Numbers are going up fast because the number of people infected is large.
I have been thinking about how different people interpret data differently. And made this xkcd style graphic to illustrate this. pic.twitter.com/a8LvlmZxT7
— Jens von Bergmann (@vb_jens) March 17, 2021
Exponential growth has that characteristic shape in the cartoon linked above, but the growth rate as a proportion is constant throughout. The left-hand end of the graph looks tame whereas the right-hand end looks very scary, but it’s all the same rate of growth, eg:
1 person infects 2
2 people infect 4
4 people infect 8
8 people infect 16…
32, 64, 128, 256, 512, 1024, 2048, 4096, 8192, 16384, 32768…
The absolute numbers go up faster and faster, but it’s always doubling each cycle.
If you’re interested enough, you could check the growth rate over time. Take this week’s figure and divide by last week’s. Then take last week’s and divide by the week before, and so on. The results are weekly growth rates; see how they vary.
For a finer-grained inspection you can take today’s (ie. Saturday’s) figure, and divide by last Saturday’s. Then take yesterday’s figure and divide by last Friday’s. Then Thursdays, etc. etc. You again get weekly growth rates, but rolling day by day.
Or you can do it the other way around. Look at today’s infection figure, and then see how many days you have to go back for it to halve. Then do the same for yesterday, or last week.
Covid multiplies like crazy. I bet a lot of it is the weather getting colder and wetter making people congregate indoors and closing the windows, sharing air.