Looking for Group: Live Streaming Programming for Small Audiences
Live streams are used by some people to broadcast themselves doing creative work such as programming. To understand why individuals choose to stream themselves writing code, we interviewed eight streamers with small audiences of ten or fewer viewers. Several of these individuals were in a transitionary stage that supported a streaming lifestyle, and were seeking feedback and live companionship. These findings guide a discussion of the implications of creative live streams for people undergoing life transitions, and how learners might use streams to support their learning objectives.
KeywordsPersonal learning environments Live streaming Twitch
Live streaming, a practice of broadcasting real-time video of oneself, has become a popular medium for showing video game-play for the entertainment of viewers. However, people also use the medium for broadcasting skilled crafts as seen in the “creative” stream genre. In creative streams, people stream themselves engaged in skilled work such as knitting, music creation, or programming. Developers who host programming-focused streams broadcast their computer’s screen as they design, build, and debug software, typically augmented by a web camera featuring their face and voice.
Previous research has studied consistently popular streamers and their associated communities, but little is known about the experiences and motivations of streamers with smaller audiences who make up the majority of streamers on live streaming platforms. Streaming to small audiences can be demotivating to many individuals , but appears to be a standard mode of operation for these creative streamers. To understand what motivates these streamers to continue to broadcast, we interviewed eight streamers who broadcast their work in the programming directory on Twitch. We found that many of these individuals were in a transitionary life period that supported streaming, and that they streamed to get feedback on their programming and for companionship.
2 Background and Related Literature
Live streaming is broadcasting online video (‘streams’), typically of a person (the ‘streamer’) engaged in a task for others to watch. Often an audience interacts with the streamer and other audience members via text chat. While the predominant media featured on streaming sites like Twitch has been video games, non-game related content is also streamed for others’ entertainment and has been endorsed on the Twitch since 2015 with the addition of the Twitch ‘creative’ section .
As live streaming has gained popularity, researchers across the information disciplines have begun to study live streaming in earnest. One key theme in these early studies is that a strong motivating drive to host or view a stream on a live streaming platform is the opportunity for socialization [5, 8]. Additionally, viewers use Twitch to learn from observing the broadcasts . Due to this learning need, streams appear to be an early part of some users personal learning environments (PLE), helping them to find and answer questions . These PLEs are constructed of resources found across the internet, making them an example of networked individualism where individuals connect to different communities on the web for different needs .
Creative live streams found on Twitch have other educational uses. The programming and game development communities on Twitch share the social orientation of other streams and also feature a focus on skills desired by employers, such as web development or database programming. Additionally, researchers have found that many viewers return to future broadcasts in the programming directory of Twitch, and eventual form learning communities that support both the community members and stream host in growing to master the practice of programming .
In this paper, we extend this prior work to provide insight into the contexts of use that allow individuals to become streamers, before mentorship communities form.
In this study, we sought to understand how and why Twitch hosts start to stream themselves writing code, and their perceived benefits and continued motivators for streaming. We focused on programming-focused Twitch streamers who had small audiences (i.e., defined as less than 10 average audience members). This number was selected as it would place the selected streamers beneath the most popular channels in the programming directory, who averaged twelve or more viewers per broadcast.
Twitch’s listing of streamers in a given directory is limited to only those who are currently broadcasting. Thus, it is hard to get a sense of all of the streamers who have been broadcasting in a directory even from visiting the listing over the course of a few different hours every day. To get a broader picture of the individuals who were broadcasting and to identify small audience streamers, we used Twitch’s public API to gather an hourly listing of all the streamers who were actively streaming in the programming directory over the course of four weeks. We filtered the data to select streamers who had at least three unique broadcasts throughout each week and an average of ten viewers or less during those broadcasts.
We used the direct message feature of Twitch to invite eleven streamers who met these criteria to participate in a semi-structured interview. Eight individuals were interviewed via either text chat (3) or audio (5) based on their preference. Interviews conducted via text ranged from paragraph-length responses to one line responses, favoring single lines. Interviews done via voice lasted for thirty minutes. Demographics were not collected, though six of the eight appeared to be male-presenting based on their face cameras. Interview themes focused on motivations to stream, and how they or their streamers learned through engagement with the stream. Interviews were analyzed via thematic analysis, starting with an open coding pass. A subsequent axial coding pass was conducted with codes focused on personal learning environments.
Half of the streamers interviewed were in a transitionary life stage that supported a streaming-oriented lifestyle. The streamers reported two main motivations for streaming: for help on their work, and for companionship they were not able to get elsewhere.
4.1 Streaming During a Transitionary Life Stage
While all of the streamers used streams to enhance their working experience, half of the streamers were in a transitionary period between jobs, cities, or lifestyles. For these individuals, streams were a way to address deficiencies they encountered while they explored new interests and developed skills. Due to Twitch’s gaming focus, three of these individuals in transition between jobs or spaces were learning how to program games. After leaving a job, S5 chose to develop a game and stream in his newfound free time: “I had time. I decided to try something I’ve always wanted to do.”
While streaming could be used to hone job-related skills, most of these streamers were not seeking to enter the games industry. S6, when discussing the ability for his project in helping him to get a job, assumed he would be working in an area not related to games: “I’m probably not going to go to a AAA studio.” Managing the difference between developing the skills to get a job and becoming an effective streamer proved difficult for some. S1, for instance, had seen others struggle with this split in skills: “They introspect themselves and there’s a turning point – am I streamer or am I going to make a game?”
These four streamers were working to find new or steady employment, which had an effect of placing a deadline on their ability to be able to keep their demanding streaming schedules. Two of these ‘transitionary’ streamers had accepted full-time work by the time interviews were conducted. These streamers were unable to maintain their normal schedule as streamers when they started at their new jobs. S8 gave up streaming entirely, running one final “goodbye” stream for his audience before letting his channel become inactive. S6, who had also accepted a job, intended to continue streaming when he had free time. He felt an obligation to continue streaming for his regular viewers, but also noted that he would not be able to stream has much as he had in the past.
4.2 Motivations to Stream
While all the streamers were driven by some amount of curiosity about streaming, several also expressed a desire for assistance on their work or to find companionship they could not get in their current circumstances.
Help and Feedback.
Five streamers ran broadcasts with the goal of gathering help from viewers as they worked. S1, for example, began streaming for this reason, sharing that “the initial goal of streaming was to get feedback.” Three of these five were novice programmers or new to the development context that they had chosen to practice on stream. S7 had made a habit of using streams to seek assistance when learning something new. He perceived Twitch as a place where there was a lot of ambient expertise among the potential viewers that could help him when he was starting a new project: “I venture into new languages a lot, and at some point I’ll be a newbie at something.”
The streamers had success in obtaining viewers who were willing to provide help and who would return to assist during future streams. S5 was surprised by the dedication of some of the people who offered help: “There was one person who was there for every stream, and checking in on my progress.” This degree of investment from his viewers changed S5’s outlook on learning as a whole: “I didn’t feel bad learning programming. It was entertaining.”
Though many of the streamers used their streams to find help, three of the five help-seeking streamers found their viewers sometimes unable to offer quality assistance. S1 found the initial feedback useful, but as his skill developed the advice became counterproductive: “…you get a bit better, and then the average feedback is nice to have, but it’s not as useful.” S6 found himself often having to providing assistance in lieu of receiving help or advice: “There’s definitely an epidemic of young computer science students who stay up way too late and go to Twitch for help.”
Six streamers noted that they began streaming to address their solitary working conditions. S3 was looking to fill a social gap in his life: “I never really looked into doing anything with it other than attract like-minded people to communicate with while I did whatever I was streaming at the moment.” The streamers commented on the difficulties of working from home or being in situations where they could not form a social group. S1 and S3 were working from home and used streams to simulate the experience of being among coworkers in a shared workspace. For S1, streaming filled this gap by providing social interaction, or “some way to have some social time while you’re alone and in front of the computer.”
While some were looking only for companionship, others were interested in gathering viewers that were part of a specific community they did not have access to in their current location. S5 and S6 sought to create a community of viewers who were specifically interested in game development. For them, streaming was a method of advertising themselves and finding other people who could come and chat with them about a shared passion. S5 talked about how streaming had brought him a set of friends he was unable to find elsewhere: “I’ve met good friends through doing this. In my personal life, a lot of them don’t have the same core interests.”
In this paper, we describe motivations and practices of less established programming streamers on the streaming site Twitch. In the following section, we discuss challenges that streaming sites need to consider should they choose to support creative streamers. Additionally, we touch upon the emerging use of these streaming sites to gather help and discuss how streaming could play a part in a learner’s attempts to create a personal learning environment for themselves.
5.1 Design for Individuals Experiencing Transition
Half of our participants took up streaming during life transitions, though streaming sites are not necessarily designed to support individuals during these changes. Streaming sites could support streamers during their times of difficult transitions, including times when streamers’ lives become busier again and they are less able to stream. Streamers actively working to find a balance for their life may move past their streaming stage entirely as they get jobs or their life situations change. This creates several design implications for such transitionary use and communities.
First, the streamer’s needs during their transition should be considered. For the streamers, it appears that streaming supports a life transition by providing a daily ritual that helps to establish a “new normal” for the streamer . This ritual actively encourages the streamer to engage with people and tasks that will help ground themselves and give them access to social capital to reconnect and reskill. The practice of streaming could be positioned as a way for individuals who are in a transitionary state to get started on reskilling by guiding viewers to find and assist these streamers.
Not all transition tools are equally as useful for bringing an individual to a better life state. The life course approach notes that some experiences or circumstances will impact the quality of those circumstances that follow . Though streaming may be an effective tool for an individual to use to reskill, once they complete their transition the streamer will be forced to decide on the future of their channel. The streamer can either continue to stream to support the companionship needs of their viewers, or choose to disband their stream and lose the support of their community. Choosing to continue to stream after the transition has led to new circumstances like a job could be a large burden on the streamer as they will need to stream after work hours, potentially resulting in less access to future developmental opportunities.
5.2 Streams as Personal Learning Environments
Streaming was not designed to be a personal learning environment (PLE) for its users, but it is nonetheless used as such. As streamers broadcast they passively piece together the people, resources, and technology that make up PLE . PLEs often rely on the employment of interpersonal literacies that can grant access to individuals who can share their own knowledge . Streaming, with its focus on social interaction, appears to be an emerging literacy that can be used by learners to grow a learning environment centered on themselves, a process well-suited for the age of networked individualism where people use the internet to find disparate communities to fill different needs . The viewers of a stream appear to work together to fill several needs of the streamer simultaneously, such as those in this study receiving both learning and social input from their communities. Unlike other literacies that are proactive such as posting on a message board, streaming is initially passive for the streamer. Teachers or online learning platforms could encourage individuals to learn the basics of streaming and to run a series of broadcasts early on in their learning process, growing the student’s capability for interpersonal literacies along with giving them access to a set of tutors in their viewers.
5.3 Employing Amicable Liveness
Presence on live streams focuses on being welcoming and warm, which has implications for how the media could be used in both educational contexts. Many learners, such as the ones who came to broadcasts looking for help on projects, were seeking individuals who would be willing to assist on a project immediately. It appears as if streams were chosen by these students due to their combination of having someone who is currently on (or ‘live’) and had a helpful affect, something we have termed “amicable liveness”. While this type of presence is currently found in live streams and other, assistance-oriented contexts like paid peer tutoring, it could also be transferred to learning contexts. For instance, in a distance education context we imagine a user to be able to mark themselves “looking to help”, indicating that they would be happy to interact with others who are looking to work through a problem.
5.4 Limitations and Future Work
This study had several limitations that could be addressed in future work. The sample we gathered was limited to the four weeks that we mined the programming directory of streamers and may have changed in the next months. Most of the streamers we interviewed were male-presenting, as we did not strategically sample for gender. Our selection criteria of at least three broadcasts per week could have inadvertently lost early stage streamers who chose not to continue their streaming practice. Thus, this limited sample is likely not representative of all creative streamers, and future work could focus on strategically sampling populations that are likely to refrain from continued streaming. Finally, future work could focus on the types and outcomes of transitions that are best supported or most prevalent on live streaming platforms.
In this paper, we presented study of less popular programming streamers on the live streaming site Twitch. We found that these streamers were sometimes in a transitionary state, and used streaming as a way to seek help and companionship. Drawing from these findings, we identified ways in which streaming sites could explicitly design for people who take up streaming during a life transition, and potential applications of streams in supporting the learning needs for both streamers and viewers.
- 1.Attwell, G.: Personal learning environments-the future of eLearning? eLearn. Pap. 2(1), 1–8 (2007)Google Scholar
- 3.Faas, T., et al.: Watch me code: programming mentorship communities on Twitch.tv. In: Proceedings of the ACM on Human-Computer Interaction. CSCW, vol. 2, pp. 50:1–50:18 (2018)Google Scholar
- 4.Haaranen, L.: Programming as a performance: live-streaming and its implications for computer science education. In: Proceedings of the 2017 ACM Conference on Innovation and Technology in Computer Science Education, pp. 353–358. ACM, New York (2017)Google Scholar
- 5.Hamilton, W.A., et al.: Streaming on twitch: fostering participatory communities of play within live mixed media. In: Proceedings of the 32nd Annual ACM Conference on Human Factors in Computing Systems, pp. 1315–1324. ACM, New York (2014)Google Scholar
- 6.Kop, R.: The challenges to connectivist learning on open online networks: learning experiences during a massive open online course. Int. Rev. Res. Open Distrib. Learn. 12(3), 19–38 (2011)Google Scholar
- 7.Massimi, M., et al.: Finding a new normal: the role of technology in life disruptions. In: Proceedings of the ACM 2012 Conference on Computer Supported Cooperative Work, pp. 719–728. ACM, New York (2012)Google Scholar
- 8.Pellicone, A.J., Ahn, J.: The game of performing play: understanding streaming as cultural production. In: Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems, pp. 4863–4874 ACM, New York (2017)Google Scholar
- 9.Sjöblom, M., Hamari, J.: Why do people watch others play video games? An Empirical Study on the Motivations of Twitch Users. Social Science Research Network, Rochester, NY (2016)Google Scholar
- 10.Twitch: Introducing Twitch Creative. https://blog.twitch.tv/introducing-twitch-creative-fbfe23b4a114