[surfdoc]

[ QUOTE ]
I'm a mid-stakes limit n00b here, but with regards to 3-bet/folding vs. 3-bet/calling:

Given the size of the pot and the assumption that given the description of Villain, Villain's hand is consistent with JJ,KK+,AK, AcQc,AcTc, it looks to me like the cost of 3-bet/calling vs. 3-bet/folding is only 0.1BB, assuming Villain NEVER 4bets worse on the river and that I did the math right.

Since the EV is so close between the two choices in the worst-case scenario (only 4bets the nuts), does it really matter which choice we make on the river, practically speaking?

[/ QUOTE ]

What you are missing here is the application of bayesian thinking. We have new and additional information. We now have the opportunity to make our decision based on that new information. There comes a time when hand ranges get narrowed down enough that you can no longer assume equal probablities to each combination. When we get checkraised on this river and then 3 bet this guys set is going to be horrified of the backdoor flush as well as straight on the board. He didn't 3 bet the turn with these same holdings. Why is he suddenly 4 betting the river? Because something has changed. The card on the river is the only thing that changed and the simplest explanation is that it helped him. The only way it could help him is if he made a flush.