It’s Simple: The Google Wants to be Human
Consumed => “Matt Cutts Describes How Content Is Ranked Without Many Inbound Links” by Matt Southern on Search Engine Journal.
=> My value add (i.e., left a comment)…
Obviously, Matt Cutts is the authority. That said, the rule of thumb (that I lean on) is pretty simple: The Google – yes, I meant to say The Google :) – wants to be human, but a highly efficient and automated human. So if you ask yourself, “What would I (i.e., a human) do?” some of the answers become more obvious.
For example, once The Google identifies the subject matter, I would presume it compares newly indexed content to established / ranked content on the same and/or similar subject matter.
I would also presume it starts with the site that the new content is on. For example, the new content is about The World Cup. The Google also finds another page on your site also about The World Cup. It seems pretty reasonable that old page helps The Google analyze and weight / rank the new page’s content.
Since its signals are somewhat limited, The Google has to make the best use of what it does know – just as you or I would under a similar circumstance.
Makes sense, yes?