Merry Crisis

[coalcube]

On the 7^th night of Hanukkah my vindictive author gave to me....

Madness Opens: Mon 26 Dec
Author Reveals: Sun Jan 1

Mini-Challenges:
Yuleporn| Crueltide | Queering The Tide | Transtide
Two for One | Seasons Treatings | IF | Wrapping Paper | Yumadrin 
 

yuletide  | yuletide_admin 
2022 Collection | 2022 Madness
2022 App (bonus + challenges) | Letter Post

Flat / Recent | Top-Level (Last) | FAQ | Search

Comments

[(Anonymous)]

For each person in Yuletide, what is the average weighted interactions with their fics per word per year? So for each person, you'd find all their yuletide collection fics, then sum up, say, hits + 2 * kudos + 4 * comments (with some finagling to avoid self-comments and maybe only count named kudos, not guest kudos), and divide by the total words and then again by the total number of years they did yuletide. And then rank everyone!

[(Anonymous)]

Why would you do per words *and* years instead of just per words? That seems like a huge penalty for people who've been doing yuletide awhile. Somebody who had ten comments on a 19000 word fic and a 1000 word treat in two different years shouldn't get 1/2 the score of someone with the same comments on one 19000 word fic and a treat in the same year, surely.

You might want to put some kind of statistical penalty on older stories that have had more time to accumulate score, but it wouldn't scale that neatly. I think you'd have to do each year with a normalization factor based on average score per word for that year's fic.

[(Anonymous)]

Yes, ideally you'd do some analysis to see how hits etc drop off on yuletide fics over time and then use that to normalise the older fics, but I guess if you could do that, you could also only count hits etc during a particular period around each yuletide and then average each year, and then average the averages for people who have done multiple years. Although now I am wondering if there aren't people who did really well one particular year and really badly other years.

[(Anonymous)]

The older fics are linked up in the first thread on this page complete with a snapshot of comments. No kudos on the old site though.

[(Anonymous)]

Well, once you had the data, it would be pretty easy to get average kudos per words per year over the whole archive - if 2009 has average .005 kudos per word and 2012 has average .004 kudos per word, you could make a kudos on a 2009 fic worth 4/5 of a 2012 one for everybody who had fic in those years. I don't think there's any way to get the data on when a particular kudos was left, at least unless you have real deep database access, but using average per year would probably do ok.

There are definitely people who do way better some years than others, though. It depends a lot on what fandoms you write and how you hit the zeitgeist. I got about 60 kudos before reveals last year and about 6 this year and I'm pretty sure this year's fic was better, but I knew this fandom was much less widely appealing.

[(Anonymous)]

If I look at direct numbers, last year was massively better than this year, but if I go by kudos:hits ratio, last year I got read roughly twenty-five times for each kudos and this year I got read roughly twice per kudos. Yet I got massively more comments:hits last year than this year. That's why I think there needs to be some kind of weighted combination of hits, kudos and comments as a sort of interaction index. I'd need to be able to try different things against multiple random samples of fics to judge what seems like a good formula. So much easier with an API. :(