Artificial Intelligence RPS

Rock Paper Scissors is a simple game played by two players using hand gestures. At the same time, each player plays a gesture with the scoring as follows:

Rock beats Scissors
Paper beats Rock
Scissors beats Paper

If the game is played only once, no dominating strategy exists and the result is down to chance. If the game is played multiple times the chance element reduces as players’ next gesture is influenced by two factors. The first being the previous result (Win, Lose, Draw). The second being the previous gesture (Rock, Paper, Scissors).

Previous attempts to create an Artificial Intelligence for Rock Paper Scissors have relied on simple frequency analysis (User has played rock the most, therefore AI must play paper next) or history matching (match the last 4 four rounds to a big database of previous games and determine what the user will play next, then the AI simply counters it). In an actuarial spirit, I take a stochastic approach and instead of determining the next gesture with certainty, I flex the probabilities for Rock, Paper, Scissors and then let my AI randomly choose one. How I flex the probabilities is based on previous results and previous gestures.