No one has suggested it, but I’ve certainly thought about it! I’d love to add it but it will be a little while I think… have to get audiovisual (movies / tv shows / anime) going first and then I want to step back and do a bit more work to solidify everything before moving onto even more content types.
As you might imagine, this sort of approach could potentially be applied to a lot of things, but you have to move somewhat slowly and do things well. I hear you though - a lot of people (especially in the Japanese community) use video games and it’d be great to have
Well, you know, you could always add more things to Natively
I would love to bring back intra series comparisons (i used to allow sequential volume comparisons). I do hear that certain series change difficulty quite a bit especially ones with somewhat separate sub series (like 本好き). Would love to map those things out, but decided to focus on inter comparisons first.
And @seanblue (and others) you’ll be happy to hear I think at the very least I’ve figured out a way to adjust the K value to allow lots of comparisons in a way i’m comfortable - simply decrease the K value dependent on how many times that user has graded the book before. It would mean the 30th time you’ve compared 本好き may not update 本好き a lot but it could, at the very least, change the book you’re comparing it to. That limits a user’s power on any one book.
Granted, this way does still have warts and the pairwise approach would be better if it works out, but it gives me confidence to at least explore fixing the algorithm sooner so that at the very least Natively could allow people to do more comparisons if they want to : - )
Edit: Now upon rereading this, I’m a bit wary of reopening the algorithm discussion but I just wanted to mention I am thinking about it and am more confident I can figure out something sooner.