Optimization heuristics have been pursued in recent years as a viable approach in cryptanalysis. Even in simple ciphers where brute force method is successful, use of these techniques demonstrates their potential application in at- tacks of complex ciphers. This paper establishes the applicability of a couple of optimization heuristics to cryptanalysis studies; one based on thermostatistical persistency applied to simulated annealing and the other one based on particle swarm principle. Though both methods lead to successful attacks, our improvised version of group swarm optimization yields better performance. As a vehicle of demonstration of our concept, we choose simple yet representative block ciphers such as computationally tractable versions of DES, for our studies.