Heuristics in theorem provers are often parameterised. Modern theorem provers such as Vampire (Kovács & Voronkov, 2013) utilise a wide array of heuristics to control the search space explosion, thereby requiring optimisation of a large set of parameters. An exhaustive search in this multi-dimensional parameter space is intractable in most cases, yet the performance of the provers is highly dependent on the parameter assignment. In this work, we introduce a principled probablistic framework for heuristics optimisation in theorem provers. We present results using a heuristic for premise selection integrated with Vampire and The Archive of Formal Proofs (AFP) (Jaskelioff & Merz, 2005) as a case study.