The No Miracles Argument (NMA) is one of the most influential arguments for scientific realism. It argues that the predictive success of science would be a miracle if mature scientific theories didn’t tend to be approximately true. Colin Howson has offered a Bayesian reconstruction of the NMA. He comes to the conclusion that the NMA commits the base rate fallacy and therefore is logically flawed. In my talk, I will distinguish two forms of the NMA, which I call the individual theory based NMA and the frequency based NMA. I will show that Howson only reconstructs the–less substantial–individual theory based version. I will further demonstrate that a formal reconstruction of the frequency based NMA can in fact alleviate other worries that have been raised about that version of the NMA.

The paper is based on joint work with Stephan Hartmann.