Hacker News Reader

Comment by supperburg | Hacker News Reader

Yes absolutely. But do you admit that when the other guy is more than an order of magnitude wrong (4%), that’s he is spreading bullshit even more than I was? And do you admit that nobody seems to call him out like they ought to? Maybe he would offer a swift correction like me and restore his honor and the whole situation could be better that way.

View on HN

You keep on using the top of the estimate as if it was the only estimate. They said 1%-4%. I agree that is high, but I also don't think the real rate is 0.2% either as that requires the total COVID death number to be inflated. I don't think there is evidence for that. I think the real rate is probably somewhere in the in the 0.4%-0.8% range. That means the 1%-4% estimate is pretty close at the low end. And if we are going to err on one side or the other, the side that makes people more cautious during a global pandemic rather than less cautious is probably the way to go.

You could also read what Stross wrote slightly differently:

> SARS1 has 20% mortality among patients, MERS (aka SARS2) is up around 35-40% fatal, SARS-NCoV19, aka SARS3, is down around the 1-4% fatality level.

"Among patients" would make me assume that Stross is talking about case fatality rates, not infection fatality rates.

1-4% is a reasonable range for CFR, even if it's too high for IFR.

(Some quick Googling after the fact also shows that while MERS had a CFR in the 30-40% range, its IFR was estimated at more like 0.3-0.7% -- roughly the same range as COVID. Large numbers of undetected mild cases were cited as the reason for the large difference.)

Reply on news.ycombinator.com