Yeah, my hypothesis would be that that sort of analysis will find a cluster of very tightly similar ballots that are the Puppies, probably divisible into two sub-clusters for sad and rabid; and then a smaller, much less similar but still significant and discernable cluster of pixel-stained technopeasants at the 70-100 level, which have a lot of spread but overlap on several things in most categories. And then maybe a few very small clusters that are, like, very specific subgenres or groups of friends, but everything else is pretty scattered.

I hope we get some good stats out of this! Because stats + SFF fandom can only lead to good things.