A personal data visualisation project built around my own chess games, played on chess.com between January and May 2026.
Instead of looking outward for improvement, more tutorials, more theory, I decided to turn the lens on myself and ask: what does my data actually say about how I play?
To run the notebook, install the required packages by uncommenting and executing the first cell:
!pip install pandas numpy matplotlib plotly kaleido requests
!plotly_get_chrome
Plotly renders are set to "colab" throughout. If you are running the notebook in a local Jupyter environment, replace pio.renderers.default = "colab" with pio.renderers.default = "svg" in each cell.
ChessQuest.ipynb
The main Jupyter Notebook. Contains the full analysis, from data gathering and cleaning to visualisation and interpretation. Structured around two core questions: an exploratory overview of my general statistics, and a knowledge discovery investigation into whether losing streaks affect my accuracy.
chess_games.csv
The cleaned dataset exported from the chess.com API. Contains 166 games spanning January–May 2026, with attributes including game outcome, player ratings, accuracy scores, opening, colour played, and temporal metadata.
visualisation.jpg
A summary visualisation of the project's key findings, designed in Canva. Intended as a standalone visual overview of the most actionable insights from the analysis.
Data was retrieved via the chess.com Public API using Python. Visualisations in the notebook were coded with the assistance of Claude (Anthropic). All written analysis and interpretations are my own. The summary visual was designed independently in Canva.