FQI: Full Press Quarterback Impact
FQI Full Press Quarterback Impact
Opponent-adjusted QB value, accuracy, consistency and avoidance.
FQI Leaderboard
Loading current rankings…
Early-season current rankings use a 75-dropback qualification threshold. Historical season leaderboards use 300+ dropbacks.
Historical FQI Explorer
Choose a season to view every qualified quarterback under the same v1.4 framework.
Top FQI Seasons
The highest season-relative FQI scores in the exact historical database. Sort any column to inspect the same seasons through a different lens.
Top QBV/Play Seasons
The best direct quarterback-created value per play seasons in the database. FQI remains beside it so raw value and contextual evaluation can be compared directly.
Individual Game Grades
Explore a quarterback’s performance in one game and download its player card.
Game FQI uses the v1.4 weights and grade colors, standardized across qualifying games in the same season. Minimum: 20 dropbacks with CPOE data. Game grades can change as the season’s comparison pool grows. Season rankings are calculated separately.
Choose a season to explore games.
| Week | Quarterback / matchup | Game FQI | Grade | Dropbacks | QBV/Play | OQBV/Play | Player card |
|---|
FQI Comparison Lab
Compare a single season, selected seasons or an available career. Each profile can use a different quarterback and time span.
Regular-season data · 2006 onward · qualifying seasons only. Career includes every available qualifying season for the selected quarterback. Period FQI is a dropback-weighted average of season scores.
The FQI v1.4 Method
Read the complete methodology & FAQ →
FQI is built to answer a different question than traditional passing statistics: how well did the quarterback himself play once throw difficulty, expected YAC, opponent strength, consistency and catastrophic outcomes are put into context?
Value
Primary pillar built on OQBV/play: QBV/play adjusted for opponent defensive strength using leave-one-game-out, QB-adjusted defensive effects. Raw QBV/play remains visible alongside it.
Accuracy
Primary pillar built on CPOE, rewarding completion performance relative to expected completion probability.
Consistency+
Context modifier: actual positive-play success minus the rate expected for that OQBV/play level.
Avoidance+
Context modifier: drive-killer avoidance above expectation for that OQBV/play level.
OQBV/play = QBV/play − opponent defensive effect
CORE = .50×OQBV_Z + .30×Accuracy_Z
RESIDUAL = clamp(.10×Consistency+_Z + .10×Avoidance+_Z, −.50, +.50)
FQI_RAW = CORE + RESIDUAL
FQI = 50 + 15×Z(FQI_RAW)
Game-state context
FQI v1.4 does not automatically discount garbage-time plays in the score. The research sweep found that doing so slightly worsened same-QB ordering and next-year validation. Instead, FQI exposes Competitive QBV/play, Low-Leverage QBV/play and Low-Leverage Share as transparent context diagnostics.
FQI vs. FQV
FQI is rate/quality. FQV is cumulative value above the league QB-play baseline. A quarterback can have an elite FQI in a smaller sample while another creates more total FQV over a full season.
Validated Across 20 NFL Seasons
FQI v1.4 was tested on 601 qualified QB seasons from 2006–2025 and audited across all 2,049 same-QB season pairs. The calibration reduces residual-driven ordering anomalies while keeping FQI descriptive rather than turning it into a copy of another metric.
Opponent adjustment is intentionally modest: across all 601 historical seasons, v1.4 and v1.3 correlate at 0.9992, with an average absolute FQI change of only 0.45 points. Year-to-year stability improves from .472 to .479 and next-year EPA association from .418 to .424. Garbage-time/leverage splits are reported separately because forcing leverage weighting into FQI slightly worsened the longitudinal audit.
