Two Steps Forward, One Back - Publicly Available Data in Football

Last week, OPTA and FBRef’s agreement was terminated, removing free access to some of the most valuable and advanced football data publicly available. This data was of value not only to myself, but to the wider football analytics community — including those employed by clubs, those not employed by clubs but working with data out of passion, and students using the data for school projects or to build personal portfolios to develop and demonstrate their skill sets.

While there is still a lot of publicly available football data, the discussion, engagement, and overall “plus” that came with having access to current-season advanced data is now more or less gone or significantly reduced. Football is constantly evolving, with new tactics, trends, and prospects emerging every season, and losing the ability to analyze these developments in real time is a major setback for casual and aspiring analysts.

I am grateful that my analytics journey began at a time when access to free football data was arguably at its peak. In many ways, this makes the loss even more frustrating, especially in a sport and profession that already lags behind others in terms of analytical openness and accessibility. FBRef’s access to advanced data, and its mission to make that data publicly available — first through StatsBomb and later through Opta just before the 2022 World Cup — played a major role in my development.

That access allowed me to develop my data analysis and coding skills in R, largely self-taught, with support from individuals across the football analytics community, particularly on Twitter/X. Through this, I was able to build live Team and Player dashboards for the 2022 World Cup using Tableau, and eventually develop my own ShinyR applications focused on team and player performance. FBRef’s publicly shared data was also a major and key part of my Graduate Thesis which I developed and analyze the impacts and abilities of Head Coaches around the world. I in fact had plans to update and enhance the thesis study, but that is on hold.

Those dashboards will remain available on my personal website (tinyurl.com/stevendabdoub), but they will not be updated until I am able to establish a clear way forward.

In the interim, for any posts or blog content I produce going forward, I plan to rely on multiple alternative data sources. While this approach provides some coverage, it also means having less control over how data can be used, interpreted, and analyzed across different methodologies and perspectives of the game. Some of the platforms I plan to use include:

• The Analyst Stats Hub • WhoScored • Understat • FotMob • SofaScore • MarkStats • DataMB • One Versus One • Ben Griffis’ Football Match Reports and Best XI App

Granted, none of these sources offer the same combination of centralization, ease of use, depth, coverage, or level of customization that Opta data via FBRef provided. As I work toward a solution to hopefully revive my performance dashboard Shiny apps — with improved features — and as I prepare to do previews, analysis, and coverage around the World Cup this summer, much of my focus will lie on these platforms for data and insights.

Beyond my own work, this situation represents a significant blow to the football analytics community as a whole. FBRef served as one of the few true bridges between elite, proprietary data and the public domain. Its removal effectively gridlocks and further setbacks the analytics community in a football ecosystem where we have made a lot of progress on but in some areas are still behind and frowned upon or not appreciated.

Publicly available data has been a key driver in the growth of football analytics, encouraging passion, discussion, collaboration, and innovation.

For aspiring analysts, I am always open to conversation and have no issue helping guide you toward resources and data sources that can support your development. I encourage you to explore the platforms mentioned above and feel free to reach out with any questions or requests.