Football or soccer for some in certain parts of this world, is a game that has deep roots in tradition. For the longest time, it was a game analyzed through the lens of goals, assists and possession. But those were simpler times, and as the game has evolved, so did the methods used to understand and evaluate player performance and team strategies. The rise of “advanced stats” in football is a journey of technological advancements, changing philosophies, and a modern obsession with optimization. Where sites like Sports Loci are now using advanced stats in mainstream ways and providing users access to such stats.
Early Days: The Birth of Data Collection
The birth of data collection in football can be traced back to mid-20th century. In this stage we are talking about a rudimentary form of data. Analysts would manually track basic metrics during games, recording actions such as passes, shots, and tackles. Charles Reep, a retired Royal Air Force officer, is often credited as one of the pioneers in football analytics. Reep in the 1950s, started to collect data to support his theories of long-ball strategies in the game, and whether or not it was the most efficient way to score goals. His findings are controversial and perhaps even no longer relevant in the modern game due to rule changes, but whether he knew it or not he laid the foundations for data collection in football.
Pre-Advanced Stats Era
In between then and the modern day advanced analytics that we see today and will discuss later, we have what I like to call the pre-advanced stats era of football. During this time, stats were far more advanced than when Reep took his paper and pencil and started jotting down stats. Stats became commonplace, everyone became familiar with stats as an integral part of the game, they were discussed in mainstream media and many stats became defacto ways of understanding the game. As time progressed more and more stats started to be used to further our analysis of the game. This era of stats remained the case for quite a while, until eventually we reach a point of renewed innovation.
The Opta Revolution
The late 1990s brought on the advent of companies like Opta among others, which began to systematically tracking in-game events. Detailed stats were starting to immerge and the depth to which we could understand the game began to change once again. Opta’s detailed data collection allowed analysts to quantify aspects of the game that were previously immeasurable, such as pass completion rates, defensive actions, and expected goals (xG).
At first these stats were used in more internal ways, even now there are certain depths of stats that are not necessarily considered mainstream, and a more so tools used by training staff to improve performances. But we will see as time progressed that some of these stats have started to filter their way to mainstream media and have started to become part of common place discussions.
Expected Goals (xG): A Paradigm Shift to Mainstream
Expected Goals (xG) is a metric that quantifies the quality of a scoring chance depending on where a shot was taken. xG emerged in the 2010s as one of the most revolutionary advanced stats. By assigning a probability to every shot based on factors like distance, angle, and type of assist, xG provides a more nuanced understanding of attacking performance.
xG is important to mention because it was the first “advanced stat” that became part of mainstream conversations. Whereas as we mentioned before, advanced stats were confined to training room conversations and optimizations. Little to say that the introduction of xG challenged conventional narratives and opened the eyes of many to a whole new world.
Tracking Data and Machine Learning
The 2010s also saw the rise of tracking data, which uses cameras and sensors to monitor player and ball movements at a granular level. Companies like StatsBomb, Catapult, and Second Spectrum have leveraged this data to create new metrics such as pressures, passing networks, and expected assists (xA).
Machine learning has further enhanced the analytical landscape. By analyzing vast datasets, algorithms can identify patterns and make predictions about player performance, tactical effectiveness, and injury risks. This technology has been embraced by elite clubs such as Liverpool, Manchester City, and Bayern Munich, who use data to gain a competitive edge.
Conclusion
The history of advanced stats in football is a testament to the sport’s evolution. From Charles Reep’s hand-collected data to today’s machine learning algorithms, the journey reflects humanity’s relentless pursuit of understanding and improvement. While traditionalists may yearn for the simplicity of yesteryears, the integration of advanced stats ensures that football remains dynamic, innovative, and endlessly fascinating.
