I predicted 1,333,803 cases and 27,423 deaths from Covid-19 in the U.S. in September. Now we see that the actual cases was 1,235,486 and the actual deaths was 24,004. I was off by 7.37% for cases and 12.47% for deaths. Actual cases and deaths was lower than predicted. However, this does not imply that the winter predictions will also be lower.

The vitamin D levels in the population are a sine wave. Lower lows means higher highs. And for Covid-19 cases and deaths, Vitamin D has a large influence. So I would expect cases and deaths to be the same or higher than what I previously predicted for October:
53,865 US deaths from Covid-19 in October
and
1,795,505 US Covid cases in October.

Of course, those numbers come from the math used to predict the cases/deaths, which is based on human coronavirus seasonality studies, and using that ratio of one month to another to predict cases, and then the death rate is estimated based on an expected fall of vitamin D levels in winter. The above estimate is based on a death rate of 3%. If it is 2% instead, then the deaths would be 35,910 instead of over 53,000. But I think the death rate will rise as vitamin D levels fall in the general population.

RLCJ