Thread by @ToshiAkima

Dear Journalists. Please take note and hold politicians accountable by challenging their falsehoods. This is basic high-school level maths, not rocket science. In this thread, I will explain it in easy-to-grasp terms

https://twitter.com/YouAreLobbyLud/status/1432623290040799233

This is what "linear growth" looks like. It goes 2-4-6-8-10-12 etc. It takes quite a few steps to get from 2 to 12. That means growth over time often tends to be slower in a linear growth model. The graph looks like this. Linear = straight line

With exponential growth, you get serial doublings. So if you start at 2, you go 2-4-8-16-32-64. Exponential growth allows you to get from 2-12 in far fewer steps than with linear growth. The curve looks like this

The NSW case numbers are growing like this (this is the AU national data but the VAST majority of the cases come from NSW). Does it look like a linear growth curve to you? Or does it look like an exponential growth curve to you?

Journalists are encouraged to make a poster-size printout of the NSW growth curve in case numbers and ask politicians to point out on it what type of growth curve it shows. Linear or exponential? And to please explain why

Mathematicians who do predictive modelling also find that the basic model of exponential growth best fits the NSW case data. For example look here

https://twitter.com/juliette_io/status/1432585825141751812?s=20

An effective R (Reff) value of > 1 defines exponential growth despite throwing lockdowns, masks, vaccines at the virus. So if the Reff > 1 as it currently is, sitting steadily at around 1.2, this means that we are by definition in exponential growth

https://twitter.com/juliette_io/status/1432578593687695360?s=20

If politicians say we aren't in exponential growth what are the options? It's not linear growth or the line would be straight. It's not exponential decline (with a Reff < 1) or else numbers would be dwindling fast. That just ain't happening

Ok, mine is a dummies version. I quickly go running to my institute's resident statistician when the maths gets fancy. Here's one bit of cool modelling which shows the NSW growth curves to be SUPER-EXPONENTIAL! 😱

https://twitter.com/BruceTabor8/status/1431945772648435715?s=20

My dummies understanding of superexponential growth is that it's like a tsunami coming at you with the serial doubling time shortening as the wave nears. On the NSW frontlines, this is exactly what we are seeing. We're going to be hammered. Brace for impact

Let's go back now to see why viruses spread exponentially. If a virus spreads with an R-value of 2, each infected person infects 2 others, who each infect 2 others (ie 4 in total) etc. So cases increase 1-2-4-8-16-32-64...which is by exponential doubling (see above!)

Let's suppose we then introduce lockdowns, masks etc, and you reduce the effective R-value to 1.2. That means each infected person now spreads to an *average* over the population of 1.2 people. Numbers overall are still going to increase exponentially, only much slower

This means that as long as the Reff is > 1, even if it's only 1.2, the increase in case numbers can only be exponential. If COVID spreads, with numbers on the rise, it always does so exponentially. Like 1 + 1 = 2, it just isn't *possible* for it to be any other way!

So please everyone if you found this thread to be a nice easy explanation, show it to journalists and ask them to hold politicians accountable when they tell you that 1 + 1 = 3. If you break it down the maths is quite easy and they can't blind you with pseudoscience