Deus Ex Google

It brightened my evening to write this email reply:


On Fri, 2006-08-25 at 00:18 -0400, RXXXXXX4@aol.com wrote:
> A colleague from Fresno State sent me this unbelievable fluke:
>
> 1. - Go to http://www.google.com
>
> 2- Type in the word "Failure."
>
> 3- Look at it the first listing and laugh at what comes up first
>
> JR

Just to add insult to injury, I'd like to elaborate a bit to investigate
this "fluke". A fluke is something has an element of accident or chance
happening. Yet I don't think this is a fluke at all. I think Google's
trying to tell us something... something important and illuminating.

Let's investigate, shall we? Enter "miserable failure", doubling the
words and multiplying the number of pages that will form our data set.
Google is lightning quick to return the same answer as before. This
isn't another page about the same person, this is the exact same link to
the exact same page. This provides us with new and important
information. Google is telling us that he's not only a failure, he's in
fact a special type of failure, a miserable one.

But wait! Isn't it logical that exact link would be the first result for
both queries, given that "miserable failure" is just a subset of
"failure"? Doesn't the method of aggregating data sets explain how we
get the same results?

Not having the source code to Google PageRank personally, I can't prove
to you that it doesn't work that way. But I can provide the same proof
by changing that variable from miserable to something else. Does he come
up first under "sexy failure"? What about "forgivable failure"? No sir,
he's not a part of the result in any form with these queries.

So there's some evidence to support the argument that Google is being
clear, specific, and consistent in what it's saying.

Now I'm going back to play with my Ouija board some more. :)

-- Craig

Posted on August 25, 2006 in other things that matter . | 669 Trackbacks, 0 Comments

Comments

Post a comment




Remember Me?

(you may use HTML tags for style)