Empirical Earth · August 13, 2026

Nobody floated.
The Sun came back on schedule.

Yesterday, millions of people were told that gravity would switch off. Here is what happened instead, and why it was never in doubt.

The scoreboard

Two predictions. One date.

Both of these were public before August 12. Both named a time. Only one of them held.

Project Anchor
“Gravity will fail for 7 seconds.”
  • Time: 14:33 UTC
  • Duration: 7 seconds
  • Deaths: 40 million
  • Budget: $89 billion
  • Source: a caption
Did not happen
Celestial mechanics
“The Moon’s shadow will cross Iceland and Spain.”
  • Greatest: 17:45:54 UTC
  • Totality: 2m 18s max
  • Shadow width: 294 km
  • Saros series: 126
  • Source: NASA/TP–2006–214141
Happened

Nobody rose off the ground. No ceilings were hit. The tide came in, as tides do, because the Moon has mass and mass is what gravity is made of. A shadow is not a force. It never was.

The tell

The hoax got the time wrong by three hours

The viral post said the gravity failure would strike at 14:33 UTC. Greatest eclipse was at 17:45:54 UTC. That is a gap of more than three hours.

Think about what that means. The claim stapled itself to a real astronomical event to borrow its credibility, and then did not bother to look up when that event occurred. Anyone could have caught this in a minute, on any eclipse site, at any point in the seven months the post was circulating.

A claim that will not check its own homework is not asking you to think. It is asking you to forward.

How the real prediction was made

The map of yesterday’s shadow was printed in 2006

Not modeled. Not estimated. Printed, in a catalogue, twenty years early, alongside 11,897 other eclipses.

11,898
solar eclipses mapped in the Five Millennium Canon
2006
the year that catalogue was published
5,000 yr
span it covers, 2000 BCE to 3000 CE
< 1 min
how far off the timing was allowed to be

The arithmetic behind that catalogue is public. The Sun’s position comes from a solution called VSOP87. The Moon’s comes from ELP–2000/82. Feed those two orbits into the geometry of a rotating sphere, and the shadow track falls out: where it lands, how wide it is, how long it lasts, and to what second.

This is the part worth sitting with. Every one of those inputs is a claim about the shape and motion of the Earth. The tilt. The spin. The radius. The Moon’s distance and the tilt of its orbit. Get any of them wrong and the shadow lands in the wrong ocean.

Yesterday the shadow landed where the globe model said it would, when the globe model said it would. That is not a coincidence you can shrug off. It is a measurement, and it was made in public, by anyone standing in the path with a watch.

Why this eclipse in particular

The path ran backwards, and only one model allows that

Most eclipse tracks sweep west to east. This one climbed over the Arctic and came down the other side.

A shadow that begins at midnight, crosses the top of the world, and ends at sunset is not a thing you can draw on a flat map. It requires a sphere, tilted 23.4°, spinning under a Moon whose orbit is tilted about 5° against the plane of that spin. Those numbers were not chosen after the fact to make the story work. They were used, years ago, to make the forecast that was checked yesterday against reality.

What happens next

A prediction you can check within a week

Here is where we go on record, so that this page can be held to the same standard we are holding the hoax to.

The claim will not be retracted. It will be re-dated. Failed prophecies rarely die. They get moved. Watch for the same numbers, the same 7 seconds, the same $89 billion, attached to a new date: another eclipse, another conjunction, another anniversary. Watch for the explanation that the event was averted, or classified, or delayed.

If instead the accounts that spread it post corrections, we will say so here, and we will have been wrong about them.

That is the difference this whole site is built on. A claim worth anything tells you in advance what would prove it false. Yesterday, one claim did that and survived. The other one told you nothing that could be checked until the moment it was too late to matter, and then it quietly moved on.

If you were scared

You were not being stupid. You were being targeted.

The post was engineered. A real date, a real agency, a real branch of physics, a plausible-sounding project name, a budget precise enough to sound leaked, and a casualty figure large enough to feel urgent. Every one of those is a trust signal, borrowed and stuck onto a lie.

Being taken in by that is not a failure of intelligence. It is what happens when someone builds a thing specifically to get past your defenses. The useful question is not how did I fall for it. It is what could I have checked.

In this case: the time. Three hours off, and the whole thing falls apart. That is all it would have taken.

Bring that question with you

Empirical Earth takes every flat-Earth and eclipse claim, states it at its strongest, then tests it against evidence you can check yourself. Including the ones we get asked about most.