On the one hand, sports are an engaging type of entertainment that connects different cultures around the world. On the other hand, many types of sports are often complex multidimensional systems that incorporate a plethora of data points that make one team or athlete better than the other. AI-enabled technologies like computer vision are poised to have an impact on all parts of this equation.
Let’s see how the use of computer vision software in sports is changing fans’ viewing experience and impacts the ways athletes analyze their performance.
In many team sports, in-depth analytics enable teams to have the edge in strategy, scouting, and performance benchmarking. Conventional game statistics that provide dry summaries of goals, assists, and shots are not sufficient to understand the performance on a granular level. With AI and computer vision, we can now understand the root cause of a game outcome, be it a specific player’s condition or move.
One of the first companies that started to dig deeper into AI-enabled analytics is Sportlogiq. Headquartered in Montreal and founded by former Olympic skater Craig Buntin and Ph.D. graduates in computer vision and machine learning, Sportlogiq helps hockey, soccer, and football teams make smarter decisions by utilizing deeper insights.
The company uses computer vision-enabled cameras that analyze specific game events and nuances like players’ movements, ball trajectories, shots, and passes. Then this data is converted into meaningful reports for teams’ coaches, commentators, scouts, analysts, and even sports betting companies. Sportlogiq’s value is indisputable as now its technology is used by more than 60 teams in professional hockey, football and soccer teams.
Ensuring Racing Safety
With an average speed of 200 mph, at least one fatal accident during the NASCAR season shouldn’t come as a surprise. Every driver that participates in the competition fully understands the risk they are taking. However, getting in an accident because of a malfunctioning car is a whole other story. Ford in collaboration with Argo AI has managed to create a deep learning neural network that helps increase NASCAR safety measures.
2. Fundamentals of AI, ML and Deep Learning for Product Managers
3. Roadmap to Data Science
4. Work on Artificial Intelligence Projects
The reason for Ford’s foray into the AI world was originally all about self-driving cars. During development, the team noticed that its algorithm was able to effectively identify the specific models of cars, especially when it came to blurred images. Given that NASCAR pushes the limit in terms of maximum speeds, this capability of the algorithm has proven to be especially relevant. The network can now quickly detect cars with malfunctions, which allows them to prevent serious accidents and reduce risks.
Making Smarter Scouting Decisions
Besides intuition, the major factor of sports scouting success is making sense of lavish amounts of data presented to analytics departments. Last year, the Orlando Magic turned to sports analytics and technology company STATS to find the next big NBA star. The AutoStats software collects players’ data directly from video footage with the help of AI and computer vision.
Capturing and analyzing players’ data from cameras is not new in big sports. However, previously, it was only possible with special in-venue cameras installed on the premises of official NBA courts and few colleges, which made analysis limited. AutoSats can now source players’ data from any recorded game, which significantly expands the analytics capabilities and makes the scouting process much less complex and more reliable.
The global player tracking market is expected to grow at a CAGR of 24.9% during the next five years. A few months ago, Boston Celtics Executive Director of Performance Phil Coles said that he believes there is a ‘significant place’ for AI and data in the NBA. However, it’s too early to make any definite statements about these technologies and their impact on scouting outcomes as only the Orlando Magic has truly committed so far.
Enhancing Shooting Accuracy
Given that the NBA’s shooting accuracy percentages haven’t been improving for the past twenty years, Alan Marty, the Noah Shooting System’s founder, has recognized that the players might need an AI-assisted boost for improving their shots.
The Noah System uses small computer vision-enabled cameras and sensors that are installed in the rafters above the baskets and track dozens of shooting analytics including players’ locations on the court, ball rotation speed, shot’s arc and shot depth. The Noah Shooting System gives instant audio feedback, allowing players to correct their shots during training in real time. For example, if the System is set to assess the player’s shooting arc, it might say just ‘39’, given that 45 degrees is the optimal number.
With over 100 thousand shots recorded across many courts every day, the System is continuously improving, allowing coaches and players to understand shooting on a granular level. The System now not only records ball trajectory data but also analyzes the players’ shooting biomechanics. Currently, the Noah System has been adopted by nearly half of the NBA teams and by about 30% of the NCAA. With teams reporting significant improvements, it’s only a matter of time when the Noah Shooting System becomes a standard tool in the professional basketball.
Improving Strategy and Viewing Experience
When it comes to professional leagues, basketball is much more about strategy and team play than the right shooting technique. Second Spectrum, an Official Tracking Provider for the NBA, uses computer vision, AR, and players’ historic data to superimpose graphics and statistics, like the probability for a player to make a successful shot. Second Spectrum’s Court Vision makes basketball broadcasts much more engaging for fans and provides comprehensive analytics for teams on the court.
With such advanced technologies disrupting strategy-heavy games like baseball and basketball, there is that nagging question: will the competition between teams reduce to the competition between algorithms?
Well, while the best chess player in the world is currently a computer, this won’t ever happen in games where the human factor is critical. AI algorithms can help coaches adjust the strategy but will never be able to motivate the players or predict Curry’s every long-range three-pointer. There is simply too much randomness in athletic sports for AI algorithms to understand.