Before we get started, I need to make a little disclaimer: this post is meant to be fun and is not based on any predictive analytics, but rather simply educated guesses based on the data we have available.
Part of the draw of Game of Thrones (GoT) are the challenges to normal storytelling tropes employed by both George R R Martin and the TV show creators, David Benioff and D.B. Weiss.
You know, things like killing the main character in the first season, or killing a main character immediately after he had a huge, meaningful showdown with magical villains, proving to be the only one who could stop them. Alright yes, one of them came back from the dead, but even so, these deaths were still pretty shocking.
But these surprises are part of what makes the show so intriguing. Game of Thrones always keeps us on our toes, never allowing viewers to guess what will happen next... But that doesn’t mean we can’t try! Let's see if we can guess Game of Thrones’s next death.
In looking through the data in Looker, we noticed a few trends, which we have extrapolated to form predictions about what to expect in the final two seasons. As with any piece of creative work, data can only go so far, and while we do have data on six seasons, only one season is not based on books and entirely created by the current writers, as the next two will be... That said, it doesn’t hurt to take a step back and see what we can find in the data to help us #prepareforwinter.
So here we go: our data-based, totally unscientific predictions for Season 7 of Game of Thrones.
As we mentioned in our first post, the three most recent seasons’ story arcs all leave the deaths for the season finale. We are predicting that trend will continue with season 7 episode 7 being the deadliest episode in the season.
In the past few seasons, the mid-season episode has been surprisingly deadly. Since the new season only has 7 episodes, this puts Episode 4 in the mid-season spot.
The data on deaths in the previous seasons shows a clear trend of deadly season followed by not that deadly season. Since the last season was so deadly, we think they will put off the biggest kills for the final season of the show. This also makes logical sense as they will probably want to go out with a bloody bang.
In our first post about GoT character screen time, we mentioned that Eddison Tollett — Jon’s right hand man who he appoints as the Commander of the Night’s Watch when he leaves for Winterfell — has a surprisingly high episode count.
This cannot be a coincidence. There has to be a reason he is in so many episodes... But with everything coming his way (cough White Walkers) I’m sure there will be many opportunities to give him some storyline to go with his plentiful episode count.
Since writing our first post, we had the opportunity to put some more granular data into our model, including screen time by season. From this GoT line graph, a few surprising insights provide hints to what the show runners are leading up to now that the show and the book have completely separated stories.
Based on how the data is trending, the following are our guesses for what will happen to some well-known characters...
As you can see in the visualization above, the only two characters of the main character group whose screen time has increased consistently over the recent seasons are Jon and Sansa. We interpret this finding to mean that they are leading up to something with their storylines, and that this season their time on screen will continue to increase.
Another interesting insight from this screen-time-by-season data, is the mirrored screen time of Varys and Petyr Baelish (Littlefinger):
This is an interesting look at how these two schemers are being shown on screen as foils of one another. When one character gets a lot of screen time in a season, the other is working in the shadows... What does this mean for their respective plots for the future of the Iron Throne? Again, the data doesn’t tell us that, but based on this chart, we are predicting that...
As we mentioned in our first post, Varys is a prominent member of the group of characters with high episode counts but low overall screen time. This is a group made up of helpers and schemers - of which he is definitely the latter - and since his screen time has been slowly increasing over the past few seasons, we predict that this is the season when his big plan will finally come to light.
There are lots of theories around the internet about the fate of Petyr Baelish… But one thing that seems to be consistent is that this is our last season with him.
And we agree... But only if he spends a lot more time on screen.
This is based on an interesting insight from our new data set. The visualization below shows that most of the characters killed in the last season had positively trending screen time in their final season:
With Littlefinger’s screen time trending down (see the first viz above), it would be a big break from the trend for the show creators to kill him off without giving him a good amount of time on screen first.
So if we start seeing more of our conniving friend, we all should prepare for the worst... If he remains in the shadows, then we should expect him to make it to the final season.
Share your thoughts on this post as well as Part I and Part II by using the hashtag #dataofthrones and/or tagging @lookerdata. Who do you think will be the next death? We would love to hear what you think is coming over the next seven weeks.
We’ll be checking back in after the season is over to see how we did on our predictions… Enjoy watching this weekend! #winterishere Interested to see what you can learn about your data with Looker? Request a demo today!
Disclaimer: Game of Thrones belongs to HBO and is not affiliated with Looker in anyway.