In 2025 Pre-Season Predictions and Information, Parts 4 and 4B, for the individual teams I showed the relationship between predicted NCAA RPI ranks and Strength of Schedule Contribution ranks under the NCAA RPI formula, both for the individual teams and for their opponents. In this article, I will show the same information, but for each conference. This gives a good picture of how the NCAA RPI discriminates among conferences because of the defective way it calculates Strength of Schedule.
This table has the conferences in NCAA RPI rank order, based on the average rating of their teams. See below the table for comments.
In the table, the first two green-highlighted columns on the left show, for each conference, the difference between its teams' average NCAA RPI rank and its teams' average Strength of Schedule contributor rank under the NCAA RPI formula. As you read down the table from the strongest conferences at the top to the weakest at the bottom, you can see the clear pattern: For stronger conferences, the conference teams' Strength of Schedule contributor ranks are poorer than the teams' actual ranks say they should be; and for weaker conferences they are better than they should be.
The next two salmon-highlighted columns look at how this plays out for the conference teams' schedules. The first of those columns shows the conferences' teams' opponents' average ranks and the second column shows those opponents' average ranks as Strength of Schedule contributors. The pattern here is the same: Stronger conferences' opponents' Strength of Schedule Contributor ranks are poorer than the opponents' actual ranks say they should be; and the opposite is true for the weaker conferences.
The next four columns break the numbers for the conference teams' schedules down into conference opponents (green-highlighted) and non-conference opponents (salmon-highlighted). Given that in conference play, the conferences' teams are playing each other, it is no surprise that the contrasts between the conference opponents' NCAA RPI ranks and their ranks as Strength of Schedule contributors follow the same basic pattern. For the non-conference opponents, where the individual teams have more control over their schedules, the pattern is similar but less extreme and with a little more variability.
It is important here to point out that coaches in top tier and most coaches in middle tier conferences are aware of these patterns and often take them into consideration in their non-conference scheduling. They also are aware, however, that in the NCAA Tournament seeding and at large selection processes, good results against highly ranked opponents matter, including against highly ranked non-conference opponents. Further, coaches of teams with NCAA Tournament aspirations often want to play at least some strong non-conference opponents. This means that they sometimes decide to schedule opponents whose Strength of Schedule contributions are likely to be poorer than their RPI ranks say they should be, essentially deciding to take a potential RPI "hit" in exchange for the potential of a good result against a highly ranked opponent.
NOTE: Being aware of the scheduling dilemma I just described, I designed my Balanced RPI, which is a modification of the NCAA RPI, with the specific objective of eliminating the difference between teams' ranks and their ranks as Strength of Schedule contributors. Thus under the Balanced RPI, if a team has a rank of X, that also is either exactly or very close to exactly the team's rank as a Strength of Schedule contributor. In other words, if the NCAA were to use the Balanced RPI, coaches no longer would have this scheduling dilemma. (As an additional benefit, the RPI no longer would discriminate among conferences in relation to conference strength.)
No comments:
Post a Comment