Every qualified quarterback ranked on two play-by-play measures: expected points added (EPA) per dropback and completion percentage over expected (CPOE). Moves forward after every completed week.
Through week 9 of 2026, Joe Burrow of the Cincinnati Bengals is the No. 1 quarterback on our EPA + CPOE composite: +.278 EPA per dropback and a CPOE of +9.7 points. He also has the best CPOE; Brock Purdy of the San Francisco 49ers has the best EPA per dropback at +.329. The top five: Joe Burrow, Brock Purdy, Drake Maye, Caleb Williams, Josh Allen. 31 passers qualify (at least 14 pass attempts per team game) over 136 games.
Joe Burrow was No. 2 after week 8, when Caleb Williams led. Biggest climb this week: Jordan Love, up 4 (No. 15 to No. 11). Biggest drop: Caleb Williams and Geno Smith, down 3 (No. 1 to No. 4; No. 7 to No. 10). New to the table this week: Baker Mayfield.
Ranked on our composite: the average of each passer’s standard score in EPA per dropback and in CPOE among this week’s 31 qualifiers, so 0 is the qualifier average and +1 is one standard deviation better. The two measures agree more often than not (correlation 0.72 across the qualifiers), and the rank in each is shown beside it. The arrow is the change in composite rank since the same table through week 8.
| Rank | Move | Quarterback | Team | Composite | EPA/dropback (rank) | CPOE (rank) | Success | Dropbacks | Attempts |
|---|---|---|---|---|---|---|---|---|---|
| 1 | ▲ 1 | Joe Burrow | CIN | +1.92 | +.278 (3) | +9.7 (1) | 55.2% | 337 | 296 |
| 2 | ▲ 3 | Brock Purdy | SF | +1.44 | +.329 (1) | +4.9 (5) | 54.9% | 275 | 236 |
| 3 | = | Drake Maye | NE | +1.30 | +.219 (7) | +6.7 (2) | 51.3% | 337 | 272 |
| 4 | ▼ 3 | Caleb Williams | CHI | +1.28 | +.226 (6) | +6.4 (3) | 50.4% | 238 | 195 |
| 5 | ▼ 1 | Josh Allen | BUF | +1.18 | +.289 (2) | +4.0 (7) | 48.7% | 300 | 236 |
| 6 | = | Trevor Lawrence | JAX | +0.90 | +.258 (4) | +2.8 (8) | 50.0% | 260 | 224 |
| 7 | ▲ 2 | Jared Goff | DET | +0.65 | +.244 (5) | +1.3 (12) | 50.8% | 309 | 273 |
| 8 | = | Jayden Daniels | WAS | +0.65 | +.106 (14) | +4.9 (4) | 46.3% | 231 | 180 |
| 9 | ▲ 1 | C.J. Stroud | HOU | +0.56 | +.198 (11) | +1.8 (11) | 49.1% | 332 | 282 |
| 10 | ▼ 3 | Geno Smith | NYJ | +0.54 | +.102 (15) | +4.2 (6) | 47.1% | 331 | 271 |
| 11 | ▲ 4 | Jordan Love | GB | +0.51 | +.213 (9) | +1.1 (14) | 44.7% | 385 | 341 |
| 12 | ▲ 1 | Patrick Mahomes | KC | +0.43 | +.149 (12) | +2.2 (9) | 49.8% | 305 | 263 |
| 13 | ▼ 2 | Dak Prescott | DAL | +0.38 | +.215 (8) | +0.1 (16) | 48.8% | 400 | 336 |
| 14 | ▼ 2 | Kirk Cousins | LV | +0.35 | +.202 (10) | +0.2 (15) | 49.7% | 358 | 315 |
| 15 | ▲ 1 | Bryce Young | CAR | +0.06 | +.082 (17) | +1.3 (13) | 44.4% | 358 | 308 |
| 16 | ▼ 2 | Matthew Stafford | LA | -0.02 | +.107 (13) | +0.0 (17) | 49.3% | 383 | 346 |
| 17 | = | Kyler Murray | MIN | -0.22 | -.023 (27) | +2.0 (10) | 39.7% | 199 | 154 |
| 18 | = | Tyler Shough | NO | -0.24 | +.055 (19) | -0.2 (19) | 47.0% | 362 | 308 |
| 19 | ▲ 3 | Bo Nix | DEN | -0.25 | +.085 (16) | -1.1 (22) | 42.5% | 341 | 287 |
| 20 | ▼ 1 | Deshaun Watson | CLE | -0.26 | +.043 (20) | -0.1 (18) | 46.6% | 328 | 249 |
| 21 | ▼ 1 | Jacoby Brissett | ARI | -0.52 | -.020 (25) | -0.3 (20) | 44.9% | 352 | 305 |
| 22 | ▼ 1 | Cam Ward | TEN | -0.59 | -.018 (24) | -0.9 (21) | 43.6% | 298 | 248 |
| 23 | = | Jalen Hurts | PHI | -0.70 | +.002 (22) | -2.2 (25) | 38.3% | 188 | 154 |
| 24 | = | Michael Penix Jr. | ATL | -0.71 | +.068 (18) | -4.0 (27) | 46.8% | 235 | 199 |
| 25 | = | Sam Darnold | SEA | -0.76 | -.005 (23) | -2.5 (26) | 45.2% | 248 | 211 |
| 26 | = | Aaron Rodgers | PIT | -0.85 | -.051 (28) | -1.9 (24) | 42.2% | 334 | 288 |
| 27 | = | Malik Willis | MIA | -1.14 | -.020 (26) | -4.8 (29) | 42.1% | 278 | 208 |
| 28 | ▲ 2 | Jameis Winston | NYG | -1.23 | +.003 (21) | -6.1 (31) | 43.7% | 199 | 166 |
| 29 | - | Baker Mayfield | TB | -1.40 | -.220 (31) | -1.5 (23) | 37.0% | 184 | 142 |
| 30 | ▼ 1 | Daniel Jones | IND | -1.57 | -.159 (29) | -4.3 (28) | 41.5% | 349 | 302 |
| 31 | = | Justin Herbert | LAC | -1.69 | -.171 (30) | -4.9 (30) | 37.5% | 309 | 235 |
Passers with at least 100 dropbacks but fewer attempts than the rule asks. Their numbers use the same definitions but are left out of the ranks and the standard scores.
Split each qualifier’s season into two halves by alternate dropbacks (his first, third, fifth and so on against his second, fourth, sixth) and the two halves rank the passers alike only in part: the correlation between the halves is 0.37 for EPA per dropback, 0.33 for CPOE and 0.39 for the composite. Scaled up to the full sample (the Spearman-Brown formula, 2r / (1 + r)), that is a reliability of about 0.54, 0.50 and 0.56. In plain terms, about 54% of the differences in EPA per dropback between passers are lasting (the passer together with his line, receivers and play-calling) and the rest is the luck of a few hundred plays.
The margin on a single number is wide. A typical qualifier’s EPA per dropback carries a 95% margin of about ±0.16 (±0.14 to ±0.25 across the 31), so two passers a few places apart in the middle of the table are not clearly different. Measured against the qualifier average of +.090, 3 of the 31 are clearly above it on EPA per dropback (by more than their own margin), 3 clearly below, and the other 25 cannot yet be told apart from average on that measure alone.
Kept in this browser only.
Pick the games left and see both conferences’ seeds under the NFL’s tie-breakers.