2026 F1 Season · Through Belgium · Where Cadillac F1 Actually Stands
Ten races into 2026, Cadillac still holds 10th of 11 — and Spa-Francorchamps, the fastest and most downforce-hungry circuit of the year, told the clearest story yet. Bottas ran the full Belgian Grand Prix and actually climbed off the bottom of the driver table to 21 / 22; Pérez retired on lap 13 with a rear-suspension failure, so his short, heavy-fuel sample is set aside from the season average (at a circuit this fuel-sensitive, a 13-lap read isn't representative). The season figures barely moved — PER −3.0 km/h, BOT −5.9 km/h below the field-median apex speed — and Cadillac stretched its cushion over Aston Martin to 1.1 km/h. The gap to the field-leading team, Mercedes, is −8.0 km/h. But the real Belgium story isn't the average — it's where the deficit lives.
Through the first ten 2026 races — Australia, China, Japan, Miami, Canada, Monaco, Barcelona, Austria, Britain, and Belgium (Saudi and Bahrain were cancelled) — we measured every driver's minimum apex speed at every detected corner, using each driver's single fastest clean lap. The metric is honest: vmin at the apex, in km/h, compared corner-by-corner against the field-median driver. Higher = better grip = more downforce; the driver who carries the most apex speed has the most car under them. Sergio Perez and Valtteri Bottas sit at the bottom of the 22-driver field, with PER averaging −3.0 km/h below the median apex speed and BOT −5.9 km/h. The Cadillac team rank is 10 of 11, clear of Aston Martin for the sixth race running — and the margin grew to 1.1 km/h. The gap to the field-leader (Mercedes) is 8.0 km/h of average minimum apex speed. Spa is the tell: its average corner gap (−4.6 km/h) is actually a touch tighter than Silverstone's, because Cadillac claws pace back at the slow corners — but at the fast corners it loses more than anywhere all year. That split reconciles the gap tracker, where Spa's long straights made the lap-time deficit look like Cadillac's best of the season while the corners told the opposite story. Britain counts: both cars finished (PER 15th, BOT 17th) and set representative clean laps, so Silverstone feeds the Cadillac averages — unlike Austria. Austria asterisk: both Cadillac cars retired very early with brake fires (PER lap 4, BOT lap 2). Neither set a representative lap, so Austria is excluded from the Cadillac season averages entirely — the rest of the field, who completed the race, do include it, which is why Pérez's corner count sits at 114 while the busiest field runners reach 139. Canada methodology note: Sergio retired on lap 39 with a smouldering chassis fire; to stop the rest of the field's lower-fuel late-stint laps from inflating their relative pace, we cap every driver's best-5-lap pool at lap 39 for Canada only. Barcelona note: Bottas retired on lap 15 — no cap there (it would discard ~77% of the race on a circuit where clean fast laps come later in the stint), but his 15-lap heavy-fuel sample is flagged. Belgium note: Pérez retired on lap 13 (rear suspension); Bottas ran the full race and anchors the team number, so no field-wide cap — but Spa is far more fuel-sensitive than Monaco or Barcelona, so a 13-lap heavy-fuel read isn't representative and PER's Belgium is excluded from his season average (the field, and the per-corner Spa table, still include it). See the methodology page for all four rationales.
Cadillac F1 is in their first season as a Formula 1 constructor in 2026, racing alongside ten established teams. Where do they actually stand? V2 of this analysis ranks every driver in the field on a single physics-anchored downforce indicator: minimum apex speed.
For each driver we pick their single fastest clean lap of the race weekend, resample its telemetry onto a uniform 5 m grid, and identify the corners as local peaks in track curvature. At each corner we read off two numbers: the minimum speed they carried through the apex, and the time they spent in the corner phase versus the straight-line phase on that lap. The driver who carries the highest minimum apex speed has the most grip — which on these new 2026 ground-effect-reduced cars correlates almost entirely with how much downforce the chassis produces.
The phase split satisfies a strict closure invariant: cornering_time + straight_time = best_lap_time, exactly. No averaging, no fudge factors. When you add the two phases back together you get the lap. Read the full methodology for detection thresholds, the throttle/brake phase-classification rule, the per-corner κ math, and what the analysis deliberately doesn't capture (banking, weight transfer, dirty air).
Season avg gap to field median min apex speed (km/h) · 22 drivers · 9 races · 92–122 corners per driver (Cadillac 114 — Austria excluded)
Antonelli holds the individual top spot, and Mercedes now leads on the team metric too. Cadillac's two drivers sit near the bottom in silver with red outlines — PER at #19, BOT at #22. The 3.2 km/h gap between PER's and BOT's apex-speed averages is consistent across races: one driver is reliably extracting more out of the same car.
Pick a race · 22 dots per corner · field median = dashed line · Cadillac silver/red
Click race tabs to switch. The dashed line is the field median for each corner — every driver below the line is losing to the field at that corner. Cadillac's silver and red dots sit below the dashed line at almost every corner of every race; the magnitude of the gap varies by corner type (low-speed hairpins are forgiving, medium-speed sweepers expose the deficit).
Each driver = one dot on best clean lap · diagonals are constant-lap-time iso-curves
Each dot is one driver's best lap. X + Y = lap time exactly (closure invariant). Drivers down-and-left have shorter total laps. Drivers up are corner-fast but straight-line-slow; drivers right are straight-fast but slow through corners. Cadillac's two silver-with-red dots sit up-and-right — slower in both phases than the field's frontrunners.
Per-corner min-speed gap to field median (km/h) · 114 corners across 8 scored races · Austria excluded (both DNF'd) · PER silver / BOT shadow
PER outpaces BOT in 75 of 114 corners (66 %). Both drivers are below the field median at most corners (bars below zero), but PER's bars are consistently shorter — he's squeezing 3.2 km/h more apex speed out of the same car. Where they invert (BOT bar above PER) is usually a corner where Bottas's experience and smoothness pays. Canada corners are sampled from the same fuel window as the rest of the field (lap 1–39); BOT's Barcelona corners come from his 15 laps before retirement, all on heavy fuel. Austria contributes nothing here — both cars were out by lap 4 with brake fires.
Median best-lap speed across the lap · pick a race · 20-driver gray hairball · Cadillac silver/red bold
The gray hairball is the rest of the 20-driver field. Cadillac's two drivers and the field median are pulled out in bold. Look for the dips — that's where the field is in corners and Cadillac's lines fall below the dashed median. The deficit is at the bottom of every dip, not at the top of the straights.
Top 12 corners across all 8 scored races (Austria excluded — both DNF'd) where Cadillac's median apex speed lags the field median most
Albert Park (sky blue) and Suzuka (red) still dominate the worst-corners list — high-load circuits where downforce deficit hurts most. The gap at the worst corner can be 25+ km/h. Monte Carlo (gold) lands just one corner on the list despite being wall-to-wall corners — at low speed the whole field is slow, and the deficit compresses. The pattern holds: the team's deficit is biggest where lateral grip matters most.
Cadillac stays 10 / 11 for the sixth race running, and at Spa the cushion over Aston Martin grew to 1.1 km/h. Bottas ran the full Belgian GP and climbed off the bottom of the driver table to 21 / 22 (PER stays 19th); the season figures barely moved (PER −3.0 km/h, BOT −5.9 km/h). The gap to the field-leading team, Mercedes, is −8.0 km/h.
Austria's story wasn't downforce, it was brake fires. Both MAC-26s retired inside four laps — PER on lap 4, BOT on lap 2 — before either set a representative clean lap. A two-lap, full-fuel, cold-brake sample tells us nothing about the car's cornering pace, so Austria is asterisked and excluded from the Cadillac season averages. The field, who ran the full distance, still counts it.
Spa splits cleaner than any track this year. At the slow corners — La Source, the Bus Stop chicane — Cadillac is level with the field or quicker, which flatters its average Belgium gap (−4.6 km/h) to a hair inside Silverstone's. But at the fast, downforce-hungry corners it posts the biggest single-corner deficits of the season: Eau Rouge / Raidillon (−16 and −12 km/h) and Pouhon (−16.5 km/h), all taken well north of 250 km/h. That's the aero shortfall with nowhere to hide — and it's why Spa flattered the lap-time gap tracker (long straights) while exposing the car through the corners.
PER outpaces BOT in 66 % of detected corners (75 / 114) and averages 3.2 km/h more apex speed across the field. Same car, same setup, same engineers — Perez is extracting more out of the platform than Bottas. With this kind of car, that 3.2 km/h is meaningful.
This isn't a power-unit story or a setup story. The phase split shows Cadillac's slowest moment is consistently the apex, on every track type, against every team. The car needs more downforce. Eight races in, the picture is unchanged — and Barcelona, the truest aero test on the calendar, still confirms it loudest (Austria was lost to brake fires before either car could run).
Austria gave us no Cadillac pace data — both cars were out by lap 4 with brake fires before they could run, so the downforce picture is frozen at Barcelona. That makes Silverstone (R9) the next real aero test, and a demanding one: the British GP is wall-to-wall high-speed, sustained-load corners — Copse, Maggotts–Becketts, Stowe — the kind of layout that exposes a downforce shortfall even more ruthlessly than Barcelona. The questions for Cadillac are simple: can the cars finish, and does the structural aero gap look any different on the calendar's fastest corners? We'll re-run the pipeline after the race and update the leaderboard.