Quote:
Originally Posted by jukofyork
Because the ICM function is massively underestimating the extra equity the big-stacked players can get by keeping the bubble alive and pwning the other players... Therefore the NE will have them calling much wider than they would if they were to realize the advantage of keeping the bubble alive.
The failing is with the (limitations of) ICM function here and not the algorithm used to find the NE (just in the same was as putting duff ranges into SNGWiz will give you duff solutions).
Juk
First of, I'd like to thank Juk. In my brief time here, I have found his posts to be most helpful and correct and they have done much to deepen my understanding (esp on the topic being discussed here).
What he says here is no exception. Its provoked these thoughts that are probably obvious but may hopefully prove helpful
1)In a game such as Fifty/50s in which the antes are so significant I think it is very important to think of stacks in terms of M rather than BB.
2)There is a range of M where push/fold ICM is most useful
3)I've found that range to be roughly between 8 and 2(in other words,fairly shallow but not tiny stack sizes)
4)Above M of about 8 push/fold is rarely correct(risk too much to gain to little)
5)Below about 2 the considerations Juk cites limits its effectiveness.
P.S. Juk, feel better man. I noticed in one of your earlier posts that you are under the weather.