Hint for modelers, who often seem to fail due to over-complexifying things: the winter Covid surge (and, yes, there will be one), will be best modeled by using vaccine rate with a small (1.2x) multiplier for the likely-greater transmissibility of the Delta variant, and comparing this with what happened last fall/winter.

Thus, if there is a 70% full vaccination rate, and last year’s winter case count was 40,000 per day, then winter 2021/2022 will have a case count of approximately 14,000 per day. That is how you do it if you want at all accurate results, but the temptation is to make models as complex as possible to demonstrate intelligence and bona fides.

But there’s a reason I’ve been right and, mostly, all those mooks have been wrong. And it’s gonna happen again.

Note that my numbers above aren’t predictive. I haven’t actually used case counts because I don’t have time to look them up. Those counts are for demonstration purposes only of how you’d produce a model and forecast that doesn’t totally fall on its face.