SantoriniBot

Creating some AIs to play the board game, Santorini

Note: I do not claim any affiliation with Santorini. All credit goes to the developers.

Agents (from easiest to mediumest difficulty):

• RandomAgent (very easy)
  • See Random.py for implementation.
  • Just chooses random actions!
• FSAgent (easy)
  • See FS.py for implementation.
  • Implements Markov decision process (MDP) forward search.
  • Enumerates every possible state to a certain depth but doesn't consider your actions
  • Will win if left alone but is easy to beat if you're mean to it!
• MiniMaxAgent (medium)
  • See MiniMax.py for implementation.
  • Implements mini-max search with alpha-beta pruning.
  • Enumerates every possible state to a certain depth and assumes you take optimal actions every turn!
  • Plays a decent game but plays very conservatively and can only look a few turns ahead.

Launch the game:

• Play against RandomAgent: python3 Game.py --agent Random
• Play against FSAgent: python3 Game.py --agent FS
• Play against MiniMaxAgent: python3 Game.py --agent MiniMax

Player markers:

• Your player: B
• The opponent: O

Controls:

For either moving or building:

• Up (u)
• Down (d)
• Left (l)
• Right (r)
• Up-Left (ul)
• Up-Right (ur)
• Down-Left (dl)
• Down-Right (dr)