Modeling draw and search effects
Enter access cards in play order without overstating their contribution.
Draws are not deterministic searches
A draw takes a random top card; a deterministic search selects a matching card. Simply adding searchers to target copies ignores cases where the target is already in hand, no target remains in the deck, or the cost cannot be paid.
Assign cards a role tag and use the same tag as the search target. Different names can then represent the same functional role.
Describe top-N selection completely
Enter how many cards are viewed and taken, the matching tags, and the destination and order of the remainder. Bottoming in the same order and bottoming randomly create different future deck states. Record any approximation, such as treating an optional take as mandatory.
Include costs and action order
Configure discarded cards and resource payments. Drawing before searching and searching before drawing can produce different decks. Check that the sequence does not reuse a played card or consumed resource.
Audit an unexpectedly large gain
Calculate a no-effect baseline, then enable one effect at a time. If one effect produces a surprising jump, check its tag, take amount, cost, and shuffle setting. Re-run with the same seed to make the comparison reproducible.