Hugh Hunt's movies page
The motion here is governed by Newton's Laws of Motion which can be restated as:
The ball has the dynamic properties of a thin spherical shell.
OK, it's true that the ball probably didn't follow exactly this path, and it's almost certain that Lampard didn't put so much backspin on the ball.
Here is another scenario:
Again the ball has the dynamic properties of a thin spherical shell.
Now that there is some slip we see that Lampard didn't have to have put much backspin on the ball at all. But the inclusion of friction means that I've had to guess a value for the coefficient of friction. In fact, I can now produce just about any motion I like by trading off friction for initial spin. This means that we can't use the model to tell us what actually happened - unless we do some detailed experiments. Another thing is that I've assumed μ = 0.3 at both the cross bar and at the ground. It's very unlikely to be the same for both, so there's even more scope for "fitting" the results with two different values of μ . Note also that the value of μ will depend on how dry the ball, crossbar and grass are, how well inflated the ball is and also how fast it is going - the ball deforms more if it is going faster and so it will wrap itself around the cross bar by varying amounts.
What this all means is that football is an unpredictable game!