Models Are Not Evidence

The point is so elementary that it should not be necessary to state: a model is not evidence. It is a theory–a hypothesis–expressed in arithmetic terms. A theory is either validated or disproved by observation. A model that is contradicted by experience is simply wrong, and is useless. History is littered with theories that sounded plausible at the time, but were invalidated by experience.

Which brings us to covid-19. A number of modelers around the world, particularly in the U.K. and the U.S.A., tried to predict the spread and impact of the virus under various conditions. Those models–certainly the ones that were most prominently publicized–turned out to be not just wrong, but grotesquely wrong. That is why we haven’t heard much about them lately.

Here in Minnesota, home-grown models #1, #2 and #3 were rolled out, hyped, and then shelved in short order as each turned out to be wildly off the mark. Though having served, allegedly, as the principal basis for policy decisions that destroyed thousands of small businesses and blighted the lives of many thousands of children, they have been quietly forgotten as an embarrassment.

In Great Britain, models apparently have retained some currency, despite being always wrong. The Telegraph brings us up to date: “How the UK’s Covid reopening has proved Imperial’s pessimistic modelling wrong.”

Asked on BBC Radio Four’s Today Programme about the latest data showing Britain is enjoying an eight-month low in coronavirus deaths and infections, Professor Neil Ferguson said on Tuesday: “The data is very encouraging, and very much in line with what we expected.”

This is the guy who predicted millions of Wuhan deaths but nevertheless violated his own guidelines to arrange a liaison with his illicit mistress.

As it was on the radio it was impossible to tell whether this was said with a straight face, but we must assume it was.

It is a wonder that nobody choked on their morning toast, for if Imperial modelling has stood for anything in this crisis, it is relentless pessimism. Plummeting figures were certainly not predicted by its researchers.

The difference this time is that the Government has pressed ahead with reopening despite the doom-mongering, and so has proven the models wrong.

Details at the link. Ferguson and his Imperial model were wrong, wildly wrong, about everything.

Under the February modelling which informed the roadmap, hospitalisations should be starting to tick up around now, but there is no evidence of that happening, with cases down 11 per cent in the past week.

On Tuesday, Prof Ferguson admitted that the link between infections and deaths had now been broken, and that up to 90 per cent of people who would have ordinarily ended up in hospital after contracting Covid would no longer be admitted.

Hey, off by a factor of 90%? That isn’t bad for a liberal.

Ferguson’s Imperial model grossly overestimated the impact of re-opening Britain’s schools, an empirical fact that would be important for the U.S. if corrupt teachers’ unions did not run public education in this country. Then there are the model’s inept projections of the impact of vaccinations:

Vaccine efficacy was badly underestimated, too. In the February modelling, one dose of AstraZeneca was assumed to protect between 56 and 70 per cent of people from hospitalisation and death. In reality, it was up to 94 per cent.

In fact, the only thing that Imperial modelling has going for it, was it was slightly more optimistic than the London School of Hygiene and Tropical Medicine, which predicted an even more devastating third wave.

Well, there’s that. It is curious, but the one thing that all the models I know of had in common is that they grossly overestimated the health consequences of the Wuhan virus under various conditions. They thus served as the basis for hysterical government actions that limited personal freedom to an unprecedented degree, but that turned out to do no good at all–a fact that no model predicted, and about which not a single government entity, seemingly, is embarrassed.

Coincidence? I don’t think so. If there is a silver lining to the covid fiasco, it could be increased public skepticism of models and other theories that lack empirical support. But for such skepticism to take root would likely require honest news reporting, so don’t hold your breath.

Notice: All comments are subject to moderation. Our comments are intended to be a forum for civil discourse bearing on the subject under discussion. Commenters who stray beyond the bounds of civility or employ what we deem gratuitous vulgarity in a comment — including, but not limited to, “s***,” “f***,” “a*******,” or one of their many variants — will be banned without further notice in the sole discretion of the site moderator.

Responses