Artificial intelligence in sport

Deputy sport editor Henry Mallin makes a strong case for the further integration of artificial intelligence in sport, especially in football.

Photo: Wikimedia Commons

Artificial intelligence (AI) in any field is always a contentious topic. Even the name artificial intelligence sparks fears of a dystopia styled like Ex Machina or I, Robot. However, in the case of sport, there is strong argument to suggest that an increasing dependence on AI can only be a good thing.

Firstly, I want to take you back in time. The scene is 27 June 2010. Matthew Upson has given England a lifeline in its Round of 16 World Cup match against Germany at the Bloemfontein Stadium. The ball then drops to Frank Lampard, who effortlessly lobs Manuel Neuer with a sublime finish that had essences of Geoff Hurst’s World Cup-winning goal. Fabio Capello leads the celebrations from the dugout, and England is back in the tie.

But wait. Jorge Larrionda has waved play on and the goal has not been counted. Jeers ring around the stadium, and England fans are in disbelief.

Seven years later, this moment still haunts me. Video evidence clearly showed that Lampard’s shot had crossed the line –– the goal should have stood. The decision changed the game, and England’s years of hurt continued as a result of human error.

This is not the first time such a mistake has happened in football. Famous examples such as Maradona’s hand of God, Luis Garcia’s ghost goal, and Henry handing the ball into the goal in a decisive game in the World Cup qualifiers all live on in the memories of football fans.

In the aftermath of the 2010 World Cup, Sepp Blatter, FIFA president at the time, issued a public apology and stated that technological means would have to be reconsidered as viable tools to aid referees. The result is that goal line-technology has been put in place for major international fixtures and introduced in the Premier League.

Football is far from the only sport moving in this technological direction. It is now a rarity that a try in rugby is scored without the referee going to the third official for confirmation. Whilst this can slow the game down fractionally, the result is greater accuracy in the decisions of the officials, a result that benefits all. Similar technologies have been introduced in many sports; tennis uses the hawk eye system, and cricket umpires have a third umpire.

But can we take this further? Are we getting to the point in which we could and should remove human error from sporting officiation and move towards referees with artificial intelligence?

My mind immediately thinks about the game FIFA, something that happens far too much in my daily life, but on this occasion I feel it’s useful. I have distinct memories of playing luscious through balls, splitting the defence, and leaving a one-on-one opportunity with the opposition’s keeper just to find out that I had drifted into an offside position. Anger erupts inside of me, and I shout at the TV for making such a terrible decision before being subdued by the clear evidence of my player being half a yard offside. As always, the game proves correct, and I quickly accept my mistake.

Whilst I know video games are not reflections of real-life situations, there certainly are parallels that can be made. Using an AI referee, similar in style to that of a referee on a FIFA game, would remove human error from referee’s decision making; furthermore, we could stop episodes such as Lampard’s disallowed goal.

Human referees are subjected to an onslaught of abuse from fans and players alike. An artificially intelligent referee who is immune to human error would be a more adequate solution to this problem.

The counter-argument is that we require human referees to provide a better interpretation of the rules in certain situations. This logic is weak, and at best it fails to effectively counterbalance the criticism of human error in referee’s decisions.

On a slightly separate note, artificial intelligence has proven an immensely powerful tool in improving sporting performance. The use of it in elite sport has become almost universal. This ranges from reviewing player performance to assessing opponent’s strengths and weaknesses. Artificial intelligence differs from statistics. Statistics in their simplest form do not capture the complexity of sport.

Artificial intelligence uses algorithms that can make sense of random patterns in data sets to provide meaningful outcomes. Bayern Munich has an official partnership with SAP, the German software giant, for detailed post-match analysis highlighting the increasing impact of AI and the reliance teams have on it.

With this new analysis of data sets, artificial intelligence is set to provide a vital role in the development of sport, and the increasing trend can only mean positive things for the future.


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.