Wikipedia talk:Desysoppings by month

Colours
This is an interesting bit of information, but I think it would be more easily interpreted as a hot cold matrix similar to my successful RFAs one. Would you mind if I added some colours?  Ϣere Spiel  Chequers  21:38, 21 April 2013 (UTC)
 * Thanks. I'm certainly for the idea, but left the colors out because of the difficulty of coming up with a scheme that accurately coped with the two "eras" of desysopping - before and after the inactivity rule - and not being distorted by the massive bulge of the first month of the second era. Perhaps it would be possible to have two schemes, that corresponded in brightness but differed in color. What do you think? — Scott • talk 09:54, 22 April 2013 (UTC)
 * The month with 230 is a problem and would go way off the scale. But as for the rest I would go with one scale and a text explanation as to the difference between the eras. If you don't mind I will try and see what it looks like on the same scale as RFAs by month.  Ϣere Spiel  Chequers  10:10, 22 April 2013 (UTC)
 * Go for it. — Scott • talk 10:52, 22 April 2013 (UTC)
 * OK I've done it, though with a flattened scale to reflect the lower numbers in most months. What do you think?  Ϣere Spiel  Chequers  12:43, 22 April 2013 (UTC)
 * Nice work. Thank you. — Scott • talk 13:07, 22 April 2013 (UTC)

Question
I'm asking you, because I figure you'll know the answer – this table represents "Total desysoppings", i.e. both due to "inactivity" and due to ArbCom (etc.) action, correct? Has anyone broken out the data for "inactivity desysoppings" vs. "ArbCom (etc.) desysoppings" over time? Thanks in advance. --IJBall (contribs • talk) 16:29, 26 July 2015 (UTC)
 * hi IJBall, I believe you're right - these figures contain both sorts of desysops, otherwise they wouldn't have both the spike from when inactives were first desysops and the steady flow of ones beforehand. I don't know if anyone has tracked them separately, but the raw data is at Category:Wikipedia_inactive_administrators.  Ϣere Spiel  Chequers  17:33, 26 July 2015 (UTC)
 * Note: At this topic at Liz's Talk page, Liz provides some excellent links to some handy resources for these questions... --IJBall (contribs • talk) —Preceding undated comment added 19:13, 26 July 2015 (UTC)

Approaching break even
The top chart shows the closest recent year to stability as 2013 with a net loss of 23 admins. But the bottom chart shows 2014 as closer to full replacement as 2013.  Ϣere Spiel  Chequers  11:52, 1 September 2015 (UTC)
 * Yeah, the 'Visualization' chart does not match the 'Yearly change in admins' figures from the top chart. So I'm guessing the 'Visualization' chart's figures come from somewhere else., care to comment?... --IJBall (contribs • talk) 00:35, 3 September 2015 (UTC)
 * I made it based on whatever the figures were at the time, I'll make a new version when I have time soon. —  Scott  •  talk  21:30, 3 September 2015 (UTC)
 * Done. —  Scott  •  talk  23:44, 7 September 2015 (UTC)
 * Note: The previous 2013 figure seems to have been in error – I have corrected it, so it looks like there was a net loss of just 20 Admins in 2013... --IJBall (contribs • talk) 01:02, 3 September 2015 (UTC)
 * Thanks for that. I don't know anything about the charts, but I've tried to fix the figures and make them more consistent. I obviously missed one. Graham 87 03:24, 3 September 2015 (UTC)
 * I have to admit – I'm not sure where the 'Total restorations' data comes from. I mean, I sourced it to WP:RESYSOPS, but I did some checking and the data didn't seem to match up. So do you know where the 'Total restorations' data comes from?... (OK, I just read the text at the bottom, and it partially explains this, but I'm still wondering exactly how it's calculated, as it may be the key piece of data on this chart...) --IJBall (contribs • talk) 03:44, 3 September 2015 (UTC)
 * Pretty much what the note at the bottom says. I went through and checked things by hand ... if the desysop/restoration process was very raid, I didn't count it in the list. The problem is that desysoppings aren't cleary logged. Graham 87 04:02, 3 September 2015 (UTC)

Table error
I see that you are updating the table. In the "2014" column, it says: But, -79+16+22=-41 (not -40).  Vanjagenije  (talk)  21:09, 18 December 2015 (UTC)
 * Total removals: 79
 * Total restorations: 16
 * Total promotions: 22
 * Yearly change in admins: -40
 * Thanks for pointing this out. I've fixed it; [//en.wikipedia.org/w/index.php?title=Wikipedia:Desysoppings_by_month&diff=690754746&oldid=690598874 my bad]. Graham 87 04:47, 19 December 2015 (UTC)

Bot automation
I believe certain things shouldn't be done by humans, and this seems like one of them. It can't be fun updating this, right? I'd love to rope in MusikBot, if you all are happy to have the help :) Pinging primary contributor &mdash;  MusikAnimal  talk  23:46, 30 January 2017 (UTC)
 * Thanks for the offer. It's not generally a bother, as long as I keep on top of it and I don't have to adjust colours/adjust the table for a new year. However on thinking about it, it does lend itself well to a bot (I wouldn't have to remember to do it after each successful RFA for instance!), so I'd therefore appreciate the help. Graham 87 01:48, 31 January 2017 (UTC)

July 2011
What happened then to create a 'night of the long knives,' as it were? Was that when the inactivity rule came in, so all thse admns who had already been inactive for a year got automatically removed in one go? That's the only thing I can think of wich doesn't involve some kind of crisis! &mdash; fortuna  velut luna  16:02, 13 October 2017 (UTC)
 * Yep, that's when the inactivity rule came in. Graham 87 04:03, 14 October 2017 (UTC)
 * In January 2023 we will implement new policy about inactive admins so we would expect a downfall in number of admins at that time... Thingofme (talk) 09:51, 11 August 2022 (UTC)
 * Actually 105 admins has been removed in this date. Thingofme (talk) 03:12, 1 January 2023 (UTC)
 * 101, actually. Graham 87 15:59, 1 January 2023 (UTC)
 * ...Dalmations? :) --Hammersoft (talk)
 * Lol, naturally. Graham 87 07:11, 2 January 2023 (UTC)

Is July 2011 Wrong?
Looking at the first inactivity purge (July 2011) the number listed here doesn't match the number of admins removed on the meta log? With 257 admins removed (258 with 1 bot) log link and other admins removed after the mass removal aswell, the stats here seem to be coming up ~30-40 short. Terasail [✉️] 15:23, 1 June 2023 (UTC)


 * Also, there is a difference of ~+180 between the number of admins the numbers in the table would suggest (1074) compared to the actual number (888 without bots). Terasail [✉️] 15:31, 1 June 2023 (UTC)
 * Maybe ask Scott, who initially put the list together and maintained it until July 2014. Some of the earlier months' desysoppings don't count instances when admins were desysopped and resysopped again in the same calendar year, which are listed at List of resysopped users. That explains some (but maybe not all?) of the discrepancies. That might explain the error you picked up in your recent edits. Any double/triple-checking of the numbers would be appreciated. Graham 87 08:30, 2 June 2023 (UTC)
 * Re-pinging Scott. Maybe the numbers should be re-adjusted to count *all* desysops of human accounts, excluding renames. Graham 87 08:32, 2 June 2023 (UTC)
 * Yeah I don't mind going through these things slowly, it will take time (If you are going to do it, you might as well do it properly). Between local and meta logs its a bit of a pain but I was thinking of making a duplicate table for "Inactivity desysops" stating in 2011 to track using the bot logs as a new section 1 for anyone interested and 2 for easily verifiable stats while the main table is corrected. Terasail [✉️] 14:13, 2 June 2023 (UTC)
 * No objections here. Also see a 2015 discussion about this topic I had with IJBall. Graham 87 15:42, 2 June 2023 (UTC)
 * Hi both, thanks for the ping Graham. I would be completely unsurprised if there are errors in the early data, I remember finding it very awkward to come up with the right figures given various flavours of losing and regaining the bit being present and needing to be accounted for. As Terasail puts it, it was a bit of a pain! I'm very happy to hear that an offer is being made to check my numbers, it's probably very overdue. Thanks Terasail! —  Scott  •  talk  20:51, 2 June 2023 (UTC)
 * P.S. When you're confident that the numbers are correct, I'll regenerate the graph which I'm well overdue doing as well. —  Scott  •  talk  20:52, 2 June 2023 (UTC)