IBE's selection criteria for 'best' explanation already presuppose that unobserved regularities beha...