crobb305 - 8:03 pm on Apr 16, 2011 (gmt 0)
I really shouldn't have bothered with the post.
This is a good post! Don't get frustrated. This discussion gets us thinking about the numerical/statistical aspects versus conspiracy and single-factor theories. Google has a over a decade's worth of data, patents, equations, filters, billions of documents on the web, enormous language/semantic datasets, etc... With so much data and experience, it's evolving into a complex exercise in statistical modeling at this point.