What you really want is the new version of the Sampler that we are working on now, which allows you to have multiple stages of selection linked together, so you could set up both “decks”, have it pull a card from the first deck and then the second deck, producing a single row of output from both.
Here are a couple of hacky ways you might get around that. In this document I set up two Samplers which would allow you to simulate each deck. (Note, I had to modify the first table some so that the second Sampler didn’t try to put it’s output into that same table.). Then I calculated a lookup value from the first table to combine the data from both “decks.”
In this version, there is one Sampler, but I added a categorical “index” attribute, so you could graph the result of each round of card selection.
Hope this helps.