Google Cloud's machine learning meets NCAA Final Four games

Google says it will ​produce and air real-time ads during half time that will attempt to anticipate what will happen in the second half of the Final Four games.
Written by Natalie Gagliordi, Contributor
Todd Taulman, Getty Images/iStockphoto

Google Cloud's wide-ranging partnership with the NCAA will be on full display this weekend during the Final Four men's basketball tournament.

Using a predictive analytics architecture and workflow built over the last few months, Google says it will produce and air real-time ads during half time that will attempt to anticipate what will happen in the second half of the Final Four games.

The end-to-end analytics architecture leverages Cloud Spanner, Cloud Datalab and BigQuery to ingest, load, and analyze observations from the first half of each game, along with decades of historical NCAA data, to make predictions about the second half. Google said the predictions will include things like the number of three-point shots each team might attempt.

"As halftime starts, the real work begins. We'll have only minutes to turn our prediction into a TV spot," Google Cloud team member Courtney Blacker wrote in a blog post. "Our creative team will take the prediction generated by our team of data scientists and data analysts and create the ad right there in the Alamodome, using a real-time rendering system built by Cloneless and Eleven Inc."

Cloud AI wars: Re:Invent 2017: AWS all about capturing data flows via AI, Alexa, database, IoT cloud services | Google I/O 2017: Here's what we learned | Google bets on AI-first as computer vision, voice recognition, machine learning improve | How Microsoft plans to turn Azure into an 'AI cloud' | IBM enhances Watson Data Platform, with an eye towards AI

TechRepublic: Amazon AI: The smart person's guide | AI: How big a deal is Google's latest AlphaGo breakthrough?

"This is likely the first time a company has used its own real-time predictive analytics to create ads during a live televised sporting event," Blacker said.

Beyond game play predictions, Google's real aim here is to show off the possibilities of predictive analytics and machine learning in an example that's relatable to the general population.

The NCAA partnered with Google Cloud last year and has been migrating 80-plus years worth of game data across 24 sports and 19,000 teams over to Google's architecture. With the NCAA's team and game data on Google's cloud, Google's arsenal of machine learning technology has also been used to improve the tournament selection and team seeding process.


Cloud AutoML: How Google aims to simplify the grunt work behind AI and machine learning models

Google's Cloud AutoML uses the company's research and technology to enable enterprises to customize models and tune algorithms with their proprietary data.

March Madness: How to watch 2018 NCAA men's tournament games without cable

If you don't have cable, you can still stream for free all 67 NCAA tournament games through the iPhone, Android phones, web, tablet, and set-top boxes (including Apple TV and more). Here's how to watch online -- plus, all the March Madness Live dates and schedules you need to know.

Editorial standards