I have been a Wikipedia editor since February 2005 and was an admin on this site from August 2007 until November 2024. I lost my admin status after failing a re-request for adminship (RRFA) following the first ever admin recall, largely due to my aggressive blocking and heavy-handed treatment of new users. The RRFA is the subject of this article, in which I offer explanations for what led to it and what it was like to undergo it. This is not a general overview of everything I've done on Wikipedia; for that, see my 2017 Signpost interview and my personal Wikipedia timeline.

I became a Wikipedia administrator in August 2007 following a unanimous request for adminship discussion. I have had four main preoccupations on Wikipedia: wikiarchaeology (investigating the old history of the site), mostly history merges and imports (the latter being the reason I obtained the importer user right in 2013), my daily skim-reading of selected Wikipedia noticeboards (such as the technical village pump, the main admins' noticeboards, and later the bureaucrats' noticeboard), Wikipedia's accessibility for blind screen reader users like myself, and, most importantly for the purposes of this article because it's where I found many of the users I'd blocked, my (until recently ever-expanding) watchlist. It contained not only articles that I'd worked on, but also many articles subject to vandalism/deleterious editing that had not been undone for more than a couple of days. When I became an admin I tried my hand at dealing with regular backlogs (like speedy/proposed deletion, which I found unfulfilling), and I did not regularly check pages where users requested protection, asked for vandalism blocks, or reported bad usernames because I prefer to be self-directed, so my watchlist became my Wikipedia fortress, so to speak. I would often check all the recent edits (or those that were the top edit) of any new user/IP to appear there. As I encountered more contributors making various types of problematic edits over the years, I became more paranoid about new users and their possible motives, leading me to become more and more aggressive about blocking. I became especially afraid of new users who I thought were gaming auto/extended confirmed permissions by making small edits (such as overlinking and often unsuccessful attempts to copyedit articles) to increase their edit count and school IPs (as many articles were on my watchlist due to vandalism from schools). I fairly frequently blocked users/IP's who I thought were problematic with little or no warning because I felt that allowing them to continue editing would waste the community's time and I valued the integrity of the encyclopedia above just about everything else. My friends, especially the occasional Wikipedian Codeofdusk, half-seriously called me a "Wikipedia meanie" for my approach to dealing with new users. Most articles on my watchlist were there because no other active users were watching them, so most of the time I obtained relatively little feedback about my admin actions except from the new users themselves (which I generally discounted) and the occasional admin/established user who either praised me for doing a good job or let me know that I was getting out of line with community expectations; in the latter case, I'd try to heed their advice.

That all changed with two admins' incidents noticeboard discussions started in September 2024: "Overzealous blocking by Graham87", about one of my blocks, and "Inappropriate blocks and WP:BITE by Graham87", about my general approach to blocking and new users; I took on board the advice in both of those threads. There are more details on my personal Wikipedia timeline subpage about how I lost my adminship, which details several sliding doors moments in which if I'd made better decisions, I could've kept my adminship; the subpage also includes a link to a relevant thread on Wikipediocracy. Meanwhile, the foundations were being laid for the admin recall process, in which 25 users had to sign a petition to force someone to undergo an RRFA within 30 days, and I happened to be the first person to be subjected to it. It was going about as well as a first try at a completely unfamiliar process can go on Wikipedia (i.e. not particularly well), but in the first nine days it had only received 12 out of the needed 25 signatures ... until 5 November when this edit adding a good-faith, but irrelevant reference to the Sleepover article appeared on my watchlist. I checked the other edits by the user who made it, was so incensed by the quantity and nature of their edits (as they pressed so many of my buttons at once) that I was hyperfocused on them and didn't even think about the recall petition, and eventually gave them a remarkably poorly executed and communicated indefinite block ... which meant that all hell broke loose. My petition reached 27 signatures in over half a day, and I needed to face a new request for adminship if I had any chance of keeping my admin status.

After the above-mentioned block and its fallout, I had a huge amount of soul-searching to do. I wanted to keep my adminship largely for wiki-archaeological work which sometimes requires it. I realised that to have even a sliver of a chance of keeping it, I should rid my watchlist of articles I wasn't interested in that were potential newbie magnets and pledge to avoid blocking. I originally saw my watchlist purge as a dereliction of duty (though the idea had been at the back of my mind for many years beforehand), but I later saw it as liberation. I spent seven hours sorting through my watchlist, removing 1,173 pages, which calmed it down considerably. As many voters in my RRFA did, I knew that my no-blocking pledge was technically unenforceable and that my attempt to hide the block links in my common.css was akin to putting an alcoholic's drink of choice in a slightly more difficult-to-reach part of the fridge. But without the political will to separate the block button from the admin tools, that's all I could do.

I also had to find potential nominators, by asking people I'd worked with in the past to endorse me for adminship. It took some time but finally my re-RFA was ready and I was relatively optimistic. What happened next is best summed up at the timeline of how I lost my adminship, the relevant parts of which I have copied here:

• 17 November: my re-request for adminship went live. Unlike a normal RFA whose threshold is 70%, this one was to have a passing threshold of 60% with any result between 50 and 60% up for bureaucrat discretion. By the time I'd signed off that night, it was at 42/5/2, so I was feeling pretty optimistic.
• 18 November, by the time I'd answered my latest batch of questions, my RRFA was at 67/33/5. Not as good as the night before, but still not the end of the world.
• 19 November: By the time I'd surfaced, we were at 105/109/7 with rumblings about withdrawing. We were below the 50% cutoff that meant failure. I'd earlier decided to withdraw at 45% (but didn't want to make the exact figure public at the time). I held on to hope that maybe things would turn around.
• 20 November (local time): The day I withdrew my RFA. At the start we were at 115/131/9 by the time I got up, not much different from 111/124/8 the night before, so I kept on holding on. After getting back from the gym later that day, the RRFA looked something like this, with a discussion of another harsh block I'd made and a very long oppose that gave me pause. I then spent a few hours thinking about what to do before I finally withdrew the RFA, gave myself some user rights that I'd need later (but avoided giving myself rollback because some people in my RFA objected to my use of that right), and let the bureaucrats know. An RFC was started later that day to allow non-admin importers (I'm the only one) to be able to history-merge pages and it was implemented nearly a week later.

In short, by blocking users to avoid wasting the community's time, I ended up creating a colossal time sink discussing my adminship status; I'm very sorry about that.

The last few months have certainly been a wild ride but the support of my friends and family along with music have helped immensely. The fact that I spent my first two-and-a-half years as a Wikipedian without admin tools (which was a long time by the standards of 2007, the year I gained adminship) makes it a bit easier to adjust; I've just revived my old habit of adding speedy deletion tags like {{db-G6}} for uncontroversial maintenance (like pages left-over from history merges) rather than deleting the pages myself. I've been missing the tools in the strangest ways. For instance, when I notice an unfamiliar user, I've often been in the habit of comparing their account-creation date to the date of their first edit and checking their deleted edits if the two don't match; I can't do that now. I also didn't realise until I lost the rollback tool that I sometimes used it for orientation; my screen reader puts links like the rollback link on their own line and that was helpful when navigating lists of edits ... and a line like "rollback: 3 edits" let me know that there were three edits by a particular user to navigate past before I find other edits to potentially check. That's a bizarre way to use rollback and I doubt I'd use it much for its intended purpose these days after the watchlist purge. If someone wants to grant me the rollback right, I'd be OK with that, but I also don't want it to cause any controversy. I've missed the tools in more tangential ways too, like having to ask for a page undeletion to facilitate a history merge. As for blocks, as of the publication date of this Signpost piece, three out of the four blocks I've requested since losing adminship have been carried out.

I can't lay all the blame on this, but I sometimes wonder if the newcomer homepage has a hand in some of the edits I've noticed in recent years. One thing it encourages people to do is copyedit, which is a great first editing task for a competent native speaker of English, but not a good one for people whose English is more shaky than they realise (compare the Dunning–Kruger effect).

The situation I was in reminds me a little of the 2005 Stevertigo arbitration case, one of the Committee's first desysopping cases, that I've never forgotten because of the debacle of the reconfirmation RFA the committee forced him to undergo; they've never done that sort of thing since. I do hope a re-RFA is always an option for admins with certified recall petitions though; I've implied this elsewhere.

Despite the setback of losing my adminship, I hope to continue editing Wikipedia for many years to come. I'm ranked #5 among human editors on the list of Wikipedians by longest consecutive daily editing streaks and I intend to keep that up.