This content was originally included in an issue of The Duelist magazine. The original article can be accessed via Internet Archive here.

---

Main Magazine Page: The Duelist #0

By Richard Garfield

The Magic: The Gathering design and playtest lore is so rich that the couple of pages I am going to write on it cannot possibly do it justice. It couldn’t be any other way; for the last couple of years it has probably been the biggest thing in my life. The amount of people-hours we spent on it is boggling. Repeatedly I have been frustrated by the realization that I am the world’s foremost expert on Magic, yet I’m clueless so much of the time; which is better, Plan A or B? Who knows—nothing like this has been done before.

What prevents someone from purchasing ten decks and assembling a collection that no one can beat? That may actually have been my main concern in design. Oh, I had a lot of theories about what would prevent this, all of which were partially true and none of which were entirely true. Maybe my most compelling theory was growth over time; if you have a deck that beats mine nine times out of ten, your ante will usually be more valuable than mine, and my deck will get stronger each of the few times I beat you. I also suspected that those who bought power rather than dueling or trading for it wouldn’t have the strategic understanding of someone who earned it the hard way. On top of This was the sentiment that buying a lot of poker chips doesn’t Make you a winner. When the smoke all cleared, though, the answer turned out to be simple: Magic is a fun game, and it really doesn’t matter so much how you got your deck. In playtesting, if one person’s deck got too powerful, people wouldn’t play for ante against it until that person took some form of handicap, or until the other decks were competitive. We also developed leagues in which players were limited to one deck apiece, or had to assemble decks by drafting from a large, common pool of cards. One league allowed players to design “wish decks” of any cards, with some liberal constraints.

In some sense, after that point Magic was alive. I didn’t really do much more design, just gave it room and time to evolve on its own. Normally in game design I tell the game what I want to see, it acquires those characteristics and a few of its own, and that’s it. With Magic, however, the game often stole the driver’s seat.

One surprising thing I realized well into the second year of playtesting was that Magic was one of the best economic simulations I had ever seen. We had a free-market economy with all the necessary ingredients for interesting dynamics. People valued different cards in different ways—sometimes because they just weren’t evaluating them accurately, but more often because some cards actually were better for one player than another, giving lots of opportunity for arbitrage.

Sometimes the value of a card would fluctuate based on new uses, or even suspected uses of a card. For example, when Charlie Cantino was collecting all the available spells that produced black mana, we began to get concerned. Those cards soon demanded higher and higher prices, and people generally wondered what he could possibly need all that black mana for. Even trade embargoes appeared; at one point a powerful faction of players would not trade with anyone who traded with Skaff Elias. I actually heard conversations such as:

A (to B): I will give you card 2 for card 1. Skaff (watching): That is a ridiculous trade, I will give cards 2, 3, 4, 5, and 6 for card 1. A: We are not trading with you, Skaff.

Needless to say, Skaff was perhaps a bit too successful in his Early duels and trades.

[Wooly Mammoth blank text box] “. . . when my collection of Mammoths topped seven I ran out of dwarves to name them after . . .” — Richard Garfield

Magic is probably closer to roleplaying than any other card or board game. Each player defines his or her deck by deciding what collection of cards to assemble, and people’s decks and cards have often evolved enough character that the owners felt compelled to name them. My favorite personal collection as a deck named The Reanimator, and when my collection of Mammoths topped seven I ran out of dwarves to name them after, so I started making up new dwarf names like Cheesy and Hungry.

In a way, even the object of the game is up to the player, like in roleplaying. Certainly the object of a duel is to win, but the duel is such a minor piece of the game. Trading and assembling a deck are more important than any number of duels, in most circles. Some people play to put together bizarre decks, others to get one of each card if they can, others to win competitions, others to collect as many of one card or a certain set of cards as possible. All these objectives were present in our two years of playtesting, and I am sure they will be present out in the real world with real decks. I don’t think there is a person on my playtest list who doesn’t have a scummy pack of teeny, homemade cards stashed somewhere. If in real life people are half as enthusiastic about collecting and playing with the beautiful cards Jesper Myrfors put together, well, the game environment is bound to be wild.

There has never been a game more difficult to playtest than Magic. I sincerely believe this. The basic framework seems fairly robust, and it’s possible to include many, many cards in the pool, but which cards? And at what frequency? Common cards have to be simple and easy to understand, but not necessarily less powerful than other cards; if only rare cards were powerful, you’d have to be lucky or rich to get a decent deck. Sometimes we made a card rare because it would be too powerful in large numbers, but more often we used this designation for more intricate or specialized spells, ones you wouldn’t want a lot of anyway. Still, these design guidelines only got us so far. The game’s whole flavor would change if a few seemingly innocent cards were eliminated, or merely reduced in frequency. The number of combinations and permutations is boggling. When it came down to actually deciding what to include and what not to include, I began to feel like a chef obliged to cook a dish for ten thousand people using three hundred ingredients.

Sometimes innocent-sounding cards can combine into something truly frightening. A good part of the playtest effort was to rout out cards that contributed to “degenerate” decks: narrow, Powerful decks that are difficult to beat and usually boring to play. Without a doubt the most striking was Tom Fontatine’s amazing Deck of Sooner Than Instant Death, which was renowned for being able to generate upwards of eight large monsters on the second or third turn. Many of the cards he used Are still part of the game, but they have been made rare so that those interested in forming such decks will really have their work cut out for them.

In the end, we decided that part of the fun of the game was Creating the so-called degenerate decks. People would assemble Them, play with them until they got bored or their regular opponents refused to play with them, and then either retire the deck or trade off its components for something new, a sort of Magic version of putting the champion out to stud. Most people end up treating their degenerate decks much like roleplayers treat their most successful characters; when you eventually get tired of them, you relegate them to the background, occasionally pulling them out and blowing off the dust for a new duel.

During playtesting, Mikhail came to me and said, “I like my new deck; I have the most powerful card in the game. When I play it, I win on the next turn.” I wracked my brains trying to figure out what he was talking about. I couldn’t think of anything so powerful that he could always win a turn after playing it. So I asked him, and he showed me a card that would make his opponent skip a turn. I was confused until I examined the wording more carefully: “Opponent loses next turn.” It was my first real lesson in how hard it is to word these cards unambiguously.

One interesting thing about playtesting Magic was the huge amount of dissent among the playtesters about which cards should and shouldn’t be part of the game. Many of them felt strongly enough about it to make their own versions for playtesting, a significant task that involves making and shuffling about four thousand cards. Each version had a different flavor, so we decided that each new release for Magic will have all-new art and different, though possibly similar, mix of cards. The playtesters have yet to tire of these new card mixes; each has its own quirks and secrets, and each one gives the opportunity to wheel, deal, and duel in a new environment.