Is it "science?"
Jerry Pournelle and others have been discussing the Mann "Hockey Stick" curve. You can check it out, along with links to the original article, but what I found interesting was this statement from Pournelle:
If you can't mail it in a letter to a colleague so that he gets the same result you do, it isn't science.
That's quite a generalization, and given the rule of "reproducable results," not new, either. What do you think? Given some leeway on the idea of being able to fit all forms of data into something as compact as a "letter," is this basically true?