Neverending Covid-19 Coronavirus

MA just made it a law today that super markets and pharmacies must dedicate at least an hour a day for people over the age of 60 only.

We have that in FL too. Problem is that the stores open for the elderly for the first hour, so guess what isn’t there for everyone else.....
 
I went golfing today.

Paid online. Checked in outside by showing a confirmation. Used a scorecard app on my phone. Walked and carried my own clubs. It was nice.

EDIT: Oh, and the rule was - put it NEAR the hole, but not IN the hole, and if you do, DON'T TOUCH THE PIN.
 



Essentially 1/2 of people who test positive never show any symptoms. This drastically cuts back projects of worst case scenarios.
Proposal that we don’t uncritically repeat The Daily Wire:
 
The San Francisco subway is closing until the shelter in place order is lifted. The effected routes will be covered by above-ground transit services.
 
Proposal that we don’t uncritically repeat The Daily Wire:

I agree - you should always look at data critically. Models provide results using the best available information at the time. When new information becomes available, you need to run it again. It doesn't make the model wrong. It's especially important when you're looking at a pre-print manuscript like this was. The public and the media don't always realize that these things are a work in progress. The main conclusion that "social distancing" is the best approach should stand despite that change in input data.

In this case, the fact that 50% of the population is asymptomatic was not clear even a week ago. That affects things in two ways - it lowers the death rate but it also increases the Ro quite a bit. It will be interesting how that changes things. Still a lot of deaths in the "do nothing curve" I'm sure, but I suspect it's less and the curve should be tighter.

There is criticism that the effects of testing & tracing wasn't considered but that's irrelevant to the "do nothing" curve. It's just another simulation that must be run.
 
Anyone in need of TP and don’t want to leave their homes check out this site which automatically shows several kinds toilet paper when it goes back in stock on Amazon. Amazon recently removed the “notify me” option, probably because people abused it for essential goods.

Was not able to get any from the very kind Instacart delivery person today. So I’m trying my luck here. Prices are normal and not gouged. I’ve got 5 left so ideally getting some sooner than later. I only plan on purchasing one pack if I can get any.

 
Anyone in need of TP and don’t want to leave their homes check out this site which automatically shows several kinds toilet paper when it goes back in stock on Amazon. Amazon recently removed the “notify me” option, probably because people abused it for essential goods.

Was not able to get any from the very kind Instacart delivery person today. So I’m trying my luck here. Prices are normal and not gouged. I’ve got 5 left so ideally getting some sooner than later. I only plan on purchasing one pack if I can get any.

This works FYI! Just scored a 30 pack for $20. Doesn’t go in stock until the 29th but it’ll come in two days after that 💩
 
For me it says everything is out of stock. The 2 that are listed as pre-order bring me to the amazon homepage and not the product page so yeah.

No luck with me. But at least I have 3 spare rolls still.
 
I agree - you should always look at data critically. Models provide results using the best available information at the time. When new information becomes available, you need to run it again. It doesn't make the model wrong. It's especially important when you're looking at a pre-print manuscript like this was. The public and the media don't always realize that these things are a work in progress. The main conclusion that "social distancing" is the best approach should stand despite that change in input data.

In this case, the fact that 50% of the population is asymptomatic was not clear even a week ago. That affects things in two ways - it lowers the death rate but it also increases the Ro quite a bit. It will be interesting how that changes things. Still a lot of deaths in the "do nothing curve" I'm sure, but I suspect it's less and the curve should be tighter.

There is criticism that the effects of testing & tracing wasn't considered but that's irrelevant to the "do nothing" curve. It's just another simulation that must be run.

In my particular modeling community we have a saying (maybe all modeling disciplines have it - i don't know)...
all models are wrong but some models are useful
why? because they're only as good as the data, assumptions, and conceptual models that fuel them

that doesn't mean they don't provide us with useful information... they're just not crystal balls or nostredamus as some people believe
 
For me it says everything is out of stock. The 2 that are listed as pre-order bring me to the amazon homepage and not the product page so yeah.

No luck with me. But at least I have 3 spare rolls still.
You just have to keep checking and checking and refreshing that page. Try all the links that say either Pre-Order or in stock. It updates every 5 minutes so check back often.
 
In my particular modeling community we have a saying (maybe all modeling disciplines have it - i don't know)...
all models are wrong but some models are useful
why? because they're only as good as the data, assumptions, and conceptual models that fuel them

that doesn't mean they don't provide us with useful information... they're just not crystal balls or nostredamus as some people believe
I model buildings. I won't say my models are never wrong but they are so fucking useful that if they were wrong it would cost 10's to 100's of thousands of dollars for the smallest mistake. I would concede that there is a level of detail we never try to approach with building models because the payoff for the effort does not exist (e.g.we model a dimensional window but not the composition and flashing system of said window). That said, as model authoring software increases in sophistication and the ability of computer maintenance and ROI analysis system grows, more and more detail is showing up in construction models as a single point of reference to feed those systems.

My models are no less models of the type you speak of. The current vernacular refers to them as digitial twins and we literally plan construction operations virtually before committing to again building the real thing in meat space. The difference I think is that the purpose, utility, and scope of my models are largely understood across the board for those who participate in it.
 
I model buildings.

My models are no less models of the type you speak of. The current vernacular refers to them as digitial twins and we literally plan construction operations virtually before committing to again building the real thing in meat space. The difference I think is that the purpose, utility, and scope of my models are largely understood across the board for those who participate in it.

I think the inputs to your models are more well-understood in that there is less variability in some, many?, of the factors. They're no less complicated but perhaps the factors are more static?

I'm not an epidemiologist so I can't speak to their models as an expert, but my assumption is that human behaviors / choices are more difficult to account for and there is a heavy reliance on past trends without fully understanding the human behaviors that may be driving them, even with large sample sizes.

As far as the natural sciences I'm more familiar with there are a number of phenomena that you just can't see with your own two eyes easily because they are too big or too small or happen so quickly or slowly that they make modeling results more variable and perhaps more dependent on the quality and quantity of the observations you do have...but they people who use them do understand.... the general public and too often the people responsible for making decisions don't because science literacy and education are poor, and technical people do a poor job of communicating it. It's pretty difficult to have a digital twin of something that is constantly changing with time.
 
Back
Top