F1 Data Does Not Tell the Truth on Its Own: The Cost of Analysis Built on a Void

In the summer of 2026, at the age of 48, I sat in an analysis room in Milan with motion data from 20 Serie A matches of the 2026-17 season. The xG figure at home in San Siro read 1.85, far above the 1.02 recorded away, yet the team's actual goal tally was identical. Many people would rush to conclude something about competitive psychology. I did not. Cross-checking the video, I found a sensor in the south-west corner lagging by 0.2 seconds, which recorded every goalkeeper-initiated build-up in the wrong position. My 14-page internal report recommended recalibrating the equipment, and the team won 5 of its last 8 matches to claim a Europa League place. I carried that lesson straight into Formula 1: before trusting a number, ask what instrument measured it.

Modern F1 is a data-generating machine. A race car carries hundreds of sensors, streaming more than a hundred telemetry channels to the team in real time, plus GPS position data, aerodynamic loads, tyre temperature and pressure, fuel consumption and degradation models. Every Grand Prix weekend, a team collects a volume of data that nobody could have imagined decades ago. I have followed F1 since 2026 and have not missed a single race, with more than 500 covered. What I have learned after three decades is this: volume of data is not the same as truth. A transmission can drop packets. Sampling frequency can be too low to catch a hard braking event. Sensor placement can produce a systematic error that repeats every lap. The problem is not that teams lack data, but that they trust data that has not been verified. In a regular season where the gap between teams is measured in thousandths of a second, that error is points.

Start with a simple calculation. At 300 km/h, an F1 car covers about 83 metres per second. If the position signal lags by 0.2 seconds, exactly the lag I once found at San Siro, then on the telemetry map the car sits nearly 17 metres away from where it really is. On a racing circuit, 17 metres is the distance between braking at the right point and braking late, between taking a corner cleanly and running wide onto the kerb. A tactical analysis built on uncalibrated data is nothing more than a guess dressed in scientific clothing.

F1 teams understand this better than anyone. They maintain an entire engineering department just to answer one question: can this number be trusted? Before every weekend, wind tunnel and CFD simulation data must be correlated against on-track data. When a model promises an upgrade worth 0.3 seconds per lap but the circuit returns only 0.1, the problem usually does not lie with the upgrade. It lies in the input assumptions. One wrong correlation coefficient, one incorrect boundary condition, and an entire technical conclusion collapses. With cost caps and severely restricted wind tunnel hours, a development direction derailed by dirty data does not merely waste time. It removes an opportunity the team cannot buy back.

The same happens with tyre degradation models. An algorithm may predict that the next lap will be 0.4 seconds slower, but if actual track temperature is 6 degrees Celsius above the assumption, the prediction is meaningless before the car even rolls. Engineers at the leading teams never ask what the model says and then simply follow it. They ask what the model says, and under which conditions it says it. Data only tells part of the story; the rest lies in knowing how to listen.

There is another layer of data that spreadsheets never display: radio. I spent years inside the paddock and learned to hear the rhythm of a race engineer's voice. When an engineer says hold the plan in a steady tone, that is confidence grounded in data. When the same sentence is delivered after a short pause, that is a sign they are reconsidering. Those pauses appear in no telemetry file. They often forecast a pit wall decision before that decision takes shape.

I drew a lesson about translating numbers into spatial images after the Germany versus South Korea match at the 2026 World Cup. With Germany's defensive line pushing an average of 68 metres high and 17 failed presses, I did not write out what the number was. I wrote that the back line had unzipped like a broken zipper reaching all the way to the valve box. That phrasing preserved what a spreadsheet drops: the sensation of a structure slowly coming apart. In F1, an aerodynamically unbalanced car says nothing through a downforce figure. But when you see it slide lightly at the entry of a fast corner for three consecutive laps, you are seeing the data with your eyes.

The most common blind spot in F1 analysis is that people always hunt for the culprit downstream. A collapsed strategy gets blamed on the pit wall. A poor start gets blamed on the driver. A slow lap gets blamed on the tyres. Very few look upstream, where data is created, transmitted and calibrated. Every collapse has a precondition; few simply choose to look beforehand.

The greatest temptation is to worship a single metric. I have seen analysts build an entire conclusion about a team's strength on one xG figure, or about a car's speed on one downforce number, without placing it alongside tyre wear, fuel load, wind direction and track state. A single metric always looks convincing because it is tidy. But that tidiness is usually a sign that context has been stripped away.

Nor do I use memories of older seasons to compare with the present. Memory is a poor data source, easily distorted by emotion and past glory. Whenever I am forced to compare eras, I anchor it to a verifiable fact: a record, a date, a specific head-to-head history. Without that anchor, every comparison is just a sigh.

Every tracking number belongs on the operating table, not on the altar.

When a team fails in the next race, look back at the three rounds before it rather than only the final result. Listen to the rhythm of the radio, note when the data was recorded, and ask yourself under which conditions that number was measured. A collapse on track rarely begins on the lap where it becomes visible.