http://twoskeletons.livejournal.com/ ([identity profile] twoskeletons.livejournal.com) wrote in [personal profile] whynot 2009-07-22 06:16 am (UTC)

TAKE 2

HAHA. Okay. FIRST, what outcome variable will we choose to represent fandom popularity? We'll go with 'number of comments'. What input variables will we put into the model? The model itself would be structured a little something like this:

number of comments = (fic length * x1) + (main ship * x2) + (fic rating * x3) + standard error [which is sort of like standard deviation]

Then you rekerfooble the beta values (viz. FOR AN IMPROBABLE EXAMPLE, the beta value of 'fic length' is 20, because an increase of 1 word is correlated to an increase of 20 comments). Say we have a sample of 200 Merlin fics (n = 200). If we were to choose one Merlin fic from this sample with the following specifications:

x1 = 1000 words
x2 = Merlin/Arthur
x3 = R

--what would be the projected number of comments that this fic would have, based on analysis of the aforementioned 200 fics? You'd want to recode the input variables into numbers, so after some rekerfoobling, your projection for how many comments one particular fic (out of the 200) might get would SOMEHOW look something like this (improbability of the numbers aside):

number of comments = 20*1000 + 25*0 + 30*3 + standard error


AND MAY THE LORD GOD HAVE MERCY ON MY SOUL. D: (i'm sorry, professorrrrr)


ahahaha EXACTLY. It's the brothel where you can do filthy things that your fic/wife at home won't let you do to her. The O11/Narnia AU is still up, though discontinued indefinitely. We just... faded. [livejournal.com profile] themusingstream has it up in pretty much chronological order, courtesy of the awesomeness of [livejournal.com profile] lazaefair.

Post a comment in response:

This account has disabled anonymous posting.
If you don't have an account you can create one now.
HTML doesn't work in the subject.
More info about formatting