It's also a common misconception that Everett's many worlds involves an infinite number of universes.

And that it involves multiple outcomes for macro objects like lottery balls.

It only means multiple 'worlds' specifically for quantum outcomes, so in OP's case their winning or not winning the lottery would need to be dependent on a superposition of quanta (i.e. Schrodinger's lottery ticket).

And given the prevailing thinking is that there's a finite number of quanta in the universe, there cannot be an infinite number of parallel worlds. (There could only be an infinite number of aggregate worlds if time is infinite and there's perpetual quantum 'foam' in its final state perpetuating multiple possibilities).

The theory is much less interesting than is often depicted in mass media (though as of recently is a fair bit more interesting given the way many worlds as a theory would mirror what backpropagation of the physical universe might look like).

I thought many worlds meant every single possible divergent quantum thingy gets its own universe. There is a universe where a single potassium atom in a banana in my kitchen doesn't decay and there is another universe where that same potassium atom does decay. Multiply that for every single particle in the universe, right?

I guess even if that was true on a macroscopic level that's not going to guarantee that every possible thing happens?