I don’t know a lot about social network analysis but I do know how to plug numbers into software I don’t entirely understand. So there’s no blistering insights from the following graph, I just thought it looked neat.
This is the data from the previous post converted into a network graph. The thickness of the bar are the proportion of points that re-allocated as a nominee was eliminated. The data is essentially incomplete because EPH stops when the six finalists have been identified. In theory the algorithm could keep going until there was one ‘winner’. So, for example, Cora and Paul don’t have a connection because they were both finalists. I also turned it into an undirected graph because I messed up how I imported the data into Gephi (https://gephi.org/ ) and the arrows were pointing the wrong way round.
The main thing is that it ended up looking pretty, like a faceted gemstone. It also shows that EPH functions in a way that is somehow both competitive but also mutually supportive.