Tourney time: 4 decades of March Madness data

It’s that time of year. In a lot of places around the U.S., the snow begins to melt, the days get longer, and people start to get back outside. There’s a special kind of excitement that comes with March and the onset of spring. But some of that excitement has nothing to do with the change of seasons. Instead, it’s generated by one of the great annual sporting events in the entire world: March Madness.

For anyone who’s been living under a rock, that’s the term for the qualifying and main stages of the NCAA men’s college basketball tournament. With conference tournaments already underway, and Selection Sunday only a few days away, we here at Domo decided to take a look at some historical data surrounding what’s also referred to as the “Big Dance”—and show just how powerful the Domo platform is at bringing data together from different sources, and cleansing data that either has errors or doesn’t quite match up properly.

We started with tournament data from 1982, which, interestingly enough, was the first year there was no third-place game. And we will update the data as the event goes along.

The first thing everybody wants to talk about, of course, is upsets, which is when a lower seeded team beats a higher seeded team. So we’ll start there, with our “March Madness Upset Assistant” dashboard (directly below), which breaks down upsets over the years in a number of ways. If you want even more data on upsets, scroll to the bottom of the main dashboard at the end of this post.

Speaking of the main dashboard, you’ll notice that it’s broken down into three distinct sections. On top, we have information on total tournament wins and national championships. Next, we have some stats on individual teams. (Select a school, and you’ll see win percentages and historical game information for that particular team.) And, as previously mentioned, we have yet another section on upsets.

As any college hoops fan might expect, North Carolina and Duke are the beacons of NCAA tourney success. But as you’ll see, other perennial participants such as Connecticut, Florida, Kentucky, and Louisville aren’t far behind.

There will be a follow-up post that talks more in-depth about our data connection and data transformation process. When it is up, we will link to it here. In the meantime, enjoy the next couple weeks of buzzer-beaters and nail-biters. And may you select enough winners to come out on top of any betting pool you enter.