4 ms·

https://threader.app/thread/1252581933835575297 https://threader.app/thread/1252581933835575297 This is a super interesting point of view.

war1025 6y ago
I read through most of that but will admit that I don't quite understand what he was trying to get at. Can you summarize?
- dboreham 6y ago
  I only read the beginning, but it seemed to be about transmission not being completely transitive.
- baq 6y ago
  it more or less says that people who get met a lot (...directed graph, remember) somewhat counter-intuitively increase the total percentage of people infected more than you'd expect even if they don't transmit the disease after they contract it more than an 'average' person (again, directed graph).
  - war1025 6y ago
    I guess that doesn't seem counter-intuitive to me. That's the whole reason behind social distancing right? To reduce the likelihood that you will contract the virus from someone else. It's not so much about keeping from getting other people sick, it's about keeping yourself from getting sick. As someone with young kids and a more frequent exposure to norovirus and other stomach bugs than I'd like to admit, once that gets into the family, everyone gets it. The way to avoid it is to not go places where your kid is going to pick it up (public library usually), but in general you still go to those places because they are convenient for other reasons. I understand the focus on handwashing and sanitizing things, but that doesn't do anything for direct exposure. If the stomach bug is near you, you're getting it. I assume coronavirus is going to be much the same.
klmadfejno 6y ago
I think this thread does a terrible job of explaining its point. I've no idea if it's accurate, but the premise seems to be that people who are highly likely to get the virus do more to spread it than people who are likely to connect with many people conditioned on getting the virus in the first place, with acknowledgements that the two labels are likely to co-exist. Just throwing it out there, I think this is likely going to be true, and the populations with higher proportions of old people are going to see greater infection rates than populations with fewer old people. That is to say, those who are vulnerable to the disease have a greater likelihood of getting other people sick than people who are naturally resistant to the disease but will behave irresponsibly. If you had a city with nobody over age 50, not only would the death rates be trivial due to the health of the population, but the spread of the virus would be very small. I also don't think the views here hold once a non trivial proportion of the population is infected, like we have now, so it probably doesn't matter and is hard to act on anyway.
baq 6y ago
it is interesting but it's also kind of common sense: of course people who contact a lot, whether they initiate the contact or not, are at risk. the thread gives some interesting results about models but epidemiologists, from what i hear, don't talk about superspreaders as in people but superspreader events which result in a lot of infected in a short time span. i don't think simple models like SIR capture these things at all.

war1025 6y ago

I read through most of that but will admit that I don't quite understand what he was trying to get at. Can you summarize?

dboreham 6y ago
I only read the beginning, but it seemed to be about transmission not being completely transitive.
baq 6y ago
it more or less says that people who get met a lot (...directed graph, remember) somewhat counter-intuitively increase the total percentage of people infected more than you'd expect even if they don't transmit the disease after they contract it more than an 'average' person (again, directed graph).
- war1025 6y ago
  I guess that doesn't seem counter-intuitive to me. That's the whole reason behind social distancing right? To reduce the likelihood that you will contract the virus from someone else. It's not so much about keeping from getting other people sick, it's about keeping yourself from getting sick. As someone with young kids and a more frequent exposure to norovirus and other stomach bugs than I'd like to admit, once that gets into the family, everyone gets it. The way to avoid it is to not go places where your kid is going to pick it up (public library usually), but in general you still go to those places because they are convenient for other reasons. I understand the focus on handwashing and sanitizing things, but that doesn't do anything for direct exposure. If the stomach bug is near you, you're getting it. I assume coronavirus is going to be much the same.

klmadfejno 6y ago

I think this thread does a terrible job of explaining its point. I've no idea if it's accurate, but the premise seems to be that people who are highly likely to get the virus do more to spread it than people who are likely to connect with many people conditioned on getting the virus in the first place, with acknowledgements that the two labels are likely to co-exist. Just throwing it out there, I think this is likely going to be true, and the populations with higher proportions of old people are going to see greater infection rates than populations with fewer old people. That is to say, those who are vulnerable to the disease have a greater likelihood of getting other people sick than people who are naturally resistant to the disease but will behave irresponsibly. If you had a city with nobody over age 50, not only would the death rates be trivial due to the health of the population, but the spread of the virus would be very small. I also don't think the views here hold once a non trivial proportion of the population is infected, like we have now, so it probably doesn't matter and is hard to act on anyway.

baq 6y ago

it is interesting but it's also kind of common sense: of course people who contact a lot, whether they initiate the contact or not, are at risk. the thread gives some interesting results about models but epidemiologists, from what i hear, don't talk about superspreaders as in people but superspreader events which result in a lot of infected in a short time span. i don't think simple models like SIR capture these things at all.