Race Winner Prediction

6.003 km lap length20 cornersMedium overtaking difficulty

Race date September 26, 2026 · generated Sep 26, 2026, 3:45 PM UTC · rain scenario weighted at 0%

Predicted vs. Actual

Predicted winner

RUS (34.5%)

Actual winner

RUS (correct)

Podium hits

1 / 3

Brier score (win)

0.0226

The model gave the actual winner (RUS) a 34.5% chance beforehand. Predicted live, before the race. See the track record page for the model's accuracy across every graded race.

Our model: 0.0226 · Grid-based baseline (always pick the polesitter): 0.0000 (baseline beats the model this time — lower is better)

Where the podium call missed

HAM (predicted podium) finished P6 instead.
LEC (predicted podium) finished P4 instead.
VER finished P2 — the model projected around P7.
HAD finished P3 — the model projected around P7.

Predicted vs. actual finishing position, every driver

Green, right — finished better than the model expected (underrated). Red, left — finished worse (overrated). Sorted by size of the miss.

Win Probability — Top 10

Error bars show ± 1 standard deviation of win probability across batches of the Monte Carlo simulation — how much this estimate would wobble on a re-run, not a formal confidence interval.

Full Prediction Table

Click a row to see what drove that driver's score.

Pos	Driver	Team	Win %	Podium %	Points %	Exp. Pos	Actual
1	▸RUS	Mercedes	34.5% ± 0.7%	68.8%	99%	2.99	1
2	▸HAM	Ferrari	14.4% ± 0.4%	42.6%	94.7%	4.73	6
3	▸LEC	Ferrari	11.9% ± 0.4%	39.1%	94.9%	4.89	4
4	▸NOR	McLaren	11.8% ± 0.5%	37.8%	93.6%	5.06	19
5	▸ANT	Mercedes	10.9% ± 0.5%	34.6%	91.5%	5.41	5
6	▸PIA	McLaren	5.2% ± 0.3%	21.9%	87.2%	6.46	14
7	▸VER	Red Bull	5% ± 0.3%	20.5%	82.9%	6.9	2
8	▸HAD	Red Bull	3.1% ± 0.2%	15.2%	80.6%	7.38	3
9	▸LAW	RB F1 Team	1.2% ± 0.1%	6.4%	56.4%	9.88	12
10	▸GAS	Alpine F1 Team	1.1% ± 0.1%	6.3%	60%	9.55	18
11	▸COL	Alpine F1 Team	0.3% ± 0.1%	2.3%	37.6%	11.74	17
12	▸BEA	Haas F1 Team	0.2% ± 0.1%	1.6%	30.3%	12.57	9
13	▸LIN	RB F1 Team	0.2% ± 0.1%	1.8%	33.9%	12.12	7
14	▸SAI	Williams	0.1%	0.5%	17%	14.22	10
15	▸OCO	Haas F1 Team	0%	0.1%	4.8%	16.93	8
16	▸ALB	Williams	0%	0.2%	9.5%	15.73	20
17	▸BOR	Audi	0%	0.3%	13.9%	14.65	13
18	▸HUL	Audi	0%	0.1%	8.9%	15.6	11
19	▸ALO	Aston Martin	0%	0%	2.2%	17.79	21
20	▸PER	Cadillac F1 Team	0%	0%	0.6%	19.23	15
21	▸STR	Aston Martin	0%	0%	0.3%	19.55	22
22	▸BOT	Cadillac F1 Team	0%	0%	0.3%	19.64	16

Methodology

Each driver's simulated race pace is sampled from a distribution weighted by qualifying pace percentile, top-speed ranking, season championship standings, historical finishing position at this circuit, tire degradation, and pit-stop delta. The simulation runs 100,000 times, with 0% of runs treated as a wet-race scenario (estimated from this circuit's own rainfall history).

This circuit's weighting — Azerbaijan Grand Prix: overtaking difficulty Medium, downforce dependency Low, tire severity Medium. Higher overtaking difficulty shifts weight toward grid position; lower downforce dependency shifts weight toward straight-line top speed; higher tire severity shifts weight toward tire-degradation management. See circuit_profiles.py for every circuit's numbers and how they translate into weights.