March Madness Heading Into The Final Four
By: Stu Kendall
2 minute read
The Elite 8 round was an incredible weekend of basketball full of buzzer beaters, monster dunks, and hard-fought upsets. Only 4 of our 85 participants in the Narrative Science Employee March Madness Bracket Challenge correctly predicted a Michigan State over Duke upset to advance to the Final Four. It’s worth mentioning all four participants just so happen to be Michigan State alumni. Go figure.
Despite the large amounts of data available of past bracket performance, crafting together a perfect bracket remains to be a near impossible feat. After the Elite 8 round, there isn’t a single person on earth that has a documented perfect bracket left. A Duke Math Professor says the odds of predicting a perfect bracket are roughly one in 2.4 trillion! Watch the video below to learn more about how he got to that number.
Even with insurmountable odds of trying to predict perfection, this isn’t stopping Narrative Science employees (like me) from using data and analytics to help give them an edge. Before the tournament started, one of our Analytics Managers took data from basketball-reference.com and built out a dashboard to highlight how certain variables impact performance throughout the tournament.
But there was a problem – not everyone at Narrative Science is skilled in Analytics. Including myself. When interpreting lengthy bar charts or hearty tables of data, it is extremely difficult to walk away with anything memorable or actionable. And to add to this, who really has the time? Sure, I can download the data to excel myself – filter, cross-tab, and cut the data to pull out an insight – but that may take me 20 or 30 minutes. I’m looking for something easier and faster.
And now here’s my shameless plug to solve this problem. We embedded our natural language technology to help tell the story of the data. If you venture over to the dashboard, you’ll see insights written out in plain-English by our Quill Extension. The summaries automatically update as you click around and filter the table. I sit in the Marketing organization here at Narrative Science and I’m still blown away by this technology every single day.
You can go from something that looks like this:
Explained to you in plain-English:
Pretty cool, right?