The analysis used by the Google Cloud team for the predictions is based on the interesting facts uncovered in the analysis of NCAA data: "Everything blocked by more shots per minute (for the record: Juniors) whether teams with a particular type of animal mascot cause more March Madness inconvenience (note: meow), "a blog post.
But the team wondered if they could do more if they used machine learning and their analysis of what could happen during a game, based on how the teams played before half-time. That's why they're trying this half-time experiment with Google Cloud Platform Tech. You can find out more about what they are doing on their blog.
This is certainly exciting for basketball fans, but the Cloud Platform team hopes that the techniques they use and what they learn from it will be broadly applicable. The team will break down their results after the Final Four