• If I'm reading this right, and 2018 saw 10 total successful RfAs, 6 inactivity desysops, 1 resignation, and 2 (permanent) emergency desysops… By my count the net change in the number of brooms being pushed on the project for 2018 is 1. --Xover (talk) 15:11, 24 December 2018 (UTC)[reply]

• Not quite, you are getting monthly and annual numbers mixed up. For instance there were three more inactivity desysops and one resignation listed in the December 1 News and notes. The totals are strongly negative for 2018. "Down 50 in the last twelve months" said Kudpung in his May Op-ed. ☆ Bri (talk) 16:23, 24 December 2018 (UTC)[reply]

• @Bri: - I was wondering if we had any idea of number of properly active admins (say, at least 5 actions in the last month etc)? Nosebagbear (talk) 16:29, 24 December 2018 (UTC)[reply]

• Yes, the roster is kept at Wikipedia:List of administrators/Active. It stands at 514 admins currently, vice 539 on January 1. I'd also invite readers to go back to my February 20 News and notes item, "Wandering in the RfA desert of 2018" and the reader comments, especially Widefox's excellent graphs illustrating the problem. ☆ Bri (talk) 17:12, 24 December 2018 (UTC)[reply]

• "Active" isn't an ideal measure as it looks at logged editing activity, so you could have an active admin who hasn't used the tools in years. But it is the measure we have available to us, and it has the virtue of having been collected consistently over time. It is showing a drop of about 24 year on year - that's almost a 5% annual contraction. ϢereSpielChequers 17:15, 24 December 2018 (UTC)[reply]

• You might be right. I did a spot check of the "active" admins starting with 'A' and there are plenty who aren't really very active. Some are just doing routine stuff that could be automated, like deleting empty categories. At least one did just two logged actions of any type in 2018 (they were both page creations). Many are just doing page moves (which is unbundled [I have that right myself]). I'm sure there's some gaming the system to keep the bit. A better picture would be "who did more than N of [block|page protection|revdel]" where N is a significant number like 20 or more. ☆ Bri (talk) 18:30, 24 December 2018 (UTC)[reply]

• At the risk of tooting my own horn, this is sort of what I created User:Amorymeltzer/s-index to look at. There are plenty of failings with it, but it tries to get some measure of logged sysop activity over time by calculating S, the number of sysops who made S logged actions in a given time period. ~ Amory (u • t • c) 02:02, 25 December 2018 (UTC)[reply]

The biggest question, though, is whether there are insufficient admins to actually keep the place afloat. Or anything close to that. As far as I can tell most backlogs that require admin intervention are not hopelessly long, but I may be mistaken. On another note it's worth bearing in mind that some admins concentrate on tasks that don't show up in the "administrative action" count, such as closing discussions or editing the main page.  — Amakuru (talk) 22:43, 24 December 2018 (UTC)[reply]

Factcheck (again): Could you be more specific about the sockfarm you mention during Galobtter's RfA? As I understood it there was one support vote (which was counted) from the Sagecandor/Yetishawl/Cirt sockfarm, and one yes-voter and one no-voter (neither of which was counted) from an unrelated sockfarm. ~ 🐝 ~ SashiRolls ^{t · c} 18:16, 25 December 2018 (UTC)[reply]

SashiRolls, See Wikipedia:Sockpuppet investigations/Korruski/Archive. (something you could have checked yourself). The purpose of mentioning the socking is not specifically to discuss SPI, but to draw attention once more to how corrupt RfA can be and that it is essential to be vigilant. Kudpung กุดผึ้ง (talk) 00:36, 29 December 2018 (UTC)[reply]

Yes, Kudpung, I saw this. What I found strange was that you mentioned the sock whose vote was struck/stricken, but didn't mention the sock whose vote *was* counted in the election. Who knows, maybe being the main author of the Wikipedia page on the Signpost brings some special perks with it, like not having your sock drawer reported in the paper of the same name. ^^ SashiRolls ^{t · c} 00:49, 29 December 2018 (UTC)[reply]

'Special perks', SashiRolls? I doubt it, maybe this thread will explain what your own research missed. Kudpung กุดผึ้ง (talk) 12:52, 29 December 2018 (UTC)[reply]

• It's notable that one admin has made about 1,000,000 deletions. I'm not sure that if we lose that one we have enough resource to fill the breach. In fact, I have often thought that (and a small sample quite a few years ago seemed to confirm) we are not sufficiently careful about speedy deletions, even when we had far more active admins. All the best: Rich Farmbrough, 14:21, 31 December 2018 (UTC).[reply]
  

• I don't agree that we've a shortage of admins now. I remember huge administrative backlogs being dealt with during the first few years after I got the mop in 2007, and those backlogs are gone. There was more work for a larger corps then, not so much now. – Athaenara ✉ 04:30, 29 December 2018 (UTC)[reply]
• I'm glad to see a dark mode added to Wikipedia (I was hoping for something like this before, and installed a plugin into my browser to do pretty much that!) That said - and correct me if I'm wrong - but unless the account's been hacked or is otherwise being abused, desysopping an inactive admin seems like a solution in search of a problem. PrussianOwl (talk) 19:18, 7 January 2019 (UTC)[reply]

• @PrussianOwl: Please forgive me for whatever my problem is, but what is "dark mode" in this context? Is it reverse light/dark in text display so letters are light on a dark background, or something more nuanced or subtle or political? – Athaenara ✉ 08:42, 12 January 2019 (UTC)[reply]

Never mind, scratch that, I see now "toggleable dark or night-mode" up there. – Athaenara ✉ 08:44, 12 January 2019 (UTC)[reply]