FANTASY GRIDIRON · METHODOLOGY

Matchup ratings

Schedule context, not projections or start/sit advice. Ratings estimate how a defense has changed same-position fantasy production after accounting for the offenses it faced.

Data

nflverse weekly player statistics and schedules. QB, RB, WR, and TE are modeled separately.

Pregame baseline

The opponent team-position group's mean in its other eligible completed 2026 games before the rating cutoff. The evaluated game is excluded. Missing expectations are omitted rather than fabricated.

Opponent adjustment

Actual whole-team position-group fantasy points allowed minus expected points, averaged by defense and position. Positive is favorable; negative is difficult.

Rolling weight

All eligible completed 2026 games before the rating cutoff receive equal weight. No recency weighting or prior-season fallback is used.

No future leakage

Every historical backtest rating uses only games available before the game being rated. Current 2026 matchup grades use only completed 2026 games before the published rating cutoff, and each evaluated game is excluded from its own offense-position baseline.

Matchup percentile

Values are ranked separately for QB, RB, WR, and TE within each scoring format. Percentile bands are: at least 0.90 Elite, at least 0.75 Awesome, at least 0.50 Great, at least 0.30 Good, at least 0.10 Average, and below 0.10 Trash.

Confidence

Low with fewer than four current-season defense games, Medium from four games, and High from eight games when sample variance is available.

Schedule swing

Following-two-week average minus the first upcoming matchup percentile; ±0.20 is required for Improving or Worsening.