When do people read fimfiction? · 8:20pm Jul 29th, 2012
A graph showing total views, in 3-hour bins, for all stories under one-half-week old, for the past week.
I generated this by reading the last 5 pages of the story index (lists view counts for the last 1000 stories) every half-hour for a week:
The red line is views. The blue line is views divided by new stories released. It goes up to meet the red line during times when no new stories were published. The hash marks on the x-axis are at midnight. Sorry I can't make the font any bigger - DBD::Chart::Plot doesn't let you. The line at the bottom saying "541 stories" is, I think, the number of new stories published during the week.
People read the most noon to midnight Eastern time, except for Friday night, when they're out having fun. (I think the big spike on the morning of the 26th is due to my connection to fimfiction going bad and a bunch of views from the night before being missed.)
I don't think this means people preferentially read noon to midnight. I think it means people read fimfiction whenever they're awake, but most of them are in America, and noon to midnight is the time when everybody from coast to coast is awake.
I never did ask, how do you get this data?
257579
Magic. Magical data theft.
No, just kidding. I hired some Ukranians to hack into fimfiction. They're the best.
I believe you completely.
257621
What? You don't believe that Ukrainians are the best hackers?
At the top of the post, I wrote, "I generated this by reading the last 5 pages of the story index (lists view counts for the last 1000 stories) every half-hour for a week". I wrote a Perl script that sends a "browse stories" request to fimfiction every half hour, and reads the view counts out of the page returned. (I'm shutting it down now, since I have no need to keep gathering this info.)
257628
I trust my Slavic brethren to be nothing short of awesome.
I've checked the source of the browse page and it's actually quite clean and eminently suitable to being parsed with perl. I thought you had some sort of access to the database which is why I was asking. Well, access aside from 'read the HTML it spits out and extract meaning out of it using regular expressions'.
257680
Eh, it's not valid XML. None of the pages on the site are, due to a few unclosed tags. So it has to be pre-processed a lot before an XML parser will handle it. And you have to set a cookie to get listings of 'mature' stories.
As for me, I work the graveyard shift. That's when I do most my reading.