Methodology
March Metrics uses publicly available data and historical pattern matching to estimate each team's tournament potential. Here's how the Power Rating works.
Power Rating (0-100)
The Power Rating is derived from Bart Torvik's Barthag metric, which estimates a team's probability of beating an average Division I team. It uses a single, elegant formula based on adjusted efficiency:
This produces a 0-100 scale where higher is better. A rating of 95+ puts a team in championship contender territory. Every national champion since 2008 scored at least 89.9, and the median champion scores around 96.
Barthag outperforms custom composite formulas because it naturally captures the interaction between offense and defense through the power-law exponent, without needing hand-tuned weights or normalization. The 11.5 exponent has been empirically calibrated against decades of college basketball results.
Key Metrics Explained
Points scored per 100 possessions, adjusted for opponent quality. Higher is better. Champion average: ~121.
Points allowed per 100 possessions, adjusted for opponent quality. Lower is better. Champion average: ~90.
Field goal percentage adjusted for the extra value of three-pointers. Accounts for shot quality, not just volume.
Percentage of possessions ending in a turnover. Lower is better. Ball security under tournament pressure separates survivors.
Percentage of available offensive rebounds secured. Creates second-chance points, which are critical in low-scoring tournament games.
Free throw attempts per field goal attempt. Getting to the line reflects aggressiveness and the ability to draw fouls.
Possessions per game. Neither inherently good nor bad, but extreme tempo can become erratic against elite half-court defense.
Composite rating reflecting the difficulty of a team's schedule. Battle-tested teams historically perform better in March.
Historical Pattern Matching
Each team is compared against 17 years of tournament teams (2008-2025) using normalized Euclidean distance across seven efficiency metrics: adjOE, adjDE, tempo, eFG%, TO%, ORB%, and FT Rate. The three closest historical matches are shown, along with a similarity percentage and their actual tournament result. This provides context: if a team's closest match won a championship, that's a strong signal.
Tournament Probability Estimates
Round-by-round probabilities are estimated using a seed-based baseline modified by Power Rating and historical pattern similarity. The Power Rating maps to an "implied seed," which is blended with the actual seed to produce adjusted probabilities. These are not betting odds — they are analytical estimates designed to contextualize each team's championship path.
Data Sources & Attribution
The Power Rating is based on the Barthag metric created by Bart Torvik of barttorvik.com. All efficiency data uses pre-tournament Torvik Time Machine snapshots to ensure consistency with the data available at selection time.
- •Bart Torvik / T-Rank (primary efficiency data, Barthag, and Time Machine snapshots)
- •Sports Reference / College Basketball Reference
- •NCAA NET Rankings (official NCAA ranking system)
- •KenPom.com efficiency ratings (inspiration, not directly used)
- •Historical tournament results (2008-2025, 17 seasons)
March Metrics is an independent analytical project using publicly available data. Statistical data is sourced from Bart Torvik's T-Rank. Team logos are trademarks of their respective universities and conferences, used for identification purposes only. This site is not affiliated with, endorsed by, or sponsored by any university, conference, or the NCAA. All probability estimates are for educational and entertainment purposes only — not for wagering.