The Road To DC: Your Delver Primer

Leading up to Grand Prix Washington DC this weekend, Drew Levin will be providing in-depth Legacy content every day! Today he continues with the three varieties of Delver.

To a casual observer, all Delver decks look the same. When I first saw a Canadian Threshold list (the predecessor to RUG Delver), it scanned like this:

4 Nimble Mongoose

4 Tarmogoyf

Like 32 blue cards (seriously)

4 Lightning Bolt

4 Wasteland

Some blue dual lands and fetch lands, but wow that’s a small number of lands


Some removal spells

Some counterspells

Some graveyard hate

What does that card even do? That’s a weird one-of; why would you want that?

And so on.

I can appreciate the dizzy feeling that comes on after reading through the second or third Delver of Secrets / Volcanic Island list in a row. I want to help you understand the differences between the green, white, and black versions of the deck, their strengths and weaknesses, and the matchups in which each deck excels. I’ll start by discussing what all three decks share, and then I’ll break each deck down by its unique strategic elements.

All three Delver decks—Grixis, RUG, and U/W/R—are so-called “tempo” decks. What does that mean in a practical sense?

They’re all built with the expectation of trading cards on a one-for-one basis and being advantaged in an ensuing low-resource situation. Here’s a thought experiment that might illustrate the point:

You and your friends are playing a match of Legacy. You both agree as a joke to mulligan to zero. This obviously leaves a lot to chance, but who has the best odds of winning the game—the person with Delver of Secrets or the person with Jace, the Mind Sculptor?

All three of these decks play a lot of cards that cost one or zero mana and trade for a broad cross-section of playable Legacy cards. Because cards like Lightning Bolt, Spell Pierce, Stifle, Wasteland, and Daze are so inexpensive—a game developer would say their “rate” is very good, as they are more powerful than their cost lets on—they often trade “above rate,” meaning that they trade one-for-one on cards but you spend less mana than your opponent in that interaction.

Here’s an example:

You play a Taiga and a Wild Nacatl. Your opponent plays a Forest.

You play Plateau, attack for three, cast another Wild Nacatl, and pass.

Your opponent plays another Forest and a Grizzly Bears. You Lightning Bolt their Grizzly Bears using your Plateau.

The outcome? You have a huge board advantage, which you will convert into a life total advantage very quickly. You may have used one more card than your opponent so far, but if you can keep trading cards with your opponent—your Path to Exile for their next creature, your Pacifism for the one after that—you can kill them before their cards have a meaningful impact on the game.

The idea, of course, is that you built your deck to create this situation over and over again.

Except you don’t just play removal spells for creatures. You play removal spells for lands, creatures, and noncreatures, and they cost zero or one mana each.

So we know that if you spend one mana to answer their three-drop, you’re effectively “saving” two mana. How do you realize that gain?

Maybe you didn’t play a third land. If someone is paying three mana for their creature and you’re paying one to kill it, why would you need a third land? Every time they play a land, that’s one spell fewer that they can play.

As you will see, every Delver tempo deck plays eighteen-to-twenty lands with a curve that stops at two, while most midrange and control decks play 21-23 and a curve that stops at four.

These decks use Wasteland the best. Wasteland is an incredibly powerful card capable of generating free wins by virtue of Legacy’s land-light nature. People have to keep land-light hands and hope to draw out of it—that’s how the format is built. Mana bases are powerful but unstable, and Delver decks exploit that fact every round by Wastelanding people off of the board.

One side of the coin is that once you’re done trading off cards with your opponent you’ll both be low on lands and cards. Your spells have to cost very little in order to support your plan to thrive in a low-resource situation.

The other side of the coin is that your threats have to be just as cheap as your spells or you’ll regularly face agonizing turns where you choose between interacting with their position and developing your own. If you have three lands, Spell Pierce, Lightning Bolt, and a creature that you want to cast this turn, would you rather your creature cost one, two, or three mana?

This incidentally is why Delver of Secrets and True-Name Nemesis are unlikely to be great friends; Delver of Secrets wants you to build a deck that trades off cards, while True-Name Nemesis wants you to build a deck that allows you to ignore a meaningful subset of your opponent’s cards, negating their impact without spending cards or mana on that negation.

Now let’s go over the core of a Volcanic Island / Delver of Secrets tempo deck:

4 Brainstorm, 4 Ponder, 4 Lightning Bolt, 4 Force of Will, 4 Delver of Secrets, 4 Wasteland: These 24 cards are the defining cards of this archetype. Cutting any of them requires a very good reason—and even then you’re probably still wrong to do it.

Brainstorm and Ponder fix your mana and find you the right card for a given situation. Lightning Bolt is the best red card in Legacy, especially since you aren’t just playing it as a three-point removal spell. A lot of the time Lightning Bolt lets you win races against an overwhelming board disadvantage because Insectile Aberration doesn’t care about how many, say, Tarmogoyfs they have.

Force of Will is a huge reason to play blue in this format. It lets you fight over important spells even as you’re deploying threats or taking care of other problems. The card that it costs you is meaningful, but your goal is to end the game before they get to capitalize on their card advantage.

Wasteland, as I discussed earlier, is a great way to get free wins. Play four, Wasteland them on the draw when they don’t have a threat in play, and remember that it even taps for mana!

Every Delver of Secrets deck plays six or seven dual lands and eight or nine fetch lands in addition to the aforementioned Wastelands. Although most people don’t take much time to consider which fetch lands they put in their deck, it matters once you’re in a match.

If your opponent leads on Polluted Delta, what goes through your mind? Is it meaningfully different from what goes through your mind if they lead on Scalding Tarn or Misty Rainforest?

What about Wooded Foothills? Bloodstained Mire? Arid Mesa?

People are conditioned by their experiences, and it takes very few iterations of “response, crack Scalding Tarn, Stifle your fetch land” for someone to start fearing Stifle whenever they see a Misty Rainforest or Scalding Tarn. Getting your first land Stifled is a hugely negative experience, and people try to avoid those whenever possible. People will develop heuristics to understand when they’re in such a situation.

It also takes relatively little Legacy experience to fear a combo deck—whether it’s Storm, Reanimator, or something else—whenever you see a Polluted Delta and no play.

Contrast these reactions with the thoughts that “Wooded Foothills, pass” inspires. No one gets killed on turn 2 after someone plays a Wooded Foothills. Wooded Foothills finds two of the fairest colors in Magic. Nothing bad happens to someone playing against Wooded Foothills. Maybe you’ll get attacked a bit, but that’s it.

Same thing with Bloodstained Mire—except nowadays people will likely put you on being Jund and having a Hymn to Tourach or Dark Confidant on your second turn.

I bring this up because no three-color Delver of Secrets deck plays basic lands or nonblue dual lands, so any shard or wedge combination has five fetch lands that it can use: the four blue fetch lands and the one that finds both of its nonblue colors.

This is useful information if you care to exploit it. I use a very straightforward level 1 heuristic for choosing fetch lands in Delver decks:

If I’m playing Stifle, I play zero Misty Rainforests and zero Scalding Tarns. If I’m not playing Stifle, I play only Misty Rainforests and Scalding Tarns.

This seeks to exploit the pattern recognition latent in many players’ minds regarding RUG Delver and Stifle. Since almost all RUG Delver decks play Stifle and since RUG Delver was the most prominent Delver deck for a very long time, people tend to associate blue-red and blue-green fetch lands with RUG Delver, which they then associate with the card Stifle.

If I’m playing Stifle, I don’t want people to consider the possibility that I have it. I want to catch them unaware, snagging their only land off of my Wooded Foothills activation.

If I’m not playing Stifle, I want people to constantly play around it. I want them to take suboptimal lines because they fear losing to Stifle. I’ll play as though I have it, selling them their own fearful narrative—“but what if he has Stifle and I lose my only land?”—and then capitalizing on all of the time that they wasted by beating them with a Dark Confidant and two Dazes, a Young Pyromancer that immediately churns out three tokens, or a Stoneforge Mystic and a Spell Pierce.

After those core 38-40 cards, the most typical four-of card is Daze.

Daze is not an objectively powerful card. If you only read what the card does, it’s a worse Stymied Hopes.

The power of Daze is that like Force of Will you cast it for its alternative cost over 95% of the time. When your curve stops at two, picking up your second land is often costless and can even be advantageous. Consider the following:

You have three lands in play. You cast a Tarmogoyf. Your opponent taps out to Counterspell it. You Daze it, returning a tapped dual land. Your opponent cannot pay, and your Tarmogoyf resolves. With your last untapped land, you Brainstorm, putting back your dual land. You play a fetch land and shuffle away the two lands on top of your deck.

In this (fairly common) scenario, your Daze effectively drew you a card on top of being a zero-mana Counterspell. It turned your useless fourth land into a fresh card from the top of your deck, and your fetch land locked in that exchange. If you want to know how Delver decks out-attrition Jace decks, it has a lot to do with managing land drops and maximizing Daze’s “drawback” in conjunction with Brainstorm and fetch lands.

I would not recommend cutting any Dazes from your Delver deck either.

This leaves the core of a Volcanic Island / Delver of Secrets tempo deck at:

8 fetch lands

6 dual lands

4 Wasteland

4 Delver of Secrets

4 Brainstorm

4 Ponder

4 Lightning Bolt

4 Daze

4 Force of Will

Leaving us with eighteen slots to mess around with. Of those eighteen slots, we want at least eight creatures. This is because dating back to the invention of Canadian Threshold in the mid-2000s these types of decks have always played at least ten creatures. The early versions played four Nimble Mongooses and four Werebears, thus the moniker “Threshold.”

White versions took the theme seriously, topping off the curve with a pair of Mystic Enforcers and the occasional pair of maindeck Meddling Mages. Red versions played Fledgling Dragon, and black versions didn’t exist because black doesn’t have a one-mana removal spell anywhere close to as good as Lightning Bolt or Swords to Plowshares.

But I digress. My point is that the bare minimum for threats is ten, with twelve being the industry standard and more than twelve creatures creating negative returns for your Delver of Secrets triggers.

Let’s get into the deck-by-deck specifics.

I’ll be upfront about this: if I were playing a Delver deck in Washington, it would be the following decklist.

The green version of Delver is notable for having a playable in-color one-drop: Nimble Mongoose. I could write an entire article about why Nimble Mongoose is an excellent creature in a Delver tempo deck, but I’ll keep it to a few paragraphs.

To understand why Nimble Mongoose is good, you have to understand how certain decks try to beat tempo decks. Delver decks play relatively few threats, as I mentioned above. This is to make sure that they have the right balance of threats and cards that keep an opponent busy while the threat(s) kill them. If you draw all your threats, your opponent might just execute their game plan, and you could die before they do. That’s no good.

This means that Delver decks usually have to fight over a few opposing removal spells or else they’re wasting all of the time that they create by trading cards back and forth. Nimble Mongoose dodges a lot of commonly played removal as well as Jace, the Mind Sculptor’s onerous -1 ability. This lets the Delver deck focus on fighting over blockers, expensive high-impact spells, and very little else. Opposing pinpoint removal gets stranded by Mongoose’s shroud keyword. That’s a huge deal.

It’s also a big deal that Nimble Mongoose only costs one mana. Since it only costs one mana, you can cast it and another spell on turn 2. You can set up your hand with a cantrip, kill an opposing two-drop with Lightning Bolt, counter a powerful spell with Spell Pierce, or Stifle their second land. The ability to deploy a threat and interact on turn 2 is a huge improvement over something like Geist of Saint Traft, Dark Confidant, or Young Pyromancer—these are all powerful cards in their own rights, but their respective casting costs have meaningful implications for ensuing deckbuilding decisions.

Tarmogoyf is your other creature, and it is still the best creature in Magic in the category “has power and toughness.” Since we care very much about how fast our opponent dies, playing four Tarmogoyfs is non-negotiable.

That leaves ten slots for spells.

This is where having Nimble Mongoose in the deck influences our choices.

Because we have Nimble Mongoose, we want to play cards that cost one mana. Most of our threats cost one mana, so it’s far more likely that we’ll have mana available on our opponent’s turn. Given that reality, we should look for the best one-mana blue instants.

Stifle is a high-utility, low-power card that fits this deck’s needs. Here is the list of things it does that this deck wants to do:

While it may not perform vital functions in every matchup, it is a usefully disruptive card that scares the daylights out of people. Once someone knows that you have Stifle in your deck, they will respect it constantly. This manifests differently for different people, but people tend to make mistakes—typically by skewing too conservative in their play style—when they play against Stifle.

The other card that I would play four of is Spell Pierce. I simply don’t understand why Delver decks play two or three of the card when it’s incredibly powerful at every point of the game. Early on it counters cantrips. In the midgame it counters removal spells. In the late game it counters important combo pieces and Jaces.

Every deck in Legacy is tight on mana, and every deck in Legacy plays noncreature spells. Even against Elves—a deck with eleven noncreature spells—Spell Pierce is still good because it cuts off important mana against Glimpse of Nature, it counters early Green Sun’s Zeniths, and it counters decisive Natural Orders.

Instead of maxing out on Spell Pierce, people have been playing two-to-three Gitaxian Probes. Now, I appreciate a Phyrexian mana spell as much as the next person, but playing Peek when you could play Spell Pierce is mind-boggling. I understand the argument that you want to know if the coast is clear to tap low for your threat against combo, but why not just have a Spell Pierce in hand and use it to counter the first spell they play, immediately cast your threat, and hope that you aren’t dead?

Spell Pierceing a combo deck’s Preordain is a very worthy use of a counterspell. I don’t understand why people let cantrips resolve with the intention of countering the Infernal Tutor or Show and Tell or Reanimate or whatever only to lose counter wars because the combo deck got to manipulate their draws every turn. Just counter their cantrips! Those are really good cards. You play them because they’re really good. Don’t let theirs resolve. Play enough Spell Pierces to counter their cantrips and then have some left over to counter their Sneak Attack.

Don’t get me wrong about Gitaxian Probe—it’s a reasonable card that’s playable in Legacy, but I never ever want to draw two of them against anyone if I’m a RUG Delver deck. I would play one, but it gets sideboarded out against any deck that’s attacking me.

The last card needs to be a red removal spell. The candidates are Chain Lightning, Forked Bolt, and Fire // Ice. Since Elves and Death and Taxes have gotten more popular, I want either Forked Bolt or Fire // Ice because both kill two cards—an important quality for a removal spell when their deck is chock full of cheap x/1 creatures.

The tiebreaker is Sneak and Show’s prominence. While Forked Bolt is better against Mother of Runes, RUG Delver never breaks through on the ground against Mother of Runes decks. It flips a Delver, protects it at all costs, and fights through whatever fliers they might deploy as blockers. On the other hand, Sneak and Show can put in a Griselbrand, which Ice can tap. Against Sneak and Show, tapping their Griselbrand is about as good as killing it.

The sideboard is as always a melange of one- and two-ofs. Rough // Tumble headlines as the only three-of in the sideboard of a deck that has only five removal spells. It is an excellent card against the other two tempo decks in addition to being a complete blowout against Elves and Empty the Warrens. As with Limited, you can play Wrath of God in your aggro deck as long as you take even a little bit of care to set it up properly.

Vendilion Clique is in the sideboard for its ability to actually disrupt a Sneak and Show opponent without tapping out on your turn, which is a combination of upsides that no other card can offer RUG Delver.

Pithing Needle names Sensei’s Divining Top and Sneak Attack.

Scavenging Ooze comes in against other tempo decks. You are the control deck. Pyroblast their Delvers aggressively—you have a better long game. Life from the Loam is a red herring. Tarmogoyf superiority is the truly relevant factor, and Ooze is another Tarmogoyf.

Grafdigger’s Cage stops Green Sun’s Zenith and Natural Order as well as Reanimate.

In a Grand Prix situation, I want to split up my utility cards—one Ancient Grudge and one Krosan Grip, not two of one and none of the other.

Don’t make me explain Sulfur Elemental to you. Same goes for Submerge.

So what are RUG Delver’s upsides? It is the fastest of the Delver decks, it interacts on the stack the best, and it has the best creatures. Tarmogoyf is the best blocker in the game, and Nimble Mongoose is very hard to kill.

Its weakness centers on the deck’s vulnerability to Rest in Peace. You can expect any white deck to board in Rest in Peace against you, as it neuters eight of your twelve creatures. Vendilion Clique can help you out here, but this is another reason why Krosan Grip is in the sideboard.

If you’re looking to play a tempo deck on Saturday, I recommend playing the above decklist.

If you don’t like RUG Delver’s relatively flat power curve, situationally powerful Stifles, and weakness to Rest in Peace, you may prefer Grixis Delver:

I understand that I am somewhat biased in my love of Grixis Delver/Pyromancer, and I want to acknowledge that bias.

With that out of the way, this deck is built to crush aggressive decks in game 1s and crush combo decks after sideboarding. It doesn’t mess around trying to Stifle fetch lands, and it doesn’t need to cross its fingers and pray every time a Mother of Runes shows up on the other side of the table.

Grixis Delver is a tempo deck that can act like a midrange decks against smaller creature-oriented strategies. It plays a very capable control role against everything from Death and Taxes to RUG Delver to tribal strategies. Against control decks it has eight must-kill two-drops that run away with a game if left alone. What makes this deck so different from RUG Delver?

It plays a lot more removal for starters. RUG Delver stops trying to kill creatures after its four Lightning Bolts and singleton Fire // Ice. Grixis doesn’t stop at the Fire // Ice though—it plays two Grim Lavamancers, two Dismembers, and a Forked Bolt. It wants to fight creature-heavy strategies, and it has staying power in doing so. But why exactly does this deck want to cast so many removal spells?

Because of Young Pyromancer. Every spell that kills a creature also makes a 1/1 Elemental, and they tend to create a fast clock after the third one hits play. How else does the deck maximize Young Pyromancer?

Four Gitaxian Probes allow the deck to get value out of a Young Pyromancer that may have been played straight into the maw of a pinpoint removal spell. If an opponent doesn’t have a removal spell, they’re in even worse shape; Gitaxian Probe might draw into a Daze or another Probe, which means more Elementals, which means less time to find an answer. But not just any answer will do, mind you, since a Lightning Bolt or Swords to Plowshares that kills Young Pyromancer will still leave behind all of those annoying 1/1s.

So why if the deck wants to maximize Young Pyromancer would it not maindeck Cabal Therapy? It has obvious synergy with Pyromancer and Probe, it interacts with combo decks, and it’s great against Stoneforge Mystic.

The answer is that this deck creates tempo advantage by triggering Young Pyromancer. Contrast the following situations:

1. You have a Young Pyromancer. They play a creature, and you kill it, making a 1/1 Elemental.
2. You have a Young Pyromancer. You cast a Cabal Therapy, miss, flash it back off of your newly created Elemental, hit a creature, and pass. They play their other creature.

In both situations you end up with a one-mana spell that answered one of their creatures and made a 1/1 Elemental. In the first situation they also spent a turn investing in that creature. In the second situation they didn’t. Now you don’t have a good attack—your play was tempo neutral, not tempo positive.

Every time you kill a blocker with a removal spell and attack with Pyromancer and friends, you’re snowballing your tempo advantage. Every time you cast a discard spell and make them discard a redundant copy of a card, you’re letting them get back in the game.

“But Drew, what if your discard spell hits their only creature?

Then I guess they would’ve been even worse off if they had cast it and you had killed it.

The reason why you can afford to be so poorly positioned against combo decks in game 1s is that your sideboard is incredible against them. The two things go hand in hand; you can’t sideboard a bunch of awesome anti-combo cards and have a great combo matchup in game 1s or else you’re giving up important edges elsewhere. By building your deck to beat up on creature strategies in game 1s, you get to sideboard in cards that are good against both control and combo decks and end up cutting removal that’s bad against those control and combo decks.

To wit, here’s your sideboarded deck against Sneak and Show:

Zero inefficient cards left in the deck and you have a million angles of attack. You get to look at their hand a ton, run wild on them with a bunch of 1/1 Elementals, attack their hand, attack their mana, attack their Show and Tells, and even attack their Sneak Attacks.

This is incidentally one of only two playable Dark Confidant decks in Legacy. This one is notable for using Dark Confidant as bait for their removal spells so that Young Pyromancer can survive.

Let’s talk about what happens if they’re not a combo deck.

Against aggressive decks you board in even more removal—Liliana of the Veil against green decks with huge creatures, Fire Covenant against Elves and Death and Taxes, and Umezawa’s Jitte against creature-heavy and removal-light decks. Badlands comes in with any three-mana card if they have Wastelands. Your sideboarded three-drops against Wasteland decks are red and black, so you want Badlands over Volcanic Island or Underground Sea.

Don’t make me explain Sulfur Elemental.

Moving on, we come to U/W/R Delver—a deck that has received praise in recent weeks and whose success has eclipsed that of RUG Delver.

I’ll be completely honest: I think the deck isn’t good. If you want to play a Stoneforge Mystic deck, that’s fine, but play Jace, the Mind Sculptor with it. If you want to play Geist of Saint Traft, consider playing a more powerful three-drop, like Show and Tell.

Or at least a three-drop that does something besides attack. Vendilion Clique and Knight of the Reliquary are options. Terravore is a lot worse in a Rest in Peace / Deathrite Shaman world, but it has done work before.

Geist of Saint Traft to be perfectly honest is not a great card. It works well in this deck if their plan is to block it with their Boltable, Plowable creatures. If their plan is “kill you while you’re tapped out,” you’re out of luck. If their plan is “use the window while you’re tapped out to resolve a backbreaking noncreature permanent,” you’re also out of luck.

At the end of the day, if you want to play so fair that you think your best way to win is equipping Umezawa’s Jitte to your untargetable three-drop, I would start with this:

There are a lot of things to dislike about Ty Thomason’s list, but “playing Delver of Secrets and a three-drop in the same starting 60″ isn’t one of them.

The reason why I’m getting this feisty about Geist of Saint Traft is that it highlights a real problem with U/W/R Delver: all of the threats are weak. RUG and Grixis both have twelve threats, any of which can win the game on their own from an early turn.

U/W/R Delver plays four copies of a card that requires four mana and doesn’t attack until turn 4, when it attacks for five if they didn’t have Stifle or any removal spell that kills a 1/2 white creature. It also plays two copies of a three-mana 2/2 hexproof creature, which if it resolves is going to close a game out in three turns if you only use Lightning Bolts to clear the way for it.

If you use Swords to Plowshares, you’re probably giving them another draw step.

I’m writing about this deck because it has done well recently. I would not recommend it for the Grand Prix, and I dislike almost everything about it.

It plays too few threats (ten versus twelve), and the ones it plays are incredibly mana-hungry and mostly fragile.

Outside of Delver of Secrets it has a fair plan that would be better backed up by Jace, the Mind Sculptor. Delver of Secrets is a much worse draw on turn 7.

Good luck hard casting Batterskull. Also, Spell Pierce it.

I can’t imagine playing Stifle in a deck with Stoneforge Mystic and Geist of Saint Traft, as you have way too many points where you have tap out to get your threats in play. This means that you’re rarely mana screwing them like RUG yet you’re also not flooding the board like Grixis.

Attack you with Delver five times, Swords to Plowshares your Tarmogoyf is a miserable sequence. I get that your deck is better at killing Tarmogoyf than other Delver decks, but having Plow in the same deck as Bolt has a tangible negative impact on Bolt’s ability to close games. It has the same stench as the old Modern U/W decks that played Mana Leak, Tectonic Edge, and Path to Exile. What are you doing here? Just play cards that work well together.

If you’re dead set on playing U/W/R Delver, I would play the above list. It’s very close to Owen Turtenwald list from Indianapolis. He wrote a much more enthusiastic article about the archetype, which again I can’t recommend despite knowing several (usually) intelligent people who plan on playing the deck this Saturday.

The deck neither keeps the game in a state of low resources like RUG Delver nor grinds people out like Grixis. It has a terrible game plan against Sneak Attack (Meddling Mage versus Pyroclasm got called for Pyroclasm like ten years ago). I can’t articulate its game plan in a sentence, and if I tried it would sound a lot like old-school U/W Stoneblade’s modus operandi.

Old-school U/W Stoneblade by the way played Jace over Moon Heron. Just a thought for those of you who want to sleeve up Geist of Saint Traft in a world where people are showing up with Show and Tell.

If you have any questions, comments, or inarticulate screams about how good U/W/R Delver is, I’d be happy to address them in the comments or on Twitter.

Join me tomorrow, when I’ll talk about two very stock midrange decks and one hot fresh import from Europe. See you in 24 hours!

Until then,

Drew Levin

@drewlevin on Twitter