Pokercat Guide
Exploit
The Exploit engine models a specific opponent type and searches for the maximally profitable response — not the unexploitable one. Where GTO+ nudges the equilibrium, Exploit is free to leave it entirely.
What makes it different from GTO+
| GTO+ | Exploit | |
|---|---|---|
| Starting point | Equilibrium strategy | Opponent model |
| Deviation | Small, anchored shift | As far as the model justifies |
| Risk | Low — stays near unexploitable | Higher — a wrong read costs more |
| Best against | Unknown or decent opponents | Clearly-typed opponents with stable leaks |
GTO+ asks “what small adjustment beats this player type without giving much up?” Exploit asks “what is the single most profitable line against this player type, full stop?” Against a calling station that means merging value bets thinner and bluffing less; against a nit it means over-folding to aggression and attacking capped ranges relentlessly.
How the opponent model works
Exploit is built on the same opponent profiles used across Pokercat, taken to their full depth:
- Profile as prior — the archetype you assign (or tune) describes how that player type reacts to actions: what they call with, raise with, and give up with, by street and by situation.
- Range reconstruction — as the hand plays out, the villain’s range is re-weighted action by action according to the profile, instead of assuming equilibrium play.
- Best response — the engine then solves for the highest-EV line against that reconstructed range.
The profile parameters are tunable, so a profile can be sharpened as you gather reads on a specific player pool.
Status and rollout
The Exploit mode is visible in the Dashboard’s mode selector today as a placeholder. It will roll out gradually, and hands played with it are already tracked as their own engine in the Data Hub and on the Ranking page, so you will be able to measure whether the aggression pays from day one.