Anonymising data is surprisingly difficult. I'm wondering if there is the "falsehoods programmers believe about" about this, as there are for topics such as names, addresses, time, etc.?