Prediction of DOTA 2 Match Result by Using Analytical Hierarchy Process Method

Abstract

This research discuss about how to predict a match result by using data mining as the foundation to make a conclusion. In this research, Analytical Hierarchy Process (AHP) is used as a main data mining method for making a prediction. AHP works by using pre-existing data events as training data to obtain predictions of future data results. This research uses four parameters to predict the result of international Dota 2 match. Those parameters are experience per minutes (XPM) and gold per minutes (GPM), matchmaking ratio (MMR) point, 3 head to head matches result, and the last 10 results matches. In its implementation, the four parameters are given different rules and priority levels to support the performance of the AHP method. The result of this study has a satisfactory level of accuracy regarding the victory estimation of the match.