We have a natural tendency to reject mistakes. When we experiment, we consider unexpected results as errors (wrong method, wrong processing…), as noise that must be eliminated rather than as signals of a possible clue. That’s a mistake…