By using AWS re:Post, you agree to the AWS re:Post Terms of Use

Win Probability and Gold Difference powered by AWS

1

I am a Computer Science student. I am really excited about "Win Probability" and "Gold Difference" powered by AWS being used in LOL

I have a presentation in school next month to introduce AWS, and I would like to choose this product as the presentation topic

Can anyone share more information about how to develop WP/GD and deploy it live?

My X: https://x.com/CL0UDER_MAN

My LinkedIn: https://www.linkedin.com/in/duy-chinh-hackerman-nguyen/

My Facebook: https://www.facebook.com/Truong.Vo.Ky.Clouder/

Thank you so much!

1 Answer
3

The "Win Probability" feature for League of Legends (LoL) esports broadcasts, developed in collaboration between Riot Games and Amazon Web Services (AWS), is an exciting application of machine learning in gaming. This feature analyzes key data points throughout gameplay to predict a team's chances of winning in real-time, enhancing the viewing experience for fans.

To develop and deploy the Win Probability feature, Riot Games worked with AWS to leverage in-game data and make real-time predictions using Amazon SageMaker. The process involves:

  1. Data Collection: Utilizing LoL Live Stats to continuously stream data from every game in the tournament as it's broadcast.

  2. Data Analysis: Evaluating various data points to track performance across individual players and overall team progress, including factors like gold accumulation, downtime, and unique gameplay events such as Elder Dragon kills.

  3. Model Training: Using historical data from the offline database to train the machine learning model.

  4. Real-time Predictions: Generating new predictions for every second of gameplay.

  5. Deployment: Implementing the model to provide live insights during broadcasts.

  6. Continuous Improvement: Regularly training the model with current data and using Amazon SageMaker to identify and compensate for data drift, ensuring the predictions remain accurate as the game evolves.

The collaboration between Riot and AWS resulted in a self-serve machine learning workflow that allows Riot's data scientists to focus on data exploration, modeling, and experimentation while ensuring accurate and reliable scaling.

For your presentation on AWS, you could highlight how this project demonstrates the power of cloud computing and machine learning in enhancing real-time sports analytics. It showcases AWS's capabilities in handling large-scale data processing, machine learning model deployment, and real-time predictions, which are applicable across various industries beyond gaming.
Sources
Riot Games and AWS bring esports ‘Win Probability’ stat to 2023 League of Legends World Championships broadcasts | AWS for Games Blog

profile picture
answered 18 days ago
profile picture
EXPERT
reviewed 18 days ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions