Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

And they both use the same scare statistics. (covid) 2 million Americans will die. (global warming) If we don't stop emitting CO2 the world will collapse by 2030.


Two million was at the very high end of only a very few models (one?).

That said, Italy is at the top of their curve (hopefully, apparently) with 16,000 deaths. However, their excess deaths this past month (compared to baseline for March) are extremely high, and equal almost as many deaths as the official Covid-19 death toll. According to at least some Italian demographers and epidemiologists, it's likely most of those deaths are due to Covid-19 and were uncounted. But let's say that only half of those deaths were Covid-19, and that the curve turns out to be relatively balanced, as with most epidemics that don't come in waves. If that's the case, Italy will see around 50,000 deaths from Covid-19. Spain will likely see around the same proportion of Covid deaths.

If the US were to experience a similar proportion of deaths, we'd be at around 500,000 deaths. I don't think we'll see nearly so many deaths from Covid-19 now but in early March, it just wasn't clear. It's not 2 million, but my point here is that this was a model that likely turned out high, but it was within the same order of magnitude. And again, I'm using the actual deaths that have occurred in Italy at the top of the curve to make that estimate. It's possible that both the curve will suddenly drop precipitously, and that most of those excess deaths were caused by some other phenomenon, but it's highly unlikely based on behavior of past respiratory pandemics.

Many more models people have discussed have predicted anywhere from 48,000 to (most commonly) around 200,000 deaths in the US from Covid-19. I think we'll probably end up around the 50,000 death figure, and that's with us shutting down our economy. I think it could have easily been much, much more (although probably nowhere near 2 million) if we had carried on business as usual (and could still end up like that if we're not careful).

So complain that the media latched on to the most dire model, that's OK. I agree the media sensationalizes. "If it bleeds, it leads".

But don't complain about the model itself. Those papers are always very clear about the degree of uncertainty.

I don't know of any serious scientist who is predicting global collapse by 2030 from global warming. Do you have someone specific in mind?


You're exaggerating, which is just as bad.

NYT reported that "200,000 to 1.7 million" people could die on March 13th [1]. Most of those calculations had to do with the fact that the hospital system would get overwhelmed at some point, compounding the deaths.

None of that is "scare statistics". It's useful modeling.

I think what you consider to be "scare statistics" is more lack of education. Similar to how people saw the Trump win as meaning "The media/polls were wrong! They had Clinton winning at 85%! They can't be trusted!". When in reality, flipping a coin and getting 3 heads in a row is all it takes for Trump to win. Or rolling a standard 6-sided die and correctly guessing the number once. It suddenly doesn't sound outlandish.

[1] https://www.nytimes.com/2020/03/13/us/coronavirus-deaths-est...

Edit: The downvotes are concerning. If you have a problem with my argument, please respond.


Exactly. Both the media and the anti-media/anti-expert crowd both have a poor understanding of probability and confidence intervals.

With the media, I think it's sometimes intentional and not ignorance, as a ploy to get clicks.

With the anti-media, anti-expert crowd, I have to believe it's usually simple ignorance.


Mostly agree. Some media is certainly better than others. The NYT article I linked has a very responsible title (IMO) of "Worst-Case Estimates for U.S. Coronavirus Deaths".

It clearly says that the estimates they're about to talk about are worst-case and estimates.


True, I should have qualified that with "most" or "much" media. I agree there are still some responsible outlets out there.


Do you realize that if the media write "could die", readers will interpret it as "will die" ?


If you actually read the NYT article, it's really hard to come out of it thinking "1.7 million people will die". This is because of a few reasons:

1. They have a good, responsible title.

2. They bury the 1.7 million number a bit so you're forced to at least understand some of it.

3. They talk about how and why the estimates will change over time:

>“When people change their behavior," said Lauren Gardner, an associate professor at the Johns Hopkins Whiting School of Engineering who models epidemics, “those model parameters are no longer applicable,” so short-term forecasts are likely to be more accurate. “There is a lot of room for improvement if we act appropriately.”

I see no reason why this article shouldn't have been posted in its current form.

Of course, as you get further away from good sources of news, you're going to get less responsible reporting. NYT is an example of very high quality reporting.

CNN at the time probably had a breaking news line item on TV of "1.7 million could die". This would be irresponsible.

Fox News at the time probably had a breaking news line item on TV of "It's just the flu". This would be irresponsible.

And that's certainly a problem.




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

Search: