SoupNazi2001 said:
BohunkAg said:
KlinkerAg11 said:
Aren't rapid tests really inaccurate?
Yes. Wildly.
Most are 93% accuracy. Hardly highly inaccurate.
Bayes theorem coming in hot! 93% is more than bad enough to overwhelm a city with false positives if testing is done on a large enough scale. Random samplings of coronavirus should show an infected rate of like .01% of our population? Its a super low number. This isn't the spanish flu where 1/3 people got it. Run this test randomly* 100,000 times and your false positive rate will overwhelm the number of those actually with it.
People actually with corona = .001*100,000=
100 People with a false positive of corona = .07*100,000=
7,000This is a vast oversimplification, but this is why tests need to be 99.9+% accurate. 93% accuracy means more than 1/20 people are getting a false positive. Thats dog*****
*we do not have random sampling. We've got high risk people self selecting into testing. And even then we are still way way down.