Machine Learning and Artificial Intelligence now assist in everything from coaching to refereeing and broadcasting to betting in sports. Complexity in rules and AI biases delay the full-scale implementation of AI in various aspects of sports. However, neural learning has been improving the overall situation. Favorably, in sports, there is no shortage of data to be interpreted.
It was only yesterday that Japan’s top soccer league collaborated with Israeli AI company WSC Sports to capture key match highlights automatically. In England, IBM Watson has been helping Wimbledon do the same. In the US, a few weeks back, the automated track strike zone was implemented in baseball. A platform by Trackman also tracks various pitching and hitting parameters in baseball that include tilt, hang time, launch angle and bearing. In the game of golf, Trackman already track parameters such as smash factor, club speed, ball speed, attach angle and face angle. Let’s see the popular use cases of AI in sports.
Player Scouting and Player Analysis
Clubs and countries spend millions of dollars and euros to bring ace players into their teams. It is humanly difficult to apply quantitative metrics throughout the matches to analyze individual players. AI solutions use data provided by video coverages, wearables, and devices installed in stadia to analyze players. The individual parameters of players can be measured. With software, teams can check whether a player’s ability matches that of an expert the team wants. Similarly, with modern AI-based solutions, one can analyze the physiology of each player and how they function under stress.
Solutions That Help Coaches, Players
AI can be used to enhance the performance of players. For instance, the application Home court developed by NEX technology uses Computer Vision and Machine Learning to analyze Basketball players’ skill levels. It calculates the shot accuracy, progress over time and key performance metrics such as speed, vertical jump, release time and ball handling to help players improve. Similarly, US-based French inventor Grégoire Gentil designed a device called Tennis In/Out. The system uses Computer Vision to detect the speed and placement of a tennis shot. Similarly, in the game of cricket, sensor-enabled bats by Spektacom analyze batsmen’s performance. Parameters like the speed of the ball on impact, the twist on impact, point of impact, quality of the shot, etc. will be available on a real-time basis to coaches.
IBM collaborates with the U.S. Tennis Association (USTA) to provide insights to coaches. It can even quantify a player’s physical exertion and endurance to predict players’ performance in a match. Imitation learning is used by Soccer teams like Chelsea. The teams can understand the underlying decision-making policy of a player to judge his quality.
Formula One (FI) is among the most data-driven sports today. For instance, Mercedez race cars have around 200 sensors. In earlier days, engineers used to continuously monitor data from sensors for information on acceleration, braking, tyre temperature and wind speed to optimize performance. Today, AI does the job of interpreting this data. It gives the teams visualization on when to take pit stops. Similarly, Honda uses IBM’s Watson to analyze its hybrid engines. A car generates several hundreds of GBs of data during races. Amazon Web Services (AWS), now provides a cloud computing platform for FI teams to store these. Six decades of race data is stored by AWS and they are analyzed by teams to figure out the best tactics.
AI in Match Predictions
Machine Learning can be used to predict the results of matches. For instance, in soccer where massive sets of data are available, a model outcome can be created to predict future confrontations. An application named Kick-off claims it can predict the percentage of chances of winning sides of both sides. Similarly, Swarm AI Technology uses a hybrid AI where networks of humans think together with machines to forecasts match results. Another Tennis oriented solutions WinnerOdds helps in betting by calculating the real probabilities of all the tennis matches.
AI Boosts Sports Revenue, Fan Experience
Companies like Nielsen and SAS provide fan intelligence to stakeholders and can even analyze fan engagements. Corporations can measure the effectiveness of their branding on media using logo recognition and media monitoring technologies. This will help them run more effective campaigns and improve revenue.
Nielsen scans 500,000,000 social media posts and 1,100,000 interviews annually to offer fan insights to stakeholders. For this, they have acquired start-ups like Repucom and V brand which had come out with path-breaking AI solutions for sports analytics.
The take away for viewers from this Data Analytics are more personalized services. For instance, they can gain seats next to the most roaring spectators’ section of a stadium. Similarly, those who best utilize the season tickets can be identified and given updates over smartphones. According to a report by research and development company NTT around 54% of viewers were not satisfied with the sports viewing experiences. Analytics can provide crucial data to improve visual engagements through media.
VR and AR Engage Spectators
The upcoming Tokyo Olympics (2020) is expected to witness the large scale use of Virtual Reality, Augmented Reality and Mixed Reality to enhance spectators’ experience. Olympic Games committee in association with Intel will provide facilities like 3D tracking of athletes for spectators in Games venues. This will ensure that the audience will have readymade data on the profile of athletes competing in each event.
AI-based Facial recognition software called Neoface will be used in Tokyo to identify around 300,000 people including volunteers, media personnel, athletes and security personnel who are part of the games. Rather than training in empty stadiums, the staff and volunteers are being trained using VR. The games committee is also using Mixed Reality technology (which is a fine mix of Augmented Reality and Virtual Reality). Intel’s True VR was already used for broadcasting in the 2018 Winter Olympics held at Pyeongchang, South Korea. The Tokyo Olympics is expected to offer a much-advanced version of this.
5G to Help Connect Viewers and Technology
“Sport will be perhaps the first direct area of consumer impact for 5G technology, both on-field and for the consumers. The introduction of 5G brings with it a suite of products that can enhance the experience for both players and the viewers, including better intelligence gathering, game predictions, and most importantly game tracking,” said Jamal Hassim, the Founder and CEO of BOLT. Global.
“With digital streaming platforms entering the high-quality live content space, one of the critical areas they are focusing on is the viewer experience. Visualizations in sporting events have been largely two-dimensional in standard television broadcasting. Under the age of immersive experiences, 5G will enable for better filming technology rendering multi-dimensional viewing experience. Under current network speeds, it’s difficult to relay live TV and sport in multi-dimensions,” Hassim said.
The 2020 Olympics in Tokyo is truly going to be a technological spectacle as well as being a sporting one, with Japan already leading the efforts of integrating 8K broadcasting, 5G network capabilities, and 3D Athlete tracking. We can also expect VR to play a big role in immersive experiences in Tokyo,” he added
AI in Sports Broadcast
IBM Watson helps Wimbledon to extract highlights of the matches. Elements such as players’ reactions (celebrations) are analyzed to identify milestones in the game. For Wimbledon 2019, IBM claims it trained Watson to utilize acoustics data. It detects each strike of the ball for tighter automated cropping of highlights. Similarly, Pixellot video solutions flaunt special hardware with AI-based software to record and telecast matches. Devices with multiple lenses do a panoramic recording of the game and track the ball and player trajectory. The AI identifies game highlights like baskets or goals to automatically produce highlight clips. The solution comes with embedded editing tools for broadcast. Such software also offers coach modules for trainers to review players.
Today, every sporting body from NBA to NHL and NFL to NASCAR uses AI to boost the sporting experience. Apart from this, it adds to the spectator’s experience through ticketing software. Most of the sporting organizations in the West which are part of the $620billion global sports industry crave to expand to a newer market. It is evident that AI will be their best friend to achieve these goals.