googlewhack point system proposal (was Re: [thechat] interesting blog)

[spinhead]

> here's a three word googlewhacks i came up with:
>
> with the bonus of each word beginning with the same letter
> http://www.google.com/search?hl=en&lr=lang_en&q=myopic+mongrel+mongoloid
>
> score:
> myopic - 55,800
> mongrel - 44,200
> mongoloid - 14,200
>
> 55,800 x 44,200 x 14,200 = 35,022,312,000,000
>
> .jeff
>

This point calculation system doesn't work because it values MORE words in
the googlewhack higher than LESS words, which is clearly backwards, it being
more difficult to gwhackfewer words. While multiplying the original word
count this way makes sense (more hits on each word INCREASES the value of
the gwhack, since a gwhackof common words is cooler than one of rare words)
there's an inverseness we need to compensate for in the number of words in
the gwhack. I think three-worders should be valued at one-third the cube
root of the calculated value, and two-worders at one-half the square root,
thus introducing an inverse proportion to re-order the valuation correctly.

The math would be ( [wordcount(1) * wordcount(2) * . . . . wordcount(n)]
^-n) / n

So, your myopic mongoloid mongrel becomes worth 10905 while my cerulean
melisma is worth 14737, which is clearly more meaningful. If anyone would
like to consider a motion to divide further by 1024 (resulting in values, in
the above example, of 10.7 and 14.4 [a significant number, in my opinion])
that would be worth discussing, I think.

By the way, if anyone would like their googlewhacks calculated by the new
Spinhead Reverse Osmosis Googlewhack Calculator, I'd be glad to oblige.

joel