Posted in the Lounge as it is simply a thought.
A method to test many people with few tests:
say for a probability of 1% of infection cases, take a set of 7 samples from 100 people. Mix all the 1st samples in 2 groups of 50 people. Statistically, one group will be positive, one group negative. With 2 tests, you have already eliminated 50 people who have been tested negative. Separate the Pos. group in 2 groups of 25 people & use their 2nd samples. Again 1 group will be pos. the other neg. You have eliminated another 25 people with so far 4 tests (total 75 people). Continue like this until finding the single positive. You will need 14 tests for testing 100 people if my calculations are correct.
What do you think?