Quote:
Originally Posted by baudib1
mmm it is a serious problem if he can bet ATC profitably, we are surely folding too often.
This surprises most people the first time they hear it but its not true at all.
What you're talking about is the naive 1-alpha defending range discussed in mathematics of poker. It doesn't actually hold up at all in real poker.
The driving factor is that it is using the indifference principle to say that at equilibrium we want our opponent to be indifferent between bluffing and not. So indeed on the river if we are merged and he is polar and he bets pot, we need to defend 50% to stop him from bluffing 100% (a few other assumptions must hold, but mostly true). However, on the flop he doesn't have zero equity when he checks back, so if we want to make him indifferent we need to give him a better payout on his cbets. Consider this, if we defend 1-alpha and he doesn't autoprofit from cbets, it would be correct for him to never ever cbet bluff and only value bet us, since every weak hand he can check back for the off chance he turns a pair! That can't be right. We are comparing cbetting to checking back, so we only need to defend enough to incentivize him to check back, not to make his cbets -ev.
(The offsetting factor is that when he cbets he stops us from realizing additional equity with weak hands on the turn, but it turns out that factor is not enough to make him want to cbet significantly more. The main point is, the naive 1-alpha approach is not a sound poker concept)