The Rise of AI in Sports: How Algorithms Are Changing the Game

When Billy Beane introduced an analytics-based approach to the 2002 Oakland Athletics, he didn’t know he’d change baseball and sports forever. After all, you’ve probably heard the term Moneyball or watched the Brad Pitt movie.

Back in the day, sports were mostly about vibes. Scouts went with their gut, looking at players through the lens of the so-called eye-test. That all changed with the advent of advanced statistics, with data scientists proving that crunching numbers can be just as important as watching games.

20 years later, we’ve seen another revolution. With AI taking the world by storm, the number-filled world of sports is also sure to feel the impact of new technology. In this article, I’ll take a look at the impact AI has on the sports world – from how teams leverage it to make their work easier, to how it can impact the fan experience, and sports betting.

Key takeaways

AI’s impact on sports is ever increasing
AI turns more and more previously intangible qualities into analyzable numbers
By leveraging AI players can avoid serious injuries
Sports betting has seen AI at the forefront of a battle between regulators and matchfixing rings
AI makes producing content more cost-effective than ever for smaller teams
Whoever finds the best way to leverage AI in sports may bring about a revolution in how we think about sports

Turning data into plays

Ever since the advent of Moneyball, sports have quickly risen to the cutting edge of statistics and analysis. While in 1999, teams would sign players based on scouts telling them that “they have that dawg in them”, now scouts are expected to present an advanced metric proving that dawg is there.

Of course, this also varies from sport to sport. The more repeatable and formulaic a sport, the more analytics play a role in it. That’s why baseball has become a near-total numbers game, basketball is not far behind, but soccer or hockey still feature lots of intangibles that can’t quite be explained with a number with a memorable acronym.

AI, however, can fill in a lot of these gaps. For example, for years, a lot of data analysis was done based on pure statistics. As machine learning became stronger, these statistics were expanded with new additions, breaking down the game into more and more detailed statistical elements. In fact, sports analytics was one of the first fields where AI concepts were being used on a regular basis.

The biggest advantage of AI is that it allows for rapid analysis of complex data sets in a way that was previously impossible. While analysts could come up with some of the same conclusions after some time, AI can basically figure it out on a play-by-play basis, giving advice in terms of biomechanics, player form, and more.

For example, a few years back, the NBA introduced CourtOptix – an AI system that captures over 10 million data points every game to generate stats like a defensive pressure score. Now, that sounds simple, right? Well, it’s not. Defense is famously hard to quantify in basketball, especially when it comes to the team game. Thanks to CourtOptix, coaches can have precise data on each player’s positioning and their impact on the offense.

Similarly, baseball coaches are now using AI to quickly analyze each player’s batting tendencies across multiple types of pitches, allowing managers to adjust pitching strategies on the fly.

Getting healthy

With the ever-increasing load on players’ bodies, many sports have seen a massive increase in fatigue-based injuries. This doesn’t just ruin a team’s season, but also hurts fans who often spend weeks eagerly awaiting seeing a star, only to find out that they’re not gonna see their favorite player because of a nasty tear or sprain.

More and more teams are introducing AI-based computer vision into their training regimens to help their players align their biomechanics in a way that will reduce load on their bodies and help them avoid injuries or recover from them quickly.

It’s very possible that as the influence of AI over sports increases, the amount of injuries will finally start decreasing. However, the question is as always, how much will the players and coaches adhere to a machine’s words.

A double-edged sports betting sword

While on the field, AI is improving overall experience and acumen; off the field, it’s become a two-way battle between betting rings and regulators. Essentially, it’s become an arms race.

Sports betting has become a plague on sports. Match fixing is notorious in lower leagues, and in sports like basketball, betting rings can easily influence betting results by bribing a single player to score less points than average, or commit more fouls than they usually would.

With AI, betting rings can spot opportunities to make money and plan rigging games in a way that will be harder to flag by authorities… who are also using AI to detect betting discrepancies, and sudden spikes in betting volumes.

So far, the authorities have gotten ahead in this battle, with AI systems flagging thousands of suspicious bets and decreasing the overall impact of match fixing on sports. Then again, as betting rings discover how the system works, they’re likely to use AI to find a way around it. It’s going to be a near-constant cat and mouse game.

Finally, AI may also give confidence to people who don’t usually gamble. The prevalence of sports gambling has increased in the last few years overall, and it’s gotten to the point where recently Toronto Blue Jays Pitcher Trey Yesavage revealed that his family was harassed, which many people attribute to gamblers being disappointed by the rookie pitching an excellent game in the playoffs.

With chatbots like ChatGPT or Gemini being able to analyze games, more people might be inclined to test these predictions out with bookmakers, which may lead to them losing money, or like in the case of Yesavage, their temper.

A content revolution

A final way AI can change sports is by revolutionizing the way we watch it. New tools allow fans to create their own way of watching and interacting with sports. For example, if you’re a Nikola Jokić superfan, AI automation will allow you to quickly find all his plays from the previous game. Broadcasters also gain a lot of insights thanks to AI, and camera software decreases their reliance on manual camera control, allowing broadcast teams more time to work on presentation.

The same stats I’ve talked about before can also be available to broadcast teams. With AI commentators will be able to give more insight than ever to viewers. This in turn will result in less dead air during games, and help viewers understand the intricacies of the game – often benefiting less flashy players.

Finally, AI is an amazing opportunity for smaller teams to be able to set up their own broadcasts without a dedicated media team. Companies like Pixellot and Veo Sports offer specialized cameras that allow smaller organizations to create a fully automated broadcast based on as little as one camera. This in turn can help them build a following and fund future success.

The future

While sports have been on the cutting edge of technology for a while, there are still many ways in which AI can impact sports. Some of them may even seem dystopian, and will surely raise ethical concerns, but are possibilities nonetheless.

The deepest of dives

AI can allow sports teams to dive as deep as getting genetic data in order to tailor training programs to their players needs. It’s not a great reach to say that in the future, similar data may be used in prospect evaluation, and may result in some serious ethical concerns, both on the side of teams using such evaluators, and on the side of sports-centric parents who may want to use genetic tools to push their children to success.

Million-to-one

Another option is quite a bit less privacy-invading. With the amount of data gathered in modern sports, it’s not hard to imagine creating fully-fledged digital doubles of top athletes. These can then be used for things as simple as video games, but also to simulate the best outcomes for training regimens, in-game tactics, and more. Basically, a powerful-enough AI can turn coaches into Dr. Strange, finding that one perfect tactic to defeat their opponents.

Robo refs and AI umps

Finally, another option often talked about is automating refereeing. Technology already aids a lot of decision processes on the field, with goal-line technology in soccer, and play reviews available in most major sports. But AI can take it up another notch. With the right amount of data, a long VAR offside check in soccer, can turn into a quick decision made by a computer model that just knows whether a player was offside. No drawing lines, no discussion about the moment of the kick. Similarly, in sports like baseball, an AI can be far more effective than an umpire at distinguishing balls from strikes.

Then again, there’s an inherent risk here. Refereeing is as much about binary decision making as it is about understanding the flow of the game. Recognizing when a line is crossed can be hard for an AI. After all, in many sports players are constantly fouling each other on paper, but the referee only steps in when he feels that a certain level of physicality has been reached. And yeah, they make mistakes sometimes, but would you rather lose due to human error, or a buggy line of code?

Risky business

Risks related to referees aren’t the only risks that come up when talking about AI in sports. As always, AI has a major impact on privacy, and with how much sensitive biometric data is collected about players, you have to wonder when it’s simply too much.

Finally, a badly trained algorithm can impact the quality of not just a game, but a player’s entire career. As sports become more and more based on stats, rather than human touch, a missed aspect of a game can cost players millions of dollars. This was even recognized by the NBA, as this season it stopped registering end-of-quarter half-court heaves as missed shots. Why? Because if they missed, their 3-point shooting percentage would go down, and a mindless algorithm could present them as a lesser player to decision makers. Now, these shots won’t count towards their stats.

Similar issues can occur with virtually every other aspect of gameplay. After all, sports are also based on intangibles. Will an AI be able to measure the impact of a veteran presence in the locker room? Or a positive player who always seems to be able to pick his team up after a hard loss? I don’t think that these will be quantifiable with any amount of AI.

Conclusion

Sports is one of the fields where AI basics like machine learning have been used the longest, and yet a lot of untapped potential remains. Teams and players that learn to leverage AI the most are the most likely to succeed, while those less inclined to do so, might end failing similarly to how teams that haven’t adapted to the analytics era have.

Just like the story of Moneyball changed the way teams looked at stats and analytics, it’s very possible that soon another Billy Beane will find a way to optimize AI for maximum impact, bringing about an AI era to a specific sport, or in fact, all sports.

Until then, AI will be what it is in most cases. A tool that can be used for success, both by people who deserve it, and people who deserve it far less. To see which way it goes, I guess we all have to stay tuned.