Analysis of Real World Football Performance on Sorare NFT Cards

Introduction

During my third semester in the Masters program, my teammate and I embarked on a project for our "Databases" course. We chose to analyze Sorare NFT cards, focusing on how real-world football performance data correlates with trading behavior on these digital collectibles.

Tech Stack

• Data Analysis: Python, for data analysis.
• Database Management: PostgreSQL, utilized for managing and querying the collected data efficiently.

Project Insights

The project involved collecting data via APIs from Sorare and various football statistics sources. We then modeled this data using an ER diagram and integrated it into a PostgreSQL database. A significant challenge was matching the NFT cards to real-time player statistics. We conducted analyses like daily trading volume, highest card sales, and trading volume by league and position, using descriptive statistics and data visualization. The process included several matching methods to align player data from the APIs with the Sorare cards, utilizing strategies like exact matching and comparing with Levenshtein distance. The project highlighted trends and correlations in the trading behavior of Sorare NFT cards relative to real-world football performances.

Outcome

The project was highly successful, earning top grades and leading to an invitation for me to write a Master's thesis in the Computer Science faculty, despite my primary association with the Economics faculty. The detailed analyses, Python scripts, presentation, and final paper can be found in my Github repository.