Prose and Khans, Part 1

Posted in Making Magic on September 15, 2014

By Mark Rosewater

Working in R&D since '95, Mark became Magic head designer in '03. His hobbies: spending time with family, writing about Magic in all mediums, and creating short bios.

The Khans of Tarkir preview weeks are over, which means it's time to start telling some card-by-card design stories. This set in particular also has a bunch of cycle-by-cycle design stories and I promise to get to those as well. Enough set-up, let's get to the part where I start telling design stories.

Abzan Ascendancy/Jeskai Ascendancy/Sultai Ascendancy/Mardu Ascendancy/Temur Ascendancy

This was one of those cycles that went through a lot of changes in design and development because we weren't quite sure what we wanted. So let's begin with what this cycle's role was supposed to be. Once we knew we were making a set with a wedge theme, the design team had a meeting where we listed everything players would expect. Down below, I'll get to some of the more obvious cycles. This one, though, was a little vaguer.

Here's what we knew we wanted. We wanted a rare wedge enchantment cycle that played thematically into the identity of each clan. We also liked the idea of a cycle that played up the idea of a warlord world where the khans were constantly fighting over territory. This led us to make all sorts of crazy versions of this cycle. In one version, the cards represented battlefields that players fought over and the effect went back and forth between players. Another version had the enchantment build up, sort of like a Planeswalker, gaining abilities over time. A different cycle sought to break major rules about the game, some we had never broken before. Another cycle attached to lands and made players literally fight over land.

Development played around with several of these crazy ideas but made a pretty startling discovery—they decided to look up what wedge enchantments we had done in the past, and here's what they found:

Abzan and Mardu each had one enchantment from Apocalypse and that was it. Jeskai, Sultai, and Temur had zero. Perhaps the role of these cards wasn't to do anything crazy, but just to make nice, clean wedge enchantments that played up the flavor of the clans. Development made sure that they were pushed enough to be relevant in various formats. Sometimes, when you are exploring new space, you don't need to be fancy, you just need to cover the essentials. I'm very happy with how this cycle turned out.

Abzan Banner/Jeskai Banner/Sultai Banner/ Mardu Banner/Temur Banner

Many of the mana-producing cycles I'm talking about today were pretty much slam dunks. This cycle, to be blunt, was a pain in the neck. We knew we wanted two common mana-fixing cycles. The other common cycle were lands capable of producing two colors. That meant two things. First, this cycle probably wanted to produce three colors and, second, it most likely wanted to be an artifact cycle rather than a land cycle.

Here was the problem: the tri-lands were sitting at uncommon. If this cycle simply came into play untapped and tapped for three colors of mana, we were afraid it would look weak versus the tri-lands, which don't cost anything to play. Also, playtesting showed that the cards were sitting in an awkward place. At two mana, they were too powerful and needed a drawback. The tri-lands were already in that space. At three mana, they were a bit on the weak side and needed a bonus. When a bonus of the right level was found, players were taking other clans' banners. Tapping for two of your colors was often good enough.

Erik Lauer, head developer and the lead developer of Khans of Tarkir, then suggested that the bonus have an activation cost that required all three colors. That also allowed us to up the bonus, which became having the ability to "cycle" the cards from the battlefield, turning into new cards in the late game. Playtesting showed that this did all the things we needed for the cards to do and, thus, that's how they stayed.

Abzan Charm/Jeskai Charm/Sultai Charm/Mardu Charm/Temur Charm

The day we added the fifth clan and it became apparent that we were going to be making a wedge set, the very first thing that popped into my head was, "We're going to have to make wedge charms." You see, the original charms appeared in Mirage. They were single-color instants for one mana that had three small spell effects you could choose from. We then did a second cycle of monocolor charms in Visions.

In Planeshift, we made three-color charms to match the five three-color dragons from Invasion. Then, when we went to Alara, we made another cycle of three-color charms, this time tied to the shards. On returning to Ravnica, we made two-color charms tied the guilds. These charms had three effects, one in each color and one that was an effect that overlapped the two colors.

As we had done three-color shard charms before, the pattern was already set for us. The charms would all be instants that cost exactly three mana, one of each of the colors in the wedge. The strongest ability would be tied to the color that was the center of the wedge. Design did a first pass on the effects with development later doing a second pass.

The one new thing actually came from templating. To make the cards easier to parse, we added bullet points so each spell effect got its own line. Note that this is not just a change for charms but for all modal spells that have two or more distinct effects.

Abzan Guide/Efreet Weaponmaster/Abomination of Gudul/Ponyback Brigade/Snowhorn Rider

As I explained last week, morph has a role in the set that is larger than any one clan. As such, we made sure to use morph as a way to show off the differences between the clans. One way to do this was through a cycle. If the set was going to have both morph and wedge clans, we needed to find a nice, simple way to tie these two themes together at common. The answer was pretty straightforward: common wedge morph creatures.

The idea was simple. Each clan would get a three-color creature that you could play face-down as a 2/2. The first version of this cycle had a morph cost using only the centered color of the clan. The idea was that this card would work better in a clan deck because you would have the ability to hard-cast it but could be played in any deck using that central color. Playtesting with other people from the company showed, though, that players were hesitant to put a Mardu morph creature in their Jeskai deck, even though having access to red mana meant they had the ability to get the card face-up on the battlefield.

We spent a little time seeing if the design could help encourage this behavior but in the end decided that we were fighting a losing battle, so we changed the morph creatures such that they required all three of the clan colors to turn face-up.

Alpine Grizzly

I don't think most people think of vanilla creature design as being something designers spend much time on. Find a void, fill it with a vanilla creature on the curve, and be done with it, right? Vanilla creatures are like any part of the design in that the best ones are an extension of the larger design. Alpine Grizzly is a great example. Why is this specific vanilla creature in the set?

If you said to help with ferocious, you get a cookie. Yes, the 4/2 for 2G was put into the set specifically as a way to help enable ferocious. What's the cheapest way to get a 4-powered creature onto the battlefield? Have it be a vanilla creature with a high power and low toughness to allow you to get the card to be as cheap as possible. I should note that the card was originally designed as a 1G 3/1 because, at the time, ferocious cared about having 3 power and not 4.

Anafenza, the Foremost

The challenge with this design is that we wanted Anafenza to both reinforce the Abzan strategy while also staying true to her flavor as a khan who leads her forces into battle. How can a creature be both defensive and offensive at the same time? I was very happy with the solution to this problem. Anafenza builds others up but only through attacking. This ability was also done in such a way that she can both pump other attackers who are attacking with her (well ones without vigilance) or an any outlast creature that is staying behind to beef itself up for a future fight.

Anafenza's second ability was added in development as a way to address some metagame concerns they had.

Bear's Companion

This card was also made with ferocious in mind. The idea was to have a Temur multicolor card that created not one creature that helped ferocious, but two. Now, remember that back in design, ferocious cared about 3 power and not 4. So this card, at the same cost, was a 3/3 that came into play with a 3/3 token. The idea was even if your opponent got rid of one 3/3, you'd still have a backup so your ferocious spells would keep working.

When ferocious got changed from 3 to 4 in development, I assume the development team tried to figure out a way to keep this card relevant for ferocious. You weren't paying enough for two 4/4s so the decision was made to just have one of the creatures be a 4/4. The token was left a 4/4, I'm pretty sure, because it worked better flavor-wise and there were more shenanigans you could do (like unsummon the Bear's Companion).

Bellowing Saddlebrute

Both raid and ferocious are what we call threshold mechanics, meaning that if you have either done a particular action that turn or have reached a particular state, you get an upgraded version of your spell or permanent. The most obvious design space with threshold cards is to have them do effect X normally but at threshold do effect Y instead, where Y is greater than X. A card might, for example, deal 2 damage normally, but deal 4 damage if you've reached the threshold.

Bellowing Saddlebrute is another area of design we've started messing around with threshold cards. These cards come with a drawback unless you're at threshold. Instead of gaining a positive, you lose a negative. Now, we don't want too many of these types of cards, as drawbacks work best in small amounts, but it is an interesting space to explore as we do more threshold mechanics.

Briber's Purse

In design, this card was called Bag of Gold. It was added as a top-down card to show the ruthlessness of the warlord world. Interestingly it was in the set before the ruthlessness clan was. The card always had a mana cost of X but the original version didn't require the 1 mana to activate. It proved to be too strong during playtest, so we had to add the activation. This is one of the cards that I believe went unchanged throughout development.

Chief of the Edge

One of the things we've started doing over the last few years is making sure that every two-color combination has an identity in Draft. As Khans of Tarkir is a wedge block, we had to lean a little less on two-color pairings, as three-color wedges are a thing. Erik made sure, though, that each color pair had an identity—especially the enemy-color pairs.

As I explained last week, the enemy-color pairs are important in Draft because starting with one leaves a player open to two different wedges. Ally-color pairs only go with a single wedge. As such, Erik worked extra hard on the enemy-color-pair identities. He was having problems with the white-black combination, so he came to me.

Because it was a warlord world and I like to have some tribal in every set, we had included a small "Warrior-matters" theme. Would I mind if the set ratcheted up the Warrior-matters theme? Erik needed something for white and black to do so he was interested in what I thought of him turning up the volume on it. I said it fit the set well, so I was fine with him adding more.

Chief of the Edge is an example of a card created in development to help push this theme in the white-black Draft archetype. So, if you find yourself starting to take white and black cards, it might be worth your while to count how many Warriors you have.

Dragon Throne of Tarkir

We had an interesting challenge creating Tarkir. It's a world without dragons but was once teeming with dragons. How could we both allow the players to feel the significance that the dragons once played without having any actual dragons in the set? The answer was through a bunch of different cards that showed the effects of the dragons were still being felt, this card being one of them.

The Dragon Throne of Tarkir is a legendary artifact. Why? Because it is literally carved out of the skull of a dragon. And not just any dragon, but once one of the most vicious dragons in all of Tarkir. Now its skull serves as Zurgo Helmsmasher's throne.


This card first came about because Gavin Verhey was putting together From the Vault: Annihilation. He came to me because the plan at the time was to include a preview card from Khans of Tarkir. He was having trouble finding green cards in theme, so he asked me if the preview card could possibly have some green in it. As we hadn't designed it yet, I said sure.

So in one design meeting, I gave as an assignment the following: make me a mass creature-destruction spell of which at least one of the colors was green. What I got back was Duneblast. Later, it was decided that From the Vault wasn't going to have a preview card (getting a card done early logistically causes issues) but I really liked Duneblast, so I kept it in the file even though the reason it was designed went away. It was still a fine addition for Abzan.

Empty the Pits

As someone who has worked on Magic design for a long time, I'm always fascinated by cards that have a unique mana cost—that is, a mana cost that doesn't appear on any other of the other 14,000 Magic cards in existence. Empty the Pits is once such a card. In fact, Empty the Pits is the first black card to ever have XX in its mana cost. Red's done it the most (on four cards—Bonfire of the Damned, Builder's Bane, Conflagrate, Meteor Shower). White and blue are tied for seconds with two apiece (Decree of Justice and Entreat the Angels for white and Part Water and Recall for blue). Green has just one (with Gelatinous Genesis). Artifacts have two artifacts with XX as their mana cost (Chalice of the Void and Orochi Hatchery) and one with XXX (Astral Cornucopia).

In general, we don't do XX very much because it tends to confuse a lot of players. In fact, for many years the number one rules question our game support would get was about how X spells worked. But as long as we keep them out of common and uncommon, the occasional XX can be seen as a special treat. I'm happy to see black has finally joined the party.

Flooded Strand/Polluted Delta/Bloodstained Mire/Wooded Foothills/Windswept Heath

Of all the questions I was asked about Khans of Tarkir before it came out, no question was asked more than, "Is it going to have fetch lands?" (So for those still curious, yes.) This leads to the next question, "Now that we know the set has them, how early in design was it decided that they would be in the set?" And the answer is, "Before design even began."

Let's start by making something very clear. Dual lands are more of a developmental responsibility than a design one. Erik Lauer figures out ahead of time what mana fixing has to be in what set so Standard has what it needs. He also works with design to make sure that Limited has what it needs. This means that design always asks development early in design what development wants for the mana fixing.

R&D had made a commitment to support the Modern format and that meant reprinting things that were necessary. Now, technically, this cycle of fetch lands is from Onslaught and thus isn't in Modern, but it helps relieve the same pressure that the other fetch lands do.

Highland Game

One of the cool things about Magic is how different players get to fall in love with different cards. A card that you might not think about twice might have brought hours and hours of fun to another player. For one such player, that card was Tarpan.

Tarpan goes all the way back to Ice Age. It's a 1/1 for G that grants you 1 life when it dies. One day, I was approached by someone who works at Wizards (for the life of me I can't remember who it was). That employee told me he (or she) and a brother loved Tarpan as kids and they always talked about us making a bigger Tarpan one day. It would be a 2/2 for 1G that when it died would grant the controller 2 life. Would it be possible to ever put that into a set?

I tucked the idea away like I do all requests. Then one day, during Khans of Tarkir design, I needed a two-drop green card. It basically wanted to be Grizzly Bears with a little upside. That's when I remember Greater Tarpan, which I put into the file. The card got turned over to development as a 2/2 but somewhere along the way got weakened to a 2/1, probably for Limited balance issues. So close, so very close.

"Khan I Buy a Vowel?"

That's all the time I have for today. As always, I'm eager to hear any feedback about today's article. Feel free to send me an email me or contact me through any of my social media (Twitter, Tumblr, Google+, Instagram).

Join me next week for the second and final part of my card-by-card design stories of Khans of Tarkir.

Until then, may you make some of your own Khans of Tarkir stories.

Drive to Work #156—Vanguard

For my first podcast today, I talk about a supplemental Magic product from way back in the day.

Drive to Work #157—San Diego Comic-Con

For my second podcast, I talk about the home of one of the annual Magic panels. I share my history with the convention and then talk about Magic's.

Latest Making Magic Articles


January 17, 2022

The Big Picture by, Mark Rosewater

Welcome, everyone. Regular column/blog readers, or podcast listeners, have often heard me say "Magic is not one game, but many games." Today, I plan to dive a little deeper into what that...

Learn More


January 10, 2022

Even More Words with R&D by, Mark Rosewater

In 2005, I wrote an article called "A Few Words with R&D" where I talked about many of the slang and vocabulary words R&D uses. In 2016, I wrote a follow-up called "A Few More Wor...

Learn More



Making Magic Archive

Consult the archives for more articles!

See All