We Counted The Ones That Did Not Come True. There Were 1,411.
Every few months a film is found to have predicted something. Nobody counts the films that predicted nothing. We took 240 titles from one decade, logged every specific catastrophe depicted in them, and checked all of them.
A months-long reporting project. Documents cited below are held in The Vault and available to readers.
LONDON — A film depicts a disaster. Years later something similar happens. Somebody notices, and the clip goes round.
Nobody has ever counted the films that depicted a disaster and nothing happened.
That number is 1,411.
Building the denominator
Two researchers spent five months on 240 titles from a single decade. They worked blind to each other. For every specific catastrophe depicted they logged the place, the mechanism and the approximate date.
They coded 1,428 events. They agreed on 91 per cent of them; the 9 per cent they disagreed on is flagged in the published list rather than settled by us.
Only after all the coding was finished did anybody check what had actually happened in the world.
The result
Seventeen arguable matches. Four on the stricter coding, which required the mechanism and the place.
One thousand four hundred and eleven depicted catastrophes that simply did not occur — the volcano under the city, the second civil war, the comet, the sea wall, the eleven separate pandemics with the wrong pathogen in the wrong year.
Nobody has ever written about any of those. There is no reason to. That is exactly the problem.
Whether four beats chance
Our statistician's answer is no, and he asked that the reasoning be printed rather than the verdict.
Fiction depicts catastrophe constantly, because catastrophe is what plots are made of. The world also produces catastrophe constantly. Two large sets drawn from the same small library of human fears will overlap, and the expected overlap for 1,428 depictions across thirty years is — on his model, with his stated assumptions — larger than four.
"You do not have an unexplained signal," he wrote. "You have slightly fewer coincidences than I would have guessed, which is the least publishable sentence in English."
Where the ideas came from
We interviewed a screenwriter who has two "predictions" to his name and who has been sent clips of his own work for fifteen years.
Both, he says, came from the same place: a newspaper.
"I read about a near miss in 1994 and thought, what if it had not been a near miss," he said. "That is the whole job. Everybody in this trade reads the news and asks what if it had been worse. Then sometimes it is worse, later, and somebody edits my scene next to a photograph."
He was asked whether he found it flattering.
"I found it flattering for about a week in 2011. Now a man emails me every few months to tell me I am part of it. He is very polite. I do not know what to say to him and I have stopped trying."
The hindsight machine
Professor Halvard Nkemelu-Reyes supplied the mechanism, which is not mysterious and is very well studied.
After an event, memory reorganises around it. Details that were ambiguous become foreshadowing. A map on a wall in a film becomes a place name. You are not lying and you are not stupid; you are running a piece of standard equipment.
"The interesting question is never why people see the pattern," he said. "It is why anybody expects them not to."
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
The claim has never had a denominator, so we built one. Two researchers, working blind to each other, watched or read 240 titles from 1990–1999 and logged every specific depicted catastrophe — place, mechanism, approximate date. The analysis plan was written by a statistician before coding began. Only then did we check outcomes.
Who we spoke to
- Two research assistants, Named in our notes; coded independently and blind to each other. Five months; 240 titles; 1,428 coded events February–June 2026 Inter-coder agreement 91%. The 9% they disagreed on is flagged in the published list rather than resolved by us.
- Dr. Emeka Sandoval-Trần, Statistician. Wrote the analysis plan before coding began; report published January 2026
- Prof. Halvard Nkemelu-Reyes, Psychologist, hindsight bias. Interviewed by video call May 2026
- A screenwriter with two 'predictions' to his name, Named with consent. Interviewed in person June 2026 Explains where the ideas came from, which is not where anybody thinks.
Documents
- PX-2035 — 1,428 depicted catastrophes coded from 240 titles, 1990–1999 accepted
What we could not confirm
- Whether another decade would give the same rate. One decade, one language, one coding scheme.
- The 9% of events our two coders disagreed on. Both codings are published and we did not adjudicate.
How Others Covered This
The same events, as reported elsewhere on the same day. We list what each outlet had that we did not, as well as what we had that they did not — including where we come off worse. Why we print this.
-
The Meridian TelegraphTHE FILMS THAT KNEW
Listed nine hits and no denominator.
Had that we did not
Nothing.
Left out
Every film that depicted a catastrophe which did not happen, which is 1,411 of them.
-
They Buriedthis newspaperWe Counted The Ones That Did Not Come True. There Were 1,411.
Built the denominator, which took two researchers five months and is the only new thing here.
Had that we did not
The full coded list, published.
Left out
One decade, one language, and a coding scheme we invented. A different decade might give a different rate and we cannot say it would not. — V. Ashcombe-Doyle, standards editor
How was this story?
We publish the result, whatever it is. Reader verdicts appear on the front page and in our newsroom metrics.
Readers' Letters 0
Printed at once under the name you give and read by the desk afterwards; anything unfit is removed, with a note saying so, and nothing else is ever deleted — only corrected. Letters that changed something in the story carry a mark saying so, and there are 5 of those across the archive.