Interaction Density and Why Cube Games Feel Good

Cube interaction is the set of cards that let one player affect what the other is doing: removal, counterspells, discard, bounce, combat tricks, and the creatures that fight. How many of those cards a cube carries, and which kinds, decides more about how the games feel than any other single design choice. Too little and every game is two players goldfishing at each other. Too much and nothing sticks and nobody gets to do their thing.

This is a piece about the middle, and about why “how much removal” is the wrong question. The better question is what the interaction lets the other player do in response.

What does interaction do for a game?

A Magic game has a shape: both players build toward something, and the question is whether they get there. Interaction is what makes that question a question. Without it, the game is decided by who drew the better curve, and both players know it by turn four.

But interaction does something subtler than stopping threats. It creates decisions on both sides. If I know you might have Lightning Bolt, I have to think about whether to play Dark Confidant into it or wait. If you know I might have Counterspell, you have to think about whether to run out Wrath of God or bait first. Those decisions are the game. A cube with the right interaction has them on every turn. A cube with the wrong interaction, or none, has them only at the start and the end.

So the goal is not “enough removal to kill the threats.” It is enough interaction, of the right shapes, that both players are making real choices most turns.

How much interaction does a cube need?

A house guideline, and we will flag it as one: in a 360-card cube, somewhere around 25 to 30 percent of the non-land cards should be interaction of some kind. That is roughly 80 to 95 cards. In a 540, the proportion holds.

Broken out by color, a reasonable spread for a 50-card color section:

Green gets less because green’s interaction is weaker and because green decks are supposed to be the ones making threats. That is a feature. A cube where every color has twelve removal spells is a cube where every color plays the same.

The percentage matters less than the distribution across the draft. If 80 removal spells are concentrated in black and red, then Azorius and Selesnya decks have none and their games are races. The question to ask of each lane is: when this deck faces a resolved threat it cannot beat in combat, what are its outs?

Which kinds of interaction feel good?

This is the part most counts miss. Not all interaction creates the same decisions.

Cheap, conditional removal creates the best games. Fatal Push, Lightning Bolt, Swords to Plowshares. One mana, kills most things, does not kill everything. The attacker has to think about which threats to expose. The defender has to think about when to spend it. Both players play around it.

Unconditional removal at fair cost is fine. Go for the Throat, Hero’s Downfall, Murderous Rider. Two or three mana, kills the thing. Less interesting than Push because there is less to play around, but it keeps midrange threats honest.

Sweepers are necessary and should be rationed. Wrath of God, Damnation, Anger of the Gods. They punish overextending, which is a lesson worth teaching, but a cube with too many of them makes aggro unplayable, which we wrote about in aggro in cube. Two or three per color that gets them, and no more.

Soft counters beat hard counters for game feel. Mana Leak, Memory Lapse, Remand, Miscalculation. They stop the spell now but not forever, and the caster has a path: play around it, or wait a turn and pay. Counterspell and Force of Will are fine in moderation, but a cube where blue has eight hard counters is a cube where non-blue players feel they are not allowed to play.

Discard is interaction with information. Thoughtseize, Inquisition of Kozilek, Duress. The caster learns the hand. The victim loses a card but keeps their plan. These are good cube cards because they create knowledge and knowledge creates decisions. Hymn to Tourach is interaction too, but it creates no decisions for the victim and is better left out of most lists.

Creature-based interaction is underrated. Flametongue Kavu, Shriekmaw, Ravenous Chupacabra, Skyclave Apparition, Brazen Borrower. A creature that removes a creature puts a body on the board, which means the game continues with both players having something. These are the cards that make midrange decks feel like decks.

If this is useful, the newsletter sends more like it.

What happens with too little interaction?

The games become races. Each player’s turn is “play my best card,” and the opponent’s turn is the same. The decision space collapses to sequencing and mulligans, and the better draw wins. Players notice this quickly, even if they cannot name it. The symptom is games that end on turn five or six with nobody having cast a spell at the other person.

The low-interaction failure usually happens in cubes that are built from a “best cards” list. The best cards are disproportionately threats, because threats are what get written about. Owners end up with 15 planeswalkers and 6 removal spells that can hit them.

The other version is a cube built around “fun” where the owner has cut all the interaction because killing people’s creatures is mean. The result is that games become boring, which is a worse outcome than games becoming mean.

What happens with too much?

The games become attrition. Every threat dies, every spell is countered, and the winner is whoever runs out of answers last. This is not nothing; some players love it. But a cube built that way punishes the drafters who want to do a thing, and the thing they want to do is the reason most of them came.

Signs: games regularly go to turn twelve. Nobody plays aggro because every creature dies. Blue is the best color every week because blue’s interaction stops the most things. The deck that wins is the one with the most two-for-ones.

Many cube designers have, over time, leaned away from cheap unconditional sweepers and toward interaction that is conditional or costs tempo, because they want threats to matter. That is the direction most well-maintained cubes have drifted, and it is the direction we have taken ours.

How do you tune it?

Interaction density is adjusted with the same tools as everything else: notes and swaps. But a few specifics help.

1. Count your answers to your best threats. List the ten scariest permanents in the cube. For each, count how many cards in the cube answer it at parity or better. If a threat has fewer than ten answers, it is probably too strong for the cube’s interaction, or the interaction is too narrow. A planeswalker that only dies to Hero’s Downfall and attacking is a problem.
2. Check the one-mana count. A cube with fewer than ten one-mana interaction spells across all colors will play slow. The cheap removal is what lets a player with a worse hand still have a game.
3. Watch the blue decks. If blue wins more than its share, it is usually because blue’s interaction is too good relative to the threats. The fix is usually to swap a hard counter for a soft one or a threat that dodges counters.
4. Ask the losers. After a game, the loser can tell you whether they lost to a decision or to a draw. “I should not have played into Wrath” is a good loss. “I never had a chance to do anything” is a density problem.

Track this in the same changelog you use for everything, described in how to update a cube. Interaction counts go up or down one or two cards per update, never by ten.

FAQ

How much removal should a cube have?

A common house guideline is 25 to 30 percent of non-land cards being interaction of some kind, spread unevenly by color: blue, black, and red carry the most, white a little less, green the least.

What is the best kind of removal for cube?

Cheap conditional removal like Fatal Push and Lightning Bolt creates the most decisions for both players. Unconditional removal is fine in moderation; sweepers should be rationed.

Are counterspells bad for cube?

No, but hard counters in volume make non-blue decks feel powerless. Soft counters like Mana Leak and Remand keep blue interactive without shutting the opponent out.

How do I know if my cube has too little interaction?

Games end by turn six with neither player having cast a spell at the other, and the winner is whoever drew the better curve. That is a race, not a game.

Give both players a turn

Interaction is what makes the other player’s turn matter to you. Get the density right, choose kinds that create decisions on both sides, and the games will feel like games. The game is the reason. The people are the point.

Subscribe to The Ludarium newsletter for new pieces on cube design and the groups that play it. If you are near Indianapolis, Cube Commonwealth, our cube group, is relaunching as part of The Ludarium; subscribers hear first when the first night is set.