The quality of video game localization depends not only on linguistic decisions but also on the conditions under which localized content is produced. In markets where official localization is limited, unofficial initiatives contribute to making international games accessible to local audiences, yet their textual quality remains underexplored. This study compares two Persian-dubbed games produced under different localization arrangements: the officially authorized Prince of Persia: The Lost Crown (Ubisoft, 2024) and the unofficial Resident Evil 3 (Capcom, 2020) localization by Darinoos. Drawing on Spiteri Miggiani’s (2022) Textual Parameters (TP) model, the study adopts a comparative, product-oriented approach to assess five dimensions of dubbing quality: fidelity, naturalness, lip synchronization, cohesion with visuals, and phonaesthetics. Quantitative error assessment is complemented by qualitative analysis of selected dialogue sequences. The Prince of Persia sample contains 11 minor errors and no major errors across 461 words, yielding a score of 97.61/100. The Resident Evil 3 sample contains 44 major and 25 minor errors across 421 words, yielding a score of 62.71/100. The findings point to differences particularly in fidelity, cohesion, and synchronization, with qualitative analysis revealing omissions, semantic distortions, terminological inconsistencies, and difficulties in rendering character-specific dialogue. The study demonstrates the analytical potential of textual parameters for assessing video game dubbing while highlighting the need to consider the interactive and multimedial characteristics of games when applying established dubbing assessment frameworks.