Basketball Analytics Glossary
Definitions of the key metrics and concepts used throughout TruRanking.
This glossary covers the basketball analytics terms you will encounter on TruRanking's rankings, predictions, bracketology, and tournament simulation pages. For a complete explanation of how these concepts fit together in our system, see the Methodology page.
Rating (TruRanking Rating)
A team's overall strength score. It blends four parts, weighted most heavily toward the first two: a game-results score (each game moves it up or down by how much the team won or lost by compared with what the two teams' ratings and home court predicted), efficiency margin, strength of schedule, and a small free-throw component. This is the primary ranking metric on the Full Rankings page. Higher ratings indicate stronger teams. The rating is recalculated daily after all new game results are processed.
Offensive Efficiency
Points a team scores per possession, compared with the Division I average and scaled so 100 is average: an Offense of 112 means the team scores 12% more points per possession than an average D-I team. Because it is measured per possession, it removes the effect of pace, so it means the same whether a team's games end in the 60s or the 90s.
Defensive Efficiency
Points a team allows per possession, on the same scale (100 = Division I average). Lower is better: a Defense of 92 means opponents score 8% fewer points per possession than against an average team. Like offensive efficiency, this metric is pace-independent, so a team that plays slowly and allows 58 points per game is not automatically a better defensive team than one that plays fast and allows 72.
Efficiency Margin
Offensive efficiency minus defensive efficiency. A team with an Offense of 110 and a Defense of 95 has an efficiency margin of +15. This single number captures a team's net performance on both ends of the floor and is one of the strongest predictors of success in college basketball.
Strength of Schedule (SOS)
How tough a team's schedule has been, based on opponents' records: two parts opponents' winning percentage to one part their opponents' winning percentage, counting games between Division I teams. It is shown as a percentile among Division I teams, so 100 is the toughest schedule and 0 the easiest. A higher SOS indicates that a team has played tougher competition. This metric is factored into each team's overall rating to ensure that records compiled against different levels of competition are weighted appropriately.
Pace
How many possessions a team's games have compared with the Division I average, scaled so 100 is average: a Pace of 105 means about 5% more possessions per game than average. Pace describes how fast or slow a team plays but does not by itself indicate quality. TruRanking's efficiency metrics are calculated on a per-possession basis specifically to remove the influence of pace from team evaluations.
Tempo-Free Analysis
An analytical approach that measures basketball performance independent of game pace. By evaluating teams on a per-possession basis rather than per-game totals, tempo-free analysis allows fair comparisons between teams that play at different speeds. This is the foundation of the entire TruRanking system. See the Methodology page for details on how tempo-free efficiency ratings are calculated.
Win Probability
The estimated likelihood that a given team will win a specific game, expressed as a percentage. On the TruPredictions, win probabilities are generated by feeding the TruRanking rating differential between two teams (adjusted for home-court advantage) into a logistic regression model. A team listed at 72% win probability is expected to win that game roughly 72 times out of 100 if the matchup were repeated.
Projected Spread
The expected margin of victory for the favored team in a given game, derived from the TruRanking model. A projected spread of -6.5 for the home team means the model expects them to win by approximately 6 to 7 points. The TruPredictions shows how the model's projected spread compares to the current market consensus line where available.
Against the Spread (ATS)
A measure of how often a team or model's predictions beat the point spread set by oddsmakers. If a team is favored by 7 points and wins by 10, they covered the spread. ATS records are useful for evaluating the model's ability to identify value relative to market expectations, beyond simply picking winners.
Automatic Qualifier (AQ)
A team that earns an NCAA Tournament bid by winning its conference tournament (or being projected to do so). There are 32 Division I conferences, each awarding one automatic bid. On TruBrackets, automatic qualifiers are projected by selecting the highest-rated team in each conference as the likely conference champion.
At-Large Bid
An NCAA Tournament invitation awarded to a team that did not win its conference tournament but has a strong enough overall profile to merit selection. The Selection Committee fills 36 at-large spots after the 32 automatic qualifiers are determined. TruBrackets projects at-large selections by taking the next-best available teams by overall TruRanking rating.
Seed
A number from 1 to 16 assigned to each team in the NCAA Tournament bracket, indicating relative strength within a region. The top four overall teams receive No. 1 seeds across four regions, the next four receive No. 2 seeds, and so on. Seeding determines first-round matchups: a No. 1 seed plays a No. 16 seed, a No. 2 plays a No. 15, and so forth.
Bubble Team
A team whose NCAA Tournament selection status is uncertain heading into Selection Sunday. Bubble teams are typically ranked in the range just inside or just outside the projected at-large cutoff. The TruBrackets Bubble Watch section shows the Last Four In (final at-large teams that made the projected field) and First Four Out (next-best teams that just missed).
Monte Carlo Simulation
A statistical technique that uses repeated random sampling to estimate outcomes. TruTourney runs over 10,000 Monte Carlo simulations of the full NCAA Tournament bracket, resolving each game based on win probabilities derived from TruRanking ratings. The frequency with which each team advances to a given round across all simulations produces its advancement probability estimates.
Logistic Regression
A statistical model used to predict the probability of a binary outcome (in this case, win or loss) based on one or more input variables. TruRanking's logistic model takes the rating differential between two teams as input and outputs a win probability. The model has been trained on thousands of historical Division I game outcomes to ensure reliable calibration.
Home-Court Advantage
The statistical benefit that a team receives from playing on its own court. In college basketball, the home team wins at a significantly higher rate than the away team, even after accounting for team quality differences. TruRanking's prediction model applies a home-court adjustment to the rating differential before generating win probabilities, reflecting this well-documented effect. Tournament games played at neutral sites receive no home-court adjustment.
Conference Strength Rating
The average TruRanking rating of all teams in a given conference, used to rank the 32 Division I conferences by overall quality. A conference with one dominant team but many weak programs will rate lower than a conference with consistent quality across the board. The Conference Rankings page displays these ratings alongside its members' combined record in games against Division I teams, average efficiency metrics, and mean strength of schedule for each league.
NET Rankings
The NCAA Evaluation Tool, used by the NCAA Selection Committee as one input in their tournament selection and seeding process. NET incorporates team results, strength of schedule, game location, scoring margin, and net efficiency. TruRanking is an independent system with its own methodology; while both use efficiency-based principles, TruRanking ratings are calculated separately and may differ from NET rankings for individual teams.