With the ever-increasing use of games, game developers are expected to write efficient code supporting several qualities such as security, maintainability, and performance. However, lack of time and continuous need to update the features of games may result in bad practices that may affect the functional and non-functional requirements of the game. These bad practices are often termed as Anti-patterns, which can cause technical debt, poor program comprehension, and can lead to several issues during software maintenance. While there exists empirical research on games, we are not aware of any work on understanding and cataloguing anti-patterns in games. Thus, we propose a catalogue of game-specific anti-patterns by mining commits, pull requests, and issues from 229 popular GitHub game repositories. We use regular expressions to create an initial dataset of open-source games and did thematic analysis on text records for cataloguing game-specific anti-patterns. We applied LDA (Latent Dirichlet Allocation) to validate further and refine the categories. We present a catalogue of 11 anti-patterns with a total of 20 subcategories involved in it and provide examples for them. We believe this catalogue would help game developers in making educated decisions during game development that can result in enhanced quality of games.

中文翻译：

---

特定于游戏的反模式目录

随着游戏使用量的不断增加，游戏开发人员需要编写支持安全性、可维护性和性能等多种品质的高效代码。然而，缺乏时间和持续更新游戏功能可能会导致不良做法，从而影响游戏的功能和非功能要求。这些不良做法通常被称为反模式，它会导致技术债务、程序理解能力差，并可能导致软件维护期间出现多个问题。虽然存在关于游戏的实证研究，但我们不知道有任何关于游戏中反模式理解和分类的工作。因此，我们通过从 229 个流行的 GitHub 游戏存储库中挖掘提交、拉取请求和问题，提出了一个特定于游戏的反模式目录。我们使用正则表达式来创建开源游戏的初始数据集，并对文本记录进行主题分析，以对特定于游戏的反模式进行编目。我们应用 LDA（潜在狄利克雷分配）来进一步验证和细化类别。我们提供了一个包含 11 个反模式的目录，总共涉及 20 个子类别，并为它们提供示例。我们相信该目录将帮助游戏开发者在游戏开发过程中做出明智的决定，从而提高游戏质量。