I wonder how the front page rankings would differ if Hacker News (secretly?) only counted upvotes if the user actually clicked the article link (or at least loaded the comments thread and read for X minutes) before voting.