A football match might seem impossible to model, but with the right data and a powerful analytics platform, teams are beginning to develop algorithms for success.
When Liverpool FC’s Philippe Coutinho scored a free kick against Brighton in December 2017 by firing his shot underneath the jumping defensive wall, Jurgen Klopp credited his analysts for pointing out the opportunity.
Data was also instrumental when Hearts ended Celtic’s 69 game unbeaten run last year. Hearts Manager Craig Levein revealed that Anderlecht’s 1-0 victory over Celtic in the Champions League had pointed the way.
“The analysis of the Anderlecht game,” revealed the BBC, “showed that the Belgians played a pressing 4-3-3 and that the three midfielders and three attacking players covered an average of 11.8km per man when hustling and harrying Celtic out of possession.”
Thanks to their own club analytics, Levein and his coaching staff knew that they had players who could cover that sort of distance. So the strategy was to “hit the 11.8km average, frustrate Celtic, press them high up the pitch, cut out the passing route to their most dangerous players, refuse to let them settle and force some errors.”
Hearts ran out 4-0 winners.
Football has become increasingly data-driven and Hearts drew upon InStat’s vast database of football statistics to help mastermind their victory.
More than 1,500 clubs and national teams use the system, which features information on more than 400,000 players. According to InStat, over 6,000 new video records are added to the database every month and the company employs 500 people in its Production and Data Center to analyse and document every sprint, kick, dribble, challenge and pass.
While InStat data can prove vital in areas such as opposition analysis, match preparation, scouting and player recruitment, clubs also assess the performance of their own players using wearable sensors and analytics.
A typical Catapult Sports wearable vest with an embedded OptimEye S5 device, for example, can collect up to 1,000 data points per second. It can measure fitness and skill levels, response to specific training techniques, tactical performance and risk of injury.
The STATsports Apex can calculate over 50 metrics in real time on the device
The STATsports Apex pitches itself as the ‘most powerful wearable in sport’ and can calculate over 50 metrics in real time on the device. In addition to core performance indicators such as max speed, sprints, heart rate and step balance, it can also number crunch factors including High Metabolic Distance and Dynamic Stress Load.
“These show how hard a player is working,” say STATsports, “and how much a session has affected them. GPS tracking isn’t just about how far a player has run, but how they have been affected by the exercise. Crucially, this data can be viewed as training is happening, due to the on-board data processing power [of the Apex module.]”
Data from the Apex modules is stored in the STATsports cloud infrastructure, allowing coaches and managers to access the information anywhere, anytime and on any device.
It’s never been easier to generate and access footballing data. But it’s what you do with the data that counts. To some extent, an analysis of past performance can inform predictions for the future – which players to buy, formations that work best against a certain opponent, the percentage chance that crosses (rather than through balls) will result in a goal.
You’ve got to ask the right questions.
“If I have a football club and our objective is to get into the Champions League,” analyst Rory Campbell told the Independent, “there are loads of things that are going to control whether or not that happens.
“How many points do we need to get into the top four? We can work that out. Then how many goals are we going to need to score? What is our goal difference going to be? Then let’s drill down to think about exactly how many shots we are going to need to have, and how many shots we are going to concede. And work back from there.”
With measurable objectives to aim for, a football club could then assess its own resources and see how much of a gap there is between where they are today and where they want to be tomorrow. If a team isn’t creating enough goal-scoring opportunities, for example, can that be improved by a change of formation or a new striker with a higher shots percentage?
Of course, with so many variables in the average football match, analytics isn’t a magic bullet. Statistical predictions can be upended by simple human error – a missed shot, a poor pass, a clever dive or a contentious refereeing decision. Or they can just as easily be defined by a moment of brilliance from a world class player.
But, used intelligently, analytics technology can give teams an edge over their rivals. Sensor data from training sessions — along with what players eat and how much they sleep – can be combined with cloud-hosted, instant-access statistics from the likes of Opta, InStat and Scout7 to help identify in-form players, define personalised training regimes and improve physical performance.
A football match seems impossible to model – 22 players (with their own strengths and weaknesses), three officials (ditto), contrasting formations, luck, randomness, pitch conditions, weather effects, crowd noise, the cumulative behind-the-scenes impact of coaches, doctors, physiotherapists, masseurs, kit managers, nutritionists, analysts and sport scientists.
But with a combination of cloud computing and big data analytics, teams are beginning to understand the probability of success and, more importantly, what they can do to improve it.