What Is The Internet Hiding?, page 3

reply to post by jackieisinlove

But it seems odd that there was a discrepancy in the amount of search results both people from the UK got on the first page of this thread. They also both did the search in the same time period, so this still doesn't seem like a perfect explanation to me.

Well, it's all speculation as Google's formulas for deriving search results are held in confidentiality to degrees that are on par (if not superior) to U.S. Top Secret materials handling procedures. In the past - these things were simple and rather open. Then, entire businesses came about deriving ways to exploit these formulas as a form of advertising for sites (you pay these companies to, effectively, get you listed on the front page of various search engine searches... which the search engine developers tend to not take kindly to).

In that sense - it all comes down to faith as the topic is, to a wide degree, unknowable. Just as we cannot know if our 'me-ness' transcends death or not - we cannot know exactly how Google's searches work. People who believe there's a conspiracy and that they are being denied information will believe that. Generally, if you believe you are being oppressed by one region or authority - you will be more likely to believe it in others. That means this site is going to have a fairly large portion of the users who will believe there is a conspiracy and there is something being hidden from them. And the people who don't are simply not going to see the deal.

As for that specific observation, again - do these people have Google Toolbar downloaded and enabled? What are their individual search criteria set to? If I have safe search enabled, then I am going to not be given returns linked to 'adult' material. If you're a pervert and have safe search disabled for the obvious reasons - searching for "egypt" is going to bring up a different variety of site that would be excluded from the other results completely (not simply shifted to the bottom).

Further - some of Google's functions use your search history as well as the cache and cookies in your browser. I run two different browsers and have noticed a difference in search results between the two (I normally use one for business related stuff, and the other for general purpose or compatibility - so there's a distinct difference between my trends with each browser).

If you really want to try and get useful data from this type of test... you have to control the variables. As it stands "search google and post your results" is about as controlled as "go drive your car until the tank runs dry and report how many miles it took." How full was their tank to begin with? What kind of car/engine? What kind of driving habits does each driver have? What path are they driving? Are they driving all at once, or just driving it back and forth to the store and work (starting it up and shutting it down repeatedly)?

Without any kind of controls - you can't make any kind of determination.

Also - there's considerable margin of error involved. Google will say: "About *# results found." ... What's "About" mean? In all likelihood, to get fast searches, your search doesn't search -all- listings. The listings are probably not all contained on one server. Your initial search for "egypt" goes as far as google's closest server to provide you a link. As that server begins to run out of results to display to you, it begins reaching out to more servers - that begin reaching out to more - until results are returned or found to not exist. Thus - the number you see is probably not how many results there are in total - but how many results your server estimates there are based on what information it has stored, locally (IE - it is simply estimating what all is out there - not really getting a hard number).

Again - that gets into the mechanics of Google that are guarded pretty closely to prevent people from taking advantage of their relevance ranking systems.