This is an implementation of the mechanics of the board game "Deep Sea Adventure" by Jun & Goro Sasaki. The point it to computationally explore the space of possible strategies and evaluate them over thousands of games.
The board game consists of an "air-tank" with 25 units of air, 6 character tokens, 48 treasure tokens and two special 6-sided dice. The faces of the dice are numbered 1-3, with opposite faces having the same number. The treasure tokens are split up into five different types:
- 16 empty tokens
- 8 tokens labelled with 1 dot
- 8 tokens labelled with 2 dots
- 8 tokens labelled with 3 dots
- 8 tokens labelled with 4 dots
The probabilities of getting a given number between 2-6 when rolling two of the special dice are as follows:
- 2: 1/9
- 3: 2/9
- 4: 3/9
- 5: 2/9
- 6: 1/9
The point of the game is to dive into the ocean and search for treasure. The more treasure you pick up, the slower you move, and the faster the communal air runs out! The player with the most treasure after 3 rounds of diving is the winner. The actual rules of the game won't be described here, but if you're interested I'd suggest you either buy the game (it's fun!) or read the source code (less fun).
Once I've done an analysis of a few strategies, I'll write up some results.