jmccormac - 2:40 am on Aug 19, 2012 (gmt 0)
There are some parallels with the "mythical" Google sandbox in the first claim. But most search engines would require a sequence of crawls of a site/document to build an efficient crawl schedule. (It is as waste of resources spidering and indexing a document that only changes once a year.) It is the assumption that continually modified pages are generally modified to gain page rank that is wrong. From what I've seen of building a small country level search engine, most of the webpages and websites are brochureware that do not change dramatically over a period of a few months. SEOed pages change but this flawed approach from Google is also going to flag A/B testing.