Beyond Winning: Margin of Victory Relative to Expectation Unlocks Accurate Skill Ratings

2025-05-31

AI Summary

Traditional skill rating methods (e.g., ELO) rely solely on win/loss outcomes, disregarding margin-of-victory (MOV)—a critical performance signal—while existing MOV-augmented models fail to adequately model the expected MOV. This work proposes MOVDA, a novel framework that for the first time incorporates both the observed MOV and its deviation from a domain-adaptive, nonlinear expectation as a weighted update signal. MOVDA further introduces a learnable scaled tanh function to capture MOV saturation effects and home-court advantage. Built upon nonlinear regression and ELO-style online updating, MOVDA is evaluated on NBA temporal data from 2013–2023. Compared to TrueSkill, it reduces Brier score error by 1.54%, improves win/loss prediction accuracy by 0.58%, accelerates rating convergence by 13.5%, and retains ELO-level computational efficiency.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Game Theory and Economic Paradigms: Adversarial Learning
• Cognitive Modeling & Cognitive Systems: Adaptive Behavior

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Knowledge of accurate relative skills in any competitive system is essential, but foundational approaches such as ELO discard extremely relevant performance data by concentrating exclusively on binary outcomes. While margin of victory (MOV) extensions exist, they often lack a definitive method for incorporating this information. We introduce Margin of Victory Differential Analysis (MOVDA), a framework that enhances traditional rating systems by using the deviation between the true MOV and a $ extit{modeled expectation}$. MOVDA learns a domain-specific, non-linear function (a scaled hyperbolic tangent that captures saturation effects and home advantage) to predict expected MOV based on rating differentials. Crucially, the $ extit{difference}$ between the true and expected MOV provides a subtle and weighted signal for rating updates, highlighting informative deviations in all levels of contests. Extensive experiments on professional NBA basketball data (from 2013 to 2023, with 13,619 games) show that MOVDA significantly outperforms standard ELO and Bayesian baselines. MOVDA reduces Brier score prediction error by $1.54%$ compared to TrueSkill, increases outcome accuracy by $0.58%$, and most importantly accelerates rating convergence by $13.5%$, while maintaining the computational efficiency of the original ELO updates. MOVDA offers a theoretically motivated, empirically superior, and computationally lean approach to integrating performance magnitude into skill rating for competitive environments like the NBA.

Problem

• Improving skill ratings by incorporating margin of victory data
• Developing a method to model expected performance differentials
• Enhancing rating accuracy and convergence in competitive systems

Innovation

• Uses Margin of Victory Differential Analysis (MOVDA)
• Learns domain-specific non-linear function for predictions
• Enhances rating systems with true vs expected MOV difference