How to build a poker bot (Part 1 Counterfactual Regret minimization)

Learn Texas Holdem Video Source & Info:

Using counterfactual regret minimization you can build your own poker bot that can beat the best pros

Code: https://github.com/IanSullivan/PokerCFR

Support : https://www.buymeacoffee.com/iansul

Part 2 https://www.youtube.com/watch?v=Qz3kSJv_9mE

Libratus paper https://www.cs.cmu.edu/~noamb/papers/17-IJCAI-Libratus.pdf

Solved heads up limit poker http://poker.srv.ualberta.ca/

Source: YouTube

Share this video:

4 thoughts on “How to build a poker bot (Part 1 Counterfactual Regret minimization)”

Is it possible that I can DM you

Updated Regret for paper in round 2 must be 1+1 =2

Is there a newer try on this called "pluberius"?
I am referring to https://www.cs.cmu.edu/~noamb/papers/19-Science-Superhuman.pdf ,
https://science.sciencemag.org/content/sci/suppl/2019/07/10/science.aay2400.DC1/aay2400-Brown-SM.pdf

Any ideas on actually implementing this? 😀

I tried using this algorithm for rock paper scissor and I noticed that the final sum of regrets seems to look pretty random. But the final average strategy and sum of strategies is good!! Can anyone explain why the sum of regrets isn't similar (or at least correlated) to the sum of strategies? Also great videos btw 🙂 !!!!

Comments are closed.