Making sense of massive amounts of data is critical to every organization today — this time of year, no more so than to the National Collegiate Athletic Association.

With the March Madness college basketball tournament season starting next week, Google LLC is expanding its work with the NCAA to crunch the association’s 80 years of game and athlete performance data, with an aim to extract more insights and make predictions.

During last year’s March Madness tournament, Google assembled a team of data scientists and basketball fans called The Wolfpack. The team built a data processing workflow using Google Cloud technologies such as its BigQuery analytics data warehouse and its Cloud Datalab interactive data insights tool.

The team built models to discover things such as who blocks more shots per minute and make predictions such as how many three-point shots a particular team might try in the second half of a game. Then Google ran real-time ads with those predictions, creating them in just minutes on the fly.

Since then, it even built a basketball court (pictured) at last year’s Google Cloud Next conference at Moscone Center in San Francisco to understand the mechanics of various shots.

This year, the search giant and cloud computing contender detailed today in a couple of blog posts, it’s expanding those efforts, mainly by recruiting 30 college students who are both hoops fans and comfortable with wrangling data.

The student group took a new look at all those decades of data to find new ways to look at games. Specifically, the students are looking at two potential metrics to evaluate teams.

One is “explosiveness,” the way a team can go on an extended run of point-scoring, and how that changes team competitiveness. If Duke University goes on a run, for example, it may predict an eventual win better than it does for another college that does the same thing. Students are also analyzing the data to determine overall “competitiveness,” or how well a team can control the score during a game.

The students’ models will be used in a new tool for fans ahead of Selection Sunday this coming weekend. Google is also supplementing last year’s predictions on three-point shots, rebound estimates and ball turnover rates with new ones on dunks, scoring runs and player contributions. And it will again run predictive TV spots during the Final Four games.

Google clearly doesn’t see the NCAA as a special case for large-scale cloud analytics, so it’s using its work with the association to show what it can do for other large organizations.

“It’s what you’d see at a classic enterprise,” for instance the processes of extracting, transforming and loading data into systems and scaling up data warehouses, Eric Schmidt, Google’s cloud advocate for big data and data science, told SiliconANGLE. (FYI, he’s not related to Alphabet Inc.’s technical adviser and former executive chairman.)

Google has emphasized its expertise in AI and analytics in its cloud services, which trail behind market leaders Amazon Web Services Inc. and Microsoft Corp.’s Azure in revenue.

Immediately after the Final Four games in early April, Google will hold its annual Cloud Next conference, where it will host a couple of bootcamps that will use NCAA data to demonstrate how to build a data science setup.

Photo and image: Google