Generating Gloomhaven dungeons through evolutionary game design

In this thesis, evolutionary algorithms were used to generate unique and enjoyable dungeons of the board game Gloomhaven. Using Python, existing Gloomhaven dungeons were stored as objects, and went through an evolutionary cycle of selection, crossover and mutation to generate new dungeons. These dungeons were checked for playability and judged on several factors based on previous research on game enjoyability. The highest-scoring dungeons that came out of this process were then tested in a user test, returning mainly positive results.