Borussia Dortmund vs Real Madrid: passing network analysis

Borussia Dortmund defeated Real Madrid 4-1.

Below is the passing network for the match. The positions of the players are loosely based on the formations played by the two teams, although some creative license is employed for clarity. It is important to note that these are fixed positions, which will not always be representative of where a player passed/received the ball. Only the starting eleven is shown as the substitutes had little impact in a passing sense.

Passing network for Bayern Munich and Barcelona from the Champions League match at the Allianz Arena on the 23rd April 2013. Only completed passes are shown. Darker and thicker arrows indicate more passes between each player. The player markers are sized according to their passing influence, the larger the marker, the greater their influence. The size and colour of the markers is relative to the players on their own team i.e. they are on different scales for each team. Only the starting eleven is shown. Click on the image for a larger view.

Passing networks for Borussia Dortmund and Real Madrid from the Champions League match at the Westfalenstadion on the 24th April 2013. Only completed passes are shown. Darker and thicker arrows indicate more passes between each player. The player markers are sized according to their passing influence, the larger the marker, the greater their influence. The size and colour of the markers is relative to the players on their own team i.e. they are on different scales for each team. Only the starting eleven is shown. Click on the image for a larger view.

The most striking difference between the sides respective passing networks was that Real had a greater emphasis down the flanks, with strong links between the wide players and their full backs. Dortmund were quite balanced in their passing approach with much of their play going through the trio of Hummels, Gundogan and Gotze.

Influential potential

Dortmund’s number ‘ten’ (Gotze) had a greater influence on proceedings than Modric did for Real, with Gotze coming second only to Gundogan in terms of passing influence for Dortmund. Ozil was far more influential than Modric, although he rarely combined with Higuain and Ronaldo. Modric was well down the pecking order for Madrid with the likes of Pepe, Varane and Coentrao ahead of him. On its own, this might not have been a problem but aside from Ramos and Lopez, the only other Real players with less influence were Higuain and Ronaldo. This contrasts directly with Dortmund, where Reus and Lewandowski played an important linking roles.

In summary, Dortmund’s attacking players were among their most influential passing performers; Real Madrid’s were not.

——————————————————————————————————————–

Passing matrices from Uefa.com press kits.

Assessing team playing styles

The perceived playing style of a football team is a much debated topic with conversations often revolving around whether a particular style is “good/bad” or “entertaining/boring”. Such perceptions are usually based upon subjective criteria and personal opinions. The question is whether the playing style of a team can be assessed using data to categorise and compare different teams.

WhoScored report several variables (e.g. data on passing, shooting, tackling) for the teams in the top league in England, Spain, Italy, Germany and France. I’ve collated these variables for last season (2011/12) in order to examine whether they can be used to assess the playing style of these sides. In total there are 15 variables, which are somewhat limited in scope but should serve as a starting point for such an analysis. Goals scored or conceded are not included as the interest here is how teams actually play, rather than how it necessarily translates into goals. The first step is to combine the data in some form in order to simplify their interpretation.

Principal Component Analysis

One method for exploring datasets with multiple variables is Principal Component Analysis (PCA), which is a mathematical technique that attempts to find the most common patterns within a dataset. Such patterns are known as ‘principal components’, which describe a certain amount of the variability in the overall dataset. These principal components are numbered according to the amount of variance in the dataset that they account for. Generally this means that only the first few principal components are examined as they account for the greatest percentage variance in the dataset. Furthermore, the object is to simplify the dataset so examining a large number of principal components would somewhat negate the point of the analysis.

The video below gives a good explanation of how PCA might be applied to an everyday object.

Below is a graph showing the first and second principal components plotted against each other. Each data point represents a single team from each of the top leagues in England, Spain, Italy, Germany and Italy. The question though is what do each of these principal components represent and what can they tell us about the football teams included in the analysis?

Principal component analysis of all teams in the top division in England, Spain, Italy, Germany and France. Input variables are taken from WhoScored.com for the 2011/12 season.

The first principal component accounts for 37% of the variance in the dataset, which means that just over a third of the spread in the data is described by this component. This component is represented predominantly by data relating to shooting and passing, which can be seen in the graph below. Passing accuracy and the average number of short passes attempted per game are both strongly negatively-correlated (r=-0.93 for both) with this principal component, which suggests that teams positioned closer to the bottom of the graph retain possession more and attempt more short passes; unsurprisingly Barcelona are at the extreme end here. Total shots per game and total shots on target per game are also strongly negatively-correlated (r=-0.88 for both) with the first principal component. Attempted through-balls per game are also negatively correlated (r=-0.62). In contrast, total shots conceded per game and total aerial duels won per game are positively-correlated (r=0.65 & 0.59 respectively). So in summary, teams towards the top of the graph typically concede more shots and win more aerial duels, while as you move down the graph, teams attempt more short passes with greater accuracy and have more attempts at goal.

The first principal component is reminiscent of a relationship that I’ve written about previously, where the ratio of shots attempted:conceded was well correlated with the number of short passes per game. This could be interpreted as a measure of how “proactive” a team is with the ball in terms of passing and how this transfers to a large number of shots on goal, while also conceding fewer shots. Such teams tend to have a greater passing accuracy also. These teams tend to control the game in terms of possession and shots.

The second principal component accounts for a further 18% of the variance in the dataset [by convention the principal components are numbered according to the amount of variance described]. This component is positively correlated with tackles (0.77), interceptions (0.52), fouls won (0.68), fouls conceded (0.74), attempted dribbles (0.59) and offsides won (0.63). In essence, teams further to the right of the graph attempt more tackles, interceptions and dribbles which unsurprisingly leads to more fouls taking place during their matches.

The second principal component appears to relate to changes in possession or possession duels, although the data only relates to attempted tackles, so there isn’t any information on how successful these are and whether possession is retained. Without more detail, it’s difficult to sum up what this component represents but we can describe the characteristics of teams and leagues in relation to this component.

Correlation score graph for the principal component analysis. PS stands for Pass Success.

The first and second components together account for 55% of the variance in the dataset. Adding more and more components to the solution would drive this figure upwards but in ever diminishing amounts e.g. the third component accounts for 8% and the fourth accounts for 7%. For simplicity and due to the further components adding little further interpretative value, the analysis is limited to just the first two components.

Assessing team playing styles

So what do these principal components mean and how can we use them to interpret team styles of play? Putting all of the above together, we can see that there are significant differences between teams within single leagues and when comparing all five as a whole.

Within the English league, there is a distinct separation between more proactive sides (Liverpool, Spurs, Chelsea, Manchester United, Arsenal and Manchester City) and the rest of the league. Swansea are somewhat atypical, falling between the more reactive English teams and the proactive 6 mentioned previously. Stoke could be classed as the most “reactive” side in the league based on this measure.

There isn’t a particularly large range in the second principal component for the English sides, probably due the multiple correlations embedded within this component. One interesting aspect is how all of the English teams are clustered to the left of the second principal component, which suggests that English teams attempt fewer tackles, make fewer interceptions and win/concede fewer fouls compared with the rest of Europe. Inspection of the raw data supports this. This contrasts with the clichéd blood and thunder approach associated with football in England, whereby crunching tackles fly in and new foreign players struggle to adapt to the intense tackling approach. No doubt there is more subtlety inherent in this area and the current analysis doesn’t include anything about the types of tackles/interceptions/fouls, where on the pitch they occur or who perpetrates them but this is an interesting feature pointed out by the analysis worthy of further exploration in the future.

The substantial gulf in quality between the top two sides in La Liga from the rest is well documented but this analysis shows how much they differed in style with the rest of the league last season. Real Madrid and Barcelona have more of the ball, take more shots and concede far fewer shots compared with their Spanish peers. However, in terms of style, La Liga is split into three groups: Barcelona, Real Madrid and the rest. PCA is very good at evaluating differences in a dataset and with this in mind we could describe Barcelona as the most “different” football team in these five leagues. Based on the first principal component, Barcelona are the most proactive team in terms of possession and this translates to their ratio of shots attempted:conceded; no team conceded fewer shots than Barcelona last season. This is combined with their pressing style without the ball, as they attempt more tackles and interceptions relative to many of their peers across Europe.

Teams from the Bundesliga are predominantly grouped to the right-hand-side of the second principal component, which suggests that teams in Germany are keen to regain possession relative to the other leagues analysed. The Spanish, Italian and French tend to fall between the two extremes of the German and English teams in terms of this component.

All models are wrong, but some are useful

The interpretation of the dataset is the major challenge here; Principal Component Analysis is purely a mathematical construct that doesn’t know anything about football! While the initial results presented here show potential, the analysis could be significantly improved with more granular data. For example, the second principal component could be improved by including information on where the tackles and interceptions are being attempted. Do teams in England sit back more compared with German teams? Does this explain the lower number of tackles/interceptions in England relative to other leagues? Furthermore, the passing and shooting variables could be improved with more context; where are the passes and shots being attempted?

The results are encouraging here in a broad sense – Barcelona do play a different style compared with Stoke and they are not at all like Swansea! There are many interesting features within the analysis, which are worthy of further investigation. This analysis has concentrated on the contrasts between different teams, rather than whether one style is more successful or “better” than another (the subject of a future post?). With that in mind, I’ll finish with this quote from Andrés Iniesta from his interview with Sid Lowe for the Guardian from the weekend.

…the football that Spain and Barcelona play is not the only kind of football there is. Counter-attacking football, for example, has just as much merit. The way Barcelona play and the way Spain play isn’t the only way. Different styles make this such a wonderful sport.

____________________________________________________________________

Background reading on Principal Component Analysis

  1. RealClimate

Assessing forward involvement

One of the more interesting innovations from an analytical standpoint at the current European Championship has been the measuring of the amount of time that a player spends with the ball per game. This measure of player involvement has in particular been applied to forward players, such as Mario Gomez. Gomez managed to score 3 goals from 6 shots in 2 games despite only having the ball for 22 seconds, according to Prozone. This contrasted with Robin Van Persie, who was seemingly more involved in general play, scoring 1 goal from 10 shots in 106 seconds.

This prompts the question: can we assess such player involvement on a wider level, with particular focus on forward players?

Without having access to the time in possession statistics, another measure is required. The number of passes per game should give a reasonable approximation of how involved a forward is in general play. Contrasting this with the number of shots attempted per game should provide a comparison between a forwards goal scoring duties and his overall involvement in play.

Top European League analysis

Below is a comparison of the number of shots a forward attempts per game vs the number of passes he attempts per game. The data is taken from WhoScored.com and is for all players classified as forwards and have started 10 games or more in the top division in England, Spain, Italy, Germany and France. The graph includes players who have played in a non-forward role at some point in the season, as defined by WhoScored. For example, Cristiano Ronaldo is classified as playing as both a left-sided attacking midfielder and forward, although in this case the distinction is likely irrelevant. Including players who have at some point played outside of the forward line makes little impact upon the general trend and averages (see table below).

Relationship between number of shots attempted per game vs number of passes attempted per game by forward players in the top division in England, Spain, Italy, Germany and France. The points are coloured by the number of goals scored by each player. The vertical dashed grey line indicates the average number of passes per game by these players, while the horizontal dashed grey line indicates the average number of shots attempted by these players. The text boxes (Z1, Z2, Z3, Z4) designate the zones of interest referred to in the text. All data is taken from WhoScored.com for the 2011/12 season. An interactive version of the plot is available here, where you can find any of the forwards included in the study.

Filter Players Shots/game Passes/game Goals
Forwards only 130 2.06±0.76 18.72±6.24 7.96±5.85
Mixed 135 2.10±1.01 24.67±8.82 8.23±7.57
All 265 2.08±0.89 21.75±8.21 8.10±6.77

Comparison of the different player position classifications prescribed by WhoScored. The mean and standard deviation for shots/game, passes/game and goals scored are given for each group. Mixed refers to players who have been classed as playing as both a forward and another position (generally as an attacking midfielder) at some point in the 2011/12 season.

In general, there is a weak positive relationship between shots attempted and passes attempted by forward players (correlation coefficient of 0.46 if you are that way inclined). The major feature though is that there is a great deal of variability across the forward players in terms of their involvement in player relative to their goal scoring attempts. An interactive version of the plot is available here, where you can find any of the forwards included in the study.

Players such as Mario Gomez and Jermain Defoe take an above average number of shots relative to the number of passes they attempt (Zone 1), with Gomez in particular being prolific for Bayern Munich with 26 goals in 30 Bundesliga starts. Other notable forwards with these traits include Antonio Di Natale, Robert Lewandowski, Edison Cavani, Mario Balotelli and Falcao who attempt a slightly below average number of passes but still attempt a large number of shots per game. Fernando Llorente and Andy Carroll also reside in this zone, with similar values for shots attempted and passes attempted. Players in this zone score 9.6 goals on average.

Several “star” forwards reside in Zone 2, where forwards take an above average number of shots and attempt an above average number of passes. The two extremes here are unsurprisingly Lionel Messi and Cristiano Ronaldo, who attempt the most passes and take the most shots respectively out of all of the forwards in the study. Messi ranks 34th for the number of passes across the top five European leagues, some 30 passes behind his Barcelona team-mate Xavi. Clearly, Messi’s false-nine role for Barcelona allows him to become extremely involved in general play and to even dictate it at times. He combines this with being Barcelona’s primary provider of shots on goal and indeed goals. Ronaldo is also involved significantly in Real Madrid’s play and incredibly attempts almost 7 shots per game. Several other notable forwards in this zone include Francesco Totti, Wayne Rooney, Zlatan Ibrahimovic, Raúl, Luis Suárez and Robin Van Persie with some of these forwards being more prolific than others. Clint Dempsey is an example of someone who generally plays outside of the forward line but is included here as he did play up-front for Fulham this season (scoring 5 goals in 5 games according to WhoScored). Players in this zone score 13.7 goals on average, although this is somewhat skewed by the exploits of Messi and Ronaldo (12.6 goals on average when excluding them).

Out of the 265 players included, 98 attempt both a lower than average number of shots and passes per game. In general, the number of goals scored in this group (Zone 3) is unremarkable, with the average goals scored per player being 5. However, there is one significant over-perfomer; Gonzalo Higuaín scored 22 league goals from 60 shots last season. In most squads, this would guarantee more games but he was up against Karim Benzema, who by comparison scored a paltry 21 goals from 100 shots. However, an added benefit of Benzema based on this analysis is that he is far more involved in general play.

The last group (Zone 4) includes players who take fewer shots than average but attempt more passes than average. Many of these players are more attacking midfield players than forwards, such as Dirk Kuyt. Again, a Fulham player is a good example of a player who rarely plays as a forward being included in the analysis, as Moussa Dembélé generally plays in midfield. Players in this zone score 5.3 goals on average, essentially the same as those in Zone 3.

Finishing the jigsaw

Clearly there is a large variation in how involved a forward player is in general play versus how often he attempts to score. Such differences are likely driven by both the individual player in terms of their skills and style of play alongside their tactical role within the team. Mario Gomez for instance has very similar numbers from the current European Championship for Germany as he does for his club side, although this could be a statistical quirk given the small sample size. It would be interesting to analyse how an individual performs by these measures across multiple games in multiple tactical systems.

There isn’t necessarily a better “zone” in this analysis but teams should bear these traits in mind when attempting to improve their squad. For example, Liverpool’s woes in front of goal last season led for calls for a simple poacher to be brought in who would simply “stick the ball in the net”. However, if by bringing in a poacher, Liverpool were to lose the passing and creativity provided by players in other areas, then you could end up exchanging one problem for another. Balance is key in such decisions; hopefully Brendan Rodgers can solve Liverpool’s goal scoring issues and at least maintain the quality of their chance creation next season.