Quote:
Originally Posted by trojanrabbit
Correct, you don't need to do the normalization step, but I included it so that you don't get any problems trying to raise a small rand to a power of thousands or millions.
Tysen
Actually, I found that I get different results with and without stack normalization. Without stack normalization, I had results closer to ICM. With normalization, I had errors:
Without normalization (100000 trials):
Player player6 Stack = 6000
Player player2 Stack = 2000
Player player8 Stack = 8000
Player player1 Stack = 1000
Player player7 Stack = 7000
Player player3 Stack = 3000
Player player9 Stack = 9000
Player player4 Stack = 4000
Player player5 Stack = 5000
Avg stack = 5000
Player player6 wins = 13476.5
Player player2 wins = 4845.3
Player player8 wins = 17086.6
Player player1 wins = 2489.1
Player player7 wins = 15343.7
Player player3 wins = 7168.5
Player player9 wins = 18802.6
Player player4 wins = 9349.5
Player player5 wins = 11438.2
Stack normalization (100000 trials):
Player player6 Stack = 6000
Player player2 Stack = 2000
Player player8 Stack = 8000
Player player1 Stack = 1000
Player player7 Stack = 7000
Player player3 Stack = 3000
Player player9 Stack = 9000
Player player4 Stack = 4000
Player player5 Stack = 5000
Avg stack = 5000
Player player6 wins = 12176.2
Player player2 wins = 5656.8
Player player8 wins = 15443.1
Player player1 wins = 10243.8
Player player7 wins = 13936
Player player3 wins = 6676.2
Player player9 wins = 17091.4
Player player4 wins = 8474.1
Player player5 wins = 10302.4