Forum Moderators: Robert Charlton & goodroi
Here are my starting contributions .
------------------------------------
Q. Yahoo and MSN show a much higher number of links than Google. Why is that?
A. With the link:example.com operator, Google intentionally shows only a small sampling of all the links that they know about and use. The working theory is that the Google algorithm is much more sensitive to links than other search engines and they don't want to expose too much data -- or else it would be too easy to reverse engineer their secret sauce.
------------------------------------
Q. My site disappeared overnight. Why did this happen and what can I do?
A. There are many technical issues that can make this happen. Two excellent reference threads for further study:
Dropped from Google - a checklist to find out why [webmasterworld.com]
Dropped Site Checklist [webmasterworld.com]
------------------------------------
Q. My rankings were excellent for a long time and then suddenly dropped. Why?
A. Again, there are many possible factors. Here's another link to a solid thread from the library:
Checklist for Sudden Drops in Rank [webmasterworld.com]
A. PageRank is a name for a patented system used by Google that assigns a number from 0 to 10 to a web url.
PR is page-specific -- there is no such thing as "PageRank" for a site or domain. PageRank is also independent of any measure of content. It is defined exclusively by the existence of links, which you can think of as "votes" that one page makes for another page.
-- THE MATH OF PR --
More precisely, when one page (the source) links to another page (the target), the linking page's PR is divided by the total number of links on the source page. That value is multiplied by a constant -- it's the the same value in every case -- and the resulting number is the "weight" of that particular vote, that link.
If you think about that definition, you can see it is kind of circular. That is, some kind of PR must already be known for the linking page, in order to calculate the weight of its vote for the target page. And in many cases there will be a link trail that eventually comes back to the original page.
So circular is a pretty good word, here. PR for all the indexed urls on the web does need to be calculated over and over again, around and around the entire index, until the values for the pages begin to settle down to a limit and further rounds are not generating major changes. Mathematicians call this an "iterated" calculation. That little constant mentioned earlier is what keeps PR from flying off to infinity and keeps it approaching a limit instead.
-- ONLY ONE PART OF RANKING --
Since PR is not dependent on content, it can only be one factor, one part, of the total algorithm that Google uses to generate search results. And indeed, Google tells us that there are over 100 factors used for the final calculation.
Google also reserves the right to decide that a certain page or set of pages should no longer "pass" PR to their link targets. This can happen even though the source page and the target page both show a PR value on the toolbar.
-- TOOLBAR PR --
The final real value for a url's PR is calculated out to many decimal points. The toolbar only shows a rounded off value, a whole number, for any url -- and that value is almost always taken from several months ago. But real PR is being recalculated continually, and the new values are then folded into the final ranking seen in the search results.
