Austin briefing: Antonelli on pole, four teams within three tenths, and a first stop that decides it
Antonelli on pole by 0.18 s from Verstappen, Hadjar third, the Ferraris fourth and fifth and Russell sixth – four teams inside 0.28 s. Our plan is one stop, soft → medium around lap 27–29: the two tyres that were level on race pace, a long first stint because this season's tyres have outlasted every Friday forecast, and no hard. Mercedes is expected to go medium → hard, and the first stop around lap 27 is where the podium is decided.
The grid: Antonelli, Verstappen, Hadjar, Leclerc, Hamilton, Russell, Piastri, Norris, Lindblad, Lawson, Gasly, Hülkenberg, Alonso, Bearman, Ocon, Stroll, Sainz, Albon, Pérez, Bottas, Colapinto (10 places), Bortoleto (60 places, back of the grid). Pole 1:34.035 on a 38.5 °C track; Verstappen 0.18 s back, Hadjar 0.22, Leclerc 0.23, Hamilton 0.28, Russell 0.27 – Leclerc's and Russell's Q2 laps were faster than their Q3 laps, so the grid understates both.
Our plan is soft → medium, one stop, around lap 27–29. That is the planner's answer on Friday's measured wear: soft → medium with the stop on lap 29 at 5,682 s over 56 laps, medium → soft on lap 27 within a second of it, and soft → hard 22 s further back. It is also what the season says. Seventeen races in, the tyres have lasted longer than Pirelli and every paddock forecast expected in all but the hottest races – Melbourne's hard ran 44 laps against a window at 20–26, Miami's 51, Madrid's mediums had twenty laps left when Pirelli said ten – and a long first stint on the soft is the way to cash that in: the soft wins the start and the restarts on a grid where the overtaking threshold is 0.8 s/lap, Verstappen's eleven-lap soft run on Friday wore at 0.07 s/lap, the best in the field, and a 28-lap soft stint followed by 28 on the medium uses the two tyres that were level on race pace and leaves out the one that was not.
Ruth Buscombe reads it the same way from the other side: soft → medium for Verstappen, soft → medium → soft as the aggressive version, and Mercedes, she expects, on medium → hard. That last plan is the one we would not run. It keeps a stop in hand for a Safety Car, which is its whole appeal, but it finishes the race on the hard, which lost 0.23 s per lap of age on Friday, worst of the three by a distance, and it hands Verstappen and Leclerc the undercut on fresh mediums around lap 27 with 28 laps of better rubber to use it. Pirelli's simulation has soft → hard and medium → hard as its fastest plans (windows 20–26 and 21–27) and medium → soft at 26–32 as Marrafuschi's personal pick because the drivers do not like the hard; we agree with the drivers and with the pick, not with the simulation.
The two-stop – soft → medium → soft, stops around 19 and 37 – is 29 s quicker in our planner if the race wears exactly as Friday did, and 2 s slower in Pirelli's. Both are right about their inputs; what separates them is whether Friday's slope survives Sunday, and on this season's record it does not: Fridays have overstated the race by two to eight times, and Pirelli's own line is that a stable 40 °C track will not wear faster than Friday. So the two-stop is the drawer plan: live if the leaders' softs are visibly gone by lap 15, otherwise not. Teams want five seconds in hand before they add a stop; nobody has that here.
The field simulation from this grid says what the plan does not: position. Its one-stop consensus has the top six stopping between laps 23 and 30, the window three laps wide, and inside it the order is set by who stops first – Verstappen from P2 beats Antonelli by stopping a lap earlier, Leclerc from P4 does the same, Hamilton and Russell reach second by stopping a lap before the car ahead. On soft → medium the undercut on a 21 s pit lane is worth about two seconds, which is the gap between P1 and P4 on this grid. The first stop is the race; the second, if it comes, is the lottery.
Pit loss 21.2 s under green (2025, 18 stops measured), 4.6 s under a VSC; overtaking threshold about 0.8 s/lap. A second stop costs 21 s over 56 laps, and the car that makes it can pass the cars it comes out behind: COTA is the first circuit this autumn where the two-stop is a plan rather than a lottery ticket.
Pirelli brings C2, C3 and C4. The FIA tyre notes (document 2) require at least one of C2/C3 in the race, make the C4 the Q3 tyre, and set minimum starting pressures of 26.5 psi front and 25.5 rear – the highest of the autumn (Bahrain 25.0/24.5, Singapore 24.0/23.0). Camber limits −3.0° front, −1.75° rear.
Friday, 41 °C track in both sessions (the forecast said over 50; the cloud came): the long runs were long enough to trust the shape, if not the level. Soft 0.152 s per lap of age over 99 clean laps in 12 runs, R² 0.80; medium 0.163 over 61 laps; hard 0.232 over 36 laps in four runs, R² 0.92. Norris lost 0.25 s a lap on an eight-lap hard stint, Colapinto 0.25 over twelve. Verstappen's eleven-lap soft run wore at 0.07, the best in the field; Leclerc's at 0.16 on the same compound and length. In race trim our long-run fit has the three compounds level on pace – medium and hard within 0.06 s/lap of the soft – where Pirelli quotes half a second from soft to medium and nine tenths from medium to hard over a lap; both can be true, one is a qualifying lap and the other a stint. Either way the soft is the race tyre from the start and the hard has little to offer: slower to warm, the worst wear, and Pirelli's own "limited role". The high pressures said so on Wednesday, and the drivers said so around the coffee machine: Marrafuschi's reason for picking medium → soft over his own simulation's hard-stint plans is driver feedback. On Friday F1.com had Mercedes and Red Bull with a fresh soft each saved for the race; our own count after qualifying, below, says only Red Bull still has it. Pirelli's other warning is about the bumps: an aero imbalance over them shows up as front locking into the heavy braking zones, which is exactly what Mercedes (Shovlin) and Haas (Ocon) spent FP1 fighting.
What is left in the garages, counted from the timing data – new sets started in FP1, FP2, FP3 and qualifying, out of the weekend's 8 soft, 3 medium and 2 hard: Verstappen and Hadjar have two new softs, one new medium and one new hard each, the only front-running pair with a new soft for the start and another for a late stint. Leclerc has one new soft, one new medium and one new hard; Hamilton has no new soft but two new mediums. Antonelli and Russell have no new soft left – four each went in qualifying – and one new medium and one new hard apiece, which is Ruth Buscombe's medium → hard reading of Mercedes written in rubber. Norris has no new soft and two new mediums, Piastri no new soft, one medium and one hard. Of the midfield only Ocon kept three new softs; most of the rest have two. A soft from Q1 with three laps on it is a perfectly good race start, so no new soft is not the same as no soft; it moves the first stint from fresh to scrubbed, three laps of life already spent. Which used sets were handed back after practice does not show in the timing data, so the count of used sets is a ceiling. For the plan: Red Bull and Leclerc can run our soft → medium on new rubber; Mercedes' version is a scrubbed soft → new medium, or the new medium → new hard the drivers dislike; Hamilton and Norris have two new mediums and could go medium → medium or scrubbed soft → medium.
2026 is a new tyre regulation and this trio has now raced five times: Shanghai and Madrid at 0.005 s/lap on a cool track, Spa at 0.02–0.03, Sepang 0.03–0.05 with the hard worst, Barcelona – a 50 °C permanent circuit with a long fast-corner sequence – 0.21/0.16/0.10 and a three-stop race. Friday's COTA numbers sit at the Barcelona end on a Shanghai-temperature track, which is the esses doing the work rather than the heat. The season's other lesson applies: Pirelli's Saturday window has been the earliest safe stop, not the fastest, and the C2 has lasted a race distance whenever the track stayed under 45 °C – but that was on circuits where it wore at nothing on Friday too. We will read Pirelli's note tonight against these numbers; if they say one stop, the question to ask is on what wear.
Our temperature rule: the practice α counts in the plan because Sunday is forecast within 6 °C of Friday's 41. If it is a 50 °C afternoon the library level takes over and the race moves towards the one-stop end of the range.
The FIA's power unit document (document 8) is out, and it is kinder than Singapore's. Race recharge 8.5 MJ per lap, 9.0 with Overtake active; qualifying 8.5, practice 9.0. The two main overtaking zones – Turn 11 to 12 and Turn 20 to Turn 1, the same two zones F1's guide lists for Straight Mode – run on the base curve: the full 350 kW up to about 290 km/h, tapering to nothing at about 345, or to 360 with Overtake. The "alternative" curve applies through Turns 4–8 (1300–1850 m) and the whole of Turns 12–20 (3700–5250 m), but unlike Singapore's 250 kW cap it is 350 kW up to about 300 km/h with a gentler taper – at the speeds those corners are taken, that is full power. So there is no capped sector at COTA worth the name; the budget is the whole constraint, and the budget is 8.5 MJ against a lap with over 30 s of full throttle. Overtake detection is at 5326 m, after Turn 20, activation at 5420 m: one zone, into Turn 1, with a 1.0 s detection gap.
Two things to read into that. First, the base curve's own taper means every car loses electric power above 290 km/h by regulation, which is part of why the long-straight fade we measured at Baku looked like drag: at 330 km/h the rule allows about 160 kW, not 350. The Turn 11–12 straight tops out around 320–330, so the end of it is regulated as much as budgeted; the knee that tells us a car has run out comes earlier and sharper than that. Second, with full power allowed through the stadium, the place to save is the back straight itself, and the car behind with Overtake into Turn 1 gets 350 kW to 340 km/h against the leader's taper from 290: that is a real straight-line advantage, which is why spending everything on Turn 11–12 is the way to be passed at Turn 1.
Friday's back straight, measured: every driver peaks 320–440 m before the Turn 12 braking point and then loses speed with the throttle flat. In FP2 Hadjar lost 5 km/h and Verstappen 6, on the lowest top speed of the front-runners (318); Piastri 12, Hamilton 14, Norris 15, Russell 18.5, Leclerc 20, Antonelli 21. FP1 was more extreme (Antonelli 32, Verstappen 7) and included a Ferrari map test – Leclerc's speed trap went from 309 in FP1 to 333 in FP2 – so FP2 is the picture to keep. The pattern is the one from Monza, Baku and Singapore: Mercedes gives the end of the straight away earliest, Red Bull holds everything to the brake.
What it means tactically, given the power-unit document: the only Overtake zone is into Turn 1, with detection after Turn 20, so a pass into Turn 12 costs energy the attacker then lacks through the stadium and up to Turn 1, where the car he just passed has the mode and the charge. Mercedes' shape is the map that saves for Turn 1; Red Bull's is the map that defends Turn 12. Whether Mercedes' is a choice or a shortage, Sunday's sector-3 times will tell. And with a 21 s pit stop and an 0.8 s threshold, the cleanest pass here is the undercut, which costs no energy at all.
The FIA's Car Presentation Submissions are in: 20 updates across eight teams, and the headline is Ferrari. A full floor package – floor body, floor board, a sidepod inlet with more lower volume and a reshaped undercut, and a revised diffuser – all filed under one sentence: more floor downforce front and rear, and more energy and flow stability sent to the rear. That is the biggest single submission from a top team since Red Bull's Baku floor, and it comes to the circuit where floor load through the esses is worth most. Mercedes submitted nothing new: the "latest package" Wolff spoke of is the one it has been trying to make work since Sepang. McLaren brings three items – front corner, rear corner and diffuser, all flow conditioning with "local load across the speed range" on the diffuser – which is the floor-and-brake-duct update Marshall described. Red Bull lists a front-wing option with a tyre-temperature turning vane, for evaluation, and a reliability fix to a suspension gaiter. Racing Bulls four items (front wing endplate, rear wing flaps, floor, engine cover), Audi three, Alpine two, Cadillac two; Williams, Aston Martin and Haas nothing.
Friday's verdicts, from the teams' own debriefs (F1.com): Leclerc, "our upgrades worked as expected", with the warning that Antonelli and Verstappen have more to unlock; Hamilton tried a different set-up direction and was uncomfortable over the bumps. Mercedes had a poor FP1 – Shovlin: locking while exploring a wider set-up window with the package on both cars – and fixed it for FP2, both cars now on the same specification and Antonelli within 0.04 s of Leclerc. Red Bull (Wache) is running a less stiff set-up than last year and still not compliant enough over the bumps, with the race as the priority and low-speed performance the focus; Hadjar's FP2 changes made the soft worse than the medium. McLaren (Marshall): closer than FP2 suggests, long runs positive, the small package delivering, low-speed corners the weakness – Norris says the pace is there but hard to repeat lap after lap, Piastri says grip. Everyone names everyone else: Shovlin would not rule out any of the top four for pole, Marshall names Ferrari and Verstappen. That is a Friday where nobody is sandbagging and nobody is sure.
The question for this track is the one Piastri put on Thursday: Ferrari is often the quickest car through the corners in the teams' analyses and loses it on the straights, and he could not say whether that is drag or power. Qualifying speed traps: Gasly 341 and Colapinto 341 (Alpine with the wing off for Sunday's overtaking), Antonelli and Norris 330, Russell and Piastri 328, Verstappen, Hadjar and Hamilton 326, Leclerc 318 – the pole lap was set with 12 km/h more than Leclerc's at the trap, and Ferrari's corner speed closed the rest to 0.23 s. Red Bull qualified second and third with the second-lowest top speed on the grid and no part of the lap where it loses to anyone. COTA has 2 km of full throttle in two pieces and 3 km of corners; it is the first track since Monza where that trade is tested both ways on the same lap.
Open-Meteo for COTA, fetched Thursday: Friday 27–28 °C in the air during FP1 and FP2, humidity under 30 %, wind 11–16 km/h from the south-east, dry; FP2 under thicker cloud than FP1, which would put the track several degrees cooler at 16:00 than at 12:30. Saturday 29–30 °C, south-easterly 20–25 km/h with gusts, dry. For Sunday F1's race-day note says scattered cloud, air around 28 °C, track around 40 °C – the old tarmac does not absorb heat – dry, and the Gulf wind gusting to 30 km/h as it has all weekend. The wind is the weather story: it decided several Q3 laps, it will feed track-limit deletions at Turn 20, and the slides it causes through the esses are what spike rear tyre temperatures and turn a one-stop into a two.
Two consequences. Track temperature on Friday will exceed 50 °C in FP1 and may be cooler in FP2; if Sunday is in the high forties, FP2 is the session whose degradation we trust. And a south-easterly at 25 km/h is a crosswind through the esses and a tailwind down the back straight, which raises top speed and pushes the braking point for Turn 12 towards the barrier – the sector-1 balance will move with the gusts, and the cars that suffer are the ones already on the edge of the front tyres.
5,513 m, 56 laps or 120 minutes, start 15:00 local, which is 21:00 in Sweden after the clocks go back on Saturday night. The start line is 296 m into the lap, up the hill before Turn 1 (FIA Competition Visa). Pit entry is before the garages; the pit lane is short, which is where the 21 s pit loss comes from. In the two races we have measured, 3 laps per race were neutralised; F1 counts a Safety Car in 29 % and a VSC in 43 % of the last seven races here, and 71 overtakes in 2025. Our overtaking threshold for the circuit is 0.82 s/lap of pace advantage – the DRS-era figures were 0.63 and 0.98 – against a season median of 1.0. Track position is worth less here than at any track since Monza.
The two straights have different jobs. Turn 11 to 12 is where energy is banked or spent; Turn 20 to Turn 1 is where the Overtake detection sits, so a car attacked into Turn 1 is a car that spent too much on the back straight.
Changes to the circuit this year, from the Race Director's notes: resurfaced from the entry of Turn 9 to the entry of Turn 11, the bumps before Turn 12 and Turn 19 smoothed, the painted "kerb" at the exit of Turn 11 removed, and new blue track-limit lines at most corners. Smoother braking zones for the two heaviest stops on the lap, and a cleaner exit from the hairpin: the back straight will be a little faster than last year before anyone touches the energy map. If a race is suspended, the Safety Car waits for the field at Turn 11 for the resumption.
Race form going in, our measure (gap to the fastest car, weighted to the last three races): Mercedes 0.16 %, Red Bull 0.29, Ferrari 0.33, McLaren 0.34. The trend is Red Bull's: −0.64 points since the Baku package, two wins in a row, and Hadjar saying on Thursday that Ferrari had the fastest car in Singapore anyway. Verstappen himself warned that Sepang and Singapore were hot and humid and Texas is not, and that two wins do not make the fastest car in Austin. Friday's long runs, our fit: Verstappen and Leclerc level, Norris 0.1 % back on the hard, then Russell, Hadjar and Hamilton 0.5–0.7 % off, Piastri and Antonelli about 1 % – Antonelli's on an 11-lap medium run that wore at 0.07. F1's own race-simulation chart reads the same sessions tighter: Ferrari, Red Bull +0.05 %, McLaren +0.07, Mercedes +0.13, four teams inside 0.13 %, which matches Shovlin's "any of the top four for pole" better than our Mercedes number does; the difference is method (we fuel-correct a single long stint, they take the team's best), and we flag it rather than pick. Over one lap F1's ideal-lap chart has Leclerc, Verstappen, Russell and Antonelli within 0.15 s, with Antonelli leaving 0.04 s on the table against Verstappen's 0.28, Russell's 0.57 and Norris's 0.58 – Ferrari and Red Bull have shown their hand, Mercedes and McLaren have tenths left in a clean lap. Ferrari says the Austin parts complete the package that began in Madrid four races ago, which is what the submissions look like.
Wolff after qualifying doubted aloud whether the car has the race pace to beat Red Bull and Ferrari, and called the pole lap one the team cannot yet explain; Shovlin named Turns 12–15 as the Mercedes' weakest sector after big overnight changes. Verstappen said the rear did not give him the grip he expected in the slow corners and that the wind made the car hard to read. The Race's long-run averages from Friday, uncorrected: Norris 1:41.289 on the hard (9 laps), Verstappen 1:41.481 soft (11), Leclerc 1:41.503 soft (15), Antonelli 1:41.521 medium (12), Russell 1:41.602 medium (6), Hamilton 1:41.949, Piastri 1:41.997 – and their telemetry has Verstappen gaining about six tenths on Leclerc on the straights over a lap, which is drag, map or exit speed, not necessarily engine. F1 puts the top four within 0.12 s on long-run pace. Antonelli's pole is Mercedes' first since the package arrived, on both cars in the same specification for the first time, and it comes with Russell sixth on a Q2 lap that would have been fourth – the car is there, the lap was not. Red Bull put both cars on row two from the same session in which Hadjar called his FP2 laps disgusting; the overnight changes worked. Ferrari's Q3 was slower than its Q2, which is either a track that went away or two cars that peaked early.
The title: Antonelli can only settle it here by winning with Hamilton fifth or lower and Russell ninth or lower – "theoretically possible but in practical terms, this isn't the match point weekend" (Lord). Antonelli leads Russell by 102 points with six Grands Prix left; Mercedes leads Ferrari by 134 in the constructors'. The championship cannot be settled here, so nobody is driving for points yet – which is when teams try things.
Qualifying gaps to pole: Verstappen 0.179, Leclerc 0.163 in Q2 and 0.232 in Q3, Hadjar 0.215, Russell 0.270 in Q2 and 0.441 in Q3, Hamilton 0.276, Norris 0.504 in Q2 and 1.014 in Q3, Piastri 0.768. The pole-time bet: our 1:35.6 from the season's 2026/2025 ratio was 1.6 s out; Marek Lubieniecki's 1:34.58 from a physics model 0.55 s out; the FP2-based 1:33.9 we posted on Saturday 0.14 s under. Lesson filed: at a track with two kilometres of straights the ratio method is wrong by a second before anyone drives. Our driver form after Singapore: Verstappen 0.24 % off the fastest car, Antonelli 0.28, Leclerc 0.30, Norris and Piastri 0.54, Russell 0.56, Hamilton 0.59. Verstappen has won the last two and started the last two from pole; Leclerc was within 0.054 s of him in Singapore qualifying and finished second on a different strategy. Hamilton's Singapore – a medium start that was slower than the soft and a missed pit entry under the VSC – is the kind of weekend that makes a driver take the aggressive option next time, and Ferrari's stated habit of splitting strategies gives him the chance. On a circuit where the overtaking threshold is 0.8 s/lap, the driver measures that matter are our attack and defence factors: Red Bull +0.26 in attack, Mercedes +0.21 and +0.92 in defence, Ferrari +0.17 and +0.63 – the Mercedes is the hardest car to pass on the grid and the Red Bull the one that passes.
Twelve FIA documents so far, including the entry list, self-scrutineering, the car presentation submissions and the power-unit count; the earlier ones: the Competition Visa, the timetable (v4), the Pirelli notes, the circuit map, the Race Director's notes, the power unit information, the car display procedure and a clean post-race inspection of Bearman's clutch from Singapore. Track limits: during qualifying, a lap on which a driver fails to negotiate Turn 20 may be deleted together with the following lap – the last corner, right before the Overtake detection line. The maximum time between the Safety Car lines for slow laps is announced after FP2. Safety Car boards and waved yellows are not used; the light panels and the timing messages are the signal. Race tyres must be fitted four hours before the start. Penalties: Bortoleto's bill grew on Saturday to 60 places – a fifth engine, turbo, exhaust, energy store and control electronics, a fourth MGU-K and a seventh ancillary – so Audi has taken the whole season's worth of parts at the cheapest circuit left, and starts last; Colapinto 10 places for a fourth control electronics unit, 22nd. Lindblad's sixth ICE is on the Technical Delegate's list without a decision yet. Bortoleto was also warned for impeding Hamilton at Turn 12 in FP3; Hülkenberg cleared for the practice-start incident with Stroll; Russell cleared for the Turn 18 moment with Leclerc in FP2; the 1:55.0 maximum between the Safety Car lines applies on the reconnaissance laps, and Ocon and Piastri were cleared after exceeding it in Q1. The power-unit count (Technical Delegate, 23 October) is the one to keep on the wall for the last six races: Verstappen, Leclerc and Hamilton have used four ICEs, four turbos, four exhausts and three of each of the MGU-K, energy store and control electronics – every allowance spent, so the next fresh element on any of those three cars is a grid penalty. Piastri is on his fourth ICE too. Antonelli and Russell are already on five ICEs (Russell took the penalty in Singapore), Hadjar on his eighth, Aston Martin on seven MGU-Ks per car. If a team has to take a penalty before Abu Dhabi, a circuit with a 21 s pit stop and an 0.8 s overtaking threshold is where it hurts least – that is a reason to watch the Friday technical reports here more than most weekends.
A soft that goes off the way Friday's slope said, 0.15 s/lap: then the 28-lap first stint is a 20-lap one, the medium → hard cars inherit the race and the two-stop comes out of the drawer. A hard that comes alive at race pressures and temperatures, as this trio's hard did at Shanghai and Madrid: then Mercedes' medium → hard is the right plan and ours ends on worn mediums behind it. A Safety Car in the first 20 laps, 29 % likely on F1's count: then the medium starters get their stop for 5 s and everyone on a soft has already spent the start advantage. And a 20 % rain shower, which would make all of it background.
How to read this: the briefing is written before the race from practice and qualifying data, FIA documents and Pirelli's information. The sections mirror how a Formula 1 team splits the work; the numbers are ours and each carries its source. After the race the Aftermath says where we were right and where we were wrong.