I type in, for example, “Maris Roger asterisk home run record” press return, and in less that a quarter of a second (0.18 seconds, to be exact), I get pages and pages of clickable links for websites that have info on Roger Maris’ record-setting home run season in 1961.
How the ^#@#!*& does Google manage to sort my query so fast?
The searchable database of keywords maintained by Google must be immense, and it boggles my mind that it can be searched so quickly. And not only that, but the search results are almost instantly constructed into an HTML-formatted webpage for me to view.
Google has a spidering program which periodically goes through gazillions of pages on the web and indexes them. The exact nature of the index is a closely guarded secret, but typical indexing techniques involve removing pronouns, conjunctions and other common words, and building a database of what terms appear at what URLs with the most frequency. This is how most basic search engines work, though Google goes a bit further. When you submit a query, Google looks in their index for all the pages that contain the words you entered. It then looks in another index to see how many pages in that group link to other pages in that group. Google assumes that the most linked-to pages are the best, and puts them first in the list.
As for speed, it comes down to two things:
Very efficient algorithms.
A really really really really big number of very very fast servers.
