Rational Cooperation in the Finitely-Repeated Prisoners' Dilemma