Artificial intelligence can transform modern-day cricket: Dhammika Perera

Artificial intelligence can transform modern-day cricket: Dhammika Perera

11 June 2019 03:35 pm

Business leader and artificial intelligence champion Dhammika Perera says AI can transform modern-day cricket, and Sri Lanka Cricket must quickly embrace this revolutionary technology to improve its performance.  

“AI is hugely disrupting the way sports are being played in recent times. One such sport that has been actively using AI is cricket. AI has been extensively used in this game to record players’ activities and propose improvements, predict outcomes of the matches and much more,” said Perera.

According to him, data science is the study of how data is obtained, deciphered and analysed to come up with useful information invisible to the eye. In cricket, the on-field information gathered is quantified and analysed by the team managers, to evaluate the strengths and shortcomings of a given player or any team precisely. For instance, to select a suitable batsman for a particular match against a specific team, it would be very difficult to analyse various parameters in short span like the score breakdown against bowling type (fast/spin/slow), run/strike rate against bowling line and length, average scores in over range, dismissal criteria (bowled/ LBW, etc.), performance in specific venue/ weather conditions, etc. “Data science has made all of this possible now to analyse millions of combinations in split seconds, to give insights which could be used by the teams to take data-driven actions,” said Perera, who recently spoke extensively on AI at a forum organised by the Computer Society of Sri Lanka.

“AI can be used to vastly improve the performance of batsman, bowlers and fielders as well,” said Perera, adding that data scientists can evaluate the batsman performance in training/ practice matches/ main matches on parameters like the number of runs scored, the number of balls faced, the number of fours and sixes, strike rate, runs scored against specific bowlers, etc. to give information on the areas the batsman is weak on and in the type of pitches the batsman can provide optimal performance. The information concerning his body movement in response to different measures of length, line and bowling speed in easy to understand graphs can show his weaknesses and how he can work with them to refine his performance.

The similar analysis can also be done for bowlers by combining his past performance with multiple other factors such as pitch data and length, weather conditions, and ground information to provide insights which could help him in improving his performance.

Perera revealed that cricket teams from the England and Wales are the classic examples which have used data analytics to decipher real-time data collected from matches played on ECB grounds. “A lot of teams in T20 domestic leagues in India have started using the strategic timeouts where they use real-time data analytics to create the strategy for their next few overs and action them on the ground,” he added.

“For us to uplift the level of Sri Lankan cricket, we need to take steps now to start driving the data-driven mindset right from the school level. We need to educate teams, coaches and management about this way of thinking, use technology like Internet of Things (IoT) to enable real-time data-gathering island-wide for the players, and then use data science to help in making right decisions,” stressed Perera.

He emphasised that use of AI can provide an edge to the cricket teams as they can now take much more informed decisions which can uplift their entire game to a whole new level. Knowledge about your own and opponents’ weaknesses and developing strategies in real time harnessing data driven insights can become a game-changer for any team.

Reaching the level of creating and executing strategies on the ground using data driven insights is a journey which Sri Lanka has to start now, else we will just depend on the prowess of a few and be left far behind in the cricket arena, added Perera.