1. notimp

    notimp Well-Known Member
    Member

    Joined:
    Sep 18, 2007
    Messages:
    5,323
    Country:
    Laos
    Thats not the point you give up at. You find a region in your country thats roughly statistically representative of your entire country (on average), and you test the entire population. Then you end up at case fatality rates of about 0.375% (you never end up at 0.1% (fake news?)) but as explained before - that means - if 70% of the population have it - the 'everyone knows a family, where one person has died of Covid _every year_' case (four grandparents, eight parents, eight kids, with eight partners, ...) is in play.

    edit: Here as a discussion amongst mathematically minded people. :)
    https://statmodeling.stat.columbia....that-someone-you-know-will-die-from-covid-19/

    edit2: And the 'every year' part (without vaccination) is important. Because usually people you know dont die lets say 'of a carcrash' -- every year.


    Also all the values listed in the articles above - are absolute death numbers. So no Case fatality rate needed.

    You cant refute them with 'but case fatality rates - unknown', or 'small'.


    edit3: Also Case Fatality Rate in the US is actually estimated to be around 5% for whatever reason.
    https://www.mdpi.com/2227-9032/8/3/330/pdf

    Dont know why. No healthcare system worth the name? Overestimation?

    edit4: The first one rather, and hospitals being overwhelmed:
    https://www.cebm.net/covid-19/global-covid-19-case-fatality-rates/
     
    Last edited by notimp, Nov 25, 2020
  2. tabzer

    tabzer GBAtemp Maniac
    Member

    Joined:
    Feb 15, 2019
    Messages:
    1,486
    Country:
    Japan
    There is a moral problem with incentivizing hospitals to pump their numbers.
     
    wartutor and jimbo13 like this.
  3. notimp

    notimp Well-Known Member
    Member

    Joined:
    Sep 18, 2007
    Messages:
    5,323
    Country:
    Laos
    So lets do that calculation again at a Case Fatality Rate of 5%. Which apparently is the going rate in the US.

    Family: 4 grandparents, 8 parents, 8 children + 8 partners = 28
    28 - 30% (because lets say 70% of people in your country get it (not currently, but in the free spread horror scenario) every year, some with no symptoms) = 20

    20*5% case fatality rate = 1

    So what follows is: Dont let it spread to 70% of the population. Easier and cheaper than to lower the CFR rate to the rest of the civilized world, when not overwhelmed (hospitals operational) of about 0.4-1% - by building hospitals, paying for payed sick leave, and healthcare for everyone (problem with pandemics is - you cant combat it in just 10% of your population, when 70% of them have it).
    ---------------------------

    edit2: Just remembered the obvious. Those CFR percentages usually originate from 'number of known infected people/number of known deaths'. As you dont test the entire population - number of 'actual' infections is MUCH higher. (Vs. known deaths.)

    Which is why you select a 'statistically representative region' in your country - and test EVERYONE. That way you end up at an actual CFR of probably closer to 0.375%.

    So closer to 'at 70% infection rate' 'everyone knows a person who has died from it every year' rather than - 'every family has one death'.

    The calculation above should basically be invalid. What 'real' CFR for the US is - I dont know. For germany its around 0.375%. IF hospitals remain fully operational.
    -------------------------------


    edit: True CFR for the US (as in we tested and entire region, and compared with death figures) for the US is estimated to be 0.5%:
    https://www.motherjones.com/kevin-drum/2020/04/the-true-cfr-of-the-coronavirus-may-be-0-5-or-less/
    -------------------------------

    20*0.5% = 0.1

    So lets say every family knows 10 other families, so you "know" one person that has died from Covid _every year_ - IF 70% of your country get infected. (Which is what you are trying to prevent.)


    Next question to ask is: Is there really the possibility that 70% of people get infected within a year, with no measures taken? And this is where population density and 'near exponential' vs 'really exponential' comes in.
    At that point, _I_ am stopping my google fu. But you can go on further along those lines, if you want to.
    (Earlier problematic point is hospitals becoming overtaxed, at which point even 'real CFR' starts to climb as a percentage).


    But "we dont know if its 0.1% or 40%" is simply not true - and grossly misrepresenting reality.
     
    Last edited by notimp, Nov 25, 2020
  4. smf

    smf GBAtemp Psycho!
    Member

    Joined:
    Feb 23, 2009
    Messages:
    4,433
    Country:
    United Kingdom
    You just admitted to not reading his post.

    We were talking about fake registrations of voters and then making absentee ballots in those names. There has been no suggestion that Russia did that.

    Known deaths is likely to be closer to the correct figure. Known infections can only be close to symptomatic infections if people actually report it, and not wanting to be proven wrong about your conspiracy theory is a good argument for why you'd try to keep it a secret (some people keep it secret for financial reasons as they want to keep getting paid).

    The asymptomatic numbers are the issue, because those people are still able to infect people who then become symptomatic.
    And you can become symptomatic more than once, which increases your chances of dying (in one known case the second time was worse).

    So without wide spread testing, which people are nervous about because of lost income & conspiracy theories, then you can't know what the risk is going to be. The accuracy of the mass testing methods is also lower than what is possible using the slower lab tests.
     
    Last edited by smf, Nov 25, 2020
  5. shamzie

    shamzie Conform comrades
    Member

    Joined:
    Mar 19, 2014
    Messages:
    1,139
    Country:
    United Kingdom
    If you have a twitter account and you're tired of being jailed for wrongthink by our liberal tech overlords. Join parler. That it, thats the whole post.
     
    UltraSUPRA and gregory-samba like this.
  6. smf

    smf GBAtemp Psycho!
    Member

    Joined:
    Feb 23, 2009
    Messages:
    4,433
    Country:
    United Kingdom
    Everyone has a 100% chance of dying anyway, its the timing.

    I accept that for a small percentage of people their covid infection coincided with their time being up, but I doubt it's statistically significant. If someone is living with heart disease and they die after a covid infection, then the covid infection is the cause.

    If you have a twitter account and like lies then join parler, the echo chamber for your lies.

    https://people.com/human-interest/parler-app-compared-echo-chamber/

    It's just a bunch of outraged people daring each other to be more outrageous. Imagine the self harm you can inflict.
     
    Last edited by smf, Nov 25, 2020
    KingVamp and AmandaRose like this.
  7. Darth Meteos

    Darth Meteos Entertainer
    Member

    Joined:
    Jan 6, 2015
    Messages:
    1,247
    Country:
    United States
    imagine being such a cringelord that 250k people dying in your country alone isn't enough for you to listen to reason
    this guy would be on the titanic shouting that everyone's a pussy, it's just some cold water
     
  8. notimp

    notimp Well-Known Member
    Member

    Joined:
    Sep 18, 2007
    Messages:
    5,323
    Country:
    Laos
    I know, and I also followed that - logic wise - in my crude calculation examples.

    This is the reason why 'real CFR' falls down to 0.5% from the 'known cases' calculated CFR of 5% in the US.

    In the end, only 'real CFR' matters in societal impact statistical modeling.

    "Known cases CFR matters for hospitals, and supply calculations" (on part of the state or county).
     
    Last edited by notimp, Nov 25, 2020
  9. Bladexdsl

    Bladexdsl ZOMG my posts...it's over 9000!!!
    Member

    Joined:
    Nov 17, 2008
    Messages:
    19,284
    Country:
    Australia
    i can't wait to see the military dragging trump out of the white house when he doesn't concede! it's going to make one of the best world headlines to date :lol:
     
    IncredulousP and AmandaRose like this.
  10. wartutor

    wartutor GBAtemp Advanced Fan
    Member

    Joined:
    Dec 25, 2012
    Messages:
    505
    Country:
    United States
    Yes but those numbers are off due to the fact you cant tell how many would of died anyway due to heart disease or from the common flu. Lumping them all under covid fatalaties because they tested positive isnt the correct way to do it. Before this heart disease was the #1 killer in the US with obessity and resporatory high up there (expecially with the elderly that died from pnemonia due to covid and not covid directly. Aka would of died anyway if they caught the flu.)
     
  11. gregory-samba

    gregory-samba GBAtemp Advanced Fan
    Member

    Joined:
    Sep 11, 2020
    Messages:
    535
    Country:
    United States
    I just made reservations for Thanksgiving Dinner at a top notch restaurant. I'll be just like Newsnom, sitting at a table with no masks surrounded by a bunch of family members. This is exactly how I've been acting since the start of this pandemic and frankly, I probably already had it and didn't know it. I'm not scared of a simply not-so-deadly virus.
     
    jimbo13 likes this.
  12. smf

    smf GBAtemp Psycho!
    Member

    Joined:
    Feb 23, 2009
    Messages:
    4,433
    Country:
    United Kingdom
    Sure, if people want to keep their infection secret and then die then hospitals don't need to care about them.

    I may be wrong, but I don't think asymptomatic covid can cause death as it appears to be the symptoms that kill you. So in terms of hospitals they are less relevant in themselves. But it's possible that you could get an undetected surge in asymptomatic infections which causes a huge demand on hospitals if they mix with people who (for whatever reason) become symptomatic.

    I'm sure in the futures someone will try to figure out what makes people symptomatic or asymptomatic.
     
    Last edited by smf, Nov 25, 2020
  13. notimp

    notimp Well-Known Member
    Member

    Joined:
    Sep 18, 2007
    Messages:
    5,323
    Country:
    Laos
    High percentage of infected are symptomless. But still infectious.

    Here:

    [​IMG]
    src: https://www.cdc.gov/mmwr/volumes/69/wr/mm6915e3.htm

    Thats hospitalization rate - not testing rate per agegroup. Testing rate per agegroup is probably higher - but not that much higher. (Depending on agegroup, maybe 10x? (thinking about the younger demographics.)
     
    Last edited by notimp, Nov 25, 2020
  14. smf

    smf GBAtemp Psycho!
    Member

    Joined:
    Feb 23, 2009
    Messages:
    4,433
    Country:
    United Kingdom
    Everyone will die anyway, it's just a matter of what contributes to their death. A large number (but not all) of them have underlying health conditions.

    "Lumping" them all under covid fatalities is the correct way to do it.

    If someone had been given a week to live and then gets infected, or someone shoots them in the head shortly after a covid infection then I don't think they would be included in the figures. But even if they were then it's a small percentage that is not statistically significant.

    Right, but that is a separate group. They aren't keeping it a secret, there is no evidence of their infection.
     
    Last edited by smf, Nov 25, 2020
  15. notimp

    notimp Well-Known Member
    Member

    Joined:
    Sep 18, 2007
    Messages:
    5,323
    Country:
    Laos
    They dont show up in the 'tested covid cases' statistics, because they dont get tested. (Thats the group I'm talking about.)

    They are the reason, that you have high "known cases CFR" and low "real CFR".

    Nothing about 'keeping it a secret' (well, not necessarily - I'm not interested in motivation (people that have to work to get income from week to week)), dont need it for the argument.
     
    Last edited by notimp, Nov 25, 2020
  16. smf

    smf GBAtemp Psycho!
    Member

    Joined:
    Feb 23, 2009
    Messages:
    4,433
    Country:
    United Kingdom
    Right, but if you can't know about them then it's not worth worrying about them as you can't plan for a number you don't know.

    All you can do is come up with a reasonable worse case prediction of spikes in detected cases & hope you aren't too high or low in your guess.

    I would assume that detected and non detected cases go up and down at roughly the same rate. It might not be true, but how would we know?
     
    Last edited by smf, Nov 25, 2020
  17. g00s3y

    g00s3y Asshole
    Member

    Joined:
    Aug 6, 2008
    Messages:
    702
    Country:
    United States
    165 pages of Trump supporters & anti maskers proving just how fucking stupid they are.

    It's great.
     
    IncredulousP, Ektif and AmandaRose like this.
  18. gregory-samba

    gregory-samba GBAtemp Advanced Fan
    Member

    Joined:
    Sep 11, 2020
    Messages:
    535
    Country:
    United States
    That depends on your point of view. I found it to be 165 pages of dumb Liberals proving just how stupid they are.
     
    wartutor likes this.
  19. tabzer

    tabzer GBAtemp Maniac
    Member

    Joined:
    Feb 15, 2019
    Messages:
    1,486
    Country:
    Japan


    Anyone want to place bets?
     
  20. notimp

    notimp Well-Known Member
    Member

    Joined:
    Sep 18, 2007
    Messages:
    5,323
    Country:
    Laos
    No, no, no... :)

    Here is how statistical sampling works. As a whole. Take 500 people. Make sure they are 'normal in distribution' (according to some criteria (age, income, health, population density..). Ask them questions. Then say "the same would be true for the entire population".

    If you do it with 5 people - it doesnt work. If you do it with 500 it kinda does. But you maybe arent that certain. If you do it with 5000 you are pretty certain. (Many issues alongside of 'stuff can change' (predictions are hard), or 'people not telling you stuff openly').
    --

    But if you only start planning, when the problem is _there_ you are dead, dead - when it comes to a problem that has exponential growth rate.

    So you use statistical models.

    How reliable they are and how accurate and, .... all valid questions. But you start with preparation in the 'everything is fine' state - 'knowing about the nature of the problem'.

    (Virus isnt going to kill more people - just because its monday. If everything impacting death criteria doesnt change much (hospitals stay operational). Virus isn not going to infect less people, if you do nothing.
    (Actually it is - because of only near exponential growth (at some point it plateuas), and stuff like herd immunity (but the issue is, that if you get there too quickly, all your hospital infrastructure is going to fail)). )
    ---

    'Exponential' is this hockey stick curve thingy:
    [​IMG]

    That comes as a result of 'on person "converting" tree others" - in a short period of time'.
     
    Last edited by notimp, Nov 25, 2020
Draft saved Draft deleted
Loading...

Hide similar threads Similar threads with keywords - president, Becomes, United