- Sport Business
- Members Log In
A Leaders Performance Institute article brought to you in association with our Supporters
By John Portch
It would be unthinkable today for the head coach of the Bundesliga champions to be scratching around for video footage of opponents ahead of a crucial Uefa Champions League tie.
Yet that very scenario faced Ottmar Hitzfeld throughout his stint at Borussia Dortmund between 1991 and 1997. During that timeframe, BVB reached the Uefa Cup [a forerunner of the Europa League] three times, twice competed in the Champions League and won the competition in 1997.
At first, the scraps of overseas footage that Hitzfeld and his assistant, Michael Henke, found on German television did not really cut the mustard, while scouting trips that covered just one or two matches did not provide meaningful insights that could be shared with other members of staff.
Dortmund, like most of their Bundesliga opponents, are a sports club of which football is but one division. The solution would lie with the Borussia Dortmund women’s handball team.
The head coach of Dortmund’s supremely successful women’s handball side of the era had instructed the team’s analyst Markus Schultz and a team of sports science students, to edit video material that had been painstakingly acquired from abroad. Henke now approached them for help doing the same with foreign opponents of the men’s football team.
Henke and Schultz would go on to found their own sports analytics company and now, as Christoph Biermann writes in his new book Football Hackers: The Science and Art of a Data Revolution: ‘A Bundesliga coach sitting on the team bus on the way home from an away game was suddenly able to watch the match back on a CD-ROM on his laptop. A few years later, he would receive an email with a data code, making it even easier for him to zoom in on defining situations, to concentrate on corners or on his side’s misplaced passes in their own half.’
Biermann goes on to explain that the VCR has influenced football more than any other technology over the past 25 years. ‘Coaches found it much easier to learn from their best colleagues and analyse their own games,’ he observes. ‘They could show players what they are doing right or wrong rather than simply talking about it. Visual learning engendered great improvement across the board. What’s more, no one ever went into matches blindly any more, not knowing what to expect. Players could literally see what type of opponent and patterns of play they were up against in advance.’
Here, the Leaders Performance Institute, through the prism of Football Hackers, explores how video footage and complex mathematical models are enabling a more profound analysis of the game and inspiring a different way of thinking about a sport often perceived as too dynamic and fluid to fit neatly into analytical patterns.
The coach as an analyst
‘[Pep] Guardiola is representative of a type of management that would be inconceivable without video technology,’ writes Biermann. Guardiola, currently the manager of double Premier League champions Manchester City, served as Head Coach of Bayern Munich between 2013 and 2016, which saw him working closely with a young analyst called Michael Niemeyer. ‘Pep took meetings that lasted for hours. With us analysts and with the players. They were a footballing epiphany for me,’ Niemeyer tells Biermann.
‘He’s all about finding a way into the box and creating superiority, with a position game. He would think for hours on end, sometimes throughout the night, discussing it with the analysis team. He would watch games himself as well, to come up with ideas for a strategy. Perhaps some work was done twice that way but it was necessary.’
As Biermann explains, Niemeyer’s recollections are part of the myth that has developed around one of the sport’s most respected thinkers, and Niemeyer goes on to says that Guardiola, in his view, has changed the job description of a football manager in the process.
‘For me, a good coach is also an analyst,’ he declares. ‘After working with the kind of brilliant teachers that we have had, you, as an analyst, start thinking like a coach, too. Not in the sense that I will direct the players to do some exercises from the touchline or explain to them how they should cross the ball – but in the way you understand the game.’
Niemeyer notes that coaches such as Guardiola or Paris St-Germain Head Coach Thomas Tuchel have also made the game more tactical, with increased focus on positions, passing lanes and pressing patterns. The question for data and analytics, then, is to be better represent these fields in numbers that provide answers and insight while promoting greater understanding.
New metrics, new clues
While this all sounds very promising for video analysis, Biermann also highlights the roadblocks that data analysis faces, with Guardiola himself having reservations. To demonstrate the problem with statistics, Biermann takes Germany’s seismic 7-1 defeat of hosts Brazil in the semi-finals of the 2014 World Cup, and tells as that by a vast array of metrics, such as possession, dangerous attacks, tackles won, lost balls, shots on target, Brazil were ahead on every count after 90 minutes.
‘We, as analysts, disregarded almost everything we had looked at in previous years, from time spent on the ball to the number of vertical passes and so on,’ Christofer Clemens, Head Analyst of the German men’s team admitted to Biermann of that tournament. ‘We’re increasingly convinced that there’s lack of data that provides real information about the things that make you successful in football.’
Biermann states that numbers could be useful in the field of fitness but that: ‘data’s usability for devising successful strategies with the ball and tactics was less clear-cut.’
Enter a bank auditor from Northern Ireland called Colin Trainor, who back in the winter of 2014-15 became interested in the seemingly inexplicable slump of Jürgen Klopp’s Borussia Dortmund that season and decided to conduct a deep dive analysis.
Trainor was not a Dortmund fan and had not seen a single match, but was still able to devise metrics, namely ‘After Expected Shots’ and ‘Passes Per Defensive Action’ [PPDA], which Biermann delves into in greater detail in Football Hackers.
Biermann also explains that attempts to provide clues about the game’s inner machinations, such as those proffered by Trainor, ‘[have] become one of the key analysis trends in recent years, [which] more often than not derived from work done by obsessive members of the global data underground.’
Elsewhere, Biermann discusses the rise of ‘Packing’ and ‘Dangerousity’, metrics and models that all point to the passing of football’s ‘big data hangover’. He writes: ‘Right now, advanced game data might only be a sideshow when it comes to analysing team play; the calculations are not quite eye-catching enough to fundamentally alter game analytics. But the later will once models […] become suitable for daily use, and when every action on the pitch will have a specific mathematical value. At that point everything we thought we knew about football will be called into question again.’
Biermann adds that data will also play a crucial role in the automation of visual game analysis. ‘A situation on the pitch is codified, data gets searched along these parameters and the corresponding video sequence gets played out.’
Then: ‘Instead of manually searching through videos for hours on end, analysts will leave computers to do that menial work and focus on more productive tasks instead.’
The revolution is already underway.
Football Hackers: The Science and Art of a Data Revolution is available now from Blink Publishing.
Looking for more data insight?
Download our most recent Special Report: Navigating the Data Maze featuring Texas Rangers, Orlando Magic and Australian Institute of Sport.