diff --git a/kaggle_environments/envs/chess/chess.py b/kaggle_environments/envs/chess/chess.py
index d62c1f2a..91fc109a 100644
--- a/kaggle_environments/envs/chess/chess.py
+++ b/kaggle_environments/envs/chess/chess.py
@@ -1,3 +1,4 @@
+import math
 import random
 import json
 from os import path
@@ -11,6 +12,29 @@
 ACTIVE = "ACTIVE"
 WHITE = "white"
 
+OPENINGS = [
+    "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1",
+    "rnbqkbnr/1p2pp1p/p2p2p1/8/2PNP3/8/PP3PPP/RNBQKB1R w KQkq - 0 6",
+    "r1b1kb1r/ppppq1pp/2n2n2/1B2p3/4N3/5N2/PPPPQPPP/R1B1K2R w KQkq - 3 7",
+    "rnbqkb1r/p2ppppp/5n2/2pP4/2p5/2N5/PP2PPPP/R1BQKBNR w KQkq - 0 5",
+    "rnbqk1nr/p1p1bppp/1p2p3/3pP3/3P4/2N5/PPP2PPP/R1BQKBNR w KQkq - 0 5",
+    "r2qk1nr/ppp2pp1/2np3p/2b1p3/2B1P1b1/2PP1N2/PP3PPP/RNBQ1RK1 w kq - 0 7",
+    "rn1qk1nr/pp2ppbp/3p2p1/2p5/2PP2b1/2N1PN2/PP3PPP/R1BQKB1R w KQkq c6 0 6",
+    "rnbqkbnr/1p2pp1p/p2p2p1/8/2PNP3/8/PP3PPP/RNBQKB1R w KQkq - 0 6",
+]
+
+
+MOVES = [
+    "",
+    "e2e4 c7c5 g1f3 d7d6 d2d4 c5d4 f3d4 a7a6 c2c4 g7g6",
+    "e2e4 e7e5 g1f3 b8c6 f1b5 f7f5 b1c3 f5e4 c3e4 g8f6 d1e2 d8e7",
+    "d2d4 g8f6 c2c4 c7c5 d4d5 b7b5 b1c3 b5c4",
+    "e2e4 e7e6 d2d4 d7d5 b1c3 f8e7 e4e5 b7b6",
+    "e2e4 e7e5 g1f3 b8c6 f1c4 f8c5 e1g1 d7d6 c2c3 c8g4 d2d3 h7h6",
+    "d2d4 g7g6 c2c4 f8g7 b1c3 d7d6 g1f3 c8g4 e2e3 c7c5",
+    "e2e4 c7c5 g1f3 d7d6 d2d4 c5d4 f3d4 a7a6 c2c4 g7g6"
+]
+
 
 def random_agent(obs):
     """
@@ -20,7 +44,7 @@ def random_agent(obs):
     """
     game = Game(obs.board)
     moves = list(game.get_moves())
-    return random.choice(moves)
+    return random.choice(moves) if moves else None
 
 
 def king_shuffle_agent(obs):
@@ -142,20 +166,25 @@ def square_str_to_int(square_str):
 
 seen_positions = defaultdict(int)
 game_one_complete = False
+game_start_position = math.floor(random.random() * len(OPENINGS))
 
 def interpreter(state, env):
     global seen_positions
     global game_one_complete
+    global game_start_position
     if env.done:
         game_one_complete = False
         seen_positions = defaultdict(int)
+        game_start_position = math.floor(random.random() * len(OPENINGS))
+        state[0].observation.board = OPENINGS[game_start_position]
+        state[1].observation.board = OPENINGS[game_start_position]
         return state
 
     if state[0].status == ACTIVE and state[1].status == ACTIVE:
         # set up new game
         state[0].observation.mark, state[1].observation.mark = state[1].observation.mark, state[0].observation.mark
-        state[0].observation.board = Game().get_fen()
-        state[1].observation.board = Game().get_fen()
+        state[0].observation.board = OPENINGS[game_start_position]
+        state[1].observation.board = OPENINGS[game_start_position]
         state[0].status = ACTIVE if state[0].observation.mark == WHITE else INACTIVE
         state[0].status = ACTIVE if state[0].observation.mark == WHITE else INACTIVE
         return state
@@ -203,7 +232,7 @@ def interpreter(state, env):
         inactive.status = terminal_state
         game_one_complete = True
     elif seen_positions[board_str] >= 3 or game.status == Game.STALEMATE:
-        active.status = terminal_state 
+        active.status = terminal_state
         inactive.status = terminal_state
         game_one_complete = True
     elif game.status == Game.CHECKMATE: