Neutrinos beat light by 60 nanoseconds — the fault was a loose cable

In September 2011 the OPERA collaboration announced that muon neutrinos sent 731 km through the Earth from CERN to the Gran Sasso laboratory in Italy arrived 60.7 nanoseconds early — six sigma, faster than light. Spokesman Antonio Ereditato hedged, asking the community to 'help us understand our crazy result', but the caveat was no match for the story: relativity broken, Einstein wrong, one of the pillars of modern science tumbling down.

The anomaly had been sitting in the data since March, and the six months between were spent cross-checking — re-running analyses, hunting 'trivial mistakes, more complicated mistakes, or nasty effects', and finding none. What nobody did was physically inspect the timing chain: the cable carrying the GPS signal to the experiment's master clock. In November they reran with a different beam and got the same early arrival, 57.8 nanoseconds — a second signature from the same broken clock.

On 22 February 2012 the collaboration announced two faults: a loose fibre-optic connector on the GPS link, which shaved 73 nanoseconds off the measured flight time, and an oscillator ticking faster than its specification, which added some back. Photographs showed the cable had been loose since at least 13 October — before the confirming rerun. Co-located ICARUS measured neutrinos at exactly light speed. Ereditato and coordinator Dario Autiero resigned at the end of March; in June CERN declared the result dead, and OPERA's corrected re-analysis settled at 6.5 ± 15 nanoseconds — no anomaly.

The damage was real but bounded: eight months of global headlines, theoretical papers inventing physics to fit a cable, and an Italian minister's press release praising the tunnel between CERN and Gran Sasso that never existed. The collaboration paid its bill in public, and the affair became the textbook case of a systematic error dressed as a discovery.