Three Tyre Compounds and an Aborted Formation Lap: What Sepang's Data Says About Wet-Weather Strategy

The formation lap at Sepang ran longer than planned, and the culprit was not the drivers. On a surface still holding water, a group of drivers could not maintain the minimum speed the regulations require when tyres have not reached their thermal window. Race control suspended the start and ordered an extra formation lap. Across 44 years of covering this sport, I can count on one hand the number of times a start has been delayed for a purely thermal reason.

I was in London, in front of four screens, rewinding the formation-lap footage three times before writing my first note. That habit dates to 2026, when I sat at the Motoring News desk: do not trust the first impression, count again. The first impression of an afternoon like that is always chaos. Data does not know the word.

The three-compound problem

On the grid, nine drivers chose softs. Four drivers chose mediums. The rest chose intermediates. Three compounds shared one starting grid on a surface described as drying. This is the classic crossover problem: the moment the track shifts from wet enough for intermediates to dry enough for slicks decides who is right and who is wrong, and nobody knows the answer before the lights go out.

Intermediates are grooved tyres that sit between full wets and slicks in both water-clearing ability and grip. Slicks have no tread, offering maximum dry grip but almost no water clearance. When a track is drying, these two tyre types sit only a narrow band apart in performance, and that narrow band is what produces the strategic split.

Splitting across three compounds is not impulsive behaviour. It is standard behaviour when the crossover point between wet and dry tyres is unknown. Under uncertainty, hedging is the theoretically optimal play in game-theory terms. The problem is that hedging reduces heavy losses but also reduces heavy gains. And in a race with a suspended start, the reward for guessing right is compressed.

I built a small table comparing the three choices across four criteria: standing-start grip, warm-up speed, water clearance, and tyre life in damp conditions. Intermediates won the first two whenever the surface still held water. Softs won only when the track was dry enough for slicks to bite from the very first lap. What Sepang showed leaned toward intermediates.

The thermal window and the soft-tyre trap

The clearest technical signal of that afternoon was not on the grid but on the formation lap. Several drivers could not hold speed, and one driver pitted at the end of lap one for a grip-related reason. This is a tyre warm-up issue — the thermal window — not a single-component fault. Slicks on a cold, wet surface slide before they bite, and when a car slides at low speed, the driver loses control in the least expected place: the formation lap.

One thing needs stating plainly: the thermal window is the temperature range in which rubber delivers its designed grip. Below it, the tyre is hard as plastic and slides. Above it, the compound grains and loses grip. In cold, wet conditions, both the track and the air drag the tyre below the threshold, and that is the technical reason a car fast enough on paper cannot hold minimum speed on a formation lap.

The conclusion is fairly clear. Grip at the start was closer to wet than to drying. If that holds, intermediates were the sensible baseline tyre and softs were a gamble. And when a soft-tyre gamble fails, the cheapest way to correct it is to pit under the Safety Car, because pit loss under yellow flags is far lower than under green.

Pit loss under yellow flags

To understand why an early stop makes sense, you have to understand pit loss. At racing speed, a pit stop costs roughly twenty seconds relative to staying out, depending on pit-lane length. Under yellow flags, with the field bunched and running slowly, that loss shrinks to a fraction, sometimes just a few seconds. In other words, a yellow flag is the cheapest moment to fix a wrong tyre call.

That is exactly what happened. One driver pitted at the end of lap one, behind the safety car, to switch to intermediates. A lap later, four more drivers did the same. This is a follow-on wave, and it shows intermediates becoming the consensus choice. When four cars change tyres within a single lap, it is no longer an individual decision; it is a market signal.

I still tell younger colleagues: data is never in a hurry, but people always are. At Sepang, the hurry of the track itself — drying faster or slower than forecast — forced half the grid to correct within the first two laps.

When teammates choose alike

One detail is easy to miss when reading the tyre list: drivers on the same team mostly chose alike. At Ferrari, both drivers started on softs. At McLaren, both chose mediums. At Aston Martin, both chose mediums. At the team of Bortoleto and Hulkenberg, both chose softs.

That alignment is not random. If each driver decided alone, the probability of two teammates picking the same compound out of three would be far lower than what was observed. When both cars of a team head in one direction, the decision was most likely made at team level, using radar data and rain forecasts, rather than by each driver.

That changes how to read the chaos. If teams met and set the compound per car before going out, the story is not eighteen bewildered drivers but nine teams making nine different calls against one variable: how fast the surface dries. That is an organised game, not an uncontrolled crowd.

The one notable outlier is the driver who pitted at the end of lap one. He started on softs, then switched to intermediates after a single lap. That is the most valuable behavioural data point of the afternoon: a driver willing to admit a mistake immediately instead of enduring it in hope of another yellow flag. In such conditions, a fast reaction is usually cheaper than a correct one.

A grid of two generations

Looking at the driver list at Sepang, one structural feature stands out: a group of young drivers newly arrived in the sport, alongside veterans who have lived through several technical cycles. Bearman, Bortoleto, Colapinto, Gasly, Hulkenberg, Stroll, Alonso, Hamilton, Leclerc, Norris, Piastri — a span running from the rookie generation to those who have raced more than a decade.

That generational mix has strategic meaning. Rookies tend to follow the team line more often, because they lack enough personal data to challenge an engineer's decision. Veterans tend to lean on their own sensory experience. In a race where the surface changes lap by lap, the gap between the two groups can produce different tyre choices even within one team. At Sepang that barely happened, and the fact that it barely happened is itself information.

What the data does not say

One principle I have kept throughout my career: knowing what you do not know matters as much as knowing what you do. For the Sepang race, the available data leaves three large gaps.

Without finishing positions and gaps between cars, I cannot calculate the net effect of the tyre decision. A driver can make the right stop and still lose position because the yellow flag ends too soon. Without gap data, any conclusion about strategic effectiveness is structural, not conclusive.

Without each car's pit-stop time, I cannot assess execution quality. Half a second in the pit lane can reverse the order on track.

And without the official classification of the start sequence, I cannot settle the effect of the pit wave. The data I have describes a situation where one group of cars waited in the pit lane while the rest took a standing start. If that was a restart, the advantage of the early-stopping cars is compressed substantially.

The risk sits in the source, not the track

In the risk table of any analysis, I always keep a row for source quality. Here, that row is high. The driver list given does not fully match the official calendar, and the event name is mixed between two different countries. That kind of error caps the reliability of everything downstream.

This point needs stating to set the right level of certainty for each conclusion. The tyre-strategy part is solid, because it rests on the physics of rubber and water. The event-identification part is weak, because it depends on administrative data. Blending the two without distinction is a methodological error.

Brentford does not read the future, it simply reads data more carefully than others — and a careful reader of data must know where their own limits lie.

The gap between expectation and data

A wet start always creates big expectations: plenty of overtakes, plenty of surprises, plenty of big names falling. Those expectations are usually right emotionally and wrong probabilistically. Most wet races do not produce significantly more overtakes than dry ones; they merely make the few overtakes more memorable, and a spectator's memory does not distinguish between the two.

There is a notable psychological effect here: under uncertain conditions, viewers attribute every outcome to individual skill, because skill is the most easily imagined variable. Data attributes most of the outcome to timing — when the rain came, when the yellow flag flew, when the tyre fell out of its window. The distance between those two attributions is the distance between storytelling and analysis.

Chaos or management?

The story the media told after the race was an emotional one: a dramatic show, a crazy start, fans treated to a spectacle. I do not deny its appeal. But set beside the actual decisions, the picture looks different.

The suspended start was not a random accident but a procedure already written into the sporting regulations. When the grid cannot move safely, race control has the right to abort the formation lap and add an extra one. That is a discretionary call by the stewards, and here it worked exactly as designed. A procedure functioning, not a procedure collapsing.

Pitting under the Safety Car is a value-maximising move, not a reckless one. It turns a wrong tyre gamble into an almost free correction. Anyone who calls that panic has never calculated pit loss.

And here is the counter-intuitive point: correlation is not causation. We see rain, a yellow flag, cars pitting, and then assume some driver handled it poorly. But the variables that dominated the outcome were the timing of the rain and the timing of the yellow flag — both beyond any driver's control. Attributing the result to individual character in a race like that is misreading the equation.

At 60, I no longer believe in luck, only in numbers that have not yet spoken — and in this case, those numbers still sit inside the data table the original article has not published.

What to watch next race weekend

Three signals will decide how this race is read back.

The official classification of the start sequence tops the list. A suspended start followed by a standing restart has a very different effect from a start under yellow flags. If race control confirms a restart, the advantage of the early-stopping cars is compressed, and the value of an early tyre change falls.

Next is the table of positions gained and lost on the opening lap. That is the genuinely valuable data, and it has not been published. With that table, the question of which tyre was right gets a quantitative answer instead of an emotional one.

The remaining signal is track temperature at the start. A tyre's thermal window depends directly on track temperature, and this is a measurable variable, not a guess. Another wet race will allow a comparison: whether teams repeat the three-compound split, or have learned from Sepang.

Every cycle in football imitates the data of the previous cycle, yet nobody learns. The Sepang start was a small, cheap lesson, and therefore easy to ignore. Another wet race will show whether teams truly re-read their own data, or once again let the track decide for them.