Playing under high pressure and ball retention

1. Introduction

Among its many uses, the game intelligence data produced by the French company SkillCorner can be deployed to analyse players' performances when they are subjected to different levels of pressure from their opponents. A player is considered to be under pressure when he is in possession of the ball and has at least one opponent close to him trying to recover it or limit his playing options.

For each of these situations, SkillCorner determines the intensity of the pressure by taking into account the speed of the players exerting it, their distance from the player in possession and the angle of their movement. In this report, we will focus on high-intensity pressure, i.e. pressure exerted with the greatest speed and at the greatest proximity to the opponent (more informationhere).

The ability to keep the ball is obviously linked to the player's individual talent (level of technique, orientation in space, perception of the game, etc.), but also to the style of play of the employer team (proximity of teammates, passing options, compactness, etc.). Regardless of his intrinsic qualities, a player under pressure and on his own will find it harder to retain possession than a colleague with more passing options available.

This report contextualises the SkillCorner data on pressure using statistical models that highlight the link between the teams' styles of play, the level of pressure faced by players and their ability to retain possession. The report is based on a sample of 7,050 matches played during the 2023 or 2023/24 seasons (up to 4 March 2024) in 28 leagues across Europe and the Americas.

2. Pressure frequency

During the 2023 and 2023/24 seasons, outfield players from teams in the 28 leagues studied experienced an average of 9.2 instances of high-intensity pressure per match. The highest values were measured for dominant teams, producing a game based on ball possession and collective pressure. Topping the list is Manchester City (15.2 pressures per game), followed by Bayer Leverkusen (13.9) and Columbus Crew (13.5).

As mentioned above, the number of instances of high-intensity pressure experienced by each team is dependent on the style of play used. Four style-related variables explain 66% of the differences observed between teams according to the regression model built for this purpose. The model shows that teams under high pressure play a short game, complete a lot of passes and dribbles (Wyscout data) and perform many high-speed sprints (>25 km/h for at least 0.7 seconds, SkillCorner data).

Within the same team, the pressure undergone by players varies considerably depending on the position they occupy. The values per position shown in Figure 4 refer to the averages and medians observed for 5,101 footballers who played in the same position for at least 1,000 minutes in the 28 leagues studied during the period in question.

On average, centre backs experience the fewest instances of high-intensity pressure per game (7.6). With an average value of 9.3, centre forwards are put under more pressure, but this figure remains low compared to other positions, mainly because they touch fewer balls than their teammates. Full backs face an average of 10.6 instances of high-intensity pressure per match, a lower figure than for the three most exposed positions: defensive midfielders (12.3), wingers (13.7) and attacking midfielders (15.6).

3. Ball retention

SkillCorner data can also be used to measure a player's ability to cope with pressure from opponents, particularly in terms of their ability to keep possession of the ball. Either the player is always in control of the ball, thanks to a successful dribble for example, or his team maintains possession thanks to a pass. The indicator of the ball retention under high pressure shows the proportion of possessions maintained by the team in relation to the number of instances of high-intensity pressure undergone.

Figure 5: correlation between the number of incidences of high-intensity pressure and the percentage of ball retention, by team

The ranking of the teams that kept the ball best under high pressure is fairly close to that of the teams that were subjected to the most high-intensity pressure instances, the two variables being highly correlated with a coefficient of determination of 0.5 (Figure 5). In both cases, the indicators reflect the style of play practised by clubs. Five teams in Europe's big-5 have the highest percentages of ball retention under high pressure, with a record of 84.2% for Manchester City, followed by Real Madrid, Paris St-Germain, Arsenal and Bayer Leverkusen (Figure 6).

The ability to keep the ball also depends on the position of the players on the pitch. The highest values were recorded for midfielders, both defensive (76.5%) and attacking (72.9%), as well as for centre backs (70.7%). Conversely, the lowest proportion was measured for centre forwards (66.8%), which is linked to the area covered and the conditions in which passes are received. For defensive positions, we note a greater heterogeneity in the values (Figure 8), which refers to the more or less offensive roles attributed to these players depending on the team, particularly at full/wing back level.

Figure 8: statistical distribution of ball retention percentage under high-intensity pressure, by position

4. Modelling at individual level

At player level, over and above the raw ball retention statistics, it is particularly useful to take into account the context in which the footballers play, both in terms of the style of play of the employing team, defined by the number of instances of high-intensity pressures experienced per match, and the position occupied. This allows for a better assessment and comparison of their performance.

The multivariate statistical model developed for this purpose, which includes 6,444 players active in the 28 leagues analysed during the season under review and who experienced at least 100 high-pressure actions while playing in the same position, explains almost 38% of the differences in values observed between players. All the variables considered are highly significant.

We applied this model to the 7,633 players in the leagues studied who had been subjected to at least 80 instances of high-intensity pressure, and calculated a theoretical value in proportion to the different positions occupied. This enabled us to obtain a residual value for each player, representing the deviation from the theoretical value. The higher this value, the best the player's ball retention ability given the pressure faced by his team and his position on the pitch.

Figures 10a to 10f show the most positive high-pressure ball retention residuals by position. This analysis notably highlights the Croatian Emir Dilaver (HNK Rijeka) among centre backs, Márcio Rafinha (São Paulo FC) among full or wing backs, Sergiy Rybalka (LNZ Cherkasy) among central or defensive midfielders, Caleb Zady Sery (SM Caen, transferred to FK Vojvodina) for attacking midfielders, Hugo Vallejo (Huesca SD) for wingers and Dominique Badji (FC Cincinnati, transferred to Bandirmaspor) for centre forwards.

5. Conclusion

The frequency with which players are exposed to high-intensity pressure depends very much on their team's style of play. The more teams play a short, elaborate game that favours possession, the more their footballers will be exposed to high-pressure situations from their opponents.

Similarly, there is a correlation between the average number of high-intensity pressure situations experienced by a team and the proportion of ball retention by players. This result reflects the importance of the role of the team as a whole in individual performances. Footballers in dominant teams are not only particularly gifted, but they also play in a context that is conducive to the full expression of their qualities.

The analysis of the residuals from the statistical model developed by taking into account the teams' style of play and the players' position in order to explain their ability to retain possession under high pressure highlights footballers who perform better than expected. This approach is particularly useful for scouting purposes to target recruits with an unexploited potential.