Poker-AI.org

I'm looking to trade the preflops I've created for 4, 6, 9, 12, 20, and/or 26 BB games, crunched on a machine with 144gb of RAM. If you have a descent ROI at HUNLHE SNGs and are willing to trade your HHs (including hole cards and post-flop play), please PM me.

An example of one of the preflops (6BB) can be found here:
viewtopic.php?f=24&t=2581

Interesting. How you calculated postflop spots when button limps and BB checks?

The same way every other spot in the abstraction was calculated, with CFRM.

The preflops include actions for fold, call/check, min, 2, 3, 4BB, and all-in bets, up to four betting rounds. Post-flop play was also calculated but not included (as it would be a terabyte+ of data for all games). The post-flop betting abstraction is mentioned in the link I included above.

So as I understand it, all preflop calls are not garaunteed to be optimal (profitable) as they highly depend on postflop play which you don't provide?

I'm asking because I wonder how it will be usefull for the bot which is already doing ok.

Nothing is guaranteed.

The postflop was solved along with the preflop, E-Nash optimal to it's own abstraction. Some people thought knowing what a preflop EQ looks like might be useful, especially one crunched with a very large subset of memory. I thought others might find it useful too.

Any others?