Is that really the best of methods? Shouldn't you be counting out the nonactive accounts or the accounts that don't have any posts, and all the accounts that are multis/old versions of active people's accounts? I mean, that number is kind of misleading, when you take into account how many people have old versions of their accounts which they have lost the password or email for, and how some people used to be fond of creating fake new accounts for pranks, and how many of the people were just random new people who left after making a thread or two.