We laughed, we cried, we closed as "no consensus"
Hell of a year, huh? Or, perhaps, hell of a "twelve months that didn't really feel like a year had gone by"? We had one here on the Internet, that's for sure. And nowhere was it wackier than the free encyclopedia that anybody could edit. In 2021, we've spent our days hashing it out over nearly 22,000 individual nominations (around 60 per day). Most of them were not very tumultuous. Notability standards are invoked, disputed and interpreted in much the same way across many similar AfDs — the obscure villages, minor-league athletes, unreleased movies, minor high schools, groundbreaking startups, and tiny stubs start to blur together after a while. But every once in a while, something really special happens, the planets align just right, and a bunch of people type out thousands of words of text in a deletion discussion.
Sometimes it is just a protracted debate about arcane details of little relevance, and sometimes it is the extension of some long-roiling personal dispute between participants, full of sound and fury signifying nothing — but sometimes questions are raised that cut down to the core of what we think Wikipedia is, what we think it isn't, what we think it means, and what we think it should be. Questions that redefine what the largest encyclopedia in the history of human civilization wants to be about, and what role it wants to play in that civilization as a de facto adjudicator of which subjects are noteworthy.
Much money and power is spent in attempts to influence that adjudication — whether through outright paid editing, the volunteered advocacy of earnest believers, or our own beliefs as editors. We live in a society, but that society lives within us as well, and what we choose to cover as an encyclopedia is part of an intricate dance with that society. Love it or hate it, it matters.
Over the years, deletion processes have changed, because notability standards have changed, because Wikipedia has changed, because the Internet has changed, because the world has changed. One wonders, of course, to what extent the reverse order of this causal chain is true — not very, but probably at least a little bit.
So how many times have we brave, strange, monastic defenders of knowledge and freedom undertaken our solemn ritual? According to the database, a little under half a million (482,316 to be precise).
It seems like something that's this big a part of our project ought to have some attention paid to it, and probably more attention than human beings can pay on their own, so earlier this year I wrote some software to help with the task. The Oracle's unblinking electric eye is able to keep watch over all of those AfDs, from 2005 up until today, and from it we can glean some insights into the process which shapes our encyclopedia. So what are they?
From 2005 to 2020, there have been an average of around 28,000 deletion discussions per year; a lot more detail about this (including impressive graphs) can be found at the Oracle's all-time stats page. Suffice it to say, however, that 2006 was the all-time peak (54,606) and since 2014 the yearly average has hovered somewhere around 24,000. This year, there were almost 22,000, slightly fewer than in 2020 (24,000) but considerably more than 2019 (18,000).
As has been the case in every year since 2005, the majority (about 15,600, or 60%) closed as "delete" or "speedy delete". About 5,400 (25%) closed as "keep", "speedy keep" or "no consensus". About 2,700 (13%) closed as "merge" or "redirect". The rest closed as "userfy", "draftify", "withdrawn", or "transwiki". 134 AfDs (0.01%) had a close that my software couldn't parse.
There have been a number of surprising, contentious, unprecedented and downright strange deletion discussions in 2021. In March, we saw an ArbCom case related to mass creation of non-notable "geostubs", and in October Lewis (baseball), a featured article (which had appeared on the Main Page just several months earlier), was merged into 1890 Buffalo Bisons season. Last month, we had the largest AfD in history. Besides that, a lot of individual discussions have proved to be very active: hot-button political topics, detailed analyses of academic publishing trends, hot-button political topics, Internet memes-du-jour, chess tournaments, and hot-button political topics. Let's get to 'em.