4 ms·
The situation of [2] is crazy. Because both the author of [2] and the people using Ferguson’s model are wrong. And the author's advice in one of the last paragr
by enaaem 1mo ago
The situation of [2] is crazy. Because both the author of [2] and the people using Ferguson’s model are wrong. And the author's advice in one of the last paragraphs of we just need better programmers is also wrong.
The issue is that pandemics are extremely fat-tailed in terms of fatalities [3]. You need to collect a shit tons of data to estimate any reliable parameters and no one has time for that. The only thing you need to know is: Is this a super contagious deadly disease? If yes, lock everything down! This is pretty much what they did in Asia, and countries like Vietnam had zero cases for months and was open while the rest of the world was closed.
How to decide a disease is deadly and super contagious? You need disease experts (who most likely also academics) who understand the nature of the disease, and not better modellers and programmers as suggested in [2]. The issue is that western countries will never let any expert to make that call, so a lot of convincing needs to be done through complicated modelling and real life suffering.
[3] https://arxiv.org/pdf/2004.08658 https://arxiv.org/pdf/2004.08658
- lazide 1mo agoWestern countries won’t let an expert make the call, because there hasn’t been recent enough pain and suffering to put it in recent memory the necessity for it. After all, why suffer if it isn’t actually necessary? Asia had recent experience with the first SARS, which had very real and extremely high fatality rates, so for them it was a relatively easy (not an easy) call. For the west, when was the last time we had a real pandemic that actually killed a lot of people? Almost a century ago. Somewhat ironically because the west listened to experts for a very long time. Notably, the last time it had a major issue also coincided with a world war.