/data-visualization-project-2021-rookies

Data Vizualisation of a Lichess.org games dataset, all viz are available on the Website (link in the README)

Primary LanguageJupyter Notebook

Chess Analysis

This project has been realized in the context of EPFL Data Visualization (COM-480) course.

Go to Website

Check out the screencast

Student's name SCIPER
Benhaim Julien 284558
Michelet Elisa 282651
Vignoud Julien 282142

Introduction

The dataset used throughout the project is a collection of chess games that occurred on the online chess platform called Lichess. Accounting around one million games, we took advantage of the dataset size to draw some tips and conclusions with a data-driven approach. We set our main objective towards providing an overview of the main aspects and strategies of chess for each player level. In other words, we aspire to give a general and visual understanding of the game that is tailored to users' familiarity with chess. For instance, beginners are more interested in basic tactics and common moves, intermediates with the most useful openings and advanced players with some examples of games from the best players in the dataset. The visualizations allow each player to explore and get insights from thousands of games, letting them take a new look at chess, use it as a tool to improve their future strategies or spark a new interest for the game.

Reports

You can find the process book as well as the different milestones below.

Process bookMilestone 1Milestone 2

Project architecture

├───img                 Images used in milestones
├───reports             Milestones and process book
├───notebooks           Notebooks used for preprocessing
├───assets              Images used in the website
└───docs                Website top folder
     ├───libraries      Libraries used in the project       
     ├───viz            Folder containing the 5 visualizations
     ├───styles.css     Root stylesheet
     └───index.html     Website homepage

Setup

The website has been created to be standalone, such that there is no need to install any libraries beforehand. To start contributing, clone the repository and start a server from index.html.