This article refers to this ongoing process of calculating same-sex dating goals through data gaming on Blued as algorithmic sociality

Drawing on the author’s own experiences during three years of using Blued, together with interview data of ordinary users and live streamers in Beijing and Shanghai, this article explores users’ use of algorithms as ritual tools, an angle that has been largely absent from recent discussions on algorithms. The first section discusses data, algorithms, and sociality in gay dating apps. The second section examines the ways that dating goals are algorithmically gamed on Blued browsing. The third section shifts to the analysis of Blued live streaming to investigate how the yanzhi algorithm is used to calculate dating preferences by both live streamers and viewers. In so doing, on the one hand, this article shifts away from the perception of algorithms being merely technical codes whose operation is black-boxed, secreted, and engineer-privatized, proposing instead that algorithms should be approached as users’ ritual tools. On the other hand, it facilitates understandings of algorithmic sociality and their implications for sexuality.

Algorithmic sociality

While algorithms have been primarily understood in a technical sense, critical scholars have proposed approaching algorithms politically, socially, and culturally (Beer, 2017 ; Gillespie, 2016a ; Kitchin, 2017 ; Neyland, 2015 ; Ziewitz, 2016 ). From a social scientific point of view, algorithms exist not only in codes but also in the social consciousness as part of a knowledge apparatus (Beer, 2017 ; Gillespie, 2014 ). Thus, algorithms should be examined within their discursive practices and framings as sensitizing devices rather than mere computational artifacts (Beer, 2017 ; Ziewitz, 2016 ).

Published online:

Before an algorithm can function meaningfully, it has to be paired with a given database (Gillespie, 2014 ). After all, algorithms are a result of translating items, actions, and processes into calculable and malleable units or data points (Willson, 2017 ). In the case of dating apps, there are two channels (user profiles and user activity) for building best places to live in Virginia Beach for singles a database (Albury, Burgess, Light, Race, Wilken, 2017 ). In the sign-up process, users are formalized into a few fixed data sets epitomized by a standard profile template including a headshot, age, height, weight, and so forth. User activity can also be rendered into calculable models in the interface design, breaking down their dating preferences into specific categories. For example, Blued browsing indexes user personalities according to 12 tags and divides gay live streamers into 4 categories, including ‘new stars’, ‘muscular’, ‘bears’, 4 and ‘groups’. In this way, users and their activities are tweaked into fixed data standards, protocols, and formats to be compatible with the calculative system upon which the algorithms act (Bucher, 2012b ; Neyland, 2015 ; Totaro Ninno, 2014 ).

However, the algorithms currently studied tend to portray users as passive audiences in their diverse forms of searching and matching, trending, recommendations, and newsfeeds (Bucher, 2012a , 2017 ; Gillespie, 2016b ; Striphas, 2015 ). However, studies on dating apps have found that users are in fact familiar with the data structures and sociotechnical operations of the apps and thereby can actively act on the data to shape the algorithmic results (Albury et al., 2017 ; David Cambre, 2016 ). For example, on Tinder, the number of profiles users are allowed to swipe for free is limited and a period of waiting is enforced before the next swipe. However, by changing the settings of sexual preferences, new profiles become available immediately (David Cambre, 2016 ). This is also a focus of this article – how users utilize the data they provide to dating apps, be it information about their physical body or visual images on the screen, to game and play with their algorithmic formations to shape their outcomes.