Yet another use of the WebGL renderer I wrote, this time attached to a rock, paper, scissors bot inspired by poking around at http://www.rpscontest.com/ This one aims to win about 60% of the time. It resets the model when a new game starts, but within a game the key to winning is to train it to expect one thing and then to change strategy without warning.
Show HN: WebGL rock, paper, scissors against a Markov Chain in a sombrero | Heykuki News