First off, the frequency analysis is not going to work out exactly like it does in traditional english. For one thing, there are going to be a great deal more Zs. Included is a comparison of a Lolcat analysis (generated from icanhascheeseburger entries) and an English analysis (taken from here).
Second, many of the common words are going to be unique - if it's 3 letters, starts with 't' and ends with 'h,' chances are you've got 'teh.'
And so on.