I just got done reading a great post by Hamlet Batista over at SEOmoz.org and damn my brain hurts.

It talks about some early adjustments to the Google Page Rank algorithm to help combat Rank Sinks and Page Hording. Basically the random surfer of Google Page Rank.

Here what I came to find out.

G = α S + (1 - α) E

G = Google matrix

α = scalar between 0 and 1 = 0.85

S = matrix stochastic = 1/n eT

eT = a row vector of all 1s

E = teleportation matrix = 1/n eeT

G = .85 * 1/n eT + (1 - .85) 1/n eeT

or

Google Matrix equals the product of a scalar between 0 and 1 and the Matrix stochastic Plus the product of 1 minus a a scalar between 0-1 and the teleportation matrix

Which means that 85% of the time the surfer is following links at random, and 15% of the time he is entering new URLs in the browser bar.

Interesting