Double risk requirement is ‘a methodological error’