On the night of 18–19 August 2024, Bayesian, a 56-metre sailing yacht built by Perini Navi, capsized and sank within minutes while at anchor off Porticello, Sicily. Seven of the twenty-two people on board died, including British tech entrepreneur Mike Lynch and his daughter. In early May 2026, leaked conclusions from a preliminary report commissioned by the Termini Imerese public prosecutor’s office reframed the case: in the view of the prosecutors’ experts, that night’s weather does not explain the sinking. The crew, however, is directly in the frame.
”Little more than a squall”: the prosecution experts’ thesis
According to the findings reported in early May by the British press — the document itself has not been made public at the time of writing — the experts appointed by the prosecutors describe the weather that night as “little more than a squall”, a sudden increase in wind speed that precedes thunderstorms and downpours. In other words: a phenomenon a professional crew should have anticipated and managed.
The preliminary report is said to identify three failings: improper actions by the crew, an underestimation of the weather, and safety devices that were not activated correctly. Three people — the captain and two crew members — remain under investigation and could face charges of negligent shipwreck and manslaughter.
Caution is warranted: as of this writing, no formal charges have been confirmed, the prosecutor’s office has not released the report, and those concerned are presumed innocent. But the direction of the Italian criminal investigation is now explicit.
A reading squarely at odds with the MAIB’s
This is what sets the case apart: the thesis contradicts the interim report published in May 2025 by the MAIB, the UK’s Marine Accident Investigation Branch — Bayesian flew the British flag. Drawing on a stability study by the Wolfson Unit at the University of Southampton — a preliminary desktop assessment, to be revisited once the wreck had been examined — the MAIB estimated that in the yacht’s configuration that night (sails furled, centreboard raised), a gust in excess of 63.4 knots on the beam would likely have been enough to knock the yacht past its angle of vanishing stability, calculated at 70.6 degrees. Beyond that angle, there was no coming back. The report added that a similar vulnerability at lower wind speeds could not be ruled out.
Crucially, the MAIB stressed that none of these limits appeared in the stability information book approved in 2008: neither the owner nor the crew could have known them. As for the weather, the Met Office study commissioned by the MAIB, together with local observations, pointed to a highly probable mesocyclonic thunderstorm, with transient hurricane-force winds well in excess of 64 knots — and possibly considerably more near the surface.
The reconstructed timeline conveys how sudden and violent the event was: standing orders to wake the captain if the wind exceeded 20 knots, 8 knots logged at 0300, 30 knots at 0355, the captain woken at around 0400 — and at 0406, a violent 90-degree knockdown to starboard in under fifteen seconds.
The two investigations follow different logics: the MAIB’s is a no-blame safety investigation whose report is inadmissible in any proceedings seeking to apportion blame or liability; the prosecutors’ investigation aims squarely at establishing criminal fault. As things stand, however, the contrast between a squall the crew should have managed and hurricane-force winds meeting an undocumented vulnerability looks irreconcilable.
What captains and managers should take away
Whatever the judicial outcome, the case already sets out four practical lessons for professionals.
Know the yacht’s real stability, in every configuration. An approved stability book may say nothing about squall vulnerability in the motoring configuration with the centreboard raised. For large sailing yachts in particular, owners and managers should require the naval architect or classification society to provide limiting wind curves for the conditions actually operated in — including at anchor.
Tighten the anchor watch. Going from 30 to more than 70 knots in eleven minutes, after a quiet night at 8 knots, is not a textbook hypothetical: it is what the MAIB documents. Wake-up thresholds, standing orders for forecast convective storms, criteria for getting under way as a precaution: these procedures must be written, known and drilled — not improvised at four in the morning.
Document, even where regulation does not require it. A yacht operated privately generally falls outside the mandatory scope of the ISM Code. Yet it is precisely documentation — procedures, logbooks, weather records, briefings — that protects a crew when a casualty becomes a criminal case. A voluntary, living safety management system (SMS) is not administrative overhead: it is a line of defence.
Prepare for the legal aftermath. The Italian criminal investigation is proceeding independently of the British safety investigation, an inquest into the deaths is under way in the United Kingdom, and civil proceedings have been reported in the press. For crew and owner alike, insurance cover — including criminal defence — and access to legal assistance should be verified before an incident, not after.
The MAIB’s final report, awaited with no announced date, will incorporate the examination of the wreck, raised in June 2025. Until then, the Bayesian case is an uncomfortable reminder: at anchor too, the line between weather hazard and crew fault will be drawn, in court, by the quality of shipboard procedures and documentation.