I cannot know if some human in history was my genetic duplicate, just by random chance. But, I can hope that, for humanity’s sake, I only happened once.
Actually shuffling properly randomly doesn’t produce an order of cards that is convincingly shuffled. A problem for early bridge programs, in that real randomness has longer runs and more uneven suit distribution than a classic shuffle. So the actual number of “shuffled” card orders is way smaller than 52! and you may, indeed, see the same game more than once.
It is, and the number of possible branches of the wavefunction so far is staggering, even though it’s probably dwarfed by the potential number, which may or may not be infinite.
However, this isn’t how branching works in the Everettian many-worlds interpretation of quantum mechanics. Cards and decks thereof are classical objects. Their permutations don’t branch the wavefunction (branching here referring to parts of the wavefunction that are almost certain never to be able to interfere with each other again). Only quantum decoherence branches the wavefunction. So although the number of timelines in MWI is indeed more than large enough to have all possible permutations of all extent decks of cards, there’s no guarantee that any specific permutation of such classical objects has or will occur.
This is all assuming Everettian MWI of quantum mechanics is correct, which is far from settled.
Don’t mean to be a wet-blanket. The misunderstanding over classical vs quantum objects is just so common that it’s good to clarify when it comes up, since this is something that science communicators have usually done a poor job of communicating.