The race where the medium tyre died – and ten drivers ran ten strategies
Pirelli's plan for Zandvoort was one stop, medium to hard around lap 30. Then it rained before the start, Verstappen crashed on the opening lap, the race was red-flagged and the medium degraded more than twice as fast as the year before. The top ten finishers ran ten different strategies. Norris won on what was left when the plans ran out: race pace.
Before the race Pirelli rated medium → hard with one stop around laps 27–33 as fastest, with soft → hard a second behind over the distance; two stops could be quick in clean air, but Zandvoort is hard to overtake on and track position therefore valuable, per F1.com. Then came the rain before the start. Pirelli noted afterwards, per press.pirelli.com, that the track was slipperier than expected and degradation considerably higher than forecast.
The nomination was the hardest on the calendar: C1 as hard, C2 as medium, C3 as soft. Our data says how much: the medium (C2) degraded at 0.075 s/lap against 0.033 the year before – more than double – and the hard 0.036 against 0.028. The soft (C3) actually degraded less than in 2025 (0.039 against 0.053). The tyre meant to carry the first stint became the worst of the three, and that is where the one-stop died. The race ended with 68 tyre changes, the red flag's included, and ten different strategies in the top ten.
Verstappen crashed on lap 1 and the race was red-flagged on lap 2. Everyone changed tyres for free. Norris, Antonelli, Russell and Piastri had started on mediums; after the stoppage four different choices stood on the grid: Norris on softs, Piastri on hards, Mercedes keeping mediums on both cars.
Antonelli passed Norris within two laps of the restart. Then the thing McLaren had hoped for did not happen: Norris's softs gave nothing. The gap grew from 0.9 s on lap 5 to 2.3 s on lap 20, the soft lost 0.10 s a lap and was no quicker than Antonelli's medium (77.19 against 77.09 s fuel-corrected). The red-flag call was, in practice, neutral.
Both stopped on lap 21 for hards. Norris crept to within 0.7 s – on lap 30 and again on lap 39 – but could not pass; Zandvoort's overtaking threshold of 1.08 s/lap is among the calendar's highest. Antonelli pitted from the lead on lap 40 for a new set of hards. Norris ran seven more laps on his old ones and came out behind Antonelli, 1.7 s on lap 50. Mercedes' early stop had worked.
But seven laps' fresher rubber on a high-degradation track is an advantage that grows every lap. Norris closed the gap in four laps and passed on track on lap 54, just before the VSC. His second hard stint, 76.78 s fuel-corrected, was the fastest hard stint of the race. That is where the race was won – not in the pits.
When the VSC came on laps 54–56 Mercedes brought Antonelli in for softs; Russell stayed out. Antonelli rejoined behind his team-mate, Mercedes swapped them so the fresher softs could attack, and Antonelli finished second, 11.5 s behind Norris. He called the decision risky himself, per F1.com.
The simulator says it made no difference. Click a scenario to load it in Simulate below.
| Scenario | Actual | Simulated | Race time | Comment |
|---|---|---|---|---|
| ANT without the soft stop under the VSC | P2 | P2 | +0.5 s | Same place, same time. |
| ANT stays out to lap 47 like Norris | P2 | P2 | −1.7 s | The early stop gained nothing net. |
| NOR covers Antonelli on lap 40 | P1 | P1 | +6 s | Still wins – on pace. |
| RUS softs under the VSC like Antonelli | P3 | P3 | +0.8 s | Neutral – Mercedes had nothing to gain from the risk. |
| HAM without the soft stop on lap 55 | P4 | P5 | +7 s | Ferrari's soft stop was right. |
Hamilton stretched his softs to lap 25 and then had thirty laps on hards; Leclerc ran mediums in the middle of the race and could attack – he passed Piastri, among others. The problem came when the two strategies met on the same piece of track: Hamilton had the fresher tyres but was left behind Leclerc, asked whether he was being used as a guinea pig, and Fred Vasseur admitted afterwards that the situation cost time, per F1.com. Fourth and fifth at the flag, Hamilton 0.9 s behind Russell in the fight for the podium. The soft stop under the VSC was right, though: without it Hamilton would have been fifth.
Tsunoda ran tenth from the restart, stopped on lap 18 for hards and went back to softs on lap 44 – an offensive final stint rather than a defence on old hards. The thinking is right for a midfield car at Zandvoort: if you have to overtake anyway you need a tyre offset, track position is not enough against a car as quick as yours. This time it gave nothing – eleventh, lapped. The simulator suggests hards to the flag would have been better, but it does not handle blue flags, so we leave that number aside.
| Situation | Strategy |
|---|---|
| Before the race | Medium → hard, one stop |
| After the rain | The one-stop died with the medium |
| After the red flag | Soft or medium → hard, with room for more stops |
| In traffic | Fresh softs/mediums became very valuable |
| In the lead | Track position + hard |
| Late VSC | A discounted soft stop if position allowed |
The conclusion is not a tyre sequence. The smartest strategy was not to have one: to accept more stops once the degradation showed itself, to use the hard where it bought track position and to have the race pace not to get stuck. McLaren did that best – but it was Norris's second hard stint, not the softs at the red flag, that decided it. That is exactly the kind of race our coverage exists to find: the difference between the strategy that was fastest on Sunday morning and the one that became fastest when the race changed the terms.
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; race pace = 25th percentile per stint. The simulator runs everyone else on their actual lap times and models the changed driver; it cannot recreate the red flag. Sources: F1.com (Pirelli's pre-race assessment, driver and Vasseur comments), press.pirelli.com (Pirelli's post-race note), FIA timing.
Simulate
Change one driver's pit stops and let everyone else run exactly as they did. The altered driver is modelled with his own pace per compound from the race, degradation, fuel, traffic and the VSC discount – cars that get stuck behind him lose time.