- Using data to the organization’s advantage. The sports entertainment industry captures an enormous amount of data - as organizations continue to be more digital its important that they use this data to their advantage.
- Improving processes. Implementing AI and ML to improve processes within the business is essential to drive efficiencies and improve the overall game.
- Mitigating risks with technology. AI and ML can be used within many areas of the business, such as officiating where the risk of gambling scandals are increasing due to the rising popularity of sports betting in North America.
Our Advice
Critical Insight
Digitally transforming different processes through technological advancements such as officiating is critical to improve the integrity and fairness of sports, where strategically combining people and technology processes will enhance the lives of officials.
Impact and Result
- Learn how AI and ML work, their algorithms, capabilities, and how they affect the entire business and officiating.
- Review the considerations, potentials, and limitations any organization should be aware of when implementing AI and ML for on-field decision-making.
- Consider the type of architecture, implementation model, vendor, and strategy needed to successfully transform officiating.
- Visit Info-Tech's AI Research Center to kick start an AI and ML implementation process.
Digitally Transform the Process of Officiating in Sports
Utilize AI and ML to Innovate On-Field Processes
Analyst Perspective
Transform sports through digital and data driven processes
Elizabeth Silva
Research Analyst, Sports Entertainment Industry
Info-Tech Research Group
Artificial intelligence (AI) and machine learning (ML) are quickly changing the landscape of the sports industry all the way from media and fan experience to management and operations, as $1.6 billion has been invested in AI and ML within the sports industry to date (Neoteric, 2022).
AI and ML are both effective changes for the future; they will have an influential impact on digital transformation and innovation within many industries, including sports entertainment. As AI and ML have many uses within sports, its important to understand each use case individually to determine if it is a fit for the specific organization.
Officiating in sports has always been controversial; fans and spectators can be skeptical of some of the decisions officials make during a game. Unfortunately, there is a history of gambling scandals and other bad decisions made by referees that set a precedent for skepticism.
This poses a great challenge for technology to overcome by using digital, data, AI, and ML to drive more efficient and trustworthy processes. By combining people and technology processes, the integrity and fairness of sports will improve, while enhancing the quality of the game.
Executive Summary
Your Challenge
Using data to the organization's advantage. The sports entertainment industry captures an enormous amount of data. As organizations continue to digitize it's important that they use this data to their advantage.
Improving processes. Implementing AI and ML to improve processes within the business is essential to drive efficiencies and improve the overall game.
Mitigating risks with technology. AI and ML can be used in many areas of the business, such as officiating where the risk of gambling scandals are increasing due to the rising popularity of sports betting in North America.
Common Obstacles
AI and ML are not a low-cost investment. This creates an intimidating investment for a business to decide on before fully investigating its current state and potential capabilities for the organization.
It is difficult to understand how AI and ML work. Due to their advancements over the years it can be challenging to comprehend.
AI and ML are not broadly used for officiating. This can make it challenging to determine how it could or would be used to assist officials in making decisions.
Info-Tech's Approach
Review the considerations, potentials, and limitations any organization should be aware of when implementing AI and ML for on-field decision-making.
Consider the type of architecture, implementation model, vendor, and strategy needed to successfully transform officiating.
Visit Info-Tech's AI Research Center to kick start an AI and ML implementation process.
Info-Tech Insight
Digitally transforming different processes through technological advancements such as officiating is critical to improve the integrity and fairness of sports, where strategically combining people and technology processes will enhance the quality of the game.
The future will be driven by AI and ML
Digital transformation is an at-scale change program – planned and executed over a finite time period – with the aspiration of creating material, sustainable improvement in the performance of an organization. This is done by deploying a programmatic approach to innovation, along with enabling technologies, capabilities and practices that drive efficiency and create new products, markets and business models.
Artificial intelligence (AI) and machine learning (ML) will drive the future due to the incredible impact they will have on digital transformation and innovation within the sports industry.
When considering the future of innovation for the sports market, there are various digital potentials that can improve the overall industry to drive efficiency and create new products.
A study within the United States revealed that 65.5 million people are sports viewers.
It is expected that by 2025, 90.7 million people will be considered sports viewers (Statista, 2021).
Problem solve with data-driven AI
The success of any sports league or team is determined by its financial strength. The outcome of each game is important to the overall business.
This creates significance on how fair each game is, but there are always random factors that play into the outcome of a game.
Some factors are:
- Incorrect decisions made by officials (referees, umpires, governing officials)
- Weather-related influences
- Unevenness of the pitch
- Absence of players
With all these factors, there is only so much that can be controlled. However, those factors that can be improved through the use of digital transformation, such as assistance in decision making for officials, should not be ignored. Wrong decisions by officials put every sport at a disadvantage to varying degrees. The minimization of wrong decisions is essential.
Officials may make incorrect decisions based on a variety of components, such as the following:
Match manipulation (intentional)
Social pressures, leading to psychological stress and mental illness (unintentional)
Genuine miscall from human error (unintentional)
The use of more technology can assist with these problems by improving the integrity and fairness of sports games while making the lives of officials easier.
…of fans vote for the continued use of VAR (video assistant referees) in soccer (football), despite the criticism it receives due to interruptions it causes, as VAR was created to reduce the amount of clearly wrong decisions referees make.
International Journal of Innovation and Economic Development, 2020.
Machine learning and artificial intelligence cohesively work together in decision making
Artificial intelligence performs tasks mimicking human intelligence. AI is a combination of data-driven technologies that includes tools such as machine learning, where it learns from experience and problem-solving.
Machine learning systems learn from experience. They learn patterns from data, then analyze and make predications based on past behavior and the patterns learned.
Deep learning is a class of machine learning algorithms, that can be learned in supervised or unsupervised manners. Most are based off artificial neural networks, which are inspired by the structure of the brain.
Statistical reasoning | Machine learning |
Artificial intelligence |
---|---|---|
Infer relationships between variables | Make accurate predictions | Adapt dynamically to innovation |
Statistical models are designed to find relationships between variables, and the significance of those relationships. | Machine learning is a subset of AI that discovers patterns from data without being explicitly programmed. |
AI systems choose the optimal combination of methods to solve a problem. They make assumptions, reassess the model, and revaluate the data. |
Download the Get Started with Artificial Intelligence Blueprint to Learn More
Leverage AI and ML for simple to complex decision-making
AI and ML have both been heavily developed to assist with more complex decision-making. Become familiar with the process of how AI and ML work, in order to best leverage them within the organization, beyond officiating.
Download the Get Started with Artificial Intelligence Blueprint to Learn More
Source: International Journal of Innovation and Economic Development, 2020.
AI and ML have advanced their capabilities
The Evolution of AI and ML in sports officiating
,AI has had a role in innovating the process of officiating since 2001.
The role of AI and ML has only gotten larger over the years; most sports today are using AI and ML in some sort of form.
In time, the industry will recognize the capabilities of this technology to be a necessary means for the future.
Source: Hawk-Eye Technologies, 2022
AI and ML have advanced their capabilities
The evolution of AI and ML in sports officiating continued…
37% of those working in sports entertainment are in favor of more AI and ML adoption.
Compared to
66% Of those working in the technology sector, who are in favor of more AI and ML adoption.
This showcases the lack of awareness or hesitancy sports industry leaders have on the capabilities regarding AI and ML.
Statista, August 2022.
Source: Hawk-Eye Technologies, 2022
The natural decision-making process for officials has many influencing factors
There are various factors that influence the final decisions an official may make for in-play situations. Influencing factors typically have more negative than positive effects on the outcome of a game. This is a major area that AI and ML can assist in since they are driven by data, able to make non-biased and unaffected decisions. When AI and ML are used to assist human officials, they enhance the decision-making process enabled by people and technology processes.
Illustrated below is the common decision making process a human official experiences, including influencing factors.
Adapted from: International Journal of Innovation and Economic Development, 2020.
Digitally transform the process of officiating with AI and ML
Illustrated below is the enhanced decision making process enabled by people and technology processes of AI and ML.
Adapted from: International Journal of Innovation and Economic Development, 2020.
Considerations for any AI and ML implementation
Involve all stakeholders in the change of processes
This can be spectators, fans, referees, and teams in order to gain acceptance for the implementation of any new technology in the sport.
Invest in data collection and analysis to train AI, with examples
Collecting examples and counterexamples for or against a decision is essential to train an AI. This will allow for real-time decision making through AI, as it can send results to officials on wearables like a watch, to be evaluated and implemented within the game by the referee.
AI should initially only be used for simple decision making
AI is better suited to make decisions for simple decisions (events that are strictly governed by rules).
The official, as the human being, should always have the ultimate decision-making power
AI technology should be tested for several years before complete implementation into any game. However, even after the test phase is completed, it is essential that AI processes are transparent in sports to gain acceptance. It is always recommended that AI can assist an official in decision making, but at the end of the process, the human should make the final decision.
Source: International Journal of Innovation and Economic Development, 2020.
Interpretable decisions require understanding of the situation and therefore belong to the official
Officials should always remain in power for discretionary decisions (events that are not governed by rules but are decided by people in authority, who evaluate each event).
The global artificial intelligence sports market is projected to reach a value of $19.2 billion by 2030, growing at a CAGR of 30.3% from 2021 – 2030.
(Allied Market Research, 2022).
The capabilities of AI and ML in officiating vary
AI and ML are growing at a rapid rate, with new developments and capabilities every year for a range of sports. Every new development is a step forward to improved games through data and technology.
The NBA identified that 8.19% of decisions were incorrect or missed calls. This included 26,822 plays of 1,476 games between 2015 to 2018. In those 4,297 minutes of action, the officials missed or incorrectly called 2,197 plays (8.19%). This may not seem like a large percentage, but in sports it is. (v7labs, 2022)
To improve the fairness, integrity, and quality of sports games it's important that the sports industry begin to embrace the capabilities that AI and ML can offer, to mitigate the number of incorrect calls in a game.
The following slides indicate the common capabilities that AI and ML can currently provide to a range of sports. It's also important to consider the limitations of AI and ML, as there are certain capabilities that cannot or should not be done by technology.
The common limitations of AI and ML across all sports are:
Feelings | Unavailable data | Other |
|
|
|
Source: International Journal of Innovation and Economic Development, 2020.
AI and ML have the following potential capabilities
Player ID by jersey number/facial recognition | Detection of puck/ball | Alignment and placement of body and/or team | Determining whether the game is in or out of play | Neutralization of psychological influencing factors | Determining the game is correctly resumed | Replay technology to assist reviews | |
---|---|---|---|---|---|---|---|
Hockey | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Soccer | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Basketball | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Baseball | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Football | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Golf | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
Tennis | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Car racing | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Horse racing | ✓ | ✓ | N/A | N/A | ✓ | ✓ | ✓ |
Rugby | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Cricket | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Source: International Journal of Innovation and Economic Development, 2020.
Legend : AI and ML have the potential for the capability (✓~): AI and ML have the potential for this capability, but it may not be completely effective due to the nature of the game
AI and ML have the following potential capabilities
Detection of situations outside the range of perception |
The probability analysis of an in-game point/score |
Use of analysis tools for game preparation |
Contact identification (e.g. side and finish lines, nets) |
Detection of clear offence/foul |
Contact of a tactical offence/foul |
|
---|---|---|---|---|---|---|
Hockey | ✓ | ✓ | ✓ | ✓ | (✓~) | (✓~) |
Soccer | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Basketball | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Baseball | ✓ | ✓ | ✓ | N/A | N/A | |
Football | ✓ | ✓ | ✓ | ✓ | (✓~) | (✓~) |
Golf | ✓ | ✓ | ✓ | ✓ | N/A | N/A |
Tennis | ✓ | ✓ | ✓ | N/A | N/A | |
Car racing | ✓ | ✓ | ✓ | N/A | N/A | |
Horse racing | ✓ | ✓ | ✓ | N/A | N/A | |
Rugby | ✓ | ✓ | ✓ | ✓ | (✓~) | (✓~) |
Cricket | ✓ | ✓ | ✓ | N/A | N/A |
Source: International Journal of Innovation and Economic Development, 2020.
Legend : AI and ML have the potential for the capability (✓~): AI and ML have the potential for this capability, but it may not be completely effective due to the nature of the game
AI architecture plays a critical role in enabling innovation
An organization needs an AI Architecture to provide the architect in the business with pre-built components that can be used to design the application.
Download the Create an Architecture for AI Blueprint to learn more
Note: The layers of this AI Reference Architecture may vary depending on the organization and use cases.
Explore implementation models and providers that can innovate the officiating process
There are two AI implementation models:
- AI/ML product or tools can be built from capabilities and may require training as part of the implementation.
- Off-the-shelf AI and ML models are pre-built, pre-trained, and pre-optimized for a particular task.
When deciding which model to go forward with, its important to remember that there is no need to reinvent the wheel and build a product you can buy. Using an off-the-shelf AI model enables an agile approach to system development, with faster POC and validation of ideas and approaches.
Consider the following vendors to purchase an off-the-shelf AI and ML model:
![]() |
![]() |
![]() |
![]() |
---|---|---|---|
A global leader in live sports broad production, officiating, and performance through innovative technology. |
State-of-the-art machine learning and computer vision techniques, servicing as the official optical tracking partner for several leagues. |
With a vision to build and improve the game of sports at all levels, Catapult designs wearable technologies, video analysis, performance data and video, and more. |
SciSports uses data intelligence to understand soccer, in order to improve the game on the pitch while enhancing the fan experience. |
Ensure AI and ML implementation is strategic
Consider the following components for a strategic approach to an AI/ML tool or off-the-self model.
Download the Drive Business Value with Off-the-Shelf AIBlueprint to Start your AI/ML Implementation
Next Steps
Digitally transforming the officiating process is not an easy decision or implementation.
Info-Tech has the right research and tools to strategically assist the organization through this choice and development.
Visit Info-Tech's AI Research Center to kick-start an AI/ML Implementation with the following methodology:
Visit Info-Tech's AI Research Center
Research contributors and experts
Anonymous Contributor
Business Executive
AI Solution Provider
Eman Smadi
Sports Data Scientist
Own The Podium
Any Ganesh
Technical Counselor – Architecture, Data Science, ML and AI
Info-Tech Research Group
Imad Jawadi
Senior Consulting Manager
Info-Tech Research Group
Bibliography
Beesetty, Y et al. "Artificial Intelligence in Sports Market by Component," Allied Market Research, Feb 2022. Accessed 21 July 2022.
Catapult. Accessed 15 September 2022.
Gottschalk, Cedric et al. "The Innovation of Refereeing in Football through AI," International Journal of Innovation and Economic Development, July 2020. Accessed 5 September 2022.
Hawk-Eye Innovations. "About," Accessed 15 September 2022.
Luzniak, Karolina. "AI in Sports – How Will Artificial Intelligence Change the World of Sports?" Neoteric, 13 May 2022. Accessed 21 July 2022.
Rizzoli, Alberto. "7 Game-Changing AI Applications in the Sports Industry," v7labs, 21 Oct 2022. Accessed 21 July 2022.
SciSports. Accessed 15 September 2022.
Second Spectrum. Accessed 15 September 2022.
Statista Research Department. "Number of digital live sports viewers in the United States from 2021 to 2025," Statista, 19 Oct 2021. Accessed 15 September 2022.
Thormundsson, Bergur. "In the future, do you think artificial intelligence should play a larger role in each of the following fields or industries in 2022?" Statista, 2 Aug 2022. Accessed 20 September 2022.