From IBM Watson’s analysis of tennis players’ performance at Wimbledon, to statistical analysis driving success in Major League Baseball, the use of technology has become an integral part of professional sport, with artificial intelligence at the heart of this.
However, the latest technological advances and the potential they have to aid coaches and teams in improving their game has largely been the preserve of professionals and their multi-million pound budgets.
Therefore, a partnership between Leatherhead Football Club, a semi-professional team based in the heart of Surrey, currently in the seventh tier of English football, and computing giant IBM may appear unlikely.
However, since the start of the 2018-19 season, the two have worked in collaboration with the aim of improving the club’s fortunes.
Nicknamed The Tanners, Leatherhead FC began the season in poor form, losing six of their first nine league fixtures. Assistant manager and former player Martin McCarthy explained that when joining the club at the start of the season, he and manager Nikki Bull faced several obstacles:
“We joined the club at the beginning of the season and walked into a whirlwind as the previous management team had left the team and taken every player bar two, so we had an empty changing room. We had to recruit and bring in as many new players as we possibly could.”
IBM Master Inventor Joe Pavitt also admits that the partnership is out of the ordinary for IBM:
“This project came around based on some technology we’ve been experimenting with using IBM Watson on how you can converse with data and how AI can act as an assistant in your everyday life. Generally IBM is associated with the likes of finance and retail but in this case we’ve been working with Leatherhead FC which is not a normal partner you’d expect IBM to be working with.”
Nonetheless, IBM approached the club to see just how much IBM’s technology could “augment the intelligence and expertise of Leatherhead’s coaches and players”.
How can IBM Watson help Leatherhead FC?
The strengths of AI can be used to assist pre- and post-match analysis. This works by filming each match, with the footage then broken down and analysed by IBM Watson, with numerous aspects of the game such as set pieces, possession, type of pass and formation all analysed to provide insights for each player on the pitch.
IBM Watson Discovery analyses match reports and other sources such as Twitter feeds to provide an insight into an opposition’s recent matches, providing information on factors such as possession, influential players, or the areas of the pitch that were used. This can then be used to create a plan of attack for future games.
According to IBM, the key challenge was how to introduce complex technology to users who are not the most tech-savvy. To remedy this, IBM deployed IBM Watson Assistant, which uses natural language processing to allow players or coaches to simply ask a question, such as “show our shots against Bognor”, and Watson is able to come up with relevant video footage and analysis.
Pavitt explains that this was a steep learning curve for IBM Watson:
“It grew over time, Watson’s understanding of the game. One of the first things we had to do was teach Watson football terminology as out of the box it doesn’t know what a free kick is, it doesn’t know what a corner is. It’s also important to understand that terminology in context. Travis Gregory puts a corner into the box is very different to Travis puts the ball into the corner.”
However, the result was a tool that had even more of an impact than the IBM team anticipated, with the ease of use meaning that players, as well as coaches, were able to access the insights and assess their own performance on the pitch:
“Our plan was to build a tool that Nikki and Martin would use to plan out training sessions and do a deep dive into their game. We didn’t plan upfront that the players would use it… but it’s been so easy to use and having that reward of it being used more than we planned it to upfront.”
“There’s no favouritism in the information”
McCarthy admits that to begin with, he was skeptical over the benefits IBM’s tools could offer the club:
“It’s fair to say I was skeptical over how we could use it and what benefit there would be for us. I hadn’t seen it before… but as we’ve trialled it it’s a fantastic tool if you use it in the right way and I think that’s where the interaction of certain things comes in where you have to look at it as there being a human element in certain things which we deal with, so we have to use the tool and then filter it back to the players.”
He believes that tools such as this can allow clubs that do not have the luxury of a large budget to gain a competitive advantage. In the higher tiers of English football, many clubs have access to opposition scouts and teams of people to analyse match footage, but budget restrictions mean teams like Leatherhead FC must look elsewhere.
“We were tasked to use the information that was given to help us in our everyday capacity… we used the tool and it was a very useful for us as a football club as financially we cannot compete with some of the clubs at our level and obviously levels above.”
Simply put, the tools help coaches and players see what is working and what isn’t, useful for backing up coaching decisions with concrete data which can then be fed back to players. McCarthy believes that the insights generated by IBM Watson have made it easier to communicate with players during training:
“I looked at the tool and used it to plan training sessions, used it to look at previous performances, maybe upcoming performances and as a coach it was very useful for me to use the information we got and then build relationships with the players . I think that’s a key topic in how we use AI. It took the emotion out of it; there’s no favouritism in the information.
“When it comes to planning our training during the week, we can give the message that we can see it, you as players can see it, and we have data that backs up what we’re saying. I think from a players perspective as well, sometimes it’s hard to take criticism, but when it’s constructive criticism and when it’s backed up with facts and figures, there’s no place to hide.”
“We’re not trying to replace the coaches”
Critics have claimed that an over-reliance on technology could change the way sports are played, and not necessarily for the better, causing playing styles to become formulaic.
Some have voiced fears that technology could become advanced enough to make the role of coach or referee redundant. However, Joe believes that rather than replacing coaches, AI can be used to enforce decisions, and confirm what coaches are already seeing:
“We’re not trying to replace the coaches… What we’re trying to do is use AI to build tools to help them do their day job. If they have a question they want answered, or they have a particular idea.
“They had a way of playing at the start where they were trying to enforce short passes. If the results weren’t coming, it’s easier for [the coach] to turn round and say that’s not working, we’ll try longer passing. They can say ‘Watson, what’s our longer passing game like, what’s our short passing game like?” And we can see that actually more chances are coming from the short passes so what it meant was it can be an assistant… it can be a comfort to them and enforce what they are seeing.”
“This is something that can help you get to the next level”
Leatherhead FC finished the league in 8th place, escaping the relegation zone and narrowly missing out on the playoffs. Although it is difficult to determine the extent to which IBM Watson was responsible, the coach agrees that the added insights provided by AI certainly had a positive impact.
However, for Pavitt this is only the beginning of the journey for IBM and football, believing that the realms of data produced by the sport means that it is the ideal environment for AI to be put to work:
“Sports at the moment have too much data. Up in the Premiere League they have health data, tracking data, sleep data… they don’t have enough time to go through the data they have. So the idea of using an AI or Watson to assist them in going through that data is the challenge that we’re trying to solve.”
For McCarthy embracing new technology will become increasingly important in the game. Not only has the tool helped coaches with their day-to-day responsibilities, McCarthy also believes that a willingness to engage with new technology indicates how eager a player is to learn and improve.
“If you’ve got a tool and you don’t want to use it then I would certainly think differently about a player. One of the things we finished the season with was a dressing room of young, hungry players that want to develop themselves… and we said to the players when we were talking to them. you’ve got a tool now that is accessible 24 hours a day, 365 days a year. If you don’t want to use that then what are you thinking? This is something that can help you get to the next level.”