/// Quarterback intelligence, measured by results
The QBIQ Index
Decision-making and accuracy, graded by what actually happened. Every quarterback takes the same test — regular season, playoffs, and wins — with the supporting cast acknowledged, not excused.
nflverse ·
data updated Sep 25, 2026
Philosophy weights — drag to argue
Metric weights inside performance
Timeframe
View
All stats, records and scores recompute over the selected years. Click any row for the full case file.
| # | Quarterback | QBIQ ▾ | Starts | TD% | Y/A | Cmp% | INT% | EPA/db | Sack% | Rate | W–L | Win% | Margin | Playoffs | PW% | Rush yds·TD | Ctx |
|---|
The rules — same for everyone, every argument
Most quarterback debates die because the logic changes player to player. Ours doesn't. These rules are fixed, they apply to every QB on this board at once, and if one ever changes, it changes for everybody.
One formula, applied to all. No narratives, no exceptions, no goalposts on wheels.
We grade what happened: touchdowns, yards, completions, turnovers, points, wins. Not arm talent, not highlights, not combine numbers.
Every season is measured against the league average that year. 2010 football isn't graded on 2026's curve.
Sixty minutes of consistency beats last-minute drama. The QB who's great for three quarters banks more than the one who needs a miracle because he was average for three.
January is graded against what playoff QBs actually average, not September numbers. Getting there earns a little; winning there earns a lot.
Win rate and margin of victory both count — beating teams by two scores says more than surviving by a field goal. But per-play performance outweighs record, because 550 throws tell you more about a quarterback than 17 final scores he shares with 52 teammates.
Two bars: 10 attempts in the window to be listed at all, and a qualifying bar — 50 career starts, about 8 a season in shorter windows — to lose the ✱. Scores are never shrunk. We don't diminish the output; the sample does that on its own.
Every QB carries a Context Index for his supporting cast, shown next to his score. It never silently changes the math — you can see it, edit it, and toggle it.
No cherry-picked splits, no massaged numbers. Kneel-downs count as carries and sacks count as plays, for everybody. The stat sheet is the stat sheet.
How the QBIQ is scored
Era-adjusted, not raw
Every season is compared to the league average that year (qualifying QBs, with the bar scaled to season length), so stat lines from different eras are judged against their own leagues. A QB's mark in each metric is his attempt-weighted distance from average, in standard deviations, over the years you've selected.
The metrics
TD%, yards per attempt, completion percentage, EPA per dropback (the situational-value stand-in for success rate), and INT% as the penalty. Sack rate is shown for reference — EPA already charges for sacks. The rushing toggle folds rushing TDs, yards and EPA into the same metrics per play.
Your era is your yardstick
Every quarterback is graded only against the years he actually played — a 2009 season against the 2009 league, a 2025 season against the 2025 league — weighted by his own throws. Years he wasn't in the league never touch his score, and backup years count exactly as much as he played. Win percentage and margin are measured against standards that never move: .500 and zero. The one comparative piece is playoff value: playoff wins are ranked against the other QBs over the selected years. That's why a score can shift slightly when you change the timeframe even though the QB's own stats didn't — five playoff wins mean more against the last five Januaries than against twenty of them. That's on purpose: the timeframe picks the field he's competing with, never the yardstick he's measured by.
Three buckets
Regular-season performance, playoff performance (same metrics, but graded against the playoff field — January stats are judged against what playoff QBs actually average, not the regular-season league), and winning — regular-season win percentage and average margin of victory as the starter (decisive wins count more than coin-flip escapes), plus playoff wins, with Super Bowl wins counting double and every playoff start worth a fraction of a win — getting there counts. If a QB has few playoff attempts in the window, the unused playoff weight flows back to the regular season instead of punishing him for games he never got to play.
Early-season scores swing — on purpose
In September every sample is tiny and every score is shown raw: expect hot starts and ugly ones that a full season will sand down. The Starts column and the asterisk tell you exactly how much football backs each number. And in a view where nobody has played playoff football yet (like the current season), the playoff share of the score is simply handed back to the regular season — no one is docked for games that haven't happened.
When playoffs help or hurt
In any window that includes playoff games, the rule is: getting there earns a little, winning there earns a lot, and January play is graded against the playoff field — the league's best defenses — not against easy September averages. A QB whose playoff numbers dip but who still outplays the other playoff quarterbacks gains ground, not loses it.
Small samples
Two bars, no dilution. Bar one: 10 attempts in the window — fewer and you're not listed, because a trick play isn't a sample. Bar two: the qualifying minimum — 50 combined starts for career views, scaled to about 8 starts a season for shorter windows (and prorated while a season is live). Career-scale views open with only qualified QBs; shorter windows open showing everyone, flagged. Below bar two a QB wears the asterisk and hatched bar, and one toggle hides him. His score is never shrunk or adjusted — we don't diminish the output; the sample is here to do that.
Context, acknowledged
We don't pretend to split credit play-by-play. The Context Index is now per team per season: weapons, O-line, run game and play caller are rated 1–10 (authored 2020 onward, neutral before, current season carried forward from last year), defense is derived from points allowed — rolling last 17 games for the young current season. A QB's index is the play-weighted blend of the situations he actually played in over your selected years. The optional adjusted view applies a modest nudge: one point of context ≈ one point of score. Edit any team-season in a QB's case file; edits stay in this browser.
Stats: nflverse (play-by-play–derived season totals) · Records from game results with the listed starting QB · Context ratings are editable opinions, not measurements.
© 2026 Scorigin Sports · The QBIQ Index