The problem here, as with virtually all broadly published predictions, is that there is no indication of the probability that it is correct. It would be much more useful if he said something like “30% probability of 3 big hurricanes, 20% probability of 2 hurricanes, 20% probability of 4 hurricanes”, etc.
I say there will be nine named storms, three majors, and one landfall in the U.S. with a 5% chance of error. If I'm wrong, then it is the 5%.