CFR (Counterfactual Regret Minimization)
CFR（反事实遗憾最小化）
Counterfactual Regret Minimization is an algorithm that iteratively solves imperfect-information games like poker. It works by simulating hands, calculating 'regret' for not taking the best action at each decision point, and adjusting future strategies to minimize total regret. Over millions of iterations, CFR converges to a Nash Equilibrium.
Example
TexasSolver uses CFR to compute GTO strategies, typically reaching 0.5% accuracy within seconds.
Related Terms
More Advanced Terms
Board Texture
The composition and characteristics of the community cards on the board.
Polarized Range
A range consisting of very strong hands and bluffs, with few medium-strength holdings.
Range Advantage
When one player's range has more strong hands than their opponent's range on a given board.
Solver
Software that calculates game theory optimal solutions for specific poker scenarios.
Blocker
A card in your hand that reduces the likelihood of an opponent holding a specific hand.
MDF (Minimum Defense Frequency)
The minimum percentage of your range you must defend to prevent opponent from profiting with any two cards.