Learn Texas Holdem Video Source & Info:
Solving the simple Kuhn poker with the counterfactual regret minimization algorithm learn in part 1
Part 1 https://www.youtube.com/watch?v=Wa-fRIBGZZI
Vanilla CFR https://justinsermeno.com/posts/cfr/
Code: https://github.com/IanSullivan/PokerCFR
Source: YouTube
It is really a nice video. Could you please illustrate MCCFR more, if possible? Thanks
At 4:15 why is the hypothetical rewards 2 instead of 1?
Nice! Do you have a github with the code? It's easier to read it that way
Why do you calculate node.regret_sum on line 58 and 61 by only multiplying with the opponents probability on line 58, for example: node.regret_sum += pr_2 * regrets ? Why not the total probability: node.regret_sum += pr_2 * pr_1 * regrets ?
I hope this video will be banned soon