If the evidence seems to disagree with your theory, maybe your theory has to change?