OICC Press International Journal of Mathematical Modelling & Computations 2228-6233 2026 07 03 Robust Preference Aggregation under Strategic Behavior: A Consensus-Aware Game Cross-Efficiency Framework 10.57647/ijm2c.2027.1702.10 EN Arezoo Khosravirad Department of Mathematics, Isf.C., Islamic Azad University, Isfahan, Iran Abdollah Hadi-Vencheh Department of Mathematics, Isf.C., Islamic Azad University, Isfahan, Iran Ali Jamshidi Department of Mathematics, Isf.C., Islamic Azad University, Isfahan, Iran Majid Tavassoli Kajani Department of Mathematics, Isf.C., Islamic Azad University, Isfahan, Iran Journal Article 2026 07 03 The aggregation of heterogeneous preferences in multi-stakeholder environments is often compromised by strategic behavior, scale inconsistencies, and the absence of ground truth for validation. While Data Envelopment Analysis (DEA), specifically the Game Cross-Efficiency model, provides a Nash equilibrium-based mechanism to mitigate weight arbitrariness, existing frameworks suffer from critical ordinal instability and a lack of internal verification mechanisms. This study presents a novel, integrated decision support framework that fortifies the game-theoretic foundations of DEA-based voting through endogenous normalization and exogenous stability assessment. Unlike traditional sequential approaches, the proposed methodology synthesizes three rigorous components: (i) an unmodified DEA game cross-efficiency mechanism to capture peer-evaluated performance; (ii) a Relative Ratio (RR) transformation axiomatically proven to enforce scale invariance and preserve strict dominance without altering the underlying preference structure; and (iii) a composite stability verification protocol utilizing WASPAS and COPRAS algorithms to quantify rank reversals and dominance consistency. Theoretical properties, including affine invariance and monotonicity preservation, are formally established. Furthermore, a novel Composite Stability Index (CSI) is introduced to serve as a proxy for ranking reliability. Application to benchmark voting problems demonstrates that the proposed framework significantly reduces ordinal ambiguity and provides a defensible mathematical basis for detecting unstable rankings. This research contributes to Operations Research by transforming DEA voting from a black-box evaluator into a transparent, verifiable system suitable for high-stakes policy and collective decision-making.