When school textbooks explain the Milky Way’s far-future, they often end with a drama: our spiral home colliding head-on with the equally grand Andromeda galaxy roughly five billion years from now.

However, a new study in Nature Astronomy now paints a calmer picture. Astronomers from the universities of Helsinki, Durham, and Toulouse ran 100,000 computer experiments that fold in the sharpest data yet from NASA’s Hubble Space Telescope and the European Space Agency’s Gaia mission. Their verdict: a direct hit in that timeframe is only about a two-percent gamble, not a certainty.

Lead author Till Sawala from the University of Helsinki notes that the team’s results do not overturn earlier calculations; instead, they widen the lens.

“When we tried to start from the same assumptions as previous researchers, we recovered the same results,” Sawala said. “We’ve simply been able to explore a much larger space of possibilities, taking advantage of new data. And while earlier studies only considered the most likely value for each variable, we ran many thousands of simulations, which allowed us to account for all the observational uncertainties.”

What changed?

Previous forecasts relied on the best single values available for Andromeda’s motion and mass. The new work treats those numbers as ranges, lets the computer explore every combination, and, for the first time, includes the gravitational tug of the Large Magellanic Cloud (LMC), a satellite only one-seventh the Milky Way’s mass but close enough to matter.

That sideways pull tilts our galaxy’s path just enough that, in most scenarios, Andromeda sails past on the first approach. The two spirals still feel each other’s gravity and shed orbital energy, yet they usually need eight to 10 billion years—twice the old estimate—before any eventual merger. By then the Sun will have swollen into a red giant and emptied the Solar System of life-friendly conditions.

“While some earlier works had focused on the interaction between the Milky Way, Andromeda, and the Triangulum galaxy, we also include the effect of the LMC,” Sawala said. “And while earlier studies only considered the most likely value for each variable, we ran many thousands of simulations, which allowed us to account for all the observational uncertainties.”

Across the 100,000 trials, more than half still end with the galaxies merging in the distant future, but many others show them settling into separate, stable orbits that last far longer than the present age of the Universe. Even in cases that do culminate in a union, the first swing-by may occur hundreds of thousands of light-years apart—far enough that stars, planets and dark matter halos remain largely undisturbed for billions of years.

This new uncertainty about the future of the Milky Way and Andromeda may not last, as the team are already looking ahead to researching further scenarios when even more data becomes available. The Gaia space telescope’s upcoming data release will refine Andromeda’s sideways speed—one of the hardest quantities to measure—and shrink the margins of error in the next round of simulations.

“Until now we thought this was the fate that awaited our Milky Way galaxy,” said Carlos Frenk of Durham University. “We now know that there is a very good chance that we may avoid that scary destiny. When I see the results of our calculations, I am astonished that we are able to simulate with such precision the evolution of gigantic collections of stars over billions of years and figure out their ultimate fate.”