Abstract: This paper investigates the optimization of the geometric configuration of acoustic beacons in a long baseline (LBL) acoustic positioning system. A systematic analysis is conducted on the performance characteristics and application contexts of three evaluation metrics: the Cramer-Rao lower bound (CRLB), the Fisher information matrix (FIM), and the geometric dilution of precision (GDOP). Unlike radar or satellite positioning, underwater acoustic positioning is significantly affected by the complex marine environment. This study focuses on the application differences of three evaluation factors within a 300 m × 300 m monitoring area and their particularities in the underwater acoustic environment. The results show that hydroacoustic positioning is affected by the geometric configuration of the sound reference beacon, resulting in a more complex positioning error model. CRLB can more sensitively reflect the influence of this factor on the theoretical lower bound of errors. FIM requires a combined analysis of system configuration to assess information distribution, avoiding information imbalance caused by geometric configuration defects. The dimensionless nature of GDOP can intuitively demonstrate the geometric amplification effect of the spatial distribution of acoustic beacons on the positioning error. Based on these findings, this paper proposes a multi-index combined evaluation strategy: prioritize the use of GDOP for configuration screening,combine the CRLB to verify the theoretical limit accuracy, and supported by FIM to analyze information distribution characteristics. This integrated approach effectively compensates for the limitations of single-metric evaluation in complex hydroacoustic scenarios.