My question revolves around which technique's the best when implementing a game with AndEngine. In this game a user is able to do simple touch movements: drag, pinch and tap on certain objects which are represented by just sprites at the moment. At this time, the game doesn't really use any physics. The only thing that's being done regarding the sprites is pixel perfect collision detection (between sprites) and adding and removing these Sprites at real time. Just using Sprites gets the job done, but I was wondering if using the Box2D extension would be a better fit?
Which are the positive and negative points regarding using Box2D extension vs just the Sprites? And does one outweigh the other?