City Planning, Part II

Posted in Making Magic on September 12, 2005

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.

Previously on “Making Magic”:

Charged with revisiting a multicolor theme, the Ravnica design team (Mark Rosewater, Tyler Bielman, Mike Elliott, Aaron Forsythe and Richard Garfield) tried hard to pick a design path that didn't feel like Invasion II. This led to the creation of a new type of multicolor card – the hybrid card – which allowed multiple colors to be used to pay the spell's colored mana.

But traditional multicolor cards didn't play nicely with the hybrid cards as the combination proved mentally taxing, so the hybrid cards were cut. Meanwhile, Brady Dommermuth of the Creative Team suggested a way to make the ten color pairs stand out by creating flavored guilds around them. The design team liked the idea but didn't know how to use it. The design team needs to solve the problem soon as the clock is ticking…

The Guilded Cage

So we had four months left in the design process. That may sound like a lot but for design it isn't. Not for Magic anyway. We were trying hard to not feel like Invasion which meant every obvious multicolor choice was forbidden. (Note that once the theme was established we did go back and use some cards that were more reminiscent of Invasion.) The only lead we had was the guild model. Mind you, I liked the guild model. It was flavorful. It took great advantage of the two-color multicolor cards that we wanted to focus on. And it was unique in its execution. The big question was how could we use it to structure the design?

We began, as many creative brainstorming sessions do, with the most radical suggestion. What if every card in the block belonged to one of the ten guilds? How could that work? For starters, it would mean that everything that normally appears in a block would have to find a home in a guild. Let's run down the list:

  • Keyword Mechanics – Making a keyword mechanic guild-relevant seemed pretty straightforward. Just assign one keyword mechanic to each guild. This simple solution presented two problems. First, keyword overload. If each guild had its own mechanic then the large set would have ten new keywords. (Even with our inflation over the last few years, this is still too much.) Second is the interactivity concern. If each mechanic is locked into a particular guild, you run the risk of being unable to play the majority of the mechanics.
    The second problem proved to be the easier one to solve. The solution? Making use of mono-color cards. You see, each mono-colored card can be played in four different guild decks. This means that any guild deck will have access to seven of the ten mechanics (the number is seven, rather than eight, as one guild will overlap) if each guild mechanic shows up in its two mono-colored cards. Which brings us back around to the first problem. We'll solve it in a second.
  • Cycles – One of the staples of Magic design is the cycle. A cycle is traditionally five cards that are connected and then spread out across the five colors. How do you do cycles in Guild World? Easy. Embrace the guild model and make ten-card cycles. The problem with this solution is that ten-card cycles take up a lot of room. You can't do all that many of them. Another problem we'll address soon.
  • Mechanical Focus – I've often explained Magic as a perpetually swinging pendulum (the kind swinging over a sand pit at the museum). From year to year, the designers choose different elements of the game to focus on. Sometimes this theme is large and obvious (the tribal theme in Onslaught) and other times it's subtle and small (the “cycling from play” theme of Urza's Destiny). The tricky part about applying mechanical focus to the guild model is that it needs to find a simple way to interconnect some of the guilds. The most obvious choice is through color, but there are other options available.
  • Route to Victory – There are many ways to win in the game of Magic. Even within the traditional confines of dealing twenty damage. Like mechanical focus, this element of the game needs to be shaken up year after year. The guild model wants to allow each guild its own route to victory. While this is flavorful it presents a similar problem to the keyword mechanics. Ten is just too much.
  • Evolution – One of the defining traits of blocks is that they take ideas, often mechanics, and evolve them over the course of the block. The guild model doesn't lend itself to this type of evolution. Ten guilds is a lot to keep track of. Tracking how each evolves over time would be an almost impossible task.

After sifting through the different mechanical requirements, one thing becomes clear. The guild model presents an interesting structure but it is overwhelming in its size. Second, it doesn't lend itself well to evolution. The key to solving this puzzle was the realization that those two problems have the same answer: The guild model requires space. As does evolution. By giving up the latter we can create the necessary space for the former.

Ravnica wouldn't be a traditional evolution, but rather a large design chopped into three parts.

If Ravnica wanted to be about the guild model, it had to rethink how the block was going to be designed. Ten guilds is too much to throw at the players all at once. If the design team wanted to do right by the guild model it had to give the guilds a chance to breathe. Each guild needed enough room to be showcased. To do so would require all three sets. Ravnica wouldn't be a traditional evolution, but rather a large design chopped into three parts. And once we looked at the numbers (the size of the sets, color balance, space for each guild, etc.) it became clear that the block wanted to have a 4-3-3 structure. That is, Ravnica would showcase four guilds, followed by Guildpact which would showcase three more, followed by Dissension which would showcase the last three.

The more the team looked at this solution, the better we liked it. By breaking up the guilds we allowed each set to spend sufficient time fleshing out the mechanics, focus and route to victory for each guild. (Most importantly, we could give each set new mechanics without violating the essence of the guild model.) We allowed the creative team sufficient space to show and explain what each guild was about. And it even allowed us the freedom to do a number of ten-card cycles in a way that wouldn't chew up too much space in any one set. In short, the solution was perfect. All we had to do now was convince the rest of R&D.

Mixing and Matching

My ten years in R&D has taught me many valuable lessons. One of the most important is the value of showing rather than explaining. Early on I spent a lot of time selling my ideas with words, explaining how they were going to work. The key fault with this method is that it's often very hard to explain the unknown to someone. Even if you completely understand what you're trying to “sell”, actually conveying what you mean can be tricky. But if you just do the thing you want to do and show it to them it has a much better chance of helping them understand what you mean to do. I could tell you about my friend or I could just introduce you to my friend. The latter is light years better.

This meant that the design team needed to execute the 4-3-3 plan to demonstrate what it would be like. The first step of this would be figuring out where exactly each guild went. This sounds a little less complicated than it turned out to be. You see, as you stop and think about what you need, you realize that there are a lot of constraints on how to execute it. For example, we decided that in order to convey the even treatment of the enemy color guilds that we wanted each set to have both ally and enemy color combinations. Two and two in the first set and then two and one (one of which would be one and two) in the next two sets. Next, we wanted to make sure that we had different tempos in each set. All the speedy, aggressive guilds (you know who you are) shouldn't all appear in the same expansion. Likewise for the slow guilds.

We also did a poll in Aaron's column to find out what color combinations players preferred. We then made a conscious effort to spread the most popular ones across all the sets. The Creative Team also made a list of guilds that they thought were slam dunks creatively, the guilds that they had the most confidence would be strong creatively. We consciously spread these out as well. (I'm not going to let you know what those are as the Creative Team did a slam dunk on the whole thing and I don't think you'll be able to tell.) We wanted to make sure that each of the five colors appeared in each expansion. We also did a number of things to help out draft, but I don't want to ruin anyone's fun figuring out how to draft Ravnica (and this block has its curveballs for draft) so just trust me that it was yet another factor we had to take into account.

In the end I believe we had almost twenty criteria to take into account when selecting the guild placements. It was so complicated that we had Paul Barclay (rules manager a few managers back) write a computer program to figure out our options. It turns out there was one. And I don't need to tell you because you'll be able to see it for yourself. Ravnica, by the way, has Black/Blue (Dimir), Green/White (Selesnya), Black/Green (Golgari) and Red/White (Boros). As for Guildpact and Dissension, you'll have to wait and see.

And Now For Something Completely Different

Once we had our four guilds for Ravnica it was time to flesh out how each would work. Check out City Planning, Part III to see what we did (and, of course, how we did it.)

For this week I want to flash ahead to the point where design had finished its initial pass on its four guilds and we took the idea to the rest of R&D. Because we divide up the R&D columns into design and development (Aaron's “Latest Develelopments” column on Friday is the development column - see the word “development(s)” – if you were unaware), we often show you the two sides of the department but we don't talk that much about how we interact.

Here's the short version. Whenever there's a buddy movie, there is most often a responsible one and a somewhat crazy one. The crazy one's job is to move the movie along. He does crazy things because, well, someone has to. The responsible one is there to keep the crazy one safe. Okay, and act as a comic foil. In R&D, design is the crazy one and development is the responsible one. Design has to “move the movie” along by suggesting some radical suggestions. Development then steps in as a sanity check. If Ravnica were a buddy movie, I'd imagine it would have gone something like this:

Design: So we're thinking of doing a multi-color block.
Development: That sounds great. We know the players really like multi-color.
Design: And we're thinking about eradicating the barrier between ally and enemy color combinations.
Development: Interesting idea. We need to be careful, of course.
Design: So we got this thing called the guild system. Kind of a ten-color color pie. We want to give each color combination its own flavor.
Development: Its own flavor?
Design: And to do that we only want to put four of them in the first set.
Development: I'm sorry. It sounded like you said you want only four combinations of multi-color to appear in the first set.
Design: That's right.
Development: But, but then colors won't be balanced. Two colors would have less representation. And, and draft would be… I have no idea what happens to draft.
Design: We're working on both those issues.
Development: What about the other six combinations?
Design: They'll show up later.
Development: How do we expand upon the first set of guilds?
Design: Oh, we don't. We just move on. To be fair, they will get new abilities on mono-colored cards.
Development: Design, you're talking crazy.
Design: Crazy like a fox.
Development: No, more crazy like a crazy person.
Design: You just have to step back and look at the big picture.
Development: This isn't a big picture. This is a cliff. You're asking us to jump off a cliff.
Design: Tomato… tomahto.
Development: A pretty big cliff.
Design: Magic's all about jumping off a few cliffs. Remember Urza's Saga?
Development: That broke the game in two, forced us to ban more cards than in the history of the game and set the power level back years.
Design: But it was a rush. Okay, maybe that wasn't the best example. How about Mirrodin?
Development: Slightly bent the game, forced us to ban more cards in the history of the game other than Urza's Saga and definitely affected the power of the block that followed it.
Design: But it was fun. You want Magic to be an endless run of Mercadian Masques?
Development: Masques didn't make anyone leave in disgust.
Design: Oh, yes it did. Didn't you read the flavor text to Bursting Beebles?
Development: I try not to look at Un sets.
Design: How about split cards? Or pitch cards? Heck, what about multi-color itself?
Development: Some of your crazy schemes have turned out okay.
Design: Don't forget the wacky hijinx.
Development: So how high is this cliff?
Design: Not that high.
Development: It looks high.
Design: They always look high from the top. But trust me, the view down is spectacular.
Development: I don't know how you keep talking me into this.
Design jumps; Development reluctantly follows.

The real story is that Development was reluctant at first but the more Design explained its reasons the more they accepted that the guild system was the right way to go. As these things go, this wasn't one of Design's toughest fights (see “Split Decisions” for an example of a slightly tougher fight).

As is almost always the case in design, the “crazy” idea is all about execution. And if you come back next week, I'll explain how we executed the first four guilds in Ravnica. I'll even talk about how all four of the new keywords were designed. Oh yeah, and I'll also explain how the hybrid cards ended up in Ravnica. But that's next week and this is only Part II, so…


Before I wrap up this column I do have one more responsibility. Being that this is a preview week, perhaps it would be cool if I actually previewed a card. Staying in theme with this week's column, I've chosen a traditional multicolor guild card. (Notice the pretty guild symbol in the background of the text box.) And just to show you that design is willing to act crazy even within the confines of a single card, I've chosen a card that might make you scratch your head. Here's the important thing to remember. The targetting rules changed as of Champions of Kamigawa allowing a single card to target multiple targets (if the card has multiple targetted effects). So yes, the following card can choose three unique targets. Or two. Or one. It's really up to you. Enjoy.

Join me next week when I finish out my Ravnica design story (you know, before I write the week after week Ravnica design columns).

Until then, may you hop off your own cliff.

Mark Rosewater

Latest Making Magic Articles


Vision Design, Set Design, and Play Design by, Mark Rosewater

Magic is a game of change. The cards are constantly changing. The mechanics are constantly changing. The themes are constantly changing. The worlds are constantly changing. The overall st...

Learn More


Odds & Ends: Ixalan, Part 2 by, Mark Rosewater

Last week I started my mailbag column about Ixalan. There were so many great questions that today I'm answering more in Part 2. It was absolutely intentional. One of the things we're a...

Learn More



Making Magic Archive

Consult the archives for more articles!

See All

We use cookies on this site to enhance your user experience. By clicking any link on this page or by clicking Yes, you are giving your consent for us to set cookies. (Learn more about cookies)

No, I want to find out more