STATISTICAL ANALYSIS OF THE INFUENCE OF USER-GENERATED CONTENT ON VIDEO GAME DEMAND

Introduction. Video games are part of a multi-billion dollar economy with a large network of game developers, business investors, publishers working to conceptualize, create and sell games to a community of millions of people. These include thousands of journalists, amateurs, and professionals who add descriptive text tags to each game to help describe the game to others and allow the community to navigate through terms, topics, and genres. A good sale becomes a matter of making a profit or return on investment and ensuring the continuous operation of the business, and the possibility of predicting the achievement of a certain threshold of break-even sales will help to make a more informed business decision. Purpose. The purpose of the study is to conduct a statistical analysis of the effect of a set of product marks on the attainment of a video game ownership threshold of more than 500,000 users on the Steam game distribution platform. Materials and methods. The research materials of the study are: 1) the dataset “All Steam Spiele und deren Metadaten”, which includes a complete data collection of more than 80,000 games available on the Steam platform together with their metadata; 2) the works of domestic and foreign authors conducting scientific and practical research on the video game market, in particular regarding its impact on the economy. In the process of carrying out the research, the following scientific methods were used: accumulation of facts related to the object of research (preparation of data for further analysis); analysis of facts, properties, factors and phenomena according to various indicators and criteria (estimation of the distribution of video games by the number of owners, comparison, classification and systematization of video games according to qualitative characteristics); formalization, analysis and synthesis (to build a logistic regression of the dependence of reaching the threshold of ownership of a video game by more than 500,000 users on the quality characteristics of a video game; logical generalization of the results (formulation of conclusions). Results. The article is devoted to the analysis of the influence of a set of text marks (tags) of the product on reaching the threshold of ownership of a video game on the Steam platform, to the identification of the relationship between the characteristics of games and their commercial success. The author puts forward a hypothesis regarding the influence of descriptive product marks (tags) on the demand and, accordingly, on the commercial success of video games on the Steam platform. Logistic regression models were constructed to investigate this relationship. ROC analysis was performed to evaluate the binary classification, to evaluate the sensitivity and specificity of the model. It was found that the recognition abilities of the model, taking into account sensitivity and specificity, are high, but the model better classifies negative cases. Prospects. Further research suggests focusing on additional aspects of the context, such as marketing campaigns, game release time, market competition, and other factors that may also influence game success. Including these factors in the model can improve its predictive capabilities.