Too-Perfect Theory is not an empirical theory, but a philosophical notion that a trick possibly can be "too" perfect and thus lead the audience directly to the method or too a wrong solution which gives the magician no credit. This notion was first published by Rick Johnsson in Hierophant in 1970, expanding on an idea attributed to Dai Vernon.
Rick Johnsson suggested that magicians should consciously construct their routines to lead the audience away from the actual method by allowing room for "red herrings." Also, since spectators will try to settle upon some solution (right or wrong), whenever possible, the magicians should lead them down a path where they receive the credit for the effect.
The "Too-Perfect Theory" article was republished in Genii 2001 August along with numerous articles debating the topic.
In the Books of Wonder, Tommy Wonder defined two parts in Rick Johnsson's idea. One diagnostic part and one part that suggest a solution. Tommy argued that the diagnostic part was valid, but that the suggested solution was flawed.
In the October 2001 issue of Genii, Tom Stone pointed out that the "theory" did not work as an diagnostic tool either, and gave several examples.