FIFA - Football World Cup entertainment analysis

This project analyses historical FIFA World Cup data to explore how future tournaments can be designed to maximize entertainment. It features a reproducible data pipeline for collection, cleaning, and feature engineering across match and event-level data. The analysis applies statistical modeling and exploratory analytics to uncover drivers of excitement — including scoring patterns and competitive balance.

Dataset Content

The dataset includes key features such as team names, final placements, goals for and against, wins, and total points. The dataset was gathered from Kaggle.

Business Requirements

• Give an understanding of how points, wins, draws and losses affect the competitiveness of a team
• Provide insights to how a strong defense and a strong attack compare to each other
• Be able to show if there is correlation between placing well in a World Cup and if that affacts the placing in the following one

Hypothesis and how to validate?

• Hypothesis 1: The average number of goals per team per World Cup has increased over time.
• Hypothesis 2: The distribution of points per team has become more balanced over time.
• Hypothesis 3: Teams that performed well in one World Cup (top 4) tend to perform well in the next.
• Hypothesis 4: Teams with higher goal differences (goals for - goals against) achieve more points.
• Hypothesis 5: A strong defense (fewer goals_against) contributes more to success than a strong offense (goals for).
• Hypothesis 6: Teams that draw fewer matches tend to earn more total points.

Project Plan

• Data collection from Kaggle.
• Data unified into a single dataset.
• Data cleaning and preprocessing in Jupyter notebooks.
• Exploratory analysis and feature engineering.
• Machine Learning .
• Dashboard development in Streamlit.
• Iterative testing and refinement based on feedback.

The rationale to map the business requirements to the Data Visualisations

• Average goal per team over the different World Cup tournaments.
• Points distribution of teams over the different World Cup tournaments.
• Collective points distribution over the different World Cup tournaments.
• Top 4 retention rate
• Goal difference vs Points
• Defense vs Offense importance
• Point distribution by draw count

Analysis techniques used

• Used generative AI tools (Github Copilot, ChatGPT and Claude) for code suggestions, design thinking and correction
• Matplotlib, Seaborn and Plotly plots used to visualise data
• Descriptive analysis used to summarise and explain data

Ethical considerations

• Ensure the dataset was publicly available or properly licensed.
• Be cautious not to frame smaller or lower-ranked teams as “less entertaining” purely due to lower performance metrics.
• Recognize that entertainment value is subjective and may reflect cultural or regional biases.
• Clearly communicate that statistical patterns indicate correlation, not causation.
• Consider that maximizing “entertainment” should not come at the cost of player well-being, match fairness, or accessibility for fans.

Dashboard Design

Home page includes dataset and summary statistics

Unfixed Bugs

No significant unfixed bug

Development Roadmap

• Data was separated in different CSV files, these files were put together in order to work efficiently
• Data types were changed to have a clear visualisation
• Future improvements include working on a better dashboard result

Main Data Analysis libraries

• pandas
• numpy
• plotly
• streamlit

Credits

Content

• Kaggle
• Code Institute LMS

Acknowledgements

Thanks to the Code Institute instructors and peers for feedback and support