debug
vcut debug
Every method here is cheap to run. None of them is the obvious one.
That is the whole reason this file exists: for each question below there is an instrument that looks more rigorous and proves nothing, and reaching for it is how confident wrong answers get written down. Each entry names the question, the method, and the trap.
The rule under all of it
A measurement that returns the same thing whether or not your hypothesis is true has not tested anything. Before running one, ask what its output would be if you were wrong. If the answer is "the same", pick a different instrument.
Two of the traps below were rediscovered while building the commands that replace them, by people who had already written the warning. Knowing the rule is not the same as applying it under a confusing symptom.