Imagine that you’ve got a disease that strikes one in a million people, and a test for the disease that’s 99% accurate. You administer the test to a million people, and it will be positive for around 10,000 of them – because for every hundred people, it will be wrong once (that’s what 99% accurate means). Yet, statistically, we know that there’s only one infected person in the entire sample. That means that your “99% accurate” test is wrong 9,999 times out of 10,000! – Cory Doctorow

I was thinking the other day that web stats tools have gotten stuck in a corner. Most of the tools follow trends in a fairly sterile manner, making it difficult to really understand what you care about. They offer tables and time-ordered graphs, but they fail to answer a web-meister’s basic questions. Where’s the money? What are people interested in? The current tools seem to splat it all at the same amplitude, missing out on the really interesting trends. And where are the blinking red lights? – me, May ‘07

Google explores the design of over a billion web documents. Nice to see mountains of data put to uses other than search (and it’s interesting to see what developers are doing).