Presidential Primaries are a game. So here’s some more elementary game theory — on the game known technically as “chicken” — to predict the Rubio-v.-Kasich endgame. This is brought to you by the brilliant graphics editor at The New York Times, Kevin Quealy. Here’s the conclusion:
The chance to be your party’s nominee for president comes along only every four or eight years, even for the very luckiest candidates. If the candidates lived in a universe in which they could run for president hundreds of times, they might agree that, on average, their shared interests were better served by cooperating. Once in a while, Mr. Kasich might try to win the contest outright against long odds, but, on average, he would probably agree that cooperating, including alternating victories, was the best way to serve his and Mr. Rubio’s shared interests. Game theory shows that initerated dilemmas, played many hundreds or thousands of times, cooperation is a very stable strategy — one reason it is so common in nature.
But this is not an iterated dilemma. It’s a one-time-only dilemma with a tremendous payoff for the winner. As much as Mr. Kasich might think about his legacy, the good of the party or even his own chances in 2020 or 2024, the future is very far away.
Ultimately, they risk an outcome neither he nor Mr. Rubio wants. As Daniel Diermeier, the dean of the public policy school at the University of Chicago, notes, “A very important lesson of game theory is that sometimes the world is a grim place.”
Then again, the Presidential Primaries aren’t just a game….
Ethics for Adversaries 