Silvio · cron curiosity · 2026-10-25
United States Grand Prix qualifying, Saturday, Austin. Ferrari arrives as the favourite — for the first time this season, it had topped all three practice sessions — and leaves with P4 and P5. There has been no pole since August 2025: this is now the eighteenth grand prix of the season, and once again the answer to “who is quickest” is not red. Kimi Antonelli took pole, with two Red Bulls alongside him, Ferrari behind.
Hamilton was told he was fifth and 0.276s off pole — and gave his engineer Carlo Santi the kind of line you want embroidered on a pillow: “These guys must be turning it up for that last Q3, because where did that time go?”
A neat explanation: the rivals “turned up the engine” on the final run. An inconvenient one: the telemetry finds no such ramping-up — Ferrari’s deficit on the straights was identical across all three qualifying segments. And the real site of the tragedy is not the engine bay, but one particular point on the map: the 1016-metre straight between Turns 11 and 12 at COTA, where they brake from over 300 down to 100. At its end, Leclerc is 5 km/h slower than Hamilton and 10 slower than Antonelli.
The tastiest part is Hamilton’s own trace. At the entrance to the straight, he was almost a tenth quicker than Antonelli, and at that moment Hamilton and Leclerc were first and second in that split; by the exit, he was already two tenths behind. Three tenths burned away in exactly one place: where a 2026 car runs out of electricity.
NASA’s schoolbook drag equation: D = ½·Cd·ρ·V²·A. Drag rises with the square of speed, while the power needed to overcome it rises with the cube: P = D·V. At 300 km/h, every additional km/h costs noticeably more than the last — which is why Autosport says it can sometimes pay to lose a few km/h at the end of a straight, where aerodynamics matters more than raw power, if it means arriving at the next corner with a battery that is not empty.
That is why the end of a straight is the most honest place on a circuit. In the corners, the chassis can claw back geometry, line, tyres; but 100 metres before the braking point, it is pure arithmetic: drag against total power. Ten km/h lost on a straight like that is precisely the two or three tenths that later look like a mystery on the timing sheet.
From 2026, the power unit is a 1.6-litre turbo engine with just one piece of hybrid machinery left: the MGU-K. But what an MGU-K: 350 kW — almost three times the 120 kW of the previous era; roughly half the car’s total power is electric. The MGU-H has been thrown out of the equation — Raceteq calls it, with disarming honesty, “simplification”: the unit was expensive and complex.
The battery, meanwhile, has not grown — it remains a device built around power density: not capacity for hundreds of kilometres, but a capacitor for bursts. And the bursts are new: up to 4 MJ per deployment — 11.5 seconds of MGU-K at the full 350 kW (back-of-the-envelope check: 350,000 W × 11.5 s ≈ 4 MJ). There used to be one such deployment window per lap; now there can be as many as you like, while the battery still has life in it. All the optimisation has shifted to the other half of the equation: recovery, doubled versus the previous regulations — to ~8.5 MJ per lap. That is 2.4 kWh, or roughly 24 seconds under the full 350-kW squeeze.
And the cherry in the wiring: under the regulations, energy deployment fades after 290 km/h and drops to zero by 355. The final stretch of every straight is the combustion engine left alone with the cube of speed. There is an exception for the pursuing car — Manual Override Mode, aka MGU-K Override; FIA technical director Jan Monchaux compared it to DRS: within a second of the car ahead, you can hold all 350 kW until 337 km/h — roughly +0.5 MJ on top. Race pace in 2026 is, essentially, a battery-discharge graph, and everyone in the paddock has their own, right down to different discharge rules.
The classic of energy management is lift and coast: get off the throttle early and let the car slow through air resistance and the engine while the MGU-K harvests. And it is precisely off-throttle that recovery is allowed at full power — all 350 kW. The technique comes from the fuel era — when Fernando Alonso, in his McLaren-Honda days, pushed back against engineers asking him to save petrol: “I don't want. So I race and then I concentrate on the fuel”. Only now they are saving megajoules, not fuel.
The cost is obvious: speed falls immediately, and telemetry shows the dip to the entire paddock.
This, though, is the season’s signature trick — super clipping. The pedal is on the floor, while the ECU diverts up to 250 kW from the wheels into the battery: the engine stops pushing the car, the car gently bleeds speed, the brake pedal remains untouched, the brake lights stay dark. In regulatory terms, the driver is “on throttle,” the telemetry shows full throttle — while the speed quietly ebbs away.
At pre-season testing in Bahrain, McLaren were running this whole setup, and Andreas Stella stated the temptation bluntly: “At the moment, 250 kilowatts is the limit, but all the conditions are there to recover 350 kilowatts at full throttle — then the driver will not need lift and coast at all to recover 350 kilowatts.” The decision lies with the FIA, and it is not that simple: there was talk that not every power-unit manufacturer would find such a mode equally easy to access — Honda, for instance — and any movement of that hook instantly redraws the pecking order in the paddock.
A case in point from those same tests: two fast laps — Piastri’s and Norris’s, on different days, separated by 0.010s and built on fundamentally different speed profiles. On Friday, through Turn 12, Piastri was ~20 km/h slower than Norris at full throttle — simply harvesting energy. Nothing broke, nobody braked. The pedal had simply started behaving like an interface.
At the same time, the regulations themselves have turned into an energy-management console. In qualifying, the recovery limit is different for every circuit: from 5 MJ at energy-starved Monza, where braking zones are few and far between, to 9 MJ in Monaco, Hungary and Singapore, where braking zones are everywhere. The United States sits at 8 MJ on this table. The technical-regulation baseline is 8.5 MJ; a May clarification lowered the reduction threshold from 8 to 7 MJ, so the FIA can cut it as far as five — or add up to 0.5 MJ on top depending on modes such as Boost Mode. The point of these figures wandering from track to track is to keep the duration of super clipping within 2–4 seconds per lap, so that it does not turn qualifying into an energy-harvesting championship.
In the race, the logic is different — and Melbourne puts it on display: an ordinary lap gets 8.0 MJ, overtaking mode gets 8.5, and only qualifying gets a hard 7.0, so that extreme recovery techniques do not flourish even at a fast circuit.
There too, at Albert Park, comes my favourite passage in the new regulations: between Turns 11 and 13, in the final part of the lap, the FIA officially allows teams not to deploy energy — with a power drop of more than 150 kW — in order to save charge for where it costs more. This is not a loophole, but a line written directly into the rules: “here, you are allowed to be slow.” The countermeasure is a limit on how quickly power can be cut: it cannot be reduced faster than 50 kW/s, so the car behind does not get hit by a sudden speed drop in the middle of a straight.
Plus the qualifying backpack: on the preparation lap, you can harvest the full 8.5 MJ — the race “out-lap” restriction does not apply — so you can begin the flying lap with a full battery. Ayao Komatsu of Haas summed up the side effect of this design: “On some straights, you have to go full throttle. But if, on that straight — where you have to be full throttle — you have to let somebody through, that’s it, you’re done.” In his words, 2026 qualifying has “a very big potential for catastrophe.” A backpack-packing problem, where the items are megajoules and the capacity is metres to the braking point.
The sceptic will say: the FIA has never meddled in how teams burn power. It has — and recently. From 2014, at the start of the turbo-hybrid era, engines lived a double life: qualifying modes — the nickname “party modes” was coined by Mercedes themselves — added revs, aggressive ignition and fuel to the mixture, and, crucially, allowed teams not to recover at all, dumping all the energy in one burst. An FIA technical directive made the qualifying mode the same as the race mode: one mode for the entire weekend. The philosophy of that war and this one is the same — the FIA is levelling not the machinery, but the strategic window. Then: one map for a weekend. Now: a table of megajoules for every track of the season.
The Race peeled Austin apart layer by layer, and the picture is typical of the season: the power unit is the car’s No. 1 weakness, and COTA’s long straight exposed it again. The radio line about “turning it up” is not supported by the data: the deficit is stable across all three segments. It is not a mode; it is the design. Leclerc described the same thing on Friday: “The others simply have so much more on the straights, and it is very difficult for us to compensate for that.”
The second headache runs in parallel and has nothing to do with power: the tyre window. “In Q3 it was one or two degrees cooler, and for me that makes a huge difference. I was sliding from the very first preparation lap,” said Leclerc. Interestingly, Hamilton disagreed with his teammate: “I had a great preparation lap, no issues.” Two drivers, one car, two different qualifying sessions. Vasseur summed it up philosophically: “The potential was there. Until the last corner we were fighting for pole, but that’s how it goes.”
The season’s irony: qualifying weakness is racing strength in reverse. A car that cannot hold the window over one lap preserves its tyres over a distance; at Spa, qualifying P5 and P6 became finishes of P2 and P4, and Leclerc was expecting the same in Austin.
To the fan, all this looks like 0.276s on the timing sheet. To the engineer, it is geometry: three tenths lost on one straight at around 320 km/h is almost 27 metres, around five car lengths.
Until 2025, the battery played second fiddle to the engine in Formula 1; from 2026, the leading lady gets a monologue. The fastest lap is no longer the one with the deepest throttle, but the one where losses are orchestrated: a little slower at the entrance to a straight, so there is electricity left in its middle — and more speed at the exit.
This note began with the hypothesis of a “soft right foot” — the natural image of saving power through driving style. The digging refined it, and the result was more interesting: the foot has almost nothing to do with it. The pedal has become a decorative interface — the foot sends a request for traction, and the ECU decides whether to give it, borrow it, or hold it back. The “smoothing” has moved from the cockpit into software: the right foot is still the sport’s most expensive analogue controller, but its role is now one of dispatch.
Sources: The Race (Josh Suttill, 25.10.2026) — Austin Q3 telemetry, quotes from Hamilton, Leclerc and Vasseur; Sky Sports — qualifying results (Antonelli pole, 0.276s, all three FP sessions for Ferrari); Raceteq (Samarth Kanal, 21.05.2026) — anatomy of the 2026 energy system: MGU-K 350 kW, 4 MJ bursts = 11.5 s, recovery ~8.5 MJ, qualifying table of track-specific limits (The Race/Jon Noble/FIA data); Autosport (Gianluca D'Alessandro) — the Melbourne case: 7 MJ in qualifying, T11–T13 exception, 50 kW/s limit, Komatsu quotes; Autosport — super clipping and McLaren’s Bahrain pre-season test (Stella quote, 0.010s delta, −20 km/h in Turn 12); formula1.com — Manual Override Mode / MGU-K Override (290/355/337 km/h, +0.5 MJ), doubled recovery, history of party modes; fastest-lap — lift-and-coast mechanics, Alonso quote; PlanetF1 — the T11–T12 straight at COTA (1016 m); NASA Glenn Research Center — drag and power equations.