Cadillac F1: Three Engineering Bases and the Ceiling of a New Project

After the Canadian Grand Prix, Cadillac F1 went through a ten-race stretch in which both cars finished together only twice: Britain and Italy. On debut, Valtteri Bottas retired and Sergio Pérez finished 16th. Across the next three rounds, both cars completed the race — a result Red Bull and McLaren could not claim over the same span. Then Otmar Szafnauer, former Alpine team principal, appeared on a podcast and supplied the explanation for the rest of the picture: the team keeps design at Silverstone, aerodynamics at Indianapolis, simulation in South Carolina.

I place those two facts side by side because they do not tell the same story. One is the worst finishing rate in the new-team cohort. The other is a launch phase more stable than two rivals that later improved visibly. Every decision in elite sport leaves a footprint; the work here is to read upstream and see who walked through.

Context: a debutant with race-winning drivers and three addresses

Cadillac is the Formula 1 team of TWG Motorsport, competing in its debut season and sitting last in the Constructors' Championship. The race seats belong to Pérez and Bottas, both grand prix winners. On trackside human capital, that configuration is rare for a newcomer: two drivers with multiple seasons in front-running cars, fluent in reading data and feeding technical feedback back to the team.

The engineering organisation, however, is spread across three locations. Silverstone holds design. Indianapolis holds aerodynamics and, according to the source, is still under construction. South Carolina holds simulation. Three time zones, three supply chains, three working cultures.

The stated ambition of the project sits at the top tier: fighting for the World Championship. The current standings sit at the bottom tier. The gap between those two facts creates a pressure field every newcomer must cross; only the crossing speed differs. The speed of a generation is not in their feet, it is in how they absorb pressure.

Mechanism: why geographic distance becomes a distance in time

The technical core of the story turns on one concept: correlation failure — the mismatch between data generated away from the track and the car's actual behaviour. When the wind tunnel, CFD and simulator sit in three different places, every upgrade package must pass three layers of confirmation before sign-off. Upgrade lead-time lengthens. The biggest risk is parts that score well in testing but do not convert into lap time.

That principle explains part of the data, not all of it. Across ten races after Canada, the team finished with both cars only twice. That is a severe operational statistic for any team, and more severe for one with two experienced drivers. When operators are proven, the probability of individual error falls, and the residual probability shifts toward the car and the organisation. This is probabilistic reasoning and can be wrong, but it moves the weight to where the checks belong.

Two distinct explanations must be separated here. The first is the fragility of the car design itself. The second is the quality of coordination across three sites, from data reconciliation to parts logistics. The source chose the second without ruling out the first. On the public data available, both survive equally well.

One further variable: relative trajectory. Red Bull and McLaren were recorded as clearly improving after Canada. Cadillac went the other way. An absolute position at the back surprises nobody for a newcomer; the direction of travel does. A new team can sit last and still be climbing. A new team sitting last and sliding is a different message entirely.

On the regulatory side, a dispersed structure creates two separate cost lines. Capital spending on facilities generally sits outside, or only partly inside, the FIA cost cap. Operating expenditure does not. Three sites drag duplicated operational headcount across three continents, and that duplication competes directly with the capped operating budget. The team can afford to build; it pays with the depth of its engineering department.

Aerodynamic Testing Restrictions allocate wind-tunnel and CFD runs by championship position, with lower-placed teams receiving more. Sitting last in the Constructors' Championship, Cadillac holds the largest regulatory allowance in the field. That allowance is designed to produce fast convergence. If convergence does not arrive, the organisational hypothesis gains weight.

Personnel is the least discussed risk line. Aerodynamicists in Indianapolis and simulation engineers in South Carolina can come to see themselves as peripheral to the Silverstone centre. In engineering organisations, being separated from the main campus is among the most common reasons people leave. Losing staff at a satellite site generates no headlines, but it erodes accumulated capability across a season.

One data point is rarely mentioned in the mainstream account. In the first three rounds, Cadillac finished with both cars while Red Bull and McLaren did not over the same span. For a debutant, that signals baseline reliability at a workable level. Raw engineering capability is the least suspect element. The most suspect element is coordination between departments.

Based on my experience tracking races and transfer deals, I have picked up one simple rule: when operators with proven records still produce poor collective results, the probability of failure sits with the system more often than with the individual. The rule does not replace the data. It only tells me which data to read first.

The contrarian angle: an accusation standing on a hypothetical premise

Szafnauer's argument is rhetorically strong and evidentially weak, in exactly the way a hypothetical parity exercise is. He stipulates that every variable is equal — car, engine, drivers — and isolates the structural variable for inspection. That isolates the variable, but the isolating condition is set by the speaker rather than observed. A conclusion of the form who wins, Mercedes, is a rhetorical construction and should be treated as expert opinion, not as a finding.

He is also a former director of a rival team, speaking on a podcast the source records without episode number, air date or transcript. He comments from outside, meaning he pays no accountability cost for being wrong. That does not make him dishonest; it only removes the self-correcting feedback that internal decision-makers always carry.

Across everything recorded, there is not a single statement from Cadillac leadership, engineering or drivers. A critique at this level, published with no right of reply, tends to be absorbed by audiences as an established conclusion. That is an asymmetry the source does not acknowledge.

The headline packaging also deserves attention. Pérez's name is placed at the front of a piece whose content is entirely about organisational structure. The driver is pushed to the shopfront of a critique not aimed at him. For Pérez's large readership, that packaging inflates the perceived scale of the crisis well beyond its sporting weight.

Deeper still, there is a timing mismatch. Indianapolis is still under construction. The structure under attack is the fully operational version of the three-base model — a version that has never existed. Judging a steady-state model from a transitional build phase misreads the phase.

One further layer: two things are being blended. Car reliability is measured data. Organisational coherence is an assertion. Using the first as proof of the second is an inference that has not been licensed, at least not yet.

The real governance question sits elsewhere: mandate ambiguity. Szafnauer sets two objectives side by side — make the team more American, or do everything necessary to win the World Championship. When both objectives coexist without a publicly stated priority order, every resource dispute between sites loses its referee. A racing project does not collapse because of one mistake; it collapses because a chain of decisions was inflated into a strategy without anyone daring to question the priority order.

One thing must be said clearly about the raw material here. All the result data above comes from a single source with no publication name, no publication date, and a description of a season that cannot be independently checked against any official results table. The part that can be analysed seriously is the argument about organisational structure. The rest belongs in a pending-verification drawer.

What falls next

The biggest risk in this entire story lies in acting on a single source. The three-base model may be wrong. But the evidence presented has not established that. For a project under construction, the correct response is usually patience at the structural level and urgency at the reliability level. Reversing that order is a costly and hard-to-reverse error.

Three signals to track in the rounds ahead: the both-cars-finishing rate race by race, any change to the three sites, and the appointment of a single technical authority holding cross-site decision rights. If all three shift direction together, the organisational hypothesis lives. If not, the story is just a podcast.

Insiders stay silent, outsiders guess. Between the two, I choose to read the results sheet race by race and ask one simple thing: if this project is optimised for identity rather than for trophies, who is responsible for saying so first?