What the Colley Matrix Is
The Colley Matrix is a mathematical ranking method for sports teams, developed by astrophysicist Wes Colley. It became widely known as one of the computer polls used by the Bowl Championship Series (BCS) in American college football from 2001 to 2013.
Its defining design choice is that it uses only wins and losses — no margin of victory, no home-field adjustment, no subjective weighting.
What the Colley Matrix Ranks
The method ranks teams within a defined competition — a league or division where teams play overlapping schedules.
Every team receives a rating derived from its record and the records of its opponents, producing an ordered ranking of the entire group.
Core Inputs Used by the Colley Matrix
The system requires exactly two inputs:
- Each team’s wins and losses
- The full schedule matrix — who played whom, and how many times
Deliberately excluded:
- margin of victory
- home or away venue
- injuries, weather, or any contextual factors
How Colley Matrix Ratings Are Calculated (High-Level)
The method builds on a statistical idea called Laplace’s rule of succession — a team’s expected winning probability starts at 0.5 and moves toward its actual winning percentage as games accumulate.
The key insight is adjusting that estimate for schedule:
- Each team’s effective record is rewritten in terms of its opponents’ ratings — a win over a strong team is worth more than a win over a weak one.
- Because opponents’ ratings depend on their opponents in turn, this produces a system of linear equations — the “Colley matrix.”
- Solving the system yields a rating for every team simultaneously, with strength of schedule emerging from the math rather than being fed in as an assumption.
Conceptual model: Every team’s record is re-scored against the true strength of its opponents — and everyone’s strength is solved for at once.
Update Frequency
The Colley Matrix is recalculated whenever new results are added — typically weekly during a season, as each round of games updates the matrix.
Known Limitations and Criticisms
- Ignores margin of victory — a one-point win counts the same as a blowout (a deliberate fairness choice, but a loss of information)
- Ignores venue — home and away games are treated identically
- Requires connected schedules — the method works best when teams share opponents; isolated groups are hard to compare
- Sensitivity to small samples — early in a season, ratings can be unstable
Where the Colley Matrix Is Used
The method is used in:
- College football analysis (as a BCS computer component from 2001–2013, and independently since)
- Rankings for other sports with round-robin-like schedules
- Academic teaching — it is a standard example of ranking via linear algebra
Summary
The Colley Matrix shows how much can be extracted from the humblest possible data: wins, losses, and the schedule. By solving for every team’s rating simultaneously, it turns strength of schedule from a subjective debate into a property of the equations — at the cost of ignoring everything about how games were won.
References and Sources
- Colley, W. N. Colley’s Bias Free College Football Ranking Method (method documentation).
- Wikipedia. Colley Matrix.
- Langville, A. N., & Meyer, C. D. Who’s #1? The Science of Rating and Ranking.