The Network Density in Movies
Just like your own Facebook- or real-life-network, the clusters, scenes and networks in movies are the environments that shape a story. Taking into account that you observe characters and their interactions you identify and empathize with, you can see social networks in movies to be in resonance with your personal life. Just like in movies, it makes a big difference if you have a network of 5 close buddies or 20 loosely-coupled friends. Here, we have arrived at our topic: The network density in movies.
The density of a network plays an important role in the dynamic development of a motion picture story. In a nutshell, the network density is “the proportion of edges in a network relative to the total number of possible edges”. Most probably, maybe 5% of you are now better informed than before. So, let´s go slow: Primarily, the social network is an empirically discovered structure. Edges in the network are the ties (connecting lines) between nodes, namely characters. Here, network density is the proportion of ties present relative to ties possible among alters in an character´s ﬁrst degree neighborhood. In this context, network density can be a measure of how many of my friends are friends themselves, controlling for my network´s size. A dense network´s nodes are “densely-knit” and “tightly-bounded”, meaning that the most relevant ties stay within the network. In sociology, network density is often seen as an indicator of the extent to which individuals identify with those around them (homophily) or it is noted as an important factor of social capital.
For a movie analysis, it is interesting to know who belongs to a group, as well as the patterns of relations that explain and sustain such a group. For example, dense networks have considerable direct communication among all members - This is the classic case of a romance. By contrast, few members of “sparsely-knit” networks communicate directly and frequently with each other, see The Lord of the Rings. In unison, these networks provide character with considerable room to switch between relationships (esp. sub-clusters) and scenes. As a consequence thereof, you will notice a higher diameter and clustering coefficient within a interaction design of that kind. To understand the idea of network density easily, just explain yourself, why the average value in romance movies is significantly higher than for any other genre.
Typical for romance movies is a smaller network (#nodes) and a relative higher amount of strong connections (#edges). Typically, a man and a woman play the central role in the movie and have a strong connection to each other and their peers throughout the plot. The density is higher because of the inner conflict of the characters, that reveal it to the close connections. Especially, the inner conflict is often projected outward onto something else, for example through the character expressing his or her feelings to another friend. While we can state that so far for most romance movies, we are surprised to discover that Romeo & Juliet doesn´t perfectly fit into that frame with a network density of 0.12.
More interesting -and maybe a small world premiere ?- is one of our other insights from the first (non-scientific) analysis. For example, we can observe differences among movie directors in the dynamic arrangement of their character network. Just have a look:
The chart shows each director´s average dynamic design of the character network in a movie. You will discover that Steven Spielberg and Oliver Stone have a different way to design the character network to the end of a movie than the Coen brothers and Clint Eastwood. Here, the increasing density in the last quarter of their movies is a result of the elimination of side-characters, and the general compression of the network at a higher amount of close interactions. The “solution of the conflict” and the showdown of the protagonist and antagonist in the drama is what we can see here. On the other hand, Quentin Tarantino and David Lynch are similar in their (average) network design at the half-runtime of a movie. Their network densities drop significantly because of the introduction of new side-stories and herewith side-characters and scenes into the story, increasing the overall network size.
These small insights are just the tip of the iceberg. There are a plenty of other factors that vary among directors and actors. A whole “Movie DNA” can be explored and be sure, we will. Maybe, you also have something discovered you like to share with us?
Stay tuned and have fun galaxing some movies!
Cheers, your Moviesgalaxies team