Visualizing Game of Thrones


This is my second D3 project done with my fellow classmates: Josselyn and Allie. Our data is from the Game of Thrones wikia, which is based off of the television series on HBO (not the books), and is updated to where the series is in real time. (So a character who maybe eventually dies in the books but is still alive thus far in the television series will have a status of “alive.”) This also means that some characters are different (Talisa Stark does not exist in the books).

That being said, in addition to organizing lots of character data, our visualization answers interesting questions about the television series such as: How many major characters have died thus far? Is there a large gender discrepancy in Game of Thrones? Which houses have the most major characters?

The results are pretty much what we expect from Game of Thrones. That is, nobody is safe, since the deceased includes a huge number of important characters, and that while there are twice as many male characters as female characters (a characteristic of the time period of Game of Thrones), there is still a huge number of strong female characters. Despite being outnumbered two to one, many of the top/most occurring characters are female.

Lastly, we included some text around the graph and in the two widgets to balance the visualization aesthetically and provide some light hearted fun facts! And you can see our visualization live here: (Spoiler alert if you are not caught up)

Data Collection

Using the B​eautifulSoup ​package, a library to pull data out from HTML files, we wrote Python script to collect the major characters of Game of Thrones data from its w​ikia​web pages into a csv file. In the end, we collected total of 37 major characters and 8 additional characters that are quite important in the story. The data includes the character name, allegiance, alive or dead status (death is one of the most important events in Game of Thrones), causes of death, number of appearance, origin, religion, culture and portrayal.

Data Cleaning

We then added the gender, link to their profile image and manually cleaned the raw data as below:

●  Removed the comma separated number for season data

●  Removed the “(see below)” hyperlink for appearance data

●  Removed “house” words for allegiance data

We initially thought putting age, religion and culture of the characters as the main nodes in visualization will be great, but not all characters are provided with this data on wikia. We also thought putting the killer of the characters will be interesting, but we decided not to since it will make the edges line more complicated among characters.


This project was created by me, Josselyn and Allie for a Data Visualization course at Cornell taught by David Mimno. Several open-source projects were used under the MIT License including D3, jQuery, RequireJS, and BeautifulSoup. Data courtesy of the Wikia.

Much of the inspiration for the visualization comes from Violence and guns in best-selling video games visualization appeared in The Guardian News.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s