Quantum results are hard to explain, but proven (by experiment) to be real. There's a particular mathematical/logical definition of something being 'real' and 'local', that I've still only half got my head around, and it should be true but isn't.

The main experiment is two particles that, if you check one, it affects what you'll see in the other in a particular, but subtle , way. And it's proven mathematically impossible to find an explanation where they don't either communicate faster than the speed of light (so, not 'local') but the effect actually happens ('real').

The trick is in the statistics - the pattern of results - that match up between the two particles in this very particular way. And one way to explain it is that different options are also happening, but in a different universe - i.e. every time two different things could happen, reality splits into two realities, one where this happens and one where that happens.

That's for specific quantum events, but some think those such quantum events underlie all choices and possibilities in reality. So, scale up that idea and you get 'infinite' (actually just very very many) parallel universes, one for every possibility that could ever have happened, branching off into more each time a (quantum) choice happens.