Deck Selection For SCG Invitationals

Two-time SCG Legacy Open finalist Drew Levin tells you how he would build a blue midrange deck for the expected Legacy field at the SCG Invitational in Charlotte this weekend.


I won’t be attending the SCG Invitational in Charlotte this weekend. I’ve been incredibly busy with a new job (which I love), but the downside of that is the sad reality that I don’t have the time to travel to live Magic tournaments right now.

Fortunately for you, that means that you get an inside track on exactly what I think about Legacy right now. Whereas almost every other writer for this site is headed to Charlotte and wants to keep some parts of their process hidden, I have no such incentive.

I want to start by talking about something that to my knowledge nobody has published an article about: metagaming at an SCG Invitational. I can’t tell you much about picking a Standard deck (and there are a ton of people who are more than willing to tell you their thoughts on that format), but I’m familiar with picking decks in Legacy:

(For the record, #SCGINVI in New Jersey with Omni-Tell was another Top 32.)

I follow a very straightforward deck selection process. If I’m confident in my Standard deck, I play a blue control deck. If I’m not confident in my Standard deck, I play a combo deck. I will never play a fair deck with a bad combo matchup. Why only these two choices?

Because the best players in the room consistently choose to play blue decks in Invitationals. It might be combo, it might be midrange, it might be tempo, but a common feature of Invitationals is an above average density of Brainstorm. Given that the average density of Brainstorm decks in Legacy sits around 63%, it’s reasonable to expect to face six blue decks out of eight Legacy matches.

When you’re competing for Top 8, things get even bluer. There are a lot of possible narratives that account for this phenomenon, but it follows a similar pattern in Standard with control decks—a lot of people play a control deck and show up planning on "outplaying" a control mirror. Given how many people pursue this exact strategy, some of them—let’s call them "greater fools"—are definitely wrong about their capacity to outplay the upper crust of the event. If you want to win the event, your goal is to beat the greater fools as well as the sharks. As I mentioned above, there are two ways that I can see to go about that.

The first way is to play a blue midrange, tempo, or control deck. You’ll play games with a ton of decisions, plenty of ways to interact, and lots of built-in variance reducers in Brainstorm and Ponder. Your deck—U/W/R Delver, BUG Delver, RUG Delver, Esper Deathblade, Shardless BUG, or something similar—offers plenty of ways to metagame against archetypically similar decks, and there’s no shortage of technology available to leverage against various aspects of a given matchup. You’ll rely upon a vast knowledge of the format, a deep understanding of the possibilities that various matchups generate, and a more or less instinctual understanding of what cards are important in various spots.

The other way is to play a high-powered linear combo deck and hope that people either aren’t playing or don’t draw the cards that are good against you. You won’t be as in control of your destiny as if you were playing a midrange/control/tempo deck, but you’ll get a real number of free wins based on matchups. People have to play more obscure cards if they want to really beat you. Of course things could easily backfire, and you could run into an unfortunate streak of bad pairings and bad draws. This approach is inherently high variance, and you should accept that going into the tournament.

I would play a blue midrange deck only if I love my Standard deck because the idea behind playing a low-variance deck is the assumption that you’ll be in a position to do well going into your rounds (because you’ve been winning in Standard a lot, because your deck is great) so you want to reduce variance and manage a positive record. If things break your way, you’re liable to Top 8 the tournament, but you should lock up a money finish regardless.

By picking a combo deck, you acknowledge that you need to spike. Maybe you’re not prepared enough to make all of the interaction decisions that Legacy tournaments ask blue control decks to make over eight rounds, or maybe you just want to roll the dice. Regardless, the best reason to pick a linear combo deck for an Invitational is because you need to run above average to do well. There’s no shame in this—half the room is below average, and the sooner you realize that you’re in the bottom half, the more you can do to ameliorate your station in life. Magic is a game where the better player can lose any given game, so by identifying yourself as the underdog who needs to go on a hot streak, you can pick a deck that lets you run hot and destroy peoples’ day.

Today I’ll be talking about how I would build a blue midrange deck for an Invitational field. I expect people to primarily play similar-looking decks and combo decks, with a smattering of rogue decks with highly variable power levels. Our goal here is to find an angle to leverage against blue midrange/control/tempo decks, play enough disruption to split game one against combo decks, and sideboard a ton of combo hate and removal. We should be more or less pre-sideboarded against decks like Esper Deathblade.

My thought process for this question is documented below:

We want to play Stoneforge Mystic since it’s the most powerful creature in Legacy, especially in conjunction with fliers and True-Name Nemesis. That one-two punch alone can carry against fair decks, so we can build the rest of the deck to win against the mirror and combo. How do you get an edge in a Stoneforge Mystic mirror?

Well, we actually already solved this problem in the Caw-Blade Standard era—you Preordain for Stoneforge Mystic and have discard spells for their Batterskull or Sword. When it was about Squadron Hawk, you played repeatable removal. But since it was almost always about Stoneforge Mystic, it involved playing more discard spells and casting Mystic on turn 2 more often. So we want Stoneforge, True-Name, and Thoughtseize.

So we’re Esper Deathblade—how do we get an edge in our "good stuff" mirror? As I just mentioned, you can either have more good stuff, find your best stuff more, or deny them their good stuff. Let’s do all three.

Deathblade’s problem is that it’s kind of threat light. It has Stoneforge Mystic and True-Name Nemesis and then various planeswalkers, but it gives opponents a lot of time to do their thing. What if we had another card, like Mystic, where our opponent has to have an to answer it or die to it?

I hated Delver while playing it in BUG and loved Dark Confidant in literally every deck I’ve tried to play it in for the last few months, so this is an easily answered question. Besides, Daze is awful in a deck with important three-drops, and Spell Pierce is miserable—I just want to use all of my mana.

So we’re playing Thoughtseize, Brainstorm, Swords to Plowshares, Deathrite Shaman, Dark Confidant, Stoneforge Mystic, True-Name Nemesis, some Equipment, and probably fewer planeswalkers since we want to cast multiple spells in a turn. Losing to Wasteland isn’t on my agenda, so let’s keep the curve lower than a typical Deathblade deck.

Is there anything else that can be appropriated from U/W/R Delver? Daze is bad with True-Name, Pierce is bad without a ton of instants to provide mana efficiency, and we aren’t playing red for Lightning Bolt. What about Ponder?

Deathblade sometimes plays a Ponder. Why doesn’t it play all four? It has a huge range of cards—threats, removal, planeswalkers, counters, a ton of different-colored lands—basically a perfect Ponder deck. We can’t really play a lot of soft counters, but we’re well over on the Mind Twist side of the Mind Twist versus Suppression Field paradigm of blue resource attrition anyway.

If we play more copies of Ponder, we can play fewer lands, which means we have to play fewer planeswalkers. We probably don’t even need Jace, the Mind Sculptor if we have Dark Confidant. We still want Liliana of the Veil against True-Name Nemesis and to pressure Miracles, but Jace always feels kludge in Ponder decks.

Ponder also happens to solve the eternal "how do I get to sixteen blue cards for Force of Will?" problem that comes up in this deck all the time. Okay, sold. Let’s see how many one-ofs we can get in this thing—after all, if we have a ton of selection, we’re going to get paid for playing slightly different versions of stuff.

Here’s my starting 60 for a low-variance blue strategy. What are we trying to do with all of these changes? What are our upsides?

For starters, all of these cards are incredibly powerful. Our creatures are some of the best that Legacy has to offer. If we’re in a topdecking war, we’re almost certain to out-quality the opponent.

We have as much selection as Delver decks with fewer conditional cards and one more land (but several more colored sources). Our path to victory there is to fetch duals, cantrip into lands, and resolve whatever we can. Eventually, one of our threats will line up well against an opponent with Daze or Spell Pierce.

We have more selection than other Deathblade style decks, and we have twice as many game-breaking two-drops. Our opponents will have four copies of Stoneforge Mystic and a Ponder to our four Mystics, four copies of Dark Confidant, and four copies of Ponder. We’re favored to come out of the early game ahead on board.

Our primary attrition engine costs two, not three or four. This is meaningful against combo decks since we want to start churning through our deck to our disruption and countermagic as early as possible. Waiting until turn 5 to start drawing an extra card per turn against combo isn’t really feasible.

Because we cut much of our high end, we can play fewer lands. This is a double-edged sword—we’re more vulnerable to Wasteland, but we’re better at drawing spells in the midgame and specifically better at drawing the spells we want to in the midgame because of Ponder. Our deck is built to maneuver blue mirrors toward topdecking wars with a ton of cheap interaction.

We’ve cut down on True-Name Nemesis. This can be useful against combo decks, but it’s a clear liability against fair decks: tribal strategies, Mother of Runes decks, and so on. Something has to give somewhere, and this deck is wagering that most of the field won’t be bringing the beats. If that assumption is wrong, this deck is in a worse spot. Not terrible of course—Batterskull and Umezawa’s Jitte still do good work—but missing out on a Nemesis or two is a meaningful difference.

The idea against combo decks is pretty straightforward. Against the applicable decks, find a Deathrite Shaman ASAP. Against other decks, find Vendilion Clique and discard spells and eventually Karakas. Against the rest of the decks, Ponder into Force of Will and cast a two-drop, survive, and cast a discard spell per turn for the rest of the game.

The Swamp is there so that you have an out against Blood Moon— you have seven black fetch lands, it casts Deathrite Shaman, and you can cast your disruption and Dark Confidant and activate Shaman’s Grim Lavamancer ability with it. After sideboarding you can activate Shaman + Swamp to Abrupt Decay a Blood Moon or Back to Basics.

So what does the sideboard look like? Well, for starters, what don’t we want against various decks?

4 Deathrite Shaman: Never cutting this.

4 Stoneforge Mystic and the Equipment: Same.

4 Dark Confidant: We could reasonably cut this against aggressive decks, but we don’t actually want to since its existence in our list informs our anti-aggression strategy (a barrage of targeted removal, no sweepers like Engineered Explosives or Supreme Verdict).

2 True-Name Nemesis: Is this Trained Armodon? Cut. Is this Invisible Stalker or Moat? Keep.

1 Vendilion Clique: Best against Miracles, combo, and Stoneforge Mystic. Mediocre against fliers. Weak elsewhere.

1 Snapcaster Mage: Weak against Deathrite Shaman and Rest in Peace. Good against almost everything else.

4 Brainstorm and 4 Ponder: If you want to cut these, you should reread this entire article and then check the Modern banned list and the Vintage restricted list.

4 Swords to Plowshares: Do they have creatures? Can you target them?

3 Thoughtseize, 1 Cabal Therapy: If they have Stoneforge Mystic, keep these. If they don’t have Mystic and the matchup comes down to peeling, cut them.

2 Liliana of the Veil: Bad against lots of one-drops. Decent against combo, especially desirable against midrange creature decks and low-creature control decks.

3 Force of Will: Do they play spells that you would trade two cards to stop? If so, keep Force. If not, cut Force.

Once upon a time (a long time ago—you can check my archives if you’re dedicated to finding it), I wrote about how I build sideboards. I’ll be building this deck’s sideboard in the same fashion. Let’s start by listing all of the decks we want to have a real sideboard strategy against:

BUG Delver
U/W/R Delver
Esper Deathblade
Mother of Runes garbage
U/W/(R) Miracles
Sneak Attack
Painted with Blood Moon

Now for each strategy let’s list the cards that we don’t really want:

(-8) BUG Delver: Force of Will is going to be a bad draw against Hymn to Tourach, discard is going to be mediocre to outright bad in a topdecking war (which will happen if we’re alive), and Vendilion Clique lines up poorly against all of their threats (Tombstalker, Tarmogoyf).

End result: -3 Force of Will, -4 discard, -1 Vendilion Clique

(-5 to 7) U/W/R Delver: Force of Will is a bit better here (but still not good), as True-Name Nemesis and Stoneforge Mystic are worth countering against a ten-creature strategy. Discard is fine against a deck that is far more reactive than BUG Delver. Liliana of the Veil is a juicy target for Spell Pierce and Daze, but we still want answers to True-Name Nemesis. Whether or not that’s Liliana of the Veil isn’t the issue, but we can’t cut Liliana without adding ways to kill a Nemesis. Vendilion Clique and Snapcaster Mage are both poor against what they’re up to—Snapcaster turns on their Dazes and Spell Pierces in the late game, while Clique is a three-mana Delver of Secrets that gets Red Elemental Blasted (if they decide to keep that in). We also have a reasonable incentive to cut blue cards to make Red Elemental Blast as weak as possible, so Force of Will is that much less desirable.

End result: -3 Force of Will, -1 Vendilion Clique, -1 Snapcaster Mage, possibly -2 Liliana of the Veil

(-0 to 3) Esper Deathblade: We could cut Force of Will, but we’re basically pre-boarded against this matchup.

End result: -3 Force of Will (maybe)

(-9) Mother of Runes garbage: Force of Will is pretty weak against them since they have Aether Vial and a ton of creatures that are all mediocre but annoying. Liliana of the Veil is pretty weak to Thalia, Guardian of Thraben and the fact that they have close to thirty creatures, making it an overcosted Cruel Edict. We don’t want our discard after sideboarding since our plan is to kill all of their stuff, so we’re content to use our cards to kill their creatures as opposed to discarding their creature spells.

End result: -3 Force of Will, -4 discard, -2 Liliana of the Veil

(-4) UWR Miracles: We just want to build a board position and protect it. Incidentally, our maindeck is very good at doing all of that. They don’t play many creatures, so our Swords to Plowshares are pretty bad, but the rest of it is all right. We could board out True-Name Nemesis or Umezawa’s Jitte, but I don’t love having zero creatures that are immune to Jace’s -1.

End result: -4 Swords to Plowshares

(-4) Elves: We need Force of Will for Natural Order and Glimpse of Nature, we need pinpoint discard spells to break up their game plan, and we want some number of Plague Winds for their random 1/1s. Wirewood Symbiote is their best card against us, so we need to be able to kill that on sight, but we also don’t want to spend real removal spells on 1/1s. Our game plan is to set up Umezawa’s Jitte + True-Name Nemesis and to not get comboed in the meantime.

End result: -1 Vendilion Clique, -2 Liliana of the Veil, -1 Cabal Therapy because we don’t know when we’ll need to hit Glimpse of Nature or Natural Order and don’t have the luxury of being wrong

(-6) Sneak Attack: We’re never going to Swords to Plowshares a Griselbrand, and we don’t want Trained Armodon. We’re going to have to fight through Blood Moon and possibly Pyroclasm, so we should keep that in mind.

End result: -4 Swords to Plowshares, -2 True-Name Nemesis

(-6) Storm: They don’t have creatures (and we don’t care a ton if they have Dark Confidant because we have Umezawa’s Jitte and Liliana of the Veil), so usual anti-creature measures come out.

End result: -4 Swords to Plowshares, -2 True-Name Nemesis

(-6) Reanimator: These people do have creatures, but Deathrite Shaman is a really good card here. We still don’t want Swords to Plowshares since that’s like boarding in Glare of Heresy against an Esper Control deck—sure, you’ll get to cast it, but you’re going to die to whatever else they pulled off of their draw 7. We need to fight their bounce spells and Pithing Needle and Engineered Explosives and regular old reanimation package as well as their Show and Tell package, which every Reanimator deck plays nowadays.

End result: -4 Swords to Plowshares, -2 True-Name Nemesis. While we could keep Plows to have outs to Elesh Norn, Grand Cenobite and could keep Nemesis to have outs to Griselbrand, I’d rather just fight them on the stack and in their hand and graveyard.

(-5/7) Painter: Finally, a combo deck where we really want Swords to Plowshares. I don’t actually love Force of Will against Painter since they have so many Red Elemental Blast and are going to keep them all in against us because we’re a blue midrange deck that plays Force of Will. As such, I want to cut Force of Will, let them have their turn one Blood Moon wins, and focus on answering the rest of their cards. We should build a low-curve interactive plan that lets us use nonblue cards to destroy problem permanents, relying on on-board effects whenever possible so that we don’t have to keep mana up because we’re scared of Blood Moon.

End result: -3 Force of Will, -2 True-Name Nemesis, -0-2 Liliana of the Veil simply because it’s not very good against Imperial Recruiter. Besides, they empty their hand fairly quickly, so the +1 is of variable utility. We’ll probably end up keeping it, but it’s good to keep in mind.

So we have our counts of cards that we want to cut. Since we can’t play a 60-card sideboard, we need to find a lot of places where our needs overlap. For instance, the six cards that we board in against Storm, Reanimator, and Sneak Attack should have significant overlap. They don’t have to be all the same, but it doesn’t hurt. A first cut:

2 Meddling Mage
2 Swan Song
1 Pithing Needle
1 Jace, the Mind Sculptor
1 Surgical Extraction

Against Storm and Reanimator we don’t want Pithing Needle. Against Sneak Attack we don’t want Jace since they’re likely to Red Blast it or Daze or Pierce it. These seven cards also have good crossover utility elsewhere in our matchup directory. Since we don’t want Mage, Song, or Jace against Painter, what other cards do we want?

I think we need some number of Abrupt Decay and Seal of Primordium. Both of them do good work against Painter and Blood Moon as well as against basically the entire mono-white Mother of Runes garbage deck. Let’s split them to favor Abrupt Decay—the more powerful card—and see how Seal performs.

2 Abrupt Decay
2 Meddling Mage
2 Swan Song
1 Pithing Needle
1 Jace, the Mind Sculptor
1 Surgical Extraction
1 Seal of Primordium

This is our sideboard going into our grindier matchups. We’ll end up cutting some of it and trying to find different crossover cards as we go. For instance, against Elves we want four cards. We’re happy playing Pithing Needle on Wirewood Symbiote and casting Abrupt Decay on Deathrite Shaman, although we’d really like something such as Golgari Charm that doesn’t hit our creatures.

Fortunately, they printed that card in Zendikar:

So let’s play two of those and cut an Abrupt Decay, keeping in mind that we need another card for Painter at some point now.

2 Meddling Mage
2 Swan Song
2 Marsh Casualties
1 Pithing Needle
1 Jace, the Mind Sculptor
1 Surgical Extraction
1 Seal of Primordium
1 Abrupt Decay

Let’s talk about Miracles. We pretty clearly don’t want them to cast Terminus or Entreat the Angels, so Swan Song is actually great here. We have Jace, the Mind Sculptor to try to out-grind them and to mess with their ability to flip Sensei’s Divining Top for the top card of their deck, and we have Pithing Needle to name Top. I’d say we’re actually all right on this front. Let’s keep moving.

Against tempo decks we want to cut a lot of clunky cards for low-curve stuff that doesn’t get wrecked by a Wasteland or two. A good place to start would be Disfigure since it hits Deathrite Shaman, Delver of Secrets, and Stoneforge Mystic, three cards that we don’t want to overpay on trying to kill. Against U/W/R Delver we probably can’t cut Liliana of the Veil, as our Disfigure plan complements a plan of whittling them down to exactly True-Name and then activating Liliana’s -2. We still need to protect our Liliana, though, so let’s board the fourth Thoughtseize in to both break up their Mystics and try to clear the way for Liliana.

2 Disfigure
2 Meddling Mage
2 Swan Song
2 Marsh Casualties
1 Pithing Needle
1 Jace, the Mind Sculptor
1 Surgical Extraction
1 Seal of Primordium
1 Abrupt Decay
1 Thoughtseize

Against BUG Delver we’re definitely the control deck. We have more draw power and less attack power, so we should focus on killing their stuff. To that end we already have Disfigure, Abrupt Decay, and Jace. If we keep Thoughtseize in (which I don’t love), we only need one more card against them. Since we want another decent card against Reanimator, we can play a Sower of Temptation and try to cover their Tombstalker endgame that way.

So how do we look against all of our matchups now? Let’s go through them with our sideboard for reference:

2 Disfigure
2 Meddling Mage
2 Swan Song
2 Marsh Casualties
1 Pithing Needle
1 Jace, the Mind Sculptor
1 Surgical Extraction
1 Seal of Primordium
1 Abrupt Decay
1 Thoughtseize
1 Sower of Temptation

(-5) BUG Delver: +1 Sower of Temptation; 1 Abrupt Decay; 1 Jace, the Mind Sculptor; 2 Disfigure

(-5) U/W/R Delver: +2 Disfigure, 1 Thoughtseize, 1 Abrupt Decay, 1 Seal of Primordium

(-3) Esper Deathblade: +1 Jace, the Mind Sculptor; 1 Thoughtseize; 1 Abrupt Decay

(-9) Mother of Runes garbage: +2 Disfigure; 2 Marsh Casualties; 1 Pithing Needle; 1 Jace, the Mind Sculptor; 1 Seal of Primordium, 1 Abrupt Decay, 1 Sower of Temptation

(-4) U/W/(R) Miracles: +2 Swan Song; 1 Pithing Needle, 1 Jace, the Mind Sculptor (1 Thoughtseize, 1 Abrupt Decay if you want to cut True-Name Nemesis)

(-4) Elves: +2 Marsh Casualties, 2 Disfigure (1 Thoughtseize, 1 Abrupt Decay, 1 Pithing Needle if you want to keep cutting cards)

(-6) Sneak Attack: +2 Meddling Mage, 2 Swan Song, 1 Pithing Needle, 1 Thoughtseize, 1 Sower of Temptation, 1 Seal of Primordium (cutting Cabal Therapy and Umezawa’s Jitte most likely)

(-6) Storm: +2 Meddling Mage, 2 Swan Song, 1 Thoughtseize, 1 Surgical Extraction

(-6) Reanimator: +2 Swan Song; 1 Thoughtseize; 1 Surgical Extraction; 1 Sower of Temptation; 1 Jace, the Mind Sculptor

(-5) Painter with Blood Moon: +1 Pithing Needle, 1 Surgical Extraction, 1 Seal of Primordium, 1 Abrupt Decay, 1 Thoughtseize

And that is how you build a sideboard. Yes, it requires work, but you will never ever second-guess your pre-tournament choices in the three minutes you have after game 1. You’ll know for sure that you have a plan against the opposing deck, you’ll know what you need to do to execute it, and you’ll play better as a result. Incidentally, this is one of the reasons why people hate giving out sideboarding guides—it takes forever to actually go through this process, but not doing that and just giving you all a list of fifteen sideboard cards doesn’t really help you. It fosters dependency on us as a creative community, which is good for business, but at the end of the week, you haven’t learned much.

I hope you learned something today.