Conference Realignment

While I'm having so much fun, I also took a look at the distribution of games of the power and non-power conferences' teams to see how much those two groups play each other. As power conferences, I included the ACC, American, Big East, Big Ten, Big Twelve, Pac 12, SEC, and West Coast conferences.



As you can see from this, the 23 non-power conferences' teams play about 90% of their games against each other.

As I've said before, from an RPI perspective, if you have two groups of teams that play no cross-group teams, then the groups' RPIs will match each other. In that circumstance, the RPI simply can't rate the two groups within a single system. In order to rank the two groups within a single mathematical system, there must be cross-group games (there must be "correspondence" between the groups). The ideal amount of correspondence is if every team plays every other team, i.e., there is a full round robin. This is not practical with 332 teams.

The question is: Is there enough correspondence between the power and non-power conference teams for the RPI successfully to rank both groups of teams within a single system? My answer is that the RPI has some problem doing this, which explains in part why you see very few non-power conference teams getting at large selections for the NCAA Tournament, even though they may be within the bubble of teams considered for selections. What happens is that the Women's Soccer Committee --when it gets to looking at head-to-head results, results against common opponents, and results against highly ranked teams -- gets to see that the RPI has this problem and uses these other factors to "overrule" the RPI.

When you look at the Ivy League's relative isolation from the rest of the country in the Conferences' table I posted, you can get to see how where the Ivy teams fit within the RPI is not necessarily a good measure of their strength. What you're really seeing is their strength in relation to teams in the Northeast playing pool. The Northeast playing pool is significantly weaker than the Middle, Southeast, and West pools, so one would expect the Ivy teams to be overrated if you look at their ratings in a combined national system.

I should point out, however, that I am talking about nuances of the RPI. Do these RPI issues have a significant effect on Women's Soccer Committee at large selection and seeding decisions? It all depends on what you consider to be significant.