I love secrets.frightened
It’s deliriously intoxicating to convince yourself that what you’re thinking about is worth keeping a secret.
It’s reaffirming to think that people would want to know what deck you’re playing if they could. One of the reasons anybody tries to innovate is to possess a tangible edge over somebody who lacks the same latent creativity.
I was going to keep my decklist for this tournament a secret just like any other writer.
A good friend convinced me that there was no legitimate downside to writing about my deck on the day before the tournament—if someone in the thousand-person Grand Prix field wanted to change their deck on the off chance that we play, more power to them. If someone wants to copy the list and ends up crushing with it, even better. If someone plays against me and is playing a deck that can make a mulligan decision based on knowledge of my archetype, I should be able to outplay them. If I lose, it’s because either my deck or I are worse than my opponent, not because they knew what I was playing.
So this article is going to discuss the 75 cards I’m sleeving up on Saturday and why.
I’m not the most creative deckbuilder in the room. My edge comes from how much I have researched Legacy as a format. All the decklists I’ve read, my heuristics in games, and my experience playing nearly every deck in the format give me a good background for the pacing of games in many different matchups. I haven’t been maximizing that edge lately.
I’ve played a lot of combo decks in the last year, done fine but not great with them, and always felt like I couldn’t maximize my knowledge of the format in gameplay. I knew what was going to happen, but I often couldn’t do anything to stop it.
I want to have more control over how my games play out.
I’ll be playing Miracles on Saturday.
Barring last-minute tweaks, this is the decklist I’m going to play.
I want to play Miracles because I think it is the strongest Jace, the Mind Sculptor deck in the format. I’ve done a lot of research on Magic Online, and Daily Events (may they Rest in Peace) always tell the same story: some RUG Delver, some Miracles, and some combo decks all do well.
Except that Miracles is constantly absent from Legacy Open Top 16s.
In my time as a data analyst, I have learned to trust high-volume data over low-volume data. I have also sought to understand why two very close to identical conditions for data production (Magic Online Daily Events and StarCityGames.com Legacy Opens) would create such disparate sets of results.
It has nothing to do with the price of cards since Force of Will and Lion’s Eye Diamond are still the most expensive cards on Magic Online yet blue decks are everywhere. People are clearly willing to shell out for success in the digital realm.
It has nothing to do with variance—it’s not like Miracles’ success is just a bunch of isolated incidents. It consistently gets multiple people paid off in every Daily Event. Same with RUG Delver. If you looked at just Magic Online Legacy events, you would have to conclude that Miracles is one of the best decks in the format.
Nobody even talks about it anymore.
People undervalue big data, but that’s not news.
What’s news is that people don’t talk about a control deck with a ton of removal, Sensei’s Divining Top and Counterbalance, and six basic lands in a format run by RUG Delver, Inc.
RUG Delver is weak to Rest in Peace, yet the best deck for Rest in Peace doesn’t get talked about. Come again?
I understand that Joe Lossett can’t lose with the deck, but at this point I doubt that there are other people out there trying.
Miracles is soft to midrange decks, is favored versus tempo decks, is even to slightly behind versus combo decks, and beats a lot of the random decks on sheer card power level. People playing “cute” strategies either can’t beat a one-mana Rout, the best Decree of Justice of all time, or Jace, the Mind Sculptor.
When I took a deep dive into what a typical Miracles list looks like, I found that there was no real consensus. Some lists played a lot of Rest in Peace, no creatures, and Helm of Obedience plus Enlightened Tutor. I hated how soft those decks were to literally anything. Show and Tell? Can’t beat it. Abrupt Decay? Can’t beat it. Liliana of the Veil? DEAD.
It quickly became clear that Miracles has to play some creatures in order to do things like block Delver of Secrets and kill combo decks and attack Liliana of the Veil. The problem is that Snapcaster Mage and Rest in Peace are at odds so you have to pick one and maximize that. Since Snapcaster Mage is pretty bad and Rest in Peace is awesome, that’s an easy choice to make except you still have to find a good creature.
Vendilion Clique is the perfect card for the job. It attacks for a significant amount, it’s very good against Show and Tell, it plays well with Karakas, and it has internal synergy with the miracle mechanic. The deck doesn’t want all four copies, but three is ideal.
Miracles lists have also been playing one or two Venser, Shaper Savant. I have nothing against Venser on a practical level—it’s a fine card and does what you’d expect it to do—but the lists playing a copy of Venser were only playing three copies of Jace, the Mind Sculptor.
This gave me pause.
Venser is a fine card against Show and Tell. It’s passable against other combo decks, but you basically want it as a way to bounce their permanent, which is fine except Show and Tell decks nowadays tend to cast Sneak Attack and activate it, not cast Show and Tell. If they do cast Show and Tell, they put in Sneak Attack and activate it.
This is all by way of saying that Venser, Shaper Savant is a four-drop in a blue control deck with three copies of Jace, the Mind Sculptor. At some point it seems like someone would have tried playing four Jaces and zero Vensers.
The strong adherence to previous lists was a pretty serious tipoff that people weren’t working that hard on updating the deck; they were just fiddling with a few numbers and putting up another 3-1 in a Daily Event.
So I started working on the deck.
I wanted to keep the things that make the deck good—a powerful late game, a basic-heavy mana base that can be used to get past RUG Delver’s Wastelands in the early game, a removal-heavy configuration that buys the time needed to bridge to the late game, and the ability to play four Sensei’s Divining Tops.
Miracles also gets to play Counterbalance, although it’s far worse now than it was a few years ago. Curves are more spread out, miracles compound that problem, and the lack of powerful defensive blue or white two-drops makes the two slot far more dubious than it used to be. Old Counterbalance curves looked like this:
Ones: 12-14 – Usually Brainstorm, Sensei’s Divining Top, Swords to Plowshares, and either some number of Enlightened Tutors or Spell Snares
Twos: 10-12 – Tarmogoyf, Counterbalance, and a few counterspells
Threes: 2-4 – Vendilion Clique and Firespout
Fours: 3 – Jace, the Mind Sculptor
Fives: 4 – Force of Will
You could reliably lock out ones and twos, unreliably tag some threes, and miraculously counter their Natural Order or Jace, the Mind Sculptor with your own Jace.
This deck’s curve is way worse:
Ones: 13 – Brainstorm, Sensei’s Divining Top, Swords to Plowshares, Spell Pierce
Two: 6 – Counterbalance, Counterspell, and the lone Rest in Peace
Threes: 6 – Vendilion Clique and Entreat the Angels
Fours: 5 – Jace, the Mind Sculptor and Supreme Verdict
Fives: 4 – Force of Will
Sixes: 4 – Terminus
To make matters worse, miracles tend to be cards that you want to leave on top of your deck, so Terminus has a lot of fundamental tension with Counterbalance. Entreat the Angels is better mostly because there are a lot of good three-drops that the deck wants to counter nowadays. Still, Terminus has made Counterbalance tangibly worse.
The “problem” really is that Terminus is one of the main draws to the deck. My theory of the case is that this is the best tournament for Terminus in the history of its existence. What contributes to that?
Tempo is at a high water mark, and Terminus is really hard to interact with outside of Stifle. Once you Stifle a Terminus, you then have to counter every Brainstorm they draw or get Terminused anyway.
Midrange is heavier on creatures than it used to be. They have to be that way since they need to kill combo decks in a reasonable time frame.
Death and Taxes is well positioned, and if there’s one thing that a White Weenie deck doesn’t want to see, it’s a one-mana Rout. They have ways to mitigate the damage by playing around it, but it’s still really bad for them.
Sneak and Show is built on the assumption that none of its creatures really die to removal spells. Terminus changes that paradigm.
People are jamming True-Name Nemesis in their deck as fast as they can because they think it’s unkillable. Have you met my good friend Terminus?
Seriously, though, if we’re having a fight where one of us has four Terminuses and the other one has a six-turn clock, I’ll take Terminus.
It’s so good in fact that you’ll note I would rather max out on Terminus than Swords to Plowshares. Since Plow is weak against Mother of Runes and Emrakul, I would rather play fewer of that and more Terminuses. I understand that this creates more opening hands with Terminus in them, but with eight Brainstorm / Jace, the Mind Sculptor, I’m not overly concerned.
The other reason that I want to play this deck is that Entreat the Angels is close to unbeatable for most non-combo decks. People like to talk about how Miracles can’t beat this card or that card when what they really mean is “as long as Miracles doesn’t make three 4/4 fliers in my combat step, they can’t beat this card.”
People can’t leave in dedicated creature removal against a three-creature control deck with five Wrath of Gods. They can’t. They’ll lose to Jace very easily if they do. That makes Entreat the Angels a very potent way to turn the corner—to shift the focus of the game from playing defense to playing offense. A lot of Miracles decks play only one Entreat the Angels.
I don’t think people respect the ability to Entreat for one.
Outside of Tarmogoyf, RUG Delver can’t do a lot about a single 4/4 Angel token. They can trade two Bolts or a creature and a Bolt for it, but the time that that gives Miracles to line up another answer is incredible.
Against Sneak and Show, an end-step Entreat for one forces them to actually move forward on trying to win the game.
An Angel token eats basically any non-Germ, non-Tarmogoyf attacker. People can’t just not attack into open mana—this is Legacy, not Theros Limited. Yes, Entreat the Angels is the best Divine Verdict ever printed. People still have to attack into open mana.
Playing seven Miracles instead of four or five adds variance to the early game in the form of having them in your opening hand. This isn’t unique to Miracles though—early-game variance already exists for every single other deck in the format. RUG Delver can’t mulligan a hand with a blue source. Combo decks often have to play into Force of Will. Aggressive decks can just lose to combo. I would rather maximize the power of my draw steps than my opening hands in a format where you can’t afford to mulligan a lot of the time. Besides, as I mentioned earlier, the deck does play ways to reset miracles.
Jace, the Mind Sculptor is at his actual best here. I do not believe there is a better deck in which someone could play Jace.
This is a threat-light control deck, so having more Jaces means having access to more cards that actually kill your opponent.
This is a control deck that has ten cards—Counterbalance and the seven miracles—that care about the top card of its deck. Jace casts Brainstorm on demand.
Brainstorm is also at its best here. Instead of simply drawing three desired cards and putting two dead ones back to shuffle away, this deck can often draw three good ones and turn the two worst cards in its hand into the two best cards it casts over the next two turns.
The deck’s counterspells—Spell Pierce, Counterspell, and Force of Will—are all important ways to interact with combo and control, although I constantly find myself wanting one more. It is possible that with the addition of the fourth Terminus over the fourth Swords to Plowshares the deck simply doesn’t need Supreme Verdict anymore.
The singleton Rest in Peace has overperformed in testing. It buys a lot of time against any Tarmogoyf or Deathrite Shaman deck, it wins game 1s against graveyard combo decks, and it has a lot of fringe benefits in other matchups. It isn’t so universally useful that the second copy is worth considering, but it has value both as a graveyard hoser and as a two-drop for Counterbalance.
The mana base is fairly straightforward. Given this deck’s emphasis on Sensei’s Divining Top, it wants nine-to-ten fetch lands. Since it has such a high curve, it needs to be able to fetch or naturally draw basic lands against RUG Delver. Two Plains and two Islands can cast every maindeck spell, so there are two basic Plains in this primarily blue deck. The decision to play four basic Islands is a deductive conclusion based on the reduced utility of all other options. To wit:
The deck doesn’t need more than two Tundras, as the second Tundra allows it to cast any one spell in the deck. The third Tundra simply increases the likelihood of naturally drawing a dual land. Since this deck wants to limit the utility of opposing Wastelands, naturally drawing a hand filled with non-fetch nonbasic lands is fairly bad.
The same logic applies to Volcanic Island. The deck doesn’t want a basic Mountain, but it also doesn’t want only one red source, ergo two Volcanic Islands. It is possible that Plateau is worth considering, but the deck already plays a lot of nonblue sources and you want seventeen (or more) blue sources in a deck that wants to cast Counterspell and Counterbalance.
Karakas is a very good card for this deck. It bounces our own Vendilion Cliques and opposing Thalia, Guardian of Thrabens as well as Emrakul, the Aeons Torn; Griselbrand; and Gaddock Teeg. The second one is an unfortunate draw sometimes, but the upside of having one more by playing two copies outweighs the downside of drawing both.
Mystic Gate is the weirdest land, but it functions both as Choke protection and as a way to cast Counterbalance off of a Karakas or basic Plains. A second copy is a disaster, but the first copy is basically always value.
The deck wants 22-23 lands. If I end up cutting Supreme Verdict, it will likely be for a fifth Island, as losing to Delver of Secrets is not on my weekend to-do list. The deck functions well with 22 lands, but missing land drops early is a disaster. Sensei’s Divining Top can smooth out a lot of awkward draws, but so does a 23rd land.
The sideboard concept is what pushed me toward this deck in the first place. Legacy’s metagame is as focused as it has ever been, so the power of a silver-bullet strategy is more potent than at other point in the format’s history.
Rest in Peace has a lot of value against U/G/x tempo and midrange decks, so a second copy helps stabilize against any number of Mongooses and Tarmogoyfs.
Aura of Silence is good both against Painter’s Servant strategies—existing both as a zero-mana instant-speed removal spell and a huge impediment to their ability to play their cards—and against Sneak Attack, where it can force an opponent to muster seven mana to cast and activate Sneak Attack once, at which point Karakas or Terminus can stop the creature while Aura of Silence cleans up the enchantment.
Blood Moon is a must-counter card for any three-color midrange or tempo deck. After sideboarding, my plan against tempo decks shifts from “navigate into a position to resolve a haymaker” to “keep casting spells until one sticks.” There is a much higher density of trumps against tempo and midrange decks after sideboarding, so it’s easier to let any single spell get countered.
Blood Moon. Pierce it? Fine.
Swords to Plowshares. Force it? Fine.
Entreat the Angels. Game.
Ensnaring Bridge puts Sneak Attack and Reanimator decks on bounce spells or Jace, the Mind Sculptor to even play Magic from that point on. There isn’t another card that does that except for Humility, and I don’t want Humility, Vendilion Clique, and Entreat the Angels in the same deck.
Pithing Needle names the usual suspects—Sneak Attack, Griselbrand, Liliana of the Veil, equipment, Aether Vial, and so on. It’s mostly there for Sneak Attack though.
Oblivion Ring is all-purpose tutorable permanent removal that doesn’t hit our own Jace or get Red Elemental Blasted. I do not want Detention Sphere.
The rest of the cards shore up various weaknesses, although it is also worth considering whether Swords to Plowshares is the best card we can sideboard in against creature decks.
Moat shuts off attacking decks with few to no fliers. Candidates include Elves, Goblins, Merfolk, and some True-Name Nemesis decks.
I don’t think I need to tell you what Wear // Tear or Pyroblast are there for.
I want Swan Song over Flusterstorm for Sneak Attack.
Elspeth, Knight Errant comes in against the more ponderous midrange decks that pressure us with Liliana of the Veil. I am still very much on the fence about it, as it has far less utility than the rest of the sideboard.
There are several cards that I’m interested in improving—the maindeck Supreme Verdict and the sideboard Swords to Plowshares and Elspeth, Knight-Errant—and I’m not entirely sure what they should be.
I would like either a 23rd land or another defensive two-drop of appropriate power level (no Azorius Charms, folks) in the maindeck. As far as the sideboard goes, I think it is well positioned against creature strategies and likely needs help with combo decks. It is not clear what the right cards are in that capacity. I welcome your feedback.
Thanks for reading throughout this week. I hope that this series has been able to help some of you prepare for this Grand Prix, and I look forward to battling on Saturday.
See you in Chantilly!
@drewlevin on Twitter