How hamdanovic98/moneyball-for-football-a-data-science-approach-for-recruiting-players. works
Project Type and Purpose
• This is a data science pipeline aimed at revolutionizing football player recruitment by gathering, processing, and analyzing performance data across multiple football seasons (2019–2024).
• The project collects raw data (from FBRef via dedicated notebooks), cleans and prepares that data, and then uses it (mostly in a “Players’ Replacement” module) to develop insights or suggest alternative player options.
Key Components
a. Data Collection Module
• Represents the “Data Collection” directory where raw CSV files for each season (categorized by year) along with the FBRef scraping notebooks reside.
• That this module interacts with external data sources (like FBRef).
b. Data Preparation Module
• Represents the “Data Preparation” directory. This module contains notebooks that process, clean, and combine raw CSV files to generate “Data Prepared” files.
• That outputs from the Data Collection module feed into this processing step.
c. Players’ Replacement Module
• Represents the “Players' Replacement” directory. Here, individual notebooks (organized per player and by categories like goalkeepers, midfielders etc.) handle data analysis for determining potential player replacements.
• Demonstrate that this module consumes the processed data from the Preparation module.
d. Auxiliary Resources
• A potential “Configuration/Metadata” component could be added to represent the integration of different years of data though it is primarily represented by the folder organization.
Relationships and Interactions
• The Data Collection module (raw data ingestion and scrapping) provides input for the Data Preparation module.
• The Data Preparation module cleans, integrates, and enhances the data and then supplies it to the Players’ Replacement module for analysis.
Architectural Patterns & Technologies
• This architecture follows a sequential data pipeline pattern – from data ingestion (collection) to processing (preparation) to analysis (players’ replacement).
• Though not explicit in the file tree, you might indicate that data analysis libraries (e.g., pandas, NumPy, visualization libraries) are being leveraged.