A Game of Nodes
https://www.macalester.edu/~abeverid/thrones.html
As one of the highest grossing TV series of the decade, Game of Thrones has successfully pulled in viewers with its intricate and dynamic plot lines over the 6 seasons aired so far. A very unique aspect to this show is the vast number of characters placed around the fictional world of Westeros that it follows simultaneously. The show effectively captures the many interactions that occur between the characters, making it an ideal subject for the study of networks.
The graph above shows an interaction network of the characters from Season 3 of the show. With the large number of characters placed far away from each other in the world, direct interaction rarely occurs. In addition to face-to-face contact, the graph also includes situations where characters refer to other characters and when characters are referred to together – the larger the number of these links, the thicker the edges connecting the characters. The nodes themselves are proportioned according to the relative appearances and mentions of the characters.
There are several characteristics I found interesting about this graph. The first is the Strong Triadic Closure property that states two strong ties (in this graph either positive or negative) are likely to be closed off in a triad. If a character on the graph has edges thicker than a certain degree with two or more others, it is very likely that they are connected too. This makes logical sense since the thicker edge signifies more screen time which is bound to overlap with other characters leading to some form of relationship. Another notable (and rather obvious) feature of the graph is that the characters with the largest nodes are clear fan favorites. This setup leads to formations of groups organized primarily by alliances and geography with local bridges spanning between the groups. Finally, this graph, and the representative network, is a very good example of the dynamic nature of networks in general. As the story progresses, more interactions occur while some cease, triads are formed and the groups would eventually converge. This perfectly exemplifies how graphs in networks are snapshots at a particular point in time.