Recktenwald, D. (2017). Toward a transcription and analysis of live streaming on twitch.

Recktenwald, D. (2017). Toward a transcription and analysis of live streaming on twitch. Journal of Pragmatics, doi:10.1016/j.pragma.2017.01.013

Areas/Topics/Keywords: Transcription; Online live streaming; Pivoting; Twitch; Computer-mediated communication; Video-mediated communication; Games

This article looks at live video game streaming on Twitch.tv and proposes ways to transcribe communication of both the streamer and the participants to allow for conversational analysis. The method takes into account the number of participants involved, their types of communication, and their roles in relation to game events. The author presents a short literature review that points to the main research into live streaming. It is noted that the main areas of research are audience reasons for watching streams and ethnographic studies looking at the interaction of audience, broadcaster and game. The authors use schematic illustrations to describe the Twitch environment and illustrate the roles and relationships of the broadcaster and the audience.

Data was collected as live video and audio capture, two hours in length, of six streams. Streams were varied by genre and game pace. Audience chat was captured via Internet Relay Chat and saved for analysis as a plaintext file. The author lays out the coding schema used to manually transcribe the video streams and places them in a ladder like grid to show relationships between different participants chat and game events.

In analyzing the text, specific behaviours were revealed that corresponded to game events. The author presents a model for the concept of “pivoting”. This is described as an analog to the “syntactic pivot” found in conversation analysis. The author describes basis of this concept as “Salient and notable game events elicit responses from the participants in the form of written messages, spoken utterances or embodied communication. The content of these messages attributes the game event with its local meaning.”

The conversational analysis proposed in this article could be of use to libraries in the design of systems for HII or HCI , in cooperative and collaborative work, and multimedia educational settings.