Home > Current Updates > Bracketology: Resume, metrics have Villanova No. 1

Bracketology: Resume, metrics have Villanova No. 1

Dave2015Recently, the NCAA Men’s Basketball Selection Committee made public its intent to evaluate options for integrating additional analytics and metrics into its process for selecting and seeding teams for the NCAA tournament.  How that might eventually look we don’t know.  What metrics will be included?  Will there be any weighting for certain metrics (such as the current Ratings Percentage Index or RPI) over others?  It’s an interesting twist, although not necessarily that new.  Selection Committee members have had various data points at their fingertips for years.  How each chose to use (or not use) ratings such as BPI, KenPom, Sagarin, and others is unknown; none were a part of the “official” process and rarely discussed publicly.

Bracketology   |   Seed List

The point today isn’t to have a lengthy discussion about the merits of the analytic tools.  Rather, it provides an opportunity to look at how a new system might look.  One of the concepts mentioned was the potential for a “composite” ranking that would factor in a given set of analytic metrics.  In the interest of today’s bracket update, here’s a look at some rankings and the “composite” for the top six teams on the seed list (data points were through games played on Tuesday).

The metrics included: RPI, NPI (Warren Nolan), ELO Chess, KenPom, Sagarin, and BPI.  Composite ranking is in parenthesis, based upon an average of these metrics.

  1. Villanova (2.33)
  2. Kansas (4.83)
  3. Baylor (9.00)
  4. UCLA (11.16)
  5. Kentucky (3.16)
  6. Gonzaga (6.50)

As you can see, the composite score would favor moving Kentucky and Gonzaga into spots two (2) and four (4) on the seed list, making them No. 1 seeds instead of Baylor and UCLA (in today’s bracket).  Those are not wrong conclusions.  Of course, metrics do not necessarily do a good job of including actual results.  Example: UCLA won at Kentucky.  An individual result is only one factor (usually a small one) in a team’s overall resume.  But the disparity in numbers (11.16 to 3.16) suggests a notable difference between UK and UCLA.  How the Committee might discuss those two teams given the result vs. numbers would be interesting.

The more relevant takeaway here is that Villanova is No. 1 by a significant margin.  When you factor in the math, the difference between Villanova and Kentucky is a little larger than it looks.  Thus, the Wildcats remain entrenched as the overall No. 1 seed today by almost any measure, resume and otherwise.

In keeping with our theme, we used a similar process, including a metric called Strength of Record (SOR) to help decipher an uninspiring collection of bubble teams.  With that in mind, Kansas State and Wake Forest are the final two at-large entries.  It’s completely erratic near the bottom of the bracket, so this is far from definitive.  History suggests that may not change much.

Advertisements
Categories: Current Updates
  1. No comments yet.
  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: