GTO Baseline and Exploit
Learn the robust baseline, then profit from deviations
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 98%
Boeree describes game-theory-optimal play as a robust strategic baseline. Solvers model scenarios and prescribe actions at particular frequencies, giving a player a default when they lack reliable information about an opponent. Studying those solutions builds a benchmark for high-quality play rather than a guarantee of profit against another perfect player. The practical edge appears when real opponents depart from the benchmark. Because every human player makes mistakes, the trained player can identify the direction of a deviation and adjust to exploit it. The sequence matters: without knowing the baseline, a player cannot reliably distinguish an exploitable mistake from sound variation. When evidence about the opponent is weak, the player falls back to the robust default.
Origin
Boeree describes using poker solvers such as PioSolver to study optimal play and then identify opponents' mistakes.
Core principles
- 01Establish a high-quality baseline before improvising
- 02Use the baseline when opponent-specific information is absent
- 03Measure mistakes as deviations from the benchmark
- 04Exploit predictable deviations rather than abandoning sound play
How to run it
- 1
Model a recurring scenario
Define the positions, available actions, and constraints closely enough to study a solution.
Pro tip Start with scenarios that occur frequently rather than exotic edge cases.
Watch out A poorly specified scenario produces a misleading baseline.
- 2
Learn the baseline
Use a solver or rigorous model to identify robust actions and their frequencies.
Pro tip Treat frequencies as part of the answer, not as optional detail.
Watch out Memorizing an action without its context does not transfer safely.
- 3
Practice accurate execution
Rehearse enough representative cases to approximate the baseline under real conditions.
Pro tip Compare decisions with model output after practice.
Watch out Knowledge that disappears under pressure is not yet operational.
- 4
Detect the deviation
Observe where an opponent repeatedly acts differently from the benchmark.
Pro tip Demand a pattern before making a large adjustment.
Watch out Random variation can look like a strategic flaw in a small sample.
- 5
Exploit selectively
Adjust in the direction that benefits from the identified error, while retaining the baseline elsewhere.
Pro tip Revert when the opponent adapts or the evidence weakens.
Watch out Over-adjusting can make your own play exploitable.
In the wild
Boeree says that studying and emulating solver charts over an intensive training period would give Ferriss a solid baseline. He could then look for mistakes because even professional players do not execute perfect GTO play.
→ The baseline supports competent play while making opponent-specific errors visible.
Common mistakes
Trying to exploit without a benchmark
Without a model of strong play, apparent mistakes may simply be misunderstood strategy.
Assuming the baseline itself creates an edge
Two perfect GTO players break even over the long run; profit comes from deviations.
Is it for you?
Best for
Adversarial games with modelable scenarios, repeat interactions, and observable strategic deviations.
Not ideal for
Novel environments where the model assumptions are wrong or meaningful behavior cannot be observed repeatedly.
From the transcript
“Because once you know what the optimal solutions are, then now you can you're sort of equipped with this like really solid baseline of what…”
“But you can't really know the way that they're screwing up until you know what the GTO is in the first place.”
From the episode
#611: Liv Boeree, Poker and Life — Core Strategies, Turning $500 into $1.7M, Cage Dancing, Game Theory, and Metaphysical Curiosities
Liv Boeree