Data Science, Psychology Reveal Why the ‘Game of Thrones’ Books Are So Riveting

A network model demonstrates how George R.R. Martin’s sprawling series remains comprehensible but surprising

Jon Snow and Daenerys Targaryen
In the television adaptation, Kit Harington and Emilia Clarke play point-of-view characters Jon Snow and Daenerys Targaryen. Helen Sloan / HBO

Readers may have come to George R.R. Martin’s A Song of Ice and Fire—the book series behind HBO hit “Game of Thrones”—in search of dragons and ice zombies, but they stayed for social interactions that mimic what human brains expect in real life, a new analysis by researchers at five universities across the United Kingdom and Ireland suggests.

As detailed in the Proceedings of the National Academy of Sciences, a team of physicists, psychologists and mathematicians used network science and data analytics to create a web-like model of the fantasy novels’ plots. This visualization helped them identify patterns in the sweeping narrative’s twists and turns, as well as in characters’ behavior.

“People largely make sense of the world through narratives, but we have no scientific understanding of what makes complex narratives relatable and comprehensible,” says co-author Colm Connaughton, a mathematician at the University of Warwick, in a statement. “The ideas underpinning this paper are steps towards answering this question.”

Connaughton and his colleagues found that the five Ice and Fire books published to date feature 2,007 named characters and more than 41,000 interactions. Yet each character only interacts with about 150 others at most. Per the paper, scientists estimate that this figure represents the maximum number of individuals that human brains, having evolved in relatively small communities, are equipped to deal with.

Greg Evans of Indy100 notes that Martin’s emphasis on multiple characters’ points of view may help people look at relationships and social interactions through a particular perspective. A total of 24 individuals narrate sections of the saga, with Tyrion Lannister (47 chapters across the five books) and Jon Snow (42 chapters) topping the list. At the other end of the spectrum, Melisandre and Ser Arys Oakheart each receive just one POV chapter.

The researchers also looked at Martin’s notorious habit of killing off main characters, explaining why the deaths may come across to readers as both shocking and believable.

Data Science, Psychology Reveal Why the 'Game of Thrones' Books Are So Riveting
Researchers mapped out a network showing interactions between characters in the books. University of Warwick

First, reports Julia Naftulin for Insider, the team rearranged the story in chronological order using a timeline created by fans on Reddit. Based on this data, the scientists found that the books’ pace of significant deaths followed a pattern mirroring the distribution of nonviolent events in the real world. Though this organization could have made characters’ demises feel predictable, Martin’s non-chronological structure ensured that the deaths managed to take readers by surprise.

“This study offers convincing evidence that good writers work very carefully within the psychological limits of the reader,” says co-author Robin Dunbar, an evolutionary psychologist at Oxford University, in the statement.

At Cosmos magazine, Nick Carne writes that critics have compared the fantasy series to both history and myth. He adds, “[T]he marriage of science and humanities in the recent study opens new avenues to comparative literary studies.”

Comparing the books, set mainly in a fantasy version of medieval Europe, to centuries-old European literature, for instance, the authors found that Ice and Fire has more in common with the Icelandic sagas than the Old English heroic poem Beowulf or the Irish folk epic Táin Bó Cúailnge.

“This kind of study opens up exciting new possibilities for examining the structure and design of epics in all sorts of contexts,” says co-author Ralph Kenna, a theoretical physicist at Coventry University, in the statement. “Impact of related work includes outcry over misappropriation of mythology in Ireland and flaws in the processes that led to it.”

This isn’t the first time researchers have attempted to make scientific sense of the “Game of Thrones” universe. As Jason Daley reported for Smithsonian magazine last year, a team from Australia attempted to figure out who in the television series was most likely to live or die based on gender and social status. And back in 2017, Smithsonian’s Ben Panko described how a software engineer attempted to use a neural network to write a new Ice and Fire book.

Unfortunately, artificial intelligence has yet to perfect the art of imitating human authors—one of the neural network’s nonsensical sentences reads, “The woods are gowned on bloody yellow and glass”—so most readers will simply have to wait for Martin to finish the final, much-anticipated two volumes of the series.