A custom MTG set is a themed group of cards designed to be played — and usually drafted — together. The craft is less about any single card and more about how they all fit: a shared mechanical theme, balanced colors, a sensible rarity ladder, and a Limited environment that stays fun across twenty games, not two. The good news for custom designers is that you don't need a 250-card behemoth to get there. A tight, finished small set beats a sprawling unfinished one every time.
This is the workflow — from blank page to a set your playgroup can actually draft.
Start by scoping the set
The single most common way a custom set dies is being too big. A 250-card set is a year of work, hundreds of balance decisions, and a draft environment that's genuinely hard to tune. Almost nobody finishes their first one.
Scope down. A first set of roughly 60–100 cards is finishable in a reasonable stretch and still drafts well with a pod. If even that feels heavy, design a micro-set built around a single two-color archetype — twenty cards that do one thing well. You can always expand a set people actually played; you can't salvage one you never finished.
Pick a theme: top-down or bottom-up
Every set is about something. There are two ways in.
Top-down starts with flavor: a haunted city, a pirate archipelago, a war between two planes. The mechanics then express that fantasy — a graveyard theme for the haunted city, raid-style aggression for the pirates. Top-down sets are evocative and easy to get excited about.
Bottom-up starts with a mechanic: a set built on +1/+1 counters, or on casting lots of instants and sorceries, then dresses it in whatever flavor fits. Bottom-up sets tend to be mechanically tight.
Either way, keep the count of central themes small — two or three keywords or mechanics the set returns to, not ten. And watch for parasitic mechanics, which only work alongside other cards carrying the same mechanic. One or two parasitic threads are fine; a set built entirely from them collapses if a drafter doesn't open the payoff. The color pie tells you which colors should carry each theme, and the balanced-design guide keeps the individual cards honest.
Build a set skeleton
The skeleton is the deliberate distribution of cards across colors, types, and roles — the structure you fill in rather than designing card-by-card and hoping it balances.
Start with roughly even cards per color, then make sure each color has the tools it needs. Every color wants a curve of creatures, and the set as a whole needs answers in the right places: red and white removal, black's destruction and edicts, blue's bounce and counters, green's fight effects (the color pie again). A Limited environment lives or dies on its creature curve — you need enough clean two- and three-drops that games are decided by combat and decisions, not by whoever drew their bomb.
Think in terms of roles, not just cards: bodies, removal, card advantage, tricks, fixing, and finishers. If a color is missing a role at common, drafters in that color will feel it every game.
Use rarity to teach the set
Rarity isn't just a power dial — it's how a set teaches itself to the people drafting it. (For the full breakdown, see MTG rarity explained.)
- Commons are the backbone. They should be clean and repeatable, and collectively they define what each color does in draft. Most of your set is commons, and most games are decided by them.
- Uncommons enable the two-color archetypes. The classic tool is a gold "signpost" uncommon for each color pair — a card that announces, in one read, what that pair wants to do.
- Rares and mythics carry the splashy payoffs, the build-arounds, and the legends. They're allowed to be complex and exciting because a drafter sees only a couple per draft.
The rule of thumb: push complexity and wordiness up the rarity ladder. A confusing common slows every game; a confusing mythic shows up once.
Design around the ten draft archetypes
A two-color draft set is really ten mini-strategies — one for each pair of colors. Giving each pair a clear identity is what makes a set draftable instead of a pile of good cards.
The ten pairs each "want" something different — one pair might go wide with tokens, another lean on instants and sorceries, another fill and exploit the graveyard, another simply curve out and attack. You don't have to reinvent these; the framework is standard. Your job is to make sure the commons support each archetype and an uncommon signpost points the way. When every pair has a lane, drafters can read your set and commit, which is the whole point of a Limited environment. The card types guide is a useful checklist for making sure each archetype has the pieces it needs.
Build, test, and iterate
Design the skeleton, fill it with cards, then playtest — this is where a set is actually made, not when the last card is written.
PipGlyph is built for exactly this loop. Build and tweak individual cards in the editor, keeping an eye on the skeleton as you fill it — a color short on two-drops, a rarity that's overloaded, a curve that tops out too high. Then deal booster-style hands from the finished pool and draft-test the environment with your playgroup.
Then iterate. Change one variable at a time — a cost, a common's stats, an archetype's payoff — and re-test. Sets converge slowly; trust the games over your instincts.
Share it and rally a community
When the set holds together, publish it to the gallery and hand your playgroup the link — the set page is draftable by anyone. For extra cards and momentum, run a design challenge around the set's theme and let the community pitch designs that fit. A set people contribute to is a set people want to draft.
FAQ
How many cards should a custom MTG set have?
For a first set, aim for roughly 60–100 cards, or even a single two-color archetype of about twenty. That's enough to draft and small enough to actually finish and balance. Full retail-style sets run 250-plus cards, but those take experienced teams months — scope down until you've shipped one.
Should I design a set top-down or bottom-up?
Top-down starts from flavor (a world, a story) and builds mechanics to match; bottom-up starts from a mechanic and adds flavor later. Both work — pick whichever you're more excited to build. Just keep the central themes few (two or three) so the set stays focused.
Why playtest a whole set instead of individual cards?
A set is an environment, not a list of cards. Individual cards can each look fine while the draft format is miserable — too few creatures, one dominant archetype, no answers to bombs. Only drafting and playing the set surfaces those problems, which is why the booster simulator and real games are the real design tool.
How does rarity shape a custom set?
Rarity controls how often a card appears and therefore how much complexity it can carry. Commons are frequent, so they should be clean and define the core of each color; uncommons signpost the two-color archetypes; rares and mythics deliver the splashy, build-around payoffs. Pushing complexity up the rarity ladder keeps the everyday play of the set smooth.