In doing some work recently in cooperation with the new Intercollegiate Women's Soccer Organization for Coaches (IWSOC), I developed a new, easily understandable way to do early season rankings of all teams. I'll start with an explanation of the basic method and then will show how it ranks the Top 75 teams.
Summary: The early season rankings, as a starting point, use last year’s NCAA RPI Selection Rankings -- not the individual teams’ rankings but rather the ranking
positions occupied by teams’ conferences.
Thus each conference will have the number of national pre-season ranking
positions assigned to it equal to the number of teams in the conference and the
particular national pre-season ranking positions assigned to it will be the
ranking positions its teams occupied last year.
(The Selection Rankings are the NCAA RPI rankings following completion
of the regular season including conference tournaments. These are the rankings the Women’s Soccer
Committee uses when seeding and making at large selections for the NCAA
Tournament.)
The second step is to look for how "experts" have ranked teams within each conference. For example, each conference's coaches do a pre-season poll that ranks the teams within the conference. Using those rankings as an example, when a conference’s coaches have done their pre-season
rankings for the teams in their conference, the conference’s teams will be
assigned in order to that conference’s ranking positions in the national early season ranking list.
For example, last year (2025), ACC teams occupied the #1, 2, 3, and 10 positions in the NCAA RPI Selection rankings. This year, the ACC coaches' pre-season poll had Notre Dame, Stanford, Duke, and Florida State in the top 4 in-conference positions, in that order. The method I described therefore assigns ranks to those teams of #1 Notre Dame, #2 Stanford, #3 Duke, and #10 Florida State.
Thus once I have all the conference's coaches pre-season polls, I can put all the teams in pre-season rank order, depending on how each conference's coaches have placed the teams within their conference and on which ranking positions the conferences' teams occupied last year.
Rationale for Using Last Year’s NCAA Selection Rankings
as a Baseline Starting Point:
Individual teams’ ranks, with very few exceptions, vary from year to
year. Conference strength, on the other
hand, is quite consistent from year to year.
In particular, the rank positions a conference’s teams occupy, although
not identical from year to year, are reasonably consistent. This is particularly true in the area of the
rankings that include the teams that will receive consideration for NCAA
Tournament seeds and at large positions.
The following table shows this, with an explanation below the table. (Scroll to the right to see more of the table. Alternatively, click on the table to see the entire table.)
Using the ACC at the top of the table as an example, this
table shows in the first green highlighted column – headed Number in 2024 NCAA
RPI Top 5 – that in 2024, the ACC occupied 3 of the top 5 rank positions. In the next green highlighted column – headed
Number in 2025 NCAA RPI Top 5 – the table shows that the following year, the
ACC again occupied 3 of the top 5 rank positions. Moving on to the next two columns, in 2024
the ACC occupied 5 of the top 10 rank positions and in 2025 it occupied 4 of
the top 10. Continuing all the way to
the last two columns on the right, in 2024 the ACC occupied 11 of the top 57
positions and in 2025 it occupied 10 of the top 57 positions. (The table includes the Top 57 positions
because historically #57 has been the poorest NCAA RPI ranked team to get an
NCAA Tournament at large position.)
In the table, the gold highlighted cells are for reading
convenience, simply marking where the different conferences had teams.
At the bottom of the table is a row for Matches. If you look down the Top 5 columns, you will
see that the ACC had 3 teams in the top 5 each year, the SEC had 1 team in the
top 5 each year, and the Big 10 filled the other spot in 2024 but the Big 12
filled it in 2025. Thus from a
conference perspective, 4 of the 5 top positions had conference matches over
the two years. This is represented by
the 4 in the Matches row below the Top 5 columns.
Thus looking at the Matches row across the table, there are
4 conference matches for the Top 5, 7 for the Top 10, 11 for the Top 15, and so
on across the table, to 50 for the Top 57.
The color coding in the Conference column on the left shows
with green highlighting the conferences for which using 2024 as a base for 2025
pre-season ranks would have assigned more Top 57 positions to the conferences
than they actually ended up with in 2025.
The orange highlighting shows the conferences for which using 2024 as a
base would have assigned fewer Top 57 positions than they actually ended up
with. As the two right-hand columns
show, the maximum difference for any conference is only 1 position. Further, if you compare the green and orange
highlighted conferences, there does not appear to be any patterned
discrimination between to the two groups of conferences.
As a whole, the table shows that in the end-of-season
ranking area of teams that are in the ranking range for NCAA Tournament at
large positions, using this method of assigning rank positions by conference is
a reasonable way to predict where conferences’ teams are likely to end up.
At the very bottom of the table, under the cell Matches, is
the number 271. This is the total of the
matches across the table in the Matches row. I also have done tests using the average of the prior two years’
rankings and the average of the prior three years’ rankings as a baseline
starting point. These resulted in 255
total matches using the prior two years’ rankings and 257 using the prior three
years. Thus using only the prior year’s
rankings for the assignment of baseline rank positions to conferences produces
the best match with where the final rankings will end up.
The following table shows in more detail how using the prior
one, two, and three years’ rankings as the baseline compare:
NOTE: In assigning teams’ ranking positions to conferences,
this method treats teams as being in the conferences they will be in in the
year for which the predictions are being made.
Thus, for example, Texas State’s 2025 rank was #57. In 2026, Texas State will be in the
re-constituted Pac 12 conference. The
method therefore assigns the #57 rank position to the Pac 12 for purposes of the
2026 pre-season rankings.
Rationale for Using the Coaches’ Conference Pre-Season
Rankings to Fill the Pre-Season Assigned Conference Rank Positions: In the past, there have been three sets of
pre-season rankings of the teams in each conference. The coaches in each conference rank their
teams. Chris Henderson uses a series of detailed metrics to rank the teams within each
conference. And I use historic ranking data to rank the teams within each conference. All three ranked teams from 2022 through 2024. The following table shows how each’s ranks
compared to teams’ actual end-of-season ranks within their conferences:
As the table shows, on average the coaches’ pre-season
rankings came within 2.2 positions of teams’ actual final conference
ranks. Chris Henderson’s were next in
accuracy at 2.3 positions and mine were at 2.4 positions. Thus the coaches’ ranks, with the coaches’
detailed knowledge of their conference opponents, roster changes, and other
information, are as good a set as you can get of pre-season rankings of
teams within their conferences. Chris Henderson tweaks his ranking method from time to time and I consider his rankings also to be excellent.
EARLY SEASON RANKS OF TOP 75 TEAMS
With that background, the methodology produces the following early-season ranks of the Top 75 teams, based on the conference coaches' pre-season polls. The table also includes the IWSOC ranking committee's ranks of the Top 25 teams, for comparison.
I will be updating the Top 60 every other week for IWSOC's publication, most likely keeping the early-season rank positions assigned to conferences but possibly changing the teams' positions within their conferences if results indicate that would be appropriate. You can follow these and other coach-oriented information on X at https://x.com/theIWSOC.
