Game Theory and the film ‘WarGames’
http://www.tft.ucla.edu/mediascape/blog/?p=1230
WarGames is 1983 film which I remember having seen as a child. While I was still too young to follow everything in the film the first time I saw it, WarGames was my first introduction to the concept of game theory. In the film, a failing high school student and genius hacker by the name of David Lightman, in typical Hollywood hacker fashion, unwittingly hacks into a NORAD supercomputer that has the ability to learn and is responsible for running military simulations and responding in the case of nuclear attack. He finds a number of simulations on the supercomputer and, thinking they are games, challenges the supercomputer to a match of “Global Thermonuclear War.” The supercomputer confuses reality and simulation and begins preparing to launch actual missiles for its simulated response. By this time, Lightman has gone through a series of events which has resulted in his being located in the NORAD control room, trying to stop the missiles from launching. He directs the supercomputer to simulate playing tic-tac-toe against itself repetitively. Because the supercomputer always made the ideal move for both players, each of the thousands of games it simulated resulted in a draw. From this, the supercomputer learns of the possibility of a no-win scenario. The supercomputer begins running simulations of all possible nuclear launches, concluding after each that there is no winner. The supercomputer then halts the nuclear launch, stating “A strange game. The only winning move is not to play.”
The games and simulations that the supercomputer plays through in WarGames are related to the concept of game theory that we have been discussing in class. The goal of game theory is to model situations where the payoff of a choice relies not only on an individual’s choice, but also on the choices made by others that the individual is interacting with. Tic-tac-toe, when played by two players that are competent, is a game that will always end in a draw. In game theory, a game of this type is called a “no-win situation,” a situation where no player benefits from any outcome. In WarGames, the supercomputer finds that, simply by attempting to win, both players have guaranteed the fact that neither will win. In the supercomputer’s simulation of global thermonuclear war between the US and the USSR, the entire simulation breaks down into four possibilities:
- Both US and the USSR launches
- The US launches and the USSR does not launch
- The US does not launch and the USSR launches
- Neither the US nor the USSR launches
If one country launches, the other will know and will be able to launch a retaliatory strike. In such a scenario, launching a retaliatory strike would reduce the damage done to the country launching it, so, assuming the country being attacked only cares about itself, one would expect a retaliatory strike to be launched every time. This is called a Nash Equilibrium, or a pair of strategies that are best responses to each other. In this situation, the dominant strategy is to withhold from launching an attack until an attack is launched by another country. Because of this, the supercomputer finds that the game of global thermonuclear war is a no-win situation and that, while it is possible to come out on top, it is not possible to make any sort of gain from the launch of a nuclear attack. As the supercomputer says, “The only winning move is not to play.”