Wikipedia:Wikipedia Signpost/2010-08-09/Admin stats
RfA drought worsens in 2010—wikigeneration gulf emerging
- WereSpielChequers is an editor on the English Wikipedia, the Strategy Process, and occasionally various other projects. He has been active at RfA for more than two years, and has researched its history back into the first year of Wikipedia.
The drought in RfA (Requests for Adminship) has now run for 28 months and is worsening. The number of admins promoted in 2009 was very low compared to previous years, and each of the past six months since has seen even fewer successful RfAs than the same month in 2009. The number of active admins has fallen from the peak of 1,021 in February 2009 to 802 on August 1, 2010.
Requirements have been increasing at RfA in various ways, and since the unbundling of rollback in early 2008 (see Signpost coverage) it has been difficult for candidates to succeed simply as "good vandal fighters". One result of this is that there is now a gulf growing between admins and non admins in terms of "Wiki generations". Over 90% of our admins first edited more than three and a half years ago. It is probably no surprise that there are no admins who first edited in 2010, but only nine started editing in 2009 and thirty-eight in 2008. Even the Wikipedians who joined us in 2007 are still under-represented in the admin cadre.
Month\Year | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | Totals |
---|---|---|---|---|---|---|---|---|---|---|---|
January | 2 | 13 | 14 | 44 | 23 | 36 | 6 | 6 | |||
February | 2 | 14 | 9 | 28 | 35 | 27 | 9 | 7 | |||
March | 8 | 31 | 16 | 34 | 31 | 22 | 13 | 2 | |||
April | 6 | 20 | 25 | 36 | 30 | 12 | 14 | 8 | |||
May | 10 | 23 | 17 | 30 | 54 | 16 | 12 | 8 | |||
June | [1] | 24 | 13 | 28 | 28 | 35 | 18 | 12 | 6 | ||
July | 3 | 11 | 17 | 31 | 26 | 31 | 16 | 10 | 7 | ||
August | 4 | 9 | 12 | 39 | 26 | 18 | 12 | 11 | |||
September | 0 | 17 | 29 | 32 | 22 | 34 | 6 | 8 | |||
October | 0 | 10 | 16 | 67 | 27 | 27 | 16 | 7 | |||
November | 3 | 9 | 27 | 41 | 33 | 56 | 11 | 13 | |||
December | 1 | 15 | 24 | 68 | 19 | 34 | 9 | 6 | |||
Total promoted | n/a |
44 |
123 |
239 |
387 |
353 |
408 |
201 |
121 |
44 |
1920
|
Total unsuccessful | n/a |
n/a |
n/a |
63 |
213 |
543 |
512 |
392 |
234 |
99 |
2056 [2]
|
Total RfAs including by email | n/a |
44 |
123 |
302 |
600 |
896 |
920 |
593 |
355 |
143 |
3976 [3]
|
Percentage promoted | n/a | n/a | n/a | 79.1% | 64.5% | 39.4% | 44.3% | 33.9% | 34.1% | 30.8% | 44.1%[4] |
Year | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | Totals |
---|---|---|---|---|---|---|---|---|---|---|---|
Year admins started editing[5] | 32 |
109 |
223 |
404 |
481 |
328 |
107 |
38 |
9 |
0 |
1731
|
Year active admins started editing[6] | 14 |
34 |
71 |
163 |
221 |
183 |
69 |
29 |
9 |
0 |
793
|
Admins still active % | 43.8% |
31.2% |
31.8% |
40.3% |
45.9% |
55.8% |
64.5% |
76.3% |
100% |
45.8%
|
Year | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | Today |
---|---|---|---|---|---|---|---|---|---|---|---|
Active admins at end of year [7] | n/a |
n/a |
143 |
360 |
722 |
979 |
1,005 |
943 |
870 |
802 | |
Inactive or desysopped (net) | n/a |
n/a |
24[8] |
22 |
25 |
96 |
382 |
263 |
194 |
||
Inactive or desysopped (net)% | n/a |
n/a |
14%<ref>Over three years not one |
6% |
3% |
9% |
28% |
22% |
18% |
Notes
About one new admin per day was promoted from mid-2005 to mid-March 2008; this pattern oscillated between half and a fifth of that rate until late 2009, and since then the promotion rate has dropped to one every five days.
Originally sourced from User:NoSeptember/Admin stats#Year to year comparison of promotions by months, copied here and colour-coded. Updates from Wikipedia:Unsuccessful adminship candidacies (Chronological) and active admins from User:NoSeptember/Admin stats and Revision history of Wikipedia:List of administrators
- Early 2003 from User:NoSeptember/RfA_chronological
On the counting of admins
Nobody really knows how many admin accounts there have been on the English Wikipedia, or how many actual people they represent. Some stats include admin bots, others do not. If an admin has their account renamed they are still one admin, but if they leave and return with a new account and have that sysoped then some stats count them as two admins. Nowadays developers and other staff members have a "staff" flag, but some early staff members were simply made admins, and at least one measure of our number of admins would include them if they made an admin action. If a protected page is moved, the automated moving of the protection settings is counted as a page protection and is listed under the username of the person who moved the page, whether or not they are actually an admin. Also, stewards can act as admins on enwiki. Records for the early days are incomplete. We have had 1,920 admins appointed (not including bots), we currently have 1,741 accounts with an admin flag (including bots), of whom 802 are active as editors, and as of August 2, 2010, 2,019 editors are recorded as carrying out admin actions on the English Wikipedia, but see the note above.
- ^ 33 had been appointed in 2001 and early 2002
- ^ 2004–2010
- ^ unsuccessful 2003 are n/a; the earliest phase was done by email and there is no record of the people who were turned down
- ^ This percentage represents all years from 2004 onwards.
- ^ last calculated Aug 2010 – this only includes current accounts of people who are still admins, but includes inactives.
- ^ User CBM
- ^ active admins peaked at 1021 on 28/2/2009
- ^ Over three years not one
Discuss this story
Comments
I really disagree that there is a problem. Yes, numbers may be slightly dropping. So what? Is there any evidence that CSD's are taking longer? Users take longer to be blocked? Pages go longer unprotected? Until I see substantial evidence that any of these are happening, I don't think that there is a drought. We're in the age of editing tools now. There's less admins because there are more rollbackers. There's less admins because a few can do a lot more work than they did 3 years ago. I'm also convinced that always taking a negative view on this is unhealthy for Wikipedia and the RfA environment. (X! · talk) · @092 · 01:12, 10 August 2010 (UTC)[reply]
If there is no information for 2001, why does the column gum up the table at the start? "n/a" can mean "not applicable" or "not available", and beyond that still requires explanation in the accompanying prose.
In the last table (not numbered Table 3, unfortunately), "at end of year" is stated only for the top row. Does this apply to all rows? The default is that it does not. Tony (talk) 01:18, 10 August 2010 (UTC)[reply]
This is a perennial discussion. The chart shows just how bad things have gotten though. I can't imagine standing for admin now, and I would not wish it on any friend. So the admins that do get through may well be ciphers... huge edit counts, never took a controversial stand on anything, and you have no idea what you are getting. So... perennial discussion. Serious problem. Getting worse. Will to fix it? I doubt it. ++Lar: t/c 02:46, 10 August 2010 (UTC)[reply]
Great report. I personally know of at least ten great admins who would never pass RFA now, in fact, Jimbo might not even pass RFA with the criteria that some people have. Ronk01 talk, 03:02, 10 August 2010 (UTC)[reply]
Yes, so many with so many different, often arbitrary criteria. Consider that years ago all we looked for someone who was around for a couple of months and was civil and looked trustworthy (for the most part would be willing to reverse thier own mistakes and step back and take a look at questions/criticism seriously). Having a FA was a huge plus, and there were still people had stricter criteria, but for the most part that was how it worked. 9+ months in my opinion is.... let me put it this way - I think the meta adminship process and renewal standards are light years better then en. I think part of the fears stem from the fact that en bcrats can't desysop, thus it becomes near-impossible to de-admin someone unless they do it voluntarily or get drug through the long painful arbcom process. I'm not going to mention names here, but a couple of the "infamous" admins back in the day were actually nice people who tutored me a bit when I was an admin myself - I would consider them wikifriends at the time; unfortunately, even with what looks like a perfect record, people can change over time. These admins in question went nice to wheel warring and were in no uncertain terms terse and rarely even mean to me even when I stopped by to say to hello. I think people think they can prevent this by having "strict" criteria... but as much one wishes you can't. Ryan Norton 10:34, 10 August 2010 (UTC)[reply]
I became an admin back when that was pretty easy. I was only asked one question which irritated me, and it irritated me not because I thought it in any way unreasonable but merely because the only way I could think of answering it took me half an hour or longer. I now notice that in the answer I blithely use "notability" in its everyday and not its Wikipedia sense; nobody commented. On that occasion I was not expected to be particularly knowledgable about WP. I wasn't very knowledgable about it and still am not. (I think WP is a Good Thing, but am not fanatically devoted to it.) On occasion I'm justifiably accused of ignorance of this or that, but I don't think that I've been accused of ignorance or incompetence in my, um, administratudinalizing. That's because I think I know where I'm not qualified to venture. I'm grateful to those admins who are omniscient and those who are brimming with energy to do great quantities of donkeywork (deleting expired PRODs, etc), but that's not me. I think that if there were a couple of hundred new admins who were no better informed and no more energetic than I am (but also no more reckless or stupid), this would be a mild plus for WP. I'm a rare visitor to RfA but when I do go there I'm delighted by the occasional response from a candidate that's less than ultimately dutiful. Whereas such responses cause great indignation among the denizens of RfA (indeed, the inquisitors' and voters' perceptions of bruises to dignity provide some unintended comic relief), I'd happily vote to janitorize somebody who I thought was competent, decent and level-headed, and who responded to a question: I didn't know the answer to that one five minutes ago. As I don't propose to employ any administrative kryptonite in areas that I don't fully understand, my ignorance doesn't worry me. I now see that in the last RfA but one you asked the same question and got an answer that satisfied you. I suppose I could paraphrase that answer to a point where the result wouldn't look like plagiarism, but I really can't be bothered. -- Hoary (talk) 10:41, 10 August 2010 (UTC)[reply]
Lifetime appointment=big deal. Set a reasonable limit on the period of time and you'd see way more admins accepted. It's really that simple. Gigs (talk) 14:12, 10 August 2010 (UTC)[reply]
See also Dunbar's number. WP is at the point where such appears to be noticeable. Collect (talk) 14:14, 10 August 2010 (UTC)[reply]
Adminship should be automatic to anyone with 4,000-5,000 edits and who states that they have read the admin guidelines and procedures (if there are any) and will abide by them. Cla68 (talk) 15:40, 10 August 2010 (UTC)[reply]
I disagree that this is a problem. I think the editorship is speaking to the need - there are too many admins already. The job is no longer regarded as "just a mop and bucket" by a large number of admins - it's regarded as a leadership position, or a hall monitor position. That's detrimental to the collaborative nature of the project, and the emphasis on enforcement of arcane rules at the expense of content creation and the constant use of the argument for deletionism that "it can always be recreated" - in other words admins can see it even if one of them has hidden it from other editors and from users wanting to look it up, so that makes it ok - are a serious discouragement to participation in what this project is supposed to be about. The community should not be encouraging people who are focused on critiquing others' work and punishing "vandals." Especially now that we have vandal-patrol bots, automatic notation of suspicious edits visible to all of us at "recent changes," and rollback and other tools available to those who want them, the basic premise that we should have a high number of admins is flawed. I'm heartened that people have high standards; I'm heartened that fewer people want the mop. Those are healthy signs that the project continues on the right course. I encounter admins who seem to have their hearts in the right place; and I can think of at least one from the early days who I regret is no longer involved. Wiki-burnout is bad, and lynch mobbing does happen. But building up adminship as per se good for the project leads to lynch mobbing against content, and that's what I've seen over and over since I got involved. I'd be happy to see the admin numbers reduced still further, and thrilled to have their power undermined by the enabling of a real list of deleted articles such that I could determine a redlink represented something deleted without collecting the now quite serious quantity of material to write the article ex nihilo and clicking to do so . . . and then I see that some admin deleted someone's earlier work. Yngvadottir (talk) 16:07, 10 August 2010 (UTC)[reply]
Part of the problem is vague criteria for both nomination and participation. When anyone with a grudge can weigh in on an RFA, that creates an issue. On the other hand, meatpuppetry would also be bad. I think that there needs to be clearer criteria for someone to apply in the first place -- more clarity of minimum time, minimum edits, how much in the admin areas as opposed to content (and here I thought contributing content was a GOOD thing!), so that the "s/he hasn't chased enough vandals" issues would not arise. I think that initial questions should be standardized to avoid stupid ones, and after that, perhaps only certain people (perhaps only existing admins) could ask individualized questions, or the questions could be submitted anonymously via one existing reviewer who vets them for appropriateness. Maybe some kind of limit on how often each person could comment, so that one individual doesn't highjack the nom one way or the other, etc... Just some ideas. Montanabw(talk) 23:07, 10 August 2010 (UTC)[reply]
This is a great piece of analysis. It raises some very interesting trends about our admin population. I'm not an admin myself, but the anecdotal info I've heard supports much of what's been said in this thread about the road to adminship being a tough one to go down. I'd be interested in taking this analysis and looking further up the "funnel", namely looking at how editor trends affect admin trends later in time. Also, would it be easy to identify "Wikigenerations" for users that were approved during a certain year? For example, it would be interesting to see for the admins that were approved in 2010, how many started editing in 2009, 2008, etc. Howief (talk) 01:29, 11 August 2010 (UTC)[reply]
I'm sticking my neck out, and since everyone else is discussing their (lack of) admin credentials I'll add that I simply have little knowledge of the greater detail of the admin role and don't involve myself in such affairs so I could never pass a RfA (I do at least have a bank of many years Wiki experience), but I've noticed that a feature of several comments here is that "I could do X job [usually vandal-fighting, it appears] but I would be failed based on lack of activity in other sectors". Perhaps the solution is to create some, for want of a better term, "demi-admin" roles which allow users to focus solely on a chosen aspect of adminship [vandal fighting in this case] and simply not have access to the other features, thus rejecting a candidate for demi-adminship on the premise of not being involved in other aspects becomes a non-issue. Of course, it would be a monumental task to produce (just look at how long reviewer status took to develop, for instance) but then if we can take this article at its word, the issue is rapidly becoming a monumental hurdle for Wikipedia anyway - we could be looking at losing half our active admins in the next 24 months if current trends remain. Of course, I'm aware that my idea is surely riddled with problems - not least the sheer work to make it manageable - so I would never actually suggest this properly, but solely as an exercise in thought for those debating here, would it work? Falastur2 Talk 03:21, 11 August 2010 (UTC)[reply]
Too many admins already?
I'd like to reiterate what Yngvadottir said above: I disagree that this is a problem. I think the editorship is speaking to the need - there are too many admins already. The job is no longer regarded as "just a mop and bucket" by a large number of admins - it's regarded as a leadership position, or a hall monitor position. That's detrimental to the collaborative nature of the project, and the emphasis on enforcement of arcane rules at the expense of content creation and the constant use of the argument for deletionism that "it can always be recreated" - in other words admins can see it even if one of them has hidden it from other editors and from users wanting to look it up, so that makes it ok - are a serious discouragement to participation in what this project is supposed to be about. That absolutely nails it for me. Adminship is the Cool Kids Club and to hell with that. I've got all the tools I need with Rollback and Autoconfirmed or whatever the hell it's called and I'm a content-creator and proud of it. If we doubled our number of content-creators and halved the number of Administrative sorts, that'd be getting us closer to the mark... Carrite (talk) 22:21, 11 August 2010 (UTC)[reply]
Are there minimum admin-related edit requirements still on the Commons or Wikipedia?
That was/is a really dumb requirement. The best, wisest admins (arbitrator material) are usually those with broader experience online and off, and don't always live 24/7 on Wikipedia as some admins do.
Some of the worst admins I have seen are those pushing for minimum numbers of admin-related edits. Brilliant idea. Snark mode off.
I think the biggest need is for more admins at the level of arbitrators and checkusers. Wise people are needed to settle content disputes, and not just rule implementation. Developers are needed. We need really experienced people and admins, and not just vandal fighters.
We need integrated watchlists, too, so that more editors and admins come on over more often to the Commons where there is much work to do. Admins are really needed on the Commons. --Timeshifter (talk) 04:22, 12 August 2010 (UTC)[reply]
Are there a minimum number of quarterly admin edits required to maintain admin status?
I think this is a bigger problem. Why get rid of our wisest admins when they aren't doing admin work 24/7 on Wikipedia!? Life goes on outside Wikipedia.
We need wise arbitrators. Arbitrators oftentimes come from admin ranks.
Wisdom and experience comes from things done outside Wikipedia. --Timeshifter (talk) 06:57, 13 August 2010 (UTC)[reply]
Dealing with "bad" admins
I've read some people arguing that the bar for new admins is set so high because, if a candidate misbehaves after becoming admin, it's very hard for the community to deal with the problem. Do many people take this stance? If this is a problem (is poor enforcement real or just perceived?), wouldn't it be better solved by improving the process for dealing with errant admins? (for a trivial example: Somebody who might ban another editor with whom they've disagreed over content)
bobrayner (talk) 14:38, 14 August 2010 (UTC)[reply]
Brief article/essay on this subject, and reasoning given at RFA
Based partially on my earlier views on the lacking, often harsh, and occasionally even ridiculous reasoning given for opposition to Requests for Adminship, I've written a brief article on the matter at User:Esteffect/RFA_Reason_Test, which looks to create something of a test of reason for opposition. I'm welcome to comment and so on on it, and while it's only a first draft, perhaps feel it could be part of a Wikipedia space essay or even guideline piece on the matter one day. Thanks. Esteffect (talk) 21:57, 14 August 2010 (UTC)[reply]
2001 had no admins
I find it curious to read that 32 users were apparently appointed as admins in 2001. From my recollection no admins were appointed in that year as the functionality simply did not exist within the UseModWiki software. Jimbo, Larry Sanger and Tim Shell had direct server access which made them the "admins" by default. The first group of MediaWiki admins were appointed by Jimbo somewhere around Feb/Mar of 2002. 59.101.40.40 (talk) 06:39, 18 August 2010 (UTC)[reply]
Cited in an academic paper