Highly intelligent people often have problems in predicting the likely behavior of thugs.
In his 1982 book Rethinking Systems Analysis and Design–a work whose relevance is considerably broader than might be imagined from its title–Gerald Weinberg briefly discusses a contemporary book called How Real is Real?–An Anecdotal Introduction to Communications Theory. Although he finds value is many aspects of this book, Weinberg strongly objects to a passage in which the author (Paul Watzlawick) suggests ways in which communications theory could have been used in the Patty Hearst kidnapping case. Watzlawick suggests that the authorities should have used “Erickson’s confusion technique” as follows:
Utilizing the same channels of delivery as the abductors, it would have been relatively simple for them to deliver to the mass media fake messages, contradicting the real ones but similarly threatening the life of Patricia Hearst if they were not complied with…Very quickly a situation of total confusion could have been set up. None of the threats and demands could have been believed, because every message would have been contradicted or confused by another, allegedly coming from the ‘real’ abductor.
Weinberg responds:
It’s very difficult for me to believe that Watzlawick ever thought critically about this idea for fifteen seconds, but its naivete is typical for this genre of speculative systems writing.
…and goes on to suggest that a good way to consider the possible real-world consequences of ideas like this is to imagine a movie (specifically, a thriller) based on the situation and the proposed actions, and to imagine how the plot might develop.