The selection of parameters is a critical factor in the efficacy of Evolutionary Algorithms (EA) in resolving complex constrained industrial problems, such as multi-pass turning optimization. In order to identify optimal and resilient parameter configurations prior to the application of EA, this paper introduces a novel hybrid approach that integrates CCR Data Envelopment Analysis (DEA) with both Best Practice Frontier (BPF) and Worst Practice Frontier (WPF). The methodology’s efficacy is evaluated on the basis of industrial models such as the single-pass, multi-pass turning, and tension/compression spring design. The results emphasize that the compromise desirability approach, which is derived from the geometric mean of response and time efficiency, substantially enhances parameter selection. Parameter D (random seed) is consistently significant, with optimal values of 0.0 in both BPF and WPF scenarios. Parameter C (population size) fluctuates between 15 (BPF) and 25 (WPF) as a result of its influence on time efficiency. The integration of DEA and RSM enhances the quality of EA solutions, with Parameter D consistently achieving optimal performance levels across all models, and Parameter C influencing time efficiency significantly. This resulted in an average production cost reduction of 12% across the tested models. Additionally, a modified Super- Efficiency BCC model was employed to further refine the efficiency evaluation of decision-making units. This hybrid approach is a valuable instrument for addressing complex industrial optimization challenges, as it ensures optimal and robust parameter selection through the combined use of BPF and WPF.