Everything I Have Learned About Cube Design

I spent a lot of time working through the design of my Powered Synergy Cube. Throughout that process, the cube went though many different shapes and forms (you can look through all the different changes I have made on the cube here, there are over 95 pages of updates! I recommend starting at the beginning to see what the cube looked like originally.) In this post, I’ll go through some of the stuff I have learned along the way which can hopefully help you get through the design process faster and not make the same mistakes I did.

Have Specific Goals

When designing anything, the single most important question you should be able to answer is “Why?” What goal are you trying to achieve in doing this? What types of gameplay are you trying to create?

By starting from a fundamental design goal, you can base every decision you make on whether it will ultimately help to realize that vision in your cube. This might sound like a cliché or something that should be obviously easy to follow, but it is such an important part of the cube planning process.

The more specific you can get with your vision, the easier it will be for you to achieve it. When I first started out designing my cube all I wanted was for the cube to be fun. That category is so broad that I ended up included a whole lot of dissonant pieces that did not mesh well together. Each individual card could be fun in a certain context, but the overall cube itself wasn’t that fun.

Eventually I took some time to figure out what specific elements of the game I found fun: Interactivity, flexibility and a flat power level (no game ending bombs). Using these as a guide, it was much easier to refine my cube because the criteria were a lot more specific. Conflux is not really an interactive card, nor is it flexible, so I cut that package from my cube.

Over time I continued to refine my vision of what my cube should be about even further. As a consequence of my prior three criteria, the mana curve started dropping lower. Cheaper spells replaced expensive spells and bombs turned into flexible, interactive components. So I added the new design goal of keeping the mana curve low. This substantially narrowed the context of the cube and helped me closer to finding what worked for me.

I am still continuing to learn and grow in my understanding of what specifically I am looking for in my cube, and have changed my mind on quite a few occasions which lead to some massive reorganizations of the cube archetypes. But throughout all of this, I have done my been able to iteratively refine my intentions which has allowed the cube to continually progress forward to the point where now I feel it is nearly complete.

Without these larger issues to focus on, I would have been lost in an endless string of cutting and adding cards that at the time felt fun and made little progress at all. Micro-tuning can be useful, but big picture decisions are what contribute to the overall feel and enjoyment of a cube environment.

Power and Fun are Context-Dependent

When I first started working on my cube (when my goals were just flat power level and fun) I figured all I had to do was screen every single new card I added to determine if it was too strong on its own or not fun. This meant I got to cut cards like T3feri, Jace, the Mind Sculptor and Bribery. Once I had cut all the obvious offenders I assumed I was finished, but oh dear was I wrong.

In most high-power cubes, Bribery is typically 5 mana: win the game because you can grab one of the big creatures like Emrakul, the Aeon's Torn and close the game on the spot. In my cube, Bribery is actually not good at all because I removed almost all of the expensive targets for it. So while I originally removed it for being too strong, in the final iteration of my cube the card is actually too weak.

The opposite can happen as well. In War of the Spark limited, Narset, Parter of Veils was pretty good, but not amazing. Most limited decks run a lot of creatures and her static ability didn’t come up too often. She was just a better divination for the most part, but not problematic at all. Meanwhile, in Vintage she plays out like a 3 mana Dig Through Time that also stops the opponent from drawing cards. She is so oppressive that she was restricted in that format.

What I’m trying to say is: As you make changes to your cube, the context in which the cards exist will also change. This means that the relative characteristics of each card can shift dramatically. Cards that previously did not align with your design goal may actually be ok now, and more perniciously, cards that previously aligned with your design goal can start to cause problems.

Adding or removing one card affects more about your cube than just that one card. The addition of cards like Forked Bolt can make 1 toughness creatures worse, in particular mana dorks. While it might be tempting to say “Forked Bolt is GREAT in this cube” because of all the smaller creatures for it to hit, you can also think about it as saying “Smaller creatures are BAD in this cube” because of effects like Forked Bolt.

The specifics of which two of these statements are correct depends on what the rest of your cube looks like. If elves is far too good, then adding cards like Forked Bolt will help keep them in check and lead to an overall healthier cube experience. If elves is already struggling in your cube, then including Forked Bolt might just make your cube even worse since you are weakening an already poor archetype.

So I’ll go back to my statement as the title of this section. If you are thinking about changing a card in your cube, make sure you stop and think about specifically what the consequences of adding that card might be. It can also be useful to look at groups of recent changes, as sometimes unwanted effects can happen slowly as a result of multiple changes rather than the addition of a single card.

A few examples of cards that I had to cut from my cube because I didn’t think them all the way through are:

  • Stifle. I was trying to bolster blue tempo decks, and stifle is a great interactive card for stopping combos. The problem is, the existence of this card somewhat completely invalidates storm as an archetype. What is the point of going through all that trouble to combo off if your opponent can answer your entire deck with a single blue mana.
  • Wrenn and Six. For a similar reason as forked bolt above. I originally added her to give lands decks a fun utility card. But it turns out her -1 ability kills 50% of the creatures in my cube, so she just ended up as a control powerhouse.
  • Orcish Lumberjack. I added this card to help the red-green madness deck go a little faster, but it ended up just being used in the Lands archetype which was already too strong.
  • Traxos. Originally I added this card as a fun but narrow card for the artifact aggro deck to try and enable. When I put it in the cube almost nobody played it and it underperformed. As I continued to build out the artifact aggro deck, this card went from unplayable to an absolute menace. For the most part it was just a giant creature with vigilance (that could come down on turn 2, thanks to Mishra's Workshop). I was also shifting my removal more towards damage like shock instead of unconditional removal, so he was almost impossible to kill.

The Fun Cards Are the Ones That Actually Get Played

This one hurt me so much to accept, but I am so glad that I have. No matter much you try and support your pet cards and how cool these archetypes might be, if they never actually come together they are less fun than other cards that actually would see play.

Glissa is a hard one for me to accept, as I love this card so much. The problem was just that her casting cost does not line up at all with the archetype she is supposed to play in. I added tons of support cards for her like Engineered Explosives and Pyrite Spellbomb, but the reality was that people would take these cards and end up in Jeskai colors (which are the major colors for artifacts in my cube) and pass Glissa every time.

She was good enough for the cube, had all the support she could possibly want and could play out ok if you specifically forced a deck around her but that is not a healthy place for a card to be. Cards need to fit the archetypes or environment they are in naturally, and unfortunately for Glissa that would mean splashing her in a Jeskai deck or somehow forcing artifacts into GB, neither of which came about naturally.

Another one for me was the crazy ultimatum storm archetype I had designed, where players would use things like Pyromancer's Goggles to doublecast Genesis Ultimatum and basically put their whole deck into play.

This archetype can work in other, slower cubes, but I tried adding this into a powered cube that had cards like Thalia, Guardian of Thraben, Force of Will and Ethersworn Canonist. It was hard enough to make the archetype come together normally, but the fact that your entire deck could be invalidated by any number of cards that costed four to eight times less mana ended up making the archetype a really unfun experience in practice. Most individuals who tried it once didn’t do so again.

The moral of this section is: sometimes you just have to make hard cuts for the overall health of the cube. I know better than most how tempting it can be to include cards because “it would be awesome if they worked out”. Just be willing to remove those cards when they don’t happen to work out. Playable cards are always more fun than unplayable ones.

Stronger is Not Always Better

This is a short section but I felt it was important to mention. Often, I see people speaking of cards as “strict upgrades” for their cube. Ragavan is “strictly better” than Jackal Pup, for example. But stronger may not lead to games that are more fun. Both Ragavan and Jackal Pup are great aggressive early plays that punish slower decks, but Ragavan can snowball out of control so quickly that sometimes it can feel like the game is over after a single hit. This extreme pressure to interact quickly can completely invalidate entire strategies and lead to players feeling resentment.

The fun feeling of “will the control deck be able to stabilize in time against the aggro deck” is a delicate balance, and it is entirely possible that a card like Ragavan can be too powerful to allow that tension to play out in a meaningful way like something like Jackal Pup might. Once you get hit by Ragavan and your opponent plays a 3 drop on turn 2 (potentially even from your deck) the game may have already been lost.

To attempt to balance this, I often see players sticking to the ban lists of specific formats. Making a “Modern Cube” for example. While that is a great step in the right direction, I would just keep in mind that constructed environments are very different than limited, so certain cards might be stronger or weaker in your cube than they would be in constructed Modern (Birthing Pod is a great example of a card that is very strong in constructed and difficult to make work in limited).

Just because a card is legal in the format you have decided to build your cube in and is stronger than a card you currently have in your cube does not mean you have to include it. I cut Ragavan from my Vintage cube and have never once regretted that decision. Rabbit Battery is currently filling that role and I’m loving it.

It’s Ok to Break the Rules

Following up on our discussion above, it is rather common to restrict a cube to a specific constructed environment or card rarities (a pauper cube, for example would only run commons). I think overall this is a good starting point, but there are plenty of reasons to make exemptions.

As above, there are some cards that might fit your criteria but still be far too strong for the format. There also may be some cards that don’t fit your criteria but would be fantastic inclusions. I have seen people remark things like “oh thank goodness that got shifted to common, now I can include it in my pauper cube”, when in reality there is no reason that card couldn’t have been in the cube the entire time.

Don’t be afraid of bending your rules to include a card that would lead to an overall better drafting experience. Is March of the Machines too strong for a cube based on only commons and uncommons? Probably not. If for some reason you were looking for a deck that won by attacking with mana rocks I don’t see any reason why you couldn’t include this card in a your cube if you wanted to. (You would obviously have to change the name from peasant cube to something else, but that is a problem with semantics and not cube design.)

Just keep in mind, I’m not saying to just start adding strictly better rare versions of your uncommon cards. Rather, that I think it is totally fine to bend the rules to include unique and fun effects that you otherwise may not have been able to if you were following your format criteria literally.

For me, this meant allowing myself to include silver bordered cards in my vintage cube. Blast From the Past and Mons's Goblin Waiters are actually perfect fits for my discard/madness and sacrifice themes. The Goblin Waiters in particular is probably the best possible enabler for my Second Sunrise. Ineffable Blessing is a much loved and powerful enabler in my creature combo decks as well. While technically not in the vintage format, adding these cards into my cube has made the experience overall more enjoyable.

I also broke the singleton rule to allow myself better mana. By running three sets of fetches and two sets of each dual, I am able to achieve constructed level fixing in a limited environment. It has been fantastic. While I only did this with lands, there can be times where breaking singleton on certain cards in the cube to help with consistency might be beneficial. This can be especially true when you are trying to support an archetype that doesn’t quite have enough redundant enablers. Is running two copies of Lord of Atlantis or Cursecatcher really that different than having Llanowar Elves and Elvish Mystic?

Calculate Card Density

This one is something that is so easy to do, but so often gets overlooked. In this section I’ll use removal as the example, but the general rule applies to anything you could want. It can be a great diagnostic tool for your cube.

In the design process, create what your ideal deck looks like for many of the archetypes. Perhaps you can do a few different takes of the deck, since it can be difficult to capture the variety experienced in limited by a single deck, but however you do it, establish some “ideal deck” to look toward.

Once you have the ideal, estimate the relative breakdown of the important components: Mana fixing, removal spells, card draw, creatures, etc. For example, maybe your ideal deck has 5 removal spells. Then you should try and design your cube so that the expected number of removal spells for this archetype is centered around 5.

There are many ways to estimate this, a simple approximation can be found by first calculating the relative ratio of drafted cards to the overall cards in a cube. For a 540 card cube, 360 cards are used in an 8 person draft (which is 66% of the cube). If you have 9 red removal spells in your cube, you will see 6 of them on average in any single draft. The specifics of how those cards are distributed amongst players depends on your cube, but in general there are 5 colors and 8 players, so we can estimate that each color will be shared between 8/5=1.6 players. So of the 6 red removal spells in the cube, one player can expect to see approximately 6/1.6 = 3.75.

This is a very rough estimate and does not take into account variance or any other factors such as cube design, multicolored spells, drafting styles, etc but it can at least give you an idea of where you are at and a good way to calculate how to modify your cube to meet your goals. In this case, we can set our expected number of spells to 5 and work backward. If we want one player to get 5 red removal spells, using our above estimate, we will want 5*1.6=8 of them to be present in the draft, which means running 8*3/2=12 red removal spells in the cube. So we should add three more.

Always back up these tests with experimental data. After a draft of your cube count out how many red removal spells players actually saw. There will be variance, so it is unlikely to be exactly correct, but it can give you a good idea of whether your approximations are close or not. Sometimes you will see 7, sometimes you’ll get 3 removal spells and that is ok. But if you are getting decks with 14 removal spells and you are aiming for 5, doing an estimate of card density can help you solve problems like this fairly easily.

Things get a bit more complex when talking about two-color pairings, but similar methods as above can still apply. Keep in mind that these color pairings can be additive. So for example if we expect an individual player to see 5 red removal spells and 9 black removal spells, then the BR color pair would expect an average of 14 removal spells when drafted as the open lane. It is a bit more complicated here because not every card will end up in their deck and competition is different when drafting two different colors instead of a mono-colored deck, but again the ballpark estimate can still be useful.

Parasitic Doesn’t Always Mean Bad

Parasitic is a term most people use to mean a card or archetype that forms a second smaller archetype that is self-referential. The analogy is in reference to how parasites feed off of their hosts. Rather than melding uniformly with the rest of the cards in your cube, these few stand on their own. They do work together, but not with your cube at large.

A common example of this is the storm archetype. These three cards work really well together in the storm deck, but you aren’t often playing a card like Tendrils of Agony or Lion’s Eye Diamond in other decks. They mostly work well when they are played together.

Usually this characteristic is cited as exclusively as a problem (and it definitely can be), but I think there are a lot of overlooked aspects of these cards that can be used as a design tool. I used storm as an example, but these rules apply to other parasitic mechanics like life gain, +1/+1 counters and artifacts.

1. They send clear signals

Because the storm archetype is so insular, when some of the cards in the archetype are missing you can be confidant the archetype is being cut. The opposite is also true: last pick LED is usually a clear sign the archetype is wide open. This creates a clearly defined lane for drafters to work with. When you know storm is open, you can be fairly confidant about wheeling certain cards because you are certain nobody else should want them.

Signals are an important part of drafting, and when everyone is able to read signals properly the decks are more harmonious and the experience is far more enjoyable. If you have ever done a sample draft with bots you will understand the struggles of random picks. Against bots I have wheeled a Black Lotus in a pack that I was just trying to get Scourge of Nel Toth from.

Against humans I know that Scourge only fits in a handful of decks and if I am drafting one of them it is highly unlikely anyone else could ever make use of that card. So I can confidently plan to get that card when it comes around again and take a less “parasitic” card over it. You absolutely could not make this play with something like Fatal Push, as there are a whole number of reasons someone might want that card in their deck.

By having certain cards that are narrow and therefore likely to come around when you are in their intended archetype, you can help drafters plan further ahead and maximize their picks. This helps keep the picks in the entire pack interesting and high in potential, rather than just the first few.

In a cube with only flexible “non parasitic” cards, the relative strength of the cards you see late in the pack will just be low. Rather than wheeling your cool zombie dragon because nobody else can make use of it, you’ll just get something like Doom Blade because Infernal Grasp got taken earlier. Obviously this is a bit of an exaggeration but I hope it helps state the point.

2. The decks are more likely to come together

I am running both Witch’s Oven and Cauldron Familiar in my cube. I originally just included the Oven because I was worried the cat was “too narrow” of a card for the cube, whereas the Oven is a reasonable sacrifice outlet, especially in decks that care about artifact tokens like Oni-Cult Anvil.

After a bit of thought and testing I realized the cat being narrow might actually be a useful thing. Whoever picks up Witch’s Oven is likely to want the cat far more than anyone else, allowing them to pick it up late and complete their combo. This will overall increase the chances of the Cat-Oven deck coming together, which leads to happier drafters and more fun games. Synergies based on highly competitive cards are fare less likely to happen (for example: Tinker + Mana Crypt).

Taking this and the previous point together, it means that it is totally ok to run narrow or parasitic payoff cards on occasion. They send clear signals to the drafters so that whoever might want them is likely to get them (and perhaps plan ahead while they anticipate the card coming around), while everyone else intentionally avoids them knowing they have no use to their deck in particular. Just be sure that there actually is an archetype that wants them, there is no point of running narrow cards if they aren’t valuable in a deck that actually gets played.

Additionally, you need to be careful when doing this with cubes where the entire card pool is not always drafted. If Witch’s Oven is not opened but Cauldron Familiar is, people are unlikely to want the cat and it will end up taking up space of another card that could have been useful to someone. I included other cards to help support Cauldron Familiar so this fail case would not be as much of an issue.

3. When supported enough, archetypes stop being parasitic

Another way to work with parasitic cards or archetypes is to just support them more. In general, +1/+1 counters tends to be a parasitic archetype. Sure there are flexible cards that just happen to gain counters, but the archetype specific cards like Hardened Scales are only useful in a fully dedicated counters deck. You can’t just put this card in a random green deck.

But this only holds true as long as +1/+1 counters is a minor theme within your cube. If 30% of the cards in the cube happen to interact with Hardened Scales, it is no longer a parasitic card and can be considered just a generally good option.

Think of a card like Young Pyromancer — in most cubes this is just a high value pickup and would not be considered parasitic. But that is only true because spells are so commonplace that it is easy to draft a deck where the Pyromancer does something. You can think of this like spells being an archetype that covers a large portion of the cube. If instead there were only a few instant and sorcery spells in the entire cube, Young Pyromancer would be a “parasitic” card.

4. Downsides

The biggest risk of running parasitic cards is if they are in archetypes that don’t get played. It is fine to have narrow cards that slot into major archetypes because you know they will get picked up. But if an entire archetype’s worth of cards end up in the draft pool and nobody wants them, it can feel pretty bad.

So if you do plan to incorporate parasitic cards into your cube, the most important thing to consider is whether the decks they fit into actually see play, and if the cards provide a meaningful contribution to those decks.

Often times people run parasitic cards but are afraid of having too much support for this parasitic archetype (frequently this happens with storm). The end result is an archetype that is bad AND parasitic. If you’re going to include an archetype like this in your cube you absolutely need to make sure it has enough support to function. Just don’t forget: If you are going to build an archetype, you can’t be afraid to support it. Plus if you support it enough it stops being parasitic (like in a storm cube).

Be Wary of Secondary Enablers

This is one of the more important lessons I have learned, and also the one I get asked the most questions about! Secondary enablers is a term I use to describe cards like the above three — they do absolutely nothing until another card enables them (they are enabled by enablers). These are really tricky cards because at first glance they look like good options to include in their respective decks. A common line of thinking goes something like: “Oh I’m playing a cube with a blink archetype, Panharmonicon likes enter the battlefield effects, it would be awesome in my cube to support the blink archetype”.

Let’s break down the series of events that needs to take place for Panharmonicon to work. First, you draft a deck with a sufficient number of enter the battlefield effects (a minimum of 12 probably). Then, you take the time to spend 4 mana and cast Panharmonicon. The turn you cast it, panharmonicon does absolutely nothing. But on turn 5 you untap and reap the rewards when you play mulldrifter and get your sweet Panharmonicon trigger.

Now I have to ask…was Panharmonicon really even good in that situation? Or to put it another way, how happy would you be putting Touch of Brilliance in your deck (let alone if you had to wait a turn and spend 5 additional mana to get your card draw). Because that’s all Panharmonicon did in our scenario.

Even if we play two Mulldrifters and get to copy two triggers, our Panharmonicon was not much better than Concentrate. It is only really when you get three or more enter the battlefield effects where this card starts to pull ahead of other existing effects that could be handled with a single card.

Now, the relevance of this caution is highly dependent on the speed of your format. If you think your format is truly slow enough and the archetype is supported enough for someone to be able to take a full turn off to play this card and survive long enough to play three or more cards to trigger it (meaning they have enough good ETB creatures like Mulldrifter in their deck), then it still may work out for your cube. But even in that situation there can be a better option…

Just play more proactive enablers instead. If instead of Panharmonicon, your creature blink deck ran more good cards with enters or leaves the battlefield effects, like Watcher for Tomorrow and Spellseeker, it would be more likely to come together and the deck would no longer require taking a turn off to get the reward. Instead of casting Panharmonicon into Mulldrifter, maybe you cast Venser, Shaper Savant into Mulldrifter. Or Mulldrifter into Cloudblazer.

In all of these scenarios you are likely to get as much if not more value out of your card slot running a proactive enabler than you would with a card like Panharmonicon because you can play them and get value whenever you want. Plus you don’t need to hit a critical threshold of cards to even make your stuff work. Cloudblazer will draw you cards regardless of whether you are able to play Mulldrifter immediately afterwards or not. In fact, it might draw you into Mulldrifter!

I know I largely used Panharmonicon in my examples here, but that’s just because it is specifically the one I get asked about the most. The card can be lots of fun, but singleton limited is just a very difficult place to make cards like this work. I have tried time and time again to run cards like this and they all have underperformed substantially. Even when they do “work out”, the payoff was often equal or only slightly better than the effects I could have gotten from other cards without having to wait multiple turns.

I do offer a caveat, which is that these cards can make for great aspirational or fun impromptu draft challenges. I personally am running Academy Manufactor in my cube because I have found that people love to chase after making this card go crazy (although the paper cube runs a custom version of this card that triggers off of any token, not just Clue Food and Treasure). I know it is objectively less strong than other options, but I like it and it seems like others do too, so I included it. When these cards go off, they can go huge in ways that other cards do not.

The difference is that in this situation I am running this card knowingly for what it is: a fun meme. It is totally fine to include cards for this purpose, but make sure they are supported enough to be enjoyable and try to keep them to a minimum.

Also, as I mentioned in the parasitic section above, this problem does not apply as much for situations where your cube is over supported for the card in question. Hardened Scales can work in cubes that are so full of creatures with counters that you are likely to have enough enablers for it in every single green deck (the Proliferate Cube comes to mind as a situation where this may be the case). This is a very difficult condition to meet, but it can happen. Blood Artist works for similar reasons because most cubes run plenty of creatures, and creatures tend to die naturally in most environments.

Conservation of Flexibility

By making certain aspects of your cube more flexible, you can free up other aspects to be less flexible. For example, if you make mana fixing more abundant, players won’t have to spend picks on mana as highly and can instead use early picks on something else.

My cube is full of strange synergy pieces, enablers and payoff cards. Many of these are must-picks for the players who are drafting those archetypes, so they have to take the cards whenever they show up. This significantly limits the flexibility players have during the drafting process to pick up other things like fixing, removal and card draw.

To compensate for this, I substantially increased the availability of mana fixing and I moved general utility cards like removal and card draw to mono colored. Drafting BR storm and seeing something like Lightning Helix and Witherbloom Command in a pack can be disheartening. If instead the cube was running Tarfire and Collective Brutality, I could pick a up a card that is likely to have a meaningful contribution to my deck. This leads to more cards actually ending up in the decks of players instead of the sideboard. There are just more decks running red than there are decks that want to run red and white.

This example is specific to my cube, as I am still running narrow multicolored cards like Squandered Resources. But it is because I have added flexibility to other sections of my cube by largely removing multicolored removal spells that I am able to make these other narrow cards work. Those late pick Tarfires being more flexible than Lightning Helix make the difference in helping players create a functional deck.

The point of this section is: there is some total amount of limitation that can be handled during the drafting portion and you get to decide where that comes from. If your cube is largely synergy driven like mine, perhaps consider making other aspects of the cube less restrictive. If your cube has no synergy then you can make your card draw and removal more restrictive.

Perfect Balance of Color and Archetypes is Not Possible

If you look at the color breakdown of my cube, you’ll notice that not everything works out to perfectly even or nice numbers. That is because I balance my cube by archetypes instead of colors, and I believe the two to be mutually exclusive.

Whether you plan for it or not, many cards fit into multiple archetypes. No matter how much I intended for Orcish Lumberjack to help my GR aggro decks, he will also be used in the Lands archetype (and in my real experience, he was almost exclusively used in the lands archetype). You can give each card a label for its intended archetypes and balance your cube as perfectly as possible that way, but cards are going to be played outside of their labeled roles.

Blood Artist might be labeled for BR sacrifice and Abzan aristocrats, but on occasion someone is going to use it as a win condition for their Scrap Trawler combo deck or Empty the Warrens storm list. Sometimes things get crazy.

The other part of this is that often some colors or color combinations span multiple archetypes better or worse than others. In my cube Black is highly spread across sacrifice, discard, storm, spells and aggro. Meanwhile white has a hard time doing more than supporting artifacts and having blinkable creatures. If I wanted to run as many white cards as I did black cards, I would have to over-support their respective archetypes.

If white supports 3 archetypes and black has 5, then running an equal number of cards in each color would imbalance the archetypes. I did try that for a time, and the white artifacts deck was the most absurdly dominant and over-supported deck in the cube. It somehow supported three separate drafters in one test that we did (or at least tricked them all into trying to fight over the archetype). Even though white had an equal number of cards in it to other colors, it was supporting fewer archetypes so those archetypes ended up with more cards in them.

As a result of what I mentioned above, even if you plan out for white and black to have an equal number of archetypes there are always going to be interactions and archetypes you did not intend for. Rather than 5 and 5 archetypes like planned, the reality will be something more like 5.2 and 6.8. Obviously it will be rather difficult to know these numbers exactly, but hopefully you can agree that they are there in theory. Some archetypes will have more or less support than intended due to unplanned interactions between cards.

If you insist a strict balance of colors, this will be almost impossible to resolve but if you allow for an imbalance of colors it can be solvable in theory. In this example, you can cut some white cards for black cards to make the number of archetypes 5 and 7, fully supporting the last archetype in black and cutting the unsupported archetype in white. Getting the number of cards per archetype to line up can be more complicated and would require careful tuning across all the colors involved in the archetype. In reality this will be a continuous process.

Ok that was a lot of text, but I just really wanted to encourage you that it is ok for the numbers to not be perfect (and to justify why I am not running a single Boros card in my cube. It turns out in my cube that “Boros” is actually the colorless artifacts deck, so many of the colorless cards act like Boros cards. Nobody was playing the actual colored WR cards so I cut them).

It is Easier to Cut Bad Cards Than Add Good Ones

The title says it all. Spotting cards that don’t fit well in your environment is just easier than identifying cards that do. This is true in both the planning and testing stages. I cannot tell you the number of times I have looked at a card and thought “wow, this would be perfect in my cube”, only to add it and see it do absolutely nothing (or sometimes actively make the cube environment worse, like Stifle that I mentioned above).

I have also looked at cards that I thought were unlikely to perform at all in my cube, but after a test run realize they fit perfectly. There have been cards that I looked at dozens of times, and each time I wrote them off because they were “bad”. Rotting Rats is probably the most notable example of this. All of the effects do things that my cube is wanting to do, it has the zombie creature type and it is a cheap and re-useable effect. But for some reason, the idea that this card was fundamentally bad clouded my judgement and prevented me from adding it to my cube until only recently.

The first time I saw Rotting Rats while drafting my cube I was in the mono black Zombies archetype and I was desperate for a discard outlet to support stuff like Scourge of Nel Toth and Vengevine. I snapped it up and the card played out fantastically in my deck, enabling me while also disrupting my opponent. Now it is one of the cards I am more excited to see whenever I am looking to discard or sacrifice stuff.

The takeaway I had from this was to not be afraid of trying out new cards. Anything you feel is at least questionable can be worth giving a chance. It is so hard to predict exactly how the card will play out until you see it in the draft and in your deck (or sideboard). The cost of doing this is low and it is so easy to remove these cards down the road if they fail to perform.

Additionally, when looking for cube feedback it is so much easier to ask others to point out cards that don’t fit than it is to ask for suggestions of newer cards that do. One requires them only to look at the list in front of them and spot outliers, the other requires an intimate understanding of your cube environment, what you are looking for and a substantial card knowledge to find something that might fit that objective. If that wasn’t enough, suggesting new things is a more vulnerable position to put oneself into than pointing out problems with existing cards. People often love pointing out mistakes but hate offering new ideas.

You can take advantage of these facts by using your time when others are drafting your cube or looking at it to ask them what cards they might cut. Some people may also offer suggestions on what to add, which is just a bonus. If you continuously add new cards and remove the bad ones, over time your cube list will improve. Be eager to try to things and quick to identify stuff that hasn’t worked.

Sometimes Bad Cards Just Need More Support

I offer this section as a bit of caution after reading the previous one. The strength or utility of cards is a function of the cube environment they are in. Often cards that underperform are doing so because they lack the proper amount of support. Birchlore Rangers is a magic card with a lot of potential. But in an environment without enough elves this card will do practically nothing.

I had originally tested out a tiny elves package in my cube, since I was scared of adding too many “narrow” cards. During that time, all of the elves cards looked horrible— nobody played them and they went last pick frequently. Dejected, I cut the whole elves package altogether thinking these were just not cards meant for cube.

A few months later I decided to give elves another go. This time I decided to go all out and add as many good elves as I reasonably could fit into green. This time the cards had the proper amount of support to shine and now Elves is one of the best and most consistent decks in my cube!

I learned two things from this. First, when looking to cut a card ask yourself if it is the card itself or the environment that is causing the poor performance. Sometimes a small environmental change can bring new life to a dead card.

Second, when trying new cards it can be beneficial to test them in an environment that supports them as much as possible. When I wanted to make Prized Amalgam work in my cube I threw in as many cards as I could to see if it could ever be useful (that was actually why I added Rotting Rats in the first place!) Even in the maximally supported environment (at least, given as much wiggle room as my cube offered), this card was terribly difficult to make work. Now I have no question in my mind that this card does not belong in my cube, rather than worrying if I just needed to support it more.

Quick note of caution: Don’t warp your cube too much just to make a single card work. These suggestions are mostly tools for testing, but you don’t want 10+ slots in your cube devoted to making something like Prized Amalgam work. If the card does pass your initial test, you then need to ask yourself if that level of support is reasonable for your cube environment.

Signals Matter

No matter what your intention is with adding a card to your cube, there will often be some pre-conceived notion of what that card is supposed to do. When drafters see cards like this, they can be led astray chasing an archetype that is not supported and come away feeling cheated. It can be overwhelmingly tempting to blame the drafters for not understanding your cube properly, but ultimately it is your job as the cube designer to make it clear what your cube is supposed to be doing.

While a card like Entomb can be used in other archetypes, it is almost exclusively associated with reanimator. I have often thought it would be good in my cube to enable things like Hogaak, Crucible of Worlds and Dreadhorde Arcanist. Even if it does work to help those cards out, it can easily send the wrong signals and get picked up early from people hoping to reanimate some fatties.

It can still be ok to include cards like this on occasion, especially if they are in isolation. Bigger problems arise when you have multiple cards like this pointing to an archetype that doesn’t exist.

For example, reanimate is a pretty decent value play on its own. Getting back a creature that died earlier for a single mana can be a massive tempo swing. But the existence of that card in your cube begins to suggest a reanimator archetype. If you run Reanimate and also included Entomb to support your lands or spells package, players can easily get confused.

The other side of this is also true: it can often be correct to include cards that send better signals over other ones that might technically play a bit better. Letting players know loud and clear “hey, this is an archetype” is something that provides real value to their cube experience.

This card might not be the best discard payoff ever created, but the fact that it can only be utilized in a discard deck and lets players know The Underworld Cookbook is also in the cube is useful. It tells players there is a serious deck that cares about discarding stuff, and gives them information about what other support options are in the cube. While it itself may or may not play a meaningful role in those decks, the message it gives players can help them make more informed decisions during the draft, leading to overall better decks.

Article still in progress, I’ll add more as I remember stuff

9 thoughts on “Everything I Have Learned About Cube Design

    • Wanted to write: great write-up on your thought process. Gonna try to implement some of this as I build my chromatic cube.

  1. Hi Caleb, really enjoying the content! Would you mind I translate this to another language so I can share it with my friends? (I will also give the link and your credit to them as well)

Leave a Reply to Caleb Cancel reply