Social Network Analysis & GOP Verbal Attacks

Not that I know anything about the GOP debates or candidates, but I casually saw in a CNN post this nice visualization of verbal attacks during the RL GOP Debate, and I thought that I would do a little SNA and try to draw conclusions on the debate WITHOUT actually having seen it…

let’s see how it goes and, please, if you’ve seen the debate and know better than me, let me know if I am very wrong :)

So I reversed engineer this plot into a network I could work with and then I run a modularity class algorithm too see how the candidates group (those with denser connections), and this is the result:

So the algorithm identified two clear groups. One might speculate that within the GOP there is also a bipartisan struggle and that eventually the survivor from each group will be the runner up. But again, since I do not know who the candidates are, I did a quick check on their backgrounds to see if I could identify any patterns.

If we look closely it seems like if each group clustered their attacks on their own “kind”. On the one hand we have a group with all the early generation American candidates from Latino, Czech, Indian parents or grandparents as well as non anglo-saxon sounding names like Pataki. The only exception seems to be Graham, which makes you wonder the reason why Graham’s team went into that fight… Genius/Risky move?

On the other hand, the second group is formed by old generation Americans citizens in which, apparently and given the intensity of their attacks (the connection arrow’s width), Trump and Bush had an epic battle.

We could go on and on about different aspects of this network, but let’s try to guess who will eventually be the winner from each group based on the analysis of their SNA metrics.

There are many metrics to consider (Centrality, Authority…) as well as their interactions but, in the Old America’s group, the winner in all of them is hands down Donald Trump followed in the distance by Jed Bush.

In the New America’s group there is no clear winner and it is a close call between Rick Santorum, Graham & Cruz.

Anyway, based on these metrics, if I had to put my money on who will be in the next Republican team for the 2016 election, my money would go for:

So much fun…

Update: Missing connections in a network are just as important as the present ones. If we check the attacks network we can see that Trump has no attack connections from the New America’s group with Kasich and Cruz, whereas Bush has no attack connections from the New America’s group with Rubio, Pataki, Graham, Kasich & Jindal.

This might mean that Cruz is aiming to be Trump’s sidekick and Rubio and the rest Bush’s companions, and this would also explain Graham’s strategy joining the New America’s members fight; to have a chance to be Bush’s Vice president.

Post Election Update: So after the more than interesting elections in the US I reviewed some of the debates including the one analysed in this post. Turns out that this debate was hold in two part when I thought it was just one. That is why the CNN graph has an horizontal line dividing Pataki, Santorum, Jindal & Graham from the rest.

These four candidates had way more time to talk than than the rest of candidates but they were also the four candidates with less support and could not directly address the others.

So if we eliminate these four from the analysis the close call among Santorum, Graham & Cruz turns into a clear choice in Ted Cruz among the New America’s candidates which, in hindsight, shows that the network correctly identified the two main contestants for the GOP nomination: Donald Trump & Ted Cruz.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s