4 ms·
"The prediction is based on a very simple assumption that the counts will follow an exponential curve" This reminds me of news reports from the 90s, where the
by lbj 7y ago
"The prediction is based on a very simple assumption that the counts will follow an exponential curve"
This reminds me of news reports from the 90s, where the big fear was that Chinese companies would totally take over Europe. The big Chinese breakthrough would happen 'next week', 'next year', 'next some arbitrary buffer that puts us just outside of the data'.
Corona might go exponential, but with the amount of effort being put into containment, I doubt it.
- semi-extrinsic 7y agoOn the contrary, isn't the expected behaviour for most types of growth that it will be exponential in the initial phase, then transition to an S-shaped curve? Whether it be expansion into a new market or spread of a virus, in the early days there is little resistance. Then when (competitors|health services) start noticing and working against it, the spread slows and plateaus. How quickly that happens and what level of spread the plateau is at depends on the speed and efficiency of the countermeasures.
- jhart99 7y agoAt some point everyone an infected person comes into contact with is either already infected or has previously been infected( and potentially immune). So I would think you would be exactly correct, at some point the number of infected people will peak and begin to fall. The number of dead and the number of recovered will form an sigmoid curve.
- generalpass 7y agoI'm pretty sure Gartner predicted that by 2010 there would be 15,000,000,000 Palm Pilot users.
- Tomminn 7y agoFor clarification, the coronavirus is currently exponential with a R^2 value very close to one. The question is whether of not it will stay exponential, or go sigmoidal soon.