Artificial Intelligence on the Football Pitch

August 15, 2019

Predictive analytics, based on big data, is becoming an integral part of developing and delivering successful strategies in sport. These approaches are evolving to become more autonomous – and are being described as ‘artificial intelligence’ by those in the industry. Some of the world’s most renowned football clubs are already employing whole AI departments to help gather, process and analyse data, aimed at improving all aspects of the club’s performance – from marketing and sales to the results on the pitch.

The ultimate aim is to produce comprehensive reports for the managers and empower them in delivering a winning strategy for every game.

Simon Sherrington, MD

Get in touch on 07917 541246 to discuss what we can do for your business.

SL Benfica, Lisbon is among the sports organisations in the world investing in AI, IoT. “For the past eight years we have been collecting data on every single touch on the ball, by every single player. At present, we rely on 100 GPS devices, tracking 100 variables.’’, said Joao Copeto, Benfica’s Chief Information Officer. He also admitted that the club’s experts are still working out how to make full use of all the data obtained. Benfica relies on wearable connected Internet of Things (IoT) devices to monitor various physical activities. According to Copeto, IoT is meant to provide real-time answers on players’ physical fitness, risk of injuries and levels of performance. Although, in his words, the technology still fails to keep pace with the demands of modern football.

Benfica sports scientists believe that AI comes into play when predictions are required. In the world of football that is twice a week, before every home or away game, in both the domestic championship and the European tournaments. The Portuguese club has a sophisticated, in-house database, where data is stored for analysis, run by a multidisciplinary team. According to the chief information officer, it is designed to be open to changes and constantly evolving. “The ultimate aim is to produce comprehensive reports for the managers and empower them in delivering a winning strategy for every game’’, said Nuno Maurício, Head of the Analysis and Observation Department (Professional Football). However, experts confess that there is still not enough data for AI to be trained to avoid bias, so human analysts’ help is needed.  In this case, we are talking about Machine Learning models and supervised training of them, and for most people, ML is an AI approach; it’s not “general AI”, but it is task-oriented AI. Benfica’s experts say they remain devoted to building a predictive analytics model run entirely by AI.

According to Newcastle United’s Jonathan Fitzpatrick, AI’ s main application in football is predicting scores and injuries, but he is also convinced that it can  be used to foretell talent: “Modern football is a multi-million-pound business. Clubs need to know what the long-term perspectives are for each player, so they can manage their investments the right way”. Benfica’s scientists are also trying to predict future career development of players by introducing psychological profiling. However, both clubs say AI-driven talent prediction is something to look forward to in the future. This approach is already used in other industries, such as recruitment for other jobs, where AI techniques filter applicants based on what their CVs say.

Collecting data requires consent either by the players or their parents in the case that they are under 18. We anonymise and secure all our data before sharing it with universities and research bodies for analysis.

One of the most significant issues related to the process of data harvesting is privacy. “Collecting data requires consent either by the players or their parents in the case that they are under 18. We anonymise and secure all our data before sharing it with universities and research bodies for analysis,” according to Copeto. He stressed that data gathered on players belongs to the players themselves but can be shared with another club in case of a transfer. Benfica is set to follow strict data ethics rules, he added. The club has no intentions of trading data; it will be shared with universities (for research purposes) and other clubs (in the case of transfers). Benfica says it is dedicated to building and using its own security tools rather than rely on those provided by vendors.

Benfica’s representatives revealed that some of the sophisticated methods for tracking sports activities, using IoT and AI, are also deployed to monitor opponents. AI and predictive analytics are utilised in the commercial aspect of running the football club too. This include tracking supporters’ behaviour during breaks, pre-and post-game and online. Although football data scientists find this useful, they agree that the real measurement of scientific success will be AI’s full integration with sports aspects.

[Image licensed by Ingram Image]

Other Stuff

In Pain? Don a VR Headset

In Pain? Don a VR HeadsetSoftware developers and medics around the world are working to prove that Virtual Reality (VR) powers stretch far beyond gaming and entertainment and have the potential to aid thousands suffering from cancer, anxiety, personality disorders,...

Joseph’s technocapable coat: energy harvesting for smart clothes

Smart clothes are where style and science meet, giving garments a whole host of innovative applications, such as charging depots for personal electronic gadgets, fitness trackers for capturing biometric data and colour-changing fashionable assets that go with everything.

Could AI help spot a fake Donald Trump?

spotting a photograph or video where part of the image has been manipulated. Such a challenge faces news organisations on a regular basis: sensitivity over “fake news” means responsible publishers are on heightened alert to potential manipulation.

Telecom Industry Slices 5G

Network slicing, a key feature of 5G, lets operators automatically create separate, virtual end-to-end networks over the same physical infrastructure.

Industrial Internet of Things Dangers

Today’s industrial technology settings have more interfaces than ever before, making industrial systems some of the most attractive targets for malware and ransomware attacks.

Electric Vehicles: Charging Infrastructure Policy Stifling Adoption

Financial incentives are available to encourage EV drivers with access to off-street parking to install home-charging units. Local councils have access to funding to install on-street EV chargers. The former has been successful, the latter has so far failed.

FarmTech: Application of Drones

Drones are eyes in the sky helping farmers gain insight into crop growth and about microclimates within individual fields …