[Svartalfiar]

This is an interesting area of research. There is some seasonality to death rates, partially due to influenza, and partially due to changes in accident rates by season and other similar causes.

Seasonality shouldn't affect this. It's a year vs year vs year comparison, so that same seasonality would be present in each data point being compared.

The only big issues might be if there's a significant number of deaths yet to be reported (unlikely), or they compared total numbers for the full year, when this year isn't completed yet (possible). But, the death chart shows a pretty big drop the past month or two, so I would be inclined to believe the story is accurate.

[freeandfreezing]

Seasonality shouldn't affect this. It's a year vs year vs year comparison, so that same seasonality would be present in each data point being compared. That would be true if the comparison was for full years. But once the operation of dividing by months and comparing the monthly averages was done that is not necessarily correct if there is seasonality.

If more people ordinarily die in the fall and early winter then the per month rate found by dividing the last year total by 12 overstates the death rate for the first 8 months of the year. If the death rate is higher in the spring and summer then the reverse is true.