I'd call his predictions "statistical analysis", not science (there's a thin line between the two in this instance) since only probabilities can be deduced from his work.