Corrections

2004 was a very confusing year as far as attendance goes.

2007 Analysis by Brent

Since Virgitarian-ness flows outward from established folks to newbies through invitation, it makes sense to model our relationships as a directed graph spraying outward from Ben. Seen this way, it's clear that we all have Ben as our Virgitarian uber-ancestor. (Thanks, Ben!) Downstream from Ben, the graph separates into: a 18-person Sarah/Sean/Kate section; a 3-person Kevin section; and a 42-person AJ/Mandy/Mel section.

Using an undirected graph instead, we can take the circa 2007 relationships as a given instead of a chronological development and ask: where is the social weight of the graph centered? (We're finding the graph's balance point(s), the people around whom the most relationships converge.) Removing nodes of the smallest degree and accumulating this degree along the remaining edges, we can peel the graph from the outside in. In this way the graph reduces to two centers: one between Mel and Mandy (with degree 56) and one at Kate (with degree 20). So the center has moved outside the original BM2004 five, but not by much. We're two social networks tied together as Virgitarians by history, an elliptical camp with two foci.

Either way you slice it, the junk pile clearly belongs in mel/mandy/aj's back yard, although kate/sean should probably have some too. QED :)

Is there a mathematician in the house?

If there are any Virgitarians that have some math or graph theory and know of free software packages that do clustering or weighted analyisis of graphs like these (an auto-brent, if you will) I'd be very interested.

Scilab plugins? Maybe I should steal Matlab?

sigma cluster detection with Matlab