This device is unable to play the requested video.
How artificial intelligence gives Leatherhead FC an unexpected advantage
The regular football season might be over but that doesn't mean a quiet summer for players or fans. England's men recently finished third in the inaugural Nations League and England's Lionesses are currently striving to win the Women's World Cup for the first time.
The hard work isn't just confined to the football pitch. Development work behind the scenes continues apace – and digital transformation is helping to change the administration of the game for the better, says Craig Donald, CIO at the Football Association (FA).
"Digital transformation means multiple things for us," he says. "It means simplification and making life easier for everyone connected to the FA. It's also about engagement and making sure that we're connecting with participants in the game, and that's everybody from a Sunday League player through to a member of the England Supporters Club and onto a Lioness performing in the World Cup. And it's about using information to improve the performance of our elite teams."
It's the latter element around performance development that is the current focus of Donald's development efforts, especially around the cloud and big data. The FA is using Google Cloud to create a platform for collaboration and advanced data analytics for the England teams at St George's Park, which is the organisation's national football centre.
Donald, who joined the FA last July, says there are two key elements to this cloud-enabled digital transformation. The first step involves placing the G Suite platform at the centre of the organisation's business operations, which means shifting from working in silos to fostering collaboration between coaches of all teams.
This connected way of working, says Donald, has helped to unify the way coaches train and develop England's 28 national teams at all levels of the game and at all ages. He says coaches need to be able collaborate and share data – and G Suite provides the platform for digital transformation.
The FA's coaches share information through a range of G Suite productivity applications and use Hangouts as their main communication tool. The data that coaches collect and share is accessed using the infrastructure as-a-service platform, Google Compute Engine.
"Moving to G-Suite provided us with the collaboration and Google Cloud is the structural foundation upon which we're building everything else," says Donald. "Our focus now is switching to the second part around performance management and the platform we're creating in the Google Cloud."
The FA is using the cloud to build a new tool called the Player Performance System (PPS). This system will be used to measure fitness, training and form of players at all levels in all 28 England national teams. PPS will automate near real-time data analysis, allowing the FA's coaches to better compare and analyse team and player performance.
It's a radical shift in approach for the FA, which is the custodian of English football and the oldest football association in the world. Donald says he hopes PPS will help to develop the game further through market-leading use of data analysis. The aim is to use the insight developed through PPS to boost female participation in the game, to make football more inclusive, and to help England coaches refine performance and deliver success on the pitch.
Donald says PPS gathers about 23 million data points from video highlights alone annually. These data points are combined with information from other sources, such as general observations about players, which Donald says totals about 300,000 statistics from 1,300 players across 30,000 fixtures each year. One of these data sources is the tracking information from the GPS vests that players wear during training camps.
"We want to bring all this information together and use the power of the cloud to harness and manipulate that data," he says. "We want to help our coaches and our office-based performance teams to pull knowledge together in different ways and to try and get new understanding about the teams and how players work well together. And then, ultimately, we'll use that knowledge to improve our performance on the pitch."
Donald describes PPS as a "0.9 version". The FA has released the system to all of its technical teams. Some of these teams are using it on a regular basis, while others are involved in the ongoing development of the system. Donald expects the FA to go live with the PPS system this summer.
The FA says it's not trying to create an algorithm to replace the selection process or the opinion of the coach. Instead, the organisation aims to collate data from all perspectives to better inform decisions – the data could just as easily confirm the opinion of a coach as refute it. Yet the FA also believes there's no doubt that technology is going to be a big battle ground for all sports teams and the organisation wants to be at the forefront of developments in this area.
The potential benefits of big data and artificial intelligence (AI) have already been showcased in football, albeit at a non-League level in England with semi-professional side Leatherhead FC. In a pioneering experiment, Leatherhead is working with technology giant IBM to see how the firm's Watson technology can be used to improve player and team performance.
While the test coincided with an improvement in results for Leatherhead last season, coaches at the club doubt that on-the-pitch success can be attributed directly to Watson. But AI, which is often seen as a replacement for human capability, has had an unexpected benefit: it's helped people at the club to bond, providing facts to back up coach opinions.
The Leatherhead experiment shows that big data can be a useful tool to add to a coach's kit bag. There's a similar trend in other sports, too – from the use of big data in Formula 1, to using RFID tags to track player movements in the NFL, and onto using AI to help edit video of tennis matches at Wimbledon. Across all these examples, one thing remains true - the use of technology in sport remains at a nascent level of development.
Whether the introduction of big data into football is a good thing is also a point of conjecture. While data might show a player has not made a huge amount of tackles or assists, they might have had a significant influence off-the-ball, working as part of a team to raise spirits and keep performance levels high. Facts and figures only tell you so much. As the Leatherhead example shows, the key to success is using data to boost human decision-making, not to replace it.
Specific details on the potential benefits of PPS for the FA are being kept under wraps for now. Donald was unable to give results from the early tests of the system. The IT team is keen to stay tight-lipped until the full launch later this summer. But the major benefits will be closely related to improving both team and player performance.
"I would love PPS for the England teams at St George's Park to be the first port of call when coaches are looking to build player performance data," he says. "We want to make sure that we're taking every piece of data that they have into this one system to allow us to access it centrally and then combine it in ways that we might not have thought of in the past."
Donald hopes the influence of the insight will eventually spread beyond St George's and to coaches on football grounds across England. "That's an interesting challenge because this won't be their area of expertise," he says.
"We could give them direct access to the knowledge. But what's more likely is that they'll work with some of our teams at St George's Park to articulate what they're looking for and then those teams, in conjunction with IT, will be able to visualise the results. But we certainly want feedback from the coaches about how they can use this data effectively."
Beyond performance-related improvements, Donald says "the sky is the limit" when it comes to thinking about how Google's machine-learning technology might be used across operational areas at St George's Park, including pitch-utilisation.
"There are plenty of other non-performance related things and problems that we have to solve," he says. "There are ways that we can harness the cloud to make those kinds of processes a lot easier for us as well. We think the kinds of developments we're pioneering with Google can help in aspects of development across all areas of the game."