benaj
Member of DD Central
N/A
Posts: 5,619 
Likes: 1,741
|
Post by benaj on Jan 6, 2022 13:39:25 GMT
For those planning to get covid intentionally from a complete stranger and enhance “immunity”, check this out.
|
|
adrianc
Member of DD Central
Posts: 10,019
Likes: 5,147
|
Post by adrianc on Jan 6, 2022 14:04:28 GMT
Even larger, that's not the most legible of colour combinations...  |
|
Mike
Member of DD Central
Posts: 651
Likes: 446
|
Post by Mike on Jan 6, 2022 14:13:44 GMT
For those planning to get covid intentionally from a complete stranger and enhance “immunity”, check this out. If you adjust the numbers for a country with more like 10% infections rate (the UK) then the probabilities work out such that there is no need to 'plan to get covid intentionally', it's extrememly likely regardless. Speaking to someone indoors in a well ventalated area with low occupancy with masks for a prolonged period is 19.1%. Do that on just three occassions and your chance is ~50/50 already (well, 47%).
|
|
adrianc
Member of DD Central
Posts: 10,019
Likes: 5,147
|
Post by adrianc on Jan 6, 2022 14:23:04 GMT
If you adjust the numbers for a country with more like 10% infections rate (the UK) then the probabilities work out such that there is no need to 'plan to get covid intentionally', it's extrememly likely regardless. Speaking to someone indoors in a well ventalated area with low occupancy with masks for a prolonged period is 19.1%. Do that on just three occassions and your chance is ~50/50 already (well, 47%). Eh? Prolonged speaking, indoors/well ventiated, low occupancy, masks = 2.1%. Not 19.1%. Do that on three occasions, your risk is 2.1% from each of the three. Not 6.3%.
|
|
Mike
Member of DD Central
Posts: 651
Likes: 446
|
Post by Mike on Jan 6, 2022 14:25:04 GMT
If you adjust the numbers for a country with more like 10% infections rate (the UK) then the probabilities work out such that there is no need to 'plan to get covid intentionally', it's extrememly likely regardless. Speaking to someone indoors in a well ventalated area with low occupancy with masks for a prolonged period is 19.1%. Do that on just three occassions and your chance is ~50/50 already (well, 47%). Eh? Prolonged speaking, indoors/well ventiated, low occupancy, masks = 2.1%. Not 19.1%. The 2.1% is for a population where 1% are infected. In the UK, it's more like 10%, so 2.1% -> 19.1% assuming it works like that (I can't get past the paywall) If you look at what I have done you will see that I have not made the error you suggest. Though FYI for small x; (1-x)^n goes like 1-nx so in your example where x is 0.021 the approximation of 0.063 is not that far from the real probability which is 0.06168... Taking the next term in the expansion n(n-1)x^2 / 2! gives 3*2*0.021^2 /2 = 0.001323 which, subtracted from the 0.063 first order approximation, is enough to get the same level of accuracy as the computed value of 0.06168. So even though x (=0.021) is not that small, a second order approximation turns out to be pretty good
|
|
agent69
Member of DD Central
Posts: 6,048
Likes: 4,438
Member is Online
|
Post by agent69 on Jan 6, 2022 14:42:48 GMT
Even larger, that's not the most legible of colour combinations... This is another one of those things that ask more questions than it answers. So if I am trying to use this table to assess risk:
- how is occupancy measured and where is the divide between low and high
- ditto levels of ventilation
- ditto short or prolonged time
- what constitutes contact
- are these the risk for vaccinated or un-vaccinated people
|
|
benaj
Member of DD Central
N/A
Posts: 5,619
Likes: 1,741
|
Post by benaj on Jan 6, 2022 14:48:45 GMT
In a nutshell, the louder the noise people making in a closer space, the better it is. 😅
|
|
|
|
Post by Deleted on Jan 6, 2022 14:54:16 GMT
Boris is doing anything to avoid mandated vaccines, that would split the party.
Now if only there was an opposition?
It would also split the country. well only 10:90 and 10% don't really care they are just worried about what Kylie's friends brother said happened to a cousin's balls.
|
|
|
|
Post by stevepn on Jan 6, 2022 15:35:46 GMT
Even larger, that's not the most legible of colour combinations... This is another one of those things that ask more questions than it answers. So if I am trying to use this table to assess risk:
- how is occupancy measured and where is the divide between low and high
- ditto levels of ventilation
- ditto short or prolonged time
- what constitutes contact
- are these the risk for vaccinated or un-vaccinated people
Basically this is meaningless.
|
|
daveb
Member of DD Central
Posts: 253
Likes: 210
|
Post by daveb on Jan 6, 2022 16:21:06 GMT
I'd say: unless you are prepared to be a complete recluse, you'll catch it it is at least plausible that you'll be sicker with a higher infecting dose than a lower one, so wear a mask and open a window and get vaccinated unless you are batshit crazy
|
|
|
|
Post by bernythedolt on Jan 6, 2022 16:48:36 GMT
If you adjust the numbers for a country with more like 10% infections rate (the UK) then the probabilities work out such that there is no need to 'plan to get covid intentionally', it's extrememly likely regardless. Speaking to someone indoors in a well ventalated area with low occupancy with masks for a prolonged period is 19.1%. Do that on just three occassions and your chance is ~50/50 already (well, 47%). Eh? Prolonged speaking, indoors/well ventiated, low occupancy, masks = 2.1%. Not 19.1%. Do that on three occasions, your risk is 2.1% from each of the three. Not 6.3%. I believe Mike 's conversion from 2.1% to 19.1% follows these lines... Probability of remaining uninfected with 1% background prevalence (as per the Tweet) = (100 - 2.1)% = 97.9% = 0.979 Probability of remaining uninfected with 10% background prevalence (as per UK today?) = 0.979^10 = 0.809 = 80.9% Leading to probability of infection of (100-80.9)% = 19.1% The original model in the tweet is far more complex, and this approximation may or may not be valid, but it doesn't seem unreasonable. If you accept the 19.1% as reasonable, three separate occasions at this risk would lead to... Probability of remaining uninfected = (100 - 19.1)/100 cubed, or 0.529. So, probability of being infected = 0.471, or 47% as stated. The binomial expansion is great to see, if less intuitive to the non-mathematician! 
|
|
keitha
Member of DD Central
2024, hopefully the year I get out of P2P
Posts: 4,593
Likes: 2,624
Member is Online
|
Post by keitha on Jan 6, 2022 16:48:48 GMT
It would also split the country. well only 10:90 and 10% don't really care they are just worried about what Kylie's friends brother said happened to a cousin's balls. Unfortunately that 10% group seems to be proportionately far more BAME than WASP
|
|
|
|
Post by moonraker on Jan 6, 2022 16:56:57 GMT
Glancing at the newspaper shelves during my Big Shop this morning, I spotted one headline that forecast we would be "back to normal" in three weeks' time. Which puzzled me. I opted for a different newspaper (mainly because it comes free with my Big Shop) but have just Googled. The pith of the Daily Mail article is behind a paywall, but the claim was made by Boris Johnson in another fit of bombastic optimism. I am very sceptical, though I suppose it does depend on what he means by "normal" and what I understand it to be.
|
|
adrianc
Member of DD Central
Posts: 10,019
Likes: 5,147
|
Post by adrianc on Jan 6, 2022 17:14:23 GMT
Eh? Prolonged speaking, indoors/well ventiated, low occupancy, masks = 2.1%. Not 19.1%. The 2.1% is for a population where 1% are infected. In the UK, it's more like 10%, so 2.1% -> 19.1% assuming it works like that (I can't get past the paywall) If you look at what I have done you will see that I have not made the error you suggest. Though FYI for small x; (1-x)^n goes like 1-nx so in your example where x is 0.021 the approximation of 0.063 is not that far from the real probability which is 0.06168... Taking the next term in the expansion n(n-1)x^2 / 2! gives 3*2*0.021^2 /2 = 0.001323 which, subtracted from the 0.063 first order approximation, is enough to get the same level of accuracy as the computed value of 0.06168. So even though x (=0.021) is not that small, a second order approximation turns out to be pretty good I'm with you. However, if we're working on 1% of the population have it, then the figures make zero sense. If 1% have it, then... if I go to the gym tomorrow, I have a 58% chance of catching it if I go to singing rehearsal next week, I have a 30% chance of catching it. So if the population is more like 10%, then...? Well, nope. I don't buy those figures for one second. MASSIVE over-simplification. The chance of there being anybody +ve at those events is practically zero. There's maybe half a dozen other people at the gym, and I'm not physically very close to any of them. There's about a dozen at rehearsal, and I know and trust them all to be sensible.
|
|
|
|
Post by bernythedolt on Jan 6, 2022 17:41:43 GMT
The 2.1% is for a population where 1% are infected. In the UK, it's more like 10%, so 2.1% -> 19.1% assuming it works like that (I can't get past the paywall) If you look at what I have done you will see that I have not made the error you suggest. Though FYI for small x; (1-x)^n goes like 1-nx so in your example where x is 0.021 the approximation of 0.063 is not that far from the real probability which is 0.06168... Taking the next term in the expansion n(n-1)x^2 / 2! gives 3*2*0.021^2 /2 = 0.001323 which, subtracted from the 0.063 first order approximation, is enough to get the same level of accuracy as the computed value of 0.06168. So even though x (=0.021) is not that small, a second order approximation turns out to be pretty good if I go to the gym tomorrow, I have a 58% chance of catching it Who are you trying to convince that you spend a 'prolonged time' in the gym? Nice try... 
|
|
