ESPN has tested an artificial intelligence system during its World Series of Poker broadcast that claims to detect when players are bluffing by analysing subtle physiological cues. The system, developed by AI researcher Luke Geel, was featured on ESPN’s WSOP coverage and uses computer vision to track indicators such as blink rate and posture changes.
The AI tool, described as a poker tells detector, operates by processing real-time video of players at the table. It monitors involuntary physical reactions that may signal whether a player is confident or bluffing about the strength of their hand. Blink frequency, facial muscle movements, and shifts in body position are among the data points the system evaluates.
During the ESPN broadcast, the AI overlay highlighted moments when players appeared to display stress or confidence signals, providing viewers with an additional layer of analysis alongside traditional commentary. The technology represents a convergence of computer vision, machine learning, and behavioural psychology applied to a live sporting context.
Poker has long fascinated AI researchers, with previous systems such as Libratus and Pluribus demonstrating superhuman ability at the game itself. However, this application focuses not on playing the game but on reading the human players – a distinctly different challenge that relies on interpreting non-verbal communication rather than calculating optimal strategy.
The test on ESPN’s WSOP broadcast is one of the first instances of real-time bluff detection being shown on mainstream television. It remains unclear whether ESPN plans to integrate the technology into regular coverage or offer it as a permanent feature.
This article was adapted from BeInCrypto. Read the original here.
