How do you investigate hypotheses? Do you seek to confirm your theory - looking for white swans? Or do you try to find black swans? I was startled at how hard it was for people to investigate number sets that didn't follow their hypotheses, even when their method wasn't getting them anywhere.