It’s now believed that he altered the data, since they fit the 1:3 ratios way too well for populations where each has a ¼ probability. Still, very good work considering he might not have heard of the scientific method.