Building a Regression Model for Analyzing Social Engagement in the Global Interactive Entertainment Industry

The article presents the construction of a regression model based on information about the number of computer game players in the world and finding factors that affect the growth of this indicator. To avoid inadequacy of the model, outliers and factors that are strongly related to each other were excluded after formatting the data - thus, multicollinearity was excluded in this regression analysis. The main goal of the article is to build a linear regression model that will reflect all the coefficients needed to write the function for predicting the number of players in the world, assess the adequacy of the constructed model, and evaluate the residuals of comparing the predicted number of players with real data. If an adequate model is achieved (measuring the F-test, P-significance and R-square), in which the residuals of the predicted number and the real number will be close to zero, it is acceptable to assume that it can be used to predict the number of players in the world for the next year.