Using the same Business Analytics tools that are available to you to analyze your data resources (Sales, Marketing, or Production data) we’re looking at the data that’s available for the NCAA Men’s Basketball Tournament. With the first two rounds of March Madness behind us, I’ve created a visual look at how this year compares to previous years.
2017 First Round, seeds 1-4 won every match:
The green and blue is historic data, the brown and gold show how 2017 compares.
As the chart shows, seeds 1-4 won every game in the first round. Every other seed combination split 25%/75%, one way or the other. The one standout is seed 6, where 3 out of 4 games were lost to seed 11 teams. Historically, seed 6 does more poorly than expected (displayed as a valley in the chart) but this year that valley was considerably deeper.
Round 1, for seeds 5-12, conference matchup history is rather telling
Here’s a heatmap showing conference matchups in round 1. The upper left triangle is based on historic data. The lower right is 2017.
As an example; the ACC historically loses to the big 10. In all first round 2017 games, the ACC lost to the big ten. A quick scan of the entire chart will show that the historical outcomes held up very well in the first round of 2017. A first round bracket picked based on only seed and conference match ups would have been close to perfect for 2017.
Round 2, seed matchups follow an almost predictable pattern:
Due to the way the first round is setup, there will only be certain seed combinations in the second round as well. The chart below enumerates each of these combinations along with the number of total wins for both high and low seed. I’ve included lines for the current tournament to show how it measures up against historic data.
It is important to note, that the scales are different to better show the correlations between 2017 and historic data. This year, we had 5 upsets in round 2: #8 Wisconsin beat #1 Villanova, two teams seeded #7 beat #2 teams and two #3 teams beat #11 teams. These upsets follow historical patterns rather closely.
Round 2 conference matchup for 2017 matches history fairly well:
Same format as above with a few outliers in 2017. These outliers are circled in blue:
It is important to note, that even though there are four outliers, only two games are reflected. ACC teams generally win over Big East teams in round 2, but Xavier bucked that trend this year. Similarly, Pac 12 teams generally win over Atlantic 10 teams, but it really was no surprise Oregon won against Rhode Island.
Round 3, History can’t help much now:
I leave you with two charts showing the history of round 3. Here we see the scores of round 3 compared to those of round 2.
The gold line shows the round 3 score of the winning team. Brown is the losing score in round 3. Blue and green show the scores the round 3 teams had in round 2. One might expect that the blue line is generally higher than the green line, but this just isn’t the case.
Finally, a chord chart showing how seeds move between rounds 2 and 3:
A full rainbow of seeds makes it the sweet sixteen. This rainbow is abbreviated and dominated by seeds 1 and 2 in the Elite Eight.
If you have comments about our analysis or you’d like to see how we can help you to get the most value out of your data resources, please let us know.