This study explores emotion recognition in videogames using electroencephalographic (EEG) data. Presently, emotion recognition using pattern recognition techniques has not yet been investigated in videogame play. This research is motivated by the possibility of retrieving insights into player experience from EEG signal during gameplay, which aims to contribute to Games User Research as an emerging discipline in the study of videogame design and their interaction with the players. In order to investigate emotion recognition several approaches for feature vector creation and classification algorithms were employed in order to assess which combination offered higher accuracy in classification. A maximum of 33.48% of classification accuracy was achieved by the Nearest Mean Classifier in the classification of four different emotions. Such low results suggests the collection and pre-processing of data from a dynamic activity, such as videogame play demands novel approaches for filtering the EEG, rejecting of artifacts and selection of the emotional model into which map the EEG brainwave oscillations.