Afaik it's an open question whether in an event like the statistical distribution of all bullet trajectories as we know it classically (e.g. based on some aerodynamics model) is even remotely similar to the distribution of quantum states.
In others it could be that all classical variation boils up from little quantum fluctuations. But it could also be that the vast majority of quantum fluctuations led to the same result. If you then go further back and time and run the simulation again, because the next most dense group of multiverses might be one when he didn't fire the shot at all.
A few months ago this was brought up in a Sean Caroll AMA.
To take the example of a fair coin toss: what looks like 50/50 odds classically may or may not be 50/50 odds quantum mechanically.
Or in programmer terms: all classical random number generators are deterministic. The question is whether classic random number generators always rely on (a multitude of) underlying quantum random number generators to produce their result. Or whether they just seem random enough to us but would be just as much without any quantum effect.