Night Dealings

Damage banks as theft counters, and those counters spend back out as a precise mana-value tutor: bank three, fetch a three-drop; bank six, fetch a bomb. The engine asks two questions at once. Can you deal damage to players fast enough to stock counters quicker than you burn them, and can you build a library where one named mana value is worth fetching for exactly the cost the activation prints? Firing the tutor costs as much as casting the enchantment did, so this is deliberately slow to cash out: not a set-and-forget value piece, and plain enchantment removal can strip the banked counters off the board before you ever spend them. The "another player" clause is the quiet constraint that shapes the whole deck: damage to creatures stores nothing, so the counters reward an aggressive board or a direct-burn shell rather than a grindy plan pointing its removal at blockers. That tension, an attrition tutor that only fills while you are actively winning the damage race, is why it has always read better on paper than it plays: by the time you have banked enough to fetch your best card, you are usually already the player ahead. A builder's puzzle for decks that deal incidental player damage anyway and want to convert that throughput into deterministic selection.