Stay up to date with notifications from The Independent

Notifications can be managed in browser preferences.

Game of Thrones season 8: Algorithm predicts who will die and who will survive

It does not look good for Bronn

Jack Shepherd
Saturday 13 April 2019 14:57 BST
Comments
Game of Thrones Season 8 trailer - 'Aftermath'

With Game of Thrones season eight fast approaching, fans have been feverishly speculating about the forthcoming final six episodes.

One of the biggest points of contention concerns who will survive until the bitter end. Will Jon Snow survive the Night King's impending attack? Can Arya keep on killing and not be killed?

A group of computer science students in Germany have developed an algorithm to predict answers to those questions.

Speaking to CBS, project supervisor Guy Yachdav said characters survival rates are predicted using longevity analysis similar to what scientists use to examine the effects of medical treatments.

Although the analysis "relies on data taken from the world of fantasy," he said, "the exact same artificial intelligence techniques are used in the real world."

The algorithm found that Daenerys Targaryen was the most likely to survive, with a less than 1 per cent chance of dying. Second is Tyrion Lannister (2.5 per cent chance of dying), followed by Varys (3.2 per cent), Samwell Tarly (3.3 per cent), and Jaime Lannister (4 per cent).

On the opposite end of the spectrum is Bronn, whose chance of dying is apparently 93.5 per cent. In second is Gregor Clegane (80.3 per cent), followed by Sansa Stark (73.3 per cent), Bran Stark (57.8 per cent), and Sandor Clegane (47.5 per cent).

Game of Thrones returns to TV screens 14 April. The show will be available on Sky Atlantic and NOW TV in the UK. Read everything you need to know about the forthcoming season here.

Join our commenting forum

Join thought-provoking conversations, follow other Independent readers and see their replies

Comments

Thank you for registering

Please refresh the page or navigate to another page on the site to be automatically logged inPlease refresh your browser to be logged in