Table of Contents
When talking about the use of Artificial Intelligence in sports, the first thing that comes to mind is the “Moneyball” movie and the use of sabermetrics by the general manager of the Oakland Athletics team Billie Beane. What does this movie has to do with AI, you may ask? The thing is, sabermetrics helps evaluate players’ performance by using objective statistical measurements – and this sounds close enough to how AI helps evaluate one’s posture or behavior through object recognition.
“Moneyball” is based on real-life events that took place in 2002. SInce then, the technology evolved enormously so now the approach of Billie Beane is highly welcomed while back then it faced massive criticism and even hostility. How does AI influence modern sports and more importantly, will it replace some of its crucial elements that make games so enjoyable? Let’s find out.
The main fields of AI application in the sports industry
In 2019, the PwC company published a comprehensive research on the use of Artificial Intelligence in sports. While this research featured many aspects worth your attention, we’d like to focus on the two main areas of AI application across the sports industry as defined by the PwC. They are:
- Management and operations
- Media and fan experience
Management and operations includes all processes related to sport events but not directly to the game. These processes are venue management, ticketing, club and team management, sponsorship, and payments.
Media and fan experience includes content generation, betting, fan relationship management, and eSports.
You may ask: but what about the actual games? For that, PwC defines a game lifecycle and breaks it down into four segments.
- Talent selection and scouting: screening and selection of the most suitable players based on their performance and its AI-powered analysis.
- Pre-game preparation: includes training and coaching, tactical and strategic game planning, injury prevention, team assembly.
- In-game activities: umpiring, coaching, game analysis.
- Post-game activities: game analysis, injury management, recovery, game analysis.
As you can see, Artificial Intelligence can successfully be used throughout the whole lifecyle of a sport event and caters to needs of all parties involved, including not only the team and the coach but also fans and venue owners. And while there are dozens of its use cases, below we’d like to focus on the biggest ones that already shape the world of the sports industry.
AI referees and coaches
Modern sports already sees the wide use of technology for accurate decision-making when it comes to umpiring. There is VAR (video assistant referee), hawk-eye, and slow-motion replay and all of them are aimed at judging games more accurately through computer vision analysis. VAR, for example, involves three people working together to review decisions made by the main umpire. As for hawk-eye, this technology allows tracking the trajectory of a ball so a referee can make a correct decision about the sportsman’s actions.
Now seems like AI may come into play as well in a form of wearables (i.e. glasses) or overhead cameras in order to help referees make instant, real-time decisions during the game, without the need to pause it. In this way, Artificial Intelligence can not only provide a detailed analysis of a needed action (i.e. a goal) but may even recommend referees on the decisions to be made. And while full-fledged AI assistant are not there yet, we might see them in the near future, considering how willingly the sports industry embraces technology.
We don’t need to say how much training impacts the success of a player’s performance and we don’t need to say how an individual training plan is better than a common one. Here is where Artificial Intelligence comes into play and offers vast capabilities for personalized and smart training.
Due to the computer vision and machine learning technologies, a specialized application can analyze the player’s physical condition, their overall well-being, a recommended diet plan, and areas for improvement. Here are the training areas that AI can cater to:
- Nutrition: creation of personalized meal plans based on the goals and one’s current condition. Computer vision, for example, can analyze the products via the camera and immediately provide the user with their nutritional value as well as with recommendations.
- Physical and biomechanics: computer vision analyzes one’s posture and movements and can provide recommendations on improvement. This is highly valuable for improving one’s technique when it comes to practice.
- Mental: AI-powered apps can greatly contribute to stress management and help keep one’s stress levels low. And this is one area that is often overlooked in sports.
- Injury prevention and management: by analyzing one’s current state and their movements/posture, an AI-powered app can provide personalized recommendations on how to minimize or mitigate possible injuries.
Fans crave content related to their favorite teams but unfortunately, human journalists are physically inable to cover all local games and all events happening. Considering the fact that sport events happen on a daily basis, a huge part of games remains uncovered and simply ignored in favor for bigger events.
This can be fixed with Artificial Intelligence. AI-powered platforms like Wordsmith can translate hard data (i.e. statistics or scores) into narratives and overall AI is capable of generating natural-like text that can be presented to readers. Should we also mention that such texts can be crafted within minutes while people spend hours on them?
Of course, AI-written text will lack the individual touch of the writer and may not be as engaging. On the other hand, such texts are highly similar or almost unrecognizable from the ones people write and the use of AI in journalism will allow to cover even the smallest games and local plays. And that’s something many fans will actually be delighted to see.
Scouting and talent search
If we get back to the “Moneyball” movie, Billie Beane was signing undervalued players and his scouting decisions were based solely on the sabermetrics. And as it turned out, his approach brought Oakland Athletics (and later Red Sox) many victories.
Today, scouting can be significantly facilitated by using AI for analyzing the player’s potential and predicting how they’ll perform in a game. The computer vision technology can track the player’s movements, identify their strongest and weakest areas, and forecast how they’ll perform in the future. In this way, scouts can easier find undervalued talents, better understand what roles to assign to what players, and overall, assemble robust skilled teams.
Easier ticketing with biometric recognition
Entry delays are one of the biggest challenges that venue owners face – if you need an example, search for Southampton FC refunding fans in 2021. Because there is always a great number of people at the entry, it can be difficult to quickly check their tickets and perform security scans.
Biometric recognition (i.e. face recognition) can significantly speed up the ticketing process by immediately scanning the visitor’s face and matching it to the photo from the database. In this way, venue owners can minimize or prevent bottlenecks which is also great in maintaining healthcare standrads in regards to the pandemic. In addition, computer vision at the stadiums can be used to analyze the crows density and notify employees in case a particular area needs attention.
Better fan experience
Fans are as important to sports industry as players so great fan experience is vital. And this is one more area that can be greatly improved with using AI in sports.
First, there are smart chatbots that immediately provide users with information across a wide range of topics, from info about players to stadium logistics. Some chatbots are even equipped with the Augmented Reality technology so users can easily identify hotspots and key players in real time.
Second, there are smart video highlights that are aimed at identifying the most exciting and interesting games and showing them to fans, thus increasing their engagement. The thing is, manual analysis of games and hand-picking of the most exciting ones takes too much time and sometimes it’s hard to select games that draw the most attention. With the help of AI, it is now possible to analyze games by the crowd noise or players’ emotions in order to select the best highlights and generate the most relevant content for users. In addition, such AI-powered analysis significantly expands game coverage since it may include in highlights less known games that are still enjoyable to watch.
A case from SoftTeco: a Golf Club application
SoftTeco worked on an AI project aimed at analyzing the golf players’ postures and strokes to later design individually fitting golf clubs. The client is a golf club manufacturer and before approaching our company, he already used computer analysis for analyzing how players moved and how they held their clubs. However, the work was semi-manual and the client needed to fully automate the process of posture estimation and stroke analysis so he reached out to SoftTeco for help.
SoftTeco designed an application that uses two AI models: one for detecting the player’s position and the position of the club and the second model for segmenting the club image from its surroundings. The application delivers collects various metrics and based on them, delivers accurate results that the client later uses for designing individual golf clubs. Such individual clubs, in turn, contribute to better performance of players since they allow more powerful strokes due to the clubs fitting players by physical parameters.
Are we starting to lose human element in sports?
While AI seems to make umpiring more accurate and provides smart training plans for players, do we really need AI to invade all areas of the sports industry?
For many long-time fans, passing a turnstile is an essential part of the whole fan experience that starts the moment a fan enters a venue. As well, almost all fans anticipate those precious moments when a referee has to make a critical decision or when a minor human factor turns out to be the deciding factor for the game outcome.
So is AI making sports too sterile and predictive and is it taking away the joy of not knowing what will happen next in terms of the game outcome and final scores? Probably yes and such concerns should be taken into consideration. Maybe the best-case scenario is to use AI in sports for assisting but not replacing certain processes and leaving the human element intact as it is. Because after all, its those small imperfections that make millions of people around the world hold their breath when watching their favorite game.