@jeffcliff it does provide a counterexample (laziness in determining which variables are worth considering) and there is another one - the kind of 'schrodinger-like' error introduced where you change the system by observing it, esp of humans