What Are NET Rankings?
The NET (NCAA Evaluation Tool) is the official ranking system used by the NCAA Division I Men's Basketball Committee to evaluate and seed teams for the NCAA Tournament. Introduced for the 2018-19 season, it replaced the RPI (Rating Percentage Index) that had been the standard metric since 1981 — a system widely criticized for being easy to manipulate through scheduling and for ignoring scoring margin entirely.
The NCAA partnered with Google Cloud Professional Services to develop the NET using machine learning techniques. The resulting model evaluates every Division I game through two primary lenses: Team Value Index (TVI), a result-based metric that measures the quality of wins and losses, and Adjusted Net Efficiency, which measures points scored vs. allowed per 100 possessions after adjusting for opponent strength and game location.
The exact formula is proprietary — it cannot be reverse-engineered or independently replicated. This opacity is intentional: the NCAA argues that a secret formula prevents teams from gaming the system. Critics counter that transparency would allow the public to better understand selection decisions. What is known: the NET processes every game result daily, weighting recent results alongside full-season performance to produce a ranking of all 362+ Division I teams.
How NET Rankings Are Calculated
The NET algorithm evaluates five components, each capturing a different dimension of team quality. Adjusted Net Efficiency is the most heavily weighted, but the combination of all five creates a more robust ranking than any single metric alone.
Team Value Index (TVI)
Result-based metric that evaluates the quality of each win and loss. Rewards beating quality opponents on the road and penalizes losses to weak teams at home. The backbone of how NET evaluates resume quality.
Adjusted Net Efficiency
Points scored minus points allowed per 100 possessions, adjusted for opponent strength and game location. The single most heavily weighted component. Measures how efficiently a team plays on both ends of the floor.
Winning Percentage
Simple win-loss record. Provides a baseline foundation, but carries less weight than efficiency-based metrics. A 25-6 record means more when the schedule is strong.
Adjusted Winning Pct.
Win-loss record weighted by opponent quality and game location. A road win against a top-25 team counts more than a home win against a 300+ team. Bridges the gap between raw record and resume quality.
Scoring Margin
Average margin of victory, capped at 10 points per game. The cap prevents teams from being rewarded for running up the score. A 10-point win and a 40-point blowout are treated identically.
Why Efficiency Matters Most
Adjusted Net Efficiency strips away the noise of pace and schedule strength. A team that plays 75 possessions per game and scores 80 points is not necessarily better than one that plays 65 possessions and scores 70. Per-possession efficiency is the single best predictor of how a team will perform against a given opponent, which is exactly what the Selection Committee needs when filling a 68-team bracket.
Source: NCAA.com. The exact weighting of each component is not publicly disclosed. Relative weights are estimated from published research and committee statements.
The Quadrant System
The quadrant system categorizes every game into one of four quality tiers based on the opponent's NET ranking and the game location. The thresholds shift by venue because home court advantage is real — Division I home teams win roughly 60-65% of the time. Beating a team ranked NET #60 on the road is genuinely harder than beating them at home, so the quadrant system gives road and neutral-site games more favorable thresholds.
Example: An opponent ranked NET #60 is a Quad 2 game at home (since #60 falls in the 31-75 range), but a Quad 1 game on the road (since #60 falls within the 1-75 away threshold). This means the same opponent can be a resume-building Quad 1 win if you beat them in their arena, or a less impressive Quad 2 win if you beat them in yours.
Wins that make resumes. Losses that are forgivable.
Good wins. Expected wins for contenders.
Wins that don't help. Losses that start hurting.
Wins are expected. Losses are resume killers.
Quadrants Are Dynamic
Because quadrant assignments depend on your opponent's current NET ranking, they change throughout the season. A November win that was Quad 2 at the time might become Quad 1 by March if your opponent improved, or drop to Quad 3 if they collapsed. The Selection Committee evaluates quad records based on the NET rankings as of Selection Sunday — not when the games were played.
NET vs KenPom vs BartTorvik vs RPI
The NET is not the only ranking system that matters in college basketball. Analysts, bettors, and even committee members reference multiple systems to build a complete picture. Each has different strengths, and understanding how they differ helps you interpret tournament selection decisions.
NET
FreeKenPom
$20/yearBartTorvik
FreeRPI
FreeWhich Should You Use?
For predicting tournament selection: focus on NET rankings and quad records, because that is what the committee uses. For predicting game outcomes and evaluating true team quality: KenPom and BartTorvik are more useful because their methodologies are transparent and their efficiency ratings are more predictive. The best analysts use all three — NET for selection context, KenPom for efficiency analysis, and BartTorvik for recent form.
Why NET Rankings Matter for March Madness
The Selection Committee uses NET rankings and quad records as their primary quantitative metrics when filling the 68-team bracket. While they also consider head-to-head results, road records, recent performance, and game film, the NET-driven quad record is the single most discussed data point in the selection room.
Bad losses override good wins. A team with 8 Quad 1 wins but 3 Quad 4 losses will face more scrutiny than a team with 5 Quad 1 wins and zero bad losses. The committee has consistently shown that they view Q3 and Q4 losses as disqualifying red flags — they suggest a team that cannot consistently perform against inferior competition is not tournament-worthy, regardless of their ceiling.
Historical Quad 1 Win Thresholds for At-Large Bids
Virtually guaranteed an at-large bid regardless of other factors. These teams have proven themselves against elite competition repeatedly.
2025-26: Auburn, Florida, Houston had 10+ Q1 wins
Very likely in the field. Need to avoid bad losses (Q3/Q4) and maintain a reasonable overall record to stay safe.
Typical range for 4-7 seeds on Selection Sunday
On the bubble. Need strong supporting metrics — good road record, no Q3/Q4 losses, solid NET ranking — to earn an at-large bid.
The 'last four in / first four out' conversation
Extremely unlikely to receive an at-large bid. Must win their conference tournament for an automatic qualification.
Most mid-major and low-major tournament teams
The Resume-Killer: Quad 3 and Quad 4 Losses
Even one or two Q3/Q4 losses can knock a team off the bubble entirely. The Selection Committee has repeatedly cited bad losses as the primary reason for excluding teams with otherwise strong profiles. A single road loss to a team ranked NET #200+ can undo months of quality wins. This is why coaches often say “every game matters” — in the quadrant system, it literally does.
Remember that quad records are dynamic. As opponents' NET rankings shift throughout the season, your quad record shifts with them. A November win over a team that was ranked #45 at the time might be a Quad 3 win by March if that team fell to #180. The committee evaluates the bracket based on Selection Sunday rankings, which means early-season wins against eventually-bad teams provide no shelter.
Frequently Asked Questions
How often are NET rankings updated?
NET rankings are updated daily during the college basketball season, typically in the early morning hours after the previous day's games are processed. The first NET rankings of the season are usually released in late November after teams have played enough games for meaningful evaluation. Rankings are not updated during the offseason.
What replaced RPI in college basketball?
The NET (NCAA Evaluation Tool) replaced the RPI (Rating Percentage Index) starting with the 2018-19 season. The NCAA partnered with Google Cloud Professional Services to build the NET using machine learning techniques. The NET is considered a significant improvement because it accounts for scoring margin (capped at 10 points), adjusts for game location, and uses per-possession efficiency metrics rather than raw win-loss records.
Do NET rankings determine NCAA Tournament seeding?
NET rankings are one of the most important factors the Selection Committee uses, but they do not solely determine seeding. The committee also evaluates quad records (wins and losses by quadrant), head-to-head results, road record, strength of schedule, recent performance trends, and the eye test. However, NET ranking and quad records are the two most quantifiable and heavily discussed metrics during the selection process.
Can a team with a low NET ranking still make the tournament?
Yes, through automatic qualification by winning their conference tournament. Any conference tournament champion receives an automatic bid regardless of NET ranking. At-large bids, however, heavily favor teams in the top 60-70 of the NET rankings with strong quad records. Historically, very few at-large teams have been selected with a NET ranking outside the top 75.
What is the scoring margin cap in NET rankings?
The NET caps the scoring margin at 10 points per game. This means a 10-point win and a 40-point blowout are treated identically for ranking purposes. The cap was designed to prevent teams from running up the score against weaker opponents. It encourages competitive play throughout the game rather than stat-padding in garbage time.
How does NET compare to KenPom?
Both NET and KenPom use efficiency-based metrics, but they differ in key ways. KenPom is a fully transparent, formula-based system that publishes offensive and defensive efficiency ratings, tempo, and four-factor statistics. The NET is a proprietary machine learning model whose exact formula is not publicly disclosed. KenPom tends to be more predictive of future performance, while NET has institutional authority as the NCAA's official ranking. Many analysts use KenPom for analysis and NET for tournament selection context.
Why do NET rankings sometimes disagree with the eye test?
NET rankings are purely mathematical and do not account for context that humans notice, such as injuries, lineup changes, or strength of play in the final minutes. A team that loses close games to elite opponents on the road may have a worse NET than their actual quality suggests. Conversely, a team that dominates a weak schedule may rank higher than the eye test would suggest. The Selection Committee acknowledges this by using both NET data and game film in their evaluations.