Recent findings regarding post-truth have led to increasing interest in exploring the role of social media in emotion management. Through communication strategies, media can form various models of emotional engagement of the audience. This paper focuses on the use of the emoticon by both media, all of which have a presence on Vk.com (these include, Lenta, Mayak, Meduza, Novaya Gazeta, Rossiyskaya Gazeta, and Russia Today) and commentators. With the help of a specially written Python script, we record the manifestations of commentators' emotional reactions. Using multidimensional scaling, we explore the particularity of emotional polarization.