def select_move(self, game_state: GameState):
winning_moves = []
draw_moves = []
losing_moves = []
# loop through all legal moves:
for move in game_state.legal_moves():
# state of the game after this move is applied
next_state = game_state.apply_move(move)
# determine opponent's best outcome given that state
opponent_best_outcome = best_result(next_state)
our_best_outcome = reverse_game_result(opponent_best_outcome)
if our_best_outcome == GameResult.win:
winning_moves.append(move)
elif our_best_outcome == GameResult.draw:
draw_moves.append(move)
else:
losing_moves.append(move)
# try to win, with drawing the next best choice
if winning_moves or draw_moves:
return random.choice(winning_moves or draw_moves)
# lost the game
return random.choice(draw_moves)
def best_result(game_state: GameState):
if game_state.is_over():
if game_state.winner() == game_state.next_player:
return GameResult.win
elif game_state.winner() is None:
return GameResult.draw
else:
return GameResult.loss
best_result_so_far = GameResult.loss
for candidate_move in game_state.legal_moves():
next_state = game_state.apply_move(candidate_move)
opponent_best_result = best_result(next_state)
our_result = reverse_game_result(opponent_best_result)
best_result_so_far = max(our_result, best_result_so_far)
return best_result_so_far
