In poker, GTO stands for Game Theory Optimal, a mathematically unexploitable strategy.
Game theory
Game theory is the mathematical study of strategic behavior. When we say a strategy is GTO, we mean that it is a Nash Equilibrium strategy. This means neither player can increase their expected value by adjusting their own strategy.
In a zero sum game, a GTO strategy maximizes the expected value floor, which is the amount of expected value that can be guaranteed against a perfectly countering opponent.
Blindless poker
Consider how No-Limit Hold'em would play out if players no longer had to post blinds. What strategy would you play?
The GTO strategy for blindless poker is either:
- Always fold
- Jam all-in preflop with AA only
- Any combination of the above two strategies
Players can fold hands at zero cost over and over again until they get AA. Even if you know someone is only playing AA, there is nothing you can do about it. That is a GTO strategy.
Exploits in blindless poker
What if you are playing heads-up against a player that goes all-in preflop regardless of their hand?
Folding everything except AA guarantees profit in the long term. However, just because a strategy is unexploitable does not mean it is the most profitable strategy against a specific opponent.
Assuming you are 100bb deep and your opponent always jams all-in preflop:
- If you fold everything except AA, you can expect a win rate of 32bb/100 hands
- If you jam with every hand that has over 50% equity against two random cards, you can expect a win rate of 802bb/100 hands
This analogy ties directly to poker: GTO is an unexploitable baseline, but player-specific exploits can earn us even greater profit. While this is true, it also exposes us to getting exploited ourselves.
If we are jamming all hands that have over 50% equity against two random cards, we are now exploitable if our opponent decides to tighten up their range. If we and our opponent keep adjusting our strategies to counter each other, we will eventually reach equilibrium again.
GTO in poker
Poker is far more complex than Blindless Poker. In No-Limit Hold'em, there are more possible game states than there are atoms in the universe. Instead of solving the full game, we choose what bet sizes are allowed and solve poker as though only those bet sizes were allowed. Equilibrium solutions to this limited-bet-size poker are the GTO strategies we use to study.
Even these approximated strategies are far too complex for humans to implement. If you sat a computer at a high-stakes poker table to play a GTO strategy, it would practically print money.
Since we cannot play a perfect GTO strategy, we instead study GTO to understand what unexploitable play looks like, and deviate according to the mistakes of our opponents.
How can GTO beat us when we know its strategy?
If you are playing Rock Paper Scissors against someone who tells you they are playing each option exactly one third of the time, can you beat them in the long term?
No. The highest EV you can achieve is zero.
GTO poker strategies bet with both bluffs and value, balanced precisely so that even if you know their strategy, there is nothing you can do to exploit it. This principle is baked into every decision. Even if you know the exact strategy, there is nothing you can do to beat it.
How to use GTO to get better at poker
GTO is a tool to help you understand the game better. Any high-level poker player will tell you that they regularly study solvers to continuously improve their game.
By studying GTO solutions you will notice patterns and build an intuition for what unexploitable play looks like. There is no shortcut to the process of learning and practicing, but having access to cleanly converged solutions and the best tooling available will get you there faster.
To get started, check out current options for studying GTO and see if any fit your needs.
Further reading
- How Do Poker Solvers Work?
- Solution Quality: dEV and EV Loss Explained
- Why Bet Sizes Matter in Solver Game Trees