AWS Only

Anthropic Models UsedSportsGen AI

IMG Arena Leverages Generative AI for Realistic, Engaging MLS Commentary

With the aim of enhancing its Major League Soccer (MLS) coverage, IMG Arena partnered with Firemind to develop a solution that generates engaging and realistic play-by-play commentary from raw event data. Firemind created an advanced automated system that processes match events, produces detailed commentary, and generates summaries in a natural American style, improving the viewing experience and providing valuable insights for various stakeholders.

AWS ONLY ·

IMG Arena Leverages Generative AI for Realistic, Engaging MLS Commentary

With the aim of enhancing its Major League Soccer (MLS) coverage, IMG Arena partnered with Firemind to develop a solution that generates engaging and realistic play-by-play commentary from raw event data. Firemind created an advanced automated system that processes match events, produces detailed commentary, and generates summaries in a natural American style, improving the viewing experience and providing valuable insights for various stakeholders.

At a glance


IMG Arena is a leading sports technology provider specialising in innovative content and data solutions, including live streaming, sports betting services, and data-driven insights for major sports leagues and media partners worldwide.

Challenge

IMG Arena needed a solution to generate realistic, engaging play-by-play commentary for MLS matches that maintained factual accuracy while filtering out less significant events.

Solution

Firemind developed an AI-powered system that processes match data to create natural-sounding, concise commentary, focusing on key moments and providing customisable summaries.

Services Used

Amazon S3
Amazon Bedrock
Anthropic Claude 3

Outcomes

Enhanced Viewer Engagement
Accurate and Reliable Commentary
Customisable Summaries for Versatile Use

Business challenges

Creating Engaging and Realistic Commentary for MLS Matches

IMG Arena aimed to elevate its MLS match coverage by providing play-by-play commentary that was both engaging and realistic. The challenge was to develop a system that could automatically generate high-quality commentary from raw event data without sounding mechanical or repetitive. The commentary needed to capture key moments such as goals and red cards while maintaining a natural, American sports broadcasting style. This required not only accurate data processing but also a sophisticated approach to commentary generation that avoided a monotonous, chatbot-like tone.

IMG Arena faced the challenge of developing a commentary system for Major League Soccer (MLS) that was both engaging and factually accurate. Traditional AI-generated commentary often struggled to strike the right balance between detail and entertainment, sometimes producing overly detailed or inaccurate descriptions that could detract from the viewer experience. IMG Arena needed a solution that could generate commentary in a natural, American-style voice, focusing on the most critical events of the match without overwhelming viewers with unnecessary information.

Additionally, the system had to ensure that every piece of commentary was backed by data, avoiding subjective language unless justified by the actual events. The goal was to create a commentary that felt authentic and engaging while maintaining a high standard of accuracy, ultimately enhancing the viewing experience for fans and providing valuable insights for broadcasters and other stakeholders.

"Firemind's PULSE tool was able to quickly and accurately summarise vast datasets of customer surveys, ensuring we could take action against any negative customer experience elements."

Dean Macfadyen, Data Platform Engineer — Generative AI Customer

Solution

AI-Driven Commentary with Customisable Summaries

To overcome IMG Arena’s challenges, Firemind developed an advanced AI-driven system that processes raw match data to generate realistic, engaging commentary tailored to MLS games. Using Anthropic’s Claude 3 models, the system creates human-like commentary by grouping events and filtering out less interesting actions, ensuring that the focus remains on key moments such as goals, red cards, and major plays. The commentary is designed to be concise and impactful, delivered in a natural American style that resonates with the audience.

Moreover, the system includes features for generating half-time, full-time, and custom minute-by-minute summaries, allowing users to quickly catch up on key events they may have missed. By combining accurate data processing with engaging narrative generation, the solution not only enhances the viewer experience but also offers customisable insights for various stakeholders, from broadcasters to sports betting platforms.

Realistic and Engaging Commentary:

Firemind's AI-driven system provided IMG Arena with the ability to generate play-by-play commentary that feels natural and engaging, closely mimicking the tone of an American sports commentator. The commentary focuses on key moments, avoiding unnecessary details, which keeps the audience engaged and enhances their viewing experience.

Precision and Accuracy:

The system ensured that all generated commentary was backed by accurate match data. By processing raw event data with precision, the solution delivered commentary that was both factual and contextually relevant, maintaining a high standard of accuracy across all outputs.

Customisable Summaries for Flexible Viewing

The solution included customisable features, allowing IMG Arena to generate half-time, full-time, and minute-by-minute summaries. This flexibility enables viewers to catch up on the match's most important moments quickly, providing a tailored experience that suits different viewing needs.

Model Spotlight


Claude 3 Haiku

For the IMG Arena project, we chose Claude 3 Haiku due to its optimal balance of performance, cost, and speed.

While Claude Sonnet offers enhanced reasoning capabilities, the simplicity of our prompts and the straightforward nature of the commentary generation process made Haiku a more efficient option.

Haiku was able to quickly process the raw match data, delivering high-quality, natural-sounding commentary without the need for the deeper reasoning layers that Sonnet provides. This decision allowed us to focus on speed and cost-efficiency without compromising the accuracy or engagement of the commentary.

Learn more

Why Firemind

Firemind’s ability to integrate advanced machine learning models into practical applications, coupled with their commitment to accuracy and engaging user experiences, made them an ideal partner for IMG Arena who’s objective was to transform raw match data into dynamic, realistic commentary that meets the high standards of professional sports broadcasting.

1920%

Workflow speed increase

1,000 documents in 12.5 minutes compared to 4 hours by human operator

88%

Keyword accuracy

for leading categories within data modelling & training

Added value

We were also able to take all stakeholders on a journey concerning how Cloud Adoption and Machine Learning would benefit their customers and their business continuity, all in a 12 week timeline from start to finish.

Get in touch

Want to learn more?

Seen a specific case study or insight and want to learn more? Or thinking about your next project? Drop us a message below!