You Can Pick Your (Best) Friends

Recent research has revealed many remarkably robust structural
properties of social networks: triadic closure, heavy-tailed degree
distributions, and small-world phenomena, to name just a few. Of
course, the why of these properties has generally been more
elusive. In this talk, I will present some results from a recent
collaboration with evolutionary psychologists and computer scientists
on questions of how people choose friends and prioritize among those
friends. Specifically, I will describe analysis of large sample of
MySpace profiles containing Top Friends lists, in which an
individual selects a small subset of his or her friends and organizes
them into a ranked order of that individual's choice. Different
classes of behavioral hypotheses give rise to very different
graph-theoretic structures in the best-friend network, and we can use
these ranking data to provide supporting evidence for some of these
theories.

Joint work with Peter DeScioli, Robert Kurzban, and Elizabeth Koch.