I was talking with some friends about searching and we all use google
since it works rather well... We then began to wonder what it would
take to create a search engine (or if there was one out there) that
rated pages based on how low they are ranked in google. A page that
cannot be found with googlw would have the best rating... pages that
tend to come up on the bottom of a search would have 2nd best ratings,
pages that come up on google for many search terms near the top would
have the worst ratings.

I'm just a playwright and I only know a smattering about computers so
if the sources you find are easy for novice to read that's be a big
help.

futurebird,

Unfortunately, the nature of the Internet makes it impossible to
create an "anti-google" of the type you have envisioned.  This is
because the pages that would be lowest ranked in such a system would
be the ones which are most recently created.  From the time when a new
page is first published on the internet to the time when Google's
automated web crawlers can "find" the new pages (and an undetermined
time afterward to place the new site in the Google databases), the new
sites cannot reside in Google's database, and are thus the lowest
ranking.  However, there is no Internet registry of new sites as they
are launched, so there is no way to find all of these new sites to
include in the "anti-Google" rankings.  So even if Google tried to
create a proper "Anti-Google", they would be faced by this paradox:
once they gathered the lowest ranked sites, they would no longer be
the lowest ranked sites!

If you are interested in finding "low ranking" sites for popular
terms, one strategy you might use is to deliberately enter in a
misspelling of a popular term in Google.  The "best" sites most likely
do not have such typographical errors, so you can use this technique
to separate some of the chaff from the wheat.  Try searching on
"Yhaoo", "basktball", or "shaekspeare" to see what else is out there. 
There is even a site called Searchspell that will help you generate
spelling errors for words in case your brain needs a rest.  Hope this
helps!

Sites Referenced:

Searchspell Typo
http://www.searchspell.com/typo

Google Search Terms Used:
misspelling typo
yhaoo
shaekspeare
basktball

---

Request for Answer Clarification by futurebird-ga on 15 Apr 2003 05:47 PDT

I don't think this answers my question. From what I understand google
rates pages based on how many other pages link to that page (among
other things) There must be pages that are linked to by very few other
pages . . . think of them as "hidden" pages. I don't think that
spelling has anything to do with the quality of a page ... it all
depends on what one is looking for.

---

Clarification of Answer by hailstorm-ga on 15 Apr 2003 22:02 PDT

futurebird,

Yes, but at the bottom of that pack is "orphan" pages, pages that are
not linked to by any other page.  It is these pages that would have
the lowest ranking, and also those which cannot be detected by
Google's automated data robots.  The only automated technique that
could detect unregistered "orphan" sites is checking every single URL
that can be constructed.  A Microsoft IE browser can recognize a
maximum URL length of 2,083 characters, and there are 55 standard
characters that can be used in a URL (though IE may recognize more),
so a lower bound on the number of potential URL combinations is 55 to
the 2,083rd power, a truly astronomical number that no amount of
computer processing power and networking bandwidth would ever be able
to handle.  Therefore, a "perfect anti-google" is impossible.

If you were to allow for sites with at least one link, you would still
not be able to create a good Anti Google for most words, because
Google only allows you to receive the top 1,000 ranked entries for a
search.  Because you cannot access anything below the top 1,000 ranked
sites, and because many Google searches yield more than 1,000 results,
you cannot use the Google site as is to run a "Anti Google" search,
and it would not be possible for a third party to create a true Anti
Google using the Google API either.

Sites Cited:
    Maximum URL Length Is 2,083 Characters in Internet Explorer
    http://support.microsoft.com/default.aspx?scid=KB;EN-US;Q208427

    URLs
    http://www.cs.tut.fi/~jkorpela/HTML3.2/3.5.html

Google Search Strategy:
    maximum URL length
    what characters can be used in a URL?

Since it would be difficult to engineer such a site, and few people
would find it useful enough to visit more than once (for novelty
value), I can't see why anyone would create a site of this sort.

But, of course, the Internet is full of bizarro oddities, so I guess
it's possible that someone might try this. As far as I know, it has
not been done yet.

Some search engines do act like "Anti-Google", since they overrate
pages that try to spam (e.g. repeated keywords that will be on the
page, but hidden from th display). Those are certainly not the pages
one would want to visit since they mostly contain no relevan
information.


As for creating a new, "pure" Anti-Google; one could access the Google
API, which is open for any developer, fetch a search result, and
display it in reversed rankings. For more information, see:

Google Web APIs
://www.google.com/apis/


And this one is a different type of Anti-Google -- everything's
mirrored!

elgooG
http://elgoog.rb-hosting.de/

(Sorry for the missing letters in above comment!)