Round 18 · 2026 season
United States Grand Prix
P1
ANT
Mercedes
44% to win
Finished 1st
P2
LEC
Ferrari
22% to win
Finished 2nd
P3
VER
Red Bull Racing
21% to win
Finished 3rd
ANT won as favourite at 44%3 of 3 podium spots calledBiggest surprise: STR, 20th to 15th
Polymarket had ANT at 35%, the model 44%
Result vs prediction
Predicted is our order before the race. Actual is where they finished. PodiumPoints
| Predicted | Driver | Actual | vs prediction | |
|---|---|---|---|---|
| P1 | ANTMercedes | P1 | ✓ exact | |
| P2 | LECFerrari | P2 | ✓ exact | |
| P3 | VERRed Bull Racing | P3 | ✓ exact | |
| P4 | HAMFerrari | P4 | ✓ exact | |
| P5 | HADRed Bull Racing | P6 | 1 place lower | |
| P6 | RUSMercedes | P8 | 2 places lower | |
| P7 | PIAMcLaren | P7 | ✓ exact | |
| P8 | NORMcLaren | P5 | 3 places higher | |
| P9 | LAWRacing Bulls | P9 | ✓ exact | |
| P10 | HULAudi | P11 | 1 place lower | |
| P11 | GASAlpine | P10 | 1 place higher | |
| P12 | LINRacing Bulls | P12 | ✓ exact | |
| P13 | BEAHaas F1 Team | DNF | ||
| P14 | COLAlpine | DNF | ||
| P15 | BORAudi | P13 | 2 places higher | |
| P16 | ALOAston Martin | P16 | ✓ exact | |
| P17 | OCOHaas F1 Team | P14 | 3 places higher | |
| P18 | SAIWilliams | P18 | ✓ exact | |
| P19 | ALBWilliams | P17 | 2 places higher | |
| P20 | STRAston Martin | P15 | 5 places higher | |
| P21 | PERCadillac | P19 | 2 places higher | |
| P22 | BOTCadillac | DNF |
Chances
WinPodium
Full field
| Pred | Driver | Grid | Win | Podium | Points | DNF | |
|---|---|---|---|---|---|---|---|
| 1 | ANT | 1 | 44.4 | 88.4 | 94.7 | 5.5 | |
| 2 | LEC | 4+2 | 21.9 | 73.0 | 92.8 | 7.4 | |
| 3 | VER | 2-1 | 21.4 | 72.2 | 92.2 | 8.0 | |
| 4 | HAM | 5+1 | 5.2 | 27.3 | 91.5 | 8.5 | |
| 5 | HAD | 3-2 | 3.4 | 18.7 | 92.2 | 7.6 | |
| 6 | RUS | 6 | 1.6 | 9.4 | 90.4 | 9.2 | |
| 7 | PIA | 7 | 0.9 | 5.1 | 88.2 | 10.5 | |
| 8 | NOR | 8 | 0.6 | 3.0 | 84.2 | 11.9 | |
| 9 | LAW | 9 | 0.1 | 0.6 | 53.4 | 9.7 | |
| 10 | HUL | 11+1 | 0.1 | 0.7 | 49.0 | 13.3 | |
| 11 | GAS | 10-1 | 0.1 | 0.6 | 48.5 | 12.1 | |
| 12 | LIN | 14+2 | 0.0 | 0.5 | 39.8 | 17.5 | |
| 13 | BEA | 13 | 0.0 | 0.2 | 26.2 | 14.6 | |
| 14 | COL | 20+6 | 0.0 | 0.1 | 12.4 | 24.1 | |
| 15 | BOR | 21+6 | 0.0 | 0.1 | 9.7 | 23.7 | |
| 16 | ALO | 12-4 | 0.0 | 0.1 | 9.6 | 22.5 | |
| 17 | OCO | 15-2 | 0.0 | 0.0 | 8.4 | 17.0 | |
| 18 | SAI | 16-2 | 0.0 | 0.0 | 4.9 | 19.9 | |
| 19 | ALB | 17-2 | 0.0 | 0.0 | 4.5 | 19.1 | |
| 20 | STR | 22+2 | 0.0 | 0.0 | 3.2 | 42.5 | |
| 21 | PER | 18-3 | 0.0 | 0.0 | 2.5 | 27.1 | |
| 22 | BOT | 19-3 | 0.0 | 0.0 | 1.9 | 27.4 |
The predicted order sits close to the starting grid, and that is expected. This model has not been shown to beat "everyone finishes where they start". The percentages are what it adds: checked against past races, they come out about right on average.
Grid penalties are only included if they were entered by hand. They are never in the timing data this is built from.