Quote:
Originally Posted by ohly
this. if solver advocates to do a certain action with f.e. 3 backdoor flushdraw combos 80% of the time and to do it 20% with the 9 non-backdoor flushdraw combos, just do it 100% with the backdoor flushdraw combos and 0% with the others. almost always you will end up with frequencies sufficiently close to gto.
This is very much how i think about the game.
You don't need to have perfect frequencies of perfect hands in each spots. In a lot of spots like mixing hands otf into bet and check ranges its much more important that you have certain type of hands approximately correct amount in each range.
Like having some% of TP in xb range so that when we defend enough vs probes, our calldown range doesnt become too weak which can be exploited by valuebetting thinner. So if you look a solver solution, it would for example check like 20% of all TP with some exceptions. But in practice you can just check back like few of the weakest tp combos almost always and end up with a solid strat.
Same with like wanting few FDs in ur checkback range, just figure out which fds make most sense to XB and xb those often and bet everything else.
Where rng becomes more prevelant are some river spots where you need to choose bluffing combos from a group of bluffs where all are equal value, but you cant bluff most of them let alone all of them.
One nice tip to get 50-50 mix I've used for years which is handy for pf. If you have 65s you 3bet, if you have 56s you call.
Quote:
Trying to get exact mixing frequencies down is futile. First of all remember the action is only mixed at equilibrium, in reality when you play against a human opponent one action is almost certainly better than another.
Agree mostly. You shouldn't use "mixed strategies" just for the sake of having mixed strats. But in a lot of situations esp where ranges are tight, you just need to do some mixing in order to have a solid strategy. In many spots esp wide range spots, you can play pure strat with each hand and still come up with solid strategy by just allocating different type of hands properly into each range.
Obviously this is not something you need to do if you have a clear exploit for every decision... but how often you can actually confidently say that one option is better than other?
Quote:
The important thing here is that you're using a mixed strategy with the right hands. It would take a large sample of hands for your opponent to figure out that you're betting 10% too much here or 10% too little there. Even so, the future is never set in stone, thus there's not much your opponent can do vs strategy (b).
THIS so much. In most spots villains aren't going to get good idea what our total strategy in each spot is gonna be. But what they can quite quickly realize is the frequencies of our actions in spots. It's way quicker to realize if someone checkbacks and folds to probe often, compared to what type of hands we use in that range.
Last edited by doctor877; 08-17-2018 at 03:33 AM.