There is a much higher probability that a first-class scholar should commit

There is a much higher probability that a first-class scholar should commit an error than that an author who usually writes nonsense should have one good idea.