A few weeks in and Australia is still struggling to get on top of the recent delta-variant outbreak of covid. It’s apparent that the initially muddled and delayed lockdown in Sydney was too late to keep the virus bottled up and it is now spreading in more rural areas.
As regular readers will know, I’ve tried to stick with one fairly consistent way of looking at covid numbers: cumulative confirmed cases per population size. There are no perfect numbers and these figures have the same issues in terms of being dependent on testing rates. However, my point about the graphs has been that it is not so much the magnitude as the slope of the graph. Lower testing rates may obscure or delay how that graph gets steeper but when the virus is spreading exponentially it will show up in the numbers.
On the “delay” aspect with testing, I thought I’d illustrate that with a simplified graph. The underlying data is a number that increases proportionally by 1.2 times the previous “day’s” number. Two lines show 5% of the current day’s total (blue) and 1% of the current day’s total. Obviously, testing rates do impact the number of reported cases but inevitably the numbers go wooosh regardless.
Putting the curves side-by-side helps show how the pattern is similar but these two lines look identical if plotted with different scales on the vertical axis. The underlying shapes are similar in a mathematical sense where the difference is scale.
Anyway, still locked down for the time being and the curve is still going up.