We use non-maximally entangled states (NMESs) to simulate an entangling unitary operator (EUO) with a certain probability. Given entanglement resources, the probability of the success we achieve is a decreasing function of the parameters of the EUO. Given an EUO, for certain entanglement resources the result is optimal, i.e., the probability obtains a maximal value, and for optimal result higher parameters of the EUO match more amount of entanglement resources. The probability of the success we ac...