2024 ARTICLE 3: POST-WEEK-2 NEWS AND UPDATED PREDICTIONS

I have not found final word on whether the Women's Soccer Committee's recommended changes to the RPI will be in effect this year, but based on the Competition Oversight Committee's approval of the changes, I assume they will.  We know that one recommendation was to change the RPI value of ties from 1/2 a win to 1/3 of a win.  We know that the other recommendation was to change the bonus and penalty structure, but we don't yet know the amounts of the bonuses and penalties.  Given that, I have revised my program to include the tie valuation change but have retained, for now, the previous bonus and penalty structure.

Below are my updated predictions, after incorporating all actual game results through Sunday, August 25.

To give some context and so you can make your own decision on the reliability of the predictions, here are some numbers comparing the actual results of games so far to my predicted results for those games.  Team 1 is the home team or, for neutral site games, the team whose name is first in alphabetical order:

Actual Team 1 wins:  52.3%

Actual Team 1 losses:  29.4%

Actual Team 1 ties:  18.3%

Predicted Team 1 wins:  49.3%

Predicted Team 1 losses:  29.7%

Predicted Team 1 ties:  21.1%

Looking through a different lens, here are numbers comparing (1) the actual results so far for the higher rated team in relation to my assigned pre-season ratings and adjusting for home field advantage to (2) the actual results for all games played from 2010 through 2023 in relation to the NCAA's actual end-of-season ratings and adjusting for home field advantage, with both my assigned pre-season ratings and the NCAA ratings based on valuing ties as 1/3 of a win:

Historically, higher rated team wins:  65.3%

Historically, higher rated team loses:  13.6%

Historically, higher rated team ties:  21.1%

This year, my higher rated team actually wins:  62.6%

This year, my higher rated team actually loses:  19.1%

This year, my higher rated team actually ties:  18.3%

Team NCAA RPI, NCAA Non-Conference RPI, and Balanced RPI Ranks Plus NCAA RPI and Balanced RPI Strength of Schedule Contributor Ranks

Conference NCAA RPI, NCAA Non-Conference RPI, and Balanced RPI Ranks Plus NCAA RPI Strength of Schedule Contributor Ranks

 Predicted NCAA Tournament Automatic Qualifiers, Disqualified Teams, and At Large Selection Status, All for the Top 57 Teams

The table shows Florida, Michigan State, Iowa, SMU, and Georgia all in the Top 57 but being disqualified by having a winning percentage below 0.500.  Lots will change by the end of the season, so I look at these teams only as illustrating, for now, the effect of the change from valuing ties as 1/3 of a win rather than 1/2 a win.  Of the five teams, my program shows Georgia as being below 0.500 whether the tie value is 1/3 or 1/2.  It shows all of the others being below 0.500 with the tie value set at 1/3, but above 0.500 with the value set at 1/2.

If the number of disqualified teams persists, I wonder if the Committee will expand the candidate pool beyond the historic #57 ranked team cutoff.

 

2024 ARTICLE 2: THIS WEEK'S NEWS AND UPDATED PREDICTIONS

News

There are two big news items this week.  The first is that it appears there will be changes to the NCAA RPI formula, effective this season:

1.  Rather than ties counting as half a win, as has been the case under the RPI formula previously, they now will count as 1/3 of a win.  This now will match how conferences compute conference standings.  It also will match a change already made for men's soccer.

2.  The bonus and penalty adjustment structure will change.  Under the new structure there will be three bonus tiers (rather than the previous two tiers).  The highest bonuses will be for wins and ties against teams ranked 1 to 25.  The middle bonuses will be for wins and ties against teams ranked 26 to 50.  The lowest bonuses will be for wins and ties against teams ranked 51 to 100.  These new tiers match two things:  (1) How the NCAA presents data about teams to the Committee for use in the NCAA Tournament bracket formation process; and (2) How the Committee looks at results, in terms of their quality.  Under the new structure there will continue to be two penalty tiers, but the tiers will be much broader.  The lower penalties will be for ties and losses to teams ranked 151 to 250; and the greater penalties will be for ties and losses to teams ranked 251 and poorer. 

The second item is that the Women's Soccer Committee is proposing that, effective for the 2025 season, it discontinue using the KP Index as a secondary rating system to the NCAA RPI and that it instead use the Massey ratings.  In my opinion, this would be a great change, as the Massey ratings have minimal discrimination in relation to conferences and regions, unlike the NCAA RPI and the KP Index.

Updated Predictions

Here are updated predicted end-of-season ranks, after incorporating all actual game results through Sunday, August 18.  These predictions include the RPI formula change of ties counting as 1/3 win rather than 1/2 win.  They do not include the changes to the bonus-penalty structure since we do not yet know what the new bonus and penalty amounts will be.

Team NCAA RPI, NCAA Non-Conference RPI, and Balanced RPI Ratings and Ranks

In this table, I have added two columns to what I showed last week.  The SoS Contribution Rank ARPI 2015 BPs column shows, for the NCAA RPI, each team's rank under the NCAA RPI formula as a strength of schedule contributor to its opponents' RPIs.  The SoS Contribution Rank URPI 50 50 SoS Iteration 15 column shows each team's rank under the Balanced RPI formula as a strength of schedule contributor.  If you compare teams' NCAA RPI ranks to their NCAA RPI Strength of Schedule Contributor ranks, you will see that they can be quite different.  For the Balanced RPI, you will see that the Balanced RPI ranks and Balanced RPI Strength of Schedule Contributor ranks are essentially identical.  The NCAA RPI's inconsistencies between its RPI ranks and its strength of schedule contributor ranks are a major problem and follow patterns that are the cause for the NCAA RPI's discrimination in relation to conferences and regions.

Conference NCAA RPI, NCAA Non-Conference RPI, and Balanced RPI Ratings and Ranks

In this table, I have added one column to what I showed last week.  The Conference ARPI SoS Contribution Rank column on the right shows, for the NCAA RPI, each conference's rank under the NCAA RPI formula as a strength of schedule contributor to its opponents' RPIs.  If you compare the NCAA RPI ranks to the RPI strength of schedule contributor ranks, you will see which conferences benefit from the disconnect between NCAA RPI's ranks and strength of schedule contributor ranks and which conferences are hurt.

Predicted Final NCAA Tournament Automatic Qualifiers and At Large Selection Status

These are the Top 57 in the RPI ranks at large candidate group, arranged in order of their likelihoods of getting at large selections.

2024 ARTICLE 1: PRE-SEASON PREDICTIONS FOR THE 2024 SEASON

Each year, my computer applies a program to the full season schedule to predict where teams will end up in the RPI rankings at the end of the season (including conference tournaments).  It also predicts where teams will stand in relation to automatic qualification for and at large selections (and also seeds) for the NCAA Tournament.  This article shows the program's pre-season predictions.

But first some background on the program, which I will put in this article and not repeat in my weekly updates.  You can use the background to do your own assessment of the reliability of the program's predictions.

Background

The program predicts the outcome of each game by comparing a pre-season NCAA RPI rating I have assigned to each team involved in the game and taking into account the value of home field advantage.  (Home field is worth 0.0166 in relation to the rating difference between the two opponents.)  In predicting the outcome of each game, for purposes of the rating computations, the program does not simply award the higher rated team a win.  Rather, for each game it assigns a win, tie, and loss probability based on historic probabilities in relation to opponents' location-adjusted rating differences.  This avoids, for example, having a team with a 51.0% win probability in each of 10 games being awarded a win in each game.  Instead, it creates a record for the team of 5.1 wins, 2.8 ties, and 2.1 losses, reflecting that the win probability in each game is 51.0%, the tie probability is 27.7%, and the loss probability 21.3%.  From a statistical perspective, this is the record one would expect for the 10 games (of course, rounded off to whole numbers for actual game outcomes) rather than expecting the team to win all 10.

Teams' predicted RPI ratings are based on teams' average Balanced RPI ranks over the last 7 years.  I use the last 7 years because, of the possible measures I have considered, using that average produces the best match with the next year's ratings.  The measures I have considered include average NCAA RPI ranks over the last 1, 2, 3, ... 15 years and average Balanced RPI ranks likewise over those numbers of years.  Once I have the 7 year average Balanced RPI ranks of teams, I put them in order from best to worst, assign them rank positions, and then assign each team the average historic NCAA RPI rating associated with its rank.

Using this method, the program does best at predicting where teams' ranks will end up at the top and bottom ends of the rankings and the poorest in the middle.  This is because the ratings are most spread out at the top and bottom of the rankings and most compressed in the middle.  Thus at the top and bottom of the rankings, a predicted rating "error" of X may not be sufficient to change teams' ranks, in other words will be inconsequential.  In the middle of the rankings, however, the same rating "error" may be equivalent to a significant number of rank position changes.

From an NCAA Tournament perspective, the teams that matter are the NCAA RPI Top 57 (see below).  Based on a review I did applying the program to the 2022 season, a reasonable expectation for teams ending up in the NCAA RPI Top 100 is that teams in the following rank groups will have actual end-of-season ranks, on average, within the indicated number of positions of their pre-season predicted ranks:

Teams ranked 1 through 10:  2 positions

Teams ranked 11 through 20:  3 positions

Teams ranked 21 through 50:  8 positions

Teams ranked 51 through 100:  12 positions 

The other thing to bear in mind is the NCAA RPI rank groups that, based on past history, are candidate groups for NCAA Tournament seeds and at large selections at the end of the season.  These are as follows:

#1 Seeds: teams ranked #7 or better

#2 Seeds: teams ranked #14 or better

#3 Seeds: teams ranked #23 or better

#4 Seeds: teams ranked #26 or better

#5-6 Seeds: teams ranked #30 or better

#7-8 Seeds: teams ranked #49 or better  

At Large: teams ranked #57 or better

The way the system works, the initial predicted end-of-season results I will show here are based on predicted results for all games.  At the end of each weekend, when all the previous week's actual game results are available, I change the predicted results for those games to the actual results and re-compute to get new end-of-season predictions.  Thus at the beginning of the season, the predicted end-of-season results are likely to have the reliability indicated above.  Each week, as I substitute actual results for predicted results, the predicted end-of-season results should be more reliable.

Further, after Week 5 of the season, I will stop using the pre-season predicted ratings as a basis for predicting future game results and instead will use teams' then current actual RPI ratings as the basis for predictions.  This will coincide with the NCAA's release of its first official RPI ratings and ranks for the season.

Predicted Final Ratings and Ranks

Here are this year's predicted final ratings and ranks.  They include the NCAA RPI, NCAA Non-Conference RPI, and Balanced RPI.  For detailed explanation of those three sets of ratings, use these links:

NCAA RPI

 NCAA Non-Conference RPI

Balanced RPI 

In the table, the Adjusted RPI 2015 BPs is the NCAA RPI, the Adjusted NCRPI 2015 BPs is the NCAA Non-Conference RPI, and the URPI 50 50 SoS Iteration 15 is the Balanced RPI.  The regions are based on the states where the teams are located and where the teams from their states play the majority or plurality of their games.

 Predicted Final NCAA Tournament Automatic Qualifiers and At Large Selection Status

The following table is for the Top 57 teams in the above NCAA RPI ranks, which historically has been the NCAA Tournament at large selection candidate group.  It has the teams arranged in order of their likelihood of getting at large selections.  (Later in the season, I will include some refinements related to likely at large selections and also likely seeds.)  For a detailed explanation of the table and what it means, use this link:

NCAA Tournament At Large Selection and Seed Patterns

The table also shows which teams the program identifies as Automatic Qualifiers and which ones it identifies as having below 0.500 records and thus being disqualified from at large selection.  The most likely at large selections are the Top 34 teams in the table that are not Automatic Qualifiers and not disqualified due to below 0.500 records.

Projected Final Conference Ratings and Ranks