Can we predict the future using a temporal version of the Copernican principle? [pdf]

Upvoted for this mathematician's-gauntlet-throwing snippet near the end:

"I sent an e-mail on 21 October 1999 and again on 2 December 1999 to my department's most comprehensive e-mail alias, which includes faculty, staff, and graduate students, requesting information on pet dogs. The responses were compiled and checked for accuracy on 6 December; a notarized list of the 24 dogs, including each dog's name, date of birth, and breed, and the caretaker's name, was deposited in my departmental personnel file on 21 December 1999. Gott's rule predicts that each dog will survive to twice its present age with probability 1/2. For each of the 6 dogs above 10 years old on the list, I am offering to bet Gott $1,000 US, at odds of 2:1 in his favor, that the dog will not survive to twice its age on 3 December 1999. The reason for weighting the odds in Gott's favor is to test his belief in his own predictions: given the odds, his rule says that his expected gain, at $1,000 per bet, is $6,000; moreover, the probability that he will be a net loser (by losing five or more of the six bets) is 7/64=0.109."

Here's the New Yorker article mentioned in the paper:

http://www.newyorker.com/archive/1999/07/12/1999_07_12_035_T...

The subject being discussed is better-known as the Doomsday Argument, btw:

http://en.wikipedia.org/wiki/Doomsday_argument