The high figures seem to result from the fact that the *AA argue that the file was _potentially_ shared with _millions_ of people around the internet. I wonder why they are not required to prove every single instance of downloading that they claiming damages from...