What is GTO in poker?

GTO stands for game-theory optimal. A GTO strategy is a Nash equilibrium of the game: a way to play every hand in every spot such that an opponent cannot gain anything by changing their own strategy. Poker solvers compute approximations of it for specific spots, and players study the results to learn what balanced play looks like.

An equilibrium, not a prediction

GTO does not try to guess what your opponent holds or how they like to play. It answers a different question: how to play so that no counter-strategy can beat you by more than the game itself allows. In a two-player spot — a heads-up pot on the flop, say — that has a precise meaning. Play the equilibrium, and the best your opponent can do is play the equilibrium too; any deviation on their side either costs them money or changes nothing.

That guarantee is strict only for two players. With three or more players in the pot an equilibrium still exists, but two opponents could, in principle, deviate together in a way that hurts you. In practice equilibrium play remains the standard baseline for multiway spots and preflop charts as well.

Why a solver plays the same hand in different ways

A GTO strategy is often mixed: the same hand bets 60% of the time and checks 40%. That is not indecision. A hand that is always played the same way can be read from the action; mixing keeps the ranges balanced, so the opponent's options become equally good — the opponent is indifferent.

The classic river example shows the numbers. You bet B into a pot P, and your betting range holds strong hands and pure bluffs. Your opponent pays B to win P + B, so a call needs B / (P + 2B) equity. For a pot-size bet that is 1/3. If exactly a third of your betting range is bluffs, calling and folding earn your opponent the same, and they have nothing to exploit.

The defender faces the mirror-image condition. A bluff risks B to win P, so it profits whenever the opponent folds more than B / (P + B) of the time. To stop any two cards from bluffing at a profit, the defender must continue with at least P / (P + B) of their range — the minimum defence frequency (MDF). Against a pot-size bet that is 50%; against a half-pot bet, 67%.

Exploitability: how close is “close to GTO”?

Real poker games are far too large to solve exactly, so solvers iterate towards the equilibrium — many with variants of counterfactual regret minimisation (CFR). Progress is measured by exploitability: how much a perfect counter-strategy could still win against the current solution, usually as a percentage of the pot. Zero means an exact equilibrium.

In the GTOLens Hold'em solver you choose the target: the solve stops at 1%, 0.5% or 0.3% of the pot. At 0.5% an all-knowing opponent could take back only half a percent of the pot — far less than the mistakes people make at the table.

GTO versus exploitative play

GTO is defensive by design: it loses to nobody, but it does not go out of its way to punish mistakes either. Exploitative play deviates on purpose — bluffing more against players who fold too often, value-betting thinner against players who call too much — and earns more against those players while becoming exploitable itself.

The two approaches work together. You can only tell that a player over-folds if you know how often they should fold, and the equilibrium is that reference point. Most strong players study GTO to build a sound default and deviate from it where they have evidence.

Is GTO poker the same as solved poker?

Not quite. Heads-up limit Hold'em was essentially solved in 2015 by the Cepheus project at the University of Alberta. No-limit games are vastly larger: solvers work spot by spot with a limited set of bet sizes, and the result is an equilibrium of that simplified game. That is why the bet sizes you allow — the tree — matter so much, and why GTOLens always shows the tree you solved.

Common misconceptions about GTO

• “GTO wins the most.” No: against weak players, exploitative play earns more. The equilibrium guarantees you are not beaten (in a two-player spot), not the maximum win against a particular opponent.
• “GTO is random play.” Mixing follows exact frequencies. A single decision can look random, but the range as a whole is built precisely.
• “You have to memorise solver output.” Memorising millions of frequencies is impossible and unnecessary. It pays more to learn the principles: why some hands bet and others check, how bet size changes the share of bluffs, and how ranges interact with the board.

How to study GTO for free

• Hold'em: solve any heads-up postflop spot in your browser with the free poker solver — set the ranges, board and bet sizes, then see the strategy, EV and equity for every hand.
• Omaha: browse PLO preflop charts from solved spots and open the full trees in the free viewer.
• Hand values: see which Omaha hands earn the most from early position in the PLO starting hand rankings.