We are now at the stage in training where we’re beginning to think about placement preferences, the industries we might want to explore, the tools we want to build on, and the direction we see ourselves heading in after The Data School. Thinking beyond placements can feel early, but it naturally prompted me to reflect on my longer-term career interests.
For me, one area keeps standing out is sport, and specifically football. It is something I have always been passionate about, and the idea of combining that with the data skills I’m developing is genuinely exciting. After speaking with people at TIL, the consistent advice I received was to start building personal projects and publishing them on platforms like Tableau Public and GitHub. This would allow me to showcase what I’m learning, develop a portfolio, and begin establishing myself in the football data space.
With that in mind, I decided to begin my first passion project.
My coach Serena recommended a website called SportVizSunday, which hosts datasets from various sports. I found a Premier League dataset focusing on player data from the 2021/22 season and immediately knew it would be a strong foundation for my first project.
First Steps
I began by exploring the dataset to understand what each field represented and what each row of data corresponded to. I created a rough data dictionary to organise my understanding. It became clear very quickly that the dataset contained a significant number of attacking metrics such as shots, goals, assists, xG, xA, etc. so because of this, I decided that the focus of my project would be attacking play.
Defining the User Story
Before deciding what to build, I needed to establish who the dashboard would be designed for. Potential users could include football analysts, recruitment teams, journalists, or general fans. Each group would approach the data differently, so choosing one audience was important for shaping the project clearly.
I chose to focus on football fans. Many fans judge attacking strength mainly through goals scored, but modern football data allows us to develop a far more accurate understanding of attacking performance. Teams can create a lot without finishing well, or finish well despite creating little. There are also important contributions that happen much earlier in the attacking sequence. My aim is to help fans develop a more complete and fair view of what makes a strong attack.
This led to the following user story:
As a… football fan,
I need to… understand what makes a team’s attack strong or weak,
so that I can… compare teams fairly using the full picture of their attacking contribution,
Through analysing… both total and individual metrics across finishing, chance creation, and build-up.
Breaking Down Attacking Play
Once the user story was clear, I reviewed the dataset again to group related fields and identify the main components of attacking play. This helped me structure the project into three key areas:
Insight 1: Finishing
To understand the strength of a team’s attack, the user needs to see how effective and dangerous the final product is.
Insight 2: Chance Creation
To understand how well a team creates opportunities, the user needs to see how consistently the team produces real, high-quality chances.
Insight 3: Build-Up and Involvement
To understand how a team constructs attacks, the user needs to see how involved the team is in earlier phases of the attacking sequence.
These insights form the foundation of my analysis.
Dashboard Structure
The project will consist of four dashboards:
- Overview dashboard - a summary of attacking strength
- Finishing dashboard
- Chance Creation dashboard
- Build-Up dashboard
The aim is to allow the user to navigate between them and explore different aspects of attacking performance in more depth. I have created sketches for each view to guide the layout and structure before I begin building them.
Overview dashboard:
Finishing dashboard:
Chance Creation dashboard:
Build-Up dashboard:
Current Stage
At this point, I am reviewing the dataset again to identify any issues and ensure everything is clean and consistent before moving into Tableau. Once the data is ready, I will start by building the overview dashboard and then move on to the three individual sections.
This is the first project in what I hope will become a growing football analytics portfolio. It already feels rewarding to work on something that aligns so closely with my interests, and I’m looking forward to developing it further and seeing where it leads.