Predictions about the pandemic rarely age well. In the United States, the pandemic appeared to be winding down in May 2021 amid a vigorous vaccination campaign, only to get wound up again with the rise of the new variants.
Source: Fall covid surge possible, but unlikely to be as severe – The Washington Post
As documented on this blog with numerous projection follow ups, model forecasts even just weeks ahead have been wrong and worse than tossing a coin.
My favorite example is Oregon, which found models so poor they kept switching model providers and cut the projection length to weeks. From October 28th, 2020 to January 31st, 2021, they issued 13 projections, each for 3 weeks. In 11 of the 13 projections, they had the sign wrong. They projected an increase, but cases went down, or they projected a decline and cases went up. A coin toss, at least, would have gotten half right 🙂
