Network games have become a popular social experience, while problem of in-game cheat (e.g. trade cheat, fraud and so on) has not an effective solution by now. In this paper, we propose a method to help players to prevent in-game cheat on the basis of group interaction analysis in massive multiplayer online role playing games (MMORPG). Frequency of group interactions reflects how possible two players like to play together. So, group interactions of all players can help a given player to distinguish between righteous players who have high frequency of group interactions with famous virtuous roles or his dependable friends and suspectable players who have high frequency of group interactions with notorious roles. In this paper, we develop an algorithm to visualize group interactions according to psychology of players and rules of the game system, and then we propose a simple method to prevent in-game cheat. Keywords network games, group activity, interaction, correlation, in-game cheat, degree of relationship.