Abstract: Objectives: Global Navigation Satellite Systems (GNSS) are confronted with severe security threats due to their inherent vulnerabilities. Consequently, research on the evaluation of satellite navigation countermeasures has become increasingly prominent. However, technical performance evaluation under single-source jamming scenarios has been the center of attention, with limited attention given to the multi-attribute decision-making efficacy under distributed jamming conditions. This gap results in a disconnection between technical performance and tactical effectiveness. Multi-attribute decision-making methods have suffered from issues such as highly subjective weight allocation and insufficient consideration of index contribution, thereby reducing the reliability of evaluation results. Methods: A machine learning-based combination weighting method was firstly proposed. The interval analytic hierarchy process (AHP) was improved by integrating the random forest algorithm to replace traditional expert scoring, thereby achieving subjective weighting and reducing the high subjectivity inherent in the conventional AHP. The fuzzy rough set theory was applied to enhance the interval entropy weight method (EW) by replacing its technical intervals, enabling objective weighting and addressing the issue of uncertainty measurement in the EW. Furthermore, the least squares method and interval theory were incorporated into the combination assigning method, allowing for the determination of a new adjustment factor used to calculate the final weighting results. Finally, the fuzzy grey relational analysis method was utilized and developed seven hotspot countermeasure schemes for distributed GNSS jamming and anti-jamming. With the designed countermeasures, the proposed method finished index weighting, scheme selection, and their comprehensive validation, thereby realizing a multi-attribute decision-making efficacy evaluation for distributed GNSS jamming and anti-jamming. Results: (1) Regarding the jamming capability evaluation of distributed GNSS, the scheme of adopting the dynamic field experiment is the best., yielding the most efficacy evaluation results. Under equivalent cost conditions, the field test environment satisfies the dynamic requirements of multi-source cooperative GNSS jamming, thereby enhancing the utility value of performance test results and improving the accuracy of efficacy evaluation results. (2) In evaluating GNSS user anti-jamming capabilities, the scheme relying on software simulation is the best, achieving the highest performance evaluation results and assessment grades. This indicates that software simulation offers distinct advantages in reducing the complexity of the electromagnetic environment, enhancing redundancy in anti-jamming users, and moderating the influence of subjective decision-making. (3) Compared to conventional efficacy evaluation methods including the analytic hierarchy process, the entropy weight method, the combination assigning method, the grey relational analysis, and the technique for order preference by similarity to an ideal solution, the proposed assigning method reduces the root mean square error by 0.06 m and brings the coefficient of determination closer to 1, its selected scheme is optimal. This method also marks a significant step forward in advancing the evaluation research from static technical performance evaluation to dynamic tactical efficacy evaluation. Conclusions: The presented research outcomes are particularly applicable to complex confrontation scenarios, where they can serve as a reference for enhancing the performance of confrontation equipment, optimizing the deployment of distributed GNSS jammers, and strengthening the security protection capabilities of GNSS. Furthermore, these outcomes will provide a foundational basis for studies on dynamic game theory in the context of satellite navigation confrontation, thereby supporting the formulation of optimal decision-making strategies.