Hacker News new | past | comments | ask | show | jobs | submit login

When the German chancellor is declaring 60% of the population will become infected, you start to get a sense of the scale.

If you look at just first world nations[0], you're talking a total population of 1.04 billion people. 60% of that is 630 million people. 3% of that 60% is 18 million people that could die just in the first world nations.

And that's rounding everything down.

0: https://worldpopulationreview.com/countries/first-world-coun...




technically it is not wrong, after all it is backed by actual, true science. That being said, this estimation assumes that nothing will change, so no working counter measures of any kind.

In reality, you have China. They are nowhere near these numbers. In Germany trade show get cancelled, events get cancelled, soccer games take place without spectators, schools close. Quarantine measures will certainly be put in place as well when needed. All this helps to keep infections down and slow the spreading. It worked in China after all.

I am no expert on diseaes, but I did my fair amount of number crunching. And that tell me that just taking the current average of deaths and infections, currently somewhere around 3.4%, extrapolate the number of infections worldwide and multiply the two doesn't fly. At the very least you have to brake down the world population by age groups based on the same groups we have specific mortality rates for. doing so already gives you a better picture and a more realistic number. Nobody who comes up with expected fatalities seems to bother to do so.

Obviously, that leaves the whole impact of counter-measures out of the equation. it doesn't account for time nor for any regionl clusters of people. Neither does it account for any local shortages due to an overwhelmed medical system for certain periods in certain regions. Or for any steps taken to counter these effects. To properly do so requires a full scale modelling of this thing. unfortunately the numbers are changing quite fast, so building these models is quite ifficult i suspect.

You would also have to account for margins of error and the fact that testing is a major influence in any parameter of this model. And testing varies highly between countries.

TLDR: The numbers a re a mess, calculation and modelling is difficult and everything is changing daily. Good luck with coming up with any reliable predictions.

I would love to see numbers for testing so. Especially how many tests have been conducted by region, how many where positive and how many negative. Ideally over time.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: