WebmasterWorld administrator, Tedster, posted a thread at WebmasterWorld discussing an updated Google patent named Information retrieval based on historical data. This is one of the more popular Google documents over the years, where much of the Sandbox theories came from.

In any event, Tedster pulled out several abstracts that are new in this document. I will highlight only two that I find would be very valuable to our readers.

(1) How does Google know when a site has changed enough where they should drop all the past trust and link popularity associated with that site?

So it is not just about changing the domain name registration information. That is why many folks who buy sites, try to keep the same style and category of content on that domain.

(2) We heard it before, “Don’t get links too quickly” because it seems unnatural. Well, here it is on paper:

Yes, for most sites, you don’t get 50,000 links overnight. But for some sites, it is possible for several reasons. So how does Google determine which sites naturally received these links so quickly? Well, if I understand this correctly, they look to see how quickly those links go away and the “dynamic-ness of the links.” Here are those explanations from the document:

Tedster goes a bit deeper into signs of the old supplemental index, which I did not go into over here.

Forum discussion at WebmasterWorld.