NCAA TOURNAMENT: LIKELY AT LARGE CANDIDATE POOL AND SELECTION CHANGES IF THE COMMITTEE HAD USED THE BALANCED RPI RATHER THAN THE NCAA RPI

This article is on the question: Would it make a difference in NCAA Tournament at large selections if the Women's Soccer Committee used the Balanced RPI rather than the NCAA RPI?

The period covered is 2007 through 2023 (excluding Covid-affected 2020).  To determine what the at large selections would have been if the Committee had used the Balanced RPI, I took two steps:

1.  I assumed that under the Balanced RPI all at large selections would come from the Balanced RPI Top 57 teams.  I assumed this because since 2007, all at large selections have come from the NCAA RPI Top 57 teams (based on the current NCAA RPI formula).  From a practical perspective, there is no reason to think this would be different if the Committee were using the Balanced RPI: Everything would look similar to the Committee, it simply would looking at different rating numbers.

2.  One of the factors I use in evaluating Committee decisions is teams' good results (wins or ties) against Top 50 opponents.  I score these results using a system that is very heavily weighted towards good results against very highly ranked opponents.  Then I rank teams based on these scores.  Once I have these ranks, I combine them with their RPI ranks, with each rank weighted at 50%.  I then rank teams using their combined rank scores.  When I do that, the ranks on average match the Committee at large selections for all but 2 selections per year.  With that in mind, to see what the at large selections likely would have been if the Committee were using the Balanced RPI, I determine what their Top 50 results scores would have been using the Balanced RPI, rank teams accordingly, see what their combined Balanced RPI and Top 50 results score ranks would be with each weighted 50%, and rank them accordingly just as I do for the NCAA RPI and Top 50 results ranks.  I then assume that if the Committee were using the Balanced RPI, it would select teams based on their combined Balanced RPI and Top 50 results rank.

Looking at the numbers from 2007 through 2023, a change to the Balanced RPI would result, on average, in 6.3 new teams in the Top 57 per year -- in other words about 6 new teams per year would become candidates for at large selections and a matching 6 using the NCAA RPI would not be candidates.  Further, and more important, 3.6 different teams per year would get at large selections -- in other words 3 to 4 teams per year that did not get at large selections under the NCAA RPI would get them under the Balanced RPI and 3 to 4 teams that did get at large selections under the NCAA RPI would not get them.  Thus the answer to the question at the opening of this article is

Yes, it would make a difference in NCAA Tournament at large selections if the Women's Soccer Committee used the Balanced RPI rather than the NCAA RPI, to the tune of 3 to 4 different teams per year getting at large selections, on average.

When looking at conferences, here are the results of a rating system change for the 2007 through 2023 period:

In the table, the Net At Large Gain or Loss column shows the difference between the Committee's actual at large selections using the NCAA RPI and what the selections likely would have been if the Committee had used the Balanced RPI.

As you can see, the conference hurt the most by the NCAA RPI is the Big 10, which lost 16 at large positions from 2007 through 2023 due to use of the NCAA RPI, or 1 position per year on average.  Of the five conferences hurt by the NCAA RPI, three are from the West:  the Pac12, West Coast, and Big West conferences.  The other conference hurt is the ACC.

The conferences that benefit the most from the NCAA RPI are the Big East, SEC, Colonial, American, Atlantic Ten, and Big 12, with a number of other conferences helped just once over the 16-year period.

In the table, the Net Top 57 Gain or Loss column shows the difference in the number of teams a conference had in the Top 57 candidate pools for at large selections.  Note that 4 of the 6 conferences hurt by the NCAA RPI are from the West.

When looking a geographic regions, here are the results of a rating system change:

NOTE:  There are two sets of "new" at large selections under the Balanced RPI.  One set is teams that were outside the NCAA RPI Top 57 that, on coming into the Top 57 under the Balanced RPI, get at large selections.  The second set is teams that were inside the NCAA RPI Top 57 and did not get at large selections, but that under the Balanced RPI would get at large selections.  In the table, the Middle region is an example of having some from each set.

In the next post, I will pull together key information from this and other recent posts to show how all the information inter-relates.