The binding energies for the reaction M + +CH 2 CHCH 3 →M + [CH 2 CHCH 3 ], for M=Ag, Au, are predicted using both second order perturbation theory (MP2) and coupled cluster theory (CCSD(T)). Systematically improving the quality of the atomic basis set to the complete basis set limit produces a monotonic improvement in the predicted binding energy for M=Ag relative to the experimentally determined value; the final predicted binding energy is within experimental error. Conversely, the same systematic improvement in the atomic basis set results in successively worse agreement at the MP2 level of theory.