The Hook: This morning in the Formula 1 digest, one phrase from Toto Wolff popped up that an engineer cannot pass by: "Six out of nine wins — better that than slow and reliable." The team is consciously chasing speed, sacrificing reliability, and has no intention of changing approach. I started thinking: where in the history of technology has humanity already made this exact choice — between "a reliable but slow path to the goal" and "a fast but risky blind leap"? The answer came instantly: the problem of determining longitude at sea — one of the most expensive and most famous engineering problems of the early modern era. And in it, the same fork in the road as Wolff's: European maritime nations spent decades paying for a "slow and reliable" solution (lunar distance tables, observatories, nautical almanacs), while John Harrison in 1761 rolled out "risky pocket watches" — and immediately got the prize. And then there was Taqi al-Din's Istanbul Observatory (1577–1580) — the Ottoman equivalent of a "fast leap," destroyed three years before it could deliver results. Three eras, the same dilemma.
Investigation:
Quote from the morning digest, verbatim: "How many have we won so far, six out of eight? I've already lost count. I would always prefer these results over having a slow but reliable car. Six out of nine? Well, I'd prefer it to be nine out of nine."
Key point: Wolff is not claiming that "slow and reliable" is worse on average. He's saying that a result distribution with greater variance (6 wins and 3 DNFs on a fast car) is preferable to a distribution with small variance (8 finishes in 8th-9th place on a reliable car), because world champion is determined not by expected value, but by maximum. This is not a question of averages — it's a question of distribution shape. And this is a purely engineering dilemma of "risk-adjusted performance."
After the catastrophic fleet disaster of 1707 off the Isles of Scilly (4 ships, ~2,000 men, longitude error), the British Parliament in 1714 passed the Longitude Act and established a prize of £20,000 (today ≈ £3 million) for a method giving accuracy of ½ degree (= 30 nautical miles) on the route from Britain to the West Indies.
The "slow and reliable" path is the lunar distance method: you measure the angle between the Moon and a bright star/Sun, look it up in the almanac to see where this pair should be in the sky at that moment — the difference gives you Greenwich time. Accurate, but requires a qualified observer, clear sky, sharp eye, and 30 minutes of calculation in a storm. The British Admiralty in 1765 launched the Nautical Almanac — and the lunar method became "official," "reliable," "infrastructural." It worked for another 100 years until chronometers became affordable.
The "fast and risky" path is Harrison's H4 chronometer. A self-taught carpenter from Yorkshire spent 30 years of his life and four versions of the device. H1, H2, H3 were gigantic marine clocks the size of a small cabinet — they failed sea trials. H4, completed in 1759, was a completely different solution: a pocket watch 13 cm in diameter. The Longitude Board commission was shocked: "These are just ordinary watches!" Harrison replied that compactness was the main innovation: the smaller the mechanism, the less it suffers from the ship's pitching. In March 1761, H4 passed an 81-day trial from Portsmouth → Jamaica with an error of 5 seconds (= 1⅓ nautical miles of longitude) — that is, with accuracy 22 times better than the Act required.
This was the engineering "6 out of 9": risk your entire career, all your money, all your reputation on one architectural hypothesis ("small watches instead of a huge marine chronometer"), and either immediately become the winner, or — yes, become a laughingstock.
Here the story becomes truly bitter. In 1577, Sultan Murad III ordered the construction of an observatory in Istanbul to rival Ulugh Beg's in Samarkand. The chief astronomer appointed was Taqi al-Din ibn Ma'ruf — a mathematician, optical engineer, constructor of mechanical clocks and spring drives, capable, according to several sources, of projecting an image of the celestial sphere onto the inner wall of the dome through a system of lenses (essentially a proto-planetarium).
The observatory operated for three years (1577–1580). During this time, Taqi al-Din managed to:
On January 22, 1580, the observatory was destroyed by janissaries on religious grounds — "because the predictions of astrologers caused plague, and the observatory breeds insolent thoughts about fate." The Sultan did not resist.
This is a classic "fast but unprotected leap": technologically the solution was at the cutting edge of progress (optics, mechanical clocks, ephemerides), but there was no "chassis" — institutional, political, religious protection for the project. Speed without armor = death.
| Parameter | Wolff / Mercedes-AMG (2026) | Harrison H4 (1761) | Taqi al-Din's Observatory (1577–1580) |
|---|---|---|---|
| "Fast" path | Aggressive aerodynamics, risk of DNFs | Pocket chronometer, rejection of the "correct" path of marine clocks | New optics, Jupiter's moons, mechanical clocks |
| "Slow" path | Improve reliability to 9/9 | Lunar distances + Nautical Almanac | Gradual tables, traditional muwaqqit-khanas |
| What won | 6/9 in 2026, championship dynamics | H4 passed trials 22x better than required | Perished at the stage of becoming |
| Project protection | Sponsor, FIA, shareholders | Crown, but 30 years of doubt | None: janissaries and fatwa |
| Long-term result | 2026 Drivers' Championship — TBD | Chronometer became an industry; after 100 years displaced lunar distances | Loss; in Europe — Tycho Brahe, Kepler, Galileo, Halley, Harrison |
The main thing uniting all three cases is not the question of "fast or reliable." It's the question of what exactly you are optimizing: average over distance or maximum result over the long perspective. Wolff is optimizing championship probability (probability of title with current points pool). Harrison optimized error of 1.25 miles on a transatlantic crossing — and it doesn't matter how elegant the mechanism looks. Taqi al-Din optimized ephemeris accuracy, but did not optimize the institutional resilience of the observatory to a janissary raid.
When we in systems development choose "better stable, proven, 20% slower framework" vs. "fast, trendy, 20% faster, but with a history of major releases breaking prod" — we are exactly at the same fork. Both the Harrison strategy ("6 out of 9 and 3 failures, but the 6 are breakthroughs"), and the Taqi al-Din strategy ("leap without armor — death at takeoff"), and the Wolff strategy ("more DNFs, but higher speed") — these are all risk-return profiles with different tail shapes.
For an engineer this means: when choosing architecture, don't ask "what's more reliable," but "what optimizes our actual KPI, and what's the tail cost of that". If the KPI is championship title, Wolff is right. If the KPI is team survival over 30 years (like the British Admiralty), then lunar distances and the Nautical Almanac are the right path. If the KPI is breakthrough, and there's an "armored sponsor" — take Harrison's risk. If there's no "armor" — build protection first, or you'll repeat the fate of Istanbul 1580.
Conclusions:
Subjectively: the longitude story is the best illustration that in engineering "reliability" and "speed" are not a scale, but two different dimensions, and they cannot be reduced to one. Harrison's H4 was reliable in its own way (accuracy, repeatability), but risky in another way (rejection of "proper" engineering convention, compactness that in the 18th century was perceived as technical frivolity). Lunar distances were reliable institutionally (Admiralty, almanac, trained navigators), but unreliable operationally (cloud cover, observer error, 30 minutes of calculation). Wolff today is betting that for his KPI (championship crown), distribution variance matters more than its mean — and empirically he's 6/9, meaning he's probably right.
What really got me was the Istanbul story. When technological breakthrough runs into a political/religious storm and takes the work with it, this is the scariest failure mode of engineering work. Taqi al-Din did not "lose" the competition to European astronomers — he was burned, and Europe for 30 years after that had not a single working observatory of that class. Lesson for today: architectural work without institutional buy-in is a house of cards where the first political draft destroys everything. Harrison's H4 was lucky — it was protected by the crown and the £20,000 prize. Mercedes today has sponsors and the FIA. And does your startup, Peter, have such protection?
And one more observation I cannot help but voice, though it's on the edge: this entire chain is a logical argument that in engineering there is no "safe" speed and no "fast" reliability. There are only different risk profiles, and engineering mastery is honestly choosing which profile fits the specific KPI, rather than trying to sit on both chairs. A team that is "both fast and reliable" is either already champion, or lying to itself. Like Hamilton at Ferrari in 2025: "there was no trust because I was performing poorly." Or, as Taqi al-Din said before the destruction of the observatory — "we cannot simultaneously please heaven and the sultan."