Strange related: results

[spander]

When I do a Google search for related:mydomain.com I see a totally unrelated site in first place.

But in 2nd place comes www.google.com
5th place is www.altavista.com
6th place is www.yahoo.com
9th place is www.msn.com
also in the top 20 are alltheweb.com, lycos.com, and freefind.com

The rest of the places are sites related to mine.

What's that all about?

[tedster]

Google says:

This search is like searching a bookstore to find books similar to the first Harry Potter novel. The results could include other children's books, a biography of J.K. Rowling, or a non-fiction book on children's literature. In general, use this operator to find resources that overlap. You'll get the best and most useful results if you use sites that cover a broad range of content.

Google uses several factors to determine the similarity of different sites. However, the quality of the sites returned has no impact on your ranking or on how Google indexes your site.

Google Webmaster Help [google.com]

The related: operator never meant much to me. When I use it (which is rarely) the results seem to be more about click distance and Google's link graph of the web that anything much about topics or themes.

[spander]

Yah, I never use it much either. I guess that's why I never noticed it before. Just seemed odd to me that G thot all those sites were related to mine. Obviously not very accurate.

[youfoundjake]

What really confuses me about the related operator, is that matt cutts blocked a page on his website with robots.txt, then set the meta to be noindex, and when viewing the url through google search, it shows the url (which is fine) but when you try and do a related search based on that page, google can offer up 30 different pages. How can they offer related content when the page in question is blocked by robots.txt and set to noindex? hmmm?
that may boggle your mind for awhile...

[g1smd]

Does it show URLs that are related to that page? Or related to that domain? How's it really work?

[youfoundjake]

I think that it's picking up on the domain name, mattcutts.com and then finding pages out there with matt cutts in it. Towards the bottom of the list, are some pages that contain the phrase "bear in mind" which appear nowhere on the orginating page or on the noindex page.

I'm providing the link only because it is relevant to this thread, and may lead to another thread as to how the search engines handle a noindex tag which is the purpose of his post..
[mattcutts.com...]

Moderators, feel free to examplify if deemed contrary to TOS