War of the Spark spoilers wrap up this week with the full cardlist, but unless there’s a serious surprise in there at the level of Yawgmoth’s Will, I found my favorite card on the first day of previews with Bolas’s Citadel.

The Citadel is unique, but it also has elements of past cards—none of which has precisely been a powerhouse. The casting cost is Lich’s Mastery territory; the Peek ability is the minor bonus of Sphinx of Jwar Isle. The final ability is basically two copies of Kaervek’s Spite, a niche but hilarious card or half of Sorin’s Vengeance. None of this is encouraging—but then there’s that central ability, that core of the Citadel that lets you exchange life points for mana and cards, which has historically been broken. I’m honestly surprised that Wizards printed Bolas’s Citadel—trading life for cards is something they’ll allow in Standard (Ruin Raider, Midnight Reaper); but with the Citadel, you’re only bound by your life total and only beholden to whatever you can pull off the top. That puts Bolas’s Citadel in the rarified ranks of Necropotence, Yawgmoth’s Bargain, and, to me and only me, Lich’s Mastery.

I haven’t felt like this since—well, since last year, but I fall hard and fast. Basically, this was me the first time I saw the Citadel:

The similarity to Lich’s Mastery makes me wary—3BBB casting cost, punitive life-loss mechanic, the need to sacrifice permanents—but the Citadel does so in a way that gives you greater options rather than fewer. Not to mention the fact that BBB has always been the sweet spot—Necropotence, Lich’s Mastery, Dark Ritual, Bolas’s Citadel, Bridge from Below, Doomsday. There’s something about those three skulls all lined up that calls to me and has since 1997.

That’s not the only similarity the Citadel has to Necropotence: honestly, in a vacuum, the Citadel is scarier. Not only does it let you trade life for cards, but it lets you trade life for mana, harkening back to the old Channel/Fireball combo or the Yawgmoth’s Bargain into Skirge Familiar engine. It’s also a mirror-universe double to (and superb partner for) Aetherflux Reservoir. They both rely on high-velocity, spell-heavy decks and are dangerously cavalier about pacing—no coming into play tapped, no mana costs, simply the ability to trade one resource for another and win with a grand gesture.

That’s suitable, since the word “Citadel” has—especially for those of us who grew up near the military-training facility/chamber of horrors of the same name—a kind of resonant finality. The citadel is the bunker, the besieged stronghold. We get the word itself from the Italian “cittadella,” meaning “little city.” Citadels were supposed to be the place of final retreat—the most secure and defensible area to preserve the most important aspects of a city—food stores, rulers, etc. In War of the Spark, it’s the staging grounds for Nicol Bolas’s invasion—I can’t imagine there’s much in there, since the Dreadhorde doesn’t eat, doesn’t fear death, and is ruled by Bolas himself. Nonetheless, that apocalyptic vibe carries over, and now you can make your opponent feel like they’re trapped in the bunker as they watch you build your army. Like all invasions, this requires massive resources—3BBB isn’t cheap, you only have 8-14 life to play around with, depending on how much you’ve already taken, and if you get stuck on a clump of lands, your engine shuts down.

So the goal then becomes: how quickly can we chew through a clump of lands? Wayward Swordtooth may come in here, as may Deathsprout to shuffle and ramp to the Citadel, or we could rely on cantrips to speed us past clumps. We could pair it with Liliana, Untouched by Death, to mill ourselves, or with Twilight Prophet to get extra value once we’ve shaped the top of our library. We could also pair it with the recently spoiled Tamiyo, Collector of Tales, to guarantee we’re able to name the right card. We have options.

The second question: what keeps us proactive? What’s cheap to cast—first in mana, then in life—to dig for the Citadel and then provide enough velocity to keep going? What furthers our agenda—i.e., getting ten permanents on the board—and is good enough without a Citadel out? (If your deck can’t win against Surgical Extraction, you don’t have a deck; that’s something I need to get embroidered and mounted above my brewing zone.) Do we focus on lifegain, like Lich’s Mastery before it?

As I see it, there are a few tactics we could take. First: get your opponent to ten and then have ten non-land permanents out—treating Bolas’s Citadel like Banefire. This could be done in a hyperaggressive Rakdos deck—Gutterbones, Light up the Stage, etc.—or in a B/R/X control deck that uses the Citadel as the last bit of reach. Essentially, you’d play a linear aggro deck that runs two copies of Citadel and treat it like a combination of Kaervek’s Spite and Sorin’s Vengeance.

Functional, but we can do better. Here’s what that would look like, though:

Rakdos Aggro

Creatures (22)
Gutterbones
Rix Maadi Reveler
Tillonalli's Summoner
Graveyard Marshal
Dreadhorde Butcher
Judith, the Scourge Diva
Goblin Chainwhirler

Spells (15)
Lightning Strike
Angrath's Rampage
Bedevil
Rite of Belzenlok
Bolas's Citadel
Lands (23)
Dragonskull Summit
Blood Crypt
Unclaimed Territory
Mountain
Swamp

Second: Full combo! Frank Karsten on ChannelFireball had a write up last week of a possible combo variant, so I won’t belabor that, but I think it’s possible to run a deck of engines that the Citadel facilitates.

Third: Similar to Karsten’s thoughts, a Sultai Midrange/Control deck that runs all of the hits—Hydroid Krasis, Jadelight Ranger, some variety of Vraskas/Viviens, etc.—could use Bolas’s Citadel as a combination win condition and draw engine. The life you get back from Krasis and the exploration package that speeds you past lands supplement the Citadel well, and I’d test out Tamiyo in this deck. A pretty staid concept, but one that may see a fair amount of play in post-Spark Standard. See Ari Lax’s writeup for a deck draft.

But the best concept I’ve discovered by far—the one that feels right, that plays well, that wins—is Orzhov Treasure. Critically, Bolas’s Citadel says “nonland permanents,” and not “nontoken permanents.” So what’s an easy token to create that can give us fodder for the Citadel while ramping us up to it originally?

Filthy Lucre

Creatures (3)
Angel of Grace

Spells (33)
Smothering Tithe
Kaya's Wrath
Settle the Wreckage
Revel in Riches
Ob Nixilis, the Hate-Twisted
Revival // Revenge
Consecrate // Consume
Treasure Map
Orazca Relic
Karn, Scion of Urza
Bolas's Citadel
Lands (25)
Godless Shrine
Isolated Chapel
Orzhov Guildgate
Azorius Guildgate
Plaza of Harmony
Plains
Swamp

Essentially, this deck turns gold into warfare and then warfare back into gold—it’s imperialism, baby! You have Angel of Grace as your sole creature, and one that fuels the Citadel chain—instead of spending mana on cards, just reset your life to ten and use that instead. Ob Nixilis, the Hate-Twisted taxes your opponent’s card draw, and can either kill your own Angel for graveyard exploitation, or force your opponent to draw, take two damage, and give you Treasure.

Speaking of: the Treasure Map plus Karn package is robust—you don’t have to trigger the Citadel when you’re already winning with 6/6 Construct tokens—and the Orazca Relics are surprisingly useful in the deck. There’s also Revel in Riches, because you never know.

By the end of the week, we’ll know how War of the Spark wraps up—who will survive and what will be left of Ravnica and so forth. It doesn’t matter to me, because I feel like I’ve already won—any Standard format that lets me win by turning a fistful of Treasure into a death laser aimed directly at my opponent is a good Standard format.

A lifelong resident of the Carolinas and a graduate of the University of North Carolina, Rob has played Magic since he picked a Darkling Stalker up off the soccer field at summer camp. He works for nonprofits as an educational strategies developer and, in his off-hours, enjoys writing fiction, playing games, and exploring new beers.

Don't Miss Out!

Sign up for the Hipsters Newsletter for weekly updates.