You are sitting on your living room floor on a rainy Saturday afternoon,
Some questions to ponder
One of the first decisions you'll need to make when you first build your cube is how many cards to actually include. There is, of course, no single correct answer. However, the answer that is right for you will depend on your answers to a few key
1. How will the cube be played?
2. How many players do you want to be able to support at one time?
3. How much variation do you want each time the cube is played?
4. How strong do you want decks to be overall?
5. Is there a limiting factor that caps the size of your cube?
There are a variety of methods that people commonly use to build cube decks. These include multi-person booster drafts, two-person Winston or Winchester drafts, Sealed deck, and more. Each and every method has merit, but each also brings certain constraints to the card pool. In general, booster drafts usually involve creating three 15-card piles of random cards per player. Winston drafts often involve a single pile of about 90-100 cards; Winchester drafts usually involve two piles of about 45-50 cards. Sealed deck usually involves 75-90 random cards per person. This leads us naturally on to the second question: how many players do you want to support at one time?
Number of players
If you plan to use your cube for two-person Winston or Winchester drafts exclusively, then you could get away with a cube with as few as 100 cards. If you generally have about four people and like to do 75-card Sealed pools each time you play, then you'll obviously need at least 300 cards to play with. If you want to support an eight-person booster draft, then you'll need at least 360 cards. The math is pretty straight forward.
However, the math only really tells you the minimum number of cards that you'll need. If you always booster draft with eight players and have a 360-card cube, then all cards will be drafted every time. This will create a fairly consistent environment but will also create a situation where players will tend to know that certain cards will turn up in an opponent's deck at some point. This could cause them to make certain decisions, such as always putting artifact removal in their main deck just in case they play the opponent with Sol Ring. On the other hand, the larger your cube, the greater the variation from draft-to-draft since some cards won't appear some percentage of the time. Generally, the more cards you have relative to the number of players, the greater the variation.
Counteracting the push to larger cubes - at least for "best of" cubes - is the simple fact that the more cards you have, the further down the list of top cards you have to go. There will likely be a significant difference in the overall power level between the top 10 and the 70th to 80th-best White cards, for example. The further down the list you go, the more cards you have to include of a lower power level. That's fine, of course, if you want a bit of a range of power levels of the cards in the average cube deck. But some cube designers prefer to keep the power level very high and therefore tend to prefer slightly smaller cubes. If you're playing cube to play with the best cards in Magic, then you you probably want those cards to actually turn up in a draft. If you're playing a powered cube, you might like to actually see moxen from time to time!
On the other hand, if you're building a themed cube of some sort, then you might find that there are only so many cards that fit your theme. This is likely to grow over time, but you might need to do a little research before finalising the overall size. Let's imagine that you want to build a tribal-themed cube and one of the tribes that you'd like to include is Kithkin. A quick Gatherer search shows that there are only 58 different Kithkin creatures in print, while there are 134 Merfolk and 264 Goblins. Further, an analysis of the Kithkin cards could reveal that you don't actually want to include all of them for whatever reason. Assuming you want some balance between the various tribes, at first glance it appears that Kithkin could be your limiting factor. It might make sense to work out how many of these cards you'd like to include and go build up from there.
For cube designers working with a traditional cube, one limiting factor turns out to be the number of aggressive (e.g. 2-power) one drops available. Agressive decks in a "best of" environment need to get off to a fast start, and therefore require a critical mass of these one-drops - cards like Isamaru, Hound of Konda, Goblin Guide and Gravecrawler. Unfortunately, there are only so many of these cards available. If you want to ensure that there are enough cards available in the cube to support aggressive strategies, then you can only grow your cube so big before there just aren't enough of these cards to appear with sufficient frequency to push someone into this strategy.
In practice, since booster drafts are one of the most popular ways to play cube, it is very common to select a cube size that is some multiple of 90. Common cube sizes include 360, 450, 540, 630 and 720 (with 630 being less common). Cubes tend to start at 360 in order to support an eight-person draft while 720 supports two eight-person drafts without the need to reshuffle (draft half the cube, then draft the other half). This is certainly not necessary, since any cards leftover just add to variance and can be shuffled back in for the next draft - the leftovers don't strictly need to be a multiple of 90. But it's a decent rule of thumb to consider.
My own traditional "best of" cube currently contains 450 cards, in order to support an eight-person draft with some variance (20% of the cards will sit on the bench) and still keep the overall power level fairly high. 450 also supports exactly six 75-card Sealed pools which is one of the ways that I like to play when there are only 3-4 people. In practice, I have had more than one 10-person draft and have had to exclude people in the past as well due to the limitation of the cube size; as more cube-quality cards get printed I find myself considering a shift up to 540. However, I also experience the limiting factor listed above: I don't believe that there are quite enough aggressive one-drops of cube-quality for me to make the shift up to 540 cards. This could change over time as more of this type of card get printed.
The number 450 has another useful property. In a traditional eight-player booster draft using sealed booster packs, there are 24 packs, each pack containing 10 common cards (not counting basic lands) for a total of 240 common cards. The average set contains about 100 unique commons (for example, M13 has 101). Assuming that all of the packs are from the same set, then there are approximately 2.4 copies of each common in the draft on average. This number can be scaled to any cube size as a simple ratio, but if you scale it onto a 450-card cube then you find that the number is exactly 3. This means that for a 450-card cube, three cards that do approximately the same thing are the equivalent to a common in a traditional booster draft. This provides a meaningful metric that I can use to help determine how many cards that provide a particular effect I want to include. As your cube scales this number gets larger. For example, in a 720 card cube this number is 4.8.
For this reason, the themed cube that I am designing - which assigns actual meaning to card rarity - will also be designed as a 450-card cube as a starting point.
While size matters for getting started, ultimately you can always make changes at a later date if you need to. It will involve some work to change, but it can be done. So don't stress - just pick something reasonable and get started.