Learn
Strategy articles and research notes — how the game actually plays, measured on real data.
The SolverThe Exploitability ThresholdEvery solver output carries a number nobody reads: the exploitability it stopped at. We measured where the strategies actually settle — and found a cliff, not a slope.Pokalab ResearchAug 25, 2026
The SolverThe Pokalab Postflop SolverBoard-exact Discounted CFR, no card abstraction, validated line-for-line against PioSolver on an identical tree. The engine behind the library — and the receipts.Pokalab ResearchAug 25, 2026
The LabWon-When-VPIPOf the pots you choose to play, how many do you actually win? Winrate climbs in lockstep with this one number — and the best regs in the pool own it.Pokalab ResearchAug 23, 2026
The LabPlayer Types, MeasuredHow many kinds of poker player are there? None by nature, six useful ones by measurement — and the stat that finds the fish is not the stat that finds the winners.Pokalab ResearchAug 22, 2026
The LabPopulation Ranges, MeasuredA hand history records the cards only when the hand reaches showdown. Read that sample as a range and it says the big blind folds 0% of the time — here is how we recover the other 88%.Pokalab ResearchAug 14, 2026
The LabRiver Cards, MeasuredThe last street, and the biggest surprise of the three: the river card barely moves the bettor at all. It moves the caller.Pokalab ResearchAug 12, 2026
The LabTurn Cards, MeasuredSame method, next street. The flop's rule flips — on the turn, the card is king — and the biggest cell in every traditional map has no name.Pokalab ResearchAug 11, 2026
The LabBoard Textures, MeasuredWe let 1.2 billion real decisions draw the flop-texture map from scratch. The first thing that decides how a flop gets played is not the high card.Pokalab ResearchAug 10, 2026
ArticlesTest PostLilxamJan 1, 1970