How It Works

The bot uses screen capture methods and computer vision to find in-game feature such as enemies defeated. Then it uses this metric as the Q-value for the Q-learning algorithm. It is intended to be used for any map and for any class, given a set of defined actions that the user provides the bot.