A check-in on how @IHME_UW COVID-19 projections are doing. (Spoiler: they& #39;re still completely awful).

A month ago I tweeted about IHME& #39;s latest projection for Sweden. Let& #39;s revisit those projections... (1/5) https://twitter.com/yinonw/status/1302396409774108672">https://twitter.com/yinonw/st...

A month ago IHME projected 180 deaths/day by Dec 1 and 20 deaths/day by Oct 1.

Instead by Oct 1st Sweden is at 1 death per day for the entire country. As reminder, masks are not worn and primary schools and businesses never closed at any time in the pandemic. (2/5)

So how has IHME updated their projections from this "surprising" result?

Instead of learning from reality and adjusting their models, they assume reality is wrong and their models are right. They simply shifted everything back another month.

New projection below. (3/5)

IHME projects that if Sweden dons masks then deaths will magically remain at near zero.

Assuming Sweden doesn& #39;t wear masks (they didn& #39;t at the height of pandemic so very unlikely now), will IHME admit how utterly worthless their model is? Probably not. (4/5)

Here are the early Sept and early Oct IHME models superimposed on one another. You can see IHME has adjusted nothing from their grossly incorrect projections and simply pushed their doomsday projections out by yet another month.

We can check in again in another month... (5/5)