The 26-month clock that governs everything
Earth and Mars line up for a cheap transfer once every 779.9 days. A six-week hardware delay does not cost six weeks; it costs more than two years.
Earth laps Mars every 779.9 days. That is the synodic period — twenty-five and a half months, rounded up to twenty-six — and it is the hardest constraint in this whole era, unmoved by funding, engineering or political will. Everything else in this era sits underneath it.
A minimum-energy transfer between the two orbits, the Hohmann ellipse, costs roughly 3.6 km/s of velocity change on top of low Earth orbit and takes about 259 days. Real missions trade a little extra energy for a shorter ride: Perseverance flew in 203 days, InSight 205, Curiosity 253, China's Tianwen-1 202. Arrive and you pay again — something near 2 km/s to capture into Mars orbit, unless you are willing to hit the atmosphere and let it do the braking, which every successful lander has done. Try to leave outside the window and the energy required climbs fast. A few months off and no heavy vehicle can afford it at all.
The windows themselves drift later each cycle. The current one is open — Japan's MMX leaves Tanegashima for Phobos on 20 October 2026. The next runs from about December 2028 into January 2029, the slot China's Tianwen-3 sample-return mission is built around. Then 2031. Nothing in between. The edges are soft rather than hard: you can leave early and pay in cruise time, which is the bargain ESA struck for Rosalind Franklin — October 2028, two years in transit, to land outside the dust season.
For a crew the arithmetic gets worse, because you have to come home and the window home has its own schedule. NASA's Design Reference Architecture 5.0, still the most detailed published crewed plan, works its reference mission out at 174 days outbound, 539 days on the surface waiting for the geometry to come back around, and 201 days home: 914 days, door to door. The alternative — an opposition-class trajectory, usually with a Venus swingby — cuts the total to somewhere between 500 and 650 days, but leaves only 30 to 90 days at Mars and demands far more propellant. There is no version where a crew goes for a fortnight.
This is why the record of missed Mars dates reads the way it does. A programme that slips does not slip by a season: the hardware sits in a clean room for two years and two months, ageing, while the team that built it is reassigned and the budget is re-argued from scratch. Every delay on this timeline is quantised to that number. The clock does not care.