4 ms·

Maybe its not that accurate to count it that way, but 1 in 10,000 would mean to the layman that every 10,000 times you do this you should expect 1 case of C19,

by cortic 6y ago

Maybe its not that accurate to count it that way, but 1 in 10,000 would mean to the layman that every 10,000 times you do this you should expect 1 case of C19, and i do this at least 720 times at every delivery. With more than 20 deliveries through the pandemic last year i have potentially touched 14,400 contaminated items which would practically guarantee a case of C19. Now i know not all of these would be contaminated, but i don't know which or how much. And since delivery drivers in my area rarely wear masks even at the house, i don't have a high amount of confidence in any delivered items. -To be honest i massively underestimated how often i touch some of these items in the first few days, coffee jars alone get touched 40+ times, jam and butter 12+, bread 18+, diluting squash 14+.

0x72 6y ago
It's around 23.69% if my math is correct. I mean, it's not nothing, but I wouldn't say "practically guarantee".
- cortic 6y ago
  A 23% chance of a 1 in 10,000 chance thing happening every 14,400 chances? That can't be right.
  - 0x72 6y ago
    huh, 76%... my bad, forget I said anything. p(a') = 1 - 1/10000 1 - p(a')^14400 = 0.76308930063962
bgarbiak 6y ago
If you touch a parcel a second time then the first time doesn't count to anything. It's 1/10k again. It's like with a lottery: you can buy a coupon every day for years and win nothing, because every draw resets your chances.
- cortic 6y ago
  48 million to one lottery odds, running everyday I'd expect to win once in the next 131,506 years (48m/356). -this is why you don't win for 'years'. I'd also expect one winner for every 48 million players in a day. Now a 1/10k taken twice is 2/10k, so taken 14 thousand times it is 14k/10k. Okay so this isn't technically how you add statistics, but its close enough to see the utility in washing items.
  - bgarbiak 6y ago
    The are lotteries with bigger chances, but in all of them some numbers will be repeated over the years, and yours may never occur. You would have to use all possible combination for one draw to be sure. That's why it's not unusual to have draws without winner, sometimes a few in a row. It's a flawed analogy, I admit, but with the parcel you would have to palpate its entire surface and lick the finger after every touch to infect yourself. Just touching it for the second time doesn't double the chance of infection. You may touch the same uncontaminated part you'd touched already (crossing the same numbers twice), the number of virus particles you've moved might not be enough to infect you, that you might have washed your hands in between the touches, and so on... Obviously, the more you "interact" with an item the chances grow, but certainly not as fast as you suggest. Add a time limit to all of that, and the fact that for every next item it's all back to 1/10k (second draw) - practically there's no way this number will ever grow over 1/1. Excuse the non-fact-checked argument as I can't find a link now but I recently read that couriers and mailmen actually get ill less often than factory workers (and most of them caught the virus via airborne transmission). I don't mind if you wash everything you bring home from outside, but I don't see point in that, and certainly I wouldn't recommend that to anyone. It's wiser to focus on preventing airborne transmission than to waste time and efforts on avoiding the least probable risk.

by cortic 6y ago

0x72 6y ago

It's around 23.69% if my math is correct. I mean, it's not nothing, but I wouldn't say "practically guarantee".

cortic 6y ago
A 23% chance of a 1 in 10,000 chance thing happening every 14,400 chances? That can't be right.
- 0x72 6y ago
  huh, 76%... my bad, forget I said anything. p(a') = 1 - 1/10000 1 - p(a')^14400 = 0.76308930063962

bgarbiak 6y ago

If you touch a parcel a second time then the first time doesn't count to anything. It's 1/10k again. It's like with a lottery: you can buy a coupon every day for years and win nothing, because every draw resets your chances.

cortic 6y ago
48 million to one lottery odds, running everyday I'd expect to win once in the next 131,506 years (48m/356). -this is why you don't win for 'years'. I'd also expect one winner for every 48 million players in a day. Now a 1/10k taken twice is 2/10k, so taken 14 thousand times it is 14k/10k. Okay so this isn't technically how you add statistics, but its close enough to see the utility in washing items.
- bgarbiak 6y ago
  The are lotteries with bigger chances, but in all of them some numbers will be repeated over the years, and yours may never occur. You would have to use all possible combination for one draw to be sure. That's why it's not unusual to have draws without winner, sometimes a few in a row. It's a flawed analogy, I admit, but with the parcel you would have to palpate its entire surface and lick the finger after every touch to infect yourself. Just touching it for the second time doesn't double the chance of infection. You may touch the same uncontaminated part you'd touched already (crossing the same numbers twice), the number of virus particles you've moved might not be enough to infect you, that you might have washed your hands in between the touches, and so on... Obviously, the more you "interact" with an item the chances grow, but certainly not as fast as you suggest. Add a time limit to all of that, and the fact that for every next item it's all back to 1/10k (second draw) - practically there's no way this number will ever grow over 1/1. Excuse the non-fact-checked argument as I can't find a link now but I recently read that couriers and mailmen actually get ill less often than factory workers (and most of them caught the virus via airborne transmission). I don't mind if you wash everything you bring home from outside, but I don't see point in that, and certainly I wouldn't recommend that to anyone. It's wiser to focus on preventing airborne transmission than to waste time and efforts on avoiding the least probable risk.