Oliver's twist

Oliver, a student who completed his IB last May, posted an interesting observation on facebook. He obviously hasn't started lectures yet at Uni so has been sitting in his room bouncing a ball off the wall. He noticed that the ball bounced as shown in his drawing. The interesting part is the final bounce which is quite vertical. His explanation was that the high friction between the floor (sticky with spilt beer, yeah definitely no lectures yet) hindered the horizontal component of the motion. I think this would be true if the ball was a cube but not for a ball. The friction would make the ball spin. So the first bounce causes the ball to spin clockwise. This spin remains as it bounces of the not so sticky wall (assuming beer did not splash up the wall when dropped on the floor). The clockwise spin now reduces the forward motion of the ball as it bounces for the second time. I made a simulation with Interactive Physics but I couldn't get the bounce vertical. I tried very high friction and perfectly elastic walls but still the ball went forwards a bit. If  made the ball spin by other means (setting it's properties) then I could get it to go vertical but that would assume Oliver was spinning the ball. I think its something peculiar to these very bouncy balls that can't be modelled by Interactive Physics. I've just noticed that in my simulation the ball isn't travelling fast enough to travel upwards after hitting the wall, when this happens more spin is given to the ball. I tried making a new simulation but still couldn't get it to go vertical.