The analytics that the Google Cloud team will be using to make the predictions will be based on the interesting facts they've uncovered while analyzing NCAA data: "everything from who blocks more shots per minute (for the record: juniors) to whether teams with a certain type of animal mascot cause more March Madness upsets (hint: meow)," according to a blog post.
But the team wondered whether they could do more if they used machine learning and their analytics to figure out what might happen during a game, based on how the teams were performing before halftime. That's why they're trying this halftime experiment using Google Cloud Platform tech. You can learn more about exactly what they are trying to do over at their blog.
While this is certainly exciting for basketball fans, the Cloud Platform team is hoping that the techniques they use, and what they learn from the experience, will be broadly applicable. The team will break down what their results were after the Final Four.
Update 3/30/18 12:44PM ET: This post has been updated to clarify that Google isn't picking winners during this process, but instead it will make predictions about overall performance.