Google consolidating link juice on duplicate content pages

[James77]

I've observed some strange findings on a website I am diagnosing and it has led me to a possible theory that seems to fly in the face of a lot of thinking:

My theory is:
When google see's several duplicate content pages on a website, and decides to just show one version of the page, it at the same time agrigates the link juice pointing to all the duplicate pages, and ranks the 1 duplicate content page it decides to show as if all the link juice pointing to the duplicate versions were pointing to the 1 version.

EG
Link X -> Duplicate Page A
Link Y -> Duplicate Page B

Google decides Duplicate Page A is the one that is most important and applies the following formula to decide its rank.

Link X + Link Y (Minus some dampening factor) -> Page A

I came up with the idea after I seem to have reverse engineered this - IE the website I was trying to sort out for a client had this duplicate content, issue, so we decided to put unique content on Page A and Page B (not just one page like this but many). Bizarrely after about a week, all the Page A's dropped in rankings - indicating a possibility that the old link consolidation, may have been re-correctly associated with the two pages, so now Page A would only be getting Link Value X.

Has anyone got any test/analysis to support or refute this??

[Dan-Petrovic]

I have observed some strange switches between home page and landing page on our site which kind of goes towards your theory. However I am not convinced that Google consolidates PageRank in any other way then simply by following links and mathematically assigning it throughout the site. All duplicate pages flow PageRank and pass it to the rest of the site in the same way they receive it. Sometimes Google will block a page which is seen as duplicate. That page will not show or pass any PageRank or anchor text value to other pages (dead end).