Quote:
Originally Posted by Manko898
About c-betting in particular can you be more specific as to why they were proved wrong and in what ways? What was wrong about their assumptions, or the accuracy of their methods? How does the solver deviate?
I would like to know what factors were considered when Janda/Miller made their guess. Like, what were their assumptions, and how did they suppose they were making accurate assumptions?
How much of that was based on player experience in a realm of non-perfect and non-robotic players? Might there assumptions and methods still prove accurate in a realistic player pool?
Janda laid out some maths (including multistreet value:bluff ratios) in his first book which sounded very reasonable. But solvers have a more accurate grasp of how ranges play against each other at equilibrium.
If you c-bet 70%+ of your range on every street, you have lots of low equity trash in it. Not only does that mean that a villain can exploit you by slowplaying his monsters, in order to get value from your bluffs (which would do better by
not betting), but he can also exploit you by check-raising, as he knows it will be impossible for you to continue at a high frequency, as your betting range has so much air.
Put it this way, imagine villain knows your strat is to c-bet 100% of the time
a la Barry Greenstein. Villain has some trivial exploits against such a strat: Never fold a made hand; slowplay some monsters on the flop, in order to pick up more money from a turn barrel; check-raise a lot of air to win the pot immediately.
Solvers produce ranges that can't be exploited so easily. Their ranges are balanced, which means that a player that uses them is "unpredictable" to some extent. His betting and checking ranges are hard to beat. (Indeed, many hands will be mixed between betting and checking).
In some (nitty/fit or fold) games, it's still very profitable to bet at high frequencies, because villains don't "play back" at you. Weak micro players see the flop with weak ranges and then check-fold a lot, for example. A strong player, however, sees the flop with a solid range, check-raises appropriately, and doesn't overfold post-flop, so you
can't barrel him to death. Against a tough player, you'd just value-own yourself, or be over-bluffing, if you triple-barreled at high frequencies.
Last edited by ArtyMcFly; 04-28-2019 at 04:23 PM.