I think you are missing the most salient point of the SPE. While Zimbardo may not have taken into account experimenter bias and the lack of random sampling, it is nevertheless true that by replicating prison conditions (not being able to leave and encouraging sadistic behavior from the guards) the participants did indeed conform to the sort of behaviors seen in our prison systems. Those behaviors INCLUDE faking symptoms and acting the part - things you consider to invalidate this research. Rather than work to undermine faulty logic on the part of the experimenter, I think it is more interesting to look at the results from the perspective of what it DOES tell us. Namely, that institutions based upon punishment will elicit certain predictable behaviors from prisoners and guards because it is a social situation that elicits those behaviors. The point it NOT to excuse individual choices but rather to examine and hopefully change our institutions so they create more positive outcomes.