If a field systematically makes wrong predictions (not “some experts some of the time”), you should not trust subsequent predictions from that field.