Bilbo, Thief in the Night

Izzet Prowess can naturally stock its graveyard; a discounted extra spell each attack is credible value, though a 2/2 must survive combat first.

Could be wrong if The available spell decks may prefer cheaper proactive threats, and revealed context does not establish enough graveyard payoff density.

A repeat flashback-like trigger can enable Izzet Aggro turns, but Bilbo is slow on impact and vulnerable before its first attack.

Could be wrong if Pioneer removal and stronger two-drops may make a combat-dependent 2/2 too inefficient without a dedicated payoff.

Cutter Prowess can fill its graveyard and exploit an extra cast, but Bilbo is less reliable than Emry's 71-point artifact-recursion benchmark.

Could be wrong if Modern's cheap interaction may prevent Bilbo from attacking; unknown support determines whether its discount enables a real engine.

Phoenix can use the graveyard and spell trigger, but Legacy heavily punishes a two-mana creature that must live through combat; it trails Emry's 69 benchmark.

Could be wrong if Cheap artifacts and spells could make the cost reduction substantially more exploitable than the supplied context shows.

Repeat access to sacrificed artifacts or spent spells is a strong engine in matching blue decks, but attacking with a 2/2 is unreliable in multiplayer.

Could be wrong if Its adoption depends on how often supported decks can protect Bilbo and profit from one graveyard cast per combat.

A two-drop that can cast a graveyard spell on each attack creates major card advantage; its fragile 2/2 body and setup keep it short of bomb status.

Could be wrong if Unknown removal density, speed, and graveyard support could make attacking safely or stocking targets much harder than expected.