Text-based adventure games are one of the most popular interactive genres that require players to navigate narrative environments and solve puzzles from text. This paper describes a new text-based adventure game’s design, development, and evaluation. Key contributions include using natural language processing (NLP)-enhanced command parsing to explore player interactions, state-of-the-art control for narration, and analysis of player interactions. The game was developed using Python and Twine and combines a rich narrative with an algorithmically branching story. We also explore the educational potential of games to enhance thinking and problem-solving. Detailed game research provides insight into player interactions, emotional interdependence, and competition.