Game Data Analytics using Descriptive and Predictive Mining

Abstract

The game industry is an industry that includes game development, marketing, and monetization. However, to be able to enter the game industry is not easy. Game developers must know how the market is going to be able to reap huge profits. By knowing the market situation, game developers can also determine whether the games made are in accordance with market conditions. Getting this information is not easy, especially for small game studios. In this research, we made a new application to find knowledge about games that are and will be trending. We used data mining is used to obtain this information. Data mining uses data from the Steam API to do clustering using the Hierarchical K-Means method and predictive using the Multiple Linear Regression method. The use of the Hierarchical K-Means method produces 3 clusters for the game's popularity level. The use of the Multiple Linear Regression method produces predictions of the game's popularity in the future. This new system will be able to help indie game studios to be able to obtain information about the condition of the gaming market thereby increasing the benefits that can be obtained.