March Madness by the Numbers

March Madness ended last Monday with what some are calling the greatest final ever played. It was an epic matchup between two powerhouse teams from the ACC and the Big East.

I’ve been a fan of teams in the Big 8 / Big 12 for most of my life. I always hear the case being made during the regular season that teams from the conference (minus the Kansas Jayhawks) are overrated, and they will find an early exit come tournament time.

I could never really accept the premise without any hard data. This year I decided to see if there was any available. Unfortunately, Google searches were fruitless, so I had to gather data from multiple sources. CBS Sports, College Poll Archive, and Sports Reference are excellent resources.

Many times, what starts out as a simple exercise can turn into a much bigger project. This was the case with the March Madness review. Data such as school names were not consistent from the various sources, so I had to use a simple regular expression, plus rely on good old fashioned manual verification in some cases.

Keep in mind, this isn’t about predicting outcomes, it’s only research and historical analysis.

The data is from the last 25 years of college basketball, 1992 to 2016. The data set contains, by year:

  • AP Top 25 rankings from the week before the tournament
  • Ranking by conference including strength of schedule and number of teams
  • Team ranking in conferences consistently in the top 5
  • Results from the NCAA tournament
  • Contains totals by conference and AP Top 25.

We could go to many different levels with the data that was collected to make comparisons. We tried to keep it simple, but found it necessary to address a couple questions.

Do bigger conferences have more opportunities to get teams in the tournament?

The short answer is yes, however they could be victims of dilution of strength. More teams does equal more chances to be selected.

Do smaller conferences have a proportional advantage because of the essentially two guaranteed spots in tournament?

Yes.  For example, Big 8 2/8=25%, but Big 12 2/12=16%. The smaller conference gets 25% of its teams in the big dance.

In the end, the goal is to measure Conference strength so dilution is what we want to measure.

When we considered the above, the formula that made sense was:

Teams advancing in 25 years / Total teams in 25 years

Let’s call it the Advancement Quotient (AQ), it’s the total number of teams over the last 25 years from each conference that progress to the next round (32 to 1), divided by the total number of schools in the conference over the last 25 years.

I added the total of AP Top 25 schools in the conference for the past 25 years and SRS (Sports Reference’s Simple Rating System) to give you an idea of media rankings and strength of schedule.

  Total # of  Average # Teams  Teams in the Round of
Conference Schools SRS AP 25 64 32 16 8 4 2 1 AQ
Atlantic Coast  269 13.5368 86 130 99 63 32 20 12 8 0.8698884758
Big Ten 283 12.1752 86 147 103 56 31 20 9 1 0.777385159
Big 8 / Big 12 270 11.6604 83 137 88 47 27 10 3 1 0.6518518519
Big East 330 11.0696 98 148 104 57 28 13 7 6 0.6515151515
Southeastern 308 11.0264 77 116 77 47 28 17 10 6 0.6006493506
Pacific 10/12 260 10.3856 61 111 73 46 21 8 4 2 0.5923076923

 

Using AQ, the Big 8 / 12 is a respectable 3rd. My perceptions were that the Big East and Big Ten would be reversed, but that’s why you do the work.

There is much more you can see in the data, such as how the AP Top 25 did in the tournament. Download the Excel file for yourself and take a look.

David Chavarria is a researcher for FreeState Poll.