SailingSperm

Would getting fluid samples from 2 groups of say 50 people, combining the samples and testing that, then keep halving and testing recursively the groups that return positive, be a viable way of testing large sums of…