Lap 12 in Melbourne – where the season's strategy began

Before the opener Formula 1 had changed both chassis and power-unit regulations: smaller, lighter cars, different aerodynamics, an entirely different power delivery from the new hybrid units. Pirelli's tyres were new too – the 18-inch rim stayed, but the tyres became narrower, with a smaller contact patch and smaller diameter. Testing in Barcelona and Bahrain had given data, but Pirelli said explicitly that Melbourne would be the first time anyone saw the cars' full performance and the tyres in a competitive environment, per Pirelli.

The questions queued up: how much load do the cars generate, how quickly do the rears overheat under the new power delivery, how much front graining, can the soft be raced, is one stop or two faster – and how much easier is overtaking with smaller cars and Overtake Mode? The last question governs all the others: if passing is hard, track position is worth more and one stop is more attractive; if it is easy, you can take another stop and run on fresher rubber.

Pirelli brought the three softest compounds – C3 as hard, C4 as medium, C5 as soft – reasoning that Albert Park is a low-severity circuit with smooth asphalt, where degradation has historically been about graining rather than heat. Friday's long runs showed graining on all three, mainly at the front, but low and manageable degradation, and steps of 0.3–0.4 s/lap between hard and medium and around 0.5 between medium and soft – enough for all three to play a role, per F1.com.

By Saturday evening, after Mercedes had locked out the front row with Russell ahead of Antonelli, Pirelli had a favourite: medium → hard, one stop, window laps 20–26. Soft → hard with a stop on laps 15–21 as the aggressive alternative. And since a stop cost around 21 seconds and Albert Park has historically seen a Safety Car in three races out of four, two stops could not be ruled out: M → H → M, or the truly aggressive S → M → S. Three schools faced each other before the start: the conservative, the aggressive – and the opportunistic, which keeps its flexibility and waits for the neutralisation. The last one won.

The start went wrong for Mercedes – too little battery energy after the formation lap – and right for Ferrari. Leclerc took the lead from P4, Russell sat 0.8 seconds behind lap after lap, Hamilton went from P7 to P3. Fifteen drivers had started on mediums, three on hards (Verstappen from P20, Colapinto), two on softs (Sainz, Alonso). On lap 11 it read Leclerc, Russell +1.0, Hamilton +2.1, Antonelli +2.9. The same lap Hadjar's Red Bull stopped. Virtual Safety Car.

Mercedes brought both cars in at once: Russell and Antonelli from medium to hard, eight laps before Pirelli's window. Pit loss under green at Albert Park is 23.3 seconds; under this VSC we measured 17.0. Six seconds cheaper – F1.com puts it at around ten, per F1.com. The price was that the hard would have to last 46 laps, and Russell said afterwards that Mercedes did not know at that moment whether one stop would still work. It was not we know this is right; it was the stop is cheap, we take it and try to make the tyres last.

Ferrari did the opposite. Leclerc and Hamilton stayed out. Fred Vasseur explained afterwards that their model said to extend the first stint – that tyre life and a later offset were worth more than reacting to Mercedes – and Leclerc was leading, so track position was not given up voluntarily, per F1.com. Both arguments were defensible on lap 12. Only one of them was right.

Once Mercedes had the hards on, they could watch how they degraded. The answer was: hardly at all. Russell's 46 laps on the C3 lost 0.038 s/lap, Antonelli's 0.023. But what Ferrari did not know either was that the medium degraded even less – our slope for every medium stint is negative, the track rubbered in faster than the rubber wore away, and Hamilton ran 28 laps on it without losing pace. So there was no reason to stop for the tyres' sake. There was every reason for the discount's sake.

More VSCs came – laps 16–20, a short one on lap 22, laps 32–34 – but Ferrari used none of them. Leclerc stopped from the lead on lap 25 under green, seven seconds ahead of Russell, and rejoined fifteen behind. Hamilton stopped on lap 28. From there the race pace was identical: Russell 84.13 on the hard, Leclerc 84.17. The whole gap at the flag – 15.5 seconds – is the pit decisions. Vasseur said afterwards, deliberately exaggerating, that the tyres could almost have done "300 laps", per F1.com. Click a scenario to load it in Simulate below.

Saturday's forecast – M → H, one stop – became the winning strategy. But the race showed the hard was stronger than Saturday's models: the final stint could be extended well beyond the forecast thanks to the low degradation, Pirelli noted afterwards, per Pirelli. The model was right about which strategy and conservative about how long – exactly what you should expect in the first race under a new set of regulations. Nine drivers ran one stop, seven two, two three.

More interesting than who won is what everyone knew afterwards. Before the race there was a fear that the new cars' power delivery and the narrower tyres would produce unpredictable degradation and temperature problems; Pirelli had pointed in particular to the balance between front and rear axle as something the teams had to learn. After the race everyone knew: the hard was very robust, front graining existed but was manageable, the soft could be raced, one stop was realistic – and a neutralisation could be worth more than a tyre offset. Above all: the usable stint length of the 2026 tyres was considerably greater than the conservative models assumed.

Pirelli were careful not to generalise – Melbourne is kind to tyres, and more demanding circuits and the cars' development would change the picture through the year. They were right about that: at Spielberg, ten races later, the same C4 degraded at 0.08 s/lap. But the principle from lap 12 held all season. Whoever took the neutralisation won in Melbourne, in Madrid and at Spa. Whoever waited for a tyre offset that never came lost.

How we did it: lap data from FastF1 (F1 Live Timing). Degradation = within-stint slope on green laps, fuel-corrected at 0.035 s/lap. Pit loss under VSC measured from the race's stops. The simulator runs everyone else on their actual lap times and models the changed driver on his own pace per compound from the race. Sources: press.pirelli.com (pre-race assessment, Friday analysis and post-race note), F1.com (strategy preview, Russell's, Vasseur's and McLaren's comments, post-race analysis), FIA timing.