Login | September 24, 2018

NCAA, Google Cloud team up to create stats heaven

DOUGLAS N. MASTERS
SETH A. ROSE
Law Bulletin columnists

Published: May 7, 2018

Did you know that college basketball players from North Dakota are the best three-point shooters? It’s an interesting piece of information that’s technically correct — but even more intriguing is how that assertion came about and what it really means.

Thanks to a new partnership between the National Collegiate Athletic Association and Google Cloud, fans got an education in how data works through an avalanche of fun facts and statistics about the NCAA Men’s Division I Basketball Tournament’s history and this year’s bracket lineup.

The collaboration gives the NCAA access to a ton of player and team performance data that its more than 1,100 colleges and universities can use to improve play.

At the same time, Google Cloud got a chance to show off to a massive, data-hungry audience of sports fans what its cutting-edge data analytics can really do.

The multiyear marketing and technology collaboration, which also includes Turner Sports and CBS Sports, makes Google Cloud the official NCAA Cloud Partner. It’s Google Cloud’s first corporate partnership and it appears to be a match made in data heaven.

The venture migrates to the Google Cloud Platform more than 80 years’ worth of statistical game and competition data across the 90 championships and 24 sports administered by the NCAA, according to the partners, who announced the venture last December.

Released in 2011, the Google Cloud Platform is a suite of cloud computing services that allows users to store and analyze data and build and host applications and websites over the internet instead of a physical computer hard drive.

The platform uses the same infrastructure that Google created to run products including Google Search and YouTube. Analysis of NCAA data is powered by Google Cloud Platform services designed to work with gigantic amounts of data including BigQuery, Cloud Spanner and Datalab, which enables fans and NCAA members to search, compare and analyze team and player performance.

In addition, the NCAA is capitalizing on the partnership to build descriptive, predictive and diagnostic datasets to determine and analyze the selection and seeding process across men’s and women’s sports.

From a marketing standpoint, the collaboration is a smart move for both the NCAA and Google Cloud. The tag line for the partnership’s marketing campaign, “Know what your data knows,” is at the heart of what both organizations want to convey to their audiences.

For the NCAA, using cloud computing to mine vast amounts of data for interesting trivia and up-to-the-minute stats offered a key way to engage with millions of basketball fans.

The venture also gives the college sports organization the chance to explore new marketing strategies using artificial intelligence and machine learning. For Google Cloud, March Madness presented the opportunity to reach a broad audience in a relatable context.

The college basketball data-driven marketing campaign opened the door to introducing and “normalizing” the concepts of cloud computing, data analytics and machine learning to a broad audience, including potential customers, Solutions Review, an online source reporting on the latest technology advances, points out.

The NCAA and Google Cloud capitalized on their partnership by developing innovative ways to engage directly with their combined audiences, including basketball fans, computer and data enthusiasts and college students.

For example, Google Cloud used AI to create short, real-time advertisements during the Final Four matchups that answered questions related to games being played, giving even casual spectators new perspective on the action, notes Solutions Review.

Google Cloud and the NCAA also held a contest for data lovers to build and train machine learning models to forecast the outcomes of March Madness games. The partners hosted the competition on Kaggle, the largest online community of data scientists, and provided a data set to use that contained every play-by-play moment in men’s and women’s NCAA Division I basketball since 2009.

That’s more than 40 million plays, which demonstrated the power of the cloud as well as the NCAA’s robust data history. A total of $100,000 in prize money was awarded across both the men’s and women’s competitions for the most innovative machine learning applications.

In another event, the partners went on the road to host 20 Google Cloud Advanced Bracketology campus events, from Boston University to the University of California at Davis, to introduce students at all skill levels to the Google Cloud Platform and Kaggle. Participants received Google swag, prizes, food and the opportunity to meet Google engineers.

Finally, the NCAA/Google Cloud venture presents an overarching opportunity to educate general audiences about the nature of data, how to dig deep and what the data found actually means and doesn’t mean. It’s a mission that could pay off in more knowledgeable sports fans and computer customers and even a greater appreciation for data itself.

On its website, Google Cloud offers numerous tidbits of information about NCAA play, with links to detailed explanations of how the data was found and what it means. Take for example, its startling assertion that players from North Dakota are the best three-point shooters.

Google Cloud explains it first looked at each player’s home state and three-point shooting percentages for current players and found that players from North Dakota have an accuracy rate from the three-point line of 38.4 percent, the highest of any state in the country.

By digging further, Google Cloud discovered that only 10 Division 1 students from North Dakota are playing this year, making it the least represented state in the NCAA. Further, only four out of those 10 players have even attempted three-pointers in their playing careers. In other words, just three players are responsible for that 38.4 percent statistic.

Google Cloud points out that while the data results are narrow, they still have value. The results show what’s possible in terms of data queries and can be used to develop new questions and gain new insights. It’s also a clever way of demonstrating what’s possible in the right hands with the Google Cloud Platform and could very well lead to new partnerships with other sports.


[Back]