Not All the Bots Are Created Equal: The Ordering Turing Test for the Labeling of Bots in MMORPGs

Abstract

This article contributes to the research on bots in Social Media. It takes as its starting point an emerging perspective which proposes that we should abandon the investigation of the Turing Test and the functional aspects of bots in favor of studying the authentic and cooperative relationship between humans and bots. Contrary to this view, this article argues that Turing Tests are one of the ways in which authentic relationships between humans and bots take place. To understand this, this article introduces the concept of Ordering Turing Tests: these are sort of Turing Tests proposed by social actors for purposes of achieving social order when bots produce deviant behavior. An Ordering Turing Test is method for labeling deviance, whereby social actors can use this test to tell apart rule-abiding humans and rule-breaking bots. Using examples from Massively Multiplayer Online Role-Playing Games, this article illustrates how Ordering Turing Tests are proposed and justified by players and service providers. Data for the research comes from scientific literature on Machine Learning proposed for the identification of bots and from game forums and other player produced paratexts from the case study of the game Runescape.