[Matt Flynn]

[ QUOTE ]

One small (but important, IMO) amendment which is they could still not be getting odds to call, and you'd still rather have them fold than call due to their equity, and your potential reverse implied odds. At least, I think what that AA example in the book is driving at.

I'm just thinking this out, so I actually may have it wrong. Whenever you're ahead but not a lock, your opponent by definition has some equity. So, you are hoping they'll call when they make a bigger mistake by calling than the equity you give up, and you want them to fold when they give up more pot equity than it costs them to call your bet (in terms of their new pot equity, I guess).



[/ QUOTE ]


do whatever yields you the most "implied equity." that is, look at all streets and figure out what move now will, on average, maximize your expectation for the hand.


[ QUOTE ]
I'm not sure exactly what that means from a theory standpoint, but it seems like it should factor into the upcoming REM discussion.

In Theory of Poker, Sklansky talked about optimal bluffing frequency being such that no matter what, your opponent had the same negative expectation. I wonder if there's something related in terms of bet sizing where the optimal (HU) theoretical bet size would be equal to an amount that caused your opponent to have the same cost whether they called or folded.

[/ QUOTE ]


a caution here: Sklansky is talking about nonexploitable play. some loosely call it "game theoretic" play. the general idea is that even if your opponent KNOWS what your strategy is, he cannot exploit it much. you minimize the maximum he can extract from you (called "minimax"). Bill Chen probably wrote a ton about this in Math of Poker. Some day I'm going to retire and read all these poker books, his first.

nonexploitable play is often NOT optimum. that's because your opponents have game flaws that you can exploit to make more money. for example, if your opponent is suspicious of big bets, you should make huge bets when you have the goods. or suppose optimum "game theory" bluffing amount is a third of the pot. if your opponent will call that half the time but call a 2/3 pot bet only 10% of the time, betting 1/3 the pot costs you money even though it is "game theoretically" accurate.