About NCAA Power Rankings

Why Another NCAA Power Rank?

Every season, as we approach the playoffs, a heated debate erupts over team qualifications. The discussions typically unfold like this:

"Washington won all their games - they've earned their spot!"

"But they're in a weaker conference. Look at Alabama's record in the SEC!"

"You're both missing the point - LSU beat Alabama. We need to consider quality wins!"

"That's blasphemous!..."

While each perspective has merit, the NCAA, unlike the NFL with its explicit playoff rules, needs an objective standard to ensure fairness. Even though the goal is not to complain about existing playoff rules (it is tough to make everyone happy), having a clear, data-driven approach is crucial for general audience have better understanding of a team's real caliber.

How is Power Rank Derived?

Our power ranking system uses a quantitative approach that considers multiple factors:

1. Wins and Losses

While win percentage is crucial, we weigh it differently from the current playoff ranking system, considering it alongside other important factors.

2. Win and Loss Impact

For concepts like 'quality wins' and 'head-to-head victories', we quantify each game's impact comprehensively. For example, when Notre Dame defeats USC, we evaluate:

  • USC's conference strength
  • The score differential
  • The strength of USC's victories
  • (a few other subtle ingredients)

USC's performance would factor into Notre Dame's victory. If USC is a top seed in the Big Ten, this win would significantly boost Notre Dame's power ranking and give them well-deserved recognition, even though Notre Dame competes in a relatively weaker conference.

3. Global Consideration

In a league with hundreds of teams, direct matchups between any two specific teams are rare. This often leads to conflicting opinions when analysts try to compare teams that haven't met on the field. Our solution? We analyze the entire network of games played throughout the season.

For instance, when LSU defeats both South Carolina and Ole Miss, who in turn have victories over Georgia and South Carolina, our algorithm recognizes LSU's competitive parity with Ole Miss and South Carolina, despite their four losses. At the same time, it distinguishes them from Georgia's tier of performance.

This network-based approach lets us evaluate "virtual" matchups more objectively. Take Notre Dame's impressive season, for example—the algorithm acknowledges their strength not just because they have only one loss, but because they're more aligned with a top-10 caliber sub-network, having outperformed teams in the top 10–25 sub-network.

Examples

Coming soon...