AI makes headlines in cricket: can the machine help us pick better teams?

If you caught the last edition of Crunching the Numbers , you'll know about our range of new AI products - a suite of chatbots for T20 cricket, NRL, NBA & NFL aimed at sports businesses looking for AI solutions.

We're not the only ones in sport venturing into the world of AI. If you've scanned the news lately, you might have come across a public acknowledgement of the use of AI in team selection by one of the world's top cricketing nations.

The reliance on AI for a more complex task such as team selection is a particularly interesting use case. This installment of Crunching the Numbers offers some commentary on the forms this task can take, discussing both the aforementioned example & our use of AI here at Wicky.

AI in Pakistan cricket

In the midst of a challenging period following a home Test series whitewash by Bangladesh and a group stage elimination from the 2024 T20 World Cup, Pakistan cricket is publicly pronouncing its journey into the chrome-plated future.

The Pakistan Cricket Board (PCB) made headlines recently for openly acknowledging its use of artificial intelligence in player selection. In search of the winning formula, the PCB chair Mohsin Naqvi, in a press conference at the end of August , outlined the role of AI, and its importance relative to human input, in selecting a pool of players to feature in the Champions Cup one-day competition as follows:

"These 150 players that have been selected, 80% of it has been done by AI (Artificial intelligence), and 20% using humans. Nobody can challenge that. We gave about 20% weightage to our selection committee. If we replace a player with a worse one, you'll be the first to complain. We'll have records and we'll all be able to see transparently who deserves a place in the team."

The specific purpose for involving AI is significant; the problem Naqvi highlighted was an insufficient talent pool and the shortcomings systemic in nature:

"The problem is the selection committee has no pool to turn to from which to select players," he said. "I spoke of surgery because we need to fix our problems. But when we look at how to resolve them, we don't have any solid data or player pool which we can draw from. The whole system was a mess."

What stands out right away is the very acknowledgement of the machine shouldering a significant portion of the selection responsibility. The increasingly widespread nature of analytics in world cricket especially over the past decade is a newsworthy event in itself, with teams - such as Islamabad United from Pakistan's own Super League, to name one example - making headlines for their public embrace of data-driven methods.

Viewed through that lens, even absent the details, the narrative significance of Pakistan cricket's open acknowledgement of an AI-powered approach is apparent - it's another step, as with analytics, in the direction of relying on the machine for better sporting decisions & the first public step into artificial intelligence specifically.

How Wicky's using AI for cricket

As mentioned in the last Crunching the Numbers, we've been developing an AI chatbot specifically for fantasy cricket, the prototype of which you can find here, along with all our other chatbots .

Having worked with Draftstars, Australia's only daily fantasy sports platform, we're no strangers to this space. We've had a little taste of using AI in team selection ourselves (tongue slightly in cheek) - at present, the fantasy cricket chatbot can recommend users complete fantasy teams for DFS contests between any two IPL teams, besides other features such as helping users decide between competing options.

We've also got a very exciting development in the works re the fantasy cricket chatbot. Right now, we're adding risk-based captaincy recommendations for users' fantasy teams, suggesting high, medium, or low risk options depending on the user's personal risk appetite, taking into account various relevant metrics.

It's a further layer of sophistication to our chatbot that assists users with entering multiple teams into DFS contests. Not only do our AI-powered chatbots significantly speed up the process of creating teams, but the risk-based captaincy recommendations help further refine their input of multiple teams with varying degrees of risk to cover as many outcomes as possible.

Just as Pakistan cricket isn't aiming for the complete elimination of human input, we're aiming for the user to have control over the risk profile of their DFS teams, down to their captaincy choices.

If you're a business searching for AI solutions and our sports chatbots pique your interest, we'd love to hear from you! Check out the video linked above & email us on [email protected] to schedule a demo or learn more about how we can help you stay ahead of the competition.