When a game was developed, it is impossible to predict the success of the game. Most game developers made a game play test stage in game development process. It is important factor that gamer have good emotion playing the game. There are many ways of gamer's emotion test such questionnaire suvey, EEG test and etc.. but their tests are hard to quantify the results. Gamer's emotion are evolveing according to playing the game. Gamer's emotions are one of important criteria that is used for choosing a game by gamers. Nevertheless, the emotion detection is not considered for play test, because of the difficulty. The emotions of each gamers are different because of their own skills, experiences, preferences, etc. In recent years, the emotion detection technology is evolved, but it is not enough to guarantee the accuracy. In this paper, we propose a hybrid emotion classification system by EEG(Electroencephalogram) test, questionnaire suvey and eyetracking.