The invention discloses a VR game intelligent management system based on big data feature recognition. Comprising a user head posture parameter acquisition module, a role view switching parameter comparison module, a role view operation sensitivity analysis module, a user operation gesture acquisition module, a role action conformity analysis module, a limb action parameter detection module, a limb action synchronous analysis module and a game operation test experience analysis module. According to the method, the visual field operation sensitivity coefficient, the hand action conformity and the limb action synchronous matching degree coefficient when the user operates the game role in the VR game picture are obtained, the comprehensive operation test experience coefficient of the user on the VR game is calculated, and if the comprehensive operation test experience coefficient is smaller than the preset VR game operation test experience coefficient threshold value, the corresponding VR game optimization processing measures are taken, so that the user experience is improved. Therefore, the experience feeling and the substitution feeling of the user on the VR game operation are improved, the viscosity of the user on the VR game is increased, and a foundation stone is provided for good development of the VR game.