Network science is all the rage these days. The methodology of how things - information, disease, money - spreads between nodes in a network is used to analyze classical problems such as the spread of a pandemic, war, forest fires or financial panics.
JOIN21 is one of several outfits applying network analysis to organizational theory, through its JOIN21 Network Leadership© product suite – allowing a view of the organization as it actually works as a mix of formal and informal relations.
More recently, sports teams have applied network analysis to unveil their performance metrics. A team is considered as a complex network whose nodes – the players – interact with the aim of overcoming the opponent network.
Can it explain the F.C. Barcelona team of Messi, Puyol and E’too?
Simply the best
Pep Guardiola’s Barcelona side of the 2008-9 season is widely considered one of the best teams in sporting history. They won an unprecedented six buckets, including La Liga and Champions League (a lopsided final defeat of Manchester United in which while watching you understood that football had changed forever).
Guardiola’s team revolutionized the game with its possession-oriented game, short and quick passes between players that were situated very close together. And then the immediate, hefty pressure on the opposition when they lost the ball. This tiki-taka football was the antithesis to Jose Mourinho's defensive teams (Inter won the Champions League the following season) and inspiration to managers like Jürgen Klopp.
[QuoteText]Network analysis lets you visualize exactly how this difference pays out in a given game. Measured by traditional football metrics such as passes, shots and goals Barcelona were supreme. But these metrics do not capture the style of play. Javier Buldú and colleagues at the Universidad Rey Juan Carlos in Spain went looking for the Barcelona signature by analyzing the team as a network. [/QuoteText]
Each player is represented by a node, creating a link between players every time they pass to each other. The link becomes thicker with each pass. The data also shows the player’s position when passing. By the end of the game the network map is a powerful document of how the game evolved.
Greatness as a network. The researchers found the network metrics that enhance the probability of scoring/receiving a goal, showing that not all teams behave in the same way and how the organization Guardiola’s F.C. Barcelona is different from the rest - including its clustering coefficient, shortest-path length, largest eigenvalue of the adjacency matrix, algebraic connectivity and centrality distribution.
[See how JOIN21 Network Leadership applies network analysis to gauge an organization’s performance]
Researchers have applied network analysis to football before, as summarized by MIT Technology Review in this great article. They can find that certain players are more ‘central’ or that certain patterns of play are common such as passing forming a triangle between three players. Buldú and colleagues analyzed how the network changed throughout the game, by isolating 50 passes at a time and comparing these 50-pass networks to each other. They began by generated passing networks for every team in the 380 games played in the 2009-10 La Liga season. Then they calculated known network measures such as the clustering coefficient – determining that triplets of players pass the ball far better in Barcelona than in any other team. Also, the average shortest pass through the team is shorter for Barcelona. And the strength of the network (the eigenvalue of the connectivity network) is also much higher for Barcelona than any other team.
By looking at the evolving 50-pass networks, it is revealed how the 2009 team’s centroid – its average field position – changes during the game. Barcelona’s was higher up the pitch and much more stable than other teams’.
Barcelona’s ratio of advance – how likely the team is to pass forward, backward or horizontally – shows a much higher likelihood of horizontal passes. Meaning they pass the ball back and forth looking for attacking opportunities.
Xavi, the brilliant midfielder, has the highest centrality of any player.
Can the network analysis identify footballing weaknesses?
As paraphrased by Technology Review, «the likelihood that Barcelona will concede a goal increases when the dispersion of players around the team’s centroid increases — in other words, when the players spread out. That suggests an Achilles’ heel that other teams could attempt to exploit.»
Other teams, however, behave differently. Valencia is more likely to score when dispersion around its centroid increases. This means, of course, that network analysis can not be used by other teams to copy Barcelona’s strategy and success. Guardiola’s system, however, has also proved very successful at clubs like Bayern München and Manchester City.
It is easy to imagine the many ways network analysis can add interesting dimensions to sports analysis. For instance by discovering how single players influence a game or how changing positions can influence their performance.
The new Netflix series «The Last Dance» about Michael Jordan’s great Chicago Bulls teams of the Nineties, centers for a while around coach Phil Jackson’s famous triangles. It would be interesting to see network analysis applied to the NBA.
Meanwhile, Organizational Network Analysis (ONA) is continuing to show leaders and employees how to view their organizations as a mix of the formal and informal ties that make the organization tick.