Fun idea, but I can't see any real journal ever accepting a paper where the methods would be so roughshod and unreliable.