ais against each other

main.py

@@ -96,9 +96,8 @@ class Board:
         print("#############")
 
 START_WEIGHT = 20
-WIN_REWARD = 3
-DRAW_REWARD = 0
-PUNUSHMENT = 2
+WIN_REWARD = 1
+PUNUSHMENT = 1
 
 class AiPlayer:
     def __init__(self, ch: Ch) -> None:
@@ -113,10 +112,6 @@ class AiPlayer:
         for choice in self.current_choices:
             self.choices[choice] += WIN_REWARD
 
-    def reward_draw(self):
-        for choice in self.current_choices:
-            self.choices[choice] += DRAW_REWARD
-
     def punish_loss(self):
         for choice in self.current_choices:
             self.choices[choice] -= PUNUSHMENT
@@ -141,7 +136,7 @@ class AiPlayer:
         for idx, choice in possible_choices:
             choice = self.interned_choice(choice)
             key = choice.key()
-            if not candiate_weigth or self.choices[key] > candiate_weigth:
+            if not candiate_weigth or self.choices[key] > candiate_weigth + 4:
                 candiate_weigth = self.choices[key]
                 candidate_idcs = [(idx, key)]
             elif self.choices[key] == candiate_weigth:
@@ -163,46 +158,67 @@ class AiPlayer:
             for key in vs:
                 self.choices[int(key)] = vs[key]
 
-class Game:
-    def __init__(self) -> None:
-        self.current_plays = []
-        self.weights = {}
-
-
-
 def main():
-    p0 = AiPlayer(Ch.Cross)
+    p1 = AiPlayer(Ch.Cross)
+    p2 = AiPlayer(Ch.Circle)
 
     games = 0
+    p1_wins = 0
+    p2_wins = 0
+    draws = 0
+
+    print("Training P1 against P2...")
+
+    for _ in range(1000):
+        board = Board()
+        p1.clear_choices()
+        p2.clear_choices()
+        while True:
+            p1.make_play(board)
+            if board.player_has_won(Ch.Cross):
+                p1_wins += 1
+                p1.reward_win()
+                p2.punish_loss()
+                break
+            if board.is_draw():
+                draws += 1
+                break
+            p2.make_play(board)
+            if board.player_has_won(Ch.Circle):
+                p2_wins += 1
+                p2.reward_win()
+                p1.punish_loss()
+                break
+            if board.is_draw():
+                break
+        games += 1
+
+    print(f"Games {games}  Score: P1: {p1_wins}, P2: {p2_wins}, draws: {draws}")
 
     while True:
-        print(f"\n\nGame #{games}")
+        print("\nNew game")
         board = Board()
-        p0.clear_choices()
+        p1.clear_choices()
         while True:
             print("AI's turn")
-            p0.make_play(board)
+            p1.make_play(board)
             board.print()
             if board.player_has_won(Ch.Cross):
                 print("AI won!")
-                print("Rewarding AI...")
-                p0.reward_win()
                 break
             print("Your turn (0..8)")
             if board.is_draw():
                 print("Draw!")
-                print("Rewarding AI...")
-                p0.reward_draw()
                 break
             possible_choices = board.possible_plays()
             should_restart = False
             while True:
                 text = input("> ")
                 if text == ".save":
-                    p0.save_to_file()
+                    p1.save_to_file()
                     continue
                 elif text == ".load":
-                    p0.load_from_file()
+                    p1.load_from_file()
                     choice = 0
                     should_restart = True
                     break
@@ -223,11 +239,10 @@ def main():
             board.print()
             if board.player_has_won(Ch.Circle):
                 print("Player won!")
-                print("Punishing AI...")
-                p0.punish_loss()
                 break
         games += 1
 
+
 try:
     main()
 except KeyboardInterrupt: