Stats Corner: Understanding Basketball Ranking Methods

Every year, the NCAA Selection Committee faces the monumental task of seeding the NCAA Tournament, a decision heavily influenced by complex ranking systems. But what exactly are these systems, and how do they determine a team’s fate?

For decades, the NCAA relied on the RPI, but starting in the 2019 season, a new contender emerged: the NET. This transition wasn’t without its growing pains, sparking immediate controversy among coaches and athletic directors alike. N.C. State Athletic Director Debbie

Yow has voiced her disappointment: “Based on the metrics cited above that the NCAA indicated they would use to evaluate team performance, we are disappointed for our athletes, coaches, and fans that our total body of work was not rewarded with selection to the NCAA Tournament.” Similarly, TCU coach Jamie Dixon previously publicly challenged the system: “You look at the NET, there are six teams that had lower NETs than us. They created a tool and talked about it, and then there are six teams there that are lower than us that are in.” (A clear sentiment of a “Power 5” team feeling overlooked.)

While the NET system has undergone revisions since its inception, it remains the NCAA’s official ranking tool. But what about other influential metrics like KenPom? This week on Stats Corner, we will break down the old reliable RPI, the controversial NET, and the highly respected KenPom, exploring how each system evaluates and ranks NCAA basketball teams.

RPI

Let’s begin with the RPI, the NCAA’s standard from 1981 to 2018. At its core, the RPI was a straightforward win/loss formula:

RPI = (WP * 0.25) + (OWP * 0.50) + (OOWP * 0.25)

Where:

• WP (Winning Percentage): The team’s own winning percentage, calculated by dividing wins by total games played. This accounted for 25% of the RPI.
• OWP (Opponents’ Winning Percentage): The combined winning percentage of a team’s opponents, contributing a significant 50%.
• OOWP (Opponents’ Opponents’ Winning Percentage): The winning percentage of those opponents’ opponents, making up the final 25%.

What becomes immediately clear is that a substantial 75% of a team’s RPI was determined not by their own record, but by the performance of their Strength of Schedule (SOS) – specifically, how well their opponents, and their opponents’ opponents, performed. This emphasis rewarded teams playing tougher schedules. For instance, as of now, Boise State sits 18th in RPI at 0.637, narrowly ahead of Utah State at 22^nd with 0.629, while Air Force lags at 304th with 0.423.

NET Ranking

With the 2019 season came the NET: NCAA Evaluation Tool. And while the NCAA can royally screw things up, they did nailed the acronym. Unlike the formulaic RPI, the NET is a predictive-learning model, leveraging machine learning to simulate countless game outcomes. It constantly compares expected results with actual outcomes, adjusting its model dynamically.

Upon its introduction, the NET considered five primary factors:

1. Team Value Index: A holistic measure considering opponent quality, game location, and the ultimate winner.
2. Net Efficiency: The difference between a team’s offensive and defensive efficiency.
3. Winning Percentage: A straightforward measure of wins versus losses.
4. Adjusted Win Percentage: This factor heavily weighed game location, rewarding road wins and penalizing home losses.
5. Scoring Margin: Designed to reward dominant wins, though capped at 10 points to prevent extreme blowouts from skewing results.

An important note: the NET treats all games equally, regardless of when they are played in the season. A November clash carries the same weight as a February showdown, irrespective of team evolution or injuries.

The Quadrant System: A Game Changer

One of the most talked-about aspects of the NET is its Quadrant System, which categorizes wins and losses based on the opponent’s NET ranking and game location. You’ll often hear commentators discussing “Quad 1” or “Quad 2” records, and for good reason—these classifications significantly impact a team’s resume.

The quadrants are defined as follows:

• Quadrant 1: Home 1-30; Neutral 1-50; Away 1-75
• Quadrant 2: Home 31-75; Neutral 51-100; Away 76-135
• Quadrant 3: Home 76-160; Neutral 101-200; Away 136-240
• Quadrant 4: Home 161-plus; Neutral 201-plus; Away 241-plus

Let’s look at some examples:

• New Mexico (currently 46th NET): Holds a 1-2 record in Quad 1 games, 2-1 in Quad 2, and a perfect 5-0 in Quad 3.
• Boise State (currently 44th NET): Shows a 2-4 Quad 1 record, 3-0 in Quad 2, and 2-1 in Quad 3.

While Boise State has more losses (4 losses to 8 wins) compared to New Mexico (3 losses to 11 wins), their losses are against higher-caliber, Quad 1 opponents, which mitigates the negative impact on their ranking. New Mexico, despite fewer losses, has fewer Quad 1 wins and a critical Quad 3 loss. A Quad 1 or 2 loss isn’t a “deal-breaker,” but a Quad 3 or 4 loss can severely hurt a team’s tournament aspirations.

The fluidity of these rankings throughout the season is fascinating. A win that starts as Quad 1 can drop to Quad 2 or even 3 if the opponent’s NET ranking falls. This creates an interesting dynamic where teams often find themselves rooting for their past opponents to perform well!

For example:

• Nevada, currently 75th, means a game at Nevada would be a Quad 1 opportunity for San Diego State. However, for Utah State hosting Nevada, it would be a Quad 2 game.
• If Nevada then loses to Wyoming (86th) at home, that becomes a Quad 3 loss for Nevada, which could lower their overall NET ranking. This, in turn, could downgrade San Diego State’s Quad 1 win against Nevada to a Quad 2, and Utah State’s Quad 2 game to a Quad 3.
• Similarly, Utah State’s initial Quad 1 loss to South Florida became a Quad 2 loss after the Bulls dropped to 76th following losses to UAB (110th).  If the Bulls lose at North Texas, 151th, and home to East Carolina, 307th, it may become a Quad 3 loses for the Aggies, or South Florida could beat Tulsa, 41st, on the road and go back to Utah State Quad 1 loss. The continuous shifts demonstrate why these rankings are constantly scrutinized.

NET Evolution

Recognizing some of the initial criticisms, the NCAA refined the NET in 2020, dropping winning percentage, adjusted winning percentage, and scoring margin from its core calculations. Now, the key factors influencing a team’s NET ranking are:

• Efficiency: The difference between offensive and defensive points per possession.
• Strength of Opponents Played: The quality of the teams on their schedule.
• Game Location: Where the game was played still matters for quadrant classification.
• Wins and Losses: The fundamental outcome of games.
• Winner: Who ultimately won the contest.

KenPom

Beyond the NCAA’s official metrics, independent analytics Ken Pomeroy offers his highly influential KenPom rankings. Operating from his own autonomous website, KenPom assigns every NCAA team an Adjusted Efficiency Margin (AEM), calculated as Adjusted Offensive Efficiency minus Adjusted Defensive Efficiency.

You might be thinking, “Isn’t that what the NET does?” And you’re partially right, but with two crucial distinctions:

1. KenPom came first. It’s the pioneer in this analytical approach.
2. It’s all about “Adjusted” efficiency. While team efficiency typically measures points scored per 100 possessions, KenPom’s adjustment accounts for the pace of play and the number of possessions a team accumulates per game.

Consider this: Purdue, for example, boasts an exceptional offensive efficiency, projected to score 129 points per 100 possessions. However, their adjusted team efficiency is 65.8 possessions per 40-minute game, ranking them 313th nationally in pace. This means the Boilermakers are highly effective with fewer possessions, outperforming a team like last in the nation, Colorado State (63.0 possessions) even if they run fewer plays.

Let’s explore some of KenPom’s key metrics, using Utah State (26th in KenPom, 16th in NET) as our example. Interestingly, KenPom and NET often show discrepancies; for instance, KenPom places San Diego State at 49th, Boise State at 51st, and New Mexico at 54th, while the NET has them at 60th, 44th, and 46th respectively, highlighting the differing methodologies.

• Adjusted Efficiency Margin (NetRtg): This is KenPom’s ultimate ranking metric. A higher score indicates a better team. For Utah State, an AEM of 21.17 (26th nationally) suggests they would defeat an average NCAA team by approximately 21 points.
• Adjusted Offensive Efficiency (ORtg) and Adjusted Defensive Efficiency (DRtg): These metrics, as discussed, quantify how efficiently a team scores and defends, adjusted for pace. Utah State, for example, has an ORtg of 120.7 (25th) and a DRtg of 99.5 (36th).
• Adjusted Tempo (AdjT): This measures the average number of possessions a team has per game, adjusted to a standard 40-minute contest. It’s calculated with the formula: Possessions per game = Field goals attempted – offensive rebounds + turnovers + 0.475 x attempted free throws. This adjustment is crucial for comparing “run-and-gun” teams with slower, grind-it-out squads. Utah State plays at a slightly quicker pace with an AdjT of 69.7, ranking 117th.
• Luck Rating: This fascinating metric quantifies a team’s performance in close games. Statistically, teams are expected to win roughly 50% of one-possession games. The luck rating reveals the deviation from this expectation. A positive rating means a team is “luckier” by winning more than their share of tight contests. Utah State’s luck rating of 0.071 (66th) suggests they’ve been fortunate in close finishes. In contrast, 9-5 Wichita State is the unluckiest team (-0.211), while 11-4 Tulane is the luckiest (0.254), meaning the American Conference has both the luckiest and unluckiest teams in the country.
• Strength of Schedule (SOS): KenPom’s SOS is based on the average Adjusted Efficiency Margin, Offensive Efficiency, and Defensive Efficiency of a team’s opponents. A higher SOS indicates a tougher slate of games. Utah State’s opponents, for instance, have an average AEM of 2.07, signifying the 127^th toughest schedule nationally, meaning the Aggies’s opponents typically win their games by an average of 2 points.
• Non-Conference Strength of Schedule (NCSOS): This metric rewards teams that actively seek out challenging non-conference matchups rather than “cupcake” games. Since conference schedules are largely predetermined, the NCSOS reflects a team’s commitment to building a strong resume. Utah State’s average non-conference schedule (151st with an AEM of 0.61) shows a balanced approach. Compare this to Wyoming’s 307th NCSOS (-3.95 AEM), indicating their non-conference opponents typically lost by 4 points. The range is vast, from Fordham’s easiest NCSOS (-11.68) to Alabama’s hardest (12.45).

Key Differences

While both the NET and KenPom strive to quantify team strength, their methodologies lead to distinct differences:

1. Game Location: The NET heavily integrates game location (home, neutral, away) into its quadrant system, crucially impacting win/loss value. KenPom, on the other hand, does not factor game location directly into its core efficiency rankings.
2. Efficiency Calculation: KenPom utilizes adjusted efficiency, accounting for a team’s pace of play and number of possessions per game. The NET focuses on points per 100 possessions regardless of the pace, which can sometimes favor slower teams with fewer, more efficient possessions differently.
3. Scoring Margin: KenPom does consider scoring margin as part of its efficiency calculations. The NET initially included it but has since removed it as a direct factor, though efficiency still captures some aspects of dominance.
4. Wins & Losses: KenPom primarily focuses on efficiency and predictive metrics; wins and losses are an outcome, not a direct input for its core ranking calculations. The NET, while incorporating efficiency, still fundamentally uses wins and losses, particularly through its quadrant system, as a direct input for ranking.
5. Methodology: KenPom relies on a strict, formulaic (predictive) approach to generate its rankings. The NET employs a more dynamic learning algorithm that continually compares expected outcomes against actual results, adjusting its model accordingly.

Understanding these intricate ranking systems—RPI, NET, and KenPom—is essential for any serious college basketball fan. They each offer a unique lens through which to view team performance, highlighting different strengths and weaknesses. While the NCAA’s NET system serves as the official gauge for tournament selection, independent metrics like KenPom provide valuable context and often spark informed debate. As the season progresses, watching how these numbers fluctuate and what they reveal about teams beyond their simple win-loss record will make following college basketball even more compelling.