Usman's Cube Content

(Originally posted on my Substack on 10/14/25.)
Originally, this article was started in 2016 with the whirlwind of great sets in that year. It’s now 2025 and not only do we have more sets per year, but the sets are generally more powerful on a pound-for-pound basis.

As someone who got into the world of cube partly from Evan Erwin’s videos about updating his cube and hearing his reasoning behind doing cube updates and as someone who’s enjoyed content about updating cubes, and even the Black Sabbath Song* with Ozzy Osbourne’s recent passing with, cube changes over time have been something that resonated with me. At its core, updating a cube is a simple process: remove cards and add cards that meet your design goals, but the simplicity of the phrase understates its use in shaping a cube environment.

Part of what drove me to finally write about the concept was an innocuous tweet by Ryan Overturf, that thankfully 2016 me took a screen capture of:

When Ryan tweeted this, I thought of cube design akin to maintaining a bonsai tree with very careful small changes that add up, but I’ve realized that while meticulous small changes can be one way to approach cube updates, it’s not the only way to do it, either.

I’ve been around the cube design block for a while, and usually when I see cube changes done, I tend to see them in a few broad categories.

• Level 1: Similar New Card changes: New cards in, cut cards with the same mana cost/function out that aren’t performing.
  • Level 1a: New Card changes: New cards in, cut cards that aren’t performing.
• Level 2: Including (not just new) cards to support a new card/cards.
• Level 3: Active macro-level decisioning to address issues in the metagame.
  • Level 3a: Shaping a cube’s metagame outside of issues in the metagame.

With that being said:

• Like with other cube articles I’ve written recently, the main purpose of this article is intended to be illustrative rather than prescriptive – you’re not doing something wrong if you’re not approaching cube changes like I’ll be describing, but it’s useful to keep these ideas in mind in your cube design toolbox to shape your cube’s landscape.
• When doing cube updates, you don’t have to do it all in one go and make a cube update an all-or-nothing project, since you can dynamically make decisions if something isn’t working in between releases, ala hot fixes and patches.
  • To make it easier, you can break cube updates down into parts, even down to updates of 1 card (which I’ve done before cube nights) to make small changes to a cube list. Building it up to be a massive undertaking may just be holding you back from doing updates to your cube.
    • In general, realizing that I could break cube tasks up into smaller parts really did help me a lot with cube projects.
• Cards that get cut shouldn’t automatically be thought of as being cut forever and thinking of cube changes as a “this makes it, this doesn’t” decree once done and never revisited again leads to thinking of cards in a black-and-white manner, which dismisses nuance and context in evaluation.
  • Maybeboards on CubeCobra and even physical storage via binders and boxes are great to put these cards that are on the bench, to look through for ideas.
• Also – you don’t have to do exact 1:1 cuts for adds either, adhering to strict cube sizes (usually multiples of 15) is mostly arbitrary as I frequently just add cards during preview season for testing cards out for set reviews, to then reign in after the article’s done. If being unsure of what to cut is preventing you from adding 10 cards to your cube, consider just adding them and figuring out cuts later, once you’re more familiar with the performance of the new cards.

Level 1: Similar New Card changes: New cards in, cut cards with the same mana cost/function out that aren’t performing.

Usually, cube changes have been done within the context of including new cards and looking at them in the context of cards that are similar in function to the new cards. Generally, this in the context of performance: being outside of the power level band of a cube (over or under the desired power level) non-power level barometers, aesthetics-wise or other factors like not being well-received by drafters.

Historically, it’s come from a place of cutting “bad cards” to keep up with set releases – when Tolarian Community College made a video describing basic cube design principles, he said that the ethos of a cube is one where there’s “no jank.” Although I think the statement is overly simplistic, I get what he’s going for, given cube’s long-time elevator pitch of cube being “broken and that’s the point.”

Because making changes to a cube is often done in the context of new cards and doing like-for-like changes, oftentimes, this leads to looking at cube updates from a view of maintaining the status quo to not rock the boat – resulting in an over-reliance on like-for-like swaps when looking at options, narrowing the lens of options for how to craft a cube.

This assumes that their cube metagame is in such a state where making such changes will break their format. Usually, it won’t!

When Long Goodbye got printed, some cube designers had their search for a card to cut for it begin and end with Eliminate. These kinds of changes are emblematic of the Dungeon Man from Earthbound principle** and I’ve generally been a non-fan of those kinds of changes, where these are the easy finds for replacements but there’s usually better treasure when you look deeper. (Even if Eliminate ends up being something to cut later on.)

There might just be a clunky 5-mana beatstick that isn’t performing to expectations that’s right off frame, instead of the easy-to-see card with the same mana cost and function. While it’s true that several of these changes can add up to changing a landscape, usually doing a singular or a few non-like-for-like change won’t break things – and even if it did has more macro-level changes, that’s something that can always be changed as a cube isn’t set in stone.

Level 1a: New Card changes: New cards in, cut cards that aren’t performing.

While this is a subset of the first point, I wanted to differentiate it here since it’s a different way of looking at making cube changes, by expanding the scope of cards to take out to any cards, not just similar cards.

By expanding the field of cards to add to not just “new stuff,” but the entire card pool, the overall principle is the same, but I wanted to separate this out since adding non-new cards means there’s usually less of a chance of the card after “new card hype” dies down since cards are more of a known quantity. Both come from a similar perspective, looking to include something over cards that aren’t serving their purpose.

Level 2: Including (not just new) cards to support a new card/cards.

Going back to the statement of a cube having “no jank” – it isn’t so much that a cube should have no cards that aren’t powerful, but that every card in a cube should have a purpose. Sometimes, the card isn’t living up to that purpose (underperforming by raw rate/support.) Because of that, it’s important to view cards in that context: purpose.

A card like Trinket Mage or Tezzeret, Cruel Captain has a pretty easy-to-identify purpose: they work cheap artifacts, artifact lands and Time Vault, as both cards fetch the various keys to work with Time Vault.

Sometimes a card that gets newly printed looks to support a card or types of cards that are underperforming – the landfall cards from the Final Fantasy set and the MTGO Vintage Cube update for Edge of Eternities are a good example of this: for the most part, landfall strategies weren’t doing a whole lot in singleton higher-power cubes because their payoffs weren’t very good, but getting a few more good cards with it like Tifa Lockhart helped to push that archetype into a strength for green.

In Magic discourse, it’s easy to think of cards as just being good or bad, but usually it’s more nuanced and sometimes it’s more that a card needs support to live up to its potential and purpose.

Usually, these are classes of cards that deal with card interactions and synergies: typed duals/triomes in cubes with fetchlands, cards like Stoneforge Mystic and equipment to tutor for and cards that help define those synergistic cards’ floors, ceilings and average cases. Sometimes it’s more a question of helping a deck hit critical mass, like with Kindred and aggressive strategies – decks that tend to rely on raw number counts to work.

Level 3: Active macro-level decisioning to address issues in the metagame.

Although looking at cube updates through the lens of single cards is one way to look at shaping a cube over time, a more broad high-level view helps when looking at the cube’s metagame as a macro lens and looking at a cube as a whole to address issues in the cube’s metagame, using both old and new cards to address issues, looking at a cube outside of the whole “good cards, bad cards” mind frame and looking at your cube as having a metagame with a collection of decks, not just as individual cards.

Years ago, I talked about using data for managing a cube and being realistic with looking at underperformance with cards and archetypes in cube and I still agree with most of what I’ve said: understanding what is and isn’t working even if you don’t necessarily know why those issues are occurring is incredibly useful.

Of course, you don’t have to go through the trouble of tracking card data with spreadsheets, etc. as most cube designers don’t do this. Sometimes you just know, and those who play Constructed formats from the FNM level to the Pro Tour testing house levels have played a fair share of decks with a <5 match sample size and don’t need a statistically viable data set to know when something just isn’t working.

• Part of this comes from identifying and clarifying your goals on a macro scale – determining what you want your cube’s metagame to look like in terms of overall gameplay.
• Part of it comes from identifying strengths and weaknesses in your cube decks’ archetypes and adjusting those to address those issues.

Sometimes these knobs can be adjusted with individual cards, like Pyroblast and Price of Progress to address issues with blue decks/land heavy decks being too strong, but usually those tend to be duct-tape fixes for bigger issues – they can help in the short term but don’t address systemic underlying issues. “A micro solution to a macro problem.”

Garry Kasparov had a good quote about being careful when addressing issues in a cube metagame, where you don’t want to diagnose issues incorrectly:

“Losing can persuade you to change what doesn’t need to be changed, and winning can convince you everything is fine even if you are on the brink of disaster.”

But you can always course correct over time. Changes are never truly final.

Level 3a: Shaping a cube’s metagame outside of issues in the metagame.

When I thought of curating a cube like a bonsai tree, this is the kind of thing I was thinking of – making changes to a metagame on a macro scale without it being in the context of needing to do it to fix a metagame issue.

This is where the bonsai tree analogy really shines through – it isn’t replacing things that are dead or dying, but making small adjustments to make the end product better and more cohesive.

An example would be looking at a cube and determining that there’s too many 3-drops as they’re frequently some number of them in the sideboard. It isn’t so much that those cards are poorly performing, but drafters tend to find that they don’t need all of the ones being offered in a draft.

Some other examples include:

• Realizing that decks could use more/less fixing – even if it’s just something you want to try out to see how it impacts the role of multicolor cards and archetypes.
• Adjusting sections like replacing 10 colorless cards for 10 WUBRG cards to see what effect it’d have on the colorless / artifact strategies.
• Seeing what impact including less/more player damage has on red’s clock, or how it indirectly impacts blue/red decks and the impact of decks that specialize in purely getting in damage ASAP.
• Cutting 2-mana green ramp to see if drafters end up picking up ramp from other areas or if 2-mana ramp in green is even all that necessary.

Another way to look at the levels of changes is a way that my friend and fellow Pack One Slick Ones cohost jesseb34r used, as a way to illustrate that the methods that I listed above aren’t the only way to look at making cube changes:

• Slot for slot — I have a structure that works, I don’t know why it works, I will improve by finding cooler versions of things that already exist without changing the structure.
• Add cards separate from cut cards — I’m starting to play with structure by cutting cards that aren’t playing well and adding cards I want to play with and letting structure change.
• Consider density and relative context — I’m starting to actively design structure by thinking about what themes I want to exist and what it takes to make them exist.
• Consider metagame — I’m starting to consider not just what density a theme needs, but relative power level and consistency of different themes, matchup skews, etc.
• I finally built a structure that is (I believe) exactly what I want so now I only make slot-for-slot changes to preserve that structure while improving aesthetics/cool card density.

But, like I said earlier, you don’t have to do it one way and you can pull multiple lenses from these views. The main take away is to illustrate that there’s more to cube updating than adding new cards and cutting “bad cards.”

Thanks for reading! You can find more of my Cube thoughts and my recently updated cube list on my Linktree.

*It’s hard to hold a candle to the original, but this live version from the highly underrated Tony Martin years of Black Sabbath is a solid one.

**to use an adage from excellent retro video game reviewer Jason Graves.