Sports’ AI frontier: The knowns and the unknowns

Dear reader,

It’s good to be writing after a hiatus. After a ten-day break, my suffering the flu, and a public holiday, The Playbook is back just as the ongoing Indian Premier League (IPL) season is heating up. Speaking of which, I couldn’t help but marvel at young Ashutosh Sharma’s fearless shotmaking last night. Sharma, a Punjab Kings lower-order batsman from Madhya Pradesh, almost pulled off the impossible (yet again!). But in that fifty he scored, he did something unthinkable: hit a six off Jasprit Bumrah, a rarity given the latter’s class, accuracy, and execution. 

Prolific six-hitter Sharma, who also plays for Railways in first-class cricket, was just a flight away from missing out on his own heroics. In his own words, he’d had a good trial with Punjab Kings, and booked a ticket for another trial, until a phone call from the franchise held him back. And as luck would have it, he got snapped up for a mere Rs 20 lakh during the auctions. This was down to good old talent scouting. Punjab Kings’ director of cricket Sanjay Bangar, a former Railways stalwart himself, had hit up his networks to spot Sharma—a story not dissimilar to Jasprit Bumrah’s arrival in Mumbai Indians courtesy of John Wright.

Success stories like Sharma or even Bumrah could be passé in a few years. If trends in sports such as football and baseball are any indication, artificial intelligence (AI) is already playing a key role in decision-making for large franchises, especially in talent identification and scouting. If industry chatter is anything to go by, at least two IPL franchises are using AI-based iPhone apps to evaluate players in trials. The Royal Challengers Bengaluru, for instance, has tied up with Ludimos, a 360-degree coaching platform, for their scouting needs, citing paceman Avinash Singh as a success story. With the budgets that IPL franchises have for scouting—sometimes as high as Rs 90 crore heading into the mega auctions—industry insiders feel AI models can potentially be deployed to pick the right talent for the right price and budget.

That AI is making waves in sport shouldn’t be a surprise, given the amount of data generated by each game. For instance, sports performance analysts estimate that each event in cricket can generate at least 100+ data points, of which a majority (60-80) are logged by either the on-ground performance analyst or the backend provider. Ditto for Formula 1, where per Amazon Web Services (AWS), “every F1 car contains 300 sensors, which generate 1.1 million telemetry data points per second transmitted from the cars to the pits.” 

Today, the sports AI industry is pegged at $5.93 billion and is poised to grow to $21 billion in 2029, according to market intelligence firm Mordor Intelligence. Close to a billion dollars ($842 million) has flowed into global startups that specialise in sports AI, according to Tracxn data. India alone has about 35 companies operating in the space, making it the third-largest geography for sports AI companies. Beyond startups, the growing need for data analytics has attracted several large technology companies such as SAP, IBM, and AWS, specialising in verticals dedicated to sport.

But is the hype real? Or is sport also poised to find itself in the fast-building AI bubble? The Playbook contacted Subramaniam Ramakrishnan (Ramky), as he’s known in sporting circles, to find out more. Chennai-based Ramky is the founder of SportsMechanics, a sports analytics company that works with several major cricket boards including the Board of Control for Cricket in India (BCCI). He was the Indian cricket team’s first performance analyst in 2003 when Wright was India's head coach. 

“The way we identify talent can be augmented with AI, and new age talent identification measures will spring up. Talent identification today is pretty much around the skill, but we need to go beyond and discover the god-given unique abilities,” Ramky tells The Playbook in a telephonic conversation.

The free hits and misses

Last month, the science journal Nature published a paper about TacticsAI, an AI-based model Google’s DeepMind developed with football experts at Premier League club Liverpool to analyse corner kick routines. The model, the authors said, helped Liverpool’s coaches trial and pick alternative player positionings for corner kicks, and pick the most effective ones basis the highest likelihood of success, as predicted by it.

The kicker from the study (pun intended) was this: “For 45 of the 50 situations (90%), the human raters found TacticAI’s suggestion to be favourable on average (by majority voting).” This might seem quite the leap, with in-game tactics and mundane (yet critical) events such as set-pieces becoming an important AI-led intervention in football. Similar use cases have emerged in other sports. Major League Baseball (MLB) tied up with UpLift Labs, a biomechanics company that analyses player performance and movement using 3D motion tracking and a predictive AI algorithm for player potential. 

But cricket might be a bit of a holy grail beyond basic applications such as analysing player performance. Why is that? As a sports analytics industry insider, who spoke on the condition of anonymity, puts it: “AI will never reach those levels for analysis, because there are too many variables in cricket. Baseball, basketball, and football don’t have to worry about the nature of pitches, time of day, weather etc. What AI can help with, however, is automating data collection.”

Not just that, there are other issues around rights. For instance, the International Cricket Council may not grant a green light to collect data. The other obvious challenge is the centralisation of data, prevalent in other leagues such as the NBA or MLB. That, this person adds, is a logistical nightmare, given the absence of rigorous data collection in the grassroots (cricket academies) beyond official BCCI and state association tournaments.

But the grassroots is precisely where AI could be deployed, particularly with coaching and eradicating some systemic issues. The post-pandemic eruption of sports coaching apps has, to some extent, played its part. Coaches in several leading sports academies use these apps not just to improve their wards’ performance, but also to predict future potential based on data captured from training sessions. This is what Siddhartha Datta, co-founder of sports tech company Tech At Play, calls a “paradigm shift”. 

“Only one coach attends to 60, 70, 80 kids and therefore, [they] can’t pay attention to every single one of them. So this is where you introduce that sort of automation, wherein a camera which records automatically, edits and analyses the footage, gives them [the kids] a dashboard saying, ‘You played several balls, [of which] 10 were middled, and 20 were edged’. And that edge is backed up by a video of the foot movement or what they did. It’s as simple as that. The coach can go back, look at the report, and say, ‘This is your error, and I’ll help you improve that error’.” Datta had said to The Signal Daily podcast last month. You can listen to the interview here.

A similar grassroots use case would involve eradicating age fraud—a rampant issue in India—across sporting disciplines. “Age frauds are the biggest bane in Indian sport, and AI can help solve the problem to a great extent,” says Ramky of SportsMechanics. “The central government has mandated age testing as part of its National Code Against Age Fraud in Sport. Where AI can come in is with better interpretation and augmentation of the collected data,” he elaborates.

The Wild West

While these applications might be imminently achievable, there’s a storm brewing in the UK (and the West in general) over a fundamental question: who owns the data? In 2021, former Cardiff City manager Russel Slade led an initiative called ‘Project Red Card’, where he and 850 professional footballers threatened legal action against betting companies for illegal use of player performance statistics and personal information. Slade’s bone of contention was that players do not receive payments for the usage of data they hadn’t consented to, a key aspect covered by both the UK Data Protection Law and Europe’s GDPR.

Also in sports, there are multiple kinds of data. One, public data (available through official organisations or event organisers). Two, observable data (which you can see, either via the rights holder or otherwise). And three, personal data (this would include physiological data such as heart rate or body temperature).

The Playbook spoke to Nandan Kamath, founder of LawNK, a Bengaluru-based sports law firm and author of the recently released Boundary Lab: Inside The Global Experiment Called Sport. “This is going to become a real battleground in data rights in sports,” he says. “Players feel that the data is theirs as it relates to them and simply because others can record and measure it doesn't mean they own own it. This is also the case when private biological training data is combined with observable data for insights. Historically, when a player signs for a club or a franchise, s/he signs primarily to render playing services, not as a data provider but the balance of expectations is rapidly changing.” Kamath adds. 

There’s also the AI-led menace of fake endorsements, where star cricketers have seen their images and voices used for betting and related advertisements. That again, is down to the absence of a clear legal framework. 

“What we need is a publicity right framework that recognises rights in name, image, and likeness. And a combination of data ownership and rights, where the person can control the commercial use of various private attributes,” Kamath concludes.

⚡️Quick Singles

📺📈🏏: The ongoing IPL season garnered a cumulative reach of 44.8 crore viewers over the 22 days of the tournament, according to official broadcaster DisneyStar. This, Star said, represented an 8% growth over the previous season, amounting to a cumulative watch time of 18,800 hours. The IPL’s digital broadcaster JioCinema revealed that the match between the Mumbai Indians and defending champions Chennai Super Kings (CSK) logged over 71.5 crore views. This number, the company said, was higher than the reach of the IPL 2023 final between Gujarat Titans and CSK. 

🎱🇸🇦🆕: Seven-time world champion Ronnie O’Sullivan will open a snooker academy in Saudi Arabia, while also doubling up as the Middle Eastern kingdom’s ambassador for the sport. The Irish legend participated in the Riyadh World Series Masters last year, which was also Saudi Arabia’s first-ever snooker event. O’Sullivan has signed a three-year deal with Saudi Arabia and will play in all World Snooker Tour events hosted by the country.

🏀💰 ✅: Women's basketball sensation Caitlan Clark is on the verge of signing a $20 million+ endorsement deal with Nike, The Athletic reported. The report added that Under Armour and Adidas were also courting the star with “sizeable” offers. Clark’s deal with Nike will also see her getting a “signature shoe”. Clark has been a standout performer in the women’s NCAA Tournament, where she represented the University of Iowa. She was recently signed by WNBA franchise Indiana Fever as their No.1 pick.

🛑🚫⛔️: The Anti-Corruption Unit of the BCCI reportedly evicted suspected bookies from Jaipur and Mumbai, The Hindu reported. The individuals who were evicted from Jaipur were attending Rajasthan Royals’ match against the Delhi Capitals, whereas the Mumbai ones were asked to leave during Mumbai Indians’ clash against Rajasthan Royals. While a first information report was filed against the Jaipur lot, the report added that the Mumbai evictees may have been victims of mistaken identities.

🛑⛹️💨Toronto Raptors’ player Jontay Porter was handed a lifetime ban by the NBA for allegedly gambling on the sport. This is the first time the league has banned a player after betting was legalised in the United States. The life ban was the maximum penalty handed by the league to Porter, who was revealed to have wagered on his performances while also disclosing confidential information about his health.

📖 Weekend Reading

The inside story of Portsmouth’s promotion back to the Championship [The Athletic]

Murdoch Partner Helps Asia’s Richest Man Build A Media Empire [Bloomberg]

That’s all for this week. If you enjoyed reading The Playbook, please share it with your friends, family, and colleagues. Please also subscribe to it (for free) if you haven’t already.

You can reach out to me at [email protected] with any feedback (good, bad, or ugly), tips, and ideas. I'd love to hear from you!

Thanks for reading, and see you again next Friday!