当代的游戏生产致力于改变性别与种族刻板印象，但是这种努力常常收效甚微。本研究基于“品性-能力”刻板印象模型以及计算文本的方法，对《英雄联盟》游戏角色刻板印象进行“编码-解码”分析。研究发现，在编码层面，游戏创作者致力于打破角色类型与性别种族的强关联，但是中国游戏玩家在认知角色时，依然代入了性别与种族的刻板印象，但部分刻板印象也会随着角色类型的多元化而消解。本研究构建了中文刻板印象内容词典，并对中国游戏玩家对《英雄联盟》游戏角色解读视频的弹幕文本进行计算分析。结果发现，与传统刻板印象一致，相比于女性角色，游戏玩家普遍对男性角色品性评价更低，能力评价更高；当女性被设定为英雄或亡命徒这样的传统游戏角色类型时，游戏玩家认为相关类型女性角色能力显著低于男性。但是当女性被设定为反英雄或统治者这样反传统的游戏角色类型时，游戏玩家则对其能力认知与男性同类型角色趋同，这一结果说明赋予女性角色反传统和多元的角色类型，如女性统治者和反英雄，有助于打破性别刻板印象。中国游戏玩家对黄种人和白人形象的游戏角色的能力评价显著高于其他有色人种，但是种族与角色类型对刻板印象认知影响的交互效应不显著。这说明中国游戏玩家对种族刻板印象具有持久性，而且角色类型多元化策略对种族刻板印象的改变作用不大。本研究扩展了刻板印象和“编码-解码”理论在游戏研究中的应用，通过计算文本的新方法工具说明游戏编码需要从多模态角度突破传统的刻板印象，以促进游戏玩家的多元认知。

Contemporary game production aims to change gender and racial stereotypes, but these efforts often yield minimal results. This study employs the “warmth-competence” stereotype model and computational text analysis to conduct a “encoding-decoding” analysis of character stereotypes in the game “League of Legends.” The findings reveal that at the encoding level, game creators strive to break the strong association between character types and gender/race. However, when Chinese players perceive characters, they still project gender and racial stereotypes onto them, although some stereotypes dissipate as character types diversify. This research constructs a Chinese stereotype content dictionary and conducts a computational analysis of the bullet comments from videos interpreting “League of Legends” characters by Chinese audiences. The results show that audiences rate male characters lower in warmth but higher in competence than female characters. Women portrayed in traditional roles, such as heroes, are perceived as having lower abilities, whereas those in non-traditional roles, such as anti-heroes, are rated similarly to their male counterparts. This suggests that non-traditional roles can help dismantle gender stereotypes. Additionally, Chinese audiences rate characters representing Asian and White individuals higher in competence than those representing other people of color. However, the interaction effect of race and character type on stereotype perception is not significant. This indicates that Chinese game audiences have a persistent view of racial stereotypes, and the strategy of diversifying character types has little impact on changing these racial stereotypes. This study illustrates the need for game encoding to shift away from traditional stereotypes, promoting diverse perspectives among audiences through multimodal approaches.