Reidentification of decision-maker preferences is a crucial aspect of Multi-Criteria Decision Analysis (MCDA), as it enables a structured evaluation of decision-making methodologies. This study presents a comparative assessment of three stochastic reidentification techniques: Stochastic Identification of Weights (SITW), Stochastic Fuzzy Normalization (STFN), and Stochastic Identification of Models (SITCOM). Each method models decision-maker preferences based on different paradigms: weight-based aggregation, normalization, and reference-object-based evaluation. To systematically analyze their effectiveness, we employ three benchmark preference functions: a monotonic function representing linear preference structures, a non-monotonic function with a single extremum reflecting a decision-maker with a specific optimal point, and a non-monotonic function with multiple extrema modeling complex preference structures with multiple local optima. Our findings indicate that SITW is most effective for monotonic preferences, STFN provides superior performance in single-extremum cases, and SITCOM excels in handling multiple-extrema scenarios. The comparative analysis highlights the limitations of weight-based approaches in complex decision problems, demonstrating that reference-object-based models are better suited for non-trivial preference structures. The study contributes to the understanding of how different MCDA reidentification techniques perform under varying decision-making conditions, offering practical insights into the selection of appropriate methods. Future research should focus on integrating hybrid methodologies to enhance reidentification accuracy and applying these techniques in real-world decision-making contexts.