The research is to focus on various classical questions in complexity theory, as they are posed anew in various settings of two-person probabilistic games, specifically interactive and zero knowledge interactive proofs, and games against nature. The kinds of questions considered include (a) comparing the strengths of different models of probabilism, e.g. one-sided error, two-sided error, unbounded error; (b) limiting the number of random bits used, or questions analogous to more classical questions involving pseudorandom number generation; (c) understanding the power of alternation or interaction; and, finally, (d) restricting the flow of information.
