rock-paper-scissors
Ordinarily, this much confidence would border on arrogance, but there’s no way we can stay mad at that face.
A couple of years ago we reported on a robot hand that could always win at the timeless hand game rock-paper-scissors or janken as it’s known here in Japan. After wrapping up, we confirmed that it would never lose, declared that the human race was doomed to sit in the back seats of our robot overlords, then called it a day.
Now we are surprised to learn that Tokyo University’s Ishikawa Watanabe Lab is back with an even better performing rock-paper-scissors robot, somewhat awkwardly dubbed the Janken Robot with 100% Winning Rate.
Rock-paper-scissors. Scissors-paper-stone. Roshambo. Elephant-man-ant. Whatever you call it, chances are you’ve played it at some point. In Japan, the game is known as janken, and is used to settle any kind of dispute or awkward situation, from who gets the last cookie to which parents have to sit on the PTA that year.
It’s not hard to see why janken is so popular in Japan: it’s simple, and everyone knows how to play. It’s also efficient (particularly if the thing being decided is trivial anyway). Decisions made by janken are stuck to religiously: in three years teaching Japanese schoolkids I never once saw a student complain about the result or demand a rematch. It’s seen as a fair way to make decisions, because the game is based on luck.
Or is it? A group of researchers from Chinese universities has published a paper that shows sure-fire ways to win at rock-paper-scissors. Join us after the jump as we explore how to outsmart small children at their own game!
Rock-Paper-Scissors, the longstanding arbitrator of riding shotgun or eating the last slice of pizza has been celebrated for hundreds of years for its simple yet elegant balance of psychology and chance.
It’s such a part of the human experience that a robot could never out match the human mind in the RPS arena. Until now that is, as Engineers from the University of Tokyo decided to stick their noses in and build a robot that never ever loses at Rock-Paper-Scissors – ever! So how does it do it?