Do not be fooled by the summary, this book has no true topic. It reads as if Trivers posited an initial question and then decided to write a series of tangentially related chapters to fill the book.
The introduction identifies an interesting question and then the remainder of the book fails to address that question. It almost felt like Trivers wanted to write about deception in nature and fell on the "self-deception" angle as a way to bring in readers. He posits that self-deception may be evolutionarily useful because it reduces the likelihood that our attempts to deceive others will be identified. He brings up a few studies that demonstrate the connections between cognitive load and the likelihood that deception will be uncovered; however, these studies are handled in a cursory way and stem entirely from the field of psychology. While I generally find psychology to be a useful social science, Trivers spends 15-20 pages lambasting psychology as unscientific and prone to self-deception. He also takes issue with the the laboratory based experiments that psychology has used because they fail to demonstrate the utility of self-deception/deception. It is curious that the only evidence he uses to support his theory of self-deception comes solely from a field he seems to hold so much disdain for.
The first few chapters focus solely on deception in nature. Not self-deception, and often not even intentional deception. Mimicry, camouflage, and non-human communication take center stage. When he discusses humans it is to note the deception that human genes engage in both in utero and in development in order to increase the the chance of procreation. This chapters are almost completely irrelevant to the topic at hand. While a discussion of the utility of deception in the evolutionary context would have been a useful foundation, the discussion in this text focuses solely on involuntary deceptions (i.e. camoflage) which, while advantageous, have no logical bearing on self-deception. After spending the first half of the book building up this irrelevant base about the benefits of (largely involuntary or unintentional) deception, Trivers gives a very brief overview of the ways that humans may deceive themselves. These generally fall into the cognitive biases that have been identified in psychological studies, again curious considering the issues Trivers has with the field. Immediately after this, he launches into some examples of self-deception by criticizing American foreign policy as well as historical revisionism in a few nations. He discusses self-deception as applied to religion and gives a brief explanation of where religions may have come from. He ends the book with a short chapter dealing with how to avoid self-deception, which mainly boils down to "be conscious of you actions," and "talk things over with others."
Trivers states at the beginning of the book that he wants to deal with self-deception through the lens of evolutionary biology. He fails spectacularly in this endeavor. Nowhere is there any attempt to actually explain the benefits of self-deception beyond a few meager, unstudied, connections. Trivers provides a polemic on the deficiencies he sees with social sciences, going on to call psychoanalysis a long-running fraud and specifically decrying the popularity of Freud. Despite this vehemence, Trivers relies heavily on the laboratory studies in psychology departments, the same ones he alleges are artificial and based on morality instead of evidence, to prop up his lackluster and very weak theory. Even worse, there is not a single new idea posited in this book. The vast majority of the topics he discusses with regard to deception and self-deception have been covered by the fields of cognitive and evolutionary psychology. His discussions of religion and warfare have likewise been studied by evolutionary psychologists, and his take on the political implications of self-deception are left wanting considering the in-depth coverage they have received by political scientists.