Hacker News Reader

You could go a long way with the standard scientific method - make predictions, do measurements, compare.

Compare it to what? Compare it to assumptions? How do you test it? How would you know it's correct?