Gregory S. Paul suggests that we should use statistical analysis to solve some of the big questions of religion. Of course, others have suggested the same. In 1872, Sir Francis Galton analyzed the longevity of British rulers, whose health was prayed for by millions, and concluded that prayer is ineffective. Your can read Paul’s essays at The Science of Religion:
The application of statistical scientific methodology to the sociology of religion has become less controversial in recent years, but it remains a sufficiently delicate subject that major questions of great import to the public at large remain seriously under researched. For centuries religious forces have claimed that a good deity exists, and that it is imperative for the citizens of nations to follow the instructions provided by their moral creator if they are to avoid societal collapse into godless nihilism. Despite the often strenuous advocacy of this claim, a comprehensive analysis confirming its truth has never appeared.