Even with the advances in different methods for authentication, passwords remain the mostcommon approach for authentication as well as for encryption of user data. Password guessingattacks have grown to be a vital part of computer forensics as well as penetration testing. In thispaper, we seek to provide a statistical analysis of password composition by analyzing whatcharacter sets that are most commonly used in over 1 billion leaked passwords in over 20different databases. Further, we use a survey to analyze if users that actively encrypt data differfrom the norm. The results of this study suggest that American lowercase letters and numbersare the, by far, most commonly used character sets and that users who actively encrypt data usekeyboard patterns and special characters more frequently than the average user.