Mercedes has won all three races in 2026. Russell took Australia. Antonelli won China and Japan. At 21 years old, Antonelli leads the World Championship after three rounds, and the W16 looks like the car to beat for the entire season.
Ferrari is second in the constructors' standings. Leclerc has two podiums. The SF-26 is clearly the second-fastest car. And none of that matters if you're Lewis Hamilton, who finished P6 at Suzuka and called it "pretty terrible."
Three races into a regulation reset and the narrative is already forming: Mercedes dominance, Ferrari struggling, Hamilton past it. That narrative is wrong. Or at least, it's incomplete. And Singapore in October is where it falls apart.
Problem
The 2026 regulations were supposed to reset the field. New active aerodynamics, new power units, MGU-H deleted, sustainable fuel, lighter cars. Every team starting from scratch. The theory was parity -- or at least genuine uncertainty about who would come out on top.
Three races in, it's a Mercedes benefit. The W16 is the benchmark by a margin that's uncomfortable for everyone else. Antonelli's wins in China and Japan weren't close -- 5.5 seconds at Shanghai, and a dominant lights-to-flag at Suzuka. Russell's Australia win was more contested but never truly threatened.
Ferrari's SF-26 is quick. Leclerc proved that with P3 in Australia and P3 in Japan. But "quick" and "winning" are different problems. The gap to Mercedes is real, it's consistent, and it shows up in the same place every race.
Hamilton's situation is worse. P6 at Suzuka, behind his teammate, behind a McLaren, behind cars that have no business finishing ahead of a Ferrari. His radio message was blunt: "I had a real lack of power through the second stint. I couldn't keep up."
He wasn't wrong. But the problem isn't Hamilton. The problem is physics.
Constraints
The SF-26's core issue is electrical de-rating. And understanding why requires understanding what changed in 2026.
The new power units triple the MGU-K output compared to the previous generation. The Motor Generator Unit - Kinetic now delivers roughly 470 horsepower of electrical power, up from around 160. That's not a tuning change -- it's a fundamental shift in how F1 cars make speed. The electrical system isn't supplementary anymore. It's nearly half the total power output.
Ferrari's hybrid system runs out of electrical deployment at the end of long straights. The battery drains faster than the energy recovery systems can replenish it, and the car loses electrical power exactly when it needs it most -- at the end of straights where the driver is relying on full electrical boost to maintain top speed against cars with better energy management.
At Suzuka, this is fatal. The back straight is 600+ meters. Turns 1-2 demand full power. Spoon through 130R is a sustained high-speed section where electrical deployment is critical. The SF-26 was losing time in the speed traps AND through the high-speed corners where electrical boost supplements the internal combustion engine.
Hamilton felt it in the second stint because tire degradation compounds the problem. As rear grip drops, the MGU-K harvests less energy under braking (less braking force = less kinetic energy to recover), which means less deployment available on the next straight, which means more de-rating, which means slower lap times. It's a negative feedback loop.
This is a calibration problem, not an architecture problem. The hardware works. The energy recovery systems function. The battery capacity is within regulations. What's failing is the software that decides when to deploy, how much to deploy, and how aggressively to harvest. It's an optimization problem -- and optimization problems get solved with data and development time.
System
The SF-26 deserves a closer look than the points table suggests, because the car is genuinely innovative. Ferrari didn't iterate on the previous generation. They rethought the architecture.
Active aerodynamics. The rear wing rotates -- physically flips upside down when switching from Corner Mode to Straight Mode. In Corner Mode, the wing generates maximum downforce. In Straight Mode, the wing inverts to present a flat, low-drag profile. The transition is managed by FIA-predefined zones on each circuit, not by the driver. This is the most visible change in 2026, and Ferrari's implementation is one of the most aggressive on the grid.
Pushrod suspension, front and rear. A new adoption for Ferrari, aligning with the approach Mercedes and Red Bull have used. Pushrod gives better packaging for the underfloor aerodynamics and allows tighter control of ride height -- critical when the ground effect floor generates a significant percentage of total downforce.
Exhaust-blown FTM flap. This is clever. Ferrari routes exhaust gases to assist the Front Torque Module's energy recovery, active even at lower gears when traditional recovery systems are less effective. It's an efficiency gain that compounds across a full race distance.
Diffuser hole routing. A significant opening in the lower bodywork channels airflow from external surfaces to the diffuser's inner wall, energizing the ground effect and increasing rear downforce without adding drag from wings.
The car IS fast. The chassis works. The aerodynamics are innovative and effective. Leclerc's two podiums in three races confirm that the SF-26 has the pace to fight at the front. The deficit is energy management software -- the part of the car that improves fastest during a season because it requires code changes, not manufacturing.
Execution
Why Singapore Fixes This
Marina Bay is a low-speed street circuit. 23 corners. The longest straight is roughly 400 meters -- short by F1 standards, and critically shorter than Suzuka's back straight where Ferrari's de-rating was most exposed.
Here's what changes at a circuit like Singapore:
Short straights mean less deployment demand. The battery doesn't drain as far because no straight is long enough to fully deplete the electrical reserves. Ferrari's de-rating problem -- running out of electrical power at the END of straights -- barely manifests when the straights are short. The car reaches the next braking zone before the battery hits its critical threshold.
23 corners mean more harvesting opportunities. Every braking zone is a chance for the MGU-K to recover energy. More corners, more braking, more energy harvested. At a circuit with 23 braking events per lap versus Suzuka's handful of heavy braking zones, Ferrari's energy balance tips positive.
Traction zones favor chassis quality. Singapore is about corner exit, about putting power down in second and third gear, about threading a car through barriers with millimeter precision. This is where a good chassis -- and Ferrari's chassis IS good -- matters more than peak power delivery. The SF-26's pushrod suspension and ground effect diffuser work hardest in exactly these conditions.
The Sprint format adds chaos. Singapore hosts the Sprint format for the first time in 2026. One practice session at a street circuit where grip evolves dramatically across the weekend. Less preparation favors the more adaptable team, the better driver, the car with a wider operating window. Sprint races with no pit strategy are pure pace events -- 30 minutes where the faster car wins, but "faster" at Singapore means something different than "faster" at Suzuka.
And then there's Hamilton. Hamilton at Suzuka was bad. Hamilton at a street circuit is a different proposition entirely. His record at Marina Bay is strong. His record at street circuits generally -- Monaco, Baku, Jeddah -- is among the best in the sport's history. Street circuits reward experience, precision, and racecraft. Hamilton has more of all three than almost anyone on the grid.
The Development Timeline
By Singapore in October, Ferrari will have had six months of hybrid calibration updates. Six months of data from real race conditions. Six months of software iterations on the energy management system that is the SF-26's only meaningful weakness.
Context matters here. The 2022 regulation reset looked like Ferrari dominance after three races -- Leclerc won two of the first three, led the championship comfortably. Red Bull won the title. The team that starts strongest in a new regulation era almost never finishes strongest. Launch performance is an early snapshot. Development rate is the season-long story.
Mercedes may well be the best team in 2026. The W16 may be the best car. But the gap between "best" and "second best" in F1 compresses through the season as teams converge on optimal solutions. Ferrari's deficit is in a domain -- software and calibration -- where convergence happens fastest.
Outcome
Ferrari's season is not defined by the first three races. The SF-26's architecture is sound. The aerodynamic package is innovative and effective. The energy management software is the weak link, and software is the most improvable part of a modern F1 car.
By October, the realistic expectation is a Ferrari that's closer to Mercedes. Possibly level in qualifying pace at circuits that suit its characteristics. Singapore is that circuit -- low speed, high cornering count, short straights, street circuit chaos.
Will Ferrari win Singapore? Maybe not. But will the SF-26 be meaningfully more competitive there than it was at Suzuka? Almost certainly. The physics of the circuit layout directly mitigate the car's primary weakness.
And Hamilton at a street circuit, at night, with six months of adjustment to the car and the team -- that's a different driver than the one who finished P6 in Japan. The potential for a strong result is real, not wishful.
Lessons
Circuit characteristics matter more than championship standings. Ferrari's P6 at Suzuka doesn't predict their Singapore performance. Different tracks expose different weaknesses. Marina Bay is the best possible circuit for a car with Ferrari's specific deficit -- short straights, abundant harvesting zones, and a premium on chassis quality over raw power delivery.
Regulation resets take time to settle. Three races into a new era is too early to call the season. Every major regulation change in F1 history has seen the early leader challenged as development converges. Mercedes started 2022 building a car with severe porpoising. They ended 2022 winning races. Extrapolating from March to October is poor analysis.
This is a calibration problem, not an architecture problem. The distinction matters. An architecture problem -- fundamentally wrong suspension geometry, inefficient aerodynamic concept, insufficient cooling -- requires redesign. A calibration problem requires data, simulation time, and software updates. Ferrari has all three, and the season is 24 races long.
Attend the race that suits your team. If I were a Mercedes fan, I'd go to any race and expect to celebrate. As a Ferrari supporter, I'm picking the circuit most likely to produce the result I want to see. That's not hope -- that's track analysis applied to travel planning. Singapore's layout is the best argument Ferrari has in 2026, and I'll be there to see if the argument holds.