Methodology

Overview

Our power rankings use a recursive strength-of-schedule algorithm that evaluates teams not just on their win-loss record, but on the quality of their opponents and their opponents' opponents. The algorithm iterates until ratings converge, then applies head-to-head, series, and streak adjustments.

Base Rating Formula

Each team's rating is a weighted blend of three components:

Recursive Iteration

The algorithm runs iteratively:

1. 1Round 1: Each team starts with their recency-weighted win percentage. Games later in the season count more than early-season games.
2. 2Round 2+: Ratings are recalculated using opponents' ratings from the previous round, with all components weighted by recency. Losses to low-rated teams incur an extra penalty, while road wins earn a bonus.
3. 3Convergence: The process repeats until no team's rating changes by more than 0.0001 between rounds. This typically takes 15-17 rounds.

Head-to-Head Adjustment

After the base ratings converge, we apply a head-to-head adjustment:

• When two teams have played each other and are close in rating, the series winner gets a bump.
• The adjustment scales with closeness (bigger effect when teams are nearly tied) and dominance (3-0 sweep matters more than 2-1).
• A quality gate ensures sub-.500 teams don't receive inflated h2h boosts from playing strong opponents.
• Maximum adjustment: 3% per matchup.

Series Bonus / Penalty

College baseball is defined by weekend series. We detect 3-game series (same opponent within a 3-day window) and apply adjustments:

• Winning a series (2-1): Bonus scaled by opponent quality — beating a strong team earns a bigger boost.
• Sweeping a series (3-0): Additional bonus on top of the series win reward.
• Losing a series (1-2): Penalty scaled by upset factor — losing to a weaker opponent hurts more.
• Getting swept (0-3): Additional penalty beyond the series loss.

Recency Weighting

Not all games are created equal. Games played later in the season carry more weight, allowing teams to show momentum and recover from early stumbles:

• Opening day games are weighted at 0.6x.
• The most recent games are weighted at 1.4x.
• Weight scales linearly by game date, so a March win counts roughly 2.3x more than a February win.
• This affects all three rating components: quality win %, opponent strength, and opponents' opponents.

Loss Scaling

Teams with a high number of losses receive a quadratic penalty that prevents schedule strength alone from propping up their ranking:

• Teams with 5 or more losses are penalized based on how far above 4 losses they are.
• The penalty grows quadratically: 5 losses = small hit, 8 losses = significant drop, 12 losses = major penalty.
• This ensures teams can't rank in the top tier purely on schedule strength while losing frequently.

Streak Factor

Current form matters. Teams on hot or cold streaks receive a final adjustment:

• Winning streaks of 3+ games earn a bonus that scales with streak length (diminishing returns after 8).
• Losing streaks of 3+ games incur a penalty on the same scale.
• Short streaks (1-2 games) are treated as noise and ignored.

Power Score (1-1000)

Each team receives a power score from 1 to 1000 based on their ranking position. The #1 team scores 1000, the last-place team scores 1, with linear interpolation between. Color-coded tiers:

Data

All schedule and results data is updated daily. Rankings are recomputed after each day's games are finalized.