Going back pages does not give you an accurate representation of the daily front page. Also, when I go a few pages back there is suddenly a huge swathe of non-U

Going back pages does not give you an accurate representation of the daily front page. Also, when I go a few pages back there is suddenly a huge swathe of non-URL posts for seemingly no reason. I lifted a dataset[1] from http://hackerslide.com http://hackerslide.com which has hourly archived front pages. With that and some dumb regular expressions[2] I categorized stories and how much time that story stayed on the front page. That is, two bitcoin posts on four different front-pages in a day would be 8 'story-hours'. https://spreadsheets.google.com/spreadsheet/ccc?key=t2W3xQ5pgv5bXvtdNXki0Cg&authkey=CKL_yI0O#gid=1 https://spreadsheets.google.com/spreadsheet/ccc?key=t2W3xQ5p... I'm sure this way of measuring has problems such as different times having different amount of activity on HN. To me it seems like there are constantly 1+ bitcoin posts on the front-page everyday. Most of the time the posts are about some trivial event (bitcoin hits 8.5 on mtgox, EFF starts accepting bitcoins, EFF stops accepting bitcoins) or a blog post that discusses bitcoin. In all of these circumstances the discussion seems to follow the same trends and go over the same ground. If there is ever a truly novel insight it would be hard to find in the mess. [1] https://dl.dropbox.com/s/182rbsb8kb83y4w/hn-frontpages.7z?dl=1 https://dl.dropbox.com/s/182rbsb8kb83y4w/hn-frontpages.7z?dl... [2] https://www.dropbox.com/s/d6ivp8lrj8esflr/count.py https://www.dropbox.com/s/d6ivp8lrj8esflr/count.py