Secrets of the Golden City
Three mana for two cards is a rate the game outgrew almost immediately; the ascend clause is the lever that tries to drag it back into relevance. Clear ten permanents and the same three mana buys a third card, which is where the spell stops being a Divination retread and becomes a genuine refill. The tension lives in the threshold itself: ten permanents is a board state, not a turn count, so the design rewards decks that flood the battlefield wide and shorts the control shell that would most want a midgame draw spell. That mismatch is the whole problem. The decks that reliably switch on the blessing (token swarms, go-wide aggro) rarely want to spend a turn standing still to draw, while the decks that crave the extra card are the slowest to reach the count. The blessing also sticks once earned, so a second copy of this card drawn later resolves at its upgraded rate for free, cashing in a threshold you already cleared; but the value only compounds for effects that specifically read the city's blessing, not for draw spells generally. By the time you would cast that second copy, a marginal third card is competing against far more impactful three-drops. It is honest conditional value: a floor that is playable when nothing better exists, and a meaningfully better card in exactly the deck least inclined to spend a turn casting it.
