2024 ARTICLE 11: POST-WEEK-9 ACTUAL RATINGS AND UPDATED PREDICTIONS

 Current Actual RPI Ratings, Ranks, and Related Information

The following tables show actual RPI ratings and ranks and other information based on games played through Sunday, October 13.  The first table is for teams and the second for conferences.

Note: Scroll to the right to see all the columns.

Predicted Team RPI and Balanced RPI Ranks, Plus RPI and Balanced RPI Strength of Schedule Contributor Ranks

The following table for teams and the next one for conferences show predicted end-of-season ranks based on the actual results of games played through October 13 and predicted results of games not yet played.  The predicted results of future games are based on teams' actual RPI ratings from games played through October 13.

In the table, ARPI 2015 BPs is ranks using the NCAA's 2024 RPI Formula.  URPI 50 50 SoS Iteration 15 is using the Balanced RPI formula.

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

Starting this week, I will show predicted #1 through #8 seeds and at large selections based on the Women's Soccer Committee's historic decision patterns.  With a good part of the season remaining to be played, there will be changes, including substantial changes, from the predictions.  My main purpose now is to show how the prediction system works.

The first table below is for potential #1 seeds.  The #1 seeds always have come from the teams ranked #1 through 7 in the end-of-season RPI rankings, so the table is limited to the teams predicted to fall in that rank range.  The table is based on applying history-based standards to team scores in relation to a series of factors, all of which are related to the NCAA-specified criteria the Committee is required to use in making at large decisions.  For each factor, there is a standard that says, if a team met this standard historically, the team always has gotten a #1 seed.  I refer to this as a "yes" standard.  For most of the factors, there likewise is a standard that says, if a team met this standard historically, it never has gotten a #1 seed.  This is a "no" standard.  In the table, I have sorted the Top 7 RPI #1 seed teams in order of the number of yes standards they meet and then in order of the number of no standards.

This table has the #1 seed candidates sorted in order of the number of "yes" #1 seed standards they meet, from largest number to smallest (1 Seed Total), and the number of "no" standards they meet, from smallest to largest (No 1 Seed Total).  In the table, North Carolina and Mississippi State each meets at least one "yes" standard and no "no" standards, so each historically would get a #1 seed.  Duke and Penn State meet both "yes" and "no" standards, which means they have profiles the Committee has not seen historically.  Arkansas and Southern California meet no "yes" and no "no" standards, which means their getting or not getting #1 seeds would be consistent with Committee historic patterns.  Wake Forest meets no "yes" and 5 "no" standards, which means it historically would not get a #1 seed.  Altogether this means we have North Carolina and Mississippi State as #1 seeds, with Duke, Penn State, Arkansas, and Southern California as candidates for the remaining two #1 seed positions.

When there are more #1 seed candidates remaining than there are positions to fill, historically the best indicator of which teams the Committee will pick is a standard that combines a team's RPI rank with its conference final standing position, with each weighted at 50%.  For conference standing, the standard uses the average of the team's regular season conference standing position and its conference tournament finishing position.  Using this standard produces the following table, where the lower the standard value the better:

The system thus predicts that in addition to North Carolina and Mississippi State, the Committee will give #1 seeds to Duke and Arkansas.

Going through a similar process for #2 seeds, where the candidates are teams ranked #1 through #14, produces the following initial table:

In this table, Iowa, Wake Forest, and Stanford are clear #1 seeds.  Penn State, Southern California, and Auburn are potential #2s.  The last four teams are not #2s.  The tiebreaker for #2 seeds is a standard combining teams' conference standing with their Top 60 Head to Head Results Ranks:

Thus Southern California joins Iowa, Wake Forest, and Stanford as #2 seeds.

For #3 seeds, the candidates are teams ranked #1 through #23, producing the following initial table:

Here, Penn State and Michigan State are clear #3 seeds.  Xavier, Florida State, TCU, Auburn, UCLA, and Georgetown are candidates.  The remaining teams are not #3 seeds.  For #3s, the tiebreaker is teams' ranks in terms of their common opponents results in relation to results of other Top 60 teams.

Thus TCU and Auburn join Penn State and Michigan State as #3 seeds.

For #4 seeds, the candidates are teams ranked #1 through #26, producing the following initial table:

Here, Florida State and Ohio State are clear #4 seeds.  Utah State, Xavier, UCLA, Georgetown, and Pepperdine are candidates.  The remaining teams are not #4 seeds.  For #4s, the tiebreaker is a combination of teams' ranks in terms of their results against Top 50 opponents and their conferences' ranks.

Thus UCLA and Georgetown join Florida State and Ohio State as #4 seeds.

For # 5 through #8 seeds, the candidates are teams ranked #1 through #49.  Although the data are limited since we have had those seeds for only a few years, the best indicator of which teams will get those seeds is a combination of teams' RPI ranks and their Top 60 Head to Head results ranks:

Using this table, the #5 seeds are Utah State, Xavier, West Virginia, and Vanderbilt.  The #6s are St. Louis, Virginia, Liberty, and Pittsburgh.  The #7s are Wisconsin, Notre Dame, Pepperdine, and South Carolina.  The #8s are Virginia Tech, Western Michigan, Kentucky, and Oklahoma State.

For the remaining At Large selections, the candidates run up to RPI #57, producing the following initial table:

Here, Rutgers, Texas Tech, Texas, and Minnesota are clear At Large teams, with 6 additional spots to fill.  Buffalo, Oklahoma, Connecticut, Arizona, Georgia, Colorado, BYU, Washington, and Tennessee are candidates.  The remaining teams are not at large selections.  For At Large, the tiebreaker is a combination of teams' RPI ranks and their ranks in terms of their Common Opponent results in relation to other Top 60 teams.

Thus the unseeded at large selections are Rutgers, Texas Tech, Texas, and Minnesota, joined by Arizona, Georgia, Buffalo, Oklahoma, BYU, and Colorado.  Connecticut, Washington, and Tennessee just miss getting at large positions.

What If the Committee Were Using the Balanced RPI?

If the Committee were using the Balanced RPI, which does not have the NCAA RPI's problem of discrimination in relation to conferences and regions, Washington, California (which would move inside the Top 57 and thus be a candidate), and Tennessee would get at large selections and BYU, Oklahoma, and Buffalo (which would move outside the Top 57 and thus not be a candidate) would not.  

2024 ARTICLE 10: POST-WEEK-8 ACTUAL RATINGS AND UPDATED PREDICTIONS

 Current Actual RPI Ratings, Ranks, and Related Information

The following tables show actual RPI ratings and ranks and other information based on games played through Sunday, October 6.  The first table is for teams and the second for conferences.

As an item of interest, during the non-conference phase of the season, the percentage of tie games was relatively low as compared to the historic percentage of entire season tie games.  Now that the conference phase of the season has begun, however, the percentage of tie games is increasing and appears likely to end up in the vicinity of the historic entire season percentage.  This suggests that the percentage of tie games tends to be higher for in-conference competition than for non-conference competition.  If one thinks of the conferences as groupings of relatively comparable teams and of the non-conference phase as allowing competition among less comparable teams, it makes sense there would be more ties during the conference season.

Note: Scroll to the right to see all the columns.

Predicted Team RPI and Balanced RPI Ranks, Plus RPI and Balanced RPI Strength of Schedule Contributor Ranks

The following table for teams and the next one for conferences show predicted end-of-season ranks based on the actual results of games played through September 29 and predicted results of games not yet played.  The predicted results of future games are based on teams' actual RPI ratings from games played through October 6.  I consider the current RPI ratings still to be speculative, but they should become better game result predictors each week as the season progresses.

In the table, ARPI 2015 BPs is ranks using the NCAA's 2024 RPI Formula.  URPI 50 50 SoS Iteration 15 is using the Balanced RPI formula.

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

Starting this week, I will show the candidate groups for #1 through #4 seeds and for at large selections, placed in order based on the Women's Soccer Committee's historic decision patterns.

The first table below is for potential #1 seeds.  The #1 seeds always have come from the teams ranked #1 through 7 in the end-of-season RPI rankings, so the table is limited to the teams predicted to fall in that rank range.  The table is based on applying history-based standards to team scores in relation to 118 factors, all of which are related to the NCAA-specified criteria the Committee is required to use in making at large decisions.  For each factor, there is a standard that says, if a team met this standard historically, the team always has gotten a #1 seed.  I refer to this as a "yes" factor.  For most of the factors, there likewise is a standard that says, if a team met this standard historically, it never has gotten a #1 seed.  This is a "no" factor.  In the table, I have sorted the Top 7 RPI #1 seed teams in order of the number of yes factors they meet and then in order of the number of no factors.  For each of the other seed tables and for the at large table, I have followed a similar pattern.

As I said in the preceding paragraph, I use 118 factors for the yes decisions but fewer for the no decisions.  In the past, I have used all 118 for both yes and no decisions.  I am using fewer for the no decisions due to the NCAA this year having reduced the value of ties from 1/2 a win to 1/3 a win when computing the Winning Percentage portion of the RPI formula.  This change will result in almost all teams having lower RPI ratings than they have had in the past.  Each of the "no" factors I am not using this year is a factor that incorporates teams' RPI ratings.  For example, historically, no team with an RPI rating less than 0.6433 has gotten a #1 seed, so <0.6433 is the standard for a "no" #1 seed decision.  This year, however, with ratings as a whole being lower, that standard most likely is too high.  I can review and re-set all of the no standards that incorporate RPI ratings, but that is too big and time-consuming a task to do during the season, so it will have to wait until after the season is over.  In the meantime, the best approach is simply to not use the no standards that incorporate RPI ratings.  I still use, however, all the other no standards including those that incorporate RPI ranks (as distinguished from RPI ratings).

Here is the #1 seed table:

In the table, the 1 Seed Total column shows the number of yes standards a team met.  The No 1 Seed Total shows the number of no standards it met.  The table suggests that currently, North Carolina and Wake Forest look like sure #1 seeds and Mississippi State looks a strong possibility.  After that, it could be any of Duke, Penn State, Arkansas, or Iowa, whose order in the table is not necessarily the order of likely selection.

This is the table for #2 seeds.  The historic candidates are teams ranked #1 through 14.  The table includes the teams that are candidates for #1 seeds.

This is the table for #3 seeds, with the historic candidate group being teams ranked #23 or better.

This is the table for #3 seeds, with the historic candidate group being teams ranked #26 or better.

The final table is for at large selections, with the historic candidate group being teams ranked #57 or better:

There will be 34 at large teams.  One way to look at this table is to count down the list until you get to the 34th team that is not an Automatic Qualifier.  That takes you to Wisconsin, with 0 yes and 0 no standards.  Since Minnesota, below Wisconsin on the list, also is 0-0, the list suggests that currently the most likely at large teams are all of teams from Georgia and above that are not Automatic Qualifiers, plus all but one of the 0-0 teams from Texas A&M to Minnesota.  The one question mark of the teams from Georgia and above is Santa Clara with 2 yes and 3 no standards.

Note:  If you expected to see a team on the list and it is not there, it is because the current prediction has the team ending with a rank poorer than #57.  Also, as you can see, the current prediction has SMU ending as the #53 RPI team but with a Winning Percentage below 0.500 and thus being disqualified from getting an at large position.

2024 ARTICLE 9: POST-WEEK-7 ACTUAL RATINGS AND UPDATED PREDICTIONS

Current Actual RPI Ratings, Ranks, and Related Information

The following tables show actual RPI ratings and ranks and other information based on games played through Sunday, September 29.  The first table is for teams and the second for conferences.

Note: Scroll to the right to see all the columns.

Predicted Team RPI and Balanced RPI Ranks, Plus RPI and Balanced RPI Strength of Schedule Contributor Ranks

The following table for teams and the next one for conferences show predicted ranks based on the actual results of games played through September 29 and predicted results of games not yet played.  The predicted results of future games are based on teams' actual RPI ratings from games played through September 29.  I consider the current RPI ratings still to be speculative, but they should become better game result predictors each week as the season progresses.

In the table, ARPI 2015 BPs actually is ranks using the NCAA's 2024 RPI Formula.  URPI 50 50 SoS Iteration 15 is using the Balanced RPI formula.

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

The following table has the Top 57 teams in order from the most likely to the least likely to get at large positions in the NCAA Tournament.  It uses the Top 57 because in the past, no team ranked more poorly than #57 as gotten an at large position.  For some of the teams identified as Automatic Qualifiers, from weaker conferences, the "At Large In Total" numbers are relatively meaningless.

2024 ARTICLE 8: EFFECTS OF THE 2022 NO OVERTIME RULE AND THE 2024 CHANGES TO THE RPI FORMULA

[NOTE:  Since initially publishing this article, I have refined the method for computing the relationship between conferences' and regions' ratings (and regions' percentage of tie games) and their actual performance as compared to expected performance.  The article's charts and related tables now are based on the refined method rather than the previous cruder method.]

In 2022, an NCAA rule change did away with overtime games during the regular season, except for conference tournaments.  This year, the NCAA, as proposed by the Women's Soccer Committee, has changed the RPI formula for Division I women's soccer.  The purpose of this article is to show the effects of these changes in relation to (1) NCAA Tournament at large selections and (2) how the RPI functions as a rating system.

The 2024 RPI Formula Changes

There are two basic changes to the RPI formula:

1.  The RPI formula for a team's Winning Percentage has been changed from

(Wins + 1/2 Ties)/(Wins + Losses +Ties)

to

(Wins + 1/3 Ties)/(Wins + Losses + Ties)

Notes: 

(a) Winning Percentage account's for 50% of the RPI rating's effective weight.  Strength of Schedule, made up of Opponent' Winning Percentage and Opponents' Opponents' Winning Percentage, accounts for the other 50%.  In computing Strength of Schedule, the previous valuation of ties at 1/2  remains unchanged.

(b) The NCAA requires that a team have an overall won-lost record of at least 0.500 to be eligible for an at large position in the NCAA Tournament.  If ties are part of the won-lost computation, the previous valuation of ties at 1/2 likewise remains unchanged. 

2.  The bonus and penalty regime, which applies adjustments to the RPI for good wins and ties and poor losses and ties, has changed.  The old regime had high bonuses for wins and ties against teams with Unadjusted RPI ranks from #1 to 40 and lower bonuses for wins and ties against teams ranked #41 to 80.  It had low penalties for losses and ties against the second-to-bottom 40 teams in the rankings and higher penalties for losses and ties against the bottom 40.  In the new regime, there are three bonus tiers: #1 to 25, #26 to 50, and #51 to 100.  There are bonuses for wins and ties in the first two tiers, but only for wins in the #51 to 100 tier.  The penalty tiers are #151 through 250 for the lower penalties and #251 and poorer for the higher penalties.

The bonus and penalty amounts are as follows:

Tier 1, bonuses for wins and ties v opponents ranked #1 to 25:

Wins:

Away 0.0032, Neutral 0.0030, Home 0.0028

Ties:

Away 0.0020, Neutral 0.0018, Home 0.0016

Tier 2, bonuses for wins and ties v opponents ranked #26 to 50:

Wins:

Away 0.0026, Neutral 0.0024, Home 0.0022

Ties:

Away 0.0014, Neutral 0.0012, Home 0.0010

Tier 3, bonuses for wins v opponents ranked #51 to 100:

Wins:

Away 0.0008, Neutral 0.0006, Home 0.0004

Tier 4, penalties for losses and ties v opponents ranked #151 to 250:

Ties:

Away -0.0004, Neutral -0.0006, Home -0.0008

Losses:

Away -0.0010, Neutral -0.0012, Home -0.0014

Tier 5, penalties for losses and ties v opponents ranked #251 and poorer:

Ties:

Away -0.0016, Neutral -0.0018, Home -0.0020

Losses:

Away -0.0022, Neutral -0.0024, Home -0.0026

Ranking Effect of the 2024 Bonus and Penalty Regime

The following tables show the difference between the 2024 Formula's Unadjusted RPI ranks and the Adjusted RPI ranks under the new bonus and penalty regime.  The first table covers all teams.

The next table is for only the Top 57 teams:

These tables compare teams' Unadjusted RPI ranks to their Adjusted RPI ranks and thus are indicators of how much the bonus and penalty regime affects the rankings.  The tables -- especially the one for the Top 57 -- show that the bonuses and penalties have little effect on teams' rankings.  This particularly is noteworthy since small ranking differences within rating systems such as the RPI (or any other accepted sport rating system) are relatively meaningless.

For more detail here are tables that show how the URPI to ARPI rank difference amounts are distributed for All Teams and for the Top 57:

Effect of the No Overtime and RPI Formula Changes on NCAA Tournament At Large Selections

The changes altogether will have a significant cumulative effect on NCAA Tournament at large selections.

In reaching this conclusion, I started with a basic assumption:

The candidates for at large selections will be the teams ranked #1 through #57 in the Adjusted RPI rankings.

This assumption reflects what historically has been the case: Since 2007, all at large selections have come from teams ranked #1 through #57 by the Adjusted RPI.  Although this could change, a change seems unlikely, so this assumption provides a good base for analyzing the effect of the no overtime and RPI formula changes.

The following table shows basic data on the effects of the changes on at large selections.  The table contains the source data for the summary tables that follow.  You can scroll to the right to see the entire table.

The table is based on the seasons from 2010 through 2023, excluding Covid-affected 2020.  The 2024 Formula URPI and ARPI columns are based on applying the 2022 No Overtime rule and the 2024 RPI Formula changes to the data for each season.  The Actual ARPI column is based on applying to each season the overtime rule and RPI formula in effect for that season.  The No OT [Overtime] column is based on applying the current no overtime rule to each season, but using the RPI formula in effect for that season.  Together, these columns allow a determination of the effect of each change on at large selections -- the change to no overtimes, the change to a 1/3 weight for ties in determining winning percentage, and the change in the bonus and penalty regime.

The following table summarizes the information in the above table.  Remember, the numbers are spread over the 13 seasons from 2010 through 2023.

The first three data columns show the numbers of teams that have been brought into the Adjusted RPI Top 57 and thus become candidates for NCAA Tournament at large positions.  As you can see, the No Overtime rule has the greatest effect on this, bringing 32 teams into the Top 57.  Next comes the change to a 1/3 weight for ties, bringing in 8 teams.  Last comes the new bonus and penalty regime, bringing in only 2 teams.  Since the data cover 13 years, this means the changes will bring an average of just above 3 new teams per year into the at large candidate pool.  As the fifth column indicates, one team entering the Top 57 is due to it having had a below 0.500 winning percentage under the actual rules in effect for it but an above 0.500 record if the no overtime rule were in effect.

Of the new teams brought into the at large candidate pool, it is not possible to know whether they would have received at large positions.

Next are three Actual At Large, No Longer ... columns.  These show teams that actually received at large positions but that would not have received them under the new rules, due to their having dropped out of the Top 57 candidate pool.  Of these, 16 dropped out due to the no overtime rule, 6 due to the change in tie weight to 1/3, and 1 due to the new bonus and penalty regime.  Altogether this is just under 2 teams per year that received at large positions but that would not receive them under the No Overtime rule and the 2024 RPI Formula.

Following these are three Not Actual At Large, No Longer ... columns.  These show teams that were Top 57 candidates for at large positions but that did not get them.  Of these, 24 dropped out of the Top 57 candidate group due to the no overtime rule, 4 due to the change in tie weight to 1/3, and 0 due to the new bonus and penalty regime.

Of the teams dropping out of consideration for at large positions, the fourth column shows that 9 of them went from having winning percentages of 0.500 or better under the actual rules in effect for them to having a below 0.500 record if the No Overtime rule were in effect.

(Note: The number of teams moving into the Top 57 and out of the Top 57 do not match, as some teams moving in and out were Automatic Qualifiers, which are not included in the summary numbers.)

As these numbers show, the No Overtime rule has a significant effect on at large selections.  The change in tie weight to 1/3 has a small effect.  The new bonus and penalty regime has very little effect.

Effect of the No Overtime and RPI Formula Changes on How the RPI Functions as a Rating System

The next set of information shows how the No Overtime and 2024 RPI Formula changes will affect the RPI as a rating system: Will they make it better or worse or will they have little effect?

The above table, and the ones that follow, compare how different versions of the RPI perform as rating systems.  In the above table, the first row shows how the NCAA Unadjusted RPI, as in effect in 2023, performs.  The second row is for the NCAA Adjusted RPI as in effect in 2023.  These first two rows use as their data base games played from 2010 through 2021 (about 30,000 games), to show how the systems performed when there were overtime games.

The third row likewise is for the NCAA Adjusted RPI as in effect in 2023, but with games played from 2010 through 2023 as the data base and with all overtime games treated as ties.  By comparing this row to the one above it, you can see the effect of the No Overtime rule.

The fourth and fifth rows are for the NCAA Unadjusted and Adjusted RPI with the 2024 formula changes, with 2010 through 2023 as the data base and with all overtime games treated as ties.

Finally, the sixth row is for my Balanced RPI, with 2010 through 2023 as the data base and with all overtime games treated as ties.  I have included the Balanced RPI to show what can be achieved using the NCAA RPI's architecture as a base but with multiple additional calculations.

The above table shows simply the percentages of games that the higher rated teams (after adjustment for home field advantage) won, lost, and tied for each system.  Since the bottom four systems are based on no overtime games and therefore fewer ties, the best comparison column is the Overall % Correct Disregarding Ties column.  This column looks only at games that were won or lost and shows the percentage of games that the better rated team won.

The most significant information in that column is that with the change to the the No Overtime rule, the RPI became a significantly more accurate measure of teams' performance - -- the increase in accuracy from 81.0% with overtimes to 82,8% with no overtimes is a quite large improvement as rating systems go.  This improvement is not surprising.  With overtimes, about 20% of games when to overtime with about half of those games (~10% of all games) decided by golden goals.  When the NCAA formula considered the golden goal games, it treated them just the same as other win-loss games, thus not recognizing that the two teams' performances were very close to equal.  With overtimes eliminated, the formula now treats the performances as equal, which is a much better measure of the teams' performances than the old win-loss treatment.

This table is similar to the preceding one, but is limited to games involving at least one Top 60 team -- in other words, is related more or less to teams competing for NCAA Tournament at large positions.  It is notable that again, the RPI performs significantly better with no overtimes.

It also is worth noting that under both of the above tables, with the changes to the 2024 formula, the RPI's performance is slightly poorer than under the pre-2024 formula.

The next tables and charts address how the RPI systems do at rating teams from conferences and regions in relation to teams from other conferences and regions and individual teams in relation to other individual teams.

The tables and charts are based on game result probabilities, which in turn, for each rating system, are based on a result probability table unique to that rating system.  A result probability table comes from an analysis of all games played since 2010 and is very accurate, as the following table for the 2024 RPI Formula under a No Overtime rule shows:

On the left are the probable results for teams that are higher rated after adjustments for home field advantage, with the probable results derived using the applicable result probability table.  On the right are actual results.  As you can see, over a large number of game there is almost an exact match.

Looking at how a conference's teams do in non-conference games, if the rating system properly rates the conference's teams in relation to teams from other conferences, then the conference's win-tie-loss likelihoods in those games will match the actual results.  As the following table for the 2024 RPI Formula under a No Overtime Rule shows, however, that is not the case for this rating system.  Scroll to the right to see the entire table.

In the table, the three key columns are on the right.  They show, for a conference, its actual winning percentage, its likely winning percentage using the result probability table, and the difference between the two.  For example, using the ACC at the top, its actual winning percentage is 71.8%, its likely winning percentage based on its teams' and their opponents' ratings is 67.1%, and the difference is 4.7%.  Remember, if a rating system were properly rating the ACC in relation to the other conferences, one would expect the two winning percentages to be the same and the difference to be 0%.  What this means, for the ACC, is that it outperforms its ratings, in other words is underrated, with a 4.7% over-performance being a measure of the extent of the underrating.

By looking at each conference in this way, you can see which conferences the rating system underrates and which it overrates.  In addition, you can see which conference is most underrated, which is most overrated, and the difference between the two.  Thus looking at the above table, the most underrated conference is the ACC at 4.7% and the most overrated is SWAC at -7.4%, with a spread between the two of 12.2% (4.7% + 7.4%, rounded off).  And, looking at all the conferences, you can see the extent by which all of them as a group differ from the ideal 0% difference between their actual performance and their likely performance, in this case a total difference of 64.6%.  These numbers give you measures of the rating system's fairness when it comes to rating conferences' teams in relation to teams from other conferences.  And, when going through this process for multiple rating systems, it allows you to compare them in terms of how the do at rating conferences in relation to other conferences.  For the six rating systems covered by this report, this produces the following table:

The table shows an improvement in the RPI's fairness in relation to conferences when moving to the No Overtime rule.  Then, in the 2024 change to a 1/3 weight for ties when computing Winning Percentage, there is a slight degradation.  The addition of the 2024 bonus and penalty regime slightly reduces the degradation.  As a whole, the table shows that all of the NCAA RPI systems have a conference fairness problem

The next table shows each conference's average rating and the difference between the conference's actual winning percentage and its likely winning percentage using the applicable result probability table.  In addition, it has the teams arranged in order of their average RPI ratings.

The following chart is based on the data in the above table:

In the chart, the upper orange line (actually a series of data points) shows the conferences' average ratings, which are arranged in order from the best average rating on the left to the poorest on the right.  The lower black data points are for the differences between the conferences' actual winning percentages and their likely winning percentages.  The black data points are in the places where they fit in the rating spectrum from the highest rating on the left to the lowest rating on the right.  The black straight line is a computer generated trend line that shows how the conferences' actual versus likely performance differences change as the conferences' average ratings descend.  A simple look at the chart says that for conferences with higher average ratings, their actual results tend to be better than their ratings say they should be and for conferences with lower average ratings, their actual results tend to be poorer than their ratings say they should be.  In other words, the rating system discriminates against stronger conferences and in favor of weaker conferences.

Above the black line is a formula.  This is a formula that describes the black trend line and allows a computation of the extent of the ratings' discrimination.  In the formula, x represents the different data points associated with the trend line.  For the data at the extreme left, where the strongest conference is, x=1.  For the data at the extreme right, where the weakest conference is, x=213.  Thus by using the formula with x=1, one can calculate the actual v likely performance difference at the high point on the trend line at the left, and by using the formula with x=213, one can calculate the difference at the low point on the trend line at the right.  The spread between these two differences is the extent to which the rating system discriminates against strong conferences and in favor of weak conferences.  In the case of the above chart for the 2024 RPI Formula with No Overtimes, the computations produce an overperformance on the left of 4.3% and an underperformance on the right of -6.3%, for a spread between the two of 10.6%.

By going through the processes just described for multiple rating systems, it is possible to compare how they do in rating teams from conferences in relation to conference strength.  The following table shows how the six rating systems covered in this report compare:

As this table shows, the No Overtime rule slightly improved the RPI in relation to conference strength.  The change in tie weight to 1/3 had a small mixed effect, and the 2024 bonus and penalty regime likewise had a small mixed effect.  As a whole, all of the NCAA RPI formulas have a problem fairly rating the conferences in relation to each other and all discriminate in relation to conference strength.

The next table is the general fairness table but for teams categorized by four geographic regions within which teams are located: Middle, North, South, and West:

As the table shows, the NCAA RPI versions all have a general fairness problem in relation to regions.  Again the change to No Overtime resulted in an improvement.  For regions, the 2024 RPI Formula slightly degrades the No Overtime performance from what it was under the pre-2024 Formula.

This table comes from trend charts for regions, looking at performance in relation to the regions' average ratings.  Here, the full NCAA 2024 Formula is a slight improvement over the full 2023 Formula, with the 2024 bonus and penalty regime being responsible for the improvement.  For all of the NCAA RPI systems, however, there is a problem of discrimination in relation to region strength..

For regions, I use an additional metric, which is performance in relation to the percentage of the regions' games that are ties.  I use this metric because the regions have different levels of parity and the percentage of games that are ties is a surrogate for parity -- the higher the percentage of games that are ties, the more likely there is greater parity within the region.  At the RPI for Division I Women's Soccer website, on the RPI: Regional Issues page, I discuss a problem the NCAA RPI has, which is rating teams from regions fairly in relation to each other if the regions have different levels of parity.  There, I show that the West region has a higher level of parity than the other regions, followed by the Middle and North regions, with the South having the least parity.  The following chart is like the one above for conferences, but is for regions and relates regions' performance to their percentages of ties:

The following comparison table comes from charts like this for the six rating systems covered by this report:

As the table shows, unlike for the other metrics above, the change to No Overtimes had a negative impact on how the RPI functions, increasing discrimination among regions in relation to intra-region parity.  Further, the change in tie valuation to 1/3 has increased the discrimination even more.  Both of these effects are as expected, since regions with a higher proportion if ties -- those with more parity -- will suffer more devaluations of their Winning Percentages than regions with lower parity and fewer ties.  As a whole, all of the NCAA RPI systems have a discrimination problem in relation to intra-region parity: they underrate teams from regions with high parity and overrate teams from regions with low parity.

Looking at individual teams produces the following table:

The table again shows an improvement in fairness with the change to No Overtimes, but some degradation in the change to the 2024 Formula.

The last table, below, addresses a different issue.  Because of the way the NCAA RPI calculates strength of schedule, teams' NCAA RPI ranks can be quite different than their ranks as RPI Strength of Schedule contributors to their opponents.  This makes it possible for coaches to "trick" or "game" the RPI by scheduling with a view to opponents' likely Strength of Schedule contributor ranks rather than their likely RPI ranks.  It is one of the reasons basketball stopped using the RPI.

This table shows that the various changes to the RPI over time have slightly degraded the RPI in relation to this problem:  They have slightly increased the differences between teams' RPI ratings and their ratings as RPI Strength of Schedule contributors.  In the table, towards the right, I have blue highlighted the column for Rank v SoS Rank Difference Percent 15 of Fewer Positions, which shows the percentage of teams for which the difference between their RPI rank and their rank as RPI Strength of Schedule contributors is 15 or fewer positions.  I have highlighted this column because, based on my experience assisting coaches with scheduling, there is enough year-to-year variability that when teams' ordinary difference is 15 or fewer positions, it may not be worthwhile to consider that difference as significant for purposes of selecting good future opponents.  As that column shows, for all of the NCAA RPI versions, about 1/3 of teams are in the 15 or fewer group, which leaves 2/3 in the group where it is worthwhile for coaches to consider their likely Strength of Schedule contributions during scheduling so as to artificially maximize their RPIs.  As this column shows, where the higher the percentage the better the system, the bonus and penalty regimes, in particular, degrade the RPI, with the 2024 Formula performing the worst of all the NCAA RPI systems.  The differences among the NCAA RPI systems, however, are relatively small.

Summary as to Effects of Changes on How the NCAA RPI Functions as a Rating System

Looking at the cumulative effects of the No Overtime rule and the 2024 Formula changes as compared to a 2023 Unadjusted RPI formula baseline, they are relatively small.  The changes, however, make the RPI more discriminatory against regions with higher parity and in favor of regions with lower parity.  And, they make the RPI slightly more able to be "tricked."

Conclusion

Of the changes over the past few years, the change to No Overtimes has a significant effect on NCAA Tournament at large selections.  The 2024 change of tie weights to 1/3 for RPI Winning Percentage purposes has a small effect.  The change to the 2024 bonus and penalty regime has a very small effect.  The cumulative effect of the changes is significant.

In terms of how the RPI functions as a rating system, it functions better under the No Overtime rule with two exceptions: it is more discriminatory against regions with higher parity and it makes the system very slightly more susceptible to being "tricked" through smart scheduling.  The 2024 Formula change effects are mixed and small.

2024 ARTICLE 7: POST-WEEK-6 ACTUAL RATINGS, UPDATED PREDICTIONS, AND NEWS

Current Actual RPI Ratings, Ranks, NCAA Tournament Prospects, and Related Information

The following tables show actual RPI ratings and ranks and other information based on games played through Sunday, September 22.  The first table is for teams and the second for conferences.

In this week's teams table, there are color coded columns.  These show which teams, based on past history, are potential NCAA Tournament seeds and at large selections.

For example, the orange column for #1 seeds shows that historically, the poorest ranked team as of this stage of the season to ultimately receive a #1 seed is the currently ranked #22 team.  No team ranked more poorly than #22 has received a #1 seed. 

Also, in the column for at large selections, the color coding covers teams ranked #11 through 149.  This means that as of this stage of the season, those can be thought of as "bubble" teams for at large selection.  They might or might not ultimately get selected.  On the other hand, teams ranked #1 through 10, if not automatic qualifiers, can be considered "locks" for at large positions, meaning that historically teams ranked #10 or better as of this stage of the season always have gotten at large positions.

Note: Scroll to the right to see all the columns. 

Predicted Team RPI and Balanced RPI Ranks, Plus RPI and Balanced RPI Strength of Schedule Contributor Ranks

The following table for teams and the next one for conferences show predicted ranks based on the actual results of games played through September 22 and predicted results of games not yet played.  The predicted results of future games are based on teams' actual RPI ratings from games played through September 22, which makes the predicted results pretty speculative.

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

The following table has the Top 57 teams in order from the most likely to the least likely to get at large positions in the NCAA Tournament.  It uses the Top 57 because in the past, no team ranked more poorly than #57 as gotten an at large position.