Utilizing Machine Learning Models to Improve Football Match
Utilizing Machine Learning Models to Improve Football Match Outcomes

Utilizing Machine Learning Models to Improve Football Match Outcomes

Utilizing Machine Learning Models to Improve Football Match Outcomes

The increased internationalization of football over the past two decades has further diversified the “beautiful game,” making it even more competitive as new playing styles, mindsets, and footballing philosophies have made their way onto the field. Therefore, it has become even more challenging for coaches and players to work out winning strategies in tournaments, in particular in the fiercely competitive World Cup showdowns. 

With 22 players present at the same time in a game on the pitch, each vying to leave a mark on the bottom line, working out the most optimal game plan is a cumbersome task for the coaching staff and match analysts, considering each player’s strengths, weaknesses, and other external factors. Lately, thanks to the advancement in technologies, formulating such strategies have become easier and more tangible, meaning the teams can breathe a sigh of relief.

Artificial Intelligence (AI) and Machine Learning (ML) are increasingly playing a crucial role in improving the state of play in all professional sports, especially football. It helps football management with predicting the success of  the matches through a detailed analysis of data, AI modeling, and much more. With the FIFA Qatar World Cup 2022 around the corner, all teams including Qatar’s national team, are working to improve their state of play and shape strategies that will yield favorable results when the world cup starts. 

As such, Hamad Bin Khalifa University’s (HBKU’s) College of Science and Engineering (CSE) researchers have proactively developed an AI-based model to analyze and recommend key performance metrics to win matches for the Qatar national team players. With less than two years to go, the researchers are upbeat that their study will result in tangible outcomes for the team.  

Jassim Al-Mulla, PhD student at CSE, and his supervisor, Dr. Tanvir Alam, Assistant Professor at CSE, have already developed an AI-based model to support the players, coaching staff, and team management to focus on specific performance metrics that may lead to winning matches. Being the first AI-based model in the MENA region to analyze and improve football players' performance, the research study has already been published in the journal IEEE Access:  https://ieeexplore.ieee.org/document/9261335

“As researchers, we always try to contribute to the betterment of Qatar. As the global footballing event is fast-approaching, we leveraged our skills in technologies and AI-based modeling to support the Qatari football players. We modeled the football match performance of the players from the Qatar Stars League and analyzed all the field positions (defender, midfield, and forward) of the players during the match time, and formulated this study as a classification framework under the machine-learning (ML) context to distinguish the winning team from the losing team in a match,” said  Al-Mulla. 

The study reveals some key winning insights and strategies for the teams based on the players’ technical skills as well as physical performance that could prove helpful in deciding favorable match outcomes including: 

  • Shots on target by the forwarder, distance covered by the forward and midfielder at high speed, successful passes, and shots on target by the defenders were vital performance metrics to win matches. 
  • Usually, in a football match, the team focuses on forwarders to win matches. Interestingly, the study found that the defenders also play a key role along with the forwarders, and the defenders’ role cannot be ignored for winning games. 
  • Playing fair games helped the teams win matches rather than saving them from losing matches.
  • The study also proposes a novel AI-based approach to predict match-winners in the next seasons of QSL considering the performance of the previous few seasons. This will help the teams at QSL to avoid the mistakes made in the earlier season and try to design new strategies to win matches in the upcoming season.

Dr. Mounir Hamdi, dean of CSE, says that “Experts at the College of Science and Engineering use different technologies and Artificial Intelligence (AI)-based modeling techniques to model professional football players’ performance to help improve their match performance and enhance the chance of winning matches in the World Cup.” 

Motivated by this AI-based study, Al-Mullawould like to investigate other aspects of football matches to help the Qatari football players perform better, ensuring the team’s competitive streak continues even beyond the upcoming mega footballing event in 2022.