Quote:
Originally Posted by hyperknit
What’s weird about the whole “if a solver recommends mixing it’s because both option have the same EV” argument, is that the EV of which line u take with a hand is dependent on the other hands u have in that range. For example, a weak hand in a strong range has more EV than a weak hand in a weak range.
So the EV of flatting QTs vs 3betting QTs matters heavily on how we’re constructing, and looking at the decision in isolation does no justice.
Well, I was not suggesting that if a solver mixes that its doing so because both options have the same EV. I would expect thats rarely a reason for mixing. I just wouldnt expect the solver to mix very often preflop if the EVs of raise vs call were noticeably different. Is
that incorrect?
I'm guessing here, but where I could see this happening - mixing even when EVs are different - is on the river (maybe turn) when, with specific runouts, we end up with a range that doesnt have enough bluff candidates. In that case we may have to pick some 'bluffs' even though the EV of calling/checking with an individual hand in isolation is higher. Obviously this shouldnt be a regular occurrence for a solver range but I would think its possible. I believe I saw this in a video from Finding Equilibrium, he suggested maybe a solver defect but I kind of doubt it.
I'm kind of thinking out loud here. I dont know if this is accurate so feel free to pick it apart. I would really like to know the kinds of scenarios where the solver mixes even when the EVs do not run close. I think those could be interesting spots to study.