Repeated rock, paper, scissors play reveals limits in adaptive sequential behavior.

Area of Science:

• Cognitive Science
• Game Theory
• Behavioral Economics

Background:

• Rational models predict random play in mixed strategy equilibrium (MSE) games.
• Humans often deviate from random play, attempting to outwit opponents in repeated adversarial interactions.
• Understanding the adaptive reasoning behind these deviations is crucial.

Purpose of the Study:

• To investigate human adaptive decision-making in adversarial settings.
• To identify the complexity bounds of opponent behavior detectable and modifiable by humans.
• To examine how people adapt their strategies in repeated rounds of Rock, Paper, Scissors.

Main Methods:

• Experiment 1: Participants played against bots with stable move patterns to assess detection and adaptation limits.
• Experiment 2: Bots exploited human move patterns to assess the limits of revising one's own behavior.
• Analysis focused on transition patterns between consecutive moves.

Main Results:

• Humans effectively detect and exploit simple sequential move patterns in opponents.
• Humans strategically modify their own simple sequential move patterns in response to opponents.
• Adaptive reasoning demonstrated significant limitations when faced with more complex patterns.

Conclusions:

• Human adaptive decision-making in adversarial games is robust but limited in scope.
• People exhibit flexible attention to simple transition patterns but struggle with sophisticated strategies.
• Findings provide a precise account of the boundaries of human strategic adaptation.