SBP

hello everyone,
i have generated SBP using my protein active site containing 7 features -1HBA, 3HBD, 3 hydrophobes. i then screened a set of ligands for hits. the fit values varied from 3.907 to 0.256 giving me 23 hits, in this range. however on docking the results with cdocker i have seen that compounds giving a lower fit value (2.908) gives a better cdocker energy (-41.581) as compared to the one with the highest fit value (3.907)giving me a cdocker energy of (-30.907).
could anyone please tell me why does this happen? ideally the molecule giving higher fit values should also be giving better cdocker energies.