Football managers, experts and fans around the world have always sought a clear explanation for the final scores in football matches. In some games, this can be down to pure luck, but TSZ suggests that the vast majority of matches can be explained by common denominators. The advanced access to performance indicators such as ball possession, shots, shots on target and number of penalties (amongst many more) have made it easier to form conclusive opinions about results.
Possibly the most important statistic is ball possession. The focus on using ball possession as a style of play is widely attributed to Johan Cruyff, Barcelona manager in the early 1990s. He had the idea that if his team kept possession and created movement, working the ball around the pitch, they would generate more chances and consequently this would lead to success. The technique was given the name tiki-taka. Later, Peo Guardiola took this concept to a new level throughout his management of La Liga club Barcelona between 2008 and 2012.
Last year’s Premier League winners Leicester City put a dent in Cruyff’s theory after they remarkably won the league with an average possession throughout the whole Premier League season of 44.8%. With this in mind, TSZ has decided to investigate ball possession in European football to see if it has a clear correlation with some of the other most common statistics in football relating to success.
Before we get started, there are a few considerations to point out: the data set includes data from the 2015/16 season for the top five European football leagues (English Premier League, French Ligue 1, German Bundesliga, Italian Serie A and Spanish Primera Division). The data is limited to only include games played in each respective domestic league and excludes European and other cup competitions. We also want to point out that we are aware that Bundesliga only consists of 34 games per team compared to 38 games for all other four leagues. This might have some impact on the first part of the analysis but is not relevant in the subsequent conclusions.
To get into the analysis we will start by looking at ball possession and compare this with the number of penalties for each respective team. This will be visualised through a combination of a line and bar graph (one per league). The line represents average ball possession per team and the bars represent the total number of penalties awarded. The green bars represent the league winner, the red bars relegated teams, and the remaining clubs (those who did not win, but also were not relegated) are illustrated with a yellow/gold bar. The second part of the analysis looks at ball possession and compares this to pass success rate, shots per game, shots on target (per game) and average points per game. This will be illustrated in 4 scatter plots where the colour of the winner, relegated sides and all other teams follow the same structure as in the line/bar graphs. The last scatter plot, which looks at ball possession vs. average points per game, also displays a 95% confidence interval for each of the respective colours (winners, relegated teams and other teams). This helps us see if there is a correlation between a high percentage ball possession and actually winning football games.
Looking at the graph below for the Premier League, we can see that the winners (Leicester) had the most penalties awarded at 13, significantly more than the other teams in the league, despite having a low average ball possession at 44.8%. Furthermore, it is remarkable that the three relegated teams combined (Newcastle, Aston Villa and Norwich) have an aggregate number of just six penalties. The line graph within the plot does not demonstrate any clear pattern. Therefore, we can conclude that throughout the last season in the Premier League, there was not a strong correlation between high ball possession and a high number of penalties awarded.
Moving on to Ligue 1 in the next graph, we can see that again the winners (PSG) received the most penalties at nine. Notably, this is exactly the same number of penalties as the relegated teams (Ajaccio, Reims and Troyes) combined. The line graph follows the same noisy pattern as the Premier League with no clear relationship between a high ball possession percentage and number of penalties awarded.
The next graph, which looks at Bundesliga, shows that Ingolstadt had more penalties awarded than the league winner Bayern Munich (and any other team) at ten vs. nine. This is interesting as Ingolstadt actually finished 11th. This contrasts to the two previous graphs where there was a positive correlation between teams finishing high in the league and the number of penalties awarded. The line graph in this plot seems to have a slight decay but it is very hard to say that this is down to extreme ball possession values (high and low) for Bayern Munich, Dortmund and Darmstadt.
Looking at the Serie A graph below, we can instantly tell that it is very different from the other two graphs. There is little correlation between number of penalties and finishing position. Most notable is that the relegated Carpi FC and the league winner, Juventus, ended up with the same number of total penalties at nine. This is two less than Lazio and Verona who top the chart with 11 penalties each. Verona also shows a remarkable seven penalties, which is the second highest in the dataset for a relegated side. Since the number of penalties seem to be rather randomly distributed for Serie A, it comes as no surprise as the line graph over ball possession is fluctuating up and down without any clear trend.
Moving onto the last graph in the first part of the analysis, we can see that Barcelona were awarded significantly more penalties than any other team in the Primera Division at 19. This is six penalties more than Leicester, who are the closest chasing team in the dataset. All three of the relegated teams sit towards the right hand side of the graph, representing very few penalties awarded. Atletico de Madrid – who finished third and were chasing Barcelona to the bitter end – only received a total of two penalties. When looking at their average ball possession at 47.7%, this accomplishment is even more incredible and could be likened to Leicester’s achievement in the Premier League.
The final take from these five plots is that ball possession doesn’t seem to be very strongly correlated with the number of penalties awarded. Having said that, the number of penalties awarded has been shown, to a great extent, to be linked to the winning teams in the league.
To investigate this further, looking at the table below we can see that on average the league winners receive 11.6 penalties, which is 6.9 (11.6-4.7) penalties more than “Not League Winner or Relegated” teams and 8.2 (11.6-3.4) penalties more than relegated teams at 3.4. From our data we can therefore confirm that the success of winning the league seems to be strongly linked with the number of penalties awarded.
Getting back to the focus of the article, the left graph below looks at average possession vs. pass success rate. Here we can see that Barcelona, PSG and Bayern Munich have the highest proportion of average possession, ranging between 63% and 66%. Their pass success rate is also towards the top in the league, closely competing with Fiorentina with a succession rate between 86% and 89%. Juventus (who also won the league) do not sit far behind in terms of pass succession rate but have a slightly lower ball possession figure of around 55%. Further down the graph, we find Leicester with a remarkably low ball possession as previously mentioned. Their pass succession rate is also only 71%, giving them the seventh lowest rate in the field, even behind all but two (Carpi & Frosinone) of the relegated sides. One other interesting team to note this graph is Darmstadt, who had only 37% average ball possession and a pass success rate at 56%, but yet still managed to finish 14th in the Bundesliga. Moving on to the right graph, which looks at average possession vs. shots per game, we can see that Real Madrid have the highest number of shots per game at 18.7, with Bayern Munich following closely behind at 18.4. In general, it seems like the teams who finish towards the top of the league shoot more. On the flip side, those who are not performing as well, with risk of relegation (on the lower left hand side of the graph) shoot less. This essentially proves Cruyff’s theory as teams with a high ball possession tend to shoot more often, and in the long run this leads to greater success (finishing higher up in the league).
Moving onto the last graphs, on the left hand side we look at average possession vs shots on target per game. Here we can see that Barcelona, PSG and Bayern Munich stand out with a distinct position in the top right corner. Their average shots on target per game ranges between 6.6 and 7.6. Real Madrid tops this graph with an average of 7.6 shots on target per game, which we suggest might be brought up by Ronaldo’s prowess. This is something we looked into in an earlier piece: Who Is The Best Goalscorer In European Football? As with the previous graphs, Leicester is sitting down towards the bottom left, looking like an exception to the rule vs. all other league winners. Relegated teams also sit towards the bottom left corner, which is in line with the previous shots per game plot. Moving on to the right hand graph, which shows possession vs. points per game, we will first consider the league winners. The green trend line, with 95% confidence interval, shows an upward trend. Notably Leicester, who had the lowest average ball possession of all league winners, also have the lowest average points per game at 2.1. In contrast, Bayern Munich, who have the highest ball possession in the study, also have the highest average points per game at 2.6.
Looking at the yellow/gold lines (showing teams who were not league winners but also not relegated) the trend of ball possession and points per game is even stronger, with a steeper line graph. This is due to the fact that the range of final points in the league varies much more within this group than those who win / are relegated.
When we consider the relegated side (red lines), we can see the upward trend is significantly less steep than for the other two groups. Even though the trend is much lower within this group, it is still evident that teams with higher ball possession tend to take more points.
To conclude, the number of penalties that each team is awarded doesn’t seem to have a strong correlation with average ball possession. However, it is clear that the number of penalties a team receives does seem to have a significant link to how the team performs in the league. When we drilled into ball possession we saw that in most cases, the teams winning the league also had a better pass success rate, more shots (on and off target) and as a result of this won more football games (higher average points per game). With this in mind, it seems like Johan Cruyff and Pep Guardiola’s match technique and focus has been (and still is) a good recipe for success. This also makes Leicester’s achievement even more of a fairy tale and an exception to the rule. But in all fairness, who doesn’t prefer a fairy tale?