The modelling which led to Britain’s dramatic lockdown amid doomsday predictions that 510,000 would die from coronavirus has been found to be so unreliable it is worthless.
Professor Neil Ferguson – the man from Imperial College who ignored the government’s stay at home messaging and allowed his married lover to visit him – has been further discredited in recent days.
Analysts have discovered that the modelling upon which Prof Ferguson based his calculations was decades old code similar to the programming language Fortran, which was old news 20 years ago.
It is so unreliable and full of errors that even imputing the same data produces widely different outcomes.
This means that the figures Prof Ferguson produced – that deaths would hit 510,000 on a “’reasonable worst case’’ basis if nothing was done – was completely random.
When Prof Ferguson’s model, which was not peer reviewed, was released it created immediate concern in the scientific community because of the seemingly extreme data inputs.
The modelling, as revealed by The Australian, was quickly changed two days earlier because Prof Ferguson had initially underestimated – by half – the number of people requiring intensive care caused by the coronavirus in China and Italy.
At the time he said: “We have had bad news from Italy and from early experience in UK hospitals that the intensive care requirements will be nearly twice what we had anticipated.”
That prompted him to crank up the computer and produce such scary figures that Prime Minister Boris Johnson, on the advice of his secretive scientific advisory committee, of which Prof Ferguson was a member, thrust the country into immediate and draconian lockdown, which is still largely in place. In addition to the 510,000 figure, Professor Ferguson claimed 250,000 would die if moderate lock down measures were put in place.
The Imperial model used code to look at transport links, population size, social networks and healthcare to anticipate the spread of the virus and the ability of the National Health Service to cope.
But when researchers released the code underpinning the advice, the faults became apparent.
David Richards, co-founder of tech firm WANdisco, told the UK’s Telegraph: “It’s a buggy mess that looks more like a bowl of angel hair pasta than a finely tuned piece of programming.
“In our commercial reality, we would fire anyone for developing code like this and any business that relied on it to produce software for sale would likely go bust.”
Scientists from the University of Edinburgh reported problems with the Imperial model, saying they got different results when they used different machines, and even in some cases, when they used the same machines to run the model.
“There appears to be a bug in either the creation or re-use of the network file. If we attempt two completely identical runs, only varying in that the second should use the network file produced by the first, the results are quite different,” Edinburgh scientists reported.
