Monday, September 14, 2020

The Three Immutable Laws of Magic Design

One of Magic's defining features is its mutability. Every format and every set brings with it new opportunities for players to redefine the game. Mark Rosewater calls this process of constant renewal and discovery "Crispy Hashbrown Theory" and posits that it's one of Magic's core strengths. While I generally agree with this, I have two objections. The first is that hashbrowns aren't the best example of a breakfast food you get sick of partway through and that it should be called "Pancake Theory" instead. The second is that while constant change is the lifeblood of the game, Magic has a set of qualities that should never be changed, no matter how weird or experimental you're getting. These are the three immutable laws of Magic design.


Games fulfill people by challenging them in ways that aren't achievable in real life. (I write more about the subject in this blog post.) As such, if a player's action doesn't bear a risk of failure, it isn't a satisfying action – something with no chance of failure or error is rote, a checking-off-the-boxes. This risk manifests itself in different ways, from suboptimal actions in Euro-style games to simply not having your card get picked in an impress-the-judge party game.

In Magic, this risk is represented by your opponent or opponents being able to oppose your plans. Creatures can be destroyed, diminished, or outclassed by larger creatures; spells can be countered or their targets made irrelevant. A well-played game of Magic isn't a cunning plan executed perfectly as much as a series of improvisational decisions made better than the other player.

In order to maintain this risk in Magic, strategies should always have points at which they can be interacted with. When they can't, strategies become boring and linear to play and deeply frustrating to play against. There are a lot of axes to pay attention to:

  • Does the strategy act too quickly, or in too tight a time frame, for anyone to be able to interact with it? (Copy Cat, Modern Dredge pre-Golgari Grave-Troll ban)
  • Does the strategy use so many individually strong cards that the opponent can't answer each one before you take over the game? (Cawblade, Czech Pile)
  • Does the strategy rely on card types that are difficult to interact with, or on zones that can't be interacted with through normal means? (Storm, Dredge)
  • Does the strategy prevent the opponent from interacting with it entirely? (Lantern Control; other prison and stax decks, though they generally haven't been powerful enough to be widely problematic)

While the previous bullet points refer to decks or broad strategies as a whole, individual cards can suffer from an uninteractivity problem, particularly in Limited where synergies aren't as powerful. Hexproof cards are the poster child for this issue, Jade Guardian in Ixalan being the most recent example that significantly destabilized a format, but they aren't the only way this manifests. This whole section was actually inspired by Patrick Sullivan's opinion on Ravenous Chupacabra, which virtually couldn't be interacted with because it had an extremely powerful enters-the-battlefield ability. While you could technically remove the resulting vanilla 2/2, your opponent has advanced their strategy in a way that couldn't be responded to equitably. Strong ETB abilities in general have been a concern in modern Magic design, with no easy answer as to how to fix them. There's also the case of Gitaxian Probe in Infect, where it became too easy to determine whether your opponent could interact with your big infect play.

This is arguably the hardest law to follow if you're moving into Magic design from being a long-term Magic player, because each role wants fundamentally opposite things. If you're a player, you have the right to go ahead and make your deck as hard to interact with as you can, because that's good deckbuilding. But as a designer, you need to make the players' lives hard so they can have a more fun time playing Magic. When a player feels like their strategy can be seriously opposed by their opponent, and they're able to overcome it and pull out a win, it feels much more fulfilling than if they merrily waltzed their way to victory from their opening hand.


Magic is engaging for longer than many other games because so much of it is built around variance. Even in Constructed formats, where you may have two to three optimized decks slamming into each other repeatedly, no two games are the same because all cards are drawn in a random order. Sometimes both players draw god hands and collide with each other at full force, and other times both players rip five lands off the top and the game turns into the scene from Pokémon: The First Movie with the two Pikachus lightly slapping each other because they don't have the strength to do anything else.

Maintaining a level of variance and unpredictability in Magic gameplay is critical to keeping the game fresh and interesting. When decks become too consistently able to do the same thing every game, the game becomes stale for everyone involved. Fortunately, the base level of variance in the game is high enough that in Limited, where the designer generally has the most control over what the environment looks like, it's very difficult to make every game feel the same. It's primarily designing for Constructed formats where designers have to exercise more caution.

Much of Constructed Magic is reducing variance, whether it's through an aggressive red-based deck having many efficient one- and two-drops or through a control deck playing a lot of scry and card draw. Decks that have a chance of winning on turn 1 but sometimes sputter out are usually foregone in favor of decks that can consistently win on the same turn. This puts designers in an interesting balance, where not giving tools to handle variance makes the game too reliant on luck, while giving too much control makes the game into chess.

As a specific example, the big issue with the Companion mechanic (that didn't have to do with power level) was that it violated the principle of unpredictability. If you can always cast Lurrus of the Dream-Den on turn 3 every single game, your games with that Lurrus deck are going to play in a much more homogenous fashion than if you had four copies of Lurrus in your maindeck, where you'd be able to land one on time more often than not but wouldn't be guaranteed to do so by any means. Tutors are another issue, but an obvious enough one that they haven't been a significant part of most formats since Time Spiral block and mechanics like transmute and typecycling were slowly phased out.

Conversely, you can use this principle to design cards that might function differently in different situations. Cryptic Command is beloved by players not just because it's effective in many formats, but because you don't know exactly what you might end up doing with it when you draw it. True, most of the time you'll want to counter a spell and draw a card (or bounce your Mystic Sanctuary), but there will be other times where you want to tap down your opponent's board or even just turn it into an expensive Repulse. The corner cases and weird stories that cards like Cryptic generate are what make Magic as popular as it is.


As much as Magic is a game of variance, it's also a game of combinatorics. The way different cards interact with each other is part of the game's big draw to deckbuilders and brewers, who constantly look for ways to exploit abilities in new and interesting ways. What's important, however, is that these players feel like the interactions between cards is something that they were able to figure out themselves and not something that the designer suggested to them. This is the core of the principle of subtle design.

Synergies in Magic can manifest themselves in many different ways. The most obvious are synergies that explicitly tell you to have something with X quality, such as various tribal cards. This is effective in giving players a strong idea of what to draft, and tribal strategies have always been beloved by the player base, but it risks making decisions of how to build and play feel prescriptive. Battle For Zendikar suffered from having too many explicit synergies: The major strategies in Limited were Allies, which were a creature type, and colorless-matters, which were only applicable to cards that were specifically marked as colorless.

A step up from this in subtlety are cards that reward you for taking an action that can be achieved in a number of different ways. The "draw-two" archetype that we've seen in several recent sets triggers off lots of different effects, including cantrips, looting/rummaging effects, draw triggers on creatures, and so on. When effect actions like these are available in many colors, it gives players the option to play synergistic cards in a color combination outside of the archetype, rewarding more insightful deckbuilding.

The highest level of subtle design are cards that happen to work better with other cards without explicitly spelling it out. The Mentor mechanic from Guilds of Ravnica is an excellent recent example of how this level of subtlety can create a more interesting play experience. There were lots of ways to play around with the mechanic: You could boost the power of your Mentor creatures with combat tricks to put the +1/+1 counter on a wider range of creatures, play weak creatures with great abilities like Healer's Hawk that benefitted from the counters, or chain Mentor creatures with cascading power levels to boost your board as a whole. Suddenly, power – an unassuming statistic on every creature – became a critical component to proper deckbuilding in decks that bordered Boros colors.

On the other side of the coin are cards that subtly improve your matchup against an opponent's deck. While Plummet effects and the like are useful for decks that don't have anything that can interact with key cards, the depth in first-to-three Magic is in finding chaff that works well in the matchup. One of the most illuminating moments of Magic to me was when my opponent in Magic Origins Sealed used an expensive removal spell on a Maritime Guard I had sided in against his mono-white Renown deck. In a set where the best cards were 2/2s and 2/1s that really wanted to connect with the opponent once, a vanilla 1/3 in the most defensive color was a brilliant inclusion.

If all of the synergies in a set or format are too explicit, there isn't room for creative deckbuilding or discovery of new interactions – since these are what set Magic apart from its competitors with lower skill floors, it's critical that products aimed toward new players include them. A player who discovers an interaction on their own without prompting, and who exploits that interaction to victory, is a player who's going to keep playing the game for a long time.


When I was starting to get into game design in 2015, I read an article that I can no longer find that discussed the importance of "planting your flag," or deciding on a core conceit when starting a game design so it doesn't lose its focus. The three immutable laws of Magic design – interactivity, unpredictability, and subtlety – are individual parts of Magic's three-colored flag. By planting them at the center of your designs, whether a single card or an entire set, you'll be able to maintain the spirit of the game that has captured so many people for so long.

1 comment:

  1. Great analysis, thanks for posting!

    > it should be called "Pancake Theory" instead

    I can relate! Though in my experience the problem can always be solved with more maple syrup.