F1 and the Data Gap: The Winner Is the One Who Dares to Say 'Not Enough Evidence'

On March 10, 2026, at the Bahrain International Circuit, the Mercedes W13 rolled out of the garage with its sidepods almost entirely gone. Within four hours, thousands of analyses flooded social media, each asserting one certainty: Brackley had found the key to the ground-effect era. The team had completed only a few dozen testing laps. There was no tyre-temperature data from a real race track. Nobody knew how the car would bounce once the fuel tank was full and the tyres were worn. The crowd had reached a verdict before the sensors had finished recording. Eight months later, the W13 closed the season with exactly one win. The gap between the speed of the story and the speed of the data, in that case, was the gap between a legend and a failure.

I retell this not to criticise a team. I retell it because it is a pattern repeated across a whole decade. The most expensive mistake in modern F1 analysis lies in this: people fill data gaps with a story that sounds plausible. And in a sport where every thousandth of a second is measured, that gap is never free.

I came to the racetrack in 2026, when telemetry was still a luxury and a chief engineer could hold nearly all of a race's data in one notebook. Thirty-eight years later, the car in the garage next to me transmits hundreds of signal channels per second: surface temperatures at all four tyres, gearbox oil pressure, steering angle, floor oscillation amplitude. By industry estimate, a team collects around 1.5 terabytes of data across a single Grand Prix weekend. Most of that volume is noise.

In 2026, working as a transfer market administrator in London, I spent three months screening 1,247 players from 15 European leagues to shortlist 38 worthwhile targets. The biggest lesson of that period came not from any number, but from having to write "insufficient evidence" on the remaining 1,209 rows. Football taught me that value lies in daring to discard. F1 taught me that value lies in daring to stay silent.

Two regulatory levers and a data problem

Over the past four years, the game in F1 has shifted through two regulatory levers, and both are data problems rather than driving problems.

The first lever is the cost cap. From 2026, each team could spend only 140 million dollars on racing operations; by 2026 the ceiling fell to 135 million. The cost cap turns every development decision into an investment decision. If you pour money into the wrong aerodynamic direction, you cannot compensate by pouring more money into another. You can only compensate by reading data better.

The second lever is the Aerodynamic Testing Restrictions system, known as ATR. The champion is allowed roughly 70 percent of baseline wind tunnel runs, while the last-placed team gets up to 115 percent. There is a paradox here that the media usually skips: this rule rewards correct inference and punishes trial and error. A team without a wind tunnel advantage can still win a title if every run it performs returns clean information.

McLaren is the clearest example. At the start of 2026, the Woking squad began the season near the back of the standings. By the Austrian Grand Prix that July, their upgrade package had turned the MCL60 into a permanent podium contender, and Lando Norris finished second at Silverstone just weeks later. That was the product of a chain of decisions built on correlated data, not of luck.

Aston Martin's 2026 season shows the flip side of the same story. Fernando Alonso climbed the podium six times in the first eight races. Then the pace vanished. When wind tunnel data no longer resonated with the track, the Silverstone team had no way of knowing it was slipping backwards until it had been overtaken. Correlation between simulation and reality is what holds value, not the raw volume of data.

Mercedes in 2026 and 2026 is another chapter of the same story. The W13 with its minimalist sidepods made the whole industry believe Brackley had found a shortcut. In reality, the team admitted across several press conferences that the correlation between wind tunnel data and the car's actual behaviour had broken down. The porpoising phenomenon made the car buck violently on the straights, and every one of their models underestimated the severity. When the model is wrong, every conclusion built on it is wrong too, however rigorous the process.

The Red Bull RB19, the car that won 21 of 22 races in 2026, proves the opposite. The Milton Keynes team built a system in which every new component had to answer a data question before it went on the car. The RB19's floor was so effective that rival teams took nearly a whole season to decode it. When you have clean data and a system that forbids inference beyond the evidence, you do not need to tell a good story.

Sensors do not measure everything

There is a technical detail few outside the industry know. Sensors do not measure everything. Core tyre temperature, the thing that decides grip, cannot be measured directly by a sensor mounted on the car; it has to be inferred from a thermal model. Tyre pressure can be measured, but the lateral force distribution across the tyre cannot. Engineers must interpolate between thousands of data points and hundreds of assumptions, and every assumption is a chance to be wrong. When a team says "we understand the car", it is saying its assumptions have not yet been challenged, not that they are correct.

Valuing people in a data market

As a transfer market administrator, I have watched many engineering deals priced on just a few lines in a performance report. When Adrian Newey left Red Bull to join Aston Martin, announced in September 2026, the market immediately assigned Aston Martin a leap forward in aerodynamics. But an outstanding individual only makes a difference when the system around them is clean enough to test their ideas. A person's value depends on the quality of the data infrastructure surrounding them. The transfer market is a contest in which whoever prices correctly wins.

What separates a champion team from the rest is not the volume of data it owns, but the discipline to refuse conclusions when the data is not yet sufficient. A good engineer in F1 must be able to say the hardest sentence: "I do not know yet." It is also the sentence the media almost never says.

The contrarian angle: correlation is not causation

There is a widespread belief that whichever team collects more data will win. That belief errs by confusing correlation with causation. Champion teams have more data not because they win, but because they have more resources to measure with. Measurement itself does not create speed. Speed comes from choosing the right variables to measure and having the courage to ignore the rest.

In 2026, when racetracks closed because of the pandemic, I sat in London and reviewed the entire dataset from the races without spectators. The empty stands exposed something the earlier models always got wrong: the home advantage of certain drivers vanished almost entirely, while the gap between two cars from the same team narrowed. When the external noise is switched off, what remains is real capability. At 60, I no longer believe in luck, only in the numbers that have not yet spoken.

The problem with F1 journalism lies here. I have written about this sport for five years, and I see clearly the pressure to reach a conclusion even without a basis. A test session, an interview, a newly launched upgrade package, and within a few hours there must be an angle. But a race lasts more than two hours, a season lasts nine months, and a technical cycle lasts several years. Those three time frames cannot be squeezed into the same news item.

F1 media operates in heat cycles. A new upgrade package creates a peak of attention for 48 hours, then fades. The real value of that package only shows after six to eight races, once tyre, aerodynamic and strategy data have accumulated a sufficient sample. Nobody holds attention that long. The result is that most public analysis reacts to noise, not to signal.

The 2026 power unit cycle will test all of this. The power split between the internal combustion engine and the electrical system is now almost balanced, fuel moves to a fully synthetic blend, and teams will fall into the old trap again: judging too early on the data of the first few laps. Every technical cycle in F1 imitates the data of the previous cycle, but nobody learns.

What to watch

Data never hurries, but people always rush. The winner of the 2026 season will not be the team with the most sensors, but the team that builds a process in which "insufficient evidence" counts as a valid answer. Watch how teams announce upgrades in the first half of 2026. The team that dares to say "we need more data" instead of promising a leap forward is the one on the right track.