Simultaneous Randomized Chess — LessWrong