Friendship as Contagion?

The Centola-Macy made me consider a few interesting aspects of network contagion that haven't really come up in the literature. The first is the idea of friendship as contagion. When reading their discussion about how some contagions require multiple contacts to communicate them, I began to wonder if friendship formation in networks worked similarly to many of the behaviors they thought had yet to be modeled effectively in research. That is, that the probability of becoming friends with someone (p) increases greatly the more contacts you share with that person. That is, with zero shared contacts ties would form only rarely; with one shared contact, more frequently (causing triad closure); and with two shared contacts, almost inevitably. This seems to me to be the next "level" in network studies since until now what we've read has seemed mostly concerned with the effects of networks rather than networks effects upon themselves.

Also, though this is more of an aside, I found it strange that they said urban legends become more accepted the more people you hear them from. I would argue that it actually goes in the reverse direction. The first time I heard the "dead person on the subway" urban legend, it was in Missouri. The second time it was exactly the same, but took place in Boston. I immediately became skeptical. It seems as though there are interesting potential studies out there in studying other contagions that are communicated optimally at a certain point and less effectively above and below that. Another good example could be something like door to door campaigning that could potentially backfire once people receive a certain amount of annoying visits.