ROUND_17 · MARINA BAY

Singapore GP 2026did we call it right?

Before the race we ran this GP 400 times through our engine, starting from the real qualifying grid, and published the odds. The race has now been run: here is the contrast, misses included.

Leer en español

VERDICT · RACE RUN

What we saidand what happened

Verstappen won from pole, the model's favourite, and the podium went to the three drivers the simulation had in front

The model had Max Verstappen as the favourite on 25.5% of 400 simulations, ahead of the rest of the grid. He won.

1.77

Avg. position error

across 18 finishers

0.91

Order correlation

1 = exact order

7/8

Front of the race

of the predicted top

8/10

Points scorers

called right

Actual	Driver	Grid	Predicted	Diff	Win
1	Max VerstappenRed BullStarted from pole and was the simulation favourite at 25.5%. Won by 7.3 seconds.	1	3.70	-2.70	25.5%
2	Charles LeclercFerrariThe model's second favourite (20.8%) and the best average position on the grid (P3.4). Started second and finished second.	2	3.40	-1.40	20.8%
3	Kimi AntonelliMercedesStarted fourth with a 53.5% podium chance. With Verstappen and Leclerc, they were the three best average positions in the simulation, and they are the three on the podium.	4	3.50	-0.50	17.3%
4	Lando NorrisMcLaren	5	4.40	-0.40	10.8%
5	Oscar PiastriMcLaren	6	4.50	+0.50	8.5%
6	Lewis HamiltonFerrariStarted third and the model gave him a 51.5% podium chance. He lost three places.	3	4.20	+1.80	17.3%
7	Liam LawsonRacing Bulls	8	9.70	-2.70	—
8	Nico HulkenbergAudiThe model gave him an 18.5% chance of points and an average of P13.9. He started tenth and finished eighth.	10	13.90	-5.90	—
9	George RussellMercedesThe preview had him 21st on the grid; after late changes he started 19th. The model gave him a 96.3% chance of points and an average of P7.6: he gained ten places and finished ninth.	21	7.60	+1.40	—
10	Isack HadjarRed BullStarted last. The model gave him a 30% chance of points and he made it, taking the final point.	22	12.40	-2.40	—
11	Pierre GaslyAlpineStarted seventh and the model gave him a 62.5% chance of points. He finished eleventh, a lap down.	7	10.50	+0.50	—
12	Oliver BearmanHaas	11	13.30	-1.30	—
13	Esteban OconHaas	12	14.30	-1.30	—
14	Carlos SainzWilliams	13	13.50	+0.50	—
15	Arvid LindbladRacing BullsStarted ninth with a 48.8% chance of points. He lost six places.	9	11.60	+3.40	—
16	Fernando AlonsoAston MartinThe preview had him 16th on the grid and he ended up starting 20th.	16	14.60	+1.40	—
17	Sergio PerezCadillac	19	18.80	-1.80	—
18	Valtteri BottasCadillacThe official classification has him 19th, behind Albon. Here he is 18th because Albon counts as a retirement.	20	19.90	-1.90	—
DNF	Alexander AlbonWilliamsRetired three laps from the end. The official classification has him 18th after completing 59 of the 62 laps, but here he counts as a retirement: he did not finish the race.	17	17.20	—	—
DNF	Lance StrollAston MartinRetired after completing 36 laps. The preview had him 18th on the grid and he started 21st.	18	19.40	—	—
DNF	Franco ColapintoAlpineRetired after completing 29 laps.	15	14.90	—	—
DNF	Gabriel BortoletoAudiRetired after completing 7 laps.	14	16.50	—	—

Where the model got it wrong

Nico Hulkenberg predicted around P13.90, finished 8th (5.90 places off).
Arvid Lindblad predicted around P11.60, finished 15th (3.40 places off).
Max Verstappen predicted around P3.70, finished 1st (2.70 places off).
Liam Lawson predicted around P9.70, finished 7th (2.70 places off).
Did not see Nico Hulkenberg, Isack Hadjar scoring.

The pattern is clear: the model orders the front of the race well and underrates the midfield, because each car's performance is set per season and does not capture a given weekend's form. That is the biggest known limitation, and the next one to attack.

Official result: Formula1.com. The probabilities were published BEFORE the race and have not been touched since.

The preview, exactly as published

Favourite

Max Verstappen 25.5%

Wins 25.5% of the simulations and finishes on the podium in 61%. Starts P1.

Driver by driver

Grid	Driver	Team	Win	Podium	Points	DNF	Avg. finish
P1	Max Verstappen	Red Bull	25.5%	61%	95.5%	4.5%	P3.7
P2	Charles Leclerc	Ferrari	20.8%	62%	97.5%	2.5%	P3.4
P3	Lewis Hamilton	Ferrari	17.3%	51.5%	94.5%	5.5%	P4.2
P4	Kimi Antonelli	Mercedes	17.3%	53.5%	99%	1%	P3.5
P5	Lando Norris	McLaren	10.8%	38.8%	96.5%	3.5%	P4.4
P6	Oscar Piastri	McLaren	8.5%	33.3%	98.3%	1.8%	P4.5
P7	Pierre Gasly	Alpine	0%	0%	62.5%	5.8%	P10.5
P8	Liam Lawson	Racing Bulls	0%	0%	75.5%	4%	P9.7
P9	Arvid Lindblad	Racing Bulls	0%	0%	48.8%	4.5%	P11.6
P10	Nico Hulkenberg	Audi	0%	0%	18.5%	5.3%	P13.9
P11	Oliver Bearman	Haas	0%	0%	22.8%	4.5%	P13.3
P12	Esteban Ocon	Haas	0%	0%	13.5%	4.3%	P14.3
P13	Carlos Sainz	Williams	0%	0%	19%	3.5%	P13.5
P14	Gabriel Bortoleto	Audi	0%	0%	5.3%	4.8%	P16.5
P15	Franco Colapinto	Alpine	0%	0%	9.5%	2.5%	P14.9
P16	Fernando Alonso	Aston Martin	0%	0%	12.3%	3.3%	P14.6
P17	Alexander Albon	Williams	0%	0%	3.3%	5.8%	P17.2
P18	Lance Stroll	Aston Martin	0%	0%	0.3%	3.3%	P19.4
P19	Sergio Perez	Cadillac	0%	0%	1.3%	4.3%	P18.8
P20	Valtteri Bottas	Cadillac	0%	0%	0.3%	3.5%	P19.9
P21	George Russell	Mercedes	0%	0%	96.3%	3%	P7.6
P22	Isack Hadjar	Red Bull	0%	0%	30%	2.5%	P12.4

Opinion · our reading

Verstappen starts on pole and wins one race in four, and the model sees Russell climbing to seventh from the back

The starting column is the official GRID, with the penalty applied to the qualifying order. George Russell qualified sixth and starts twenty-first for fitting his fifth engine of the year, and everyone who was behind him moves up a place. Isack Hadjar set no time and starts last.

Max Verstappen has pole and the model gives him a 25.5% chance of winning. Charles Leclerc, starting second, wins 20.8% and has the best average finish (P3.4), and Kimi Antonelli, fourth on the grid, wins as often as Lewis Hamilton, who starts third: 17.3% each. All the wins go to the top six on the grid. Marina Bay is one of the circuits where starting position counts most in our engine, and it shows: from Pierre Gasly back, nobody wins a single time in 400 races.

What we will be watching most on Monday is Russell. The model has him climbing from twenty-first to an average of P7.6, scoring in 96% of races, though he never reaches the podium. That is a lot more than history suggests: since 2008, of 64 starts from the back in Singapore only six finished in the points, and the best front-running car to start last, Vettel's Ferrari in 2016, finished fifth. This is where we think the model may be generous to the Mercedes, because beyond the weight of grid position it does not know how hard it is to pass between walls. What works in his favour is that Singapore is the circuit on the calendar with the most crashes, and every safety car bunches the field up again.

How the model works

The engine accounts for how much starting at the front is worth at each circuit, and nowhere is it worth as much as at Monaco. Since the 2026 Azerbaijan GP that weight is calibrated track by track against the real races from 2014 to 2025; earlier previews ran on a simpler formula that derived it from how easy it is to overtake on each layout. Which is why Max Verstappen comes out favourite at 25.5% without having the fastest car of the season.

A favourite who only wins one race in four is not a broken model: that is what a real Grand Prix looks like. A retirement, a safety car at the wrong moment or a slow stop flips the result, and the engine simulates all of them. That is why 6 drivers have a genuine shot rather than a single name circled in red.

The most interesting column is the average finish: George Russell starts P21 and averages P7.6, the biggest gain on the grid: the driver the model thinks is most underrated by where he qualified.

The numbers come from 400 full race simulations (62 laps) with car and driver ratings updated to this point of the season. This is not betting advice and not an official prediction: it is what our model said. The odds were published before the race and have not been touched since: the contrast with the real result is above.

Your turn

Predict the GP →

Order your own top 10 and see how close you get.

Rewrite history

What If? →

Put any driver in any car, any season, and run it.