hmm with my little experience in compressing algorithm from university, i have to say it sounds right, but Erwin is right here, more noise would be harder to compress than just informative things
or as an example:
you have a spamm post:
Code:
abdjdrhjgf,fhrfkfhednfznrfufnfbkgigfkgkdugjgudr
would be greater when compressed as this helpful post:
Code:
this is so, and you have to do so and so and so
politeness of good discussion forums is good to compress, never forget