Oh, you're absolutely right that a lot of this stuff was purely goofy, but, not having read the Lewis book, I still feel compelled to recommend the Gillian Tett book to give a better picture of how people persuaded themselves otherwise. To me (still a very layperson's point of view, of course), the problem with the CDOs is not really that the math was so very blatantly, ridiculously bad as that a horrible incentive structure combined with certain prevailing economic conditions to encourage people to try to solve problems with math that wasn't reliable enough to sustain the sheer sums risked on it in a system that, contrary to expectation, amplified rather than diffusing risk.

...Is that sentence convoluted enough???