You clearly do not understand sampling theory.
Here is a simple sample size calculator. To get a 5% margin of error at a 95% confidence level for a population of 300 million, you only need 384 samples. A 99% confidence level requires 666. A 2% confidence interval at 95% only requires 2400.
Many people think sample size rises linearly with population. That’s not true. Reduce the population to 30 million and you get the same sample size.
Getting a sample that isn’t biased in some way is much harder and more important.
Because the virus has hot spots, dead spots and light spots would this make unbiased sampling more difficult. Would the extra difficulty mean more sampling was needed??
Again, tossing more uncertainty into the mix is not friendly to your position.
Something like this one.
I don’t think it’s so ideological; it’s just “Something bad is happening…I should be angry at those people again”.
I think the conservative right has struggled to think of where our rage is supposed to be directed. It obviously can’t be at the federal government.
You can’t call it a hoax because it’s clearly not.
Antifa?
Erm…The bad guy is the one making you stay at home! They’re not letting meat factories produce delicious meat! Stay angry, and that might…help…I guess?
It has nothing to do with having a position. The position should be flexible as more info come in. My position is just based on what I see now. There does seem to be some pretty solid patterns developing.
Generally, a 1,500 sample is going to give you an error of +/- 2% in the population of the US. Here is the first article I pulled on sample size. So to take the South Korea numbers at face value (not quite a random sample but if anything biased to having a higher percentage than the general population) and they have about .1% +/- 1% of their population infected so their number of infected could range from 1.1-0.02% or 564K people infected down to 10,800. The their death rate is probably between 0.04% and 2%.
That isn’t the narrowest of ranges but it’s got a decent amount of confidence around it. So to revisit my initial point after actually doing the math (though only a single day not the series that could narrow the range)
Worst-case scenario is 2 in every 100 die and best-case scenario is 3 in 6,000 die. His best case actually wasn’t far off but the worst case doesn’t match the data. As I said earlier due to the ROK not having random outbreaks I think they are catching most of the people in their testing and so the worst case number is much more likely than the best case number.
Another issue I have is when they talk about flattening the curve. I have no idea what the capacity of hospitals in Ca has for example but lets suppose we say we would like to get 60% through this in one year. We project that California has the potential for a 1000 per million death rate and 10 times that many will wind up in the hospital for 30 days. That would come out to about 20,000 people in the hospital at any given time. Could California handle 20,000 patients?? 10,000??. What is our capacity? What ever that is shouldn’t we strive to stay open enough to at least be reasonably close to those numbers?
What I have seen is that after the uncertainty, better data is telling the modelers that opening the economy with few masks, little to no contact tracing or little testing will lead to more deaths, once again, it is more likely that the increase in deaths will shoot down the confidence from the people, that confidence that is needed to open the economy more will not be there when less restrictions are demanded.
I think many of us see this as a natural disaster like an earthquake or giant storm. The only difference is that it is slower and we have time to react. Will our fear prevent us from acting rationally or will it simply force us to keep pushing it back while absorbing more damages until we can no longer hold it back at which time it will take off on its own anyway. Now that they have us scared shitless how can they unscare us so we can go back to living at at least some semblance of normalcy.
It appears that the State of California has 27K hospital beds and ~2,000 ICU beds. What percentage is in normal use by non-COVID patients I’m not certain but I’m sure you could dig up utilization rates.
IHME, the tool mentioned in the site referenced by GIGOBuster, has a listing for ICU beds available. For California it is listed as just under 2000. Total hospital beds around 27,000.
I’ve been down this path on this board and it was pointed out that trying to maintain a certain percentage occupancy rate of hospital beds is a dangerous game. It wouldn’t take much to overwhelm the system when there are only 2000 to start. And that is in all of California.
From what I can find as of march 20th we had 88,000 beds with many more being added on so by now we should have close to 100,000 if you figure in areas not normally used for hospital beds but can be utilized. Not sure how many the medical professionals could handle but it looks like 20,000 beds could be doable in a pinch. As for being a dangerous game, it is certainly dangerous but it is not playing a game.
Surely there is a geographic aspect. If one did all the testing in Claredon, AR you would have a very different result than if your data come from NY, NY.
Indeed, it is not just beds, but enough medicines, medical equipment, besides the usual shortages of nurses and doctors reported even before the pandemic.
I don’t care what you want to call it. There isn’t going to be a lot certainty to it. You are likely to get a lot of pushback around here for those types of opinions.
Something I am curious about that maybe only a Dr. could answer. If a person is nearing the end of their life and they catch the flu and die will they be labeled a flu death? If 20 patients were in a nursing home and nearing the end of their life and the flu finished them off would they be labeled as flu victims? I know the corona virus it is being used pretty liberally.
Is there a pathologist or coroner in the house?
You “know” - from where? If anything, COVID isn’t blamed enough; search news on covid deaths undercounted for many reports. The US population isn’t being tested (because political folly) and all corpses aren’t tested or autopsied so we don’t really know who dies of COVID directly, or indirectly as from secondary infections, or from lack of medical care. When excess death rates for the COVID era are calculated, we’ll have a better picture of its impact. Till then, listen to immunologists and epidemiologists, not to self-serving political hacks.
No, better sampling.
The classic example is the Literary Digest fiasco in 1936. The did a poll for each election, and in 36 predicted that Landon would win. They had 2.2 million responses, way more than needed. Way, way more. But their sample was based on subscribers, car owners and telephone owners, all of whom were richer than average and thus more likely to vote Republican. Especially against FDR.
Besides the bed question, I don’t think it is possible to tune the opening to the number of cases. Geographically, perhaps. If you open one restaurant you have to open all of them, or else give that one restaurant a big advantage.
Do you think anyone can tune the reopening to not potentially overwhelm the system?
I agree it is basically a natural disaster. People in hurricane territory board up their windows. Only idiots would object that it it hurts their constitutional right to look outside.
I lived in tornado country once. I never heard an alert, but I knew where the basement was. If tornadoes came one after another you’d stay in the damn cellar way longer than you would want to. That’s more or less what we’re doing.
And thanks to people around here obeying the rules, I don’t feel scared shitless when I go to the grocery once a week, even though I’m over 65. If I was in one of the states that are opening before they should, I’d be a lot more worried.