Work Text:
This year has been ... I don't even have words. But I haven't written or done basically any fandom stats or anything till now, and I know many people have been far more directly impacted than me. So I've been really curious about how fandom on AO3 has been responding so far. I gathered a bunch of data about AO3 from Jan 1-Oct 1 of 2020 and compared it to past years.
Note that this is not the total number of fanworks on AO3 -- this is just the number of single-chapter works produced between Jan 1 and Oct 1 over the past decade. (I eliminated multi-chapter works because otherwise all the works that were updated recently would get included in 2020, and that would skew the numbers.)
The 2019 surge is partly due to the Chinese fanwork migration; see GGC19 stats (second half of Chapter 1) for further discussion and relevant fandom data analysis.
I mean, obviously it has. But the surges in fanworks created have also been more extreme from month to month. (Here again I'm only including single-chapter fanworks, for the same reasons.)
I thought about trying to label major real world events for comparison, but… tbh I got depressed by the nature of the recent events. But I did end up breaking it down by languages, and we can see different patterns for different languages...
First, the top 3 most common languages on AO3.
Then the rest of the top 10 languages on AO3 overall.
These are all fandoms with >10K new/updated works in 2020 -- no longer limited to single-chapter works. (the earlier works are also represented in a lighter color.) I eliminated some highly overlapping tags, like a few variations on “The Witcher” tag; most notably, “Marvel” would be the second biggest tag if I hadn’t removed it. I also shortened some tags that were very long (mostly multilingual tags).
These are the same tags, but revisualized as a word cloud because I know some folks find those easier to parse. The size of the fandom here represents the # of new/updated works in 2020.
Observations -- these are not as dominated by Western live action TV/movies as in past years (in part almost certainly due to the surge in Chinese participation). Lots of video games/mobile games. Lots of animated content A podcast and web series. Lots of Chinese and Japanese properties. Plus a bunch of K-pop.
* Defined here as fandoms with >2K new/updated works in 2020, and limited to fandoms with >50% of fandom’s works new/updated in 2020. This list is HIGHLY variable depending on the min # of works threshold, which was very arbitrary. (Please don’t stress if your fave isn’t on here.) E.g., when I drew the threshold at 1K works, the list included several very fast growing Thai fandoms (and also some of my personal faves, like Killing Eve).
Here are the same fandoms shown in the previous slide. However, as in the previous word cloud, the size of fandom represents the # of new/updated works in 2020 -- that means they’re ranked differently than on the last slide, where they were sorted by the proportion of works produced in 2020. (I thought it might be useful to also see them ranked by absolute number of new works.)
Not my image -- just a screenshot of AO3’s tag cloud...
These are the tags with the most new/updated works in 2020 -- threshold was 50K active works in 2020.
For tags that have better known subtags and/or that contain several very different subtags, I’ve also listed some of the popular tags (shown just below the relevant bar).
Note: I don’t really understand how Alternate Canon and Deviates From Canon differ -- they have lots of subtags in common but are not identical.
I didn’t end up making an alternate visualization word cloud for this one, because it was pretty similar to the bigger tags on the previous slide (AO3’s overall tag cloud).
*Defined here as tags with >5K new/updated works in 2020; >33% of works new/updated in 2020. This list is HIGHLY variable depending on the min # of works threshold, which was very arbitrary.
Hurt/Comfort was a mega tag that came closest to making it into this word cloud, with 30% of its works produced in 2020. Seems appropriate to the year.
Some of these, like Penises, Blood, and Ambiguity, contain several fairly different popular subtags.
Note that a bunch of these tags are more commonly used in mult-chapter works than single-chapter works; that makes sense because this analysis includes longer works updated in 2020 (and so contains a higher percentage of multi-chapter works than a random sample of the archive would).
As for previous word clouds, size of tag represents # of works new/updated in 2020, so these are sorted in a different order than on the previous slide.
I broke down “AO3 Tags - Freeform” into the qualifying subtags, because that one was particularly meaningless on its own (e.g., “Other Additional Tags to Be Added,” “How Do I Tag”)
... And that's it (for now, anyway)! Hope you enjoyed.