The Fidler Alternative is the name of the alternative value of p-hat, the sample proportion, which would lead to the same result for a two-tailed hypothesis test.
He told his students that the Fidler Alternative was always going to be an equal distance from p-null as p-hat was, just in the opposite direction.