Amazon Web Services, Inc. (AWS), an Amazon.com, Inc. company, and the National Hockey League (NHL) today announced Face-off Probability, a live, in-game NHL stat that will be displayed as a graphic that instantly shows the odds of a player winning a face-off and possession of the puck and displays them on screen for fans watching the broadcast of the game. Before the puck is dropped in a face-off, the Face-off Probability machine learning (ML) model identifies where on the ice a face-off is going to occur, and who will take the face-off, and determines the probability of each player winning the draw.
Read More: SalesTechStar Interview With Ryan Johnson, Chief Product Officer At CallRail
“Applying AWS’s machine learning services to game footage and official NHL data allows us to develop and share such analytics and insights as Face-off Probability, which provide new in-game analysis and predictions to enhance live broadcasts and take fans deeper into the game.”
When play stops, the Face-off Probability model begins generating win probabilities for players who it predicts will take the upcoming face-off. The probabilities are based on metrics including the players on the ice, face-off location, and current game situation. Face-off Probability is one of the first ML driven stats developed for NHL Edge IQ, powered by AWS. Fans can learn more about how AWS is transforming the hockey industry with the NHL on the AWS NHL site.
The face-off is one of the most anticipated and contested moments in hockey. Late in the third period of a close game, when two players are face to face on the ice, waiting for the puck to drop deep in the zone, the tension is palpable. Whichever player wins possession can swing the momentum to his team and directly impact the game’s outcome. With Face-off Probability, fans and broadcasters now have data to back up their predictions and determine which player is most likely to win.
“We’re excited to showcase Face-off Probability as part of NHL Edge IQ,” said Dave Lehanski, NHL Executive Vice President, Business Development and Innovation. “Applying AWS’s machine learning services to game footage and official NHL data allows us to develop and share such analytics and insights as Face-off Probability, which provide new in-game analysis and predictions to enhance live broadcasts and take fans deeper into the game.”
The NHL’s new ML-driven Face-off Probability model, powered by AWS technology and created in partnership with the AWS ML Solutions Lab, uses cloud capabilities such as analytics, storage, serverless compute, and media services to determine a player’s probability of winning a face-off and then immediately flashes the probability as a graphic on the live game broadcast for viewers. Using Amazon SageMaker (AWS’s service for building, training, and deploying ML models quickly in the cloud and at the edge), Amazon Kinesis (a service that makes it easy to collect, process, and analyze real-time, streaming data), and Amazon Simple Storage Service (Amazon S3), the NHL can gather data from the ice in real time, then analyze and visualize it for fans.
Read More: Training Industry Names ValueSelling Associates A Top 20 Sales Training Company For The 13th…
The NHL and AWS built the ML model for this new stat using the NHL’s Puck and Player Tracking data, along with data from hundreds of thousands of face-offs from over 10 years of historical NHL Hockey Information and Tracking System (HITS) stats. This data source was significantly varied and complex, incorporating such information as a player’s home and away face-off statistics; head-to-head matchup history; player characteristics such as height, weight, and handedness; and game context such as the face-off location, game score, and face-off time. This combination of historical success rates, player matchup characteristics, and game context brings together HITS and puck and player tracking data to provide a complete perspective for a face-off’s dynamics.
The Face-off Probability model is also flexible, adapting predictions based on changes to the game situation. For example, if a player is waived out of the face-off due to a violation, the model updates the prediction to the new matchup based on the real-time puck and player tracking streaming sensor data at sub-second latency. The model can also determine if a team’s probability of winning a face-off decreases or increases when the primary center player is waived out of the face-off in favor of a different player. This can highlight the importance of a particular player remaining in the face-off (pre-puck drop), adding a new point of analysis for NHL commentators and fans alike. In addition to providing the face-off win probability of the two players taking the face-off, the model provides information on other possible face-off combination pairings, including potential matchups that are most beneficial for each team. This information will be made available to on-air commentators to point out during the broadcast to enrich the fan viewing experience.
“One of the most important and least understood aspects of hockey is the battle over control of the puck during critical face-offs. This new stat brings the breadth and depth of AWS services to NHL data to better capture the details of gameplay for hockey fans, coaches, players, officials, and media partners,” said Matt Garman, Senior Vice President of Sales and Marketing at Amazon Web Services. “Sports serve as a great medium to bring the benefits of machine learning to life, and AWS is enhancing and enriching the NHL game experience through data, insights, and second-screen experiences that will make the sport more dynamic and engage a whole new generation of hockey fans.”
AWS entered a partnership with the NHL as the Official Cloud, Artificial Intelligence, and Machine Learning Infrastructure Provider of the league in February 2021, and the NHL and AWS also released shot and save analytics at the 2021 Stanley Cup Playoffs. The NHL will continue to introduce new analytics using AWS services and technology throughout the current season.