Working Draft

Posted in Making Magic on March 22, 2010

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 Draft Week. This week we'll be examining the fun world of drafting. As Making Magic is a design column, the topic seems rather straightforward: How do we design cards for drafting? Just as I was about to put words to paper, I remembered something important. This wasn't the first Draft Week. We had done the theme back in March of 2005.

In my column from that week, I asked the exact same question. The short version of the answer I gave (you can go read it if you want the long answer) was that we don't specifically design for draft. With few exceptions, we don't design a card thinking "this is for draft." In my last article I explained the different things design does for every set that can have an impact on drafting. Rereading that article, I realized that I was asking the wrong question. The better question is this: What can design and development do to enhance drafting? The answer to this question is "a lot." Design does a lot to help drafting just not so much in the card by card design. Aiding drafting in design requires much bigger picture thinking. Today's column will be examining how we do it.

    First Draft

Before I jump into the design issues, I first wanted to talk about the history of drafting in Magic. It's become a giant part of the game. How did it come about? To answer this question I did something that I luckily have the ability to do: I called Richard Garfield. I asked Richard how drafting started. Richard's memory was that it began with the "Penn group." The "Penn group" is another name for the team that designed Mirage and Visions (Bill Rose, Joel Mick, Charlie Catino, Don Felice, Howard Kahlenberg, and Elliott Segal), called such because the majority of the team attended University of Pennsylvania. It appeared the answer I needed was available within the walls of Wizards.

I asked Bill for details and here's what I found out. A number of the Penn group—Bill, Charlie, Howard and Don—were not only gamers but also fans of baseball. As such, one of the games they played was Strat-o-matic Baseball. For those unaware of the game, it allows players to choose actual baseball players to field a fantasy team with. Each player is represented by a card, and the way you choose the players is to lay them all out and draft them, much like major sports team hold drafts every year.

It's common for gamers to cross breed games, so it didn't take long for Bill and the rest to start drafting Magic cards. Here's a good trivia question for your fellow Magic players? What was the first draft format used in Magic? The answer is Rotisserie draft. Yes, the wacky format we talk about from time to time in feature articles and Serious Fun (pushed into the limelight by its inclusion in a Magic Invitational). The very first draft mimicked a Start-o-matic Baseball draft with all the cards available placed face up on the table with a snake-like drafting pattern.

As some point Richard thought about how to sell the cards and thus figured out the 15-card booster pack. Once the booster pack was a known thing, Richard began handing out cards in packs to better simulate the real game. The next evolution was Rochester draft (supposedly named, by the way, because it the first draft of this kind was done in Rochester, New York). Instead of drafting everything at once, the draft went pack by pack.

So how did this evolve into booster draft? No one knows specifically. The people involved were the Penn group, the East Coast Playtesters (Skaff Elias, Jim Lin, Dave Petty and Chris Page), Barry Reich, and Richard. Between them, what we know today as booster draft slowly evolved.

Now that I've admitted not to officially know, let me give you my best educated guess. Richard introduced booster packs to the playtesters. The first evolution of draft was Rochester draft, where the players drafted a limited Rotisserie a booster pack at a time. I believe the precursor of booster draft was probably some format where all the information wasn't open, not every player knew every card. Imagine a Rochester draft where some or all of the cards were turned upside down and only the current drafter got to look at them.

The next leap I'm guessing was the idea to keep the cards in one player's hands rather than put cards on the table face down. Opening a pack and choosing a card feels pretty elegant. At some point, I'm sure they realized that one player could open a booster pack and pass it around. The next step was to let everyone do it at the same time to cut down on how long it took to draft.

What is known is that over time, the Alpha playtesters stumbled upon this way of drafting and drafted as such before the game was ever released. It's funny that when the Pro Tour started (in early 1996), Ramp;D pushed booster and Rochester drafts into the spotlight. We knew in the end one format would end up being the preferred one, but at the time we thought it was going to be Rochester as it's open information allowed for more skill. Ironically, in the end, it was the surplus of information (and longer drafting time) that did Rochester in.

    Rough Draft

With the history out of the way, let's get back to the question at hand: What can design and development do to enhance drafting? My first drafting article covered a lot of the basics, so for today I'm going to dig a little deeper. These are elements of design that add an extra oomph to the drafting experience. These are not listed not in order of importance but in order of how much I had to say about each one.


I've briefly talked about this concept before, but today you'll get the more in-depth version. To understand layering let's begin by talking about a new axis, the number of concepts in a set. On one end of the axis is one idea. (It can't be zero as a set has to have at least one idea in it.) A set on this end of the spectrum has every card dedicated to this idea. On the other end of the spectrum is a set where every card is about a separate idea. Neither end of the spectrum makes for very good Magic.

To understand why, let's look through each extreme through the eyes of drafting. The first extreme creates a draft environment that is one note. Once the players figure out how that one idea plays, the draft is figured out and quickly gets stale. In short, it would be an environment that players didn't want to draft for long. The second extreme creates a draft where there are no inherent strategies. The set has no identity or cohesion and, as such, there are no through-lines to draft. No cards change value based upon deck types, and the draft boils down to taking the best cards, all of which are basically the same to each player. This extreme also creates a draft environment without legs that players stop drafting after a short time.

This axis is very important to designers because we have to understand how many ideas to pack into any particular set. We don't want the set becoming boring or lacking identity, so we have to create a balance between having enough for variety but not so much that it keeps the set from having cohesion. Be aware that there is no one number. I can't tell you exactly how many ideas have to go into a particular set, because it has to do with numerous factors that vary depending in the theme and mechanics.

The general key thought is this. A set has to have a major focus. That focus is going to be larger than anything else. A set at its core cannot be equally about two things, something has to pull focus and thus exist in larger quantity and mindspace. What I mean by mindspace is how much an idea captures the audience of a set. In Alliances, as an example, pitch cards (Scars of the Veteran, Force of Will, Contagion, Pyrokenesis, and Bounty of the Hunt ) were only five cards in the set, but their share of the mindspace was huge. Once you have your major focus, you then need to find things that tie into that focus but pull the game into other directions. These extra things exist in various numbers, none higher than the main focus.

As an example, let's look at Zendikar. Zendikar's main focus was lands; you can think of it as "lands matter" if you like. From the land focus came both the concept of adventure world and the need to use excess mana. The first grew from the explanation of why land mattered more than normal creatively. The second came from the reality of what happen in an environment where land is prioritized. Adventure World led to traps and quests and allies. Extra mana led to kicker, bigger activations, and more expensive spells (Rise of the Eldrazi runs with this last one).

These extensions exist in different numbers much for the same reason that different hairs are cut at different lengths in a haircut. In a haircut, hair is layered to create more volume, to create a look that ties the hair together as a single entity. In card design, effects are layered, well, pretty much for the same reason. When I was creating Zendikar, for instance, the goal of each mechanic wasn't to make it stand out but to use it as a piece to enhance the overall feel of the set.

Another key importance of layering, one that has a big impact on drafts, is that it forces the designer to branch out what the set does. Focus too much of the set on a single facet, and the set starts losing its replay value. Part of the fun of draft is exploration. If the designer does not weave enough new paths to explore, the journey becomes repetitive and boring.

In addition, the mechanics each have a particular function and thus need a different size to accomplish that task. Secondarily, although very important for draft, creating various sizes for different mechanics changes up how much they show up. This variance makes for a much more compelling draft experience. For example, let's say a set has five key components A, B, C, D amp; E (sets, especially large ones, often have more than five components, but the math is easier to demonstrate with five).

If all five components were the same size, you would get exposed to them in pretty equal numbers. There would be some variance, but players more or less would be exposed to them at the same basic volume. If a particular player is more disposed to play mechanic C (and make no mistake, most players are predisposed to play certain mechanics, specifically the ones that tie closer to their psychographic) then many more drafts would involve C. Now, bring up A bit and take down D and E, and you start creating a system with much greater swing. Yes, A shows up in most drafts (as the major focus should; this is balanced by a number of the factors I talk about below), but D and E can be absent from time to time and B and C have some swing from being minor components to being major ones. In short, there is more variance and the drafts have more replay value as there is more uniqueness from draft to draft.

Layering is a tricky design element as it's most noticeable when it isn't there. If done correctly, much like layering of hair, what stands out is the total look and not the role of the individual pieces.


So a set wants different elements in different volumes, but the needs don't stop there. These different elements cannot each exist in a vacuum separated from one another. No, each element has to exist in a larger structure that connects them. A key part of designing a Magic set is making sure that there are different facets to explore but that those facets have some relation to one another, that they have (well, the header gave this away) synergy.

Why do these different pieces have to fit together? The same reason all the seemingly unconnected elements in a story always come together. When players buy a set they aren't just looking for individual cards (okay, they are also looking for individual cards; for most this actually isn't their main motivation), they are looking for an experience. A set is much more than a sum of its parts. This means as the designers of Magic with each set we have to deliver something more. (Speaking of that, if you happened to miss the announcement of the fall set in Magic Arcana last week, take a peek. I can't comment on it yet other than to say my lip is bleeding from how hard I'm biting down on it not to say anything.)

When I think of a set design, I think of it as a singular thing with many parts. Those parts combine to create the whole that is the set. As such, it is crucial that these parts have some meaning in relation to one another. The connection isn't always mechanical; sometimes its flavor, sometimes it's a general mood. The important point is that the pieces must combine into a whole. This means that we have to make sure that the different elements have some meaning to one another.

The importance for draft is that adding synergy between the different elements opens up combinations. One of the favorite tricks in design is to make a few cards that hinge different elements together. For example, the card Harrow in Zendikar pulled double duty. First, it was a card that enabled getting lands onto the battlefield for things such as landfall. Second, it helped advance and fix your mana, which was key to the many parts of the set that pushed you to play with larger spells and effects.

When used correctly, synergy creates a jigsaw puzzle effect for draft. Different pieces can connect up to others, making cards fill different roles from draft to draft. For example, green and black could send you one direction, while green and blue take you down a slightly different path. In each draft the green has similar elements, but the pull of black or blue can change which green cards are the most important.


This brings us to a connected topic: a design element I call branching. This term is derived from another game term, a decision tree. In gaming, a decision tree is a path of what players can choose to do. A game with an extensive decision tree gives its players lots and lots of decisions. A game with a smaller decision tree limits the players choices down to a few key options. In game design, games tend to want to start with a pretty simple decision tree and build as the player advances in the game.

Note that decision trees exist in many facets of the game. Individual games have a decision tree, many cards (mostly permanents) have a decision tree, metagames can have a decision tree and, of course, drafting has a decision tree. (The reason booster draft is the default and not rotisserie, by the way, has a lot to do with the former having a much smaller decision tree at the start. ) The point of all this is that design has to be very conscious of its decision tree.

Branching is the act of creating cards that force a branch, a.k.a. a decision, into the game at key times. Since this is Draft Week, I'll talk about branching in draft. To create variety in drafting, you have to create points in the draft where players have to make a choice. Some of this comes naturally with the format. For instance, choosing your colors is branching that's going to happen without design doing much work.

What design can do is build multiple avenues into different elements, forcing a decision at some point in the draft. The most common example of this can be seen in colors. Most often, we make sure that each color supports at bare minimum two different strategies. This means that even once you've picked a color, you still have to figure out which part of the color you're most interested in. Some cards in the color can overlap both strategies, allowing drafters to stall a little, but eventually they are going to have to draft cards that push them down a particular path.

The reason branching is so important to drafting is simple. It creates more opportunities for variance. The draft in which you go down path A will be very different than the one in which you choose path B. Sure, there are many ways for path A to play out, but if a player drafts it enough, at some point the decisions become more obvious and less interesting as one starts to figure out what works and what doesn't. With branching we just cut down on this repetition, keeping the draft format fresher longer.

Linear Themes

In 2003, I wrote an article (Come Together) where I introduced the terms linear and modular. For a full run down of the definitions, please click the link. The important part for today's discussion is the importance of linear themes in sets for purposes of draft. (Linear themes, for those who refuse to click links, are themes that strongly encourage players to put certain cards with other specific cards. Allies, is the best example of a linear theme in Zendikar block.) An important part of making a fun draft experience is giving people through-lines to draft. While linear themes are not the only way to do this, they are the ones that players grab onto the quickest.

There are tons of reasons other than drafting that sets want linear themes, so it's not a great problem to add linear themes into every set. I'm just bringing it up here because linear themes are another tool in the designer's arsenal to improve drafts.

"Build Around Me's"

I said early that few cards are designed specifically for draft. Few means though that there are, well, a few. The most common card designed for draft is what I call a "build around me." These cards usually go in uncommon and their role is to create a draft deck. The idea is that if players draft this early, they will be set up to draft around it. Decks built around these cards won't happen often, but when they do they allow players to explore completely different facets of the set.

The most common trick in a "build around me" is to make the card cares about something plentiful in the set, but not something any other card cares about. This ensures that there is draft potential while not repeating draft themes already in the set.


The final tool I'll talk about today is really a developer's tool as much as a designer's tool, but we use it plenty so I feel fine talking about it. One of the ways to goose draft is to take a card that normally might sit at rare and move it down to uncommon. Obviously the card cannot be too complicated. The idea is to take a card that might be interesting in draft and lower its rarity so it actually has a larger impact on draft. Cards in this category tend to be very in-theme for the set, as the focus on the theme allows us to justify making it appear more than normal.

    Final Draft

That's all I got for today. I hope this little peek into design helps you understand the types of things we do to help make draft as fun as it can be.

Join me next week when some Eldrazi get woken up and the previews begin. (I've got a doozy of a preview card.)

Until then, may the drafter to your right stay out of your colors.

Latest Making Magic Articles


August 8, 2022

Ravnica Design Handoff Document by, Mark Rosewater

An important part of Magic design is that the set passes through multiple leads. When Vision Design finishes and hands off the set to Set Design, we create a document known as a vision de...

Learn More


August 1, 2022

State of Design 2022 by, Mark Rosewater

Back in December of 2003, I became the head designer for Magic. One of the things I decided I wanted to do was write an article at the end of every Magic "year" where I discuss all the se...

Learn More



Making Magic Archive

Consult the archives for more articles!

See All