Summarized as: “Any statistic that appears interesting or different is usually wrong”