Methodology

What Is a Resume Ranking?

The Quality Resume model is designed to evaluate what a team has accomplished during a season. It is not primarily designed to predict which team would be favored in a hypothetical game.

The central question of the model is:

To answer that question, every game is evaluated using two forms of context: the quality of the opponent and the way the team performed in the game. A victory over a strong opponent generally contributes more to a resume than a comparable victory over a weak opponent, while the margin of the result provides additional information about the quality of the performance.

Because the strength of an opponent depends on that opponent's own results, the model treats the season as a connected system rather than a collection of independent games.

1. Establishing Initial Ratings

The process begins with a schedule-based rating built from wins, losses, and who each team has played. This creates a neutral starting estimate before opponent strength and margin are introduced.

The initial ratings are based on the results and connections created by the schedule. Teams are not evaluated in isolation. Every game creates a relationship between two teams, and those relationships connect the entire college football schedule into a larger network.

For example, if Team A defeats Team B, and Team B later defeats Team C, the results involving all three teams help establish their relative positions in the initial rating system.

The matrix records the schedule: a team is connected to every opponent it played, while its own row also reflects its number of games and a small stabilizing adjustment. The right-hand side begins from the team's win-loss record. Solving the full system lets schedule relationships affect every starting rating.

Solving this system gives r₀, the initial rating vector. These values are not the final rankings; they are the first estimate used to evaluate opponent strength.

2. Calculating Opponent Strength

Once each team has a rating, that rating is used to estimate the strength of every opponent it has faced.

The raw team ratings cannot be used directly because they exist on a different numerical scale than the margin component of the model. Opponent ratings are therefore transformed onto a comparable scale between 0 and 1.

The scale is calculated from FBS ratings only. Teams are weighted by how often they appear in FBS-versus-FBS matchups, so the center represents a typical FBS opponent rather than a simple average of teams.

The model uses the middle half of that weighted distribution to measure spread, which limits the effect of unusual extremes. A bounded sigmoid then translates the standardized rating onto the opponent-strength scale. The curve is deliberately asymmetric: it preserves useful separation in the upper tier while preventing the lower tail from overwhelming the rest of the field. The exact settings are intentionally not published.

This avoids relying on a single best or worst team as a fixed endpoint. Opponent strength still responds to actual rating gaps within a season, so a separated elite tier can receive more credit without assigning values by rank.

FCS opponents are evaluated from their own completed schedules, including games against other FCS teams. Their normalized strength is pulled partway toward the lower bound to reflect the subdivision adjustment. Results can still distinguish stronger and weaker FCS opponents; they are not all assigned the same value.

3. Transforming Margin of Victory

Wins and losses alone do not describe the complete result of a game. The model therefore converts the point margin into a margin performance value between 0 and 1.

A value near 1 represents a dominant victory, while a value near 0 represents a substantial loss. Results near the middle of the scale represent close games.

The function g is built from a small set of margin anchors with linear interpolation between them. It gives extra credit for moving from a narrow result to a clear win, then applies diminishing returns and a ceiling to very large margins.

For example, the difference between winning by 1 point and winning by 10 points can provide meaningful additional information. However, the difference between winning by 50 points and winning by 60 points should not necessarily have the same impact.

The margin transformation therefore allows the value of larger margins to gradually level off rather than increasing indefinitely.

Overtime is handled separately. An overtime winner receives a modest positive margin value and the opponent receives the complementary value. This prevents a multi-overtime score from being treated like a conventional large-margin victory.

4. Calculating Game Performance

Every game is assigned a single performance score by combining opponent strength with the margin value.

Where:

The weights satisfy the relationship:

This ensures that the performance score remains a balanced combination of the two components.

The published model uses tuned weights to balance opponent quality and margin, without making either component the entire result. The performance score is continuous rather than simply a win or a loss: a strong close loss can score better than a weak narrow win, while a convincing win over a strong opponent scores especially well.

5. Converting Performances into Team Ratings

After every game has been assigned a performance score, both teams receive a fractional result. Let pᵢg be team i's performance and pⱼg its opponent's performance in game g. Team i receives pᵢg fractional wins and pⱼg fractional losses; team j receives the reverse.

A game therefore does not have a fixed amount of resume credit. Its total connection weight is the two teams' combined performances. A high-quality matchup can carry more influence than a weak-versus-weak game, while the credit remains balanced: every performance enters once as a fractional win and once as a fractional loss.

The model then uses those fractional wins and losses in the same schedule-based system that produced the starting ratings. For a team i, its result term is:

The schedule matrix keeps one row for every team and one weighted link for every game. A game's connection weight is its combined performance, so a team's own entries reflect its weighted schedule connections plus a small stabilizing adjustment, while matchup entries are negative. Solving the system, written as A r = b, produces a new rating vector. In other words, a rating is not merely the average of a team's performance scores: it also reflects how those performances connect through every opponent's schedule.

A team that consistently performs well against strong opponents will therefore tend to rise above a team whose strongest results came primarily against weaker competition. This is one of the main reasons the model is a resume evaluation system rather than a simple win-loss ranking.

6. The Iterative Process

The model contains an important circular relationship.

To evaluate a game, the model needs to know how strong the opponent is. However, an opponent's strength depends on the results and performances from its own schedule.

This means opponent strength and team ratings must be solved together through repeated calculations.

Each iteration uses the results of the previous iteration as a new estimate of team strength.

To keep the calculation from overreacting to a single pass, each newly solved rating vector is blended with the previous one before the next pass begins. This damping step makes the path to a stable ranking smoother while still allowing the schedule to reshape the ratings over time.

7. Convergence

The iterative process continues until the team ratings stabilize.

After each iteration, the model measures how much the ratings changed from the previous calculation.

The model examines the mean absolute change across all teams. When that change falls below a small convergence threshold, the ratings are treated as stable.

Early passes normally move ratings the most. Later passes make smaller adjustments as the schedule relationships settle:

As the ratings converge, the effect of the schedule becomes incorporated throughout the entire network of teams.

Why the Iteration Matters

Consider a team that defeats an opponent early in the season. At the time of that game, it may be unclear how valuable that victory should be.

If that opponent later defeats several strong teams, the quality of the original victory should increase. Conversely, if the opponent performs poorly throughout the remainder of the season, the value of the victory may decrease.

The iterative process allows these relationships to develop naturally. A game is not evaluated only by the information available on the day it was played. Its value is influenced by the complete network of results included in the ranking.

The Purpose of the Model

College football schedules are unequal. Teams play different opponents, compete in different conferences, and face varying levels of competition.

A simple win-loss record does not capture those differences. Two teams with identical records may have achieved those records against dramatically different schedules.

The Quality Resume model attempts to account for that context by asking two questions about every result:

The final ranking represents the model's attempt to combine those answers across an entire season into a single measure of resume quality.

Each update uses the games included through the selected reporting period, then recalculates the full connected schedule rather than simply adding points to the prior table. FBS and FCS teams are evaluated in the same connected model and displayed in separate ranking tabs. Subdivision membership comes from the input data for that season, so a team that changes subdivisions is ranked with the correct group for each year.

Records and game reports include completed games against other subdivisions too, but those opponents do not appear in the FBS or FCS ranking tables. Unplayed forfeits and cancelled games are not treated as performances. The 2020 season includes the FCS games played in spring 2021.

The week selector lists regular-season weeks first. Conference championship week and Army–Navy share the next entry, followed by one final postseason entry containing all bowl games and every playoff round in both subdivisions. Each entry includes the earlier reporting groups as well. FCS playoff games that take place before conference championship weekend still belong to the final postseason group, not a regular-season update. The 2020 season lists its spring regular-season weeks separately and skips the empty winter weeks.

The Complete Calculation