[texthompson]

I thought exactly the same thing. Statistics is about uncertainty, and it's very easy to be misled when you don't correct for trying lots of hypotheses.