I'm in the process of updating some of my work, some of which tracks team ranks over time. The ranks I track are the NCAA RPI, my Balanced RPI, and Massey ranks. And, I track team ranks under the RPI formula as strength of schedule contributors (which figure into their opponents' RPI ranks). The updating process is labor intensive and somewhat tedious, but it produces some interesting information. With that in mind, as I go through the updates, team by team, I'll post charts here for the upper level teams and also for some other teams I find interesting. My first is for Alabama: Things to look for in the charts are: Differences between the ARPI Ranks (NCAA RPI ranks) on the one hand and the Massey and Balanced RPI ranks on the other hand. (In most cases, the Massey and Balanced RPI ranks are quite close). Differences between the NCAA ARPI Ranks and the NCAA ARPI Strength of Schedule (SoS) Contribution Ranks. These should be the same. As you can see for Alabama, they are not.
I'll be adding, for each team, three numbers. I'll start with Alabama, from the preceding post: Average Balanced RPI rank (last 7 years): 43 Average NCAA RPI rank (last 7 years): 40 Strength of Schedule Contributor rank (based on average NCAA RPI rank): 70 The 7-year average Balanced RPI rank is the best predictor I have been able to come up with, when using only rating system past ranks, for predicting what teams' ranks will be next year. This doesn't mean it is an accurate predictor, since for DI women's soccer there seems to be no such thing, but based on comparisons with others' predictions, it is about as good as any predictor can be. At the top and bottom of the rankings, it is fairly accurate, whereas in the middle of the rankings (where there are a lot of teams of relatively equal strength) it is far less reliable. Also, the Balanced RPI ratings are a better measure of team performance than the NCAA RPI. They match overall game results better than the NCAA RPI and they rate teams from different conferences and regions properly in relation to teams from other conferences and regions, whereas the NCAA RPI discriminates against some conferences and regions and in favor of others. The way I think about these three different ratings is: The Balanced RPI is a measure of true team strength The NCAA RPI is what the NCAA tells the Women's Soccer Committee to use as a measure of team strength The Strength of Schedule Contributor rank is what the RPI considers a team's strength to be when it measures strength of schedule for RPI computation purposes Using Alabama as an example, the above numbers suggest that the NCAA RPI gets Alabama's ranking about right. But, if you play Alabama, the RPI formula is only going to give you credit for playing an opponent that is significantly weaker than they really are.
Using Brown as an example, a question to consider based on their chart: Has their mean been reset as suggested by the years 2019 through 2023? Or did they have four good years and now are in the process of reverting to a poorer ranked mean? There is no way to know at this point, but it is something to watch for over the coming years. Average Balanced RPI Rank: 74 Average NCAA RPI Rank: 69 Strength of Schedule Contributor Rank: 65
I am working my way through the teams in alphabetical order. If I skip a team but you want to see its chart, let me know and I will post it. Average Balanced RPI Rank: 24 Average NCAA RPI Rank: 32 Strength of Schedule Contributor Rank: 29
I am including Cal Baptist, since it reclassified to Division I in 2018 and is an example of the improvement that is possible for a re-classifying team. Average Balanced RPI Rank: 149 Average NCAA RPI Rank: 178 Strength of Schedule Contributor Rank: 109
I'm including Colorado State since it was a brand new team in 2013 and is an example of what a new team can do. It also may be a good example of where predicting next year's rank based on an average of the last 7 years' ranks can go awry. Average Balanced RPI Rank: 146 Average NCAA RPI Rank: 161 Strength of Schedule Contributor Rank: 158
Taken out of alphatebical order, just for you. Average Balanced RPI Rank: 55 Average NCAA RPI Rank: 79 NCAA RPI Strength of Schedule Contributor Rank: 130 I.e., their average strength rank over the last 7 years is #55. But when the Women's Soccer Committee looks at the RPI, what they see is #79. And if their NCAA RPI rank is #79, when they play an opponent (including all of their Big 10 opponents), the likely value of their contribution to the NCAA RPI strength of schedule portion of the opponent's rating is ranked #130. The NCAA provides the Committee with a lot of data. But, it does not provide the Committee with teams' ranks as Strength of Schedule contributors. I wonder why ....
Average Balanced RPI Rank: 90 Average NCAA RPI Rank: 73 NCAA RPI Strength of Schedule Contributor Rank: 87
Average Balanced RPI Rank: 107 Average NCAA RPI Rank: 112 NCAA RPI Strength of Schedule Contributor Rank: 189
Average Balanced RPI Rank: 114 Average NCAA RPI Rank: 133 NCAA RPI Strength of Schedule Contributor Rank: 126