Game of Thrones — Data Story

An interactive data analysis project exploring how Game of Thrones captured and then lost its audience across 8 seasons and 73 episodes.

What this project covers

• Ratings analysis — IMDb episode ratings mapped against key plot events to show exactly which moments triggered audience reactions
• Character network graph — interactive visualization of house allegiances, relationships, and how the power structure shifted across seasons
• Survival analysis — screen time vs survival probability modeling, with a logistic regression predicting character deaths

Live app

https://got-data-analysis-agoqm2ftftjezbqy2qiwv5.streamlit.app/

Tech stack

Python · Pandas · Plotly · NetworkX · PyVis · Scikit-learn · Streamlit

Data sources

All datasets sourced from Kaggle (Game of Thrones open datasets + IMDb ratings)

Key finding

The data shows that fan-favourite characters stopped dying at the exact point ratings began to decline — quantifiable evidence of the show trading narrative risk for audience safety in its final seasons.