Suppose 2/3 of people surveyed like the new trial flavor of chips. How many people would you have to select randomly, on average, in order to get 10 people who DO NOT like the new trial flavor?