The Research

The White Paper

Playing the player: A Heuristic Framework for Adaptive Poker AI

This paper presents Patrick, an AI built on the philosophy that the path to victory lies not in being unexploitable, but in being maximally exploitative. It details the system’s novel architecture, its prediction-anchored learning method, and its profitable performance during the trial. This document challenges the “solver” orthodoxy and makes the case for a new approach to AI in human-centric games.

[Download the White Paper here]

The Trial Data

This is the complete, unaltered 64,267-hand history from Patrick’s trial, conducted from 1st January to 26th February 2023. The data is provided in a Poker Tracker-compatible format for independent review, analysis, and validation of our results.

[Download the Trial Data (Hand Histories) here]

The Video Analysis

Data alone doesn’t tell the full story. To provide a complete picture of Patrick’s performance, we are presenting all 16-hands in two parallel series.

On the left, the ‘Easy Game’ series provides humorous, personality-driven commentary from Patrick himself. On the right, the ‘Trial Data’ series presents the raw, uncommentated footage for your own clinical and scientific review.