA was easier when I did it", or that "I passed then, but probably wouldn't now". Arguably, the "old" process would not be appropriate anymore - edit patterns have changed, the visibility of Wikipedia has changed, the "craftiness" of people has changed ... because of that, RfA has become tougher. However, is this "toughness" change been a quasi-relative response to the relative changes in threats to Wikipedia?

Also, what is the true validity of people cherry-picking their Oppose !votes. As an example "I oppose because of [blah this] comment to User:so-and-so", yet fail to either notice or point out that so-and-so and the RfA candidate discussed the issue, and have worked conderfully collaboratively since? Too late, the pile-ons have begun! RfA candidates seem to be dissuaded from responding to !votes ... in my own RfA, I ended up striking my responses because of it.

This isn't a sign of broken-ness per se, but there should be a way - maybe a chance for a candidate to respond to the opposes in a centralized spot, and !voters be notified to "review your vote" at the end of voting, but before bureacrats make a decision? I'm just throwing things out here from the top of my head ... feel free to disagree! (talk→ BWilkins ←track) 13:43, 17 November 2009 (UTC)

Those are good ideas and yes evolving in the wrong direction might be a helpful alternative way to look at it - then when we try and get consensus for change we can say were be updating RFA rather than the more emotive work reform. Broken is still a useful shorthand way to say it though. FeydHuxtable (talk) 19:46, 17 November 2009 (UTC)

The problem is the three steps that : 1/ people realise how much harm a bad admin can do 2/ they realise how difficult it is to remove one, and the lack of any lesser way to control their acts 3/ consequently, they are careful whom they appoint. Earlier, I don't think it was as clear how much harm could be done, and how easily admins might become susceptible. DGG ( talk ) 06:31, 21 November 2009 (UTC)

It is a tough problem, and I can't begin to claim I have a solution. If I'd run in the current climate, it would have been a good deal closer, to say the least. I think DGG certainly has a point in that it is difficult to remove admins that are bad, but not egregiously bad (general nasty or arrogant demeanor or repeated refusal to listen to advice, vs. deleting/vandalizing the main page or wheel warring). I don't really know the solution to that, though. The places that need administrative intervention the most are by definition contentious and often hostile, and making it too easy to take potshots at admins may discourage them from being willing to step in where it's badly needed. On the other hand, being hard to remove admins does allow that occasional bad apple to do a lot of damage. I do think ArbCom has been more willing to handle the issue of recent, and that's a positive sign, but I think the return of the "Not enough Portal Talk edits!" (or not enough featured articles, or not enough talk page work, or too much talk and too few article edits, or whatever the wherearetheeditscountitis issue of the day is) are a symptom of the problem. In the end, we should be asking a few simple questions when deciding whether to support an RFA: "Does this editor have sufficient knowledge of policy to avoid making frequent unintentional errors? Does this editor have sufficient self-control to avoid making intentional errors or attacks? Does this editor have sufficient thoughtfulness and reason to know when to ignore all the rules and get something done, and just as importantly, when that would be a really bad idea?" If I can answer yes to those questions, I support. Seraphimblade ^{Talk to me} 07:15, 21 November 2009 (UTC)

A possible explanation is that anyone interested had already went through RfA and either got the tools or got bitten the head off. OhanaUnited^{Talk page} 07:16, 24 November 2009 (UTC)

The few 2003 diffs I've examined were unstructured braindumps added without balance or references. Maybe "we have ten times as many articles and many times as many edits per day now than we did in 2003", but we also have better tools, better processes (with better guidelines that are better understood), better editors and in general a better understanding of how to write for Wikipedia. Editors who can't handle our emerging V+RS+NPOV/OR culture don't stay around for long. If Wikipedia is really at risk because the number of active admins is relatively static, perhaps we should cull some of the sacred cows that cause our admin workload. - Pointillist (talk) 00:52, 25 November 2009 (UTC)

I wouldn't call 1% decline per month relatively static, but what are the sacred cows that you speak of? Ϣere SpielChequers 01:12, 25 November 2009 (UTC)

I don't know the statistics, but if contributors are becoming reluctant to become/remain active-admins, and if events-that-require-admins aren't reducing at the same rate, an overload can be predicted. AFAIK there's no sign of this overload yet, but it could develop very quickly, because Wikipedia has an apparently unstable social model, based on corpus of inflexible rules ("sacred cows") that initially hold the doors wide open for new contributors and later use administrators to enforce fine-grained principles that only make sense to editors from a narrow educational background. The middle ground of responsible editors would dwindle rapidly if they had to defend the amount of time they spend here to their families/partners, anyway:

"What did you do at Wikipedia today, Daddy?"

"Well, I did NPP and I found that a sockpuppet of a PR company had created a c7@p article about one of their clients."

"Did you blast them out of orbit Dad?"

"No, I spent two hours looking for reliable sources and re-wrote the article so it was totally NPOV, full of perfectly crafted citations, a fine piece of work that won't get questioned by any other editor."

"Wow, you really showed them how to do their job! I bet they paid you a lot for doing that?"

"No, they don't pay me. Wikipedia doesn't pay me. I do it for the love of the thing. That's why I'm tired and cranky in the mornings."

"Hey Dad, why don't you become an administrator?"

"Er..."

Social institutions evolve or die. I expect that Wikipedia will survive despite its sacred cows (after all, there's an almost unlimited supply of university students to polish it). If it really is at risk in the way you originally said, we should re-examine its ways of working rather than just asking for a potentially unlimited increase in administrator numbers. I'd ask current administrators and editors how to reduce the workload, e.g. "If you had to stay on this project until the day you die, what would you fix, and why?" - Pointillist (talk) 03:10, 25 November 2009 (UTC)

A question

This may have already been brought up before, but the above is tl;dr. Raw numbers aside, how has the proportion of new community servants to new community members altogether changed, if it has to any significant degree at all? That, I think, is what is most relevant. Kurt Weber (Go Colts!) 21:00, 1 December 2009 (UTC)

Hit and run suggestion

There's been lots of debate about whether or not RfA is in some way broken, but I think that much of it has missed the point; the administrator role is poorly defined and confuses several aspects such as policing behavioural issues, judging behavioural issues, and dealing with content issues. Administrators have accreted to themselves more and more powers that many of them are clearly incompetent to use, but are reluctant to relinquish to those who are.

I realise though that pointing out the bleedin' obvious wins no friends around here and changes nothing. Nevertheless I feel better for having said it. :-) --Malleus Fatuorum 22:27, 26 November 2009 (UTC)

Well, you might want to make a note or three at Wikipedia talk:Community de-adminship/Draft RfC then, Malleus. -- Avi (talk) 22:29, 26 November 2009 (UTC)

What would be the point? --Malleus Fatuorum 22:46, 26 November 2009 (UTC)

The point would be to help draft the process that has the best chance of facilitating more streamlined forced relinquishment of tools when necessary. -- Avi (talk) 22:59, 26 November 2009 (UTC)

Wake me up the day that anything significant changes around here. There is absolutely no chance of the turkeys voting for Xmas. --Malleus Fatuorum 23:28, 26 November 2009 (UTC)

Isn't this more of a comment than a suggestion? –xeno^talk 22:49, 26 November 2009 (UTC)

I thought that my suggestion was implicit, but clearly not. The administrator's package needs to be split. --Malleus Fatuorum 22:57, 26 November 2009 (UTC)

Gotcha. –xeno^talk 23:00, 26 November 2009 (UTC)

What are these extra powers and when did they come in? (This is not a sarcastic remark - I can't think of any changes in the last two years other than the unbundling of Rollback, and while I've edited for more than two years, before then I was blissfully unaware of wp space let alone wt space). Ϣere SpielChequers 23:06, 26 November 2009 (UTC)

Well, there is the Edit filter... But that's already unbundled... (sorta kinda) –xeno^talk 23:21, 26 November 2009 (UTC)

There's also the promised flagged revisions, which gives administrators the ultimate call. How many administrators were voted in because of their understanding of BLP policy? --Malleus Fatuorum 23:32, 26 November 2009 (UTC)

Malleus is right. WP:UNBUNDLE is the way forward, but it won't happen any time soon. Pedro : Chat 23:11, 26 November 2009 (UTC)

Well a slight move in that direction would be to have more admin training modules, so that when you do start using the tools in an unfamiliar area there's a training module or maybe even a CTM. Ϣere SpielChequers 23:25, 26 November 2009 (UTC)

That's just about as comprehensive a misunderstanding of the issue that I've ever seen. --Malleus Fatuorum 23:33, 26 November 2009 (UTC)

If someone uses a tool incompetently then taking away the tool they misuse is one solution, training and retraining are other approaches. Personally I think all three approaches have their merits and ideally when incidents occur we should be deciding which approach is the best for that incident. That may not go down well with those whose only response to an incident is to demand that heads must roll, but in my experience being able to pick from all three approaches makes for better decisions than over-reliance on any one of them. Ϣere SpielChequers 00:02, 27 November 2009 (UTC)

As I said, wake me up when you wake up. --Malleus Fatuorum 00:24, 27 November 2009 (UTC)

"The administrator's package needs to be split." Fair enough, but just how? This discussion is pissing in the wind unless someone can come up with a way to partition the present administrator responsibilities into two distinct roles, neither of which also suffer from being "poorly defined and confusing several aspects". What are we going to do, tell some people they are alowed to revert vandalism but they can't warn or block vandals? tell others that they can warn or block people for edit-warring, but they can't protect the page being warred over? Hesperian 00:28, 27 November 2009 (UTC)

It would be very easy to do, if there was a will to do it. Blocking, for instance, is clearly a policing function that has very little to do with being able to edit abuse filters. --Malleus Fatuorum 00:33, 27 November 2009 (UTC)

Indeed. Now define for me two distinct roles, one that would need the block button but not the ability to edit abuse filters, and one that would need the ability to edit abuse filters but not the ability to block users. Can't do it? Hesperian 01:07, 27 November 2009 (UTC)

I don't think abuse filter edition has caused too much drama. Unbundling would probably work best when separating blocking, deleting and protection, the rest remaining in the common abilities pool.

With that, you get content wardens with the protection button (RfAs would examine mostly on content work and activity on the content-related noticeboards) , janitors with the delete button (RfAs examining content building, AfD participation, or specific cleanup activites/NPP/RCP), and community moderators with the block button (dispute resolution skills) . Doesn't have to be more complicated than that. MLauba (talk) 01:16, 27 November 2009 (UTC)

I don't want to waste any more of my time on yet another unproductive discussion on what must rank as the most unproductive talk page on wikipedia. Argue amongst yourselves, as the truth is plain to see. --Malleus Fatuorum 00:38, 27 November 2009 (UTC)

I wouldn't mind unbundling, as long as it was done sanely. The fact that 1,600 admins can hand out Edit Filter Manager and IPBE (two very powerful userrights) without any community review and crats can hand out Bot (a very powerless userright) only with BAG approval simply doesn't make sense to me. Some admin rights like editusercss and movefile probably could be handed out by any admin to many users, but others like block and deletedtext probably deserve a higher level of procedure and review. MBisanz ^talk 01:15, 27 November 2009 (UTC)

My final thought on this, as nothing will change, is that I fall at the first hurdle. Why should administrators be handing out "rights" that didn't exist when they were voted into power? They weren't even voted those rights themselves. --Malleus Fatuorum 01:59, 27 November 2009 (UTC)

Interesting question, which merits a separate discussion, I think. I also think the answer lies somewhere near Marbury v. Madison. Whether or not one agrees with the decision, Marbury established judicial review, and it has shaped law in these United States ever since. Whether the framers of the United States Constitution intended that or not is of academic interest (at least to some of us on this side of the pond, Malleus :-) ), but it doesn't change the way things work. Perhaps that proves your assertion (nothing will ever change), perhaps not...but I think it sort of answers the why question about administrators handing out bits that didn't exist when they became administrators. Frank | talk 02:26, 27 November 2009 (UTC)

Actually, Malleus, with the exception of AbuseFilter editor, all of the user rights that administrators can issue are part of the administrator package, and have been for a very, very, very long time. So yes, you actually did vote to give us those rights. J.delanoy ^gabs _adds 03:09, 27 November 2009 (UTC)

So your argument is that most of the admin rights were voted on, but it doesn't matter that some weren't, and that all additional "rights" are accreted to the body least able to honestly deal with their abuse? It's difficult to decide whether you're an honest idiot or just an idiot. --Malleus Fatuorum 03:20, 27 November 2009 (UTC)

Because you said, "Why should administrators be handing out "rights" that didn't exist when they were voted into power?". We aren't "handing out rights that didn't exist when we came into 'power'". Because of software updates, we were given the ability to hand out rights that were previously packaged solely into the "sysop" bit. J.delanoy ^gabs _adds 03:25, 27 November 2009 (UTC)

And if administrators are the "least competent" users to hand out those user rights, who do you suggest does it? Anyone approved by you? Simply saying that "Oh, those guys suck" and not offering any viable alternative is not good enough. Someone has to decide who gets them. J.delanoy ^gabs _adds 03:30, 27 November 2009 (UTC)

OK, you've convinced me. You're an idiot. --Malleus Fatuorum 03:34, 27 November 2009 (UTC)

Good lord, Malleus. Stop poking. Keeper | 76 03:35, 27 November 2009 (UTC)

OK, perhaps I did overstate my case a little. I offer my apology to J.delanoy for that. --Malleus Fatuorum 03:47, 27 November 2009 (UTC)

OK so Malleus has now explained that his concern isn't that administrators have been gaining tools that didn't exist when they were appointed, but that there has been some unbundling, and that this concerns him because it is the admins who have gained the ability to hand out certain user rights. I think he has identified a theoretical issue here, but I'm not convinced that we actually have a problem. I'm not very active at the Dwamah boards so may have missed the odd desysop or trouting for inappropriate handing out of Rollback or Autoreviewer. If such things are a common occurence then I'd agree that our current system needs to change. However in the absence of evidence to the contrary, I think one of the parts of RFA that works well is that we focus on whether we trust the judgement of the candidate before we hand out the mop. Admins are then trusted with the whole mop regardless of whether certain aspects were discussed in their RFA. Personally I like to see a candidate who promises to proceed with caution when using unfamiliar parts of the tools, but I'm not convinced such commitments influence many RFA !voters. Taking Autoreviewer as a case in point; I can't remember the last time a candidate was asked what their criteria were for handing out or revoking Autoreviewer rights, so I see little point being concerned if some admins got the bit before admins became able to hand out Autoreviewer to others. But reverting to my point about training, I could find no mention of Autoreviewer allocation at new admin school. I use this example because I have recently started to hand out Autoreviewer, but had no thought of getting involved there at the time of my RFA. Perhaps we should rename wp:New admin school as wp:Admin refresh training and give its contents an overhaul, after all one of the big changes in the last five years is that we've gone from a project where no admin had served for more than two years to one where most have served for more than two years and many for more than five. So perhaps we should start thinking about an ongoing training program for admins. Ϣere SpielChequers 14:17, 27 November 2009 (UTC)

As an aside, as it looks, at least on the surface, that the true beef is about who hands out / who receives page patroller rights once the system goes live, the last discussion Dank had initiated here about a month ago has petered out, and all pages tied to WP:FPPR have pretty much gone silent since mid-September (see Wikipedia talk:Reviewers). As far as I'm concerned, I'm all for granting Reviewer rights to the bulk of all WikiProject BLP participants that have even half of a clue for the trial itself, running the trial on an admin-only base would be a quite severe mistake in testing the system. MLauba (talk) 14:40, 27 November 2009 (UTC)

I think the idea of breaking the admin job down into particular roles is a great one; however, particular roles cannot be that clearly distinguished in terms of tools. I mean, yes, an enforcer of civility and editwar rules needs the blocking tool, but so does an antivandal worker. A content moderator needs the ability to protect, but so does a high-volume template editor. AFD closers need deletion, but so do the veteran CSD patrollers. Then, of course, arbcom enforcement and sanction-creation admins fall into a class all of their own - closer to proconsul with imperium inside the designated topic area than anything else. Our problem right now is that admins are all of these things. Antivandal editors are being evaluated based on how much "experience" they have with content disputes. Veteran article-writers and consensus builders are being evaluated on how well they "understand" (too often meaning: agree with question-asker's faction about) BLP or Image copyright policy. These things often occur without regard to the stated intentions of the candidate for how they'll use the admin tools. The history of RfA of late is littered with failed nominations on those grounds. I'm not sure what the lines between these separate roles might be, but I'm pretty sure we need separate roles, with understandings that, barring IAR-type circumstances, they will not step outside that role (e.g. theoretical purely antivandal admins blocking established editors). Ray^Talk 21:50, 27 November 2009 (UTC)
There is analogous problem with editing. Non-registered editors, registered editors, and autoconfirmed editors each have a toolkit that may be way more than what we need for our desired task. Some of us focus on article- and content-creation. Some focus on article organization and project management. Others focus on community-building. Still others do gnome work or vandalism-management. Do I really need the ability to edit semi-protected articles if my main focus is welcoming new users? Probably not, but the tool is there.

What is needed, and what we by and large hopefully have among active admins, is that each admin has a sense of where their competencies lie and they tread carefully when using tools outside those areas. We, as the community, have to trust that they have the good sense to know where they don't have the skill to do the job well, and that when they do use the tools in those areas they will either get a second opinion or do their homework and study the rules and precedents before taking actions that can have serious consequences. davidwr/(talk)/(contribs)/(e-mail) 22:38, 27 November 2009 (UTC)
We don't have to trust anyone, and neither should the system demand it. A total of zero administrators were elected to be the final arbitors on article content via flagged revisions. The exact same number that were elected because of their expertise with the abuse filters. Why should they be trusted? --Malleus Fatuorum 22:55, 27 November 2009 (UTC)

Heh. Yeah, but we don't have a problem with editors being rejected for confirmed status, the way we do for admins. And as things stand, admins are essentially unaccountable absent massive, arbcom levels of drama. I mean, over at the admin recall draft RFC, there's a proposal that requires a 70% margin to remove an admin that's in the lead. You need less of Congress to vote to remove the president. So what we have is extraordinarily vaguely defined role exercising a broad spectrum of power, with only minimalistic levels of accountability. And then we wonder why people hesitate to promote people to such a position. Ray^Talk 23:16, 27 November 2009 (UTC)
The difference, of course, being that 70% is a lot easier to achieve here than in Congress and that as a wiki, the evidence is clear and visible. Since most Presidents only get slightly above a majority when elected, you can argue that we're being consistent - harder to get in, but correspondingly harder to reverse that decision. ~ Amory (u • t • c) 02:06, 28 November 2009 (UTC)
And my point is that, from the point of view of maintaining editor confidence in the admin corps as a whole, and keeping the admin corps grounded in the Wikipedia community rather than separating them as a nomenklatura, neither is desireable. My attitude is that an admin who loses majority support in a weeklong publicized discussion for his retaining the bit has definitely lost the community's confidence. Ray^Talk 04:27, 28 November 2009 (UTC)
I agree. The difference between WP admins and the presidents of the USA is that POTUS is checked by many other mechanisms - Congress, the courts, journalist, etc. WP admins are in practice unaccountable, unless there is massive drama, usually with ArbCom as the final act. I've seen blocks with were ill-judged or authoritation, and a few are thoroughly incivil but get much more lenient treatment than non-admins. The currenty practice of admins regulating admins is as ineffective as any other self-regulation. ---08:30, 28 November 2009 (UTC)

I would like to request the creation of a user right that has only one ability: block Malleus. :) P.S. I would also like that user right. :) Ottava Rima (talk) 23:32, 27 November 2009 (UTC)

Actually there may already be a few folks filling that spot, but if something opens up we'll contact you. ;) — Ched : ? 23:48, 27 November 2009 (UTC)
Why block when you can delete? :) davidwr/(talk)/(contribs)/(e-mail) 23:55, 27 November 2009 (UTC)

Malleus is perfectly entitled to raise these subjects, we could indeed move to a US style system where everyone from the sheriff to the dogcatcher is elected, with separate approval systems to use the tools to allocate rollback, protect articles etc etc. Even if the tools were not technically unbundled we could move to a system whereby each admin had some parts of the toolset they were empowered to use and other parts that they were only authorised to use on designated training material in new admin school. Personally I think that would be a bureaucratic solution to a non problem, but as I said I'm open to persuasion. If people can show me examples where admins have blundered about with an unfamiliar tool and even after trouting and other attempts at retraining had persisted and had to be desysopped, then I'll support Malleus's solution. However my understanding is that all or almost all desysoppings have occurred for things that the admin should have known were wrong at the time of their RFA. As to whether Malleus should be deleted, I note that he has not yet listed himself here. Ϣere SpielChequers 14:09, 28 November 2009 (UTC)

I disagree with WereSpielChequers's "all or almost all desysoppings have occurred for things that the admin should have known were wrong at the time of their RFA":

RfA cannot predict how well the admin use the blocking power. Some have been appalling.
The are well known "how to become an admin" cookbooks, in other words, some admins have gamed the system - pretending to be more incivil / neutral / honest than their true colours appears after adminship. Gaming the system can't be prevent in advance, the only way to counter is by sanctions after their true colours are known - in other words suspension, examination and, if the admin's conduct is less than is required of non-admin editors, desysoppings. Note the phrase "the admin's conduct is less than is required of non-admin editors" - what of the current is that after RfA admins are reluctant to sanction other admins. --Philcha (talk) 14:44, 28 November 2009 (UTC)

Actually I'm not sure that you do disagree with me Philcha. Admins who feign extra civility to get through RFA probably know that being less civil after their RFA is wrong. As do Admins who block fellow editors for things which they would not have filed at AIV before their RFA. RFA has not proven an infallible way to decide whether an RFA candidate would or would not make a good admin. But blocking has been an admin power since the earliest days of RFA, RFA may not be a perfect way to appoint admins, but admins who have been appointed by it have been given the blocking power by the community. Ϣere SpielChequers 20:23, 28 November 2009 (UTC)

WereSpielChequers: I was joking about deleting Malleus. Actually I wasn't really joking, but I'll only nominate him if he's up for it. I hear the results can be deadly :) <-- note smiley. davidwr/(talk)/(contribs)/(e-mail) 15:06, 28 November 2009 (UTC)

Hi David, Absolutely, I understand you perfectly. I regard almost everything to do with EFD as a joke, but was just a bit surprised that Malleus hadn't opted out. Ϣere SpielChequers 20:23, 28 November 2009 (UTC)

Maybe he has, I didn't check that. davidwr/(talk)/(contribs)/(e-mail) 21:26, 28 November 2009 (UTC)

I did check and he hasn't. There are surprisingly few editors who have listed themselves at User:GlassCobra/Editor for deletion/Opt out, and none since January this year (which makes the four day gap at RFA practically a blip). Perhaps we should have a trawl through those who've had rollback or autoreviewer for 6 months and give them a choice of EFD or RFA? Ϣere SpielChequers 17:34, 30 November 2009 (UTC)

I haven't because I couldn't care less about it. I guess that's why so few others have either. --Malleus Fatuorum 22:09, 30 November 2009 (UTC)

From a different perspective.........

Today's Arbitration Committee no longer uses stone tablets to conduct business.

Someone above said that the issue was a poorly defined role. I wonder if I could take that a little bit further. I think a clue is in the alternative name used for admins - 'sysops'. A sysop is someone who can log in to the printer queue with admin rights and clear the buffer. A sysop is someone who can set up a new user, or delete an old account. You obviously still need plenty of that function - someone who can delete pages that someone else has requested be deleted, someone who can move user accounts to another name, that kind of stuff. But most of what your admins are now doing is a customer services job. They aren't blocking vandals, they are dealing with editors who are arguing - with each other - with anyone who tries to mediate. They are having to investigate who flung mud, and where and when, and decide what action to take, and probably get roundly criticised for whatever decision they make. As a long standing customer services officer who was promoted to work with the IT guys putting in new customer services systems, I can safely say that good sysops usually make very poor customer services officers. So Wikipedia's old admins (from 2003/4/5) are jacking it in as a bad job, and Wikipedia's recruitment process isn't picking the right people for the new job, because it hasn't defined the job.

How about, rather than looking at it as unbundling tools, looking at it as unbundling roles. Just because someone has the button, doesn't necessarily mean they need to have permission to use it in all cases. You might come up with

sysop: can delete things if there is a valid deletion request (CSD,PROD or AfD). Can move usernames, sort out disasters with redirects etc. Can semiprotect pages with valid RPPs for vandalism. In short, has tools to deal with content only.
patroller: can block drive by vandals and bad usernames, semiprotect pages that are being vandalised, warn or block for persistently recreating rubbish, spam, copyvios etc. Not inteded to undertake any action connected with editors who have been productive but have descended into fighting with each other.
moderator: can intervene in disputes, semi or fully protect pages to force discussion, warn (and lets have some beefed up warnings) or block users who edit war, refuse to discuss, troll, attack etc. Can close pages at AfD.
Umpire: clear first line reviewer for decisions taken by the first three. Block reviewer for blocks made by a moderator (patroller blocks could be peer reviewed probably )

There might be a few more categories. Unbundling by role means that one person no longer has to have in depth knowledge of everything, and admins are more encouraged to work together. --Elen of the Roads (talk) 01:16, 28 November 2009 (UTC)

Nice in theory but I'm skeptical that it will work out in practice. There's always going to be the grey areas, these will make the bulk of the ANI reports and we'll be back with 4 groups of mistrusted people instead of one. It only works as long as everyone is playing along nicely. Plus, to a certain point, we already have this in our status quo, with many admins specializing in only one or two areas of interest, just not formalized.

Unbundling the tools may be less elegant but it enforces both a separation of the roles by virtue of not having the tools, and the cooperation you rightly mention, as you'd have to go to a different group.

There could even be different promotion thresholds, for instance, page protection is probably less contentious than blocking and could work with 60% support to promote. MLauba (talk) 01:24, 28 November 2009 (UTC)

I take your point, but I do think the problem is around thinking of it as a function of the 'tools', not a function of the role. Elen of the Roads (talk) 01:44, 28 November 2009 (UTC)

There are only 2 ways this will work: 1) Administrators are encouraged to self-identify as "here is what I do with the tools" but without any software enforcement, or 2) there is some software mechanism to enforce it.

Until the software gets per-page granularity on using the tools, I don't see any progress on an enforcement-based mechanism. On the other hand, once per-page granularity is available, a whole world of possibilities opens up, including project-based administrators who can control the articles in their own projects but not the rest of the wiki, or "deleters" who can delete articles which have a deletion tag and which they themselves have not edited, etc. But we are a long way away from this capability. davidwr/(talk)/(contribs)/(e-mail) 01:48, 28 November 2009 (UTC) See here for related question. davidwr/(talk)/(contribs)/(e-mail) 02:00, 28 November 2009 (UTC)

I think if the roles were clear (and had to be advertised on the userpage for example) the community could 'enforce' it. I don't think you need automated reporting - in fact I think that would probably just make the suspicion factor ratchet up. Elen of the Roads (talk) 01:51, 28 November 2009 (UTC)

I think part of the problem there would be with consensus. What is consensus among a few individuals in a WikiProject is not necessarily consensus elsewhere. WikiProjects don't WP:OWN own the articles, especially those with a dozen attached, and how could we begin to define which projects they belong to? I think that'd create more drama than it might be worth. ~ Amory (u • t • c) 01:59, 28 November 2009 (UTC)

Mmm, this is too good an idea to ever get anywhere, sadly, since it's been said splitting tools wasn't really possible. Persons such as myself (and I'd assume Davidwr as well) might look at basically every RfA and see things look more depressing every time. Support or Oppose based on GA/FA numbers? Too many edits? Too high an automated percentage? Namespace distribution? No file uploads? Worst of all, article creation. There are persons that do not want to go through the process because they are convinced they're doomed over this. Period. No other reason. I don't blame them, and I'll turn into one of those people eventually. View being that it's punishment for actually enjoying some greasework and to take more pride having in having their user page vandalized than a row of FA stars put across the top. We might not know a great article worthy of praise but we'll tell you if it's worth adding to start with. This would be large % grouping that most people in a "checkbox" system like this hypothetical would take. ♪ daTheisen (talk) 20:34, 28 November 2009 (UTC)

(edit conflict)I often use the word sysop instead of admin because, aside from actually being the name, I often feel it more accurately describes the role. What you define, Elen, is probably better described here as our sysadmin. Truth be told, the role is really a combo of procedural stuff (protect, delete) and administrative stuff (close debates). And while I don't really see how this is different then the suggestion above or otherwise, I'd agree with MLauba that there are advantages to having everything together. Dividing your corps up means AfD closures without deletion, BLP attack pages without blocks, and an overall diminished variety of opinion. And, having been one before elsewhere, I can tell you the last thing this place needs a role named Moderator. ~ Amory (u • t • c) 01:53, 28 November 2009 (UTC)

I'd also tend to agree, and when I look at someone at RfA, I'm not evaluating whether they know a given policy—that's trivial to look up anyway. I'm looking at my trust of the person, and if I trust them, that includes trust that they'll leave complex stuff in areas they're new to to those with new experience, and start out with simple cases there if they want to start work in it. If I can't trust you to exercise good judgment, in that way as well as others, I wouldn't trust you to do anything right. If I can, I trust you to start with what you know and move into other areas at an appropriate pace, if at all. Seraphimblade ^{Talk to me} 02:03, 28 November 2009 (UTC)

From my own perspective. I've always found the "sysop" term amusing. For me, a sysop is someone who can set the switch to "off" and change RAID batteries and tape cartridges. Tim or Brion would qualify as sysops too, since they can remotely shut the whole thing down. That's a sysop.

Elen makes some useful distinctions, but none of those roles are mutually exclusive, so declaring for a role (or using software granularity) won't necessarily help things. The "moderator" role (which is what I think admins should be called anyway) is open to all, but can be usefully backed with blocking and protection tools. The "umpire" is just a trusted admin. The "patroller" role will have problems with the edge cases, if the editor has 1100 edits but the patroller knows they're going wrong, do they really have to kick it upstairs? Elen's "sysop" still needs to exercise judgement, so if they have that judgement, why are they restricted to one role?

Ultimately, RFA still comes down to "has clue? If so, yes". In my own case, since I would have had to declare "all of the above or maybe none", Elen's layout wouldn't have worked at all. (And 25% still think it didn't work :) Franamax (talk) 02:14, 28 November 2009 (UTC)

I think we would want to skip the moderator title. It inherently carries a weight within discussions (do what Jim Lehrer says or ELSE!) and that's something we want to avoid. ~ Amory (u • t • c) 03:31, 28 November 2009 (UTC)

Re name sysop: For some older than me, the name sysop means "has bank of Commodore 64s in his mothers basement with as many phone lines coming in as he can afford, and has the power to set your login level to TWIT." :) davidwr/(talk)/(contribs)/(e-mail) 02:20, 28 November 2009 (UTC)

64 KB of memory? God I would have loved that much space! In my day, I had to make daily backups by charcoal sketch of the memory core and we kept them all in a cardboard box in the middle of the road. Still, those were good times... :) Franamax (talk) 02:50, 28 November 2009 (UTC)

Way to bump the analogy up to the max, thereby disallowing any of the hilarious intermediates! ~ Amory (u • t • c) 03:31, 28 November 2009 (UTC)

Oh c'mon. I remember when I used to send my !opposes by Telex. I remember running alongside the admins shouting "get a horse". I remember when block notices were carved on actual blocks of wood. Lots of life left in that baby. :) Franamax (talk) 03:48, 28 November 2009 (UTC)

An early early vote to impose a ban on several users on a predecessor to the Greek Wikipedia. davidwr/(talk)/(contribs)/(e-mail) 04:52, 28 November 2009 (UTC)

I think that this would be a fairly good solution to a problem we don't currently have. Fairly good because a system based on different types of admins only using the parts of the mop they are empowered to use would be vulnerable to wikilawyers saying that a particular incident doesn't quite fall under the aegis of the type of admin who tried to resolve it. A solution to a problem we don't really have, because I'm not convinced RFAs often if ever fail because we trust the candidate in the area that they want to work in but not in an area they haven't touched. I think it is rather more frequent that candidates fail because of the things they've done that do relate to the type of admin they want to be. If things were to change and for example the "no featured" oppose became common enough at RFA to defeat an otherwise good candidate, then I think we should revive the idea of different types of admins who were appointed on the understanding that they would only use certain parts of the mop. Ϣere SpielChequers 17:50, 30 November 2009 (UTC)

There have been more than a few who got significant opposes because of inexperience or diffs showing a few poor decisions in a particular area other than where they said they wanted to work. Not necessarily enough to torpedo the nomination by itself, but frequently enough to make that particular RfA more of an ordeal than it needed to be. davidwr/(talk)/(contribs)/(e-mail) 22:38, 30 November 2009 (UTC)

I like this idea of "specialized" admins. 122.162.176.142 (talk) 14:30, 1 December 2009 (UTC)