Well-Defining the Commander Cube Problem

     Currently, there is no consensus on how to create the best commander cube experience and it continues to be a challenge in part due to the problem never being laid out in a clear way. The goal of this article will be to have a clear goal and guidelines for designers to begin to bring fresh ideas to the table without having to struggle with a vague notion of the problem and to knowingly tackle all problematic aspects of the commander cube.

      First, let us define a commander cube. A commander cube is a cube that is meant to be played with commander rules. That includes one commander, singleton, and commander color identity (all cards in your deck must be the color of your commander’s color identity.) What changes from regular commander is no banned list, as the cards to include are up to the cube owner, and the deck size will have to be smaller than 100 cards. Since constructed to limited is 60 to 40, commander to limited commander can be 100 to 60. That allows for decks mana bases to piggyback off already know land counts. If a number other than 40 or 60 is chosen, land counts will be less intuitive as there is no format to compare them to and having to be based on raw percentages instead.

      We also want to keep the heart of the commander concept. This includes the multiplayer aspect and the idea of building around a commander. Without these two clauses, the format is not really commander. If a single player commander cube is desired, then it is better to think of it as regular cube with some minor deck building restrictions tacked on, while forcing a player to have a commander and follow color identity does not make a commander deck.

      Now we know what a commander cube is looking for: commander style rules and gameplay. However, since there is a focus on multiplayer game play, it seems like it would be nice to add regular two-headed giant as something that could be done with a commander cube. Lots of the cards that work in commander for having multiple opponents will also work in two-headed giant. So, here is an outline for building this cube now that we have an aim in mind:

  1. Need to provide several commanders for each color combo to drafters
  2. Ideally would want all color combos to have commanders and draftable
  3. Decks should feel as built around the commander
  4. Each deck has 50% more cards than usual, how many cards should this cube have then?
  5. Decide on how to implement the draft
  6. Support two-headed giant
  7. How competitive should this be
  8. Support 4 player drafts, 8 player drafts, or 4 to 8 player drafts
  9. Land base must be able to support 3+ color decks
  10. Deal with artifact mana

      We can now look at each individually to see what issues arise to better reflect on the problems of a commander cube.

1. Need to provide several commanders for each color combo to drafters

      This means we need at least three of each color, guild, shard, and clan. We also need the cycle of four-colored commanders or ten partner commanders to replicate the same color combos. We also need two to five WUBRG commanders. Colorless will probably be impossible, but if it’s not, it can replace one of the WUBRG commanders. That brings the total to 82-85 commanders (87-90 if we use the partners.) This means we are looking at roughly 90 cards to be commanders. This is a large amount of design space that will be dedicated to commanders. However, any less of some color combination would result in too little commanders and replayability would be compromised. The alternative is to only allow certain color combinations to be drafted. For example, only shards and wedges. That would lower the number of necessary commanders overall allowing for an increase in commanders for the combinations desired returning some of the replayability that is lost by forcing specific color combinations. But again, this results in increased design space used.

      This gives us the first real guideline to building a commander cube. A large amount of design space will need to be used on commanders to ensure good replayability. Thus, a choice needs to be made on a large depth on specific color combinations or a shallower depth per color combination but support much more of them.

2. Ideally would want all color combos to have commanders and draftable

      This one ties in with the previous one. Personally, I dislike formats that push you to be a specific color and does not allow for deviation. Supporting only specific color combinations in a commander cube would remove all hopes of deviation from it because of the commander color identity rules. If there are no Izzet commanders, for example, drafting Izzet is impossible given the rules.

      Although that is not a result that I am particularly fond of, it may be that supporting all combinations is impossible to do well. With such a large amount of design space being dedicated, there might not be enough left to offer a wide range of support. This gives our first main commander cube problem: Which and how many color combinations to support while maintaining the goal of increased diversity and replayability?

3. Decks should feel as built around the commander

      Commander is more than just a deck with a card that is always in your hand. The commander is the heart of the deck. It is the driving force of the whole deck. All other cards should be working towards furthering the goal of the commander and providing great synergy with it. This raises a difficult problem for a commander cube. There will be a finite set of cards. So, cards must work with as many commanders as possible. A commander like Hapatra, Vizier of Poisons might be a great commander for having fun with -1/-1 counters. However, it requires a large amount of -1/-1 counter cards that will be tough to work in other strategies.

      With a commander cube already having a significant amount of design space devoted to commanders, finding spots for niche strategies will be impossible. Commanders must be playable, not only as commanders, but as part of the deck as well. A card like Hapatra, Vizier of Poisons does not play well with most other strategies. They will not want most -1/-1 counter cards and Hapatra, Vizier of Poisons is quite useless without them. It might be possible to tune such a black/green section to support Hapatra, Vizier of Poisons as a commander and still have other decks want those -1/-1 cards and play with Hapatra as part of the main deck. In this scenario, there is the large cost of pushing other strategies out to make such room for Hapatra, Vizier of Poisons. This also does not include the difficulty of providing other black-green commanders that will provide a different enough experience to not conflict with our already established problem of diversity and replayability. There could be cases where accepting one or two more poisonous commanders will be acceptable, but they will demand excellent design to make them worth the warp in design they require. This leads us to another problem to focus on: How to implement commanders that feel unique while staying away from the poisonous?

4. Each deck has 50% more cards than usual, how many cards should this cube have then?

      Usual cubes have 360 cards. Give 50% more and it goes up to 540. That is a good starting point. Commander cards need support, so having more cards in a cube than appear in a draft will demand a higher number of cards percentage wise than if a cube always has all its cards present in a draft. Designers will have to account for the expected value and the expected deviation from that value to ensure all commanders are properly supported (I will write an article about this (hopefully) soon.) As we know that we want high diversity, we want to minimize support for each cards as much as possible by alleviating the design demands of that where ever we can. Thus, making a cube only as big as it requires for a draft and not having to worry about expected value and deviation is a great start.

      Now, 540 is not a number set in stone. It is just a starting point. The real number will depend on the draft method that is discussed in the section below.

5. Decide on how to implement the draft

      This is a contentious problem in the current cubing community that is currently (to my knowledge) unsolved. Many designers will simply add a pack of commanders ahead of the draft to allow for the extra inclusion of commanders. Some will add the pack after the first or second pack. Others will have the commanders in the packs as normal and just draft four packs of cards. All those solutions are inadequate. Commander is about the gameplay. The draft should be a much more minor part of the play than it is in regular limited and cube. The draft time needs to be cut down. The draft decision complexity needs to be cut down. The draft format needs to be simple. Shorter draft times means not adding more packs, less decision complexity means not adding cards per pack or more information, and a simple draft experience means this new draft format must be elegant in its explanation and understanding.

      These restrictions are gigantic. We are stating that we need 50% more cards while decreasing the time, reducing decision complexity, and creating an elegant and simple draft format. This is probably the hardest problem to solve in creating excellent commander cubes. The restrictions placed on it are open. With traditional draft, the restriction of cards coming in a pack is gigantic. Rochester draft was a natural progression to creating a draft, while when the need to reduce the time and complexity came, the current regular draft of picking one and passing the pack is an elegant and obvious progression. This draft problem is not bound by packs and pack sizes. Designers must work with a blank canvas and must fight against the community stablished draft for cubes which follows the limited draft format even though the pack restrictions are non-existent. So, we come to the third problem to focus on: How to implement a draft that is shorter, has less decision complexity, is elegant and simple, and has roughly 540 cards?

6. Support two-headed giant

      This issue is a freebee. It will probably be easy to support two-headed giant as many cards that work well in commander work in other multiplayer formats. The only thing to keep in mind is balance requirements, but that does not affect the core of a commander cube design and will come into play later. As for the draft, a cube can probably revert to regular draft for two-headed giant play. It might need some adaptation to work with the lands and artifact mana issues raised below, but designers can probably work on this after the design for the primary purpose of playing commander is established well.

7. How competitive should this be

      Commander is traditionally a format that is not competitive. There are players who have competitive builds of commander decks, but most prefer to use the format to have fun. However, limited is traditionally a highly competitive format. Few players play limited and do not take it seriously. Even fewer when it comes to cube players. There is a bit of a clash going on which causes to further complicate all the issues already mentioned. Will the draft format created specifically for commander cubes be as competitive as regular draft? It does not have to be. Traditional cubes are often trying to be the best possible competitive formats they can be, given the restrictions the designer has on it. Designers need to abandon that notion for commander cubes. This new draft format and design needs to provide the best possible commander cube experience. It does not have to be able to be ported over to traditional cubes and be a good competitive traditional cube experience. It can be a worse format for those cubes, while still being perfect for commander cubes.

      The mixing of a competitive limited experience with commander must always be kept in mind throughout the design process. The competitive aspect of winners must still exist. Since we still desire multiplayer games, pods of four will probably work best (this is standard commander game size.) Score needs to be tracked and not only for the winner. Different point for first, second, third, and last places would be ideal to ensure a reward exists for coming in second and third. While the fun/casual aspect of commander needs to exist in the archetypes and commanders with having cards that encourage fun strategies, gameplay, and drafts. Strategies should feel fun to play against and with, the gameplay needs to have exciting, big moments that turn the game, and drafts should encourage players to do wacky things that they normally would not during regular competitive drafts. This leads us to another problem: Balancing the competitive limited draft experience during games while having fun commander archetypes, drafts, and cards.

8. Support 4 player drafts, 8 player drafts, or 4 to 8 player drafts

      Many play groups have fewer than eight players. Thus, a draft needs to support anywhere from four to eight players. We do not need to worry about less than three as the intent of this cube is to be a multiplayer experience. A regular cube can be used for drafts with two players. Three players are up to the designers if they want to support it or not. Luckily, we are already have to implement a different draft method so this restriction merely adds on an existing one, which becomes: How to implement a draft that is shorter, has less decision complexity, is elegant and simple, has roughly 540 cards, and supports four to eight (or three to eight) players? (Drafts with fewer than eight players might need to be below 540 cards.)

9. Land base must be able to support 3+ color decks

      Commander is about the three-, four-, and five-colored decks. But with those decks come a requirement for a good land base. In regular limited, having drafters draft lands to support such decks are part of the challenge during a draft. It enhances the experience of a competitive draft. However, commander drafts need to be less intense. It rarely feels good to pick up a land early in the draft or having two-colored lands be sitting in packs until late. These feelings need to be avoided for a commander draft. However, that is not to say that there should not be lands that are drafted. We just want to reduce how highly they are picked and to reduce how often the two-colored ones sit in packs. The latter is especially true in commander where splashing is impossible.

      How to approach this will have to be up to the designer. Give lands for free at the end of a draft? Have a package of lands as one pick in a draft? Have wild cards during the draft that can be any color combinations of a specific type of land? For example, a fetch wild card can be used for any fetch land. Which kinds of lands to include? Should commanders get lands in their color identity for free? If so, how many? There are many wildly different ways to approach this problem. The only certainty is that the regular method for including lands is insufficient. This leads to another problem for designers to answer: How to implement lands in a way that minimizes early and late land picks?

10. Deal with artifact mana

      Artifact mana is an inherent part of commander. It leads to some of the big plays that are available in the format that would not exist otherwise. However, it is dependent on some key cards. For example, every commander deck has Sol Ring. The card is too good at ramping not to include, and due to being colorless, it can be part of any commander. How does a designer implement that in a cube? Do you include one copy and it leads to drafting inequalities in decks as more such examples come up? This would lead to artifact ramp cards being early picks much the same as they are in regular powered cubes. That sounds like a similar unfun problems as early picked lands had in the section before. A designer can implement multiple copies of such cards, but then decks cannot include multiples as it breaks commander rules. This means that there will be cards in packs that cannot be picked by certain players, in which case, they can be given away for free to all players instead of including them in the draft. That would also reduce the number of cards needed to be present in a draft. Even more drastic, a designer can choose to avoid these effects all together. But that probably leads to a worse limited commander experience as it deviates further away from regular commander.

      This gives the final focus for a designer: How to implement artifact ramp to provide enough to each player without it becoming too high of a pick in draft, having a large power disparity between cards chosen, or breaking singleton in the draft?

      To summarize, these are the major commander cube problems that a designer needs to knowingly address:

  1. Which and how many color combinations to support while maintaining the goal of increased diversity and replayability?
  2. How to implement commanders that feel unique while staying away from the poisonous?
  3. How to implement a draft that is shorter, has less decision complexity, is elegant and simple, has roughly 540 cards, and supports four to eight (or three to eight) players? (Drafts with fewer than eight players might need to be below 540 cards.)
  4. Balancing the competitive limited draft experience during games while having fun commander archetypes, drafts, and cards.
  5. How to implement lands in a way that minimizes early and late land picks?
  6. How to implement artifact ramp to provide enough to each player without it becoming too high of a pick in draft, having a large power disparity between cards chosen, or breaking singleton in the draft?

(Different implementations of 5 and 6 could lead to needing less than 540 cards in 3)

      With these questions being the focus of designers as they tackle commander cubes, it should lead to better experiences than the current ad hoc way of implementing commander cubes with few adaptations that leave much to be desired. I would encourage current and future commander cube designers that have or want answers to these questions to keep encouraging the conversation in whatever forums they might visit. The cubing community is a few great ideas away from finding a great solution to the commander cube problem, and with the problem well-defined, it should be easier for designers to come to them. Remember, the goal of this is not to have a quick and easy solution that adapts the current cube formats to commander cubes, it is to create a truly great format that can answer the above questions rigorously.

      If wanting to join me in this discussion, along with fellow listeners and readers of Path to Cube, join the discord at: https://discord.gg/uEFYtxu

Fernando Lopes