Skip to Content

Algorithm predicts who will live and die in ‘Game of Thrones’ season 8

Since its premiere, HBO’s “Game of Thrones” (GOT) has been saturated with death. In fact, many fans see character deaths as a defining trademark of the hit show. Each season, viewers tune in to see if their favorite characters make it through to the end, with some even rooting for the deaths of those they despise. There is even a cottage industry of betting on GOT character deaths, with sports books raking in money each season from the show’s violent antics.

Due to GOT’s grisly nature, fans and gamblers alike want to know each characters’ odds of survival before sitting down and pledging time to another season. With HBO clamping down on leaks and spoilers to an incredible degree, how are fans expected to make accurate predictions or prepare themselves for the emotional shock of a beloved character dying?

As it turns out, a group of data scientists at a university in Germany felt the same way. They created a computer algorithm that calculates the probability of a “Game of Thrones” character dying based on multiple points of information, such as gender, family house and marriage status. The results are presented as percentage based odds that can give fans some much needed hope during season 8. Do you want to know “wins or dies”?

How to find out who dies in ‘Game of Thrones’ season 8

For those of you who hate spoilers, be aware that some of these predictions may turn out to be correct. Proceed with caution. 

In order to make the most accurate predictions about the final season of “Game of Thrones,” data scientists at the Technical University of Munich built their algorithm around machine learning and in-universe data points.

By matching characters with highly specific information about their current status in the world of Westeros, the algorithm would be able to compare and rank them in order of most survivable to least.

Some data points used were aspects like family house, alliances, marital status and location. In addition, the team pulled data from the many fan wikis for “Game of Thrones” that exist online, with frequency of appearance determining “main character” status for several on the ranking.

Once information was pulled, the algorithm analyzed trends in data in order to build its predictions. Some trends were more useful than others in building the ranking.

For example, analysis of character gender proved that male characters in Westeros are much shorter lived: having a 22% death rate compared to female characters. Some of the family houses have proved more resilient than others as well, with living members factoring in to survivability.

Predictions for which characters die in ‘Game of Thrones’

After comparing trends and ranking data, the algorithm predicted which characters were most likely to survive the long winter in season 8. Based on the list, Queen and Khaleesi Daenerys Targaryan is ranked as most likely to live.

With her Hand of the Queen, Tyrion Lannister, placing second and Varys the court spymaster placing third. The top three have a 0.9%, 2.5%, and 3.2% chance of dying, respectively.

Ranked as most likely to die is Bronn the mercenary, with a whopping 93.5% chance of death this season. Followed by Bronn is Gregor Clegane, also known as The Mountain, personal guard of Queen Cersei Lannister at 80.3%.

To the shock of her many fans, Lady of Winterfell Sansa Stark is trending toward death at a 73.3% chance of meeting an early demise. Considering her tremendous suffering at the hands of King Joffrey, Queen Cersei and Ramsay Bolton, it would be incredibly cruel to see her snuffed out so late in the game.

Then again, showrunners David Benioff and D.B. Weiss are known to be mercurial in their (sometimes literal) axing of characters.

For the complete ranking of season 8 death predictions, check out the Technical University of Munich’s dedicated website. It might help you prepare for the worst deaths yet to come. App background

Check out the free App!

Get tech updates and breaking news on the go with the App, available in the Apple and Google Play app stores.

Get it today