What is it?
Data scientists use statistics, programming and curiosity to answer questions with data. They clean messy datasets, look for patterns, build models that make predictions, and explain what they found so others can act on it.
In football, that could be finding players with a similar style to a target signing, predicting which training loads raise injury risk, or measuring which tactics actually create chances.
What you'd actually do
Questions
Work with coaches, scouts or product teams to turn a vague question into something data can answer.
Preparation
Gather data from different sources, fix errors and gaps, and shape it into something usable.
Analysis
Explore the data, test ideas with statistics and build models that predict or group things.
Communication
Turn results into clear charts, dashboards and stories that non-experts can trust and use.
What to learn
Programming
Python or R for analysis, plus SQL for getting data out of databases.
Statistics
Probability, distributions, testing ideas properly and knowing when a result is just luck.
Machine learning
Regression, classification, clustering and how to tell if a model is any good.
Communication
Data visualisation and the ability to explain findings simply.
Your roadmap
A step-by-step route into football analytics. Work down the line, one topic at a time, from querying data to building with AI.
- Databases
- SQL query basics
- Grouping functions
- Multi-table queries
- Analytical functions
- Subqueries
- CASE clause
- Create, update, delete tables
- Window functions
SQL
- Python basics
- Lists, tuples, dictionaries
- Functions and variable scope
- Control flow
- Modules and packages
- Object-oriented programming
- Exception handling
- NumPy
- Pandas
- Matplotlib
- Seaborn and Plotly
Python
- Data ingestion (different formats)
- Data modelling
- Creating visualisations
- Filters, parameters
- Dashboards
- Storytelling with data
- Connecting to a database
Visualisation
- Data cleaning and exploration
- Feature engineering
- Feature selection
- Dealing with imbalanced data
- Model building & tuning
- Model deployment
Data Science & Machine Learning
- LLM models and tokens
- Prompt engineering
- Vibe coding
- Claude Code
- Planning and context files
- Skills, custom commands
- Sub-agents
- Model Context Protocol (MCP)
GenAI
Tools you'll use
Ways in
University
Degrees in maths, statistics, computer science, economics or sports science with a strong data focus.
Apprenticeship
Data analyst and data science apprenticeships let you learn on the job.
Self-taught
Online courses and public football datasets make it possible to build a convincing portfolio.
Try it yourself
- Analyse a season of results and find which stats best predict winning.
- Build a player similarity tool that finds look-alike players from their stats.
- Create a dashboard comparing teams' attacking and defensive styles.