Fission barriers heights and excitation energies of superdeformed isomeric minima are calculated within the microscopic - macroscopic Woods - Saxon model for 75 actinide nuclei for which the experimental data are known. State - of - the - art methods were used: minimization over many deformation parameters for minima and the imaginary water flow on many - deformation energy grid for saddles, including nonaxial and reflection-asymmetric shapes. We obtain 0.82 - 0.94 MeV rms deviation between the calculated and experimental barriers and 0.53 MeV rms error in the excitation of superdeformed minima (SD). Experimental vs theory discrepancies seem to be of various nature and not easy to eliminate, especially if one cares for more than one or two observables. As an example, we show that by strengthening pairing in odd systems one can partially improve agreement in barriers, while spoiling it for masses. We also discuss the thorium anomaly and suggest its possible relation to a different way in which the Ac and Th barriers are derived from experimental data.