Quote:
Originally Posted by aner0
Solver maximizes EV for a single hand. Against the strategy it's paired against.
You both gave different answers to this. I don’t know what to believe anymore! It does make a big difference… I presume the solver is capable of both? I’m not sure of the exact calculations the solver makes.
Quote:
Originally Posted by aner0
Your opponent "personality" and "characteristics" and his "thoughts about what your range is" will ultimately manifest in a certain strategy that you can technically input into the solver, it's just very inconvenient if you want to apply it to every runout. Him changing his strategy from one hand to the next will manifest in multiple different static strategies that you can also technically input into the solver.
Nodelocks are definitely not the end all be all, they are just a simplified model of how actual poker works and they have their flaws. Use them to learn about different concepts, but ultimately use your knowledge and intuition to figure out what the best exploit is and how far it's reasonable to take it.
I like these paragraphs! You should get a job editing poker books, mine does need
another once over.
Quote:
Originally Posted by aner0
In the solver, your perceived range is just your actual range. If you wanted to know how to play a certain hand against an opponent that is perceiving a certain range, we should input that range for us.
This would show you the range a solver trained opponent will likely adjust to your strategy, which is half way there. For it to give you the L3 strategy you’d have to node lock that strategy for the opponent and then run it again to how to counter it. It’d be real easy for the solver companies to code this in. It would add shitloads to each players node locked winrates too. It’s a shame
for them that they don’t listen to me!
I think I could code the entire human thought process, including feelings and emotions too. To enhance a solver exploitatively you basically need to replicate this thought process. Each development would improve the winrate the solver can give against certain opponents. This isn’t the best application for this knowledge lol, if you can replicate a human thought process for a poker player you can do it in general. I’m working on a paper that teaches this, but with no funding nor any support it’s slow going. It’s very tough to write lol, it changes our understanding of the universe and might even prove the existence of God!