Introduction

Since Kizuna Ai broke ground as the first person to label themselves a Virtual YouTuber (Vtuber) on YouTube in 2016, the Vtuber community has been steadily growing in both numbers and popularity.¹ It's become an increasingly influential medium, especially among the younger generation worldwide, making it one of the defining media trends of the 2010s and 2020s. Vtubers come in all shapes and forms, from gaming enthusiasts to product promoters. With billions of views worldwide, the Vtuber industry has firmly embedded itself in the capitalist landscape. Our research focuses on analyzing the online comments of Vtuber audiences to understand their interests and how they engage with Vtubers and their content.

Previous studies on Vtuber content consumers primarily rely on qualitative interviews or surveys with a small sample size. In contrast, our research utilizes web scraping (Python) to gather over 200,000 YouTube comments. By utilizing a concordancer (Voyant) for analysis, our goal is to uncover recurring patterns in word usage among users, thus enriching our comprehension of user experience. By comparing our findings with existing qualitative research, we can validate our results and gain deeper insights into Vtuber-content consumers and their interactions with creators.

Fan Interaction in Qualitative Research

In our research, we specifically concentrate on the comments posted on YouTube videos and streams featuring Vtuber Usada Pekora (兎田ぺこら). These sort of comments are part of the Viewer Space, one of the three dimensions of Vtuber content outlined in Gwillim-Thomas Patrick’s triadic model. The Viewer Space encompasses the viewer's personal environment, their emotional responses to the content, and their active engagement with the virtual world created by the Vtuber. It includes aspects such as the viewer's emotional connection to the Vtuber, their interpretation of the content, and their participation in discussions or interactions within the Vtuber's community.²

In a survey centered on the Viewer Space, which involved 669 participants investigating the parasocial relationship between fans and Vtubers amid the COVID-19 pandemic, intriguing findings emerged. Gender appeared to play a significant role, with women showing a stronger parasocial attachment to Vtubers compared to men. This connection seemed to be linked to the stress relief they found in consuming Vtuber content. Furthermore, age also played a key role, as it was noted that while Vtuber content is primarily consumed by individuals under 18, those aged 26-30 exhibited the strongest parasocial attachment. ³

This study, which utilized interviews with viewers, concluded that viewers felt a personal connection to certain Vtubers due to shared interests and hobbies. These content consumers can communicate, often live, but also delayed through YouTube comments or other platforms like Reddit with the Vtubers themselves or with other fans. These conversations are often very personal and cover many aspects of daily live as if it where conversations among close friends. This results in the creation of an authentic sense of amity among Vtuber content consumers. Part of this community atmosphere can be explained by the existence of many memes or inside jokes unique to a specific Vtuber community. These inside jokes, derived from memorable moments during streams, serve as a means for viewers to connect not only with the streamer but also with fellow fans. ⁴ Through social media, these memes and popular phrases/words have become integral to the interaction among fans, fostering a sense of camaraderie and sometimes exclusiveness within the community. An example is the use of particular greetings or aisatsu unique to certain Vtubers. ⁵ This creates an communal atmosphere among active viewers, who know these unique terms, even using the terms among themselves.

Quantitative Analysis of YouTube Comments under Vtuber Pekora

The results of our analysis of over 200.000 YouTube comments confirm the qualitative research conducted in the first section of our text (tools & methods used) One notable aspect of the Vtuber community is the ability for viewers to establish a profound connection with the avatar despite its virtuality. Analyzing the comments, it is evident that the language employed by Vtuber enthusiasts is notably informal and amical. An example being the extensive usage of the honorific ちゃん, found over 16.000 times in the scraped comments. In addition many supportive comments or words can be found, exemplified by the high frequency of words like 応援 (support) or おめでとう(congrats). Showcasing the feeling of friendship experienced by the viewers. Moreover, among the most frequently used words a large number of adjectives expressing positive emotions towards Pekora can be observed, e.g. いい (good), 好き (like, love), 嬉しい (happy), 可愛い (cute, lovely), 面白い (fun, interesting), 上手 (skillfull). This shows that Pekora’s fans prefer to engage in a positive discussion of the viewed content, mostly making use of compliments and expressing their positive feelings. Further highlighting the amical atmosphere described in the qualitative research section. Additionally this could mean that many of the viewers experience positive emotions while consuming Vtuber content, suggesting that the consumption relieves stress as Tan Y explained in his qualitative research.

Furthermore, through our analysis of all of the words we found multiple of the memes or popular phrases as Lee Miles described in his research. An example under Pekora’s video’s is the use of the unique aisatsu こんぺこ,こんぺこー or おつぺこ meaning respectively こんにちは (hello) and お疲れ様です (good job). Like described earlier this showcases the deeper connection between community members, in this case frequent viewers of the Pekora YouTube channel. The word こんぺこ and ぺこ were not included in the frequency list/graph due to Voyant not recognizing the word, but this could be solved by using a different method.

Conclusion

Through our quantitative research we were able to support existing qualitative research on the parasocial landscape of the Vtuber community. Based on discovered patterns in word usage we were able to confirm the amical atmosphere in which Vtuber fan interactions takes place. Nevertheless, as we only studied the comments under one specific Vtuber, it would be beneficial to expand this research to a more diverse audience as this could deepen our conducted research and provide new insights. Furthermore, it would be helpful to utilize a parser like MeCab instead of a concordancer (Voyant) as we used, as it would allow for a more detailed and less problematic analysis of the word and sentence usage in the comments.