But can you take into account not looking both ways on before crossing the road when you've drunk a bit too much? What you have stated are variables, and the whole idea of a scientific study is to eliminate as many variables as possible.