Links on Android Authority may earn us a commission. Learn more.
Google will use machine learning for real-time Final Four predictions (Updated)
Update, March 31: We’ve update this article to clarify that Google will not, in fact, attempt to predict Final Four winners. The company will make instead certain specific predictions during halftime.
Original post, March 30: If you’ve placed any bets on the upcoming NCAA Final Four college basketball games, you might want to talk to Google before placing any more. That’s because Google will be using machine learning to make predictions regarding the Final Four games, processing both past data and data that’s happening in real time as the games progress.
Google teamed up with NCAA late last year to load almost a hundred years of sports data into Google Cloud Platform. Using that data combined with real-time information rolling in during the first halt of the games, Google will try make certain predictions about the second half. The company won’t attempt to predict the winners (which Google called “a fairly limited prospect”), but rather it will offer predictions like “We expect to see at least 25 three-point attempts combined in the 2nd half with a 78.2% probability.”
The Google NCAA predictions will be aired in an ad during halftime for each game. The video ad will be rendered on the fly, based on the predictions generated by Google’s machine learning system. The whole process will take minutes, but everything needs to work in sync, so the pressure will be high.
According to the blog post on the matter, Google will be taking into account “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).” But will the Google NCAA predictions be correct? We’ll have to wait and see.
While the notion of using machine learning to predict sports games is pretty novel, Google hopes that the experience will help it apply Google Cloud Platform learning more broadly. One could imagine using machine learning to predict natural disasters, or even help respond to areas hit by natural disasters, for example.