Wimbledon organisers deploy AI to analyse matches

The All England Lawn Tennis Club, in collaboration with IBM, has employed artificial intelligence to analyse players’ performance and forecast match outcomes. The Next Web.

To collect data on players, the company’s engineers installed a multitude of sensors on 18 Wimbledon courts. IBM also drew on historical data, automated match-tracking systems, and NLP analysis of media discussions.

The collected data were transformed by the algorithms into digital features accessible in the Wimbledon app.

They include the IBM Power Index, ranking players before and during the tournament. The forecasting model then determines their probability of winning each match.

According to an IBM spokesperson, in the early days of the tournament the predictions reached 100% accuracy on Centre Court and Court One. However, on the outer courts, predicting the outcome of matches proved more difficult.

“When you go to some of the outer courts and to less well-known players, there is less data available, so making a correct forecast is harder,” said Kevin Farrar, head of IBM’s sports partnerships in the United Kingdom.

IBM also used AI to create the Match Insights feature, which generates the tennis players’ information bulletins, and the In the Media section with analyses of expert opinions.

Fans can submit their own forecasts, access personalised recommendations, and watch personalised highlight reels of the best moments.

“All of this is designed to attract spectators so that Wimbledon can reach its goal of increasing attendance,” said Farrar.

In July, FIFA said it would deploy an AI algorithm to determine offside positions at the upcoming FIFA World Cup in Qatar.

In July 2021, AI named tennis the most “euphoric” sport.

Subscribe to ForkLog news on Telegram: ForkLog AI — all the news from the world of AI!