GoT Algorithm Has Updated Its Prediction For The Ultimate Survivor

We are just days away from the final season of Game of Thrones, and the predictions just keep rolling in.

Using artificial intelligence, students at the Technical University of Munich (TUM) have now put in their own two cents.

During a recent seminar, these budding computer scientists created an algorithm that predicts which characters will survive and which one will most likely sit upon the Iron Throne.

Scouring the web for data on the hit television series and its 2,000-some characters, the algorithm's website is a minefield for any avid fan.

Be warned, potential spoilers (and spoilers of previous seasons) ahead!

Ready? Okay, let's go.

The character with the most grim prospect is Bronn, the skilled and dangerous sellsword of Tyrion Lannister, though Gregor Clegane and Sansa Stark are not far behind.

With a 99.1 percent chance of surviving, Daenerys Targaryen is clear and away the ultimate victor. According to the results of the algorithm, being a noble and, what's more, a female who is married, are all factors that work in the dragon queen's favour.

The analysis is so extensive, in all of these categories there are comparisons to the books, including each and every character's age, whether they are dead or alive and how long they are likely to live.

It might sound far-fetched, but this sort of thing has already proven successful. Before the airing of Season six, the same seminar produced an algorithm that accurately predicted Jon Snow's resurrection.

"While the task of predicting survival chances for Game of Thrones characters relies on data taken from the world of fantasy, the exact same artificial intelligence techniques are used in the real world and are having a powerful impact on our everyday lives," explains the lead mentor of the class, Guy Yachdav.

“虽然预测《权力的游戏》个角色的幸存几率依赖于虚构世界中的数据，但这种完全相同的人工智能技术可以运用在现实世界中，并且会对我们的日常生活有很大影响。”团队的导师Guy Yachdav解释道。

In other words, this isn't just a practice in fanaticism. Survival rates for cancer patients and other medical treatments use similar longevity analyses to the GOT predictor.

In a few more weeks, we'll know how accurate it really is. And we're not prepared.