Non-random shared-space games can be solved by duplicating all of the components. E.g. you can play chess by mail if each player has their own board and two sets of pieces. Random shared-space games can be played so long as the players can have a shared awareness of the situation. E.g. you can play whist over Skype by putting your opponent's "hand" in a position where it can be seen by them, but not by you (your opponent will need to trust you to disable the reverse-facing "mirror cam" mode). Your opponent will tell you which card to play, numbered from their left (your right), e.g. "I pay the third card". @StephenBuckley's game, though, is neither random nor does it require a shared space. It doesn't actually even need a deck of cards, so long as the players have means to remember which numbers they've used. If you want to play a random, shared-space card with two separate sets, though, you'll need a real-world random seeding algorithm. I suppose it could be done, and it'd be pretty cool to try to do. For example: * All players "order" the decks - e.g. if we're playing Rummy or another shared-deck French-suited playing card game, we could put it in order from Ace to King in each suit, ordered Hearts, then Spaces, then Diamonds, then Clubs. We each then turn our deck face down.
* We "shuffle" the decks according to an algorithm produced by a random seed. This would ideally be done under the instruction of a computer program. So we both run the program and give it the same seed - a random number - and it then produces a list of instructions.
* Those instructions might be a little like this: deal cards face down into two piles, as follows - left, left, right, left, right, left, left, right, right... (etc.), place the left pile atop the right pile, deal cards face down into two piles, as follows - right, left, left, right, left, right, right... (etc.)
* Once the instructions have been followed, the two players have a pair of identically randomly-ordered decks, and can begin play.
* Play must be done carefully; e.g. I could say (via online chat, telephone, or whatever) "I deal seven cards to each of us, starting with you") and you'd know to put the top card of the deck into your hand, then the next card (face down) into my "hand", and so on, while I would do the reverse with my identical deck. We'd need to trust one another not to peep at each other's hands, and we'd need to not re-order our hands (at least; not without explaining the re-ordering to the other player), so that I could say "I play the 2nd card from my hand; the 8 of clubs" and be able to rely on the other person pulling the same card when they count the second card from their copy of my "hand". I suppose it's a little like quantum entanglement of playing cards. It'd probably feel quite spooky to have the same thing happen at both ends of an Internet connection, with physical decks. It'd probably work quite well for games like poker and blackjack, where drawing is rare, card order in hand is irrelevant, and few cards are drawn. It wouldn't work so well for games that required frequent shuffling (which would need to be done algorithmically, of course), and it wouldn't work at all for snap, of course! Is it crazy that I now have the urge to build a card shuffler that works to a seeded algorithm?