Alexander Field was not convinced of a result that I claimed in my JEP 2001 paper that in "haystack models" with non-assortative mating, if the number of descendants of founding group members is determined by an n-player prisoners' dilemma game, then the population will converge to a population of defectors. He thought that the result applied only if the groups were large. I respond with a more detailed discussion and show how the result works even when groups have only two members.