The muon is a heavy cousin of the electron that lives for just two millionths of a second. In a magnetic field, it spins like a top, but slightly faster than formulas predict. Twenty years ago, this 'slightly faster' was noticed at Brookhaven, and physicists have been arguing ever since: either the instruments were wrong, or unknown particles are hiding in the vacuum.

Now, at Fermilab, the muon's spin was measured four times more precisely, and again it turned out 'too fast.' But in parallel, theorists on supercomputers recalculated the vacuum's contribution: it turns out the vacuum isn't empty but teems with virtual particles that give the muon a slight extra twist. New calculations, accounting for the mixing of two types of particles, raised the theoretical value almost to the experimental one. The discrepancy melted away.

The Standard Model has held up again. It doesn't explain dark matter or the tiny neutrino masses, but for the muon, its predictions now match experiment. This isn't a failure but a triumph of precision: we've learned to calculate the vacuum better.