Tuesday, November 11, 2003

Backwards Induction and the Death Penalty

The logic-and-decision-theory puzzles that have been getting some attention in the blogosphere lately (from me, for example) reminded me of this story told to me by Jim Dow -- yes, the same Jim Dow who posts on this blog sometimes. But I’m sure he won’t mind my stealing his story (especially since he didn’t make it up anyhow).

An economist was sentenced to die within the next seven days. He was not told his day of execution, but he was informed by the King that whichever day it was, the economist would be surprised when it happened. Noting that the current day was Sunday, the economist thought to himself, “Well, I definitely can’t be executed next Sunday, because if I haven’t been executed by Saturday, I’ll know that Sunday is my execution day, and thus I won’t be surprised. And I can’t be executed on Saturday, either, because if I haven’t been executed by Friday, I’ll know that Saturday is my execution day (Sunday already having been eliminated), and thus I won’t be surprised.” Proceeding along these lines, he reasoned that he couldn’t be executed on Thursday, Wednesday, Tuesday, or Monday, either. He therefore concluded, happily, that he would not be executed at all.

On Monday morning, to his great surprise, he was executed.

No comments: