A set cube is a cube that rebuilds one retail draft format so you can draft it whenever you like, long after the boosters are gone. Triple Innistrad. Khans of Tarkir. Rise of the Eldrazi. Modern Horizons 2. The best draft formats last about three months in stores, and a set cube is how a playgroup keeps one.
A set cube is not a singleton cube with a flavor restriction. It deliberately includes duplicates, keeps the bad commons, and builds packs that look like real boosters. That makes it different in construction from everything else in a cube collection, and this article is about those differences: the rarity math, the pack build, what to cut, and when to cheat.
Why not just buy a box?
Because a booster box is gone after one draft and the format you loved is gone with it. A sealed box of original Innistrad sells for several hundred dollars at the time of writing, about three times what it cost in 2011, and even when a set is in print, a box is single-use. A set cube costs about one box’s worth of singles once and drafts forever.
There is also a design reason. Retail boosters are random. A set cube is curated randomness: you decide that the rare slot never produces a Griselbrand, that the best common appears at the frequency it did in retail, and that the one archetype that never came together in stores gets a slight nudge. The format you rebuild can be a little better than the one that existed.
How is a set cube different from a normal cube?
Three ways, and all three cut against normal cube instincts:
- Duplicates are the point. A draft format’s texture comes from seeing Travel Preparations in three packs a draft. Singleton kills that. Commons appear in multiples; uncommons in fewer multiples; rares once.
- Filler stays in. Normal cubes cut every weak card. A set format needs weak cards; they are what makes pick eight a decision and what makes the good commons feel good. You cut the worst of the worst and keep the rest.
- Packs are built, not dealt. A set cube’s packs follow the retail rarity slots. You do not shuffle 540 cards and deal 15; you construct packs with the right number of commons, uncommons, and rares.
What rarity ratio should a set cube use?
The useful public reference is the Triple Innistrad Set Cube on CubeCobra, which documents its math in full. It runs 848 cards built to produce 60 packs, enough for two eight-player pods and a half pod of four. Its ratios:
- Commons: five copies of each of the set’s 101 commons (505 cards), plus about 35 extras for swapping.
- Uncommons: three of each.
- Rares and mythics: one of each non-double-faced rare (53), plus 7 of the 15 mythics drawn at random each time the cube is rebuilt. The unused mythics rotate in next time.
- Double-faced cards: each paired with a checklist card so the pack does not reveal it.
Each 14-card pack then gets 9 commons, 3 uncommons, 1 rare or mythic, and 1 double-faced card of any rarity, which mirrors the retail Innistrad booster, where a dedicated double-faced slot replaced one common slot (the original packs also carried a basic land and a token or ad card, which the cube leaves out).
You do not have to go that big. A lighter build that still serves an eight-player set cube:
| Rarity | Copies each | Example set (100 C / 60 U / 53 R / 15 M) | Cards |
|---|---|---|---|
| Common | 3 | 100 × 3 | 300 |
| Uncommon | 2 | 60 × 2 | 120 |
| Rare | 1 | 53 | 53 |
| Mythic | 1 (or rotate half) | 8 to 15 | 8 to 15 |
| Total | ~480 to 490 |
That builds 24 packs of 15 cards (10 commons, 3 uncommons, 1 rare or mythic, 1 land or flex) with about 120 cards left over, which lets you rebuild packs with a different mix next time. It will not perfectly mimic retail, where a given common shows up more rarely relative to its pool, but the texture is close and the box fits in a normal 540 case.
The number to watch is commons per pack divided by unique commons. In retail, nine or ten commons per pack drawn from a pool of about 100 means a given common turns up in roughly one pack in ten. Three copies of each across 24 packs works out to about one pack in ten, which matches retail.
How do you build the packs?
Pre-building packs is the part most set cube owners underestimate. The Triple Innistrad method, adapted from a community shuffle technique:
- Sort commons by color, with artifacts, gold, and lands as a sixth pile.
- Deal each pile evenly across all packs so each pack gets about the same color distribution a real booster had (roughly two of each color plus one or two colorless).
- Deal uncommons one or two at a time, then rares one per pack.
- Slip the DFC and its checklist card in last, or put the flex slot last.
- Rubber-band or sleeve each pack. Store them built. Rebuild once every pack has been opened.
Pre-built packs are what turn a set cube into a 15-minute setup instead of a 45-minute one. Build them the day after the draft, not the night of.
For a set with a dedicated pack slot (Innistrad’s DFC, Kaldheim’s snow land slot, a bonus sheet like the Strixhaven Mystical Archive), copy the slot. The slot is part of why the format felt the way it did.
If this is useful, the newsletter sends more like it.
Which sets make good set cubes?
The sets people rebuild are the ones with a clear, deep archetype structure and a common layer good enough to carry a deck:
- Innistrad (2011). Ten clean archetypes, flashback, morbid, a DFC slot, and Travel Preparations. Probably the most-rebuilt set, and the one with the best-documented public list.
- Khans of Tarkir (2014). Five three-color clans with morph as glue, the best fixing at common of any set, and a format where the picks are hard from pack one.
- Rise of the Eldrazi (2010). A slow, strange, level-up and Eldrazi Spawn format with almost no overlap with anything else. People who loved it rebuild it because nothing else plays like it.
- Modern Horizons 2 (2021). Dense, high-power, with a bonus-sheet-free but reprint-heavy structure; expensive to assemble but a favorite.
- Dominaria (2018) and Ikoria (2020) come up often for the same reason: strong commons and a mechanic (historic, mutate) that rewards drafting around it.
Original Ravnica block and Shadowmoor are built too, usually as the whole block rather than one set, with packs mixing the sets at the retail ratio (one pack of each).
When should you cheat the set?
Purists rebuild the exact format, filler and all. Most set cubes cheat a little, and the good cheats share a rule: change the frequency, not the card pool.
- Trim the rare slot. Cut the two or three rares that ended games alone (in Innistrad, think Olivia Voldaren or Bloodline Keeper). The slot still exists; it just hits a bomb less often.
- Remove the unplayable commons that were never drafted, but keep the mediocre ones. Cutting all of them turns a set cube into a worse singleton cube.
- Fix the one archetype that never worked. If a color pair’s signpost uncommon was too weak, run three copies instead of two. That nudges without redesigning.
- Swap in a card from the same block if a slot needs it. The Triple Innistrad cube’s 35 spare commons exist for this.
What not to do: add cards from other sets because they are better. A Lightning Bolt in a Khans cube is not Khans anymore. If you want a curated best-of, build a plane cube and call it that.
How much does a set cube cost?
It depends entirely on the set. Commons and uncommons from any set older than two years are nearly free in bulk; the cost is the rare slot and the handful of chase uncommons. A Khans set cube is cheap except for the fetchlands, which most owners proxy or replace with the set’s tri-lands and common gain lands. An Innistrad set cube’s rare and mythic slots include Snapcaster Mage and Liliana of the Veil, and a few cards (Parallel Lives, Snapcaster Mage, Liliana of the Veil) make up much of the cost of the rare slot. For the pure-play case, that is a reasonable place to use proxies.
FAQ
What is a set cube in Magic?
A set cube is a curated card pool that recreates one retail set’s draft format, with duplicates of commons and uncommons and pre-built packs that follow the real booster’s rarity slots.
How many cards does a set cube need?
About 480 to 500 for an eight-player draft at three commons, two uncommons, and one rare each; one draft uses 360 of them. Full-fidelity builds like the Triple Innistrad cube run 800 or more.
Should a set cube be singleton?
No. Repeated commons are what give a draft format its texture, and singleton removes that. Rares are usually singleton; commons run three or more copies.
Which set is best for a set cube?
Innistrad is the most-built because of its ten clear archetypes and strong commons. Khans of Tarkir and Rise of the Eldrazi are the other frequent choices.
The format you miss, in a box
A set cube is the most sentimental thing in a cube collection and the one that gets the loudest table. The game is the reason. The people are the point.
Subscribe to the Ludarium newsletter for new pieces on building and running cubes, and if you are near Indianapolis, you will hear first when Cube Commonwealth, our cube group, sets its first night as part of The Ludarium.
Get the next one in your inbox.
Cube news, draft guides, and Cube Commonwealth updates from The Ludarium.