3 ms·
Just 29% lower is nothing. NOTHING. Of all people, the YC News crowd really ought to understand that O(exp(n)) overwhelms O(1) literally exponentially. Not fig
by jiggawatts 5y ago
Just 29% lower is nothing. NOTHING.
Of all people, the YC News crowd really ought to understand that O(exp(n)) overwhelms O(1) literally exponentially. Not figuratively. Literally.
If a virus variant has a constant factor reduction in its side-effects, but a higher exponential factor in infectiousness, the latter will overwhelm the former in short order, and then continue to overwhelm it even more. Exponentially.
Let me do some simple maths assuming that it spreads "just" twice as quickly as Delta.[1] Lets assume that at some point, Omicron will be the primary strain, and there are 29% less patients in hospital because of it. Assume the doubling-time of Delta is 1 month. (It's better or worse than this depending on circumstances, but start here.) The doubling time of Omicron is then 2 weeks.
So with Delta you have: 1.0 == .71 * 2^(t/30d) which solves to about 15 days to get back to a full hospital.
With Omicron the equation is 1.0 == .71 * 2^(t/14d) which means it'll be back to a full hospital in just 7 days. Then the hospitals will be at 200% capacity in 21 days, 300% capacity at 29 days, and 1000% of capacity in just 53 days.
Of course, the full mathematical model is more complicated, typically something like SEIHRD, but the point is that a mere 29% reduction in mortality is next to nothing compared to an exponential increase in infectiousness. It would have to be more like 90% less dangerous for that to matter at all in the long run.
[1] https://www.forbes.com/sites/masonbissada/2021/12/03/scientists-say-omicron-is-spreading-twice-as-quickly-as-delta-in-south-africa-and-could-be-more-likely-to-re-infect/?sh=1c87ec163c5f https://www.forbes.com/sites/masonbissada/2021/12/03/scienti...
- IAmGraydon 5y agoThat was an incredibly verbose explanation for something very simple.
- jiggawatts 5y agoIt's a simple concept that about 95-99% of the general population does not understand at any level. This includes most politicians, civil planners, even scientists and medical staff. When the pandemic first started, people were flabberghasted about how FAST! the numbers were going up. Faster and faster, new records broken every day! Meanwhile, on a logarithmic plot, it was a dead straight line. It wasn't changing, not really. People don't "get" this. Even now, two years in, people are clinging to constant factors. I've had similar conversations with people who just don't understand Moore's law, Nielsen's Law, etc... I had to use pictures to explain to IT network professionals that upgrading a 15-year-old 2 Mbps site office link to 4 Mbps is hilariously under-specced when the mobile phone in my pocket can exceed 1 Gbps while I'm sitting at the bus stop. They just hadn't "grokked" that their mental models are off by orders of magnitude. They're used to x*2 instead of x*10^2.
- icedchai 5y agoAre you surprised? People don’t understand compounding investment returns either. 4 megabits! That would be considered fast 25 years ago.
- Accacin 5y agoHi, I'll admit I didn't fully understand that, and your explanation helped. I'll also use this for when my friends and family miss the point too, thanks :)
- kbelder 5y agoThe human population is finite, and exponential growth hits its bounds pretty quickly. In the long run, if you assume that most non vaccinated and a large part of the vaccinated will eventually get COVID (probably the case), then the 29% reduction in mortality does, actually result in a 29% reduction in mortality, and the exact same number of people will have been infected either way. The difference will be in timing; the people will have been infected more quickly, with the (severe) problems that entails.