Rock Paper Scissors (RPS) traces its origins to a Chinese game called Shoushiling, documented in the Ming dynasty (17th century). The game spread to Japan as Jankenpon, then to Europe and America in the early 20th century. Today it serves as a genuine competitive sport — the World RPS Society has run international tournaments since 2002, with cash prizes and media coverage.
Against a truly random opponent, Rock Paper Scissors is mathematically unwinnable — every choice has a one-third chance of winning, drawing, and losing. But humans (and many simple AI implementations) are not truly random. Human players show measurable biases: Rock is the most commonly chosen throw (roughly 35% of throws), followed by Paper (33%), and Scissors (32%). After losing, players tend to switch to the throw that would have beaten their opponent's last throw. After winning, players tend to repeat their winning throw.
The World RPS Strategy Guide (an actual published book) documents these patterns and trains players to exploit them — essentially making RPS a game of psychology and pattern recognition rather than random chance. Professional RPS tournaments use best-of-many formats with enough throws to statistically separate skilled players from lucky ones.
This version's AI uses a combination of random choices and adaptive pattern recognition — it will start to pick up on your tendencies if you develop them. The best strategy is to avoid patterns entirely, but that's harder than it sounds under time pressure.
Can You Beat a Random Computer at RPS?
- Against truly random AI: No strategy works. Pure variance determines the outcome over any short series.
- Against adaptive AI: Switch up your patterns deliberately. If you've thrown Rock twice, throw Scissors (counter-intuitive).
- In human competition: Throw Paper in the first round — most people default to Rock on round 1 (it's the strongest-feeling choice psychologically).
- Watch for repetition patterns. Even players trying to be random tend to avoid repeating the same throw three times in a row. You can exploit this.