Mixed strategy Nash equilibrium — LessWrong