This study examined behavioral differences between male and female players in two large multiplayer online games (MMOs), EVE Online and CR3, using machine learning techniques. By analyzing server log data from over 1,700 players total across both games, the researchers found they could predict player gender with over 90% accuracy for EVE Online and over 80% accuracy for CR3. Social behaviors like number of corporations joined in EVE Online and message length in CR3 were significant predictors of gender. The findings suggest there are differences in social behaviors between male and female players in both Western and Chinese cultural contexts of the games studied.