Football coaches usually have to depend on their best level of assumption to decide which player to sign, retain or sale among other pressing daily decisions that can break or make a club, a champion. Knowing the exact ability and shortcomings of their competitor’s in the next match also requires coaches to process tons of data.
However, modern sport is less and less depending on human brain, as clubs increasingly lean on data harvesting machines. These are separate companies that undertake data analytics for every player, in every match, over various measurements and grading systems.
The harvested data is not only used by coaches but also sponsors, scouting teams and betting companies, among others.
Though the trend started almost two decades ago in Europe and the US, only now has the business set roots in Ethiopia. Currently, a number of Ethiopian football clubs are installing separate departments for data analytics, videos and written performance reports.
But for Wondwossen Zewde co-founder and operation director of R&D Groups, a decade-old football data analytics company based in Addis Ababa, Ethiopia’s sport industry is well behind the technology.
“Ethiopian football clubs have been reluctant to adopt the technology for long. It is just starting now,” said Wondwossen.
Partnering with Ortec Sport, a leading football data analytics firm based in Holland, RD analyzes between 1,200 and 1,500 football matches every month.
“We analyzed data for any league around the world including the UK, Japan and Australia among others. The data is then sent to football clubs, Federations, FIFA, scouting projects and other international clients that come through Ortec, Wondwossen said.
According to him, RD does not work with betting companies because it contradicts with the company’s principle its CSR. “We know there is huge money in betting but we do not work with them. There is a company in the US, which generates odds for betting companies there, using our data.”
Wondwosen says R&D is the largest sport data harvesting center in Africa. The company started with four data analyzers twelve years ago and currently hires 121 data analyzers.
Ortec, through which international clients come to RD, pays staff by calculating the expenses of each analyzer, per game analyzed. Its employees are assisted by software that is evolving by teaching itself and at some point, will become an artificial intelligence with close to zero error, which is currently up to two percent.
During each football game, data is registered every time a player touches the ball, passes, sprints, and the like. According to Wonwossen, for instance, when the ball is touched, the staffs register at least five parameters. This includes which player touched the ball, with which leg, where the ball went from there, effectiveness of the touch and others. These parameters have their own grading systems. Other effectiveness measurements including dribbling, location, have close to fifteen grading systems.
“We produce some 67 page report from a single football match, which is sent to the clients,” said Wondwossen.
Researches indicate that clubs that rely on accurate data analytics have better performance, less staff expenses and most incisive decision making process and predictions.
However, Wondwossen says local football clubs in Ethiopia do not understand the necessity of data analytics.
“Some years back, we generated data for St. George but the contract stopped after one year, when new coach [Egyptian coach] came to the club. Many coaches in Ethiopia do not see the power of data. Ethiopian clubs must reverse this. Of course some have their own data analysis departments for decision making,” said Wondwossen.
Addis Worku, St. George football club performance analyst, agrees.
“Data harvesting is essential in football decision making. During a match, there are tracking data’s and event data numbers that give mathematical information to the clubs. You can easily use these data events registered through videos, GPS, or other mechanisms,” said Addis.
These event data and tracking data provide passes, dribbles, passing accuracy, intensity, distance covered and goal attempts.
The technology also advanced from analyzing goal attempts to expected goals.
“Previously, all attempted goals were considered as an attempt on target and are posted on possession boards. But now includes expected goals,” Addis added.
The expected goal metric gives unparalleled insight into teams and players, shedding another light on performances. The Expected Goal (xG) calculates how many goals a team should have scored based on the quality of chances created.
However, Addis says there are performances impossible to put in parameters and some qualitative data are difficult to insert into the parameters in the technology. “It is impossible to put every movement or performance of players in the data events. All players have their quality that is impossible to put in number. Some outstanding players can decide quickly taking into consideration match contexts,” said Addis.
Currently, St. George is comparatively better in deploying data analytics, compared to other clubs in the country.
According to Wondwossen, Ethiopian football can improve leaps and bounds by harnessing the latest data harvesting technologies.
“We tried so much to adopt Ethiopian football clubs to data analytics. We struggled with the federation for three years and then dropped the idea, even sponsors need to have the data, instead of just disbursing money to clubs, ” said Wondwossen adding, “After the league company is established, we are communicating with them. The managers are happy with the innovation but they think Ethiopian football is way backwards regarding adopting technology.”
The company is considering doubling its employee numbers and scale up capacity to analyze data for up to 2,000 matches per month, according to Wondwossen.
“Since we serve the international market, we have no market problems whether local clubs become our clients or not. The biggest challenge is infrastructure. We suffered a lot because of internet blackouts and Ortec was almost near to terminate the contract. We now have subscribed for a 200 MB internet and installed two power generators,” added Wondwossen.
Following in the footsteps of RD, the number of new sport data analytics companies are being established in Ethiopia currently. Wondwossen says they are preparing to form an association, to lobby the government to introduce favorable legal frameworks and incentives for the budding service export business segment.
Contributed by Ashenafi Endale & Dawit Tolesa
