We have tried to determine whether experimental data on L=0 (t,p) and (p,t) reactions on the even calcium and nickel isotopes are consistent with a reasonable amount of configuration mixing and a distorted-wave Born-approximation treatment of the two-neutron transfer process. Our calcium calculation involves the six shells from 2s12 through 1f52; our nickel calculation involves the five shells from 1f72 through 1g92. We have simulated the configuration mixing by diagonalizing a pairing force between seniority-zero states. The calculated eigenstates are consistent with one-particle transfer data, and give a satisfactory account of the ratios of observed (t,p) and (p,t) cross sections for ground-state transitions. However, certain excited calcium states seen in the (t,p) reaction are predicted to be populated with about twice the observed cross section. These states must involve more complicated degrees of freedom than the seniority-zero components we have included. Calculated (t,p) angular distributions are in good agreement with the experimental data. Calculated (p,t) angular distributions are about 5°out of phase with the experimental data.