Destruction in video games traditionally use pre-fracturing techniques to efficiently swap out game objects at run-time, which requires artists to create multiple models in various states of destruction. Real-time methods require less design time and can produce more realistic results but often come at a cost of performance. If real-time methods can perform at similar levels as tradition pre-fracturing means, this could increase the current standards of realistic destruction in video games. This paper explores and implements real-time fracturing techniques, comparing these to traditional pre-fracturing methods in terms of visual realism and performance. The results of the implementation were then distributed in an online questionnaire, to view how participants perceived the differences of both techniques and whether the visual or performance aspects affected their preferences of one method over the other.