I've been playing SnG using the ICM theory to gain some real rather than simulated experience of the mathematics involved in it. I noticed some inconsistency in certain circumstances which encourage me to be careful about depending on it.
In hands I played using the ICM I was frequently granting more value or equity to the short stack than they actually held. The ICM calculations don't seem to consider the stack size of the next player who will post the big blind. In my standard play if the blinds are large and the small stack has to post the next big blind I would factor into my move that the short stack could be committed and have to call any all-in on the next hand with any two cards. By strictly applying the ICM I lost some games by not considering that.
While playing as middle stack using the ICM means of calculation I gave less value or equity to some of my hands than they actually held. Strictly adhering to the ICM I played and lost some hands in that situation, whereas in my standard play I would have folded and just outlasted the short stack.
When the blinds are large individual equity or value changes significantly by the chips posted blind, even if a hand is not played the small stack loses equity while the medium and large stacks gain. What is the allowance or consideration for that in the ICM method? If that is not considered correctly is the ICM method not more effective towards the outcome of the current hand rather than that of the overall tournament?
