"It uses a broken methodology to measure the internet and is, for reasons unclear to anyone, regarded as somehow definitive simply because it allows you to compare two sites with a single simple number."<p>As opposed to IQ, GPA, SATs, LOC, Meyers-Briggs, Megapixels, Megahertz, Merit Badges, Horsepower, US News Rank, Karma points, etc.?<p>Someone should really write down some theory on what can be measured and what can't be measured. You'd think we'd have some idea after 2,500+ years of philosophy, but it doesn't much seem that way.
He is absolutely right: confounding is explained in the first chapter of any book on statistics. Alexa's stats do not fall into "randomized controlled" category of studies, therefore they are pretty much guaranteed to be wrong.