São Paulo briefing, Saturday edition: Norris on pole in the rain, three winners from Friday at the back, and a grid where everyone has new softs
Qualifying was run on intermediates and Norris took pole by two tenths from Russell, with Antonelli third and Leclerc fourth; Hamilton went out in Q2, Verstappen in Q1, and then the engine penalties rearranged everything: Russell drops five places, Leclerc ten, and Hamilton, Verstappen and Pérez start from the back. Nobody used a slick in qualifying, so every driver reaches Sunday with four or more new softs (C5). Interlagos is 4.3 km of old-fashioned racing circuit: anticlockwise, 800 m above sea level, a long climb to the line and a middle sector that punishes a single locked wheel. Pirelli brings the Mexico set – hard (C3), medium (C4) and soft (C5) –, one step softer than last year's sprint weekend, and asks for 1.5 psi more in the front tyres and 2 psi more in the rears than it did in Mexico. The official forecast gives Saturday a 50 percent chance of rain, heaviest from 14:00 local time, an hour before qualifying. Six of the last seven races here have had a Safety Car. We go in with one stop as the working plan and a two-stop in the drawer, and with the question Mexico left us: was 43 laps on a soft (C5) a Mexico result or a soft (C5) result?

Saturday, and the rain came an hour early, as the forecast said it would. Qualifying was run on intermediates throughout, track 29 °C and air 21, with Q2 the quickest segment as the surface dried before the heavier rain of Q3. Norris took pole with 1:27.470 from Russell at +0.204, Antonelli +0.625, Leclerc +0.819, Lawson +0.851, Hülkenberg +1.198, Hadjar +1.293, Bearman +2.361, Alonso +2.841, Albon +2.986. Hamilton was eleventh and out in Q2, Piastri twelfth, Câmara thirteenth in his first qualifying, Lindblad and Sainz behind; Verstappen went out in Q1, sixteenth, and Bortoleto set no time. Then the stewards' documents did the rest. Russell loses five places for Friday's engine, Leclerc ten for Saturday's; Verstappen's 30 places and Pérez's 30 send them to the back; and after qualifying Ferrari fitted Hamilton with a complete new power unit – a sixth engine, turbo, exhaust, energy store, control electronics, MGU-K – for 65 places, which puts him at the back too. The official grid: Norris, Antonelli, Lawson, Hülkenberg, Hadjar, Bearman, Russell, Alonso, Albon, Piastri, Câmara, Lindblad, Sainz, Leclerc, Colapinto, Gasly, Stroll, Bottas, Bortoleto, Hamilton, Verstappen, Pérez. The FIA counts each penalty from the qualifying position without the other penalised cars' moves, so Russell is seventh, not sixth, and Leclerc fourteenth.
The wet session changes the tyre arithmetic completely. Nobody fitted a slick in qualifying, so the eight softs (C5) are intact apart from what Friday used: after the Q3 return, Norris, Russell, Antonelli, Leclerc, Lawson, Hadjar, Bearman, Alonso and Albon have four new softs (C5), Hülkenberg six, Verstappen five, Hamilton six, Piastri four. In Mexico the top six started the race with none. Every car has a new set for the start, a new set for a Safety Car and a new set to spare, and the one shortage on the grid is Norris's: both his hards (C3) were used on Friday and he has none left, which rules the medium (C4) → hard (C3) plan out for the pole-sitter.
The plan from pole is Friday's plan. In our field simulation on this grid and Friday's pace, Norris wins on soft (C5) → medium (C4) with the stop around lap 37, never in traffic; Antonelli, run as the car of interest, wins his own simulation on medium (C4) → soft (C5) with the stop around lap 33, which is the plan the medium's Friday pace suggests. The race at the front is therefore two different tyre plans from the front row, and the one that stops second has the information. Russell from seventh comes out third in the simulation but spends 37 laps behind other cars; the undercut on the Racing Bull and the Audi ahead of him is worth more to Mercedes than the tyre choice. Leclerc from fourteenth reaches sixth, Hamilton from twentieth tenth, Verstappen from twenty-first twelfth to fourteenth depending on the number of stops – the simulation lets him through the midfield slowly because its 0.8 s threshold is close to his 0.4–0.8 s advantage over it, and we think that undersells him against the Cadillacs and Aston Martins, which are more than a second off. Even allowing for that, the dry race for the three at the back is a fight for the top ten; the podium needs a Safety Car, and this track has had one in six of the last seven races.
Friday, two sessions, both dry. FP1 at 44.7 °C track, FP2 at 34.5 under cloud: Antonelli 1:10.763 on the soft (C5), Norris +0.25, Verstappen +0.26, Russell +0.27, Piastri +0.32, Hamilton +0.45, Leclerc +0.72. Sector 1 and sector 3 were within a tenth or two for the top six; sector 2, the infield that the FIA runs at 250 kW, is where Antonelli found 0.2–0.3 s on everybody, which is what a track that rewards rhythm over power was always going to do. The long runs are the other half of the session and they tell a different story: on our fuel- and age-corrected race pace, Hadjar 75.38 s, Verstappen +0.01, Norris +0.08, Piastri +0.10, Russell +0.10, Antonelli +0.14, Gasly +0.27, Bortoleto +0.42 – six cars from three teams within 0.14 s a lap, and neither Ferrari in the eight. Hamilton was ninth on the same measure, 0.42 s a lap off Hadjar, and Leclerc eleventh at 0.49, behind Gasly, Bortoleto and Lawson.
The plan, on Friday's numbers. Our planner on the season's race-day wear level, which is what has been right every weekend this autumn, gives one stop, soft (C5) → medium (C4) around lap 32 with a window of 28–37, and medium (C4) → hard (C3) around lap 39 equal to the tenth; the best two-stop, soft (C5) → medium (C4) → medium (C4) around laps 21 and 46, is 4.6 s slower. In Mexico the same comparison was 7 s and in Austin more. On Friday's raw wear instead – 0.136 s per lap of age on the soft (C5), 0.126 on the medium (C4), 0.184 on the hard (C3) – the two-stop wins by 37 s and the one-stop is a mistake. The break-even sits at roughly twice the season level, and Fridays this autumn have come in at 2.8 to 7 times the race. If the Friday-to-Sunday factor is 3 or more, as it has been every weekend since Singapore, one stop; if it is 2, two stops. That is a narrower margin than we have worked with all year, and it is why the second stop stays ready from lap 20 whatever the plan says.
The tyre that changed on Friday is the medium (C4). In Mexico the soft (C5) was 0.79 s a lap quicker than the medium (C4) in race trim; here the FP2 long-run levels have the medium (C4) at 76.01 s, the soft (C5) at 76.30 and the hard (C3) at 76.50, and the fitted base pace puts the medium (C4) 0.17 s ahead of the soft (C5) at zero age. A tyre that is quicker over a stint and wears no faster is the race tyre, and it makes the medium (C4) start the fast plan, not the overcut gamble it was in Mexico. The cars that start on the soft (C5) – which is anyone who qualifies in the top ten in the dry, with one new set at most – will need the undercut, and with a 21.8 s stop and a 0.76 s overtaking threshold they have it.
Round 20, the last of the triple-header, at the Autódromo José Carlos Pace, 4.309 km, 71 laps, anticlockwise, roughly 800 m above sea level. FP1 Friday 12:30 local, 16:30 in Sweden; FP2 16:00 local, 20:00; FP3 Saturday 11:30, 15:30; qualifying 15:00, 19:00; the race Sunday 14:00, 18:00 in Sweden, 71 laps or 120 minutes (FIA timetable v3). No sprint this year – last year's was a sprint weekend and that matters for every comparison below.
The tyres are the Mexico set: hard (C3), medium (C4), soft (C5), two, three and eight sets, one of hard (C3) or medium (C4) mandatory in a dry race, the soft (C5) as the Q3 tyre. That is one step softer than 2025, when Pirelli brought hard (C2), medium (C3) and soft (C4) and later wrote that the medium (C3) and soft (C4) "displayed significant degradation" and that two stops was the fastest race. We do not carry last year's wear numbers forward – different compounds, a sprint format, and a season of 2026 data that says Friday wear overstates Sunday by three to seven times. What we carry forward are the track's properties: a 21.8 s pit stop under green, the cheapest in the second half of the season, a Safety Car in six of the last seven races and a Virtual Safety Car in four of seven, and an overtaking threshold of 0.76 s per lap of pace advantage from our 2025 calibration, with 19.6 percent of close encounters ending in a pass – above the 2026 season median of 16.5 and a long way from Mexico's 8.8.
Those three numbers set the strategy before a wheel has turned. A cheap stop and a track where you can pass make the undercut work and make a second stop affordable; a Safety Car that comes more often than not makes the second stop free about half the time. Mexico was a race where track position was everything and the one-stop was the only answer. Interlagos is the other kind of track: position is worth something, not everything, and the car that is 0.8 s a lap faster gets through. The planner's answer on the season's race-day wear is still one stop – it always is until the tyres say otherwise – but the margin to two stops will be a few seconds, not the 7 s it was in Mexico, and a Safety Car in the first half of the race flips it.
The tyre question is the one Mexico left open. Antonelli ran 43 laps on one soft (C5) there with no measurable wear, Norris 39, Russell 33 on a used set; our library had given the compound 24 laps. We said then that we could not tell a Mexico result from a soft (C5) result, and Interlagos is the test: the same compound, 1,500 m lower, with longer corners under load and a surface that is new in two places. Pirelli's own prescriptions say what they expect. The minimum starting pressures are 27.0 psi front and 26.5 rear against 25.5 and 24.5 in Mexico, a step of 1.5 and 2.0 psi, and the pressure is Pirelli's lever against load: more load, more pressure, less footprint, more sliding at the limit. The expected running pressures of 28.0 and 27.5 are the highest we have logged this autumn. Read that as Pirelli expecting the soft (C5) to be worked here in a way it was not in Mexico, and plan for the 43-lap stint not to repeat.
Our pole estimate, for Marek Lubieniecki and anyone else keeping score: the 2026 cars have been slower than the 2025 cars over one lap all season, by a median of 3.6 percent in the spring and by less every race since – Monza +3.0, Singapore +2.5, Austin +1.65, Mexico +1.2. Last year's pole was Norris's 1:09.511, set on the Saturday of a sprint weekend. At +1.2 to +1.7 percent that gives 1:10.35 to 1:10.7; our central guess is 1:10.5. Lubieniecki's physics model says 1:10.5–1:11.2 with 1:10.8 as the point estimate, on 6.5 MJ of recharge and 5 percent less grip than his first run, which came out faster than 2025 and which he did not believe. FP2 has now spoken: 1:10.763 for Antonelli at 34.5 °C, and the season's FP2-to-pole ratio – pole 1.5 percent quicker than Friday's best – puts a dry pole at 1:09.7, three tenths under last year's. Both of us were too slow, and Lubieniecki's first simulation, the 1:09.3 he did not believe, was the closer one. The new asphalt has more grip than the gap to 2025 allowed for. If Saturday is wet, none of this applies.

FP2 at 34.5 °C track, 238 clean laps: soft (C5) 0.136 s per lap of age raw over 185 laps in 21 runs, medium (C4) 0.126 over 29 laps in 4 runs, hard (C3) 0.184 over 24 laps in 3 runs. The medium (C4) and hard (C3) samples are small and the hard (C3) runs were short and early. FP1 at 44.7 °C had the soft (C5) at 0.193 over 91 laps, so ten degrees of track took a third off the Friday wear, which is the temperature term doing what the library says it should. Nobody ran a stint longer than 13 laps; the longest were Russell's 13 on the soft (C5), Bortoleto's 10 on the soft (C5), Hamilton's 9 on the medium (C4), Verstappen's 8 on the medium (C4) and Norris's 6 on the hard (C3). Graining: not visible in the lap times of any run, which fits Pirelli's 2025 note that it was less than expected here. The stint-length question that Mexico raised – 43 laps on a soft (C5) against a library life of 24 – cannot be answered from 13-lap Friday runs; it will be answered on Sunday.
What the three compounds did last time out, in Mexico, is the only 2026 evidence we have for them together. Soft (C5) wear 0.039 s per lap of age raw and 0.036 after the fuel correction over 800 laps; medium (C4) 0.045 and 0.034 over 297; the hard (C3) unused by all 22 cars. Soft (C5) 0.79 s per lap faster than medium (C4) in race trim at equal age. Friday had said 0.108 for the soft (C5) and 0.206 for the medium (C4), overstatements of 2.8 and 4.6 times. That pattern – Friday grains, Sunday does not – has now held for five weekends, and the question for Friday here is only how large the factor is, not whether there is one.
What changes at Interlagos is the load. Mexico's thin air gave the cars a quarter less downforce and the corners were slow; here the air is only 8 percent thinner than at sea level, Turn 1 is a downhill braking zone into a long left that loads the front-right for two corners, the middle sector is a string of medium-speed corners taken in rhythm, and the lap ends with two long left-handers, Juncão and the climb to the line, that work the right side of the car under full throttle. Pirelli calls the lateral loads "broadly average" with an even split between the axles, which is fair, and it is still twice what Mexico asked. The soft (C5) that grained for a few laps in Mexico and then settled will be asked to do more, and the pressures above say Pirelli thinks so too.
The new asphalt is the unknown that Friday has to measure. Pirelli says Turns 3–5 and Turns 8–9 were resurfaced during the year: that is the exit of the Senna S onto the back straight and the entry to the middle sector. New asphalt in Mexico was 68 percent of the lap and the track went green-to-rubbered over the weekend with graining on Friday that was gone by Sunday. Here it is a few hundred metres, so the effect on wear should be smaller, but it sits at the two places where the car is hardest on the front-left – the exit of Turn 3 and the braking for Turn 8 – and a grip step between old and new surface is where a front locks.
The allocation arithmetic is the usual one and it is worth repeating because it decided the Mexico restart: eight softs (C5), one reserved for Q3 and returned afterwards, so the Q3 runners reach Sunday with one new soft (C5) at most and the Q1 and Q2 eliminations with two. Three mediums (C4), two hards (C3). If qualifying is wet, every team reaches Sunday with new softs (C5) in hand and the race opens up; if it is dry, the front of the grid starts on used softs (C5), and we measured twice this autumn – the soft (C4) at Austin, the soft (C5) in Mexico – that a scrubbed qualifying soft (C5) costs nothing in stint length and nothing measurable in pace. Hamilton did 28 laps on one in Mexico.
Our library gives the soft (C5) and medium (C4) life limits that Mexico showed to be too short in cool conditions, and we have not yet rebuilt the planner's cliff to depend on temperature. Until we have, the planner's one-stop windows will sit early; read them as "no later than", not as the optimum, and read the FP2 long runs for where the compounds actually fall away.

The FIA's power-unit document for the round came on Thursday evening (doc 3), and it is Mexico's rulebook on a different lap. Recharge 8.0 MJ per lap in the race, 8.5 with Overtake active, 6.5 in qualifying, 8.5 in practice – half a megajoule less than Mexico in the race and exactly the 6.5 MJ for qualifying that both Marek Lubieniecki and Toni Cuquerella had assumed in their simulations. The power-limited distance is 2,278 m, 53 percent of the 4,309 m lap, against 2,101 m in Mexico. The MGU-K curves are identical to Mexico's: 350 kW to 290 km/h on the base curve, tapering to nothing at 345; the Overtake curve holds 350 kW to 338 km/h; and the Alt 1 curve, a 250 kW cap up to 310 km/h, applies in the race from Turn 6 to Turn 12, lap distance 2,000–3,150 m – the whole infield, 1,150 m or 27 percent of the lap. That is the section where the lap is won in rhythm and not in power, so the cap costs little; what it does is keep the battery for the two places that matter. The Overtake detection line is at 3,250 m before Juncão (loop L17, marked "to be confirmed") and the activation line at 3,411 m at its exit (L18), so a car within a second gets 350 kW all the way up the 1.2 km climb through Turns 13 and 14 to the braking for Turn 1. Straight-mode zones are that climb and the back straight from Turn 3 to Turn 4. The base curve ends at 345 km/h and the cars will be close to that at the top of the hill in race trim; we measured in Mexico that nobody clipped on the main straight at 2,285 m, and here, 1,500 m lower with a longer full-throttle section, we expect to see the fade return, so the straight-speed measurement goes back on Friday's list. One reading of the document for the race: with 8.0 MJ and a lap where the battery is emptied twice, the leader who has to defend into Turn 1 every lap is the car spending the most energy, and the car behind with Overtake is the one with the most to spend. On a track with a 0.76 s overtaking threshold that is a real effect, not a Mexico footnote.
Friday's straight-speed measurement on the climb from Juncão to Turn 1, 930 m of full throttle: peak speeds 312–332 km/h, 700–800 m after the exit, and a fall of 4 to 9 km/h from the peak to the braking point over 14–20 laps per driver. Lawson 4, Verstappen, Norris, Leclerc, Colapinto, Gasly and Hadjar 5, Russell and Piastri 5.5, Antonelli 6, Hamilton 6.5, Bortoleto 8. That is a tighter spread than Mexico's 2–15 and a smaller fall than the base curve would produce if the cars were clipping – the taper from 290 km/h takes 100 kW away by 340 and nobody reached 340. So the fade is energy management, not the rulebook: everyone arrives at Turn 1 with a few km/h in hand that a car with Overtake behind them will not leave on the table. Full table in the Reddit follow-up.

Red Bull had the fastest car in Mexico on our race-pace measure, Verstappen 0.14 s a lap ahead of Hamilton and 0.36 ahead of Antonelli, with the cooling upgrades from the Thursday list on both cars. The form chart over the season, relative to Mercedes, has Ferrari, Red Bull and McLaren all within 0.2 s and Mercedes no longer the fastest car in any of the last four races. Interlagos is a medium-downforce track with a long climb, and the cars that were quick in Austin's fast middle sector – Ferrari, Red Bull – have the better claim on it than the cars that were quick in Mexico's slow one. The FIA's car-presentation list came on Friday: McLaren is the only team with new parts – a completely new front wing, a new rear wing and beam wing for more load, and revised floor furniture, four items – and the other ten teams declared nothing. For the third weekend running McLaren had the third- and fourth-fastest race car on Friday; the new wings are the answer to that, on a track where the wing level matters.

F1's official forecast, published 4 November: 30 percent chance of rain Friday at 25 °C, 50 percent Saturday at 27 °C with heavy rain possible after 14:00 local – an hour before qualifying – and a warning of localised flooding and standing water, 40 percent Sunday at 19 °C. Friday's track could reach 45 °C in FP2; Sunday's is likely to be nearer 30. Interlagos in November is a track where the weather is a strategy variable, not a backdrop: 2024's race was run in the wet after a Sunday-morning qualifying, 2016 was a wet race with two red flags and several Safety Car periods, and Pirelli's preview ends with the sentence that the intermediate and the full wet "could come into play in Brazil once again this year". The temperature swings are the other half of it. A dry Sunday after a wet Saturday means a green track and cooler asphalt, which is exactly the Friday-to-Sunday change that has made Friday's wear numbers useless all autumn. Friday's long runs will be run at whatever temperature Friday gives; our rule asks for Sunday within 6 °C of that before we trust them, and we broke the rule in Mexico and were right to. Here the rule may be the least of it. Friday itself showed the swing: FP1 at 44.7 °C, FP2 at 34.5 under cloud, and the soft (C5) wore a third less in the cooler session.

Pit loss 21.8 s under green from 33 stops in 2025, with F1.com's own figure at 21.83 – the two agree to a tenth. Under a Virtual Safety Car we measured 5.7 s and under a Safety Car 7.1 s; Cuquerella's simulation says 22.9 and 14.8 for green and neutralised, so the sources differ on how cheap a neutralised stop is, and ours is the measured one. A pit entry at the end of the lap, down the hill after Juncão, a short pit lane and a pit exit that rejoins before Turn 1: this is one of the cheapest stops of the year and it makes the undercut a real weapon. The run from pole to the Turn 1 braking point is 202 m, one of the shortest on the calendar, which puts a premium on the launch rather than the tow. Overtakes completed in 2025, by F1.com's count, were 96. Our own count last year had 271 close encounters and 53 passes, one in five, the main places being the end of the pit straight into Turn 1 and the end of the back straight into Turn 4. The Race Director's notes list the circuit changes: partial resurfacing from the exit of Turn 3 to the exit of Turn 5 and from the exit of Turn 8 to the exit of Turn 9; grasscrete replaced by concrete before the apexes of Turns 1, 4 and 12; the apex kerbs at Turns 2, 8 and 10 lowered; new blue track-limit lines at most corners and an asphalt patch at the exit of Turn 13. The Safety Car resumption point is before Turn 8, and under a red flag the cars stop in the fast lane. The maximum time between the Safety Car lines, the slow-lap rule, is announced after FP2.

Mercedes arrive as drivers' champions and can take the constructors' title this weekend by outscoring Ferrari by three points, which would give them both titles for the first time since 2020. Russell took a sixth engine in FP1 and starts five places back from wherever he qualifies. Their race pace has slipped from first to third over the autumn on our measure, which did not matter in Mexico because Antonelli only needed to finish. Ferrari have the fight for second in the drivers' championship on their hands, Hamilton 259 to Verstappen's 261 with Russell on 250 and Leclerc on 245, and a lap-1 Turn 4 in Mexico between their own cars that Vasseur called "a lack of respect"; Interlagos's Turn 1 is a better place to do it again. Red Bull bring the fastest race car of the last two weekends and a driver who has won three of the last four races. McLaren won here in 2024 and 2025 from pole both times with Norris, on a different car under different rules; they bring a Piastri car that had its engine and cooling package changed before the Mexico race and a form chart that points the wrong way. Racing Bulls, Alpine and Audi are the midfield fight, 108, 69 and 27 points, and Haas on 31 run their second new driver in two weekends: Rafael Câmara, who replaces Ocon for his home race ahead of his full-time seat next year, in the car Hirakawa took to a point in Mexico. He will be the only driver on the grid racing in front of his own crowd.

Race pace in Mexico, fuel- and age-corrected: Verstappen 80.20 s, Hamilton +0.14, Leclerc +0.32, Antonelli +0.36, Russell +0.55, Norris +0.69, Piastri +1.12. Three cars within 0.4 s and a fourth within 0.7 is the closest front of the field we have measured this season, and at a track with a 0.76 s overtaking threshold that means the order on the road will be decided on Sunday, not on Saturday. Hamilton led from pole for three corners in Mexico and finished 0.8 s behind after 25 laps within 1.2 s of the leader; here that car gets a run up the hill into Turn 1 every lap. Lap 1 matters less than in Mexico – 202 m to the braking point, so the grid order mostly survives to Turn 1 – and the race matters more. The launch measurement from Mexico was lost to a timing-archive fault; we will have it again here, and with 202 m to the braking point the first three seconds decide the first corner. Qualifying in the wet was McLaren's and Mercedes' and then a long gap: the Red Bull that had been quickest on Friday's long runs was seventh with Hadjar and out in Q1 with Verstappen, who never found the lap on intermediates. Câmara's thirteenth on debut, 0.5 s behind Bearman in Q2, is the quiet result of the day.

Standings after 19 of 23 rounds: Antonelli 380, champion; Verstappen 261, Hamilton 259, Russell 250, Leclerc 245, Norris 221, Piastri 144, Hadjar 107, Lawson 83. Constructors: Mercedes 630, Ferrari 504, Red Bull 382, McLaren 365, Racing Bulls 108, Alpine 69, Haas 31, Audi 27, Williams 12, Aston Martin 7, Cadillac 0. One hundred points remain for a driver, 172 for a team; Mercedes clinch with a three-point gain on Ferrari here. Grid penalties this weekend: Russell five places (sixth ICE, Friday), Leclerc ten (fifth ICE, Saturday), Verstappen 30 (fifth ICE, turbo and exhaust), Pérez 30 (fifth ICE and turbo, fourth MGU-K), Hamilton 65 (a complete new power unit after qualifying). Three cars start from the back in qualifying order: Hamilton, Verstappen, Pérez; the official grid is published. Câmara a reprimand for impeding Leclerc in FP2, Bearman a reprimand for reversing in the pit lane. Covers-on time is published; the final grid follows. Câmara races with number 51. The stewards for the weekend are Nish Shetty, Loïc Bacquelaine, Vitantonio Liuzzi and Felipe Maluhy. Seven FIA documents are out as we write: the competition visa, Pirelli's prescriptions, the timetable, the Race Director's notes, the power-unit information, the circuit maps and the car-display procedure; the car-presentation submissions follow on Friday.

Rain, first and last: a wet Saturday hands out new softs (C5) to everyone and a wet Sunday makes every number above a footnote. Second, the soft (C5) under real load. If the compound that did not wear in Mexico grains on the new asphalt here, the race is two stops as it was in 2025 and the medium (C4) start is the plan, not the exception. Third, our own planner: its life limits were fifteen to twenty laps too short in Mexico and we have not rebuilt them yet, so its stop windows this weekend are early until the FP2 long runs correct them. And fourth, the Safety Car: with six in seven races, the plan that wins here is usually the one that has a cheap stop ready when the car in front needs an expensive one. Fifth, after Saturday: a dry race run as a Grand Prix of recovery drives, where the cars we have at the back of the simulation pass the midfield faster than our 0.8 s threshold allows and arrive at the top-six train by half distance, with new softs (C5) to use against cars that started on them.
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.
