Except you have a judging team do it, based on published criteria, rather than all the participants ranking eachother. A bit like a school exam (where students are usually ranked - curve fitted - to eliminate variable difficulty of the questions that year)