PREDIKSI KEMENANGAN DAN SUSUNAN TIM PADA GAME MOBILE LEGENDS BANG BANG MENGGUNAKAN ALGORITMA NAÏVE BAYES

Abstract

Mobile Legends is one of the games made by Moonton which is a MOBA (Multiplayer Online Battle Arena) type and is commonly played via Android and iOS smartphone media. The popularity of this game makes it officially contested at the local, national, and international levels. This study aims to calculate the prediction of victory and the composition of a team through specialty, role, and counter hero using the naïve Bayes algorithm and apply it to matches so that victory can be achieved. The use of parameters for calculating the probability of winning is calculated using the naïve bayes formula P (X | Ci) = P (x1 | Ci). P (x2 | Ci). P (x3 | Ci). P (x4 | Ci) where each variable represents the total win or loss, the total winrate of the specialty hero in one team, the type of hero role used, and the presence or absence of a counter hero from the opposing team in the allied team. The series of calculations is packaged in an application made through Microsoft Excel 2016 and VBA (Visual Basic for Application) with training data derived from the results of national and international matches or tournaments. The test results obtained from the specialty winrate training data have an effect of 91%, the type of role has an effect of 86%, and the presence of a counter hero has an effect of 60%, and for the win rate obtained, the winning results are 40 times out of a total of 50 matches which means that the application is made to have an accuracy of 80%. This result is higher than the training data which has a predicted win rate of 79% from the data collected and of course this research can be said to be successful. With such results, this research is expected to be able to help Mobile Legends players to choose the right hero lineup in their team and help gamers who aim to become Mobile Legends pro player.