From wearable health and training devices to goal line assistance in football, technology has been invading the sporting field for some years. Whether by helping the athletes up their game, catching rule infringements, or improving safety, this tech has largely all fallen within the same broad aim of improving game play.

But what would happen if we used technology not to police the rules but to create them? Enter Speedgate.

Introduced by design agency AKQA as a project for Design Week, Speedgate is reportedly the first sport to be conceived by artificial intelligence (AI). It merges concepts from croquet, rugby and soccer, with six-player teams kicking a ball around a field with three gates.

Data from 400 popular sports around the world was fed into a neural network, and then crunched to create a basic framework of rules and concepts. A large number of the suggestions weren’t exactly realistic (exploding frisbees, anyone?), but Speedgate emerged as the top choice after the company test-played the leading contenders.

Luckily, the rules seem more understandable than the game’s moto, which the AI also came up with – ‘face the ball to be the ball to be above the ball’.

So how do you play it?

The company was looking for a game that was easy to learn, broadly accessible and gave players a good work out.

The Speedgate field has three pairs of six foot (1.8 m) high posts, one at either end and one in the middle. Players must pass a rugby ball through the centre posts to gain possession and can then score by getting the ball between the end posts from either direction. Teammates that catch the scoring ball and immediately kick it back through the posts convert the goal from two points to three.

The game is played in three seven-minute intervals, with three defenders and three forwards playing on each team. The ball can be passed or kicked, but cannot remain still for more than three seconds.

Technology in play

Humans have played and watched sport since ancient times, from the first Olympics in 776 BC to cave paintings predating this by thousands of years apparently depicting early wrestling and sprinting.

Many of the most popular sports today are built on years of tradition. But while sports are still undeniably human-centric, technology and AI is changing the field.

A collaboration between car manufacturer Ford and Argo AI to produce self-driving cars is providing one example, with the deep learning being applied as a means to improve the safety record of the notoriously risky NASCAR. The technology can more quickly identify specific cars than the human eye is able to – which is trickier than you might think at 200 miles an hour (322 kph) – and allows teams to identify cars with faults.

Elsewhere, as our ability to capture and crunch data relating to health and performance proliferates, the application of AI to monitor injuries and maintain athletes’ peak condition is an obvious extension of this tech.

And then of course there is a wealth of opportunity when it comes to AI applications for sports from the viewers’ perspective. Audiences may soon be treated to nearly instant highlights and action replay: by monitoring the crowd’s reaction and the commentators' excitement level, IBM’s Watson can predict which parts of a game are most relevant. Advertisers and broadcasters could also use this information to maximize exposure to ads.

This same predictive technology could also be used by coaches to identify opponent patterns they might have otherwise missed, or analyse players’ strengths and weaknesses, allowing real-time predictive analysis which could be acted on while the game is still being played.

For some purists, technology is an unwelcome intrusion into sport, and goes against its spirit (just look at the debate around VAR (video assistant referee) in the 2018 World Cup). But given the commercial benefits it undoubtedly offers, when it comes to increasing use of AI on the sports pitch, all bets are on.