diff --git a/paper.md b/paper.md index 4391afd..b18187d 100644 --- a/paper.md +++ b/paper.md @@ -60,7 +60,7 @@ When such software is available, it is often hard to use for less technically pr Esports can be treated as a subset of gaming with additional requirements for players, such as tournament presence, organized play, training, and professionalization [@Formosa2022]. The study of esports is multidisciplinary in nature [@Brock2023Interdisciplinary;@Pizzo2022Interdisciplinary]. Due to the growing academic interest in the area of gaming and esports [@Yamanaka2021Review;@Reitman2020Review;@Tang2023Review;@Bialecki2024Review], it is key to provide tools for researchers capable of simplifying the process of acquiring large datasets efficiently, not only for authors interested in the area of computer science [@Ferenczi2024sc2_serizlizer;@SmerdovLoLDataset2021]. -We focus on solving problems within the StarCraft 2 (SC2) infrastructure ecosystem. We solve the problem of ease of access to the data encoded in files with ".SC2Replay". StarCraft 2 is a real-time strategy game developed by Blizzard Entertainment. The game is known as one of the most prominent real-time strategy (RTS) esports titles [@Tyreal2020;@Dal2020]. It is also characterized by its fast-paced gameplay and a high skill ceiling [@Migliore2021]. These attributes make for a great environment for testing various AI agents [@Ma2024LLMStarCraft2,@Vinyals2019;@Samvelyan2019SMAC;@Pearce2022CSGO]. Moreover, research in StarCraft 2 is not limited to AI agents -- there are efforts to analyze the game from various perspectives and provide insights that can assist players in their gameplay [@URLSC2AICoach2022LLM;@URLSc2replaystats]. +We focus on solving problems within the StarCraft 2 (SC2) infrastructure ecosystem. We solve the problem of ease of access to the data encoded in files with ".SC2Replay". StarCraft 2 is a real-time strategy game developed by Blizzard Entertainment. The game is known as one of the most prominent real-time strategy (RTS) esports titles [@Tyreal2020;@Dal2020]. It is also characterized by its fast-paced gameplay and a high skill ceiling [@Migliore2021]. These attributes make for a great environment for testing various AI agents [@Ma2024LLMStarCraft2;@Vinyals2019;@Samvelyan2019SMAC;@Pearce2022CSGO]. Moreover, research in StarCraft 2 is not limited to AI agents -- there are efforts to analyze the game from various perspectives and provide insights that can assist players in their gameplay [@URLSC2AICoach2022LLM;@URLSc2replaystats]. So far, our software was leveraged in preparation of major datasets: "SC2ReSet" [@Bialecki2022ReSetZenodo] and "SC2EGSet" [@Bialecki2023EGSetZenodo] with an accompanying peer-reviewed and published Data Descriptor article [@Bialecki2023SC2EGSet]. Our goal for this work was to lower the technical knowledge required to obtain data from in-game replays. @@ -103,7 +103,7 @@ Main functionality of this software collection introduce a repeatable way of wor # Statement of Need -The need for similar solutions is clear, output of this software or related artifacts were used directly in varying contexts, the community cited our work, some authors following the general flow of our exploration [@Kim2024;@Ferenczi2024sc2_serizlizer]. Presented software was created to assist with the process of StarCraft 2 data processing. Mainly, the software fulfilled the research needs of our team, other collaborating research teams, and the scientific community as a whole. Additionally, as a result, the software was used to process and create a dataset shared openly [@Bialecki2023SC2EGSet]. Finally, an API interface was created to load and work with the data in PyTorch [@PyTorch2019] and PyTorch Lightning [@PyTorch_Lightning_2019]. +The need for similar solutions is clear. Output of this software or related artifacts were used directly in varying contexts, the community cited our work, some authors following the general flow of our exploration [@Kim2024;@Ferenczi2024sc2_serizlizer]. Presented software was created to assist with the process of StarCraft 2 data processing. Mainly, the software fulfilled the research needs of our team, other collaborating research teams, and the scientific community as a whole. Additionally, as a result, the software was used to process and create a dataset shared openly [@Bialecki2023SC2EGSet]. Finally, an API interface was created to load and work with the data in PyTorch [@PyTorch2019] and PyTorch Lightning [@PyTorch_Lightning_2019]. There exist many implementations built for the purpose of parsing replay files [@ZenodoSC2Reader]. These tools and libraries require expert programming skills to extract and interact with the resulting data. Many research approaches involve scientists that may not posses such expert knowledge in programming, but nonetheless interested in investigating esports (e.g., in psychology, biomechanics, social sciences and humanities -- SSH, and others) [@Kegelaers2025;@Dupuy2025;@Donghee2020]. Lowering the technical overhead needed to interact with in-game data can open gaming and esports to researchers with various non-technical backgrounds. Furthermore, integrating SSH scientists in the research process is not only a requirement in some funding programs, but also a practical necessity, if one aims to conduct socially responsible studies [@graf2019bringing;@sonetti2020only]. @@ -113,7 +113,7 @@ Before introducing our software, users were bound to write their own tools extra We conclude that despite there being some software packages available, they often require additional programming skills and knowledge. Our solution provides a simple to use executable file and a set of scripts to work with StarCraft 2 data. Additionally, we conclude that our software solves a very specific infrastructure problem that is prevalent in the gaming and esports research on StarCraft 2. -In its current version our toolset "SC2Tools" is capable of simplifying the work associated with handling files used to create StarCraft 2 datasets. We are planning to keep updating the software to include more tools, features, and functionalities. Additionally, due to the capability of our software to output JSON files, We claim full interoperability with other replay parsing solutions as long as they keep the same output format. +In its current version our toolset "SC2Tools" is capable of simplifying the work associated with handling files used to create StarCraft 2 datasets. We are planning to keep updating the software to include more tools, features, and functionalities. Additionally, due to the capability of our software to output JSON files. We claim full interoperability with other replay parsing solutions as long as they keep the same output format. # Conflict of Interest