Write a game that plays many rounds of Rock Paper Scissors. The user and computer will each choose between three items: rock (defeats scissors, but loses to paper), paper (defeats rock, but loses to scissors), and scissors (defeats paper, but loses to rock). If the player and computer choose the same item, the game is a tie.
You could extend this program to include different algorithmic strategies for choosing the best item. Should the computer pick randomly? Should it always pick a particular item or a repeating pattern of items? Should it count the number of times the opponent chooses various items and base its strategy on this history?