Wikipedia:Wikipedia Signpost/2019-07-31/Special report

Special report

Administrator cadre continues to contract

The admin count trend is down. For more graphics, see User:Widefox/editors.

Coming as no surprise to those who have been following the remarkably linear decrease in the total number of administrators, a new record was recently set: fewer than 500 active admins.^[a]

Trends

About 45% of all admins are active.

The large graphic above plots the total number of administrators (+) each month since 2011. That year was chosen as the beginning of the "inactive admin suspension rule", under which the privilege is removed for user accounts that don't make any admin actions for a period of time. The points are fitted to a linear trend line. The trend is a consistently decreasing admin count since 2011. Widefox, who has been maintaining trend graphs at User:Widefox/editors since 2017, shared this analysis: "A linear decline trend in the number of admins is a good fit with r² = 0.994 . The WP:FRAMBAN loss of admins is a significant drop in the last two years, but similar to fluctuations before 2017."

A second chart to the right shows the active editor fraction: between 40 and 50 percent of all administrator accounts are active at a given time. If you look at the official tally (raw data compiled by bots), the active editor count peaked at just over 1000 in 2008, declined over the next few years to around 550, and has oscillated steadily between 510 and 570 since early 2017. Every tally during 2019 has been in the 43–46% range.

Trends explained

Impact

Prior discussions have not resulted in consensus on how problematic a loss of administrators is; some take the view that bots and other efficiencies can take up the slack. Others like Iridescent express concern: "[T]he issue isn't so much the decline in admin activity per se, but the fact that if admin numbers continue to drop while editing continues, the admin/page ratio increases exponentially and we eventually reach the point where problematic pages need to be locked because there are insufficient admins to perform the routine maintenance." Whichever view is true, English Wikipedia now has 5.896 million articles and 48.247 million pages^[c] versus 496 active administrators – or nearly 100,000 pages per active administrator – and the ratio is growing for the foreseeable future. Whether that is a problem, or how a problem would manifest, are questions still to be answered.

For previous Signpost coverage, see my "Wandering in the RfA desert of 2018", published almost a year and a half ago, and other related articles in the sidebar.

Also see

Wikipedia:RFA by month
Wikipedia:RFA2011, aka the Great RfA Debate

Footnotes

^ "active" is defined at WP:ADMINLIST as 30 or more edits in the last 2 months.
^ boldface added by The Signpost
^ page count and article count via WP:STATS

← Previous "Special report"

Next "Special report" →

In this issue

31 July 2019 (all comments)

Special report

Discuss this story

These comments are automatically transcluded from this article's talk page. To follow comments, add the page to your watchlist. If your comment has not appeared here, you can try purging the cache.

Deadweight admins and what their inactivity means

That almost half of admins are inactive re-ifies the fact that many hold the ban hammer as a status symbol. I think this calls into question the foolish WP:NETPOSITIVE rationales at RfA when we need to really see a need for tools. As for lacking mop-swingers, I've long called for locking-down most articles. Why not put automatically put PC1 on GAs and fully lock FAs? GAs and FAs don't need the help any n00b could provide. Vandalism, especially sneaky vandalism, so easily defaces the edifice our productive contributors made that we need to do almost anything to prevent it. Chris Troutman (talk) 17:15, 31 July 2019 (UTC)[reply]

How did that "locking down most articles" work out the last time someone tried it? --Guy Macon (talk) 18:52, 31 July 2019 (UTC)[reply]

The German Wikipedia uses Flagged revisions, which has a 56 day backlog. —Kusma (t·c) 19:39, 31 July 2019 (UTC)[reply]

I don't see how the presence of inactive admins invalidates WP:NETPOSITIVE - or at least I don't see how a harmless thing (people boosting their egos by gaining adminship does not harm anyone) is supposed to matter at RfA. Galobtter (pingó mió) 05:31, 1 August 2019 (UTC)[reply]

Re "people boosting their egos by gaining adminship does not harm anyone", it is a basic principle of computer security that you don't give anyone permissions that they don't need or don't use. --Guy Macon (talk) 05:49, 1 August 2019 (UTC)[reply]

That is why we have been desysopping for inactivity since 2011. This is a nonprofit volunteer project and it would be a terrible mistake to be as stringent with permissions as a banking institution or military unit should be. Cullen³²⁸ Let's discuss it 06:02, 1 August 2019 (UTC)[reply]

Since we don't pay admins and rely on volunteers we shouldn't worry if some of our admins, like many other editors, go through active and inactive phases. I'm pretty sure that admins gaining the tools and never using them is a misconception based on our adminstats not including admin actions earlier than December 2004, the vast majority of admins appointed in the last decade have at least had a phase of being active as admins. What I'm not sure of is how much movement there is between the semiactive and active camps of admins, obviously there is a fair bit of movement between inactive and not inactive, otherwise the inactives at any one time would be a good predictor of the number of desysops for inactivity over the following 24 months. Ϣere SpielChequers 13:23, 2 August 2019 (UTC)[reply]

The solution here is for experienced users and administrators to encourage those who they think will be good admins to run, offer to nominate users, or even to help them find nominators. We have the user base needed to create new admins. They just need to ask for it. TonyBallioni (talk) 21:03, 31 July 2019 (UTC)[reply]

It would also need experienced users and administrators to model the behavior that would encourage ordinary editors to want to follow in their footsteps. MPS1992 (talk) 22:03, 31 July 2019 (UTC)[reply]

I would agree. Thankfully, I think the admin corps is generally comprised of individuals who do their best to live up to the ideals of this project, so that shouldn't be too difficult to find. TonyBallioni (talk) 22:34, 31 July 2019 (UTC)[reply]

Generally, yes, and I have seen some examples of that in recent days. Generally is not enough, though, and I would like to see better. A few bad apples is not what the cart needs. MPS1992 (talk) 00:09, 1 August 2019 (UTC)[reply]

It's time to give some of our long-time editors more tools and relieve our admins of tasks like AfDs, moves and PP. Qualifications could include GAs & FAs, time spent in the relative area, etc. Similar has been mentioned before and it makes sense to me. We have qualified editors who can help reduce the janitorial load and leave the big jobs for admins....OR maybe one of our tech-gurus can create an algorithm that will clone TonyBallioni. 😊 Oh, speaking of long-time editors, do we have a graph of active editors similar to the admin graph? ^Atsme _Talk 📧 03:07, 1 August 2019 (UTC)[reply]

MPS1992, people notice and remember the bad apples rather than remember the good ones. Every week, I run into administrators I have never encountered before who are plugging away on their corner of the project, getting work done and not frequenting the high traffic noticeboards. You don't hear about them because they have found their niche and they are devoted to it. But at this point, I think they are the majority of active admins. Liz ^{Read! Talk!} 05:02, 1 August 2019 (UTC)[reply]

As proof of my point, look over this list of active admins: Wikipedia:List of administrators/Active...I bet there are dozens whose names are unfamiliar to you. They are not admins being called out for bad behavior or attracting attention, they are just getting to the work. Liz ^{Read! Talk!} 05:17, 1 August 2019 (UTC)[reply]

I agree, Liz. When straying around the less contentious areas of the encylopedia, I often see an administrative action taken by a name I do not recognize. Looking closer, I see a productive but low profile administrator who has been plugging away doing useful maintenance work for many years. Cullen³²⁸ Let's discuss it 05:57, 1 August 2019 (UTC)[reply]

Here's my perception which may be significantly different from what others think, but that is OK. Back in the 2005-2007 timeframe, editing Wikipedia was a fad. There was lots of low hanging fruit and somebody could easily write an article about their favorite album or athlete, and tell their friends that they were editing Wikipedia. "Look at my Wikipedia's article!" Similarly with becoming an administrator. If you contributed for a few months, wrote an article or two, and weren't a complete jerk, you could easily become an administrator, and brag about it to your buddies. Many of these people made some useful contributions, but were not fully committed to the project for the long haul. When the "faddish" aspect of Wikipedia editing faded from 2007 to 2009, and it was no longer easy to write a new article, the project consolidated around a pretty solid and stable core of highly active long term editors and active administrators. The chart showing a "dramatic" decline in administrators since 2011 is less dramatic than implied by a graph that goes from 1550 to 1000 as opposed to 1550 to zero. Lies, damned lies and statistics. The chart shows the ongoing departure of those not fully committed to the project. But even that is a slight diversion. What really matters is the number of editors and administrators who are highly active and committed to the project for the long run. Those numbers have been pretty stable since about 2011, as the faddishness factor died away. Cullen³²⁸ Let's discuss it 05:12, 1 August 2019 (UTC)[reply]

That would imply that we are seeing an "admin bubble" from the early years in run-off? Anytime I look at the RfA of a long-standing admin, I am surprised at how different it was to now, and certainly there are many examples of successful RfAs back then that would not pass now? However, I don't think there is a view that the RfA process back then harmed WP?

However, I also note that many of the main "boards" in Wikipedia that needs admin management (RPP, AIV, etc.), are only kept going through 2-3 admins who do most of the work, which seems unsustainable – was it always that way? Britishfinance (talk) 12:29, 3 August 2019 (UTC)[reply]

Of those "100,000 pages per active administrator", how many pages actually see any activity, especially any that would require admin intervention? Figures like these might sound impressive on paper but don't mean much without contextualizing and analyzing them. A better metric would have been to look at the activity on the admin noticeboards. Opencooper (talk) 06:09, 1 August 2019 (UTC)[reply]
The real killer is less articles but some of the more specialised backlogs (ACC etc) that require admins. They have backlogs stretching into months, and often depend on 2-3 energetic and keen admins. This means they have an appalling bus factor - if a stream loses even 1 of them, even for a couple of weeks, backlogs spike even further. Nosebagbear (talk) 09:42, 1 August 2019 (UTC)[reply]
There are 1,146 admins and 1,063 former admins. Soon eclipsed. The majority of admins edit some corner of the project, but perhaps don't even care to comment on the most notable happenings in the community, like Framgate. It would be interesting to see an analysis of how many admins are active at AN, ANI and AE. That's where there is real power to topic ban or block users, affecting things in the long-term. For a community with 120,000 active accounts, the core number of active admins at these must be staggeringly low. There is a real risk of these central avenues being run by a small number of gatekeepers who consider each other acquaintances.--Pudeo (talk) 12:57, 1 August 2019 (UTC)[reply]
- AN and ANI are reasonably bulky for most decisions, because non-admins can participate, and consensus decisions generally seem pretty in line with the group as a whole (instead of just the Admins). AE is somewhat more power-concentrated (DS in general). Nosebagbear (talk)
I have yet to see anyone present a reason why any sane person would ever want to be a Wikipedia Administrator. First you go through hell at RfA, then you are either constantly attacked for doing your job or you get tired of the constant attacks and become a "deadweight admin". Maybe being willing to accept the position of admin should disqualify you from being an admin on grounds of insanity. Would anyone here like to try giving me one reason why I would ever want to become an admin? --Guy Macon (talk) 13:08, 3 August 2019 (UTC)[reply]
@Guy Macon: I think the answer is that the wide, wide majority of the day-to-day work that needs to get done by Wikipedia administrators is mundane. If you want to be an administrator who makes it your sole job to mediate contentious disputes and calm down angry experienced users, then I can certainly see how you might get burned out very quickly. However, the flip side is that it is also easy to be a low-profile administrator who continues doing exactly the same thing you had been doing before you were an admin, only now you don't need to ask for an admin when you spot an obvious spammer because you can delete, block, and/or protect all by yourself. Mz7 (talk) 11:01, 12 August 2019 (UTC)[reply]

Layout

fun fact: phrases like “to the right” to describe a graph do not reflect the way information is rendered across multiple platforms.~T P W 11:59, 3 August 2019 (UTC)[reply]

Thanks for the reminder. I'm usually more careful with that, and I often edit on keyboardless devices with odd layout myself. Just imagine what it's like for our voice-to-text audience or others differently abled! ☆ Bri (talk) 14:16, 3 August 2019 (UTC)[reply]

Further analysis

Wow (and wow), very interesting analysis and the trend seems unrelenting but bizarrely linear (as if there is some natural "hidden hand" behind this decay process).

It would be interesting to follow up this analysis to clarify whether this is an "admin crisis" (e.g. we need to find ways to get more admins; which I think is an accepted perennial issue), or a wider "engagement crisis" (e.g. is Wikipedia a waning, or even dying, project?).

For example, what is the graph on the ratio of pages (or page edits, or editors) per admin. If pages, edits, and editors continue to rise, we have a "higher-quality problem" (e.g. too much demand vs. resources (e.g. admins) to handle it). However, if these statistics are also declining in line with admins, then we have a more serious problem, and the fall-off in admins is only an indicator of a wider issue?

Britishfinance (talk) 12:21, 3 August 2019 (UTC)[reply]

WP:FRAM aside I was really surprised by how linear it was as well. I suppose we would expect people to apply and stop editing at roughly the same rate through the year, and thus inactivity losses. However I was surprised by how standard the resignations were (must be) as well. Nosebagbear (talk) 13:11, 3 August 2019 (UTC)[reply]

I believe the majority of admin losses are simply a result of the inactivity rules, so it's not hugely surprising to me that it would be consistent. Sam Walton (talk) 14:47, 4 August 2019 (UTC)[reply]

What decent editor would want to join a power structure that has a reputation for corruption and bullying? Having been here since 2005, I can say: it has worsened in the past two or three years. Tony (talk) 23:30, 9 August 2019 (UTC)[reply]

How about paying admins for their time for "administrative" activities and not allow them to edit articles - to get more admins

First I'll admit that I have minimal idea of what admins do (other than Noticeboard disputes, blocks, and such) so I'll be the first to admit this may be a stupid idea, since I haven't thought it through, and don't know enough about the role to do that, but I just thought I'd suggest because I hadn't seen it suggested. ---Avatar317 ^(talk) 22:01, 3 August 2019 (UTC)[reply]

Given that it takes thousands of edits for someone to know enough about wikipedia to be an Admin, I don't think any of them would be willing to give up all editing, even if they were paid for their admin actions. There's also a community desire for admins to keep a hand in so they know what it's like for editors. Nosebagbear (talk) 09:59, 4 August 2019 (UTC)[reply]

I would favor something that is somewhat the opposite of the above. I think all WMF employees and especially board members should have to, as a requirement for keeping their job, put in at least a couple of hours of editing Wikipedia every month. --Guy Macon (talk) 16:01, 4 August 2019 (UTC)[reply]

While I personally think Guy's thought is exactly right (and I've thought it since my naivety crumbled when I realised how little experience WMF staff had), but of course, some of the legal protections might not apply to their edits, which would probably discourage the WMF. Nosebagbear (talk) 16:20, 4 August 2019 (UTC)[reply]

They may have to, Guy - at the rate we're losing/disenchanting editors - does anyone have a graph that shows long term active editors or don't they matter in the grand scheme? ^Atsme _Talk 📧 22:07, 4 August 2019 (UTC)[reply]

"Active editors" (English Wikipedia editors with >100 edits per month)

@Atsme, there you go. Contrary to popular belief the "active editor" count has actually been rising slowly but steadily since the end of the 2007–2014 crash. ‑ Iridescent 13:22, 5 August 2019 (UTC)[reply]

Iridescent, thank you - did you create the chart? Do we know what kind of editing/topic areas we can attribute to the rise and fall? For example, prior to an election, the AP2 topic area attracts new editors and stimulates editing. Does the graph represent article edits, noticeboards, UTP, or ??? ^Atsme _Talk 📧 13:54, 5 August 2019 (UTC)[reply]

@Atsme, it's modified from Template:Wikipedia editor graph (100 per month) with the Y axis increased to make the three separate trends (massive expansion from 2002–2007, collapse between 2007-2014 and gradual but consistent rise 2014–2018) more apparent. The raw data from the WMF is here. The data is just a straightforward count of people who've made 100 edits in the past 30 days, and not affected by which namespace (which is as it should be; someone who makes 100 edits to an article talkpage is just as invested in Wikipedia as someone who corrects 100 instances of "doe snot" in the mainspace). The data only goes up to 2018 because in December 2018 they stopped publishing the "active editor" (i.e. 100+ edits per month); the metric they now use is 5+ edits per month, presumably on the grounds that "Wikipedia has 70,000 editors" sounds better than "Wikipedia has between 3000 and 4000 editors" to the donors.

Other than the seasonal dip in December each year, I very much doubt rises and falls in editor numbers can be ascribed to any particular event; there's some kind of election, major sporting contest, war etc going on somewhere virtually all the time. (You notice the editing spikes in AP2 because that's a topic in which you're interested, but they're dwarfed into insignificance by the editing numbers generated by sports leagues as every single game played necessarily generates at least one edit apiece to the articles on both teams and on every player involved.) At least some (but certainly not all) of the post 2007 crash can be ascribed to the combination of the post-Siegenthaler expansion of semiprotection and the introduction of Cluebot leading to a general drop in vandalism and there consequently being fewer people racking up large numbers of edits reverting vandalism. For what it's worth, the start and end of the decline correspond exactly with Sue Gardner's tenure as CEO. There's some (rather rambling) discussion on the thread on my talkpage that originally prompted Bri to write this piece. ‑ Iridescent 19:00, 5 August 2019 (UTC)[reply]

That is really interesting analysis; feels like a longer article with these metrics is worth considering? Britishfinance (talk) 21:19, 5 August 2019 (UTC)[reply]

We hear non-stop about declining admins, implying WP is a waning/dying project? However, metrics like above give a different picture. Britishfinance (talk) 21:33, 5 August 2019 (UTC)[reply]

Iridescent - something you may possibly find of interest re: declines of editors. I found it on the UP of Jorm. I'm not trying to take away from the declines we've recently experienced in our admins, but we also cannot forget that admins come from our pool of editors. ^Atsme _Talk 📧 13:21, 7 August 2019 (UTC)[reply]

The admin/sysop decline is a side-effect of the editor decline, absolutely. I remember a long conversation with fuzheado about this very thing as far back as Wikimania 2012. I have an idea of what has to happen to fix it, but it's not popular. --Jorm (talk) 21:44, 7 August 2019 (UTC)[reply]

Where do the bread crumbs lead, Hansel? ~~Gretyl 21:50, 7 August 2019 (UTC)[reply]

The gender gap. That's the biggest thing. Fixing the gender gap will solve for the decline. Fixing that is a gloppy mess of solutions but in the end they won't mean shit as long as popular/prolific editors don't get sanctioned for bad behavior, people throw the c-word around, and the community disbelieves the stories of harassment victims by default and instead turns them into subjects of interrogation.--Jorm (talk) 22:08, 7 August 2019 (UTC)[reply]

I somehow doubt "c-word" has anything to do with "crowd-sourcing" but there is a connection. I'll just say that it takes a pretty heavy coat (full-length mink?) to deflect the chilling effects of a whirring boomerang. ^Atsme _Talk 📧 00:00, 8 August 2019 (UTC)[reply]

If the decline in numbers of admins was simply a side effect of the editor decline, then 2014 should have been the low point and the 2015 rally in editing should by now have fed through to a rally in adminship. Editing volumes on at least one measure are above 2014 levels, yet last year (2018) we had less than half as many successful RFAs than in 2014. Five years ago we had over 600 active admins, now it is just under 500. If the only factor in play was the 2007-2014 decline in editing, then RFAs would not have dropped from over 400 in 2008 to 10 in 2018; If anything we would now have more active admins than in 2008, non admins would be rarer among the regular community than they were a decade ago. Numbers of new editors is down, and you'd expect that the number of RFAs would be partly linked to the number of new Wikipedians who joined us one to three years ago. But of our current admins, ignoring bots, only sixty started editing in the last ten years and only four of them in the last four years. Why isn't there a single editor who joined us in 2016 and is now an admin? Fixing the gender gap and recruiting more female editors would increase our number of admins (and fixing the bigger problem of our ethnicity gaps would help even more), but I'm pretty sure we had big gender and ethnicity gaps in 2008. Ϣere SpielChequers 21:36, 8 August 2019 (UTC)[reply]

"Why isn't there a single editor who joined us in 2016 and is now an admin?" I doubt there are many editors that joined in 2016 that are qualified for adminship. ~~Certainly if there were, they're not interested.~~ I note that Chetsford is looking at a successful RfA now, and he joined in 2017 so it can happen. Chris Troutman (talk) 21:48, 8 August 2019 (UTC)[reply]

We have three admins already from the class of 2017, anyone who started editing in 2016 has been here for more than two and a half years, and with the de facto minimum for RFA being 12 months experience we could have a successful RFA from someone in the class of 2018 (there would be opposes, but we have had one candidate pass in recent years with only a year's editing). Given the numbers I suspect the gap in 2016 could be a statistical fluke, the bigger point is that I think our wider wikigeneration divide is unhealthy. Ϣere SpielChequers 22:11, 8 August 2019 (UTC)[reply]

Last Signpost report on active editors

I'm not able to scour the entire archive right now but I found Wikipedia:Wikipedia Signpost/2009-01-03/Editing stats by Ragesoss. It included more analysis of the numbers of non-administrator editors than was provided in this 2019 item. If there is reader interest maybe we could do a follow-up report with something like that. ☆ Bri (talk) 14:40, 5 August 2019 (UTC)[reply]

That would be quite useful...especially to demonstrate trends/triggers regarding what drives editors to & from the site. Also, academic projects may be significant enough to distinguish. ^Atsme _Talk 📧 16:15, 5 August 2019 (UTC)[reply]

When I wrote Wikipedia:Wikipedia Signpost/2010-08-09/Admin stats nine years ago I talked about the Wikigeneration divide between the generation that supplied the admins and the current community. At the time over 90% of our admins had been wikipedians for more than three and half years, now that is over 99%. When I wrote that article the project was less than ten years old, now over 90% of our admins have been Wikipedians for more than ten years - though I suspect our few newish admins are doing a disproportionate amount of the admin activity. Ϣere SpielChequers 21:54, 8 August 2019 (UTC)[reply]

I would be interested in a follow-up article, especially since this one was from 2009. I generally find statistics interesting and I would definitely read a Signpost article that focused on that. Clovermoss (talk) 22:16, 22 August 2019 (UTC)[reply]

Moved Long Post

<!-In my opinion, when placing a reduction of number of administrators, as a whole, websites or entities should not post or show number reduction, for it unintentionally invites people off criminalistic behavior to be more voluntary and brazenly to change and illegally alter wording or facts to there agenda or favor. Hense in my claim, perpetrators have stolen income and military income at that as long as stalked and alter the information with redirected wording to maintain a continuous theft ring within the criminal faction that has complete control of all BUISNESS and household devices to prolong control of money and communication. If this was a rival company, this would be exactly how to illegally close BUISNESS to steer markets or marketing to illegal factions direction.

Administrators are needed not so much for policing but knowledge base to make sure policy and wrong doings are caught and sought out immediately. As on my keyboard, you can see the malware still has a control on device and nightly or daily nation wide there team on synchronized timing will use all redirects and enter my wife's, kids, and my devices at our domain and pump data along with carrier drop offs right out of our devices to be transported via music firm usually, then laundered into cash. They do this at the direction of a company that I hope for one last chance will take the situation serious enough to block the perpitrators from ever using there system again. For this company would be a dominant growth in our nation's economy and a true intavator along with companies and websites like Wikipedia and Google inc. In my opinion if you are caught performing illegal redirects or stealing data at any level or amount, should automatic be banned or disbarred from usage off website indefinitely. The actions off these criminals have invited more people to learn the way of the malware to which in turn creates more crimes of what this GROUP CALLS LEGAL THEFT THROUGH WIKIPEDIA.I don't hold no one responsible but the thieving operators.But this is the beginning, eventually the cash cow always runs out. And with narcissistic behavior, They will begin to steel social security from the elderly and veterans and so on.

Im not saying they need more oversight on Wikipedia,I think there should be marked barriers so as it be if someone is making a redirect, the system automatically notifies administrator to watch and clearly check to make sure the redirect is off practical legality.And if not said person is to be notified if infringing on websites by laws and given one warning, But that person is flagged and coded so they cannot go to a dark corner and do the same crime anonymously. I could think off other things that would make sure safer and less crime preventing, but I'm sure you have people to place and replace rules to effect criminals.Let me say this though as of to this point over a trillion dollars has been stolen through redirected writing on websites, on just my case alone, to which had shattered not only my future but my children's and all my extended family to which has even drive my wife and I to Divorce.Without the means to travel criminals have less chance of stealing!----------------------------------------------------------------> — Preceding unsigned comment added by 2604:2d80:c100:5700:d42:416d:b866:77da (talk • contribs)

@2604:2d80:c100:5700:d42:416d:b866:77da: this page is somewhat dead as it's on an old article, but I've read your very long post and I'm still no clearer as to what the issue is - I think your use of redirect is different from the one we use. Anyway. If you have an actual issue please take it to the TEAHOUSE which helps new editors with issues. Nosebagbear (talk) 10:57, 8 November 2019 (UTC)[reply]

Want the latest Signpost delivered to your talk page each month?

Home

About