How to Build a Deck

A month of so back I lamented on how people don’t really design decks anymore, so I guess I have to do it myself.

Of course, I’m not a real expert on the subject, but I do have some experience on. Furthermore, I am quite experienced in the field of software development and teaching. These two disciplines are actually quite similar and use a lot of the same principles.

I’ll be using an example of current standard, which might not be very timeless, but it is a way to showcase the process (or at least a process, because there are obviously many ways to do stthis).

Starting Point: Top-Down

All competitive decks are (or should be) built for a metagame. Many cards are individually powerful, but depending on the meta and the cards in the format, not all those cards are actually good in a context.

So, first thing: Context is everything.

Think Jeskai Ascendancy. We know its an extremely powerful card. For a short period, it ruled Modern and it was pretty good in Standard as well. However, you don’t see much of it, because the decks lost some of the tools (Raise the Alarm, Sylvan Caryatid). It still has potential (and I’m not sure there isn’t a strong deck out there) and maybe we’ll see it someday.

So, we need to identify the powerful cards in the context. Basically, what do you want to be doing in this format?

Gideon, Ally of ZendikarJace, Vryn’s ProdigyRally the AncestorsReckless BushwhackerDragonlord Kolaghan

… and here’s my current favorite, which I’m going to using as the example:

Kalitas, Traitor of Ghet

Why would we want to use it? Because its pretty good against the aforementioned cards. Again, context. It has lifelink, so it helps against fast red decks. It exiles creatures that die, so it helps against Rally decks. You can produce your own army against Gideon’s, so those puny Knights actually become a liability. Same is true for the token armies of the BR Dragons decks. Kalitas doesn’t do anything against Jace itself, but you want to play plenty of removal against him, so at least Jace dies easily, although many decks playing Jace’s don’t play many other creatures.

So, yes, we want to play Kalitas.

After identifying this starting point, we can start to identify cards we want to play in a deck with Kalitas. First, we want removal. Plenty of it. The rule here being that we want things to die, not exile. Gladly, this is exactly what black does.

Kolaghan’s CommandRuinous PathMurderous CutUltimate PriceCrackling Doom
LanguishCrux of FateFiery ImpulseWild SlashDraconic Roar

That’s a nice selection right there. Crackling Doom would be nice, but the way I’m going, splashing for white isn’t very easy, so I’ll start with being black-red.

Deeper into the Meta: Bottom-Up

What do you expect to play against? Removal is always situational. This is actually one of the reasons we want to play Kalitas. The most common removal spells in our meta are Fiery Impulse, Silkwrap and Dromoka’s Command. Kalitas dodges two of those and at least you get some value if Kalitas is hit with the third. Also, if you can attrition out your opponent, Command doesn’t do anything against you.

From our end, I think we need a pretty diverse set of removal. Since there are plenty of aggressive decks in our meta, I think I’ll need Fiery Impulse. I’m also thinking a full set of Kolaghan’s Commands, which can actually get you two zombies, if you can kill a Hangarback Walker and something else. It can also get us back the Kalitas, if needed.

Planeswalkers are also popular, so perhaps Ruinous Path is a good choice. Murderous Cut has other costs, but is also the most efficient removal in the format, so a couple of those would be nice.

The problem here is that our meta is very diverse. Identifying the right removal can be hard, when its hard to know what you’ll be playing against. Big tournaments are easy. You just know a certain percentage will be playing the most popular deck and a certain percentage will play control decks no matter how bad it is. So, its important to find cards that do a lot of different things.

The End-Point: Backwards Design

How do we see the game ending? The important part is not to identify the actual final turn of the game, but rather the point where the game is under your control and the opponent doesn’t have a way to recover, although in the case of many aggressive decks, this can be the actual final turn.

Of course, plans never survive the actual conflict, but it doesn’t mean there shouldn’t be a plan. So, how do we see this ending? The deck is going to be quite attrition focused. We exchange card for card until I can extract enough value from my cards to gain enough advantage to finish the game.

The finisher? Chandra, Flamecaller seems good. It can play into our plan as well. It simply does everything. The first ability kills fast, if the board is empty. The second ability can find you the card you need and give you card advantage, as well as fodder for Delve. The third ability can clear the board. What is there not to love?

Playing Chandra on an empty board is often game over.

Finally: Forwards Design

Lastly, we need to figure out how we get there. We have already identified removal, but we need other things as well, like card advantage.

The first few turns, we are more reactive than proactive. We just remove stuff, but I would also like disrupt the opponent in other ways. Discard isn’t very good with Kalitas, but having access to Duress and Despise can destroy your opponents game.

After the initial flurry of spells, we need to gain the advantage we need. Here I’m thinking Goblin Dark-Dwellers, because its both a good body and a flexible way to use our graveyard. I’m also thinking Read the Bones, because it works better with the Goblins than Painful Truths.

If you look at the Rock decks of the past, personlands have been a good threat. The question is, which one do we want to play. Again, I don’t want to splash white, because that complicates the manabase needlessly and I want to play a copy or two of Tasigur, the Golden Fang, which needs either blue or green to activate, so I have a choice between Wandering Fumarole and Hissing Quagmire. This largely depends on what we want to splash from those colors. I think I want more card advantage, so I have a choice between Den Protector and Jori En, Ruin Diver. I’m going with the latter for two reasons: I want to test her out, since I haven’t had a chance to play her, and I want the sideboard options blue brings. Also, she dodges certain removal Mother doesn’t, like Wild Slash and Ultimate Price. The problem is that these both do make us vulnerable to certain removal, such as Silkwrap, Fiery Impulse and Draconic Roar.

With all this in mind, we can build the manabase. Early blue isn’t very important. Maybe fourth turn or so. Early red and black is much more important, preferably emphasizing black somewhat. Surprising the opponent with a counter is also nice, which might influence the choices somewhat.

Here’s the first list.

That gives us 16 cards to recast with Dark-Dwellers, which should give us around 95% probability of having a spell in our graveyards, when we cast the first one as long as we’ve delved properly (you can see the calculations in this article from Frank Karsten).

Iterative Approach and Being Adaptive

Meta changes. Even if Kalitas remains an unknown quantity, other decks will come and go, changing what we require of our deck. Because of this, the deck needs to change with the times. However, even now, as I’ve compiled the list, we should go back and look at the list. Are there nonbos? Are there ways to make the synergies better? Do we have blind spots?

Of course, here’s where the sideboard come in. Sideboard isn’t just something you throw together. You should be taking your deck into account. There are some good generic answers, like Duress, which is good to bring in both against the both the most aggressive decks, the control decks and the ramp decks. Put cards into your sideboard that can be used to tune your strategy, but also know what to expect.

Here’s a list I came up with:

Think about the match-ups. Supposing I’m playing against ramp. What do I bring in? Depends on the flavor actually. If I’m playing against the monogreen version, which has plenty of mana producing creatures, the plan is probably this:


Disdainful StrokeDisdainful StrokeInfinite Obliteration


LanguishRuinous PathRuinous Path

… but all in all, I don’t think this will be the final sideboard. I’ll have to come up with some cohesive list. I’d rather use cards I can use against multiple decks than use silver bullets, like in the older formats, but the meta might require such cards. For now, I’ll be happy with this, although it doesn’t feel nearly as good as the mainboard.

A deck is never done. There’s always something to tinker with, as you learn more about the deck and the context. Mold your deck. Find cards you can cut. There are no sacred cards. I may find that Kalitas isn’t even good here, but the rest of the deck is. In that case, I should cut Kalitas and try something else. I don’t know what that would be, but whatever it is, its not going to find itself.

Also, read Chapin’s book, Next Level Deckbuilding. He’s much more versed in this than I am.


Think about the deck from multiple points of view. Build only after you understand the context. Don’t fall in love with your ideas and cut them if they don’t function as needed. Read Chapin.

Leave a Reply