Femslash week in review: 2019-03-19 - 2019-03-25

High-level stats for week of 2019-03-19 - 2019-03-25

• Total works categorized F/F on AO3: 3925 (+27 from last week)


• Works I classified F/F: 2446 (+97 from last week) (1138 new, 1308 continued)


• 0.97% of all 251161 AO3 works I've classified F/F were updated this week

A few callouts this week:

• First off: thanks very much everyone who participated in my feature request poll! Here's what I took away from the responses:
  
  
  1. Readers do not, directly, care about data accuracy: they care about features (analysis, recs) and are mostly not bothered at a high level by the idea that there are some bugs in my dataset or data use. Reasonable! I think a lot of the specific accuracy bugs are important precisely because they're blockers for features people might like, but when I'm thinking about what work to do, I should be starting with the desired feature request and fixing accuracy bugs on the way to that feature, and not sweating the accuracy bugs too much on their own merits.
     
  2. That said, everyone agrees that the revealed exchanges work undercounting bug is appalling, and I should fix that shit. (But, again, people presumably want that bug fixed primarily because recs from revealed exchanges are more likely to match the kind of recs people want to see more of (i.e. recs from smaller fandoms and ships).)
     
  3. A bunch of people want to see better analysis, and the most popular option was picking a fandom of the week and adding some over-time stats about that fandom. That'll take a little bit of doing (and, ahem, fixing some fandom-related bugs), but it'll be my next major project. And I'll take the point that people want more words about what's new or interesting this week: no specific project there, just try to err on the side of finding out and including that stuff.
     
  4. For recs, people really want to hear about older works I read during a given week. No problem! I'll kick that off immediately. And the next-most-popular ask was for a classification system, which doesn't take any code to spin up; I'll experiment with some category names, and add them to the FAQ once I've picked them.
     
  5. Again, thanks for participating everyone; this was super helpful! And a particular hat tip to isis for getting my current list of tag errors reported back into wrangler-space.
  
  
• As promised from last week's bug report, I did add the fix to recheck "uncommon" ships that appear in new works. I think it made a significant difference: the number of works I excluded because I couldn't find a valid ship in them fell by around 100 works. That's about 3% of all works tagged F/F this week, and after accounting for jumps in things excluded for other reasons, it looks like about 2% of F/F works are now being reported on which were wrongly excluded before. Not too shabby for one bugfix.
  
• Recs time: nothing from new shiny fandoms this week, because I've been focusing on those bugfixes and the poll. Instead how about:
  
  
  • A chewy Breq/Seivarden 1930s mob/noir AU
    
  • From the vaults: I went hunting for Emily of New Moon fic this week, and came up with the Tumblr AU (which is ridiculous but pretty damn funny), and an actual canon-divergence. Both Emily/Ilse; I want to stop and be surprised by how little there is in this canon given how much of a Yuletide staple Anne/Diana is.
  
  

( Full top-20 table and description of methodology after the jump )