Something doesn’t make sense. The only way the positivity ratio can point toward the prevalence ratio is if tests are being conducted randomly with regard to symptoms. If most of the tests are being done on sick people, the positivity ratio should be high. Even 20% sounds low. At the very least, shouldn’t the positivity ratio should be adjusted for the “tests per 1,000,000” population ratio? In other words, doesn’t S. Korea have a low positivity ratio because it has such a high tests per 1,000,000 ratio? The US did more tests on April 15 than March 15 because there were more sick people in April. It’s not a deeper dive into the same jellybeans; it’s a deeper dive into a bigger jar. What am I missing?