Bitcoin brain wallets, were way of turning nothing but a password into a keypair, at least until it was widely understood what a bad idea this was. The wake of data left behind includes a very interesting corpus of passwords to analyze and logs of attacker activity - after all blockchains never forget. This talk focuses on what we can learn from this. Do people select stronger passwords when more money is on the line? How quickly does bitcoin sent to weak brain wallets get drained? How many distinct thieves can be identified?