C'mon Innistrad, Part 2

Posted in Making Magic on September 12, 2011

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.

Welcome to Week 3 of Innistrad previews. Today I will continue the story of Innistrad design and show off a preview card that is as top down as they come. If you haven't read the previous two articles (Every Two Sides Has a Story and Part 1 of this article ), I urge you read them first, as I'm writing this article assuming you have.

    It Will Be All Whiteboard

So I started the very first design meeting of Innistrad (with my design team: Tom LaPille, Jenna Helland, Graeme Hopkins, Richard Garfield and myself) by having the team write up everything on the board that they felt was part of the horror genre. (By the way—readers really want to see this list, but unfortunately it wasn't copied, so I don't have the exact list of what we wrote down.) Last week I talked about a lot of the races on that list. Today's about everything else.


At first blush this seemed a bit broad, but when you take a step back and look at the horror genre, you realize how fundamental it is. In most stories, death is the end. When a character is killed off, that's the last you see of them (okay, the other "genre fiction" categories—science fiction, fantasy, and superheroes—also tend not to end at death). In horror, death is often the beginning. For example, humans become three of our four big monsters—vampires, zombies, and ghosts—by dying. Horror even uses the term undead to designate that there is a state after death.

Art by Randy Gallegos

On top of that, a big part of the job of monsters is to kill. They bring death. As I mentioned last week, I believe the protagonists of what I'll call the meta-story are the humans. They are trapped in a world where they are surrounded by death, watching as they slowly get turned into monsters. I felt it was important if we wanted to capture the horror genre to give death as a concept weight in the design. This led me to want to add in "death matters."

We played around with different versions but in the end, having a word that meant "bonus if something has died" seemed the simplest and most obvious path to take. In design, we called the ability word deathwatch. Really quickly for those who are unfamiliar with what an ability word is, let me explain: A keyword is used when the ability is the same or mostly the same from card to card (e.g., exalted), or when we need the rules to be able to refer to the ability by name in order for it to function (e.g., reach). An ability word is a word in italics before text that could just exist on a card by itself. We use ability words because there are things that feel like mechanics and want to have a name for flavor and referencing, but aren't consistent enough in function to make a keyword.

Another big part of deathwatch was the creation of the term dies. While the term took effect in Magic 2012, it first existed because of deathwatch. My earliest template for deathwatch was "if something has died this turn." At first, I was using it as a shorthand because it was much easier to write and was much more evocative.

Eventually, I said to myself, "Isn't this just better than the status quo?" I wasn't sure, so I brought the issue up to the rest of Ramp;D. My argument was that it was already the slang everyone used and that making the change would make the cards less wordier and more flavorful—two things Ramp;D has been working harder to do. There was much debate as there is with any issue, but in the end, Aaron made the call to adopt "dies" as new terminology.

The big design and development challenge of morbid is that it can't just be put onto any effect. For example, it worked poorly on combat mechanics, because you needed the effect before creatures would normally be dying so it made the cards "feel bads" (an Ramp;D term for cards that by their very design lead to moments that make the play frustrating and unfun). The same turned out to be true for reactive cards, as it was hard to both have things die and have the situation you need to make the reactive card work.

Another change that happened in development is the condensing of colors. Design had morbid in all the colors but slightly heavier in black and green as those seemed like the colors that were more tuned into death. Development ended up pushing it more in green as green was looking for more mechanical identity. The one other thing design did was to try to change up how different colors used morbid. We liked, for example, that green had the morbid creatures that got bigger.

The one other "death matters" mechanic we used was one we borrowed from the world of Grixis (one of the shattered worlds of Alara). Internally we call the mechanic "carnage" (although it has no actual name on the cards) and it triggers whenever anything dies. The reason we used carnage in additional to morbid is that it went on a different set of effects, allowing us to make "death matter" even more.


Well if horror is about death, then it only makes sense that a place associated with death would feature prominently in a horror-based design. Innistrad is filled with plenty of art that takes place in a graveyard, but that's not really where this word led us. While I didn't know going into design what exactly we were going to do, I was aware that the graveyard (in Magic terminology now) would be part of the design. As I explained two weeks ago, the entire idea for Innistrad came from Brady Dommermuth and me complaining about how Odyssey's creative didn't make sense with its graveyard theme.

Curse of Death's Hold | Art by Clint Cearley

The big question though was how was the graveyard supposed to matter. For starters, I knew I wanted to have at least one mechanic that worked either in the graveyard or cared greatly about the graveyard. The first place I went was flashback, because I know from our market research that flashback is one of the most popular mechanics we had ever done. It was flavorful, it was popular, and it would work well with our other themes... and I didn't want to use it.

What? My reluctance was this. I felt like Innistrad was going to be our second ever "graveyard block" (third time having a graveyard theme, as Weatherlight, while only a small set, had a strong graveyard theme). I was very eager, as I am whenever we revisit a theme, to do things differently than we had done them last time. Odyssey used flashback. I felt that if I wanted Innistrad to feel different, I would have to not use any of the mechanics we had used the last time. Besides, I already had a lead on a very cool graveyard mechanic: delve.

For those who don't know the three cards above, they come from the set Future Sight. All three are future-shifted cards from the future. (You have to image me going "Ooooooooh.") Clearly these cards showed off a mechanic that will one day be seen in larger numbers. When I created the mechanic, I actually had Innistrad in mind. I knew we were going to go to horror world and want graveyard mechanics, so I made one.

The very first playtest we had included delve cards, and I learned something important. The mechanic delve doesn't really work in a graveyard set. I understand this is counter intuitive and I didn't see it until I tried putting it into a graveyard set. So why doesn't delve work? Because graveyard sets like to care about what's in the graveyard. For example, last week I talked about how part of making Zombies work was making cards that kept returning them to the battlefield. This was key to creating the building horde of Zombies that was going to define their game play.

Delve eats up the graveyard. A set with delve ends up mostly not having a graveyard because it's the resource that delve exploits. In other words, delve fought against everything else the set wanted to do with the graveyard, and those things were the more flavorful things. That meant delve had to go. This doesn't mean we'll never use delve but it's going to be in a very different style of block interestingly than what I first created it for. This goes to show the value of playtesting in design.

I had my team submit graveyard mechanics and we played with a variety of different things none of which were exciting me. Meanwhile, I was also searching for a mechanic to bring back. Ever since Shards of Alara block, one of the guidelines for design has been that every block brings back at least one mechanic. (This is being done primarily to preserve design space, simplify learning the game and because players like things coming back.) Shards of Alara had cycling. Zendikar had kicker. Scars of Mirrodin has imprint. What would Innistrad use?

Also meanwhile, we were busy working our way through all the transformation mechanics. I think at the point we were down to day/night and double-faced cards. I was happy with our progress in this area, and our top-down design was doing great work. It was then that I made the following realization: having the horror theme overlaid on the graveyard set was going to make all the difference in the world. Innistrad was not going to feel like Odyssey. Once I accepted that, I was willing to do what I had always wanted to do: bring back flashback.

Another thing we did was to give some definition to the colors as to how they cared about the graveyard. Milling (putting cards into the graveyard directly from the library) has always worked well with graveyard sets, so we had an obvious place for blue. Add to this our desire to have a number of the Frankenstein's monster–style blue Zombies require one or more creatures out of a graveyard to cast and the milling fit right in.

Black seemed like the color that should be best at getting things back from the graveyard. It fit right into Zombies, as I talked about above, but it also made sense with other black tropes. Green ended up being the color that cared about what was in the graveyard. Green has always had a reverence for the past, so we liked the idea of green being the color that cared about what was in the graveyard. For flavor and mechanical reasons, we decided to mostly make green care about creatures in the graveyard.

While delve didn't work out, we did like the idea of having a little graveyard-as-resource theme, so we decided to let red (and blue zombies) have that tiny bit of the pie. Red didn't have a lot of synergy with the graveyard colorpiewise (I hereby trademark this term), so it seemed like a good sliver to give red. White, as the "good" color battling the four other "evil" colors (note I'm talking about this block only and not the colors in general), didn't feel like it wanted to have a lot of graveyard synergy. It has a little bit of reanimation where it fit flavorfully but it's the color that cares the least about the graveyard in the set.

The last bit of graveyard design came from Richard Garfield. I know a lot of people are excited he was on the team and want to hear all about what he did, but as a valuable team member, his role was working with everyone else. I'll have a few stories when I get to the card-by-card articles (the next two weeks), but there's not a lot to pull out that was specifically what Richard did. That said, here's one thing.

Richard noticed that one of the problems with graveyard recursion (getting cards out of the graveyard and back into your hand) is that it creates a problem Ramp;D refers to as "repetition of play." Because you can choose what you want, the same spells tend to get cast over and over, creating a more repetitive environment. Richard noticed that this doesn't happen with card drawing because the draw is always random.

Richard fixed this problem on a few cards he designed by making the graveyard recursion random. You would get a card back from your graveyard, but you wouldn't know which one. If you're unaware I'm a big fan of randomness (read this article if you care to know why), so when I saw what Richard had done I loved it—so much so that I made a design dictum that all (okay, most) of the spells and effects that put cards from your graveyard into your hand did so randomly. This helped cut down on repetition, and with the graveyard-as-resource theme in blue and red, players could often control what card they got back.

As you'll see when you finally get your hands on the set, the graveyard component is strewn carefully throughout the set. What you care about and how will be based mostly on what colors you're playing.


One of the fun parts of doing top-down designs is that you get to start with something known and ask "How would this be represented on a Magic card?" Curses seemed like a perfect fit for the horror genre. The big question was how to make them. The answer came from a place that many answers have come from: Un- sets.

Curse of the Bloody Tome | Art by Jaime Jones

A quick aside: I love the Un- sets more than most, as both sets were my babies. One of the greatest value of them, I've argued for many years, is that they are a great source of experimental design. Often we try something that seems crazy but taps into something not so crazy that actually can be used by black-bordered Magic. Case in point: enchant player.

Unglued has two cards that literally put cards onto one of the players. (Charm School balances a card on your head, while Volrath's Motion Sensor forces an opponent to balance a card on the back of his or her hand.) When trying to figure out what kind of enchantment it was, I took the silly answer: "enchant player." After all, I argued, you were actually putting something on a player.

In Dissension, we created a card (Psychic Possession) that really felt like you were putting an enchantment on an opponent, so we decided to embrace "enchant opponent" in black-bordered Magic. "Enchant player" made its black-bordered debut in Time Spiral and came back one more time in Shadowmoor.

As the design team talked about curses, I said that they sure felt like an "enchant player" card. The team agreed, and we began making curses. The rules were simple. They were all "enchant player" Auras (and the only "enchant player" Auras in the block) and they all hurt the player that was enchanted. We specifically left them as "enchant player" rather than "enchant opponent" because while normally they are nothing you'd want to enchant yourself with, we know the Johnnies out there will find reasons. (You'll enchant yourself with Curse of the Bloody Tome in certain draft decks, for example.)

You'll notice that all the Curses have Curse as a subtype. The reason we did this is that we wanted to make a few cards that referenced the Curses, and to do that we needed the subtype. Those cards aren't public yet, but you'll know them when you see them. I'm very happy with how the curses turned out, and I think they have all the flavor we hoped for when we first wrote "Curses" on the white board.

    Odds amp; Ends

I've talked about a lot of the big things on the board but it was also filled with lots of little things, such as Wooden Stake.

At the PAX Innistrad party, I was talking to a player and asked which card was his favorite. He said Wooden Stake. He explained to me that he was nervous when he heard we were doing a top-down horror set, but when he saw Wooden Stake he got what we were doing. We were figuring out all the things one would expect to find in the horror genre and making cards that mechanically matched.

Wooden Stake | Art by David Palumbo

I was happy he picked Wooden Stake, because that was one of a long list of cards from the white board. In fact, my preview card for today comes from a very similar place. I asked Jenna, the design team member who is also part of the creative team, to come up with a long list of horror card names. Then in numerous meetings we rolled a die and picked a name off the list. The team would then design a card together that matched the flavor.

Today's preview card came from one such design session. We rolled the die and the name off the list was "Cellar Door." Here's what we came up with.

Click here.

Here's why I love top-down design. There's no way this card would ever be designed by a normal process. It exists only because we wanted a card called Cellar Door.

During naming, I kept checking in with Doug Beyer, because I really felt the name for this card was important. I remember at one point asking and Doug said, "The card was concepted as Cellar Door, so I'm not sure what else I could even call it."

Cellar Door | Art by Rob Alexander

The other small fight around this card came up during development. Some of the developers thought getting the card from the bottom of the deck was unnecessary. It would functionally be easier to just take it off the top. As the lead designer of a set (and to a lesser extent as head designer), I try hard not to jump in on too many changes, but I reserve the right to ask the lead developer in a few cases to change it back. Usually unless there is a strong reason not to, they'll comply. I felt very strongly that having the cards come from the bottom of the library really mattered for the aesthetic and flavor of the card. I'm not sure if Erik agreed, but he respected my wishes and changed it back. (Thank you, Erik.)

    Innistrad and Out

That's all the time I have for today. I hope you've enjoyed this glimpse into the making of Innistrad, and I hope you make it out to the Prerelease the weekend after next to try it for yourself. If you have half the fun playing it that we had making it, you'll have a blast.

Join me next week when I go from macro to micro and tell you some stories about how individual cards were made.

Until then, boo!

Latest Making Magic Articles


June 27, 2022

A Double Take, Part 2 by, Mark Rosewater

Last week, I started telling some card-by-card design stories about cards reprinted in Double Masters 2022, but there were a lot of stories to tell, so I'm continuing today. Body Double ...

Learn More


June 20, 2022

A Double Take, Part 1 by, Mark Rosewater

Before the release of Double Masters 2022, I thought it would be fun to look back at how some of the cards were made. This set has cards from all over Magic's history, so over the next tw...

Learn More



Making Magic Archive

Consult the archives for more articles!

See All