Who Will Ultimately Die in the New Season of Game of Thrones?...

Who Will Ultimately Die in the New Season of Game of Thrones? Let’s Ask an AI

57
SHARE

With the newly released season 7 of the acclaimed fantasy drama Game of Thrones, fans, and yourself too if you closely follow the series, may be wondering if it’s their favorite character’s fate to die or not. Well, to save you from cluelessly thinking who will ultimately meet their death, an AI has painstakingly digested the large volume of information available about the series to come up with a death list.

[Image Source: HBO]

Milan Janosov, a Ph.D. candidate from the department of Network Science at the Central European European, formulated a vast social ecosystem that essentially predicts a character’s fate throughout the duration of season 7.

Using network science to predict the fate Game of Thrones characters

As an avid fan of the fantasy drama series, Janosov couldn’t help but wonder which of his beloved Game of Thrones character would ultimately face their demise in the newly released season 7. So, even before the fresh season was aired, the Ph.D. candidate took matters into his own hands by using network science to predict who will die during the course of the season.

Janosov primarily formulated a ranking status category for each character according to their social interaction patterns with their peers. He then made use of machine learning methods to predict if a particular character is likely to die or not. To make it easier to read the network, each member of the great houses are assigned and plotted with their own colors. The Starks are plotted as blue, Lannisters are assigned with the color red, and the Martells are plotted down using yellow.

[Image Source: Central European University]

By using the almost 600 scenes from the previous four seasons, Janosov was able to construct a vast network of character links and ties. The scenes are extremely useful in gauging a character’s strength within the social ecosystem. “The scenes are complete graphs, or cliques, increasing the tie strength between all pairs of people present by one”, said Janosov. Each character is a given a node which varies in size according to how popular they are in the social system.

A complete network science system of the Westeros realm

After formulating the social interaction system, a complete network science system of the Westeros realm was achieved by calculating how important nodes are. The highly important nodes (or characters) are given name labels and the least significant nodes are filtered out. Ultimately, Janosov ended up with a network of almost 400 nodes and more than 3,000 edges. From the complex network of nodes and edges, he was able to quantify a character’s possibility of dying.

Janosov revealed that the artificially intelligent machine learning method was able to digest the large volume of information and show the hidden links between various characters.

“As network measures are often very correlated, we can’t pick out one or two that are highly predictive on their own, but seemingly characters with high betweenness, low clustering, and high degree are less likely to be killed. In any case, the strength of the machine learning approach is exactly finding hidden relationships among a large number of features”.

SPOILER ALERT!

So, which beloved Game of Thrones personas are predicted to ultimately meet their death this season? According to Janosov’s network science predictions, one of the infamous Sand Snakes, Tyene Sand, and Princess Daenerys Stormborn are most likely to face their demise during the course of the new season. Whilst Cersei and Baelish are forecasted to make it through to the next and final installment of the fantasy drama.

GoT predicted death list using network science GoT predicted death list using network science

  [Image Source: Central European University]

Given those highly formulated predictions, how accurate do you think this correlates with the actual dynamics of season 7 so far? If you’re a die-hard Game of Thrones fan then we are interested to hear your opinion on this death list.

Via Central European University