Empty Cells, Full Format: The Gap Between 'No Signal' and 'No Risk' in F1 Analytics

A file opened on a Tuesday morning in the London data office. Twenty-two pages. Nine dimensions. Under every dimension a table, in every table columns, in every column the same sentence repeated: "N/A — insufficient information." The technical section had an empty box where an upgrade should have been named. The strategy section carried a formula built to calculate an undercut, but no circuit, no compound, no lap number. The risk matrix printed six categories and the same single word against all six. The only reliable figure in the entire document was the page number.

A qualifying sheet never behaves like this. Every sector carries a colour, every driver a lap time to three decimal places. If the sector times are populated and the lap time is blank, the fault sits in the timing system, not the driver. That Tuesday I was looking at exactly that failure, with the timing system replaced by an analysis pipeline. The format was complete. The payload was zero.

This piece is the audit of that zero — and of the way an empty report can be made to read like 'all clear'.

Context: a format never looks empty, which is precisely the problem

The pipeline runs in two stages. Stage one extracts facts from a source: which component arrived at which circuit, what lap-time delta was logged in which session, which administrative decision landed on which team. Stage two builds a nine-dimension analysis on top of those facts. When stage one returns nothing, stage two still writes nine dimensions — it simply fills every cell with 'insufficient information'.

The failure is presentational rather than technical. An experienced race engineer who loses data writes 'sensor failed' on the sheet. Software that loses data builds a handsome table. The reader cannot tell the two apart. Seeing 'N/A', they assume the analyst looked, checked, and concluded there was no problem. In fact the analyst never looked at anything.

In F1 that illusion costs more than in most sports, because the paper trail here is unusually dense. The FIA document portal publishes every entry list, every parc fermé sheet, every stewards' decision and every technical directive across a race weekend. Team filings sit open on Companies House in the UK. Promoter ticketing revenue, circuit contracts, administration notices — all of it is recorded somewhere. When the source arrives, verification is straightforward. When it does not arrive, the gap is invisible, because the template covers it.

My own training began in that paper trail. Working in Force India's marketing and sponsorship operation in 2026, I watched the team's India portfolio get built — which sponsor depended on which race's ticketing, which contract survived only if a particular grand prix stayed on the calendar. Car performance and team ledger sat on the same desk. The habit never left.

In 2026, at a London data vendor building a twelve-metric F1 schema for Star Sports India — lap-time delta, tyre-degradation slope, pit loss, DRS-train exposure — the hard part was not the maths. The hard part was forcing the studio to publish definitions before the race rather than after. At the 2026 British Grand Prix, Hamilton's fastest lap was 1:30.621, and the model flagged Bottas's second-place undercut risk two laps before the pit wall reacted. Since that day every preview I write carries a fixed metric box and a methodology note. The word 'momentum' appears only when it is tied to a rolling three-race lap delta.

The second lesson arrived in 2026. Force India entered administration and returned from the Belgian Grand Prix as Racing Point Force India, a new entry. At Spa, Pérez finished fifth and Ocon sixth — eighteen points. Because the entry was new, the season's accumulated 59 points were wiped. My adjusted model showed the car's race pace that weekend was still fourth-fastest despite the reset. That is where my rule came from: a crisis story does not get written until it can be tied to a specific points, payroll or pace metric.

Core: nine dimensions, and the minimum each one actually requires

The real value of that empty report was the inverse question it forced — what is the minimum input needed to open each dimension? The list matters now, because every F1 analytics product should be auditing itself against it before the 2026 regulation reset.

The technical dimension requires a component name, the circuit of introduction, and the lap-time or sector delta that followed. Without those, 'the upgrade worked' is a weather forecast with the temperature removed. The deeper issue is not the upgrade but its allowance: under ATR — the aerodynamic testing restriction — wind tunnel and CFD hours are allocated in reverse order of the previous season's constructors' standings. Backmarkers get more. A backmarker's 'big upgrade' and a front-runner's 'small upgrade' can be identical in hours, and comparing them without the ATR context is dishonest.

The strategy dimension requires circuit, session, compound, pit-lap numbers and the timing of any safety car or VSC. The whole undercut calculation rests on one inequality: the gain from fresh tyres must exceed that circuit's pit loss. Most circuits sit at 20–24 seconds of pit loss; fresh tyres typically deliver 1.5–2.5 seconds a lap for two or three laps. An undercut only exists if the rival's tyre is already fading. At the 2026 Austrian Grand Prix, Bottas took pole at 1:02.939 — a product of a short lap, not a supernatural lap. Had the safety car fallen differently that weekend, the entire race story would have changed and the lap times would have stayed identical.

The team and driver dimension offers exactly one controlled experiment: the teammate. Same car, same setup window, same engineering meetings. That is why I start every driver valuation with the teammate head-to-head rather than the championship table. In 2026 Alonso scored 81 points to Ocon's 92 at Alpine. On paper Ocon was ahead. Once retirements and DNFs were stripped out and qualifying delta, race pace, tyre management and reliability were combined into what I call the Driver Performance Surplus Index, Alonso returned a positive pace surplus. He was undervalued before the Aston Martin move, and that conclusion came from the adjustment, not the table.

The competitive landscape dimension turns on the regulation cycle. Early in a cycle, order reshuffles, because a large rule change resets incumbent strength. Late in a cycle, order ossifies, because every concept has matured and the cost cap makes a large leap difficult. 2026 is a full reset — new power unit, new aero philosophy, new car concept. Reading 2026's constructors' order as a forecast for 2026 is a category error, and any analyst offering next season's predictions from the current table is ignoring where the cycle sits.

The regulation and governance dimension needs three things watched continuously. First, technical directives — the FIA documents that tighten the interpretation of an existing rule, usually to close a grey-area design. Second, cost cap investigations and their outcomes, which change the following season's development capacity later but harder than any points penalty. Third, Super Licence penalty points, where twelve accumulated triggers a one-race ban. If none of the three appears, the governance dimension is empty no matter how neatly 'risk: low' is printed.

The driver market dimension prizes source tier above contract status. A contract story from an established paddock journalist is a lead. The same story from a fan account is noise. From a source close to a team that wants a renewal lever, it may be leak-and-inflate — seeding a rumour to move a price. Without a source tier, a transfer claim cannot be weighted at all, because an authoritative reporter and a hype account look identical.

The risk dimension has three high-frequency hard exposures: power unit component quota breaches producing grid penalties; wind tunnel and CFD data failing to correlate with track data; and core technical talent leaving, especially once gardening leave expires. Gardening leave is the enforced gap between an engineer leaving one team and starting with another, designed to erode the competitive value of what they carry. A big engineering signing therefore lands six to twelve months later, not immediately. Without a date and a job title, the assessment is impossible.

The public narrative dimension carries the most familiar trap: the winter champion. Pre-season testing hides fuel loads and engine modes, so the fastest time is often not the fastest car. A second trap is the nationality filter, where certain national media systematically overrate their own drivers — detectable only if both the outlet and the driver are known. The third, and the most instructive for me, was the empty stadium of 2026. When races ran without crowds, plenty of people assumed the atmosphere had gone and performance would follow. I built an empty-track correction from 2026 and 2026 data at the same circuits, separated track evolution from support-race effects, and published a confidence interval against every qualifying delta. The stands were empty; the stopwatch was not. The empty stadium did not silence the stopwatch; it exposed the noise. That year I scrapped three scripts that had overread the numbers.

The industry transmission dimension asks where the money is moving. FOM, Formula One Management, holds the commercial rights under Liberty Media, distributes revenue, sets broadcasting and shapes the calendar. A team's on-track performance and its commercial value are separate columns, and merging them is the most common analytical error in the sport. Buddh International Circuit is the cleanest case: roughly 95,000 race-day attendance in 2026 sliding toward 60,000 by 2026, three Vettel wins, one Tilke Grade 1 circuit — and no Indian Grand Prix since 2026. MotoGP Bharat arrived in 2026; the cars did not. My verdict keeps the same shape: the stopwatch never lied, the hosting fee did.

Seen through those nine dimensions, the empty report's real value becomes clear. It was not an analysis. It was a gate left open.

The people behind the ledger

Force India's 2026 administration, written only as a points reset, is half a story. At Spa that weekend the car was fourth-fastest, and the data proves it. Inside the garage, that week was not about points. Administrators were appointed, buyers were sought, and the contracts of hundreds of staff hung on a single document. When I write a solvency audit, I reserve one paragraph for the people who appear in the ledger only as a cost line — the mechanics, the tyre fitters, the factory shifts, the families who lived through those months of uncertainty. That is not sentiment; it is completing the reporting.

For Indian readers the same story has a second end. When F1's return to India is debated, the conversation is about hosting fees and circuit grades; the fan's cost barely enters it. From a London desk I can audit the pricing pages of platforms like FanCode, but the 1 a.m. telecast, the club-racing ladder around Chennai and the Drive to Survive generation's habits do not show up in the data. So any claim about that audience gets attributed to a measured source or a named Indian voice, never to my inference.

Contrarian: the shortage is provenance, not data

The industry assumption is that F1 lacks information. The opposite is true. A race weekend throws off enough numbers that any of them can be attached to any conclusion. What is missing is the line that should accompany each number: where it came from, who measured it, and under what conditions. Provenance scarcity, not data scarcity, is the actual crisis.

Its most dangerous form is the story with the most decimal places. When a rumour arrives wearing the shape of 'a 0.3-second gain according to sources inside the team', readers file it as data. For me it remains a lead, not a finding. The rule is simple: an anonymous whisper is not published as fact until a document, a timing sector or a named on-record quote corroborates it.

The second error is mistaking correlation for cause. A package arrives, a team finds two tenths, and the headline says the package worked. More often the circuit that weekend matched the car's setup window, or the compound was softer than a year earlier, or track temperature was suppressing degradation. The real test comes three races later, at a circuit where the package should offer nothing. A lap time is a claim; the analyst's job is to build its evidence chain.

The third, and quietest, is reading 'no signal' as 'no risk'. When a dashboard shows no red flag against a team, the person in the meeting concludes all is well. The flag is absent because the sensor was never fitted. In this sport that has a direct cost: an unlogged reliability problem, component quota pressure or a departing engineer resurfaces as next season's grid penalty, and by then nobody can call it a forecast.

The fourth, the one I find hardest to forgive, is retro-editing the locked box. Pole margin, pit-window spread, points swing — published with a timestamp before the session, graded in public afterwards. Quietly dropping a forecast that missed, or shifting its threshold after the result, is the single professional sin I cannot absorb. The model is rarely wrong; race conditions change the question. The 2026 safety car voided several strategies — that was not the model losing, that was the part of the model that should have been written down in advance.

Takeaway: three questions for any analytics product before 2026

First: how many facts here carry a source, a date and a measurement condition? If the answer is zero, the report is not analysis however handsome it looks. Second: is there a forecast in it that can be proven wrong? A claim that cannot fail is decoration, not a claim. Third: when the data was missing, did the report say so plainly, or did it hide behind 'N/A' and manufacture an impression of health?

The 2026 reset will sharpen all three. New power units, new aero rules and a new car concept will produce a flood of claims, most of them without a measured number behind them. The people who survive that flood will not be the fastest analysts. They will be the ones who decided earliest when to say nothing.

Good data governance is boring, which is why it wins championships quietly. I still have that empty twenty-two-page report. I will never throw it away, because it reminds me every morning that the most dangerous report is not the one with a wrong conclusion — it is the one that performs the act of concluding while containing nothing.

The question is not whether the data will arrive next time. It is whether your report will name the gap when it does not.