This chapter's title signifies not only that backward induction is introduced, but that some backing up is necessary to put the material of the previous chapter on a more solid foundation. The game of Matching Pennies, with and without peeking, is used to illustrate the difference between games of perfect and imperfect competition. The rules of a game of a general game of perfect information are then described using the idea of a game tree. The principle of backward induction is introduced for the case of win-or-lose games, where its application is entirely uncontroversial. This leads to the notion of the value of a win-or-lose game. As examples, the values of the games of Nim and Hex are found. Chess is then introduced as an example of a strictly competitive game with more than two outcomes. It is shown that finite strictly competitive games always have a value. The connexion between Nash equilibria and saddle points is explained. Subgame-perfect equilibria are introduced as the end-product of backward induction. It is explained why some Nash equilibria are not subgame-perfect. The chapter ends with a discussion of the rationality assumption, and why subgame-perfect play may not always be a good idea when an opponent fails to behave rationally.
Oxford Scholarship Online requires a subscription or purchase to access the full text of books within the service. Public users can however freely search the site and view the abstracts and keywords for each book and chapter.
If you think you should have access to this title, please contact your librarian.