CURRENT ACTUAL RPI RATINGS FOR GAMES THROUGH 9.18.22

From now until the end of the season, I will be doing three articles each week.

1,  The first article will be team actual RPI ratings and ranks and other information, to date.

2.  The second article will be simulated end-of-season ARPI rankings based on the actual results of games played to date and simulated results of games not yet played.  The simulated results will be based on the actual ARPI ratings of teams, to date.

3.  The third article will be simulated NCAA Tournament brackets based on the actual results of games played to date and simulated results of games not yet played.

Over the course of the season, all of these should come closer and closer to where they actually will end up.

Here is a link to this week’s workbook with actual RPI ratings and ranks, as well as other information, based on games played through September 18, 2022: 2022 RPI Report 9.18.22.

Two notes about the workbook:

1.  On the RPI Report Source page, which has team RPI data, I have included two columns on the far right, that I did not include in previous years.  Previously, on the right, I had two columns: Strength of Schedule and Strength of Schedule Rank.  I still have those columns.  I have added, next to them:

 a.  Opponents Average Rank.  The Strength of Schedule Rank column is based on ratings from the strength of schedule portion of the RPI formula.  This is not the same, however, as Opponents Average Rank and in fact can be quite different due to the structure of the RPI.  The Opponents Average Rank column gives a better picture of how strong a schedule is as compared to the Strength of Schedule Rank.

b.  Strength of Schedule Contributor Rank.  This shows how a team ranks as a strength of schedule contributor to its opponents, based on how the RPI calculates strength of schedule.  If you compare it to the RPI Rank column, you can see that in many cases a team rank is quite different from its rank as a strength of schedule contributor.

 Also, some notes about the accuracy of my numbers:  Weekly, once the NCAA has published its ratings at the RPI Archive (a different source than the rankings it publishes at NCAA.com), I compare my ratings to theirs.  If my numbers differ from theirs, I go through a checking process to find out why there are differences.  The differences can come from data errors or from programming errors.  If the errors are mine, I make corrections and re-run my numbers to be sure they match the NCAA numbers.  If the errors are theirs, I inform them of the errors and hope they will make corrections.  (I also work cooperatively with Chris Henderson to be sure his College Women’s Soccer Schedule data match mine and the NCAA’s.)

The NCAA current numbers have some errors related to penalty adjustments.  There are two tiers of penalty adjustments: the higher ones apply to teams with ties or losses against teams in the RPI bottom 40 of the rankings and the lower ones apply to teams with ties or losses against teams in the RPI next to bottom 40.  This means that when there are new teams competing in Division 1, the penalty programming must change.  For example, this year, with the number of D1 teams increasing to 348, the programming had to change so that the penalty tiers are teams ranked 269-308 and 309-348.  As has happened a number of times in the past, the NCAA has forgotten to change its penalty tiers.  Thus some teams are receiving penalties from the NCAA when they should not be.  I have let them know of this error and hope they will make a correction.

Apart from the penalty tier problem, my numbers match the NCAA numbers.