Soccer is among the most difficult sports to quantify. In baseball or basketball, every play is a well-defined event, whereas here the game flows seamlessly, the points are few and the recordable plays even fewer. Analyst Jake Wallerius describes it as a low-scoring, fluid sport with few recordable plays , thus representing the "final frontier" of sports analytics.
The numerical comparison is eloquent: in an MLB season 143 players reached at least 500 innings, in the Premier League only 7 attempted at least 100 shots and just 2 exceeded 50 shots on goal.
With so few events, more refined indicators are needed to explain performance. The process starts with shots and ends with expected goals , while a parallel discipline, performance analysis , brings the numbers into the daily work of the staff.
From Total Shots to PDO: The First Team Metrics
The first studies focused on shots. Aggregating data from many matches revealed that the best teams take more shots, especially on target, and concede fewer to their opponents. The difference between shots for and against correlates strongly with points in the standings at the end of the season.
Two simple measures describe it: the Total Shots Ratio (TSR), the share of one team's shots out of the total shots of both teams, and the Shots on Target Ratio (SoTR), calculated in the same way but only on shots on target.
These measures are derived from indices used in hockey and have predictive value. The TSR remains fairly stable from one season to the next, a sign of intrinsic quality. The PDO , the sum of the conversion rate and the save rate, hovers around 1000 but fluctuates widely and is used as an indicator of short-term luck . A very high PDO indicates a team that scores and saves more than normal, with a likely regression; a very low PDO signals bad luck and potential improvement.
Expected goals: distinguishing the result from the performance
Not all shots are equal. Twenty shots can be misleading if fifteen are from 30 meters, while a few shots from the six-yard box are much more dangerous. Expected Goals (xG) assigns each shot a value from 0 to 1, the probability of it resulting in a goal, based on distance, angle, game situation, body part, and presence of defenders. Adding them together gives the number of goals a team would have "deserved." The same applies to xGA , which refers to chances conceded, and the difference between xG minus xGA (xGD) is a more reliable indicator than simply goals scored and conceded.
Comparing actual and expected goals helps determine whether a team is overperforming or underperforming. With 20 goals scored against 15 xG after 10 matches, the team likely benefited from exceptional aim or less-than-perfect opposition goalkeeping. With 5 goals and 10 xG, they've underperformed. Those who score significantly above their xG risk slowing down, while those who waste a lot can quickly unblock themselves: these are the so-called regressions to the mean . Other metrics complete the picture: Opta's big chances, shot conversions, set-piece goals, clean sheets, blocks, and line saves. Even the xG per shot conceded is useful: in Serie B that year, Cesena had the worst value in the league, a sign that they were giving their opponents too many easy shots.
Match analysis, performance analysis and scouting
Three terms often used interchangeably refer to different activities. Match analysis systematically observes a single match (passes, shots, tactical movements). Performance analysis, on the other hand, follows an athlete or team over multiple matches or an entire season. Scouting looks outward, studying opponents or potential new signings. If a team loses possession, especially on the left flank, in the second half, a comparison with the last ten matches reveals whether it's an isolated incident or a trend.
Indicators must be chosen carefully. The distance covered by a midfielder says little without considering the phase of play, role, and intensity of his sprints. Therefore, a performance model is constructed , a coherent set of indicators related to sport, role, and objective. In swimming, for example, stroke rate makes sense when interpreted in terms of speed and strokes per length. Aspects that remain difficult to measure, such as reading the game or managing stress, require the expert eye of a coach.
Technologies, privacy and use in training
Video is the most common tool for tactical analysis. GPS and wearable devices, such as heart rate monitors and inertial sensors, record distance, speed, and acceleration, while optical tracking automatically tracks multiple athletes on the field. Different sensors can provide slightly different measurements for the same movement, so comparing data from different technologies is tricky. Physiological and movement data are also sensitive information: collection and storage must comply with regulations, as indicated by the Italian Data Protection Authority.
A report full of numbers unrelated to training remains unused. The analyst interacts with coaches, trainers, and athletes: if a player has accumulated a lot of sprints in recent games, the staff can reduce the high-intensity workload the following week. The role requires knowledge of the sport, basic statistics, proficiency with video analysis software, and communication skills. Opportunities include professional clubs, federations, applied research centers, and individual athlete consultancy; a program like the Master's Degree in Sports Science and Technology provides the theoretical and methodological foundation.
