In Development – So…Why Don’t You Lose?

The A-10 Warthog is an amazing ground attack aircraft. Developed and introduced in the 1970s, it has provided close air support to American ground forces in the first Gulf War, the Balkan conflicts, and more recently in Iraq (again) and Afghanistan. It is, perhaps, most famous for being built around a truly profound Plan A.

That’s the GAU-8 Avenger, the very large rotary cannon around which the A-10 was, essentially, built. And yes, that’s an actual VW next to it.

However, neck-and-neck with its fame for being built around a gigantic cannon, the Warthog is also well known for this:

That’s Major Kim Campbell (at the time a Captain) checking out the tail of her A-10, which took a hit from a surface-to-air missile over Baghdad in 2003. That hit took our her hydraulic systems and her backup hydraulics. Fortunately, the A-10 had a backup to the backup, and through some skilled piloting Campbell was able to fly her plane back to base using just the tertiary manual controls – which is a lot like trying to drive your car without power steering, the ability to control acceleration, or working brakes.

Two weeks ago, we looked at the offensive side of deck power – by analogy, the GAU-8 Avenger half of the equation. As many of you rightly pointed out, some decks that have tremendous raw offensive strength fold terribly when challenged by an actual living, breathing, card-wielding opponent.

Today, we’re going to check in on the second component of deck strength, the “it still flies even without hydraulics” part of the game – resilience. I’ll take a look at what we intuitively mean by resilience, the specific aspects of resilience, and a quick rule of thumb way to test your deck’s resilience.

Understanding resilience

As with offensive power or any other aspect of deck quality, we have an intuitive understanding of what it means for a deck to be resilient. That’s generally a good place to start, and from there we can come up with a more coherent definition of resilience that will lead into figuring out how to evaluate it when we’re picking, tweaking, or designing decks.

How we intuitively understand resilience

Intuitively, resilience is our deck’s ability to “keep flying” despite the attempts by your opponent and pure bad luck to knock it out of the sky. From the outside point of view, a resilient deck just keeps coming despite your best effort to take it down in a match, or hate it out of a format. In contrast, a deck lacking resilience folds up pleasingly to simple opposition.

The classic low-resilience deck might be Legacy Belcher. Viewed from the side of offensive capability, it has a tremendously explosive Plan A – killing the opponent on turn one via [card]Goblin Charbelcher[/card]. This is backed by an only slightly less impressive Plan B – generating a swarm of Goblins on turn one. But…

“Okay, land, ritual, ritual, Goblin Charbelcher – “

“Wait. Force of Will.”

“Damn it.”


“Okay, land, ritual, ritual, Manamorphose, Mox, Empty the Warrens…make ten Goblins.”

“Okay. Play my land for the turn, and Engineered Explosives at zero.”

“Hm. Attack for ten?”

“Sure. Done?”


“Play my land, pop Engineered Explosives.”

Now, there are plans around this, especially post-board, but you get the gist.

In contrast, the classic high-resilience deck might be Faeries, circa…anywhere from Lorwyn on through contemporary Extended. Faeries lacks the explosive Plan A offensive capability of a deck like Belcher, but has that obnoxious tendency to alternate between either effortlessly blocking everything you try to do or seemingly being on the edge of death…before chaining together Cryptic Commands and Mistbind Cliques to win the game.

Faeries has even proven meta-resilient, refusing to be hated out even in the wake of a series of cards specifically tailored to that task.

So these two decks – Faeries and Charbelcher – might form two reasonable extremes for our idea of what constitutes resilience. With that range bracketed, let’s take a look at a

Defining resilience

In looking at a deck’s offensive power I conceptualized it in terms of “how hard is it for your opponent to recover when you deploy your plan?” In other words, a deck’s offensive strength is reflected in its ability to knock the opponent off of their game plan.

This guides us right into a definition for resilience.

Resilience is the ability to stay on plan.

It’s the ability to stay on plan in the face of resistance from your opponent and from the luck of the draw.

More to the point, it’s the ability to deploy a winning game plan in the face of variance and active resistance.

With that definition in our pocket, it’s time to figure out how we can evaluate and understand resilience.

A quick interlude – the value of backups

The single strongest bolster to your deck’s resilience will probably always be “having more than one plan.”

At last weekend’s PTQ, I found myself facing down R/G Valakut in round three. Although my opponent managed to accelerate into more Mountains that I liked, I stil felt pretty good when my Bloodbraid Elf cascaded into a Thoughtseize. Until I saw this:


This was not an especially exciting decision. If I took Titan or Oracle, any land off the top would give my opponent the Scapeshift kill on his turn. However, when I took Scapeshift, that let him cast Primeval Titan, which gave him two active Valakuts right away and a land drop to use on that same turn. And, of course, two more Mountains coming in when he swung with Titan the turn after that.

I basically got to choose between a no-turn clock and a one-turn clock due to the presence of multiple lines of attack all leading to a Valakut kill.

Consider a typical Valakut list from the current extended season:

Valakut (as played by Ricky Calarco)

Ranked in order of decreasing offensive strength, the deck’s plans of attack are:

The first take-home from this list is that you save a lot of getting killed if you can somehow stop Valakut itself from having an impact.

The second take-home is that attacking any one of the elements still leaves you vulnerable to multiple other strong attacks. That’s resilience, and the clear value of a deck having more than one real way to win.

Evaluating resilience

Resilience is in many ways a more complex trait than offensive power. As such, it’s helpful to break it down into a couple few different traits that we can tackle one at a time.


It’s easy to forget about variance when we think about a deck’s resilience, since we’re typically so focused on beating the person sitting across from us. But the basic skeleton on which any other resilience will be built is the degree to which your deck does not simply lose to itself.

At last weekend’s PTQ, I had one round that essentially involved my opponent mulliganing to four on the play and then dying with one land on the battlefield…followed by a near-repeat performance in game two. As we were signing the match slip, I asked the obvious question:

“How many lands are you running?”


He insisted that was enough, and given that he’d just lost an incredibly frustrating round, I didn’t press. But I also didn’t agree, and think this is a clear case of a deck that lacked resilience in the absence of outside factors.

There are few well-known builds that are shaky enough to be really good examples of issues with variance. Most known archetypes have undergone enough testing to be relatively stable. Even so, let’s take a look at a Standard deck that has some issues with variance:

G/W Quest (as played by Nico Bohny at PT Paris 2011)

Whereas some aggro decks walk fairly evenly up their mana curve without regard for individual card choices, the Quest deck cares a great deal about which cards it has in its opening hand, and then the sequencing of cards after that. I tend to agree with Patrick Chapin that, for example, there’s a significant difference between a Quest deck that plays a turn one Quest versus on that doesn’t.

On his way to losing a hard-fought five-game game match in the quarters of PT Paris, Bohny mulliganed some eight times across five games, including going down to four cards in the fourth game and five cards in the fifth. Although these were probably correct mulligans, they mean that Bohny payed a tithe of from one to three cards each game so that his deck could get out of the starting gates.

Naturally, having to “spend” all of those cards to let your deck work removes many of your options against plays your opponent makes against you – making you less resilient.

Here are some things to ask when evaluating a deck’s resilience against variance:

How frequently do I have to mulligan to consistently deploy my Plan A?
How frequently do I have to mulligan to consistently deploy my Plan B?
If certain matchups depend on access to key cards, how likely am I to see them?

Essentially, all of these questions ask, “How lucky do I have to get for my deck to run the way it’s supposed to?” The more we can cut down on internal variance within a deck, the more resilient that deck becomes.

I’ll cap off this discussion by pointing out that this is one reason we tend to stick with the sixty-card minimum limit when building Constructed decks, as that maximizes our likelihood of drawing into our game plan when we need it.


The close companion to resilience against variance is structural resilience.

A line that’s always stuck with me comes from Rich Hagon’s podcast coverage of Pro Tour Valencia 2007, aka “PT Flood.” In introducing the Extended format of that time, Rich said that in evaluating your opening hand, you always had to imagine it without its best card. At the time, the culprits were Cabal Therapy and Duress; their contemporary analogs being Inquisition of Kozilek, Duress, and Thoughtseize.

The basic question asked when we evaluate structural resistance is “what happens when the opponent punches holes in our plan?”

Valakut decks, in Extended and Standard alike, have significant structural resilience. You saw what happened when I tried to punch a hole in that plan with a Thoughtseize – it wasn’t enough. In general, this has been an issue in attempting to tackle Valakut decks in the past. If you take out their acceleration…they proceed to kill you a turn or two later before you can capitalize. Kill a Primeval Titan and the deck swarms the field with plant tokens, courtesy of Avenger of Zendikar. “Turn off” the Valakut kill with Leyline of Sanctity and they just use it to kill all your creatures instead…and then kill you with a Titan.

Caw-Go decks with Sword of Feast and Famine tackle this problem by poking an endless series of generic holes in the Valakut plan via repeated discard backed with other disruptive elements. Although Valakut is highly resilient to having holes punched in it, modern Caw-Go overloads its ability to soak up the disruption – at least for now.

A deck has poor structural resilience if it has just a few, key cards that are easily “touched” by the opponent, with few ways to protect them. We could imagine, for example, a dedicated Charbelcher deck that focused entirely on ways to find and accelerate out a Goblin Charbelcher, with no Warrens plan and no disruption. That might be an incredibly fast deck…but it would auto-lose each and every game to a Force of Will, no questions asked.

Similarly, compare the following pair of lists from recent PTQ top eights:

Mythic Conscription (David Mathis, Mobile PTQ)


Bant (Reid Duke, MTGO PTQ)

Last time, I pointed to the Bant versus Mythic comparison in terms of offensive power, where Mythic is the clear winner.

However, when we turn our eye toward resilience, the situation flips around. The Mythic deck’s Plan A can be stifled by a Doom Blade or Path to Exile. Once the deck loses that first Sovereigns (you did kill the Sovereigns, right?), it might Jace its way into another copy, but it’s otherwise onto its Plan B – Bant creatures, six of them being legitimate fatties.

In contrast, the Bant deck, at least as played by Reid Duke, starts with twelve powerful creatures and two copies of Elspeth, Knight-Errant. It trades away the explosiveness of Mythic for a continuous, even flow of legitimate threats that are all pretty much interchangeable.

What happens when your Mythic deck loses its Sovereigns of Lost Alara? You cast a Knight of the Reliquary and hope to get there. Your plan has been forced to change.

What happens when your Bant deck loses its Mirran Crusader? You play a Knight, with Bant, Qasali Pridemage, and Path to Exile backup. Your plan remains the same.

Here are a couple questions to ask when checking into your deck’s structural resilience:

Can I keep playing my Plan A if I lose one of its key cards?
If not, how likely am I to draw into a replacement card that’ll let me play out my Plan A?

How good are your trades?

The third pillar of deck resilience involves how favorably you interact with opposing decks when they try to disrupt your game plan. This comes in two flavors – trading and warping.

Think about that attempt to disrupt the Valakut deck’s primary game plan by running Leyline of Sanctity. It’s a good example of both concepts.

Say you’ve sided into four copies of Leyline of Sanctity. Assuming they want to re-active the “kill you directly with Valakut” game plan, how many cards does the opposing player have to burn? Let’s check out Tomas Bilek’s list from Paris:

Valakut (Tomas Bilek, PT Paris 2011)


The answer here is “one.” Bilek has to burn one copy of Nature’s Claim or Acidic Slime to take out your Leyline, and then he’s back in business.

Depressingly enough, he even has more access to these solutions than you do to Leylines, since he can Zenith for Slime, effectively giving him five copies of Slime for seven solution cards, total.

The second consideration, warping, has to do with how much your opponent has to mess with their primary game plan in their attempt to attack yours. Once again, consider those Leylines.

They’ve cost us four spaces in the deck that are no longer devoted to our primary game plan.

They’re definitely much more effective when we get to drop one or more before the game begins, which demands some amount of mulliganing from us dedicated strictly to getting a Leyline in our opening hand. Thus, instead of paying a card tithe to variance, we pay it to trading instead. Unsurprisingly, this raises our vulnerability to variance, since we’ve invested in hitting our sideboard cards instead of buffering against hands that are bad for our primary game plan.

Mulliganing, like anything else, is a resource.

The Standard RUG deck is another good example of this trade off in action:

RUG Control (Shahar Shenhar, PT Paris 2011)

Check out all of those comforting sideboard cards versus small-critters-and-reach aggro. Pyroclasm, Ratchet Bomb, Obstinate Baloth…the only problem comes when you try to decide which cards you’d actually want to remove from the main deck.

If opponents have to significantly deform their deck to adjust it for the matchup with you, that is resilience. Over in Legacy, if an opponent is going to pack enough hate to legitimately threaten the Dredge game plan, they typically have to sloow themselves down tremendously…which in turn gives you more time to make the kill. Thus, even in an environment that has the option of running smashingly powerful graveyard hate, Dredge remains a legitimate contender…because the opponent has had to cripple their own game plan in their effort to mess with yours.

In considering trades and warp, here are some questions to ask:

How many cards do I have to spend to beat my opponent’s cards that attack my plan?
How much do my opponents have to damage their own offensive plan if they want to legitimately attack mine?

A simple test

Last time around I offered a simple suggestion for figuring out whether your offensive game plan is strong enough. This time, I can’t present anything as pithy as that one-liner, but I can suggest a simple test of resilience. Just do this:

Shuffle up your deck, then take the top third of it and set it aside. Run some playtest games. Assume cards that search (fetches, tutors) have access to the separated twenty, but otherwise just go with what you have.

If your deck works under these conditions, congrats – you’re probably not playing a horrid train wreck. If not, then it’s time to review the discussion of resilience we just went through and consider whether your deck is somehow falling down on the job.

…and with that, we’ll close out the topic of resilience, at least for now. I hope that, in combination with the prior discussion of offensive strength, these ideas will help you pick, update, or design the best possible deck for your next big event.

So what do you value most? Offensive strength? Resilience? An even mix of both? Let me know in the comments.

magic (at) alexandershearer.com
parakkum on twitter

18 thoughts on “In Development – So…Why Don’t You Lose?”

  1. Another excellent article. Thanks Alex. I loved your anecdote about Major Reed-Cambell. That is one bad-ass lady. I am trying to make my G/W Quest deck more resilient by using Mirran Crusader instead of Devout Lightcaster, as it helps in more situations. Can’t complain about a double-striker with a sword. Also, a couple copies of Lead the Stampede can help you recover from a wrath effect. Still trying to decide what artifact removal to use against Caw-Blade; giving them 4 life doesn’t help with the tempo issue. Also, I am getting more tempted to keep four Kor Firewalkers in the board (instead of two) to deal with Pyroclasms and bolts. Argentum Armor on a Firewalker usually elicits a funny face from the red mage… It feels like these considerations fit with your description of structural and variance resistance. Any thoughts?

  2. Not to be That Guy or anything, but technically, the word “tithe” means “a tenth,” so tithing one or two cards from a seven-card hand isn’t a realistic situation.

    Man, I’m such a bag.

  3. What does your test thingy even do besides cut the top 20 cards to the bottom, given you aren’t drawing 41 cards in a game?

  4. One of the issues with resilience is how difficult it can be to build a deck that is resilient to multiple game plans. Back in the Masques/Invasion Standard, people looked at Fires of Yavimaya and went crazy about how good that deck was, even though it got slammed pretty hard by Counter-Rebels and Nether-Go. A friend of mine would constantly take the Friday night standards there by bringing a weird ponza deck, and just wrecking everyone. There was no resilience to land destruction, and Fires didn’t run anything that could reasonably survive bashing through a 2 mana 4/4 wall. They either had the Saproling Burst or they lost.

    I suppose what I’m getting at is that it’s nearly impossible to cover all the angles, so you may as well just focus on either A: the most likely angles to occur. B: the angles that disrupt your plan the most. C: the angle that disrupts your plan that you feel is likely to occur the most.

  5. Without taking any merit from the article, I’d just like to point out that conceptually speaking, the ability to deploy a winning game plan in the face of variance is more correctly called consistency.

    Resilience should be applied only to the ability to deploy a winning game plan when actively disturbed, i.e. what you defined as structural resilience and interaction with opposing decks.

  6. Excellent article, Alex! This is a very interesting subject, as most of the players simply ignore this concept when choosing/building their deck.

  7. On the whole pretty insightful, although I feel compelled to point out that the test at the bottom is questionably relevant. The Standard Valakut deck, accurately cited as a paragon of resiliency, probably doesn’t actually work without it’s top third.

  8. Amazing article! That’s often the balance one must find with choosing a deck for an event. It is careful scale of raw power to resilience.

  9. I don’t get why people keep bashing gw quest. Have you ever played WITH the deck? Vengevine + squadron hawk is about as resilient as you can get. You know what deck is depenent on opening draws? Boros. WW quest. Valakut. NOT gw quest. Try it and you would probably agree.

  10. Nice article. I’d like to suggest a different test for resiliency, though, since yours, honestly, doesn’t seem that great (how does removing a random third of your deck test how resilient it is? For every time you set aside a key card and deny access to it, twice you’ll actually increase the likelihood of drawing said key card).

    Here’s my test. It’s a standard goldfish test, except that:
    a) Turn 1, your opponent plays Thoughtseize, taking your best card, and loses no life.
    b) Every third turn your opponent plays either Vindicate* or Coercion, whichever is better under the circumstances.
    c) Your opponents spells can’t be countered, have their targets changed, etc.**

    Figure out how many turns it takes you on average to beat the modified goldfish, and compare that to how many turns it takes you on average to beat a regular goldfish. If your deck’s winning within a turn or two of what it was goldfishing at, your deck’s pretty resilient. If it’s taking you an extra ten turns to do so, not so much.

    * If your deck’s a swarm deck, you might want to make Vindicate into Volcanic Fallout or Wrath for a more realistic test of what’s going to put your deck on its back foot. Adjust cards to suit.
    ** If you’re a heavy control deck with very few actual win conditions, you might need to remove this rule. If you do, though, I suggest you make your opponent cast Vindicate/Coercion more often to make up for it.

  11. I agree with GRF on both counts:

    (1) Nice article! Complements the other piece very well, particularly the way you went back over some of the same decks.

    (2) I’m not so sure about the “take away 20 cards” test, and his test sounds pretty good.

    In some sense, the question is almost “how many win conditions do you have?” It’s slightly more complex than that, of course, but it’s a useful way to think about why some cards are so good.

    For example, Planeswalkers–not only can they impact the board immediately and enable or accelerate your Plan A (or B), but they often represent a win condition themselves. Most decks that run Jace don’t think of Jace’s ultimate as Plan A or even Plan B, but his ultimate sure is a nice thing to have in your back pocket as Plan C (or whatever).

  12. good read alex, the part about the thoughtseizing was against me in the ptq i rememberd right away when you mentioned that, so i guess thanks for the indirect shout out in the article, again good read

  13. Pingback: MTGBattlefield

  14. Pingback: » In Development – One Big Stack of Death

Comments are closed.

Scroll to Top